Interactions between the Woodwasp noctilio and Co-habiting - and Woodboring Beetles, and their Fungal Associates in southern Ontario

by

Kathleen Ryan

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy

Faculty of Forestry University of Toronto

©Copyright by Kathleen Ryan 2011

Interactions between the Woodwasp Sirex noctilio and Co-habiting Phloem- and Woodboring Beetles, and their Fungal Associates in southern Ontario

Kathleen Ryan

Doctor of Philosophy

Faculty of Forestry University of Toronto

2011

Abstract

In its introduced southern hemisphere range, Sirex noctilio causes considerable mortality in

non-native forests. In its native Eurasian range however, S. noctilio is of little concern

perhaps due to interactions with a well-developed community of pine-inhabiting and

their associated microorganisms. If such interactions occur, they may limit the woodwasp’s

impact in its newly introduced range in . My research addresses two broad

questions: 1) Does S. noctilio share its habitat with other insects and if so, with whom? 2) Is

there evidence that co-habitants affect S. noctilio , and if so how might such interactions

occur?

Field studies undertaken to describe the woodwasp’s host-attack ecology in Pinus

sylvestris showed S. noctilio activity occurred between mid-July and late August, and other

phloem- and woodborers sometimes entered the tree after the woodwasp. Tree mortality

occurred from two weeks to several months after initial woodwasp symptoms. Suppressed or

intermediate trees, those with ≤ 25% residual foliage, or those with stem injury or previous

woodwasp symptoms were most likely to have symptoms of woodwasp attack.

ii

A second field study conducted to identify associated species in

S. noctilio -infested Pinus sp. showed the wasp was sometimes found alone, but usually

shared the tree with other phloem- or woodboring insects, most commonly the curculionids

Tomicus piniperda , Pissodes nemorensis and Ips grandicollis and the cerambycid

Monochamus carolinensis . I found no indication that wasps were absent when beetles were

present, but there was evidence that woodwasps were less abundant, but larger, when beetles

were present.

Experiments showed that indirect interactions can occur between the two insect

groups via fungal associates of one or both. In the laboratory, the woodwasp symbiont was

outcompeted by two beetle-associated fungi, Leptographium wingfieldii and Ophiostoma

minus, over a range of temperatures. Under field conditions the woodwasp was able to detect and avoid ovipositing in P. sylvestris inoculated with L. wingfieldii, but its oviposition was

unaffected by O. minus .

My results show that insects co-habiting pine with S. noctilio have potential to exert a

measure of biological control on the woodwasp and may help to limit its impact in North

America.

iii

In memory of Peter de Groot

Mentor and friend

iv

Acknowledgements

As with any endeavour of this scope, many people were instrumental in the completion of this project. Special thanks to my advisors Peter de Groot who provided guidance, and who shared his wisdom throughout the completion of project and Sandy Smith who taught me that anything is possible. Thanks to my advisory committee members Jean-Marc Moncalvo, Peter

Kotanen and Taylor Scarr for advice and assistance, and especially for their enthusiasm about this project.

Thank you to Chuck Davis for providing assistance and guidance on all aspects of this project, and whose contribution was immeasurable.

Data was collected with the assistance of Sheung Au, Madelaine Danby, Sarah Drabble,

Megan Evers, Sean Strong and Kate Surowiak. I am grateful for their conscientious efforts.

Data collection was facilitated by the outstanding technical assistance provided by Ian

Kennedy and John McCarron.

Many people provided essential advice for the fungal component of this project including

Simona Margaritescu, Martin Hubbes, James Reid, and Tony Ung. Ed Czerwinski provided training and suggestions on tree assessment, but more importantly showed me how to see a forest through an insect’s eyes. Hugh Evans gave advice, helped with finding sites, and patiently answered a myriad of questions.

Thanks to Isabelle Ochoa for teaching me beetle and woodborer identification. Isabelle,

Katherine Nystrom, Reg Nott and David Langor identified many of the insects used in this project, and Serge Laplante verified identifications. v

Many students at the Faculty of Forestry helped with advice or field work, of those I’d especially like to thank Laura Timms, Smith Sundar, Jeff Boone, James Dennis and Nick

Rudzik.

Thank you to my husband Sean for his love and support through this endeavour, for reminding me of the big picture, and for helping with every aspect of this project. I am forever grateful to you. Thanks to my family and friends for their patience, support and perseverance through my time completing this project, and especially for never asking, why?

Several private landowners, Sandbanks Provincial Park (Don Bucholz), Toronto Regional

Conservation Authority (Tom Hildebrand), Grey-Bruce Conservation Authority (Ken

Goldsmith), Simcoe County (Graeme Davis) and Canadian Forces Base Borden (Bill Huff) provided sites and access to trees. The Ontario Tree Seed Plant (Al Foley) provided facilities.

Funding was provided by Natural Resources Canada - Alien Program, the

National Sciences and Engineering Council of Canada and the Ontario Ministry of Natural

Resources.

vi

Table of Contents

Abstract...... ii Acknowledgements...... iv Table of Contents...... vii List of Tables ...... ix List of Figures...... x List of Appendices ...... xii Introduction...... 1 Chapter 1: Literature Review...... 5 1 Insect-fungus complex...... 5 2 The life history of Sirex noctilio...... 8 3 Host trees ...... 12 4 Population factors ...... 13 4.1 Natality Factors...... 13 4.2 Population limiting factors...... 14 5 Interaction among woodboring insects and their fungal associates...... 15 6 Fungal interactions...... 18 7. Summary...... 20 Chapter 2: Aspects of Sirex noctilio host colonization ecology of P. sylvestris; preferred host tree condition, timing of attack, host death, and spatial and temporal variation ...... 21 Abstract...... 21 1 Introduction...... 22 2 Methods...... 24 2.1 Field sites ...... 24 2.2 Symptoms of S. noctilio oviposition activity...... 27 2.3 Field methods...... 28 2.4 Data Analysis...... 30 3 Results...... 31 4 Discussion...... 39 Chapter 3: Effect of two phloem- and woodborers-vectored fungi on the on-host search and oviposition behaviour of Sirex noctilio on trees and logs...... 42 Abstract...... 42 1 Introduction...... 42 2 Methods...... 46 2.1 Fungi ...... 46 2.2 Insects ...... 46 2.3 Field sites ...... 46 2.4 Experiment 1: Effect of L. wingfieldii and O. minus on S. noctilio behaviour in live trees...... 46 2.5 Experiment 2: Effect of L. wingfieldii and A. areolatum on S. noctilio behaviour on bolts...... 51 2.6 Data Analysis...... 52 3 Results...... 53 3.1 Experiment 1...... 53 3.2 Experiment 2...... 56 4 Discussion...... 58

vii

Chapter 4: Pinus spp. colonization by Sirex noctilio and it co-habitants in southern Ontario, and evidence for interaction ...... 61 Abstract...... 61 1 Introduction...... 62 2 Methods...... 64 2.1 Data Analysis...... 69 3 Results...... 70 3.1 Within tree distribution...... 72 3.2 Effect of tree characteristics on the phloem- and woodborers community ...... 75 3.3 Evidence for interactions ...... 76 4 Discussion...... 80 Chapter 5: Interactions between areolatum and two bark beetle-vectored fungi, Leptographium wingfieldii and Ophiostoma minus ...... 85 Abstract...... 85 1 Introduction...... 86 2 Materials and methods ...... 88 2.1 Fungal isolates ...... 88 2.2 Interactions between A. areolatum and ophiostomatoid species on artificial media 89 2.3 Interactions on wood substrate and effect of temperature ...... 89 2.4 Ability to grow on pre-colonized resource ...... 90 2.5 Data analysis ...... 91 3 Results...... 91 3.1 Interactions on artificial media ...... 91 3.2 Interactions on wood substrate at 25°C ...... 94 3.3 Effect of temperature on interactions on wood substrate...... 96 3.4 Growth on pre-colonized substrate ...... 99 4 Discussion...... 99 This paper has been accepted, with revisions, to The Canadian Entomologist ...... 104 Chapter 6: Synthesis and general discussion ...... 105 References...... 111 Appendix...... 123

viii

List of Tables

Table 2.1. Stand attributes and descriptions for all Pinus sylvestris field sites in southern and central Ontario used for all Chapter 2 studies of Sirex noctilio attack phenology, chronology of tree attack between S. noctilio and other phloem- and woodborers, tree mortality after S. noctilio attack, and S. noctilio host-tree preferences in 2007-08...... 26 Table 2.2. Speaman’s rank correlation coefficients for the association of the proportion of trees with symptoms of 2007 Sirex noctilio attack ( ≥ 5 Sirex noctilio resin beads), testing (< 5 resin beads) and the two pooled, with stand attributes of eight Pinus sylvestris field sites in southern Ontario. Significant findings (p<0.05) bolded...... 34 Table 2.3. Percentage of total trees attacked ( ≥ 5 Sirex noctilio resin beads or drips), tested (< 5 resin symptoms), presumed S. noctilio killed and total dead trees in each Pinus sylvestris study site in southern Ontario in 2007 and 2008...... 35 Table 2.4. Logistic regression and multiple logistic regression results for tree characteristics as predictors of Sirex noctilio attack ( ≥ 5 resin beads or drips) and testing (< 5 resin symptoms) for 503 Pinus sylvestris from eight field sites in southern Ontario. Multiple model components reflect best fit model using backward elimination procedure from significant variables on individual regression...... 37 Table 3.1. Logistic regression results for the effect of inoculated fungus ( , Leptographium wingfieldii or control), treatment (vs. control), and the degree of adventitious fungal growth on Sirex noctilio searching and drilling behaviours and drill scars in Experiment 2. Significant results bolded...... 57 Table 4.1. Geographical location, tree attributes and year felled for all study trees in southern and central Ontario...... 66 Table 4.2. Summary statistics for Sirex noctilio and most common co-habiting species emerging from Pinus spp. trees from southern and central Ontario. Emergence averaged over the number of trees the insect species emerged from (indicated in parenthesis)...... 71 Table 4.3 Repeated measures ANOVA results for effect of beetle presence on Sirex noctilio emergence from Pinus spp. in southern and central Ontario in 2007-2008. Emergence from the lower 4 (1 metre length) sections of tree stem compared...... 79 Table 5.1. Provenance information for the strains of Amylostereum areolatum , Leptographium wingfieldii and Ophiostoma minus used in all Chapter 5 experiments...... 88 Table 5.2. Mean per cent (±SE) surface area colonized by contending fungal strain at two weeks on potato dextrose agar and on Pinus sylvestris wood chips, all stored at 25°C. ANOVA results for difference in amount of substrate colonized on control substrate vs. that with each contending species...... 95 Table 5.3. T-test values and significance results of colonization of Pinus sylvestris woodchip surface area at 2 weeks by Amylostereum areolatum vs. three strains of each of two ophiostomatoid species (log n transformed) at temperatures between 15 and 30°C. In each case the ophiostomatoid species colonized more area than the A. areolatum strain...... 97

ix

List of Figures

Figure 2.1 Locations of eight Pinus sylvestris field study sites studies in southern and central Ontario used in all Chapter 2 studies...... 25 Figure 2.2. Resin bead (approximately 5mm in width) on live Pinus sylvestris as a result of Sirex noctilio oviposition drilling. This symptom was used to assess for evidence of S. noctilio activity for all Chapter 2 studies of S. noctilio attack phenology, chronology of tree attack between S. noctilio and other phloem- and woodborers, tree mortality after S. noctilio attack, and S. noctilio host-tree preferences in 2007-2008...... 27 Figure 2.3. Accumulation curves of the number of Pinus sylvestris per southern Ontario field study site with symptoms of Sirex noctilio oviposition activity (resin beads or drips) during each site re-assessment over the 2007 season a) tree as a whole, b) lower one third of the tree stem, c) middle one third, d) upper one third. H14 n = 22 trees, E1 n = 26, E6A n = 37, E6B n = 34, K1 n = 27, H18 n = 33, H7 n = 37, H15 n = 15...... 32 Figure 2.4. Accumulation curve of the number of Pinus sylvestris dead after symptoms of Sirex noctilio oviposition activity (resin beads or drips), for each southern Ontario study site with presumed S. noctilio induced tree mortality. E1 n = 26 trees, E6A n = 37, E6B n = 34, H18 n = 33, H7 n = 37...... 33 Figure 3.1. Sirex noctilio drill scars in Pinus sylvestris : a) shallow drill scar in phloem b) scar not extending into sapwood c) deep drill scar d) deep drill scar extending into sapwood. .... 47 Figure 3.2. Sirex noctilio cage used in Experiment 1 in situ on Pinus sylvestris ...... 49 Figure 3.3. Mean number of deep Sirex noctilio drill scars per Leptographium wingfieldii , Ophiostoma minus and control section of living Pinus sylvestris trees in 2008 and 2009. Results of Tukey’s Honestly-Significant-Difference test: lower case letters indicate significant differences between treatment groups in 2008 and upper case show the same in 2009...... 53 Figure 3.4. Mean number of shallow Sirex noctilio drill scars per Leptographium wingfieldii , Ophiostoma minus and control section on living Pinus sylvestris trees in 2008 and 2009. Results of Tukey’s Honestly-Significant-Difference test: lower case letters indicate significant differences between treatment groups in 2008 and upper case show the same in 2009...... 54 Figure 3.5. Number of quadrants searched, probed or drilled (+SE) by Sirex noctilio per Leptographium wingfieldii , Ophiostoma minus and control section of living Pinus sylvestris trees in a) 2008 and b) 2009. Significance results of G-test embedded...... 55 Figure 4.1. Collection cup arrangement for phloem- and woodborers rearing containers containing Pinus sylvestris bolts...... 68 Figure 4.2. Mean male and female Sirex noctilio emergence, standardized by tree surface area, per one metre section of tree stem from a) Pinus sylvestris (n = 36 trees) b) Pinus resinosa (n = 4) and c) Pinus banksiana (n= 5) from southern and central Ontario 2007-2008. Significance results for repeated measures ANOVA for woodwasp emergence per first four 1m sections of tree trunk embedded...... 73 Figure 4.3. Mean beetle emergence, standardized by tree surface area, per one metre section of Pinus spp. tree stem (tree species pooled) a) Tomicus piniperda (n = 27 trees), b) Pissodes nemorensis (n =3 4), c) Ips grandicollis (n = 33), d) Monochamus carolinensis (n = 13), e) Gnathotrichus materiarius (n = 13), from southern and central Ontario positive in 2007- 2008...... 74 Figure 4.4 Principle Components Analysis ordination scatterplot of phloem and woodboring beetle species (represented by ≥ 15 individuals) from Sirex noctilio -positive Pinus spp. from southern and central Ontario. A. pus = Neacanthocinus pusillus , A. sex = Astylopsis sexgutta , x

G. mat = Gnathotrichus materiarius , H. opac = Hylastes opacus , H. pin = Hylurgops rugipennis pinifex , I. grand = Ips grandicollis , M. car = Monochamus carolinensis , O. cael = Orthotomicus caelatus , P. nem = Pissodes nemorensis , T. sp. n = Tetropium sp. nov., T. pin = Tomicus piniperda , T. lin = Trypodendron lineatum , S. edw = Sirex edwardsii , S. nig = Sirex nigricornis , S. noct = Sirex noctilio , X. sax = Xyleborinus saxesini ...... 75 Figure 4.5. Primary co-habitants (>10 individuals) with Sirex noctilio in beetle-positive ( ≥17 beetles per tree) Pinus spp. trees (all tree species pooled) from southern and central Ontario in 2007-2008...... 77 Figure 4.6. Mean Sirex noctilio emergence, standardized by tree surface area, per one metre tree stem section in beetle-positive ( ≥ 17 beetles per tree) and “beetle-negative” ( ≤9 beetles per tree) Pinus spp. trees in southern and central Ontario (2007-2008). Effect of beetle presence results from ANCOVA embedded...... 79 Figure 5.1. Amylostereum areolatum surface area (+SE) colonized on potato dextrose agar in the presence of ophiostomatoid competitors over 6 days a) 075 7011 growth in presence of Leptographium wingfieldii strains, b) 075 7013 growth in presence of L. wingfieldii strains, c) 075 7011 growth in presence of Ophiostoma minus strains, d) 075 7013 growth in presence of O. minus strains...... 93 Figure 5.2 Surface area (+SE) of sterilized Pinus sylvestris woodchip colonized by each of two Amylostereum areolatum strains (075 7011 and 075 7011) in the presence of ophiostomatoid competitors at 7 days a) Both A. areolatum strains growing with 3 Ophiostoma minus strains compared to control growth, b) Both A. areolatum strains growing with 3 Leptographium wingfieldii strains compared to control growth. Results of Tukey’s HSD test indicated in caps for A. areolatum strain 075 7011 and in lower case for strain 075 7013...... 94 Figure 5.3. Surface area (±SE) of sterilized Pinus sylvestris woodchip colonized at 7 days by each of two Ontario Amylostereum areolatum strains at 10-30°C...... 96 Figure 5.4. Amylostereum areolatum colony area (±SE) in the presence of contending species at different temperatures at 2 weeks on wood chip substrate a) 075 7011 growth in presence of Leptographium wingfieldii strains, b) 075 7013 growth in presence of L. wingfieldii strains, c) 075 7011 growth in presence of Ophiostoma minus strains, d) 075 7013 growth in presence of O. minus strain...... 98

xi

List of Appendices

Appendix 1. Phloem- and woodboring insects collected from each tree species from pine forests in southern and central Ontario..……………………………….……………123

xii

Introduction

Sirex noctilio Fabricius (; Symphyta; Siricidae) is a woodboring wasp that has

elicited considerable attention over several decades for its ability to kill , especially in

regions where it has been introduced. It has a wide host range and most pine species appear

to be susceptible to its attack (Morgan & Stewart 1966a; Spradbery & Kirk 1978). Sirex

noctilio is the only woodwasp species known to be capable of physiologically stressing and killing trees (Spradbery 1973), though it generally favours severely stressed and declining hosts over healthier trees.

Sirex noctilio is native to , and North Africa, where it is considered a secondary pest of various pine species (Spradbery & Kirk 1978) and is of little economic or ecological concern (Hall 1968). During the 20 th century, S. noctilio became established in

several countries in the southern hemisphere (summarized by Ciesla 2003). In regions of this

introduced range, the wasp has become a primary pest of some pine species (Rawlings 1948).

In some situations it has caused extensive mortality in plantations of introduced pines,

especially in those affected by drought or other stress (Neumann & Minko 1981). In 2004, S.

noctilio was first discovered in North America (Hoebeke et al. 2005; de Groot et al. 2006).

The extent to which this insect will be of economic concern in this region remains to be

determined.

There are a number of factors that have influenced the pest status of S. noctilio in the

southern hemisphere. Poor site conditions, drought and the scarcity of natural enemies have

all contributed to woodwasp outbreaks (Taylor 1981; Neumann et al. 1987). Another, factor

that has been speculated on but not formally studied, is the effect of interactions between the

woodwasp and bark beetles or other woodborers. Pines have been introduced into the

southern hemisphere and because of their distinctiveness from native trees in these regions,

1 have been relatively free of insect pests (Wainhouse 2004) and, if present, it is these other insects that would potentially interact with the woodwasp. In contrast, in the wasp’s native range there is a well-developed community of phloem- and woodboring insects attacking P. sylvestris along with S. noctilio (Wermelinger et al. 2008). Similarly, in the woodwasp’s new range in North America, there is an abundance of other woodboring insects attacking pine including species of curculionids, cerambycids and buprestids (see USDA 1985 for examples). The community of phloem- and woodboring insects sharing trees with S. noctilio in North America has not been examined to date, so this is a rich avenue for research.

Sirex noctilio has a unique biology that contributes to the way in which it interacts with other insect species (reviewed in Chapter 1). The development of S. noctilio, and thus its

reproductive potential, is tied to the vigour of its obligate fungal symbiont, Amylostereum

areolatum (Chaillet ex Fr.) Boidin (e.g. Coutts & Dolezal 1965). Interactions between the

fungus and a phytotoxin produced by the wasp allow the insect to breach the defences of its

living host tree and kill it. The fungus is essential for egg eclosion and larval nutrition, and it

influences adult wasp size (Coutts & Dolezal 1965; Madden 1981); therefore, if the growth

of the symbiont is impacted so too is the wasp.

Interactions with other co-habiting insects could occur at different times in the

woodwasp’s life cycle. As hypothesized by Hanson (1939) and Spradbery and Kirk (1978),

the wasp may avoid trees, or parts of trees, already colonized by phloem- or woodboring

beetles and their fungal associates, thus influencing insect distribution (investigated in

Chapters 2 and 3). Fungal associates of co-habiting insects could also inhibit the growth of

the woodwasp’s symbiotic fungus (investigated in Chapter 5), thus impeding its offspring’s

survival, development or vigour (Hanson 1939; King 1966; Titze & Stahl 1970). These

mechanisms of interaction are speculative; no formal research exists in this area.

2

Evidence of interaction between species could include inter-tree partitioning, altered within-tree distribution, or changes in insect life history traits (investigated in Chapter 4).

Intra-tree partitioning is well described in bark beetle communities where some species are restricted to certain areas of the tree if other species are present; this is thought to occur through chemical cues (e.g. Flamm et al. 1987; Schlyter & Anderbrant 1993). If species share space and compete for resources, outcomes include reduced insect body mass or progeny production, or altered sex ratio (Andersen 1961; Rankin & Borden 1991; Schlyter &

Anderbrant 1993). Changes in these traits may in turn affect population growth of the species.

My thesis consists of a series of chapters each addressing a different aspect of the broad

topic of woodwasp interaction with pine-dwelling phloem- and woodborers. The first two

chapters cover the relevant background material. Chapter 1 consists of a general literature

review of the topic and identifies areas of research opportunities. In Chapter 2, I investigate

and describe aspects of S. noctilio’s ecology that were necessary to understand before

pursuing my main research questions. Specifically, it addresses the questions: i) what are the

characteristics of the host tree condition favoured by the woodwasp, ii) what is the timing

and chronology of attack between the woodwasp and phloem- or woodborers, iii) how

quickly does a host tree die following woodwasp attack, and iv) what is the spatial and

temporal variability in woodwasp activity and host mortality.

Chapters 3, 4 and 5 each represent a separate manuscript prepared for publication and are

modified only slightly from manuscript format, thus there is inherent repetition between

them. Chapter 3 consists of two experimental investigations of the effect of selected bark

beetle associated fungi on S. noctilio . In these experiments I addressed the questions i) does

S. noctilio avoid ovipositing in wood already colonized by ophiostomatoid fungi, and ii) is its

offspring’s development affected in the presence of ophiostomatoid fungi. In Chapter 4, I

3 describe subcortical insect community of S. noctilio -infested pines and examine for evidence

of interaction between the two groups, specifically I: i) describe the community of phloem-

and woodborers habiting the tree with the woodwasp, ii) describe the distribution of the

woodwasp and the most common phloem- and woodborers within the tree, and iii) look for

signs of interguild interaction or competition including inter-tree partitioning, altered within-

tree distribution and the effect of co-habiting species on S. noctilio’s life history traits.

Chapter 5 consists of a series of laboratory experiments evaluating the outcomes of

competition between the woodwasp’s symbiotic fungus and two species of bark beetle-

vectored fungi. In this chapter I sought: i) to evaluate the outcomes of primary competition

for uncolonized substrate between A. areolatum and two bark beetle-associated fungi ii) to

examine how temperature modifies the outcomes of these interactions; iii) to evaluate the

ability of A. areolatum to establish in already occupied substrates (secondary resource

capture); and iv) to evaluate the effect of substrate on the outcomes of experiments between

wood-habiting fungi. Chapter 6 is a synthesis of the main findings of each chapter and a

general discussion.

4

Chapter 1

Literature Review

1 Insect-fungus complex Sirex noctilio’s symbiont, Amylostereum areolatum , derives considerable benefit from its relationship with the woodwasp. It is stored and protected in the wasp’s mycangia before the insect drills through the protective bark of the tree and introduces the fungus directly into a suitable, relatively competition-free, host substrate. Amylostereum areolatum primarily spreads vegetatively and several clonal lineages over wide geographic areas have been found

(e.g. Vasiliauskas et al. 1998; Vasiliauskas & Stenlid 1999; Thomsen & Koch 1999; Slippers et al. 2001). The close association of A. areolatum with its woodwasp symbiont is thought to be a reason that the fungus rarely reproduces sexually in some areas of its native range, and fruiting bodies have not, so far, been found in its introduced range (Gilbertson 1984).

Sirex noctilio is specialized to ensure the continued relationship with its symbiont.

Adult females carry the fungus within paired mycangia located at the anterior end of the ovipositor and that open into the oviduct (Boros 1968). The mycangia are lined with glands and the secretions produced appear to stimulate fungal growth (reviewed in Morgan 1968).

Upon oviposition the female woodwasp inoculates arthrospores or fragments of its fungus into the host sapwood through her ovipositor and into a separate drill beside the egg

(oviposition is described in detail in the following section). At times, no egg is deposited but the fungus and the phytotoxic mucus (discussed below) are (Coutts & Dolezal 1969). From the second instar onward, the fungus is transferred from one larval instar to the next and the adult female takes it up into her mycangia when she sheds her pupal skin, thus ensuring the continued association with the fungus (Parkin 1941; Boros 1968). The fungus is thought to

5 be absent in the wasp’s pupal stage but subsequently taken up from the wall of the pupal chamber by the adult (Francke-Grosman 1939; Parkin 1942 but see Cartwright 1938).

The woodwasp produces a phytotoxic mucus that is inoculated into the tree, along with the fungus, during oviposition drilling activity (Coutts & Dolezal 1969). Wong and

Crowden (1976) describe the mucus as a protein-polysaccharide complex containing several including amylase, esterase, phenoloxidase and proteolytic enzymes. The mucus is synthesized in a pair of secretory glands and is stored in a mucus reservoir until use; a duct connects the reservoir to the base of the wasp’s oviduct (Boros 1968). Sirex noctilio has

larger mucus glands and reservoir than other siricid species and this may contribute to its

ability to kill trees while its congenerics do not (Spradbery 1977).

The inoculation of A. areolatum into a tree on its own has no deleterious effect, but

when injected into a tree in combination with the wasp’s phytotoxic mucus, it appears to

function as part of a pathosystem (Vaartaga & King 1964; Coutts 1969a). In tandem the

fungus and the mucus cause severe physiological stress to the tree, which includes impaired

water relations and translocation and this is often followed by tree death (Coutts 1969b; Fong

& Crowden 1973). Symptoms of toxicity to the tree include foliar chlorosis, often within two

weeks, or senescence with or without chlorosis (Coutts 1969b). Needle wilting is also

described (e.g. Neumann et al. 1987) but this has not been seen in North America (Chapter

2). Sirex noctilio is the only woodwasp recognized to be capable of triggering this degree of

physiological response. This probably relates to both mucus production and behaviour; of all

of the woodwasps, S. noctilio not only produces the greatest amount of mucus, it also shows

the greatest density of oviposition drill sites (Spradbery 1973).

In addition to its role in stressing the tree, S. noctilio ’s mucus has direct effects on its

fungal symbiont. The mucus has been shown to release A. areolatum from the wax packet

6 that surrounds it while stored in the wasp’s mycangia, and stimulates its growth (Boros 1968;

Titze & Turnbull 1970).

The role of fungi in the nutrition of woodwasps is not entirely clear. Kukor and

Martin (1983) demonstrate the role of fungus-derived cellulases in the digestion of by a congeneric woodwasp species, Sirex cyaneus Fabricius. Other authors report direct consumption of mycelia-impregnated wood in S. noctilio (Büchner 1928 in Büchner 1965).

Fungal hyphae are digested by gut secretions of the larvae (Francke-Grosman 1939) and in lab conditions siricid larvae can feed on pure fungus (Cartwright 1929). The role of the fungal symbiont in woodwasp nutrition may depend on the life stage of the insect. Madden and Coutts (1979) suggest that first-, and some second-instar larvae feed exclusively on the fungus, while later larval stages feed on fungus-colonized wood. Regardless of the specific role of the fungus in larval nutrition, it is essential for the developing larvae. The adult woodwasp is thought to not feed and to survive only on stored fat-body reserves (Taylor

1981).

The presence of A. areolatum is essential throughout the development of the immature woodwasp. Egg eclosion is delayed when conditions in the tree impede fungal growth (Madden 1981) and larvae may starve if symbiont growth is inhibited by the presence of other fungi (Coutts & Dolezal 1965; King 1966). Larval development relates to fungal growth and when conditions are optimal for the fungus, larger adults are produced (Madden

1981). The fungus also modifies environmental conditions. The wood-decaying symbiont dries the wood substrate providing a more suitable micro-environment for egg and larvae development (Coutts & Dolezal 1965) and wood degradation by the fungus facilitates tunneling of the larvae (Gilmour 1965).

7

2 The life history of Sirex noctilio Sirex noctilio has a one- to three-year life cycle: the one-year life cycle is the most common throughout its introduced and native range (Morgan 1968). Development occurs above a 6.8

°C threshold and requires 2500 degree days (Madden 1981), so at a constant 30°C a woodwasp would take approximately 90 days to complete its life cycle, while at a constant

10°C it would take about 360 days. This fits with Neuman and Minko’s (1981) observation that some wasps may take as little as three months to develop in south-eastern .

Within a single tree, development times may differ; wasps in the upper trunk may develop in one year while those in the lower trunk may take two years (Morgan 1968). This difference is thought to relate to lower wood moisture conditions in the upper, compared to the lower, bole.

Adult S. noctilio emerge from early or mid-summer to early autumn. In its native

range in Eurasia, emergence is from mid-August to mid-November (Spradbery & Kirk 1978).

In the southern hemisphere, emergence times vary between regions and range between mid-

October and May (Rawlings 1948; Morgan & Stewart 1966a; Taylor 1981; Hurley et al.

2008). Emergence times and lengths vary, even within the same region, and this depends on

climate (Neumann et al. 1987). Regardless of this variation, most wasp emergence is

concentrated over a period of a few weeks. Emergence patterns vary between regions, some

have uni-modal peaks of emergence, others bimodal (Taylor 1981; Neumann et al. 1987).

When there is a second peak of emergence it is thought to be a result of the emergence of

short-cycle individuals (those with life cycles of a few months) (Neumann et al. 1987).

Emergence is favoured by above average temperatures and falling barometric pressure

(Morgan 1968; Taylor 1981). In cool, wet years emergence may be haphazard and without a

clear peak (Morgan & Stewart 1966a).

8

Emergence of adult S. noctilio is protandrous; the males emerge a few days before females (e.g. Rawlings 1948; Morgan & Stewart 1966a). Wasps rest on the bole before moving to the crown of the tree where males aggregate and are joined by females for mating

(Morgan & Stewart 1966a; Madden 1988); wasps are sexually mature at emergence

(Neumann et al. 1987). Sunny, warm conditions (>21°C) have been observed to stimulate copulation (Morgan & Stewart 1966a) and cold, wet days to hamper it (Dolezal 1967). Life spans are short; males live for up to 12 days and females up to five days (Neumann et al.

1987). Smaller wasps tend to have shorter adult life spans than larger wasps (Madden 1974).

As life-spans are short, prolonged emergence periods in cool rainy seasons may decrease mate-finding, although this may be compensated by greater longevity in cool autumn weather

(Neumann et al. 1987).

Following mating, females locate a tree for oviposition, but when mating does not

happen host location occurs after an initial period of flight activity (Madden 1988). After a

potential host tree is located, the female explores the bole with her antennae and then probes

into the sapwood with her ovipositor (Francke-Grosman 1939; Madden 1988). Trees with

phloem sap osmotic pressures > 18.0 X10 5 Pascals are rejected by the female wasp (Madden

& Coutts 1979). When a suitable host tree is found, the female typically begins laying eggs

near the bottom of the tree and works her way up, drilling single, double, treble or quadruple

sets of oviposition chambers every 7.5-50 cm along the stem (Rawlings & Wilson 1949;

Morgan & Stewart 1966a). Oviposition generally occurs along the entire stem in young trees

and between 3 and 14 m above ground in older trees (Neumann et al. 1982). Eggs are

generally laid around the stem (Coutts & Dolezal 1969). When the bark texture is uniform,

the distribution of oviposition sites on successfully attacked trees tends to be random

(Madden 1974), however, exposed areas on the lower stem may be preferred, especially if

they are sun-exposed (Rawlings & Wilson 1949). Several females can attack a single tree

9

(Bedding & Akhurst 1974). Oviposition rate is influenced by ambient conditions and increases with temperature to a maximum at around 21° C and declines at higher temperatures. Relative humidity also affects egg laying and is most favourable between 49-

52 % and declines above or below this range (Madden 1974). Eggs are laid in sapwood containing between 20 and 200 % moisture, but the woodwasp prefers a moisture content of about 60 % (Morgan & Stewart 1966a).

Oviposition is up to 19 mm into the sapwood (Gilbertson 1984) depending on the

length of the wasp’s ovipositor which ranges from 8 to 20 mm (Coutts 1965). Depending on

site quality, the female generally lays up to three eggs per drill site, each in separate tunnels

(Madden 1974; Madden 1988). A final, eggless tunnel at each drill site is inoculated with

arthrospores or fragments of A. areolatum , as well as the mucus (Coutts & Dolezal 1969).

Sirex noctilio is also known to make single drills, introducing only the mucus and fungus,

perhaps when, after testing the suitability of the tree for oviposition, she finds it unacceptable

(Coutts & Dolezal 1969; Spradbery 1977). These test drills are thought to condition the tree

and make it prone to further attack (Madden & Irvine 1971).

Egg eclosion is preceded by growth of the fungal symbiont around the egg (Madden

1981). Hatching usually occurs in 16-28 days (Morgan & Stewart 1966a) depending on

temperature conditions but ranges between 10 (at 30°C) and 60 (at 10°C) days (Madden

1981). The temperature threshold for eclosion is 6.2 °C, which differs slightly from the

developmental threshold of 6.8° C, but Madden (1981) does not consider this difference

significant. Eclosion can be delayed, up to 12 months (Waterhouse & Sands 2001), if high

moisture conditions or low temperatures, such as those found at tree bases, impede fungal

growth (Madden 1981). A wood moisture content of between 40 and 70 % is most favourable

for egg eclosion and the development of early-instar larvae (Morgan & Stewart 1966a).

10

Larvae begin tunnelling more deeply in the sapwood after the third or fourth instar

(Taylor 1981). They bore through the sapwood, excavating tunnels that have been measured up to 26 cm long (Neumann et al. 1987). Galleries are primarily oriented along the wood grain; specific shapes vary and boring patterns depend on wood moisture (Morgan 1968;

Madden 1981). There are between six and 12 larval instars, depending on conditions within the sapwood; larger adults tending to have had more instars than smaller ones (Madden

1981). Adults may differ in size, even within the same area of the tree, depending on wood moisture conditions – smaller adults emerge from drier wood (Coutts 1965; Madden &

Coutts 1979). Once larval development is complete, the tunnels outward and becomes a prepupae within 5 cm of the bark surface (Taylor 1981). The prepupal period lasts up to four weeks, while the pupal stage lasts approximately 20-28 days (Morgan & Stewart 1966a). In cooler, wetter conditions, pupation may not occur until the second or third year after eclosion

(Taylor 1981).

Natural dispersal of the woodwasp is estimated to be up to 30-50 km per year

(Haugen et al. 1990). Recent investigations using a flight mill showed high variability in flight ability between individuals. Female wasps fly between 1.1 and 49.7 km (mean 17.4 km) over a 20-hour period (Bruzzone et al. 2009). However, flight distance depends on parasitism; Villicide and Corley (2008) found that the average flight distance of uninfected females is double that of parasitized ones (30 vs. 16 km / 24h). These authors show that parasitized females are smaller and lose more weight in flight than uninfected wasps. The natural spread of woodwasps is easily augmented by the movement of infested wood because all immature life stages can be transported in unprocessed logs, and untreated lumber and solid-wood packing materials (Haugen 2006).

11

3 Host trees Sirex noctilio has a wide host range and most pine species, including species common in

Canada, for example, Pinus banksiana Lamb., P. contorta Dougl. ex Loud., P. ponderosa

Dougl. Ex P. & C. Laws., P. sylvestris L., Pinus resinosa Ait. and P. strobus L. appear to be

susceptible (Morgan & Stewart 1966a; Spradbery & Kirk 1978; Ciesla 2003). The wasp

prefers low-vigour trees weakened by other insects (Spradbery & Kirk 1978) or mechanical

damage, suppressed trees (Hall 1968) and those suffering from drought stress (Wermelinger

et al. 2008). In some cases, the woodwasp is also thought to be a primary cause of tree death

(Rawlings 1948), though the conditions under which this occurs are not described by the

author.

Madden (1977) describes the mechanism of the wasp’s attraction to a physiologically

stressed host and the subsequent positive feedback loop. Pines stressed by crown damage

have impaired translocation, along with enhanced transpiration and phloem respiration. As a

result, these trees have greater tissue permeability and a higher rate of and

water vapour loss through the bark. This in turn attracts S. noctilio to these trees and is

correlated with greater oviposition activity. Attack by the wasp, with its attendant mucus and

fungal inoculations into the tree, further impairs translocation and increases physiological

stress. These conditions favour A. areolatum colonization and reduce the defensive response

of the tree. In some regions emergence is concomitant with low soil moisture conditions and

therefore with suppressed physiological activity and lowered resistance of host trees

(Madden & Coutts 1979); this may contribute to the wasp’s ability to overcome tree

defences.

12

4 Population factors

4.1 Natality Factors The reproductive potential of the female woodwasp can be calculated with the equation: total

# eggs/female = 28.8(52.5) x, where x = prothorax width in cm (Madden 1974). Using this method, estimates of S. noctilio fecundity range from 21-458 per female (Neumann et al.

1987). This corresponds with an average of approximately 264 eggs per female found by

Spradbery and Kirk (1981) who counted the number of eggs in female ovaries on dissection.

Different factors modify potential fecundity and for S. noctilio these factors include weather and insect size. In near ideal environmental conditions (23°C, 45% RH) the female, on average, can be expected to lay 82.6% of her eggs (Neumann et al. 1987). In addition to influencing fecundity, female size may affect fertility as smaller females tend to have more unlaid eggs than larger ones (Madden 1974).

The sex ratio of Sirex noctilio shows spatial and temporal variability. Although the offspring sex ratio is 1:1 (M:F) for mated females under controlled conditions (Morgan &

Stewart 1966a), in most regions where the insect has been introduced, males exceed females by between 1.5 and 32:1 (Morgan & Stewart 1966a; Taylor 1981; Klasmer et al. 1998; Iede et al. 1998). In comparison, the sex ratio in a European study is approximately 1.82:1, corresponding with other species of woodwasps in the same region (Spradbery & Kirk 1978).

These ranges may relate to the state of the insect’s population; when the population density is increasing the sex ratio is thought to approach 1:1 but when it is declining the ratio increases

(Morgan & Stewart 1966a). Since S. noctilio is arrhenotokous we should expect to see an increasing sex ratio when the population is declining and mate location becomes more difficult. Sex ratio though may be tree specific and may depend on individual host tree susceptibility (Taylor 1981).

13

4.2 Population limiting factors Several parasitoids, primarily ibaliid or rhyssine species, are known to parasitize S. noctilio and other woodwasps; ibaliids attack the egg and first- or second-instar larvae, and rhyssines attack later larval stages (Taylor 1978). Woodwasp parasitoids are well known from Europe and North America (e.g. Cameron 1965) and this guided the selection of parasitoids for introductions in Australia where they were introduced for biocontrol (reviewed in Murphy

1998). In the southern hemisphere, parasitism by Ibalia species is < 10% in Tasmania but is

between 20 and 40% in most other regions (Taylor 1978; Neumann et al. 1987; Iede et al.

1998; Murphy 1998). Several authors show that parasitism by rhyssines has the potential to

exceed that of ibaliids (Taylor 1978; Spradbery & Kirk 1978; Nuttall 1989 in Hurley et al.

2007), and as a group may cause parasitism rates up to 70%. However, these results are

inconsistent, and in some areas of introduction rhyssines have failed completely, showed low

infestation rates, or the results are unknown (Neumann et al. 1987). Both ibaliad and rhyssine

species are already associated with trees infested by S. noctilio in North America (Long et al.

2009) but parasitism rates are not yet described.

Deladenus (Beddingia) siricidicola Bedding is the only nematode species found

parasitizing S. noctilio in Europe (Bedding & Akhurst 1978). This nematode has been used as

an effective biocontrol agent in Australia and New Zealand where it causes 70-100%

infection rates (Bedding & Akhurst 1974; Zondag 1979; Bedding & Iede 2005). It has been

used with varying degrees of success elsewhere in the southern hemisphere where infection

rates range from < 5% to 85% (reviewed in Hurley et al. 2007). This nematode species has

been detected in Canada (Yu et al. 2009); however, its prevalence and its effect on the

woodwasp are yet to be determined. Research directed toward the use of this organism for

biocontrol of the woodwasp in North America is currently being conducted (Williams &

Mastro 2008).

14

Other described mortality agents of S. noctilio are birds and viruses. Both

woodpeckers (larval predators) and aerial hunters (adult insect) predate S. noctilio but the

extent to which they reduce the woodwasp population is difficult to quantify (Madden 1982;

Spradbery 1990). A cytoplasmic polyhedral virus has been identified in S. noctilio in

Germany (Talbot 1977), however, its prevalence, effect on the insect, and presence outside of

Europe are unstudied.

Another factor, which has not been studied, is the effect of interactions with bark or woodboring beetles on the biological control of S. noctilio . Pines have been introduced to

regions in the southern hemisphere and are chemically distinctive from native trees in these

areas and this has allowed them to, at least initially, remain relatively free of insect pests

(Wainhouse 2004). Therefore, we can expect that there are relatively few subcortical insects

sharing the tree with the woodwasp in this range. In contrast, there is a well-developed

community of these insects attacking pine along with S. noctilio in the wasp’s native range

(Wermelinger et al. 2008) so these insects could be a factor in limiting populations of the

wasp in this area. Similarly, in North America there are a number of curculionids,

cerambycids and buprestids attacking pine (see USDA1985 for examples).

5 Interaction among phloem- and woodboring insects and their fungal associates Interactions among phloem- and woodboring insects may be direct or indirect. Larvae of one species may be predated by other insect larvae, although this is expected to be a result of indiscriminate feeding rather than obligatory (Schenk & Benjamin 1969; Dodds et al. 2001) and therefore may have limited effect on the population size of a species. Interactions may also occur indirectly through the inhibition of insect colonization. Vertical partitioning of the trunk of standing host trees is well documented in bark beetle communities (Paine et al.

15

1981; Flamm et al. 1987; Schlyter & Anderbrant 1993; Hui & Xue-Song 1999; Ayres et al.

2001) and is thought to be a result of a chemical feedback mechanisms functioning to reduce competition between species (Flamm et al. 1987; Schlyter & Anderbrant 1993). When different species do share space and compete for resources, possible outcomes include reduced body mass or progeny production of one or both species (Rankin & Borden 1991;

Schlyter & Anderbrant 1993).

Ophiostomatoid fungi are frequently associated with bark and longhorn beetles, or the phoretic mites associated with them, and these organisms can inoculate them into host trees during attack (Bridges and Moser 1983; Paine et al. 1997; Jankowiak and Rossa 2007;

Jankowiak and Kolařik 2010). The presence of these fungi can inhibit colonization or oviposition by bark beetle species, for example, the presence of Ophiostoma ips (Rumbold)

Nannf. reduces colonization by Ips sp. in pine (Yearian et al. 1972; Kopper et al. 2004).

Although there are few studies documenting insect avoidance of fungus-colonized-substrate in the bark beetle community, this phenomenon is also evident in other systems. The leaf beetle, Oreina cacaliae (Schrank), prefers ovipositing on leaves that are free from the rust

fungus Uromyces cacaliae (D.C.) Unger (Röder et al. 2007), and the house fly, Musca

domestica L., avoids oviposting in feces colonized by certain species of fungi harmful to it

(Lam et al. 2010). Both Hanson (1939) and Spradbery and Kirk (1978) observed that S. noctilio avoid ovipositing in trees colonized by bark beetle-vectored-fungi, so this mechanism may be important for the woodwasp as well, though such observations remain anecdotal. This may be of reproductive benefit to the wasp, as the presence of ophiostomatoid fungi is thought to cause larval (Morgan & Stewart 1966b) or pupal (Hanson

1939) mortality.

Ophiostomatoid fungi vectored by bark beetles or associated mites may have more

direct effects on an insect species, either positive or negative, if the insect colonizes the tree

16 during the time the fungus is in the tree. The primary benefit of this relationship is improved insect nutrition; fungi may also help the insect to overcome the defences of its host tree but this is controversial (Paine et al. 1997). Enzymes produced by fungal symbionts can break down wood tissue (Valiev et al 2009) making nutrients accessible to the beetles.

Ophiostomatoid fungi may also concentrate phloem , phosphorus and potassium in their mycelia providing a nutritious food source for bark beetles and in addition can provide vitamins, sterols and other growth factors (Ayres et al. 2000; Dajoz 2000; Six 2003; Bleiker and Six 2007). There are several examples of developmental benefits of incidental mycophagy in these beetles including larger body size, the production of more progeny and reduced developmental times (Whitney 1982; Bleiker and Six 2007). In contrast, the presence of certain species fungi may be detrimental to the insect offspring; reduced brood production in the southern pine beetle Dendroctonus frontalis Zimmerman is found in the presence of Ophiostoma minus (Hedgc.) Syd. & P. Syd. (Barras 1970; Hofstetter et al. 2006).

The phoretic fungus O. minus outcompetes both of the mycangial fungi, Ceratocystiopsis ranaculosa J.R. Bridges & T.J. Perry and Entomocorticium sp A (Klepzig & Wilkens 1997), associated with D. frontalis . This fungal competition can have practical implications, for

example, another ophiostomatoid species, Ophiostoma piliferum (Fr.) Syd. & P. Syd has

been investigated as a potential biological control agent for this beetle (Klepzig 1998).

As previously described, S. noctilio is especially dependent on its fungal symbiont

which is essential for egg eclosion and larval nutrition, and symbiont vigour contributes to

larger adult insect size and therefore reproductive success as well. Given this, if the growth

of A. areolatum is impeded by other fungi, this would have considerable impact on the

population dynamics of the woodwasp. It has been suggested by some authors that A. areolatum is a weak competitor with fungi vectored by phloem- or woodboring beetles and

that this negative interaction may subsequently impede the development of S. noctilio larvae

17

(Hanson 1939; Morgan & Stewart 1966b; King 1966; Titze & Stahl 1970). However, these suggestions are speculative as the only specific description of such interaction is by King

(1966) who simply notes that Trichoderma sp. and Sphaeropsis sapinea (Fr.) Dyko & B.

Sutton were strongly antagonistic to A. areolatum but did not quantify the effects.

6 Fungal interactions Much of the knowledge on fungal interactions to date is based on the work of Rayner and his associates who describe interactions between fungi as being either competitive, neutralistic or mutualistic in nature (Cooke & Rayner 1984 in Rayner & Webber 1984; Rayner & Boddy

1988; Boddy 2000). Neutralistic interactions include situations where mixtures of enzymes from two fungi growing together are more effective than each alone, or where the activity of one fungus releases nutrients for use by another. Mutualistic interactions are those where the vegetative or reproductive growth of one fungus is enhanced by the presence of another (e.g. stimulation of sporulation). Evidence of these types of interactions in culture is the intergrowth of healthy fungal mycelia; however, this evidence depends on substrate conditions and mycelial morphology. These interactions may occur when the specific nutrient sources are not directly contested by either of the two fungi.

Competitive interactions are much more extensively described by the above authors than neutralistic or mutualistic ones and these interactions may be categorized in different ways. Competitive strategies are described as being those directed toward either primary or secondary resource capture. Primary resource capture strategies are those in which the fungus rapidly exploits the uncolonized resource and gains control over it; success is due to effective dispersal to the resource, having effective enzymes to exploit the resource, and rapid growth rates. Secondary resource capture occurs when one fungal species is able to gain access to area already colonized by another fungal species and by parasitism of an earlier colonizer 18

(i.e. a selective replacement) or by non-selective replacement. Non-selective replacement is the more common of the two and occurs as pioneer species are replaced during fungal succession. Antagonistic interactions between fungal species may occur when two species meet during primary resource capture and this may result in either a deadlock between the two species, or the partial or total replacement of one species by another. Inhibition of fungal competitors can occur at a distance or via contact. Three mechanisms of antagonistic interactions are described in the above cited literature: antibiosis, hyphal contact and gross mycelial contact. The inhibition of one or both fungal participants from a distance

(antibiosis) can be mediated via fungal waste products, the alteration of substrate pH, or by the production of volatiles or other substances that diffuse through the substrate (Rayner &

Webber 1984). Sensitivity to antibiosis is variable and may be specific to certain fungal pairs

(Rayner & Webber 1984; Boddy 2000), and therefore is situation specific. Contact between hyphae of two species may cause changes in hyphal membrane permeability and eventually result in the death of one or both hyphal compartments; this occurrence is thought to be more common in certain fungal groups including wood-decaying basidiomycetes (Rayner &

Webber 1984). Gross mycelial contact between two species may also inhibit fungal colonization when a morphological change at the contact zone, such as an increase in mycelial density, prevents progression of the other species.

The results of interactions between fungi are specific to the species involved (e.g.

Klepzig & Wilkens 1997). Outcomes may be influenced by a number of factors including inoculum size (Holmer & Stenlid 1993) and environmental factors such as wood moisture

(Klepzig et al. 2004), temperature (Hofstetter et al. 2007), secondary metabolites of the host- tree (Hofstetter et al. 2005) and gases (CO 2 and O 2) (reviewed in Rayner & Boddy1988). Oral secretions from bark beetles can inhibit the growth of some ophiostomatoid species (Cardoza et al 2006) and volatiles from the bacterial symbionts of these beetles can either stimulate or

19 inhibit growth of fungal associates (Adams et al 2009) and this could also influence the outcomes of interactions.

7. Summary Sirex noctilio attacks living trees and overcomes their defences through the combined effect

of the phytotoxic mucus that the wasp manufactures and the symbiotic fungus, A. areolatum ,

it carries. Pines that are suppressed or stressed by factors such as climate conditions, injury

and pathogens or pests are particularly vulnerable to S. noctilio attack and this further

increases the attractiveness of the tree to the wasp. An abundance of these vulnerable trees,

together with insufficient natural enemies, can contribute to high populations of S. noctilio ,

as evident in its introduced range in the southern hemisphere. One factor that may also

influence the woodwasp’s population over time is the presence of co-habiting insects, as

these could affect the wasp’s distribution or its life history traits.

Amylostereum areolatum is not only essential to help the woodwasp overcome the tree’s defences, but also for the nutrition and development of the wasp larvae. Good fungal growth enhances wasp nutrition and facilitates larger offspring body sizes. Wasp longevity, fecundity and flight potential all relate to body size, and are important factors facilitating its population growth and expansion. Thus, factors influencing fungal growth will also influence

S. noctilio populations and spread rates.

The paucity of ecological information describing the biological interactions between

S. noctilio and phloem- and woodborers, and their associated fungi, especially in North

America, indicates research in this area should be conducted to explore whether and how these interactions could contribute to the biological control of this insect.

Sections of Chapter 1 will be included in Ryan, K. and B. Hurley. Life history and biology. In The Sirex Woodwasp: Expanding Frontiers - 100 years of dealing with an invasive, mutualistic insect-fungus forest pestilence.

20

Chapter 2

Aspects of Sirex noctilio colonization ecology of Pinus sylvestris ; preferred host tree condition, timing of attack, host death, and spatial and temporal variation

Abstract The woodwasp Sirex noctilio has been recently introduced into North America. Throughout much of its introduced range in the southern hemisphere it has had significant economic impact on Pinus spp. , though in its native Eurasian range it does not. The woodwasp is expected to encounter a number of other pine-inhabiting insects, both in Eurasia and in North

America, and interaction with these insects may contribute to its population control in regions where pine is naturally found. However, there is little information about the wasp’s ecology in its new range and this limits research on these interactions. In this study, several factors were clarified in order to study potential interactions between the woodwasp and other phloem- and woodborers: the wasp’s preferred host tree condition, timing of oviposition activity, chronology of attack between the woodwasp and other insects, tree mortality after attack, and the spatial and temporal variability in wasp activity. In this study,

231 Pinus sylvestris in Ontario were examined bi-weekly to describe the timing of symptoms of oviposition activity by S. noctilio , the sequence of insect feeding guilds (i.e. siricids,

curculoinids and cerambycids) entering the tree, and the occurrence and timing of tree

mortality after woodwasp activity. I assessed 503 trees for symptoms of woodwasp

oviposition activity and evaluated tree attributes that may predict the trees favourability to

the wasp for oviposition. Sirex noctilio oviposition activity was evident between mid-July

and late August and other insects entered the tree after the woodwasp; cerambycid beetles

shortly after S. noctilio attack and bark beetles the following spring. Tree death after

21 woodwasp attack varied between sites; trees died as early as two weeks after woodwasp activity in one site but in other sites the trees died over the following winter. The amount of residual tree canopy foliage was the strongest predictor of S. noctilio attack. There is great potential for the woodwasp to interact with a number of curculionid and cerambycid beetles in Pinus spp. and spatial and temporal variability of woodwasp activity may influence these interactions.

1 Introduction

Sirex noctilio is an introduced woodwasp recently detected in eastern North America

(Hoebeke et al. 2005; de Groot et al. 2006). This introduction is of particular concern because in its introduced range in the southern hemisphere it has caused serious economic losses

(Bedding & Iede 2005) and predictions indicate that there may be considerable impacts in parts of its introduced range in North America as well (Yemshanov et al. 2009). In contrast, in its native range in Eurasia and northern Africa it is a secondary pest of no particular economic significance (Hall 1968).

There are a number of factors that influence the impact of S. noctilio in the southern hemisphere. Poor site conditions, drought and the paucity of natural enemies have all contributed to woodwasp outbreaks (Taylor 1981; Neumann et al. 1987). Another factor, which has not been well studied, is the potential effect of interactions with other phloem- or woodboring insects on the biological control of the woodwasp. Pinus spp. have been

introduced to these southern hemisphere regions and because of their distinctiveness from

native trees in these regions, have been relatively free of insect pests (Wainhouse 2004). In

contrast, in the wasp’s native range there is a well developed community of phloem- and

woodboring insects attacking pine along with S. noctilio (Wermelinger et al. 2008) and

22 interactions with them could help to limit woodwasp populations. Similarly, there is an abundance of curculionids, cerambycids and buprestids attacking Pinus spp. in North

America (see USDA 1985 for examples). Knowledge of the relevance of this potential population controlling factor in North America is needed to predict the wasp’s impact here.

The first step to understanding these interactions is elucidating the ecology of the woodwasp in North America.

In its introduced range in the southern hemisphere, the woodwasp favours trees that are suppressed or those that are physiologically stressed by environmental conditions, mechanical injury or by other insect pests or pathogens (Neumann & Minko 1981; Neumann et al. 1987; Madden 1988). In order to study interactions between the woodwasp and other co-habiting insects, a sound description of the condition and attributes of trees favoured by the woodwasp in North America is required. This will allow prediction of the insect species that the woodwasp is most likely to interact with since each of these insects has certain host tree condition preferences. For example, the curculionid, Tomicus piniperda (L.) prefers to

lay eggs in living pines with declining crowns (Morgan et al. 2004) whereas the cerambycid

Monochamus species often attack already dead pines (USDA 1985) and this would mean that

these two groups could interact very differently with the woodwasp.

Interactions between S. noctilio and other insects could occur through a variety of

mechanisms. The wasp may avoid trees already colonized by phloem- or woodboring beetles

and their fungal associates (Hanson 1939; Spradbery & Kirk 1978). Alternatively, the

presence of fungal species vectored by co-habiting insects could inhibit the growth of the

woodwasp’s own symbiotic fungus, which is essential for the successful development of its

offspring (Hanson 1939; King 1966; Titze & Stahl 1970), or cause wasp larval or pupal

mortality (Morgan & Stewart 1966b). Therefore, we must know host colonization phenology,

and the chronology of attack by the woodwasp as well as any co-habitants to determine

23 which of these mechanisms of interaction warrant further exploration for its effect on the woodwasp’s behaviour or development.

The woodwasp carries a symbiotic fungus, Amylostereum areolatum , and manufactures a phytotoxic mucus and both substances are inoculated into the tree’s sapwood when the insect oviposits (Coutts & Dolezal 1969). The fungus and the mucus together cause severe physiological stress to the tree, which in turn causes foliar chlorosis or senescence, and may be followed by tree death (Coutts 1969a; Coutts 1969b; Fong & Crowden 1973).

There is a succession of phloem- and woodboring insects colonizing trees as they decline and die (Dajoz 2000). This duration of decline between S. noctilio attack and tree death could influence the composition and diversity of the insect community entering the tree after the wasp.

The objectives of this study were four-fold: i) to describe the timing of S. noctilio oviposition activity and to explore the likely chronology of entry into the tree if and when the woodwasp shares the tree with other phloem- and woodborers, ii) to describe how soon trees die after woodwasp attack, iii) to examine inter- and intra-stand variability in S. noctilio activity, and iv) to describe the host tree condition preferred by S. noctilio in North America.

2 Methods

2.1 Field sites Eight sites (Fig.2.1) were selected in south-western and central Ontario according to the following criteria: pure Pinus sylvestris overstorey forest composition; closed-canopy (based

on visual evidence of overlapping crowns within the main canopy); presence of S. noctilio based on survey trapping results; presence of trees with evidence of presumed S. noctilio attack (resin beads or drips) from the previous year (2006); and relatively unobstructed sight- lines, necessary for examination of the trees from the ground. Sites were selected to cover a 24 range in forest conditions based on tree height and tree decline (the proportion of standing dead trees). This was initially estimated visually and later quantified (Table 2.1). All sites had evidence that there were a number of other pests and pathogens present (Table 2.1).

Descriptions of each of the study sites can be found in Table 2.1. In one site (K1), shrubs were pruned to improve sight-lines.

Figure 2.1 Locations of eight Pinus sylvestris field study sites studies in southern and central Ontario used in all Chapter 2 studies.

25

Table 2.1. Stand attributes and descriptions for all Pinus sylvestris field sites in southern and central Ontario used for all Chapter 2 studies in 2007-08. Site Location & UTM General site observations Mean Basal Mean Mean Initial Understory General evidence of co-ordinates (zone age area dbh tree ht. stand pests and pathogen easting northing) (m 2/ha) (cm) (m) mortality (±SE) (±SE) (%) H14 Sauble Beach, ON Plantation 57 15.0 18.9 14.0 50 Small amount Scots pine Bark and sawyer (17T 479691 (7.5) (0.80) (0.64) regeneration beetle activity, 4944524) porcupine and sapsucker damage, storm damage H15 Elmwood, ON Plantation, in agricultural 38 20.8 18.4 13.1 34 Small amount ash and lilac Pine shoot beetle and (17T 502506 area (5.40) (0.51) (0.44) in understory, goldenrod sawyer beetle 4898199) and grasses abundant activity, gall rust H18 Orangeville, ON Plantation with many 21 13.6 13.4 7.6 24 Goldenrod and grasses Aphid and root collar (17T 574289 forked trees. (2.65) (0.52) (0.20) moderately abundant weevil activity, gall 4862585) rust and root rot, porcupine damage H7 Tottenham, ON Naturally regenerating 20 14.4 11.5 8.6 30 Small amounts of cedar in Pine shoot beetle and (17T 592582 stand. Densely stocked (0.88) (0.75) (0.39) understory aphid activity, tip 4873134) with small trees blight and gall rust, porcupine damage

K1 Bolton, ON (ca. Plantation in agricultural 34 29.2 24.0 17.7 35 Abundant buckthorn in Pine shoot beetle, 17T 604413 area. Many trees fallen by (6.69) (0.64) (0.35) understory, small amounts root rot 4859593) 2009 of elderberry and Manitoba maple E1 Cavan, ON Plantation in agricultural 42 32.7 16.3 13.6 24 Abundant buckthorn, some Pine shoot beetle and (17T 697478 area (4.08) (0.66) (0.49) cedar, Scots pine and other bark beetle 4895472) hardwoods activity, gall rust E6A Sandbanks PP, ON Plantation on sandy soil 52 10.0 10.3 6.7 24 Small amounts solomon Pine shoot beetle (ca. 18T 317667 (planted for sand dune (1.76) (0.65) (0.28) seal and poison ivy activity, tip blight 4863851) stabilization) E6B Sandbanks PP, ON Declining plantation on 59 36.0 15.0 10.0 62 Small amounts cedar in Extensive pine shoot (18T 317829 sandy soil (planted for (2.45) (0.77) (0.36) understorey beetle and pine spittle 4863693) sand dune stabilization). bug, root rot

26

2.2 Symptoms of S. noctilio oviposition activity The symptoms of S. noctilio oviposition activity I used were resinous beads (Fig. 2.2) or

drips. To test the validity of this substitute for S. noctilio activity, we felled five trees with

presumed S. noctilio -induced resin symptoms from each of three different locations, each

location adjacent to one of the field sites (15 trees total). Starting from a randomly selected

point on the stem of each tree, the first 10 resin symptoms encountered along a randomly

selected direction were preliminarily identified as being either S. noctilio -caused or non-S.

noctilio -caused symptoms. Each of the resin symptoms was then dissected to examine its

source. If drill scars, similar to that of a known S. noctilio drill site, were found below the

resin symptom then I concluded that the symptom was a result of the woodwasp. This test

showed 97 % accuracy over all three sites (150 resin symptoms in total); therefore this test

showed that a visual inspection of resin beads or drips was a good indicator of woodwasp

oviposition activity since only S. noctilio is known to oviposit in living trees, thus inducing a

defensive resin response (Spradbery 1973). Native woodwasps typically oviposit in dead

trees which are not able to produce a response.

Figure 2.2. Resin bead (approximately 5mm in width) on live Pinus sylvestris as a result of Sirex noctilio oviposition drilling. This symptom was used to assess for evidence of S. noctilio activity for all Chapter 2 studies in 2007-2008.

27

2.3 Field methods In each of the eight sites, a 100 m long X 10 m wide transect was marked with flagging tape.

Depending on the shape and size of the stand, the transect was either W-shaped or linear.

Within each transect, each presumed S. noctilio -vulnerable tree was marked with a uniquely

numbered aluminum tag that was attached to the tree with wire to avoid damage to the tree. I

expected that vulnerable trees were those that were suppressed or intermediate in dominance,

or those that had deteriorating crowns ( ≤ 25% of expected foliage) or stem injury (see

descriptions of tree characteristics below).

In total 231 trees, between 15 and 37 per site, were selected to monitor for insect

activity. Each tree was assessed for symptoms of previous insect activity ( S. noctilio resin

beads and drips , bark beetle entrance or exit holes or cerambycid oviposition pits). Residual

foliage was estimated to the nearest 5% for each tree by estimating the amount of foliage

present, compared to the amount of foliage normally expected for a tree having a full

complement of needles in the same position within the canopy. The numbered trees were

then scanned every two weeks from the first week of July to mid-September 2007 for

symptoms of insect activity. During the tree scan, the entire stem of the tree was visually

inspected in three overlapping, vertically oriented swaths from three equally spaced vantage

points around the tree using either a spotting scope mounted on a tripod or binoculars,

depending on tree height and the vantage point available for the scan. Each tree stem was

visually partitioned along its length into three equal sections, and the presence or absence of

S. noctilio or other insect evidence was recorded for each section on each bi-weekly visit. Per

cent residual foliage was re-assessed on each of these site visits. In late-August the number of

resin symptoms per one-third section of the tree was tallied (to a maximum of 5). This tally

was repeated for each tree on a subsequent visit between mid-September and late October

2007, after the hardwood trees and shrubs had dropped their leaves. This final assessment

28 was done to ensure that all symptoms were recorded because, despite my best efforts to select sites with good visibility, some sites (E1 and K1) had understory shrubs that could not be entirely removed. During the late August site visit, all non-numbered trees in each transect were also evaluated for symptoms of S. noctilio activity, in a similar manner to the marked

trees, in order to provide a site-level estimate of S. noctilio activity. After the 2007 field

season, these trees and sites were monitored for a further three visits: 1) in late April to early

May 2008 all numbered trees in each site were evaluated for evidence of spring bark beetle

attack and to re-evaluate residual foliage; 2) in mid-August 2008 I conducted a repeat tally of

S. noctilio symptoms per section on all numbered and non-numbered trees; and 3) in May

2009 I re-evaluated residual foliage of each numbered tree. Trees with no remaining green

foliage (i.e. 0% residual foliage) were deemed dead. Trees that had died and that had

evidence of S. noctilio activity (> 1-2 symptoms) were assumed to be killed by S. noctilio .

In the fall of 2007, every tree (i.e. both numbered and unnumbered trees) in each

transect was assessed. Diameter at breast height (dbh) and total tree height were measured.

Dominance class, crown closure and evidence of stem injury were recorded for each tree.

Dominance class was a rating of the canopy position of the tree as follows: dominant trees

had crowns extending above the main canopy, co-dominant tree crowns were at the level of

the main canopy, intermediate tree crowns reached the main canopy but were below it, and

suppressed trees were entirely overtopped by the main canopy. Canopy closure was the

number of quadrants of the tree crown touched or overlapped by neighbouring trees and

provided an estimate of sun exposure. Residual foliage was rated to the nearest 5%. For

numbered trees that had died over the summer of 2007, the per cent of residual foliage

estimated on the first site visit, in July 2007, was substituted. In three very densely treed sites

(H7, H18 and E6A), trees were sub-sampled and those within two metres of the centre line of

29 the transect were assessed; this was used to estimate the number of trees without S. noctilio

symptoms for site level summary statistics of woodwasp activity.

Overall stand characteristics were assessed for each site. Average stand tree height

and dbh were calculated using the first 50 trees encountered along the centre 4 m of the

transect. The proportion of these trees that were dead provided an estimate of initial stand

mortality (Table 2.1). Basal area was averaged from three factor 2 prism sweeps, one each at

the beginning, the middle and the end of the transect. Increment cores from three co-

dominant trees were averaged to estimate stand age. The presence of tree regeneration and

other plant cover was documented, as were evidence of pests and pathogens. For field sites

that had been used as Ontario Ministry of Natural Resources (MNR) 2006 Sirex noctilio

Survey sites, stand data, collected between Oct 2006 and Sept 2007, were obtained from the

MNR; for the remaining sites I collected these data in Sept 2007.

2.4 Data Analysis A G-test, calculated in Microsoft Excel 2003 was used to analyse the distribution of woodwasp symptoms. The G-test was based on the null hypothesis that the frequency of symptoms would be equally distributed. Correlations between the number of trees with wasp symptoms and the different site characteristics were tested with Spearman’s rank correlation using SYSTAT 12.0 (SYSTAT Software Inc, Chicago). The predictive value of tree characteristics for symptoms of woodwasp oviposition activity was examined with a logistic regression model using R 2.10.1 (R-Foundation, Vienna). Significant characteristics (p

<0.05) were included in a multiple model and a systematic backward elimination procedure was used to maximize model fit.

30

3 Results Of the 231 trees followed bi-weekly, 174 (75%) showed symptoms of S. noctilio oviposition

activity by the end of the 2007 season, between 10 and 31 trees per site. Symptoms of new

woodwasp activity were evident between mid-July and late August (Fig. 2.3a). In three of the

sites (E6B, H7 and K1) there was an early accumulation of trees with woodwasp symptoms

and by the end of July a large proportion of the total S. noctilio symptom-positive trees in

these sites were already detected. In five of the sites (all but E6A, E6B and H18), there were

few new trees with S. noctilio symptoms detected after mid-August. In two of the sites (E6A,

H18), there were several more new trees with woodwasp symptoms detected after this date.

Symptoms were not evenly distributed through the tree (G = 16.57, df = 2, p = 0.003); 69

lower compartments, 125 middle compartments and 102 upper compartments had signs of S. noctilio activity. In three of the sites (E6B, H7 and K1), activity was especially evident in the middle sections of the tree by the end of July (Fig.2.3c). Patterns were less clear in the lower and upper sections of the tree (Fig. 2.3b & d). Symptoms in upper sections of the tree continued to accumulate in site E6B and this curve had not levelled off by the end of the study period.

Of the 174 trees with evidence of S. noctilio activity in 2007, 31 of them subsequently died over the following 10 months. Those that died had a mean of 9.7 ±0.85 (SE) resin beads or drips per tree, those that did not die had an average of 3.0 ±0.27 (SE). In the trees that died, foliage became red, but had no evidence of needle wilting. In most sites, when trees died they did so over the following winter (Fig. 2.4). In one site, E6B, trees began dying within two weeks of initial attack and most of those that died after woodwasp attack died by eight weeks after the onset of wasp activity.

31

35 a) 30 H14 E1 25 E6A

noctilioS. 20 E6B 15 K1

symptoms H18 10 H7 5

No. trees with H15 0 2-Jul 16-Jul 30-Jul 13-Aug 27-Aug Final Week 35 b) H14 30 E1 25 E6A

noctilio S. 20 E6B

15 K1

symptoms H18 10 H7 5 No. trees with H15 0 2-Jul 16-Jul 30-Jul 13-Aug 27-Aug Final Week

35 c) 30 H14 E1 25 E6A noctilioS. 20 E6B 15 K1

symptoms H18 10 H7 5 No.trees with H15 0 2-Jul 16-Jul 30-Jul 13-Aug 27-Aug Final Week

35 d) 30 H14 E1 25 E6A S. noctilioS. 20 E6B K1 15 symptoms H18 10 H7 5

No. trees with H15 0 2-Jul 16-Jul 30-Jul 13-Aug 27-Aug Final Week

Figure 2.3. Accumulation curves of the number of Pinus sylvestris with symptoms of Sirex noctilio oviposition activity (resin beads or drips) per southern Ontario field study site during each site re- assessment over the 2007 season a) tree as a whole, b) lower one third of the tree stem, c) middle one third, d) upper one third. H14 n = 22 trees, E1 n = 26, E6A n = 37, E6B n = 34, K1 n = 27, H18 n = 33, H7 n = 37, H15 n = 15. 32

25

20 E1 & E6A

15 E6B H18 10 H7

5 Number of trees dead

0 0-2 weeks 4 wks 6 wks 8 wks 2-10 months Time after S. noctilio activity

Figure 2.4. Accumulation curve of the number of Pinus sylvestris dead after symptoms of Sirex noctilio oviposition activity (resin beads or drips), for each southern Ontario study site with presumed S. noctilio induced tree mortality. E1 n = 26 trees, E6A n = 37, E6B n = 34, H18 n = 33, H7 n = 37.

To allow more detailed exploration of the data, trees were categorized as attacked ( ≥5 resin symptoms in total) or tested (< 5 symptoms). This threshold for classification was based on the 95% confidence limits calculated for the mean number of symptoms per killed and surviving tree and equally dividing the difference of the gap between them. Of the 174 trees with S. noctilio activity, 52 of them had ≥ 5 resin signs and were deemed attacked and 122 of

them were considered tested.

Of the 231 numbered trees, several had evidence of previous insect activity: 88 had symptoms of previous S. noctilio activity, nine of bark beetle activity, six of longhorn beetle

activity and one of both bark and longhorn beetle activity. Of the 88 trees with symptoms of

previous woodwasp activity, 75 had symptoms of further S. noctilio activity in 2007. Of

those with bark or longhorn beetle symptoms, 14 of the 16 had subsequent symptoms of S.

noctilio activity; however, it was rarely in the same section of the tree (7 sections of a

possible 42).

33

Of the 174 trees with new S. noctilio resin symptoms in 2007, 30 had subsequent

symptoms of other insects, eight of bark beetles, seven of longhorn beetles and 15 had both.

In 14 of the trees, beetle symptoms were in one or more of the same compartments as the

wasp symptoms. Most of the trees (14 of 15) with both bark and longhorn beetle activity

were those attacked by the woodwasp ( ≥ 5 resin symptoms). Cerambycid beetle activity

occurred in the summer of 2007, following shortly after the initial S. noctilio activity. In one

site, E6B, Monochamus sp. were frequently seen mating on trees with S. noctilio symptoms within two weeks of woodwasp activity. Bark beetle activity was more commonly seen the following spring.

None of the stand characteristics showed a significant correlation with the proportion of

S. noctilio trees attacked, tested or both pooled (Table 2.2). There was a significant correlation between the proportion of trees with symptoms of S. noctilio activity (both types pooled) and the proportion of trees in the stand with crowns ≤ 25%; however this relationship was not significant when the site E6B was removed from the analysis.

Table 2.2. Spearman’s rank correlation coefficients for the association of the proportion of trees with symptoms of 2007 Sirex noctilio attack ( ≥ 5 Sirex noctilio resin beads), testing (< 5 resin beads) and the two pooled, with stand attributes of eight Pinus sylvestris field sites in southern Ontario. Significant findings (p<0.05) bolded.

% Trees attacked % Trees tested % Trees with S. noctilio symptoms (pooled)

Stand age (years) 0.16 0.21 0.57 Mean basal area (m 2/ha) 0.18 0.43 0.57 Mean height (m) -0.04 0.62 0.48 Mean dbh (cm) -0.06 0.57 0.41 Proportion trees with crowns ≤ 25% 0.56 0.10 0.71

There was considerable variability in S. noctilio symptoms and tree mortality among sites and between years (Table 2.3). In 2007, the proportion of trees attacked by the woodwasp was particularly high in one site (E6B) but the proportion of those tested was more similar

34 between sites. All of the trees that died in E6B by the spring of 2008 had symptoms of S.

noctilio activity, whereas in other sites tree mortality was not usually associated with the

wasp. In 2008 there tended to be less evidence of S. noctilio activity and mortality in most

sites than in 2007, with the exception of E6A and E6B, though the small number of trees

alive in E6B in the spring of 2008 made the proportional values quite high.

Table 2.3. Percentage of total trees attacked ( ≥ 5 Sirex noctilio resin beads or drips), tested (< 5 resin symptoms), presumed S. noctilio killed and total dead trees in each Pinus sylvestris study site in southern Ontario in 2007 and 2008.

% Trees % Trees % Tree % Tree killed by killed by mortality mortality % Trees % Trees % Trees % Trees S. S. (all (all attacked attacked tested tested noctilio noctilio causes) causes) Site 2007 2008 2007 2008 2007 2008 2007 2007 K1 10.6 0.0 38.3 27.7 0.0 0.0 0.0 2.1 H14 4.5 0.0 20.5 0.0 0.0 0.0 4.5 0.0 E1 1.7 0.0 23.3 8.8 1.7 0.0 5.0 1.8 E6A 4.5 13.1 20.9 10.3 1.8 8.4 2.7 11.2 E6B 58.3 60.0 20.8 15.0 58.3 50.0 58.3 50.0 H18 4.4 1.8 15.0 7.2 0.9 0.0 1.8 3.6 H7 8.9 2.7 11.8 5.4 4.9 1.1 8.4 1.6 H15 2.4 2.5 43.9 7.5 0.0 0.0 2.4 0.0

Individually, most tree characteristics, with the exception of crown closure, were

significant predictors of S. noctilio attack ( ≥ 5 symptoms) (Table 2.4). To aid in interpretation, the continuous variable residual foliage was categorized (> 25% or ≤ 25% residual foliage): trees having > 25% residual foliage were a fifth as likely to be attacked by the wasp as those with ≤ 25%. Dominance class also had predictive value; suppressed trees were 3.7 times as likely to be attacked as co-dominant trees, and intermediate trees 2.8 times as likely. Stem injury and presence of old S. noctilio attack were also significant predictors of

attack and both had relatively large effect sizes. In the multiple model, crown condition was

the only significant predictor of tree attack. Trees tested by S. noctilio were less well

predicted by the individual tree characteristics though canopy condition continued to be a

predictor. Trees with >25% residual foliage were half as likely to be tested as those with ≤ 35

25% foliage. Dominance class and stem injury were also of predictive significance for woodwasp testing.

36

Table 2.4. Logistic regression and multiple logistic regression results for tree characteristics as predictors of Sirex noctilio attack ( ≥ 5 resin beads or drips) and testing (< 5 resin symptoms) for 503 Pinus sylvestris from eight field sites in southern Ontario. Multiple model components reflect best fit model using backward elimination procedure from significant variables on individual regression. Estimate Std. error Z value p Odds ratio Conf. int. (2.5- 97.5%) Attack Test Attack Test Attack Test Attack Test Attack Test Attack Test Residual foliage (%) Intercept 0.55 -0.56 0.33 0.24 1.68 -2.30 0.09 0.02* 1.72 0.57 0.91- 0.35- 3.29 0.92 Residual -0.11 -0.02 0.02 0.01 -7.09 -2.33 <0.001*** 0.02* 0.90 0.98 0.87- 0.97- foliage 0.92 1.00 Crown category (> 25% Intercept -1.24 -0.73 0.16 0.14 -7.69 -5.12 <0.001*** <0.001*** 0.29 0.48 0.21- 0.36- or ≤ 25%) 0.39 0.63 Crown > 25% -1.57 -0.72 0.30 0.21 -5.18 -3.64 <0.001*** <0.001*** 0.21 0.48 0.11- 0.32- foliage 0.36 0.73 Dbh (cm) Intercept -1.07 -0.77 0.34 0.26 -3.17 -2.98 0.002** 0.003** 0.34 0.46 0.18- 0.28- 0.66 0.77 dbh -0.06 0.02 0.02 0.02 -2.49 -1.38 0.01* 0.17 0.95 0.98 0.90- 0.95- 0.99 1.01 Tree height (m) Intercept -1.06 -1.07 0.35 0.28 -3.00 -3.88 0.003** <0.001*** 0.35 0.34 0.17- 0.20- 0.69 0.59 Height -0.07 -0.003 0.31 0.22 -2.39 -0.13 0.02* 0.90 0.93 1.00 0.87- 0.95- 0.98 1.04 Dominance Intercept -2.17 -1.27 0.18 0.13 -11.83 -9.43 <0.001*** <0.001*** 0.11 0.28 0.08- 0.22- 0.16 0.36 Dominant -0.60 -0.27 0.62 0.39 -0.96 -0.70 0.34 0.48 0.55 0.76 0.13- 0.33- 1.61 1.57 Intermediate 0.82 0.66 0.35 0.29 2.34 2.30 0.02* 0.02* 2.8 1.94 1.12- 1.09- 4.47 3.38 Suppressed 1.31 0.74 0.35 0.31 3.74 2.36 <0.001*** 0.02* 3.70 2.09 1.83- 1.12- 7.29 3.82 Crown closure Intercept -1.94 -1.22 0.36 0.28 -5.39 -4.33 <0.001*** <0.001*** 0.14 0.29 0.07- 0.17- 0.28 0.50 Crown closure 0.02 0.04 0.12 0.09 0.17 0.46 0.86 0.65 1.02 1.04 0.81- 0.87- 1.30 1.26 Stem injury (+/-) Intercept -1.97 -1.16 0.14 0.11 -14.19 - <0.001*** <0.001*** 0.14 0.31 0.10- 0.25- 10.89 0.18 0.38 Stem injury 1.35 1.16 0.49 0.46 2.77 2.53 0.006** 0.01* 3.87 3.20 1.41- 1.28- 9.85 7.99

37

Table 2.4 continued Attack Test Attack Test Attack Test Attack Test Attack Test Attack Test History of S. noctilio Intercept -1.50 -0.30 0.22 0.17 -6.94 -1.75 <0.001*** 0.08 0.22 0.74 0.14 - 0.53 - attack (+/-) 0.33 1.03 Old S. noctilio 0.64 0.25 0.32 0.27 1.99 0.92 0.05* 0.36 1.89 1.28 1.01- 0.75 - attack 3.54 2.19 Multiple model Intercept 0.06 -0.98 0.49 0.29 0.12 -3.38 0.90 <0.001*** 1.06 0.37 0.40 - 0.21 - 2.82 0.66 Crown -0.08 -0.01 0.02 0.01 -4.58 -1.20 <0.001*** 0.23 0.92 0.99 0.89 - 0.97 - condition 0.95 1.01 Tree height 0.01 - 0.04 - 0.18 - 0.85 - 1.01 - 0.93 - - 1.08 Dominant - -0.23 - 0.40 - -0.58 - 0.56 - 0.79 - 0.34 - 1.66 Intermediate - 0.61 - 0.30 - 2.05 - 0.04* - 1.84 - 1.02- 3.28 Suppressed - 0.61 - 0.33 - 1.81 - 0.07 - 1.84 - 0.94 - 3.50 Injured stem 0.82 1.01 0.55 0.47 1.51 2.15 0.13 0.03* 2.28 2.75 0.75 - 1.08 - 6.64 7.00 History of 0.59 - 0.35 - 1.69 - 0.09 - 1.80 - 0.91 - - wasp 3.60 * p <0.05, ** p <0.01, *** p < 0.001

38

4 Discussion

As anticipated, Pinus sylvestris with declining crowns ( ≤ 25% residual crown), those that were suppressed or intermediate in dominance, and those with injured stems were more likely to show symptoms of S. noctilio activity and therefore were preferred by the wasp over

more vigorous trees. The pine shoot beetle, T. piniperda , also prefers pines with a similar crown condition (Morgan et al. 2004). Since there was evidence of this bark beetle at a number of the sites (Table 2.1), two insects may either compete for, or share the same host trees. Trees with evidence of previous woodwasp activity were also favoured for attack. The woodwasp is known to make single drills, perhaps when testing the tree for host suitability and when doing so introduces the wasp’s phytoxic mucus and its fungus to the tree (Coutts &

Dolezal 1969; Spradbery 1977). These test drills are thought to condition the tree and make it prone to further attack (Madden & Irvine 1971) and this fits with my findings.

Symptoms of new woodwasp activity were generally evident from mid-July until late

August. A number of secondary insects of pine reproduce through this time period, e.g. Ips

grandicollis (Eichhoff) (Ayres et al. 2001), Pissodes nemorensis Germar (Atkinson et al.

1988) and Monochamus carolinensis (Olivier) (Walsh & Linit 1985; Yanega 1996), and

could be expected to enter the tree around the time of, or shortly after the woodwasp. This

was evident in one site (E6B) where Monochamus sp. was regularly seen mating on trees

with fresh symptoms of S. noctilio activity. In some sites (E6B, H7, and K1), woodwasp

activity tended to occur earlier in the season than in other sites; notably, these were also the

sites with the largest proportion of trees classified as attacked by the wasp. These temporal

differences may influence interactions with specific beetle co-habitants if they do not have

the same temporal shift in these sites. These sites were geographically dispersed, so this is

not likely a climatic effect.

39

Symptoms of woodwasp activity were rare in sections of trees with pre-existing symptoms of bark, or longhorn beetles. This could be evidence that the wasp avoids ovipositing in trees colonized by beetles and their associated pathogens as Hanson (1939) and Spradbery and Kirk (1978) observed. However there were few trees meeting our selection criteria that had beetle symptoms so this pattern may be an artefact of a small sample size.

In the majority of the sites, most trees that died after S. noctilio attack did so during the winter following attack. In the site (E6B) with the highest S. noctilio -related tree

mortality, trees died soon after attack, often within a few weeks. This range in the timing of

mortality is similar to that described by Madden (1988) who attributes this variability to the

effect of location and site characteristics on tree resistance. The timing of tree death may

influence the community of phloem- and woodborers entering the tree while the S. noctilio

larvae are developing. All of these insects have host condition preferences, some prefer

stressed and dying trees, others favour newly-dead trees (see USDA 1985 for examples), and

once the tree has been dead for a time more secondary insect species colonize it (Dajoz

2000). Therefore, unless the woodwasp’s developmental time is markedly accelerated by

conditions in the dead tree, it is likely to share the tree with a greater diversity of phloem-

and woodboring species during its early larval instars when the tree dies rapidly than when it

dies more slowly. If interactions between S. noctilio and these co-habiting beetles do occur,

this could have considerable impact on the wasp.

There was substantial variability in the number of symptoms of S. noctilio activity

between sites and between years. Field sites covered a range of stand and site conditions

(Table 2.1) and this is expected to have contributed to the variability between sites. Site E6B,

which was particularly impacted by the woodwasp, was planted to stabilize a sand dune so

would be expected to experience drought stress at times, due to its well-drained substrate.

40

Poor quality sites on sandy soil and drought stress are associated with greater woodwasp activity and tree mortality (Taylor 1981; Neumann et al. 1987). In addition, this site had evidence of extensive activity by other pests and pathogens (Table 2.1) and so individual trees may have been more stressed and this could have contributed to tree mortality in this site as well.

In most sites, S. noctilio activity decreased in 2008. There was considerably more precipitation in the summer of 2008 as compared to the previous year (Environment Canada

2010) and this may have affected the wasp’s activity levels, as the wasp has been observed to refrain from ovipositing on rainy days (Dolezal 1967). It is surprising then that there were more trees with symptoms of S. noctilio attack in site E6A in 2008 than 2007. However, this stand was planted to stabilize a sand dune as well, so site factors may have superseded the effects of weather in this site. There is both spatial and temporal variability in S. noctilio activity and tree mortality; therefore, there may also be variation in the effect of bark and longhorn beetle community between sites if such factors as the timing of tree death are important in determining the community of these insects occupying the tree.

This is the first description of S. noctilio’s oviposition activity period, preferred host condition, and onset of tree mortality after attack in North America and provides essential background to aspects of the host colonization ecology of the woodwasp S. noctilio in its

introduced range in eastern North America. These descriptions provide the foundation for

further work on interactions between the extant phloem- and woodborers and the woodwasp

investigated in Chapters 3, 4 and 5.

41

Chapter 3

Effect of two bark beetle-vectored fungi on the on-host search and oviposition behaviour of Sirex noctilio on Pinus sylvestris trees and logs

Abstract The woodwasp Sirex noctilio’s obligate fungal symbiont, Amylostereum areolatum , is

required for its offspring’s development. The symbiont is a weak competitor with bark

beetle-vectored ophiostomatoid fungi so it would be beneficial to the wasp if it could detect

and avoid ovipositing in substrate colonized by these competing fungi. I investigated the

response of the woodwasp to the presence of two bark beetle-associated fungi,

Leptographium wingfieldii and Ophiostoma minus inoculated into living trees, and to L.

wingfieldii and the wasps’s own symbiont inoculated into cut logs. The wasp consistently avoided areas inoculated with L. wingfieldii and there were fewer signs of oviposition activity (drill scars) in these zones. There was no significant response of the woodwasp to O. minus or A. areolatum . Female woodwasps can detect the presence of some species of beetle- vectored fungi and make choices about oviposition sites that benefit the survival and fitness of their offspring.

1 Introduction Sirex noctilio is a woodboring wasp native to Eurasia and northern Africa and was recently

discovered in eastern North America (Hoebeke et al. 2005; de Groot et al 2006). In the

southern hemisphere, where this insect is an introduced invasive pest, it is capable of causing

extensive economic loss and ecological impact, yet in its native range it is of little ecologic or

42 economic concern (Hall 1968). The wasp’s ability to kill its host tree is a result of its complex biology. It has an obligate, species-specific symbiotic relationship with a basidiomycete fungus Amylostereum areolatum (Gaut 1969). Interactions among the two organisms and a phytotoxin produced by the wasp allow the insect to overcome the defences of living trees, killing them (Coutts 1969b).

The wasp favours Pinus spp. that are suppressed, physiologically stressed or injured

(Neumann et al. 1987; Chapter 2). After arriving at a pine, the adult female woodwasp walks over the bark of the main stem of the tree, taps the bark with her antennae, and when a potentially suitable site is found she begins drilling into the tree with her ovipositor (Francke-

Grosman 1939; Coutts & Dolezal 1969). The female may reject the site before the ovipositor reaches the cambium (Madden 1968). When the site is suitable, the female wasp oviposits into the sapwood of the host tree laying up to three eggs per drill site, each in separate tunnels, and a final eggless tunnel at each drill site is inoculated with arthrospores or fragments of A. areolatum , as well as the phytotoxic mucus produced by the wasp (Coutts &

Dolezal 1969; Madden 1974).

In Europe, S. noctilio shares the tree with curculionids, cerambycids, and buprestids that also favour and colonize weakened and stressed trees (Wermelinger et al. 2008). In the wasp’s native range, potential co-habiting species include Ernobius mollis (L.), Phaenops cyanea F., Acanthocinus aedilis L., Pissodes piniphilus Hrbst., Ips acuminatus (Gyll.),

Orthotomicus longicollis (Gyll.), Tomicus minor (Hartig) and T. piniperda (Schroeder 1987;

Wermelinger et al. 2008). In North America there is also a rich and diverse community of phloem- and woodboring insects that also colonize pine (USDA 1985; Morgan et al. 2004), the most frequent being Tomicus piniperda , Pissodes nemorensis , Ips grandicollis and

Monochamus carolinensis (Chapter 4). With the recent discovery of S. noctilio in North

America, the nature and impact of the local phloem- and woodboring insect community on

43 the survival and population dynamics of S. noctilio are unknown. It is important to

investigate this interaction as it could have significant implications for pest management, in

fact, it may determine whether S. noctilio will become a pest at all. This interaction could

occur directly between the insects, indirectly between the fungi these insects carry, or both.

Both S. noctilio and many of its phloem- and woodboring-co-habitants have

ecologically important associations with fungi. In the case of S. noctilio , the association is

obligate and development of A. areolatum is essential for woodwasp egg eclosion (Madden

1981). If fungal growth is impeded, larvae may starve (Coutts & Dolezal 1965; King 1966),

but when conditions are optimal for A. areolatum , larger adults are produced (Madden 1981).

Curculionids and cerambycids, or their associated phoretic mites, have well-documented

associations with various species of fungi, primarily ophiostomatoid (blue stain) species (e.g.

Whitney 1982; Wingfield 1987; Nevill & Alexander 1992; Kirisits 2004; Hausner et al.

2005; Jankowiak 2006; Ben Jamaa et al. 2007; Jankowiak and Rossa 2007; Jankowiak and

Kola řik 2010). The most frequent beetles that co-habit with S. noctilio have described associations with fungi. Examples include Leptographium wingfieldii Morelet , L. procerum

(Kendr.) Wingf., L. lundbergii Lag. & Melin sensu Jacobs and Winfield, Ophiostoma minus

(Hedgcock), O. ips (Rumb.) Nannf. and Sphaeropsis sapinea (Fr.) Dyko & Sutton (T.

piniperda ), O. ips ( I. grandicollis ) and L. procerum Wingf. ( P. nemorensis ). There is a range

in the degree of closeness of beetle relationships with fungi; in some species nearly all

individuals carry fungi and in other species, such as those listed above, only some individuals

are associated with it (Kirisits 2004). When fungi are present, there is often a benefit to the

fitness of the beetle (e.g. Six & Paine 1998).

Interactions between insect-vectored fungal species could influence the reproductive

fitness or distribution of an insect, and more so if the species is dependent on its fungal

associate for development. For example, southern pine beetles, Dendroctonus frontalis

44 without their fungal mutualists are less fecund than those with it (Barras 1973; Goldhammer et al. 1990) and have a lower ratio of population growth (Bridges 1983). Because the blue stain Ophiostoma piliferum outcompetes the beetle’s two mutualistic fungi it has been investigated as a potential biocontrol agent for the beetle (Klepzig 1998). The woodwasp, is highly dependent on its fungal symbiont therefore could be even more profoundly affected by competing fungal species.

Amylostereum areolatum is a weak competitor with some ophiostomatoid fungi

(Chapter 5; King 1966). This may be one reason that the woodwasp is not an economically important pest in Europe where subcortical beetles would commonly share the same host.

Given the poor competitive abilities of its fungal symbiont, it would be of reproductive benefit to the wasp if it could detect the presence of competing fungi in the tree and avoid them. The avoidance of oviposition in substrate colonized by harmful fungi has been recently documented in other insects systems (Lam et al. 2010). Hanson (1939) and Spradbery and

Kirk (1978) speculate that S. noctilio detects and avoids the beetle vectored fungi but this has never been investigated.

In this study, I aimed to address two hypotheses: i) S. noctilio avoids ovipositing in wood already colonized by ophiostomatoid fungi, and ii) the development of S. noctilio offspring will be adversely affected by the presence of ophiostomatoid fungi. I tested these hypotheses by conducting two experiments. In the first experiment I inoculated living trees with two ophiostomatoid fungi, O. minus and L. wingfieldii and caged female woodwasps on them to evaluate their behaviour, oviposition drilling patterns and larval development. In experiment two, L. wingfieldii and A. areolatum were tested in a similar fashion on cut sections of wood, to further assess wasp behaviour and offspring development in the presence of these fungi.

45

2 Methods

2.1 Fungi All fungal cultures were obtained from the Canadian Forest Service culture collection

(Accession numbers: L. wingfieldii 025 7010, O. minus 075 7007; isolate details can be found in Table 5.1) and were inoculated onto, and re-isolated from sterile Pinus sylvestris wood chips to counteract effects of storage prior to production for this experiment. Each of these fungal species was grown on potato dextrose agar (PDA) for use in the field.

2.2 Insects Sirex noctilio were reared from infested pines ( P. sylvestris or P. resinosa ) that were obtained from eight pine plantations located in south-western and central Ontario. Trees were felled in the spring or early summer, prior to woodwasp emergence, and stored in either screened tents or enclosed tubes. Wasps were collected from the rearing containers shortly after emergence

(several times per day) before mating could occur. They were stored up to two weeks at 5-

8°C.

2.3 Field sites Four P. sylvestris plantations near Barrie, Ontario were selected for the field study. Sites with an abundance of trees in the condition preferred by the woodwasp were selected, i.e., those with declining crowns (between 10 and 25% residual foliage) and those with many trees in the suppressed (crown completely below that of the general canopy) or intermediate (crown extended to general canopy but shorter than the more dominant trees) crown dominance class

(Chapter 2). One of these plantations was the source of logs used in experiment two.

2.4 Experiment 1: Effect of L. wingfieldii and O. minus on S. noctilio behaviour in live trees A preliminary trial of S. noctilio oviposition behaviour was conducted using 14 female wasps

caged onto stems of S. noctilio -favourable P. sylvestris in one of the selected field sites in

46

2008 . Half of the wasps were placed on an untreated tree, the other half on a tree that had been inoculated three weeks earlier with the two fungal pathogens L. wingfieldii and O. minus (treatment and caging are described later in this section). This trial was conducted to test the experimental protocols and particularly to develop categories of insect oviposition behaviour. The female wasps exhibited three types of behaviour: 1) tapping their antennae over the bark while walking, hereafter referred to as searching; 2) insertion of the ovipositor a short way into the stem (< 25% of the ovipositor length) for a short duration (< 60 seconds), termed “probing”; and 3) insertion of >25% (typically 75-100%) of the ovipositor into the tree and for >60 seconds (usually much longer), referred to as drilling. An examination of the drills revealed two distinct patterns: shallow drill scars extending into the phloem only (likely corresponding to probing behaviour), and deep drill scars into the sapwood (likely corresponding to drilling behaviour) (Fig.3.1a-d).

a b

c d

Figure 3.1. Sirex noctilio drill scars inDeep Pinus sylvestris drills: a) shallow drill scar in phloem b) scar not extending into sapwood c) deep drill scar d) deep drill scar extending into sapwood.

47

The wasps exhibited similar behaviour regardless of the tree treatment and these behaviours were comparable to those I observed in wild populations. These behaviour types were used to evaluate wasp activity in this and the following experiment.

An in-vivo study was conducted in 2008 and 2009 to test the ability of S. noctilio to

detect and alter its searching and oviposition drilling patterns in the presence of certain

ophiostomatoid fungi. Within each of the four experimental sites, P. sylvestris were selected that had between 10 and 25% residual crown foliage, diameter at breast height (dbh) <20cm, and were preferably suppressed or intermediate in dominance. In some cases, there were an insufficient number of trees meeting the selection criteria available in a site, so trees with a slightly greater proportion of residual foliage (30-35%) were selected and the crowns were pruned to 20-25% residual crown. The crown condition, dbh and dominance of each tree was recorded. In 2008, 48 trees (12 from each of the four sites) were selected and in 2009, 36 trees were selected (12 each from three of the original sites). To further stress each of the study trees, in mid-June of each year, a 30-cm wide section of bark and cambium was removed from around the entire circumference of each tree, the lower end extending to breast height. A 1-cm wide strip of outer sapwood was excavated to a depth of 5-10mm in the centre of the girdled area. Mechanical girdling makes trees more attractive to the woodwasp for oviposition (e.g. Spradbery & Kirk 1981).

Fungal inoculation took place in early July in both 2008 and 2009, two weeks after the trees were girdled. Twenty four trees were selected for treatment in each year; there were

24 untreated trees in 2008 and 12 in 2009. Trees were selected for treatment within each site by ranking the trees hierarchically based on crown condition followed by dominance and dbh. Every other tree was selected for inoculation in 2008 and two of three in 2009. A 90-cm length of stem was marked, at least 30 cm above the girdled area. This section was divided

48 into thirds lengthwise starting from the north. Sections were randomly selected for inoculation with the two test fungi ( L. wingfieldii or O. minus ) and the third section was left

as an untreated control. Within each section, bark plugs were cut to the phloem-sapwood

interface with a sterile # 4 cork borer and the resulting cavities inoculated, mycelium side in,

with a # 3 cork borer plug of fungus. Fungal plugs were taken from the colonized surface

layer of the growing edge of the colony. All bark plugs were re-inserted after inoculation.

This procedure was repeated every 1.5 cm in a circumferential pattern around the stem in

rings 10 cm apart. A 2 cm buffer zone between treatments was maintained. In the control

section, the holes were bored in a similar manner but not inoculated with fungus. Trees were

caged immediately after treatment (Fig. 3.2) in order to prevent colonization by other insects.

Figure 3.2. Sirex noctilio cage used in Experiment 1 in situ on Pinus sylvestris .

49

Cages consisted of nylon screen held above the bark immediately above and below the treated section of the stem with sections of cylindrical polyethylene foam approximately 7 cm in diameter (i.e. pool noodles); one side was bevelled so that the foam could be snugly attached to the tree stem. The screen was held in place at each end with wire and the open edges were overlapped, rolled and secured with metal clips. Similar sections of the untreated trees were caged in an identical manner.

Between late July and mid-August of each year, female S. noctilio were added to the

cages. Two active but host naïve females were placed in each cage and monitored for a 20-

minute period. Each of the two insect’s movements over the stem was mapped; the number

of search visits, probes and drills were tallied for each treatment section. Insects were

monitored in an identical fashion on each of the two subsequent days and removed from the

cages on the fourth day (i.e. over a 72h period). Woodwasps that died, or were injured or

compromised, were replaced.

In the fall of each year, all study trees were felled and the treated sections were

removed. Samples were taken from the outer 1 cm of sapwood adjacent to the treated area for

gravimetric wood moisture measurement. The length of the response zone (resin-infiltrated

area of active host response to injury) and the extent of fungal growth (when present) in the

sapwood above and below each inoculation point were measured in each of the treated and

control zones, and then averaged per tree. Four fungal cultures from each of the fungus-

treated sections were taken from each tree and cultured on PDA to re-isolate the inoculated

species. Woodwasp drill scars (both shallow and deep) were tabulated per treatment section.

A subset of drill sites was dissected for signs of eggs or larvae; however, none were found so

this outcome measure was abandoned.

50

In 2008, one of the treated trees could not be safely felled and one of the untreated trees in 2008 was killed by girdling resulting in excessive growth of adventitious blue stain fungi

(those not purposely inoculated): both trees were removed from the analysis.

2.5 Experiment 2: Effect of L. wingfieldii and A. areolatum on S. noctilio behaviour on bolts A second study was conducted in an outdoor insectary using cut sections of pine (bolts). By using cut bolts we sought to extensively evaluate the woodwasp behavioural responses to fungi without the effects of weather and tree health. Second, we sought to assess the effect of ophiostomatoid fungi on woodwasp development that we were not able to test in the previous study. We hypothesized that despite our efforts to create a weak and suitable host tree for S. noctilio , there may have still been sufficient residual defensive response of the tree to prevent woodwasp development and thus cut bolts would reduce this problem. In this experiment, L. wingfieldii was selected for further testing since the woodwasp demonstrated a response to this pathogen in Experiment 1, and the wasp’s own symbiont, A. areolatum , was tested as a negative control.

In 2009, 30 cm lengths were cut from freshly felled P. sylvestris and bolt ends were immediately sealed with paraffin to prevent desiccation and introduction of adventitious fungi. The treatments consisted of inoculating one side of each bolt, into the phloem- sapwood interface, with L. wingfieldii or with A. areolatum ; the other half of the bolt was left untreated. Inoculation techniques were the same as described for the previous experiment.

The third and final treatment consisted of control bolts in which one side only had holes bored (no fungal inoculations) and the other side was left untreated. All bolts were left in the insectary for two weeks before the insects were introduced in the cages.

51

Each bolt was placed upright in a small plywood and screen cage. Two inexperienced

host naïve female S. noctilio were placed on the transition zone of each log, i.e. on the border between the treated and control sections, and was observed for 10 minutes. The number of visits, probes and drill attempts (as defined above) per section were recorded. The insects were kept caged with the bolt for two days; observations were repeated on day two. There were 30 replicate bolts for each treatment and the control (a total of 90 bolts).

Growth of adventitious fungi became a problem in this experiment despite all efforts to prevent it. In addition, wasp activity was low on many of the bolts and upon peeling them drill scars were absent, so there were several zeros in this data set. These issues thwarted the original research objectives, so data collection and analysis methods were modified and used to augment those from Experiment 1. I rated the amount of adventitious blue stain fungal growth on a scale of 0 to 3, with a rating of 0 indicating no contamination; 1 corresponding to trace amounts (< 1 % of the area colonized); 2 indicating 1 – 50% of the area was colonized by these fungi, and a rating of 3 corresponding to > 50% colonization. I converted the three response variables, searching, drilling and drill scars to binary variables for subsequent statistical analyses. Fungi were cultured from the originally treated zones.

2.6 Data Analysis In Experiment 1, the number of drill scars was compared between sections of the treated trees using ANOVA and Tukey’s Honestly-Significant-Difference test. To meet the assumptions of normality and homoscedasticity, drill scar data were transformed with log n +1 for all

ANOVA tests. A G-test was used to compare the number of drill scars between treated and untreated trees and to compare the occurrence of behaviours between treatment sections. The

G-test was based on the null hypothesis that the frequency of each event would be equally distributed between treated and untreated trees or between treatment sections. Experiment 1 analyses were conducted in SYSTAT 12.0 (SYSTAT Software Inc, Chicago) (t-test, 52

ANOVA), or calculated in Microsoft Excel (2003) (G-test). Insect response data from

Experiment 2 were tested with a logistic regression model using R 2.10.1 (R-Foundation,

Vienna).

3 Results

3.1 Experiment 1 The number of deep drill scars made by S. noctilio females differed significantly among

treated sections in the treated trees in both years (2008, F 2,66 = 7.73, p = 0.001; 2009, F 2,69 =

22.94, p < 0.001). There were fewer deep drill scars in the sections treated with L. wingfieldii

than in sections treated with O. minus or in those left as untreated controls (Fig.3.3). There

was no effect of the aspect of the treated section (i.e. northwest, northeast) on the number of

deep drill scars (2008: F 2,66 = 0.04, p = 0.96; 2009: F 2,69 = 0.96, p = 0.39). There were fewer

deep drill scars found in untreated trees than in treated ones (2008: G = 204.71, p < 0.001;

2009: G = 394.65, p < 0.001).

30 A 2008 25 2009 20 a A 15 a

10 b B 5

Mean drills per section (+SE) 0

ControlControl L. wingfieldii O. minus

Treatment

Figure 3.3. Mean number of deep Sirex noctilio drill scars per Leptographium wingfieldii , Ophiostoma minus and control section of living Pinus sylvestris trees in 2008 and 2009. Results of Tukey’s Honestly- Significant-Difference test: lower case letters indicate significant differences between treatment groups in 2008 and upper case show the same in 2009. 53

When shallow drill scars were analyzed, a pattern similar to the deep drill scars was apparent. That is, there was a significant difference in the number of shallow drill scars within treated trees in both years (2008: F 2,66 = 3.35, p = 0.04; 2009: F 2,69 = 11.69, p <

0.001). There were fewer drill scars in L. wingfieldii treated sections than either of the other

treatment sections in 2009; in 2008 there were fewer scars in the L. wingfieldii treated section

than the control (Fig. 3.4). There was no effect of the aspect of the section on shallow drill

scars (2008: F 2,66 = 0.59, p = 0.56; 2009: F 2,69 = 0.24, p = 0.79). There were fewer shallow

drill scars seen in the untreated than the treated trees (2008: G = 15.99, p < 0.001; 2009: G =

177.73, < 0.001).

30 A 25 2008 2009 20 a 15 A ab

10 b 5 B

Mean drills per section (+SE) 0

Control L. wingfieldii O. minus

Treatment

Figure 3.4. Mean number of shallow Sirex noctilio drill scars per Leptographium wingfieldii , Ophiostoma minus and control section on living Pinus sylvestris trees in 2008 and 2009. Results of Tukey’s Honestly- Significant-Difference test: lower case letters indicate significant differences between treatment groups in 2008 and upper case show the same in 2009.

In the treated trees, there were no differences in number of searching visits (pooled by tree) per treatment section in either year (2008: G = 1.86, p = 0.39; 2009: G = 3.32, p = 0.20)

(Fig.3.5a-b). There were more quadrants searched in treated (all treatment sections pooled)

54 than in the untreated trees in both years (2008: G = 32.31, p < 0.001; 2009: G = 32.66, p <

0.001).

Treatment had a significant effect on probing activity in 2008 (G = 11.72, p = 0.003) but not in 2009 (G = 2.99, p = 0.22), though there were few probing events witnessed in either year (Fig.3.5a-b). The number of probes was less in treated than untreated trees in

2008 (G = 11.57, p < 0.001) and more in 2009 (G = 12.51, p < 0.001).

Search a) Search p = 0.39 Probe 8 Probe p = 0.003 7 Drill p = 0.16 Drill 6 5 4 3

or drill (+SE) 2 1

Mean no. search, probe 0 ControlControl L. wingfieldii O. minus Treatment

Search Probe 8 b) Search p = 0.20 7 Probe p = 0.22 Drill 6 Drill p = 0.007 5 4 3 drill (+SE) 2 1 0 Mean no. search, probe or ControlControl L. wingfieldii O. minus Treatment

Figure 3.5. Number of quadrants searched, probed or drilled (+SE) by Sirex noctilio per Leptographium wingfieldii , Ophiostoma minus and control section of living Pinus sylvestris trees in a) 2008 and b) 2009. Significance results of G-test embedded.

55

Drilling activity was not affected by treatment in 2008 (G = 3.71, p = 0.16) but was in

2009 (G = 9.83, p = 0.007) (Fig.3.5a-b). There was more drilling of treated than untreated trees in both years (2008: G = 18.62, p < 0.001; 2009 G: = 77.28, p < 0.001).

Mean growth of L. wingfieldii was 6.0 mm in 2008 and 10.0 mm in 2009 and that of

O. minus was 10.4 mm and 20.7 mm respectively. The inoculated fungal species were re- isolated from most of the treated sections in 2008, but L. wingfieldii was rarely re-isolated in

2009. In about half the trees O. minus was re-isolated in 2009. In the fungus treated sections,

reaction zones were longest in the O. minus treatment, 43.0 mm in 2008 and 55.3 mm in

2009, compared to 26.0 mm (2008) and 24.0 mm (2009) in the L. wingfieldii treated zones.

This compares to 18.2 mm in 2008 and 15.8 mm in 2009 in the control sections. There was

no significant difference in wood moisture between the treated and untreated trees in either

year (2008: t = 1.43, df = 44, p = 0.16; 2009: t = 0.86, df = 34, p = 0.39; sqrt transformed),

although it tended to be higher in the untreated trees ones (2008: untreated 95%, treated 85%;

2009: untreated 51%, treated 47%).

3.2 Experiment 2 The presence of L. wingfieldii , treatment (vs. control), and the degree of adventitious fungal colonization all affected the probability of the presence of drill scars in the logistic regression model (Table 3.1); all three variables were associated with the absence of drill scars. The size of the treatment effect was greatest for L. wingfieldii ; bolts treated with this

fungus were one-fifth as likely to have drill scars as those having the other treatments.

Treatment sections and bolts with adventitious fungi were about half as likely to have them

as control sections and bolts without adventitious fungi.

Searching (search or no search) was not predicted by any of the variables, but the

presence of adventitious fungi approached significance (Table 3.1).

56

The presence of L. wingfieldii and adventitious fungi both had similar negative relationship and magnitude of effect on drilling activity as it did on drill scars (Table 3.1). In contrast to the drill scar results, treatment had a positive relationship with drilling activity, which was twice as likely to occur on treated sections of the bolt as on untreated ones (Table

3.1).

Table 3.1. Logistic regression results for the effect of inoculated fungus ( Amylostereum areolatum , Leptographium wingfieldii or control), treatment (vs. control), and the degree of adventitious fungal growth on Sirex noctilio searching and drilling behaviours and drill scars in Experiment 2. Significant results bolded. Behaviour Model term Estimate Std. error z value p Odds ratio Conf. interval Search Intercept -0.72 0.47 -1.53 0.13 0.49 0.19-1.21 Fungus A. areolatum 0.45 0.38 1.17 0.24 1.56 0.74-3.33 Fungus L. wingfieldii 0.03 0.39 0.07 0.94 1.03 0.48-2.21 Treated 0.19 0.31 0.61 0.54 1.21 0.66-2.21 Adventitious fungus 0.41 0.21 0.94 0.05 1.50 1.00-2.29 Drill Intercept -0.36 0.53 -0.67 0.50 0.70 0.24-1.98 Fungus A. areolatum 0.36 0.43 0.85 0.40 1.44 0.62-3.38 Fungus L. wingfieldii -1.42 0.53 -2.67 0.008 0.24 0.08-0.66 Treated 0.77 0.38 2.01 0.04 2.16 1.03-4.68 Adventitious fungus -0.66 0.26 -2.60 0.009 0.52 0.31-0.84 Drill scar Intercept 1.28 0.52 2.47 0.01 3.61 1.33-10.34 Fungus A. areolatum -0.11 0.40 -0.26 0.79 0.90 0.41-1.98 Fungus L. wingfieldii -1.64 0.47 -3.50 <0.001 019 0.07-0.47 Treated -0.87 0.35 -2.50 0.01 0.42 0.21-0.82 Adventitious fungus -0.72 0.24 -3.00 0.003 0.49 0.30-0.77

Inoculum growth varied between the two fungal species. There was minimal growth of the A. areolatum inoculum, on average 2.1 mm per inoculation point. Mean L. wingfieldii growth was 83.0 mm. Amylostereum areolatum was never re-isolated from the bolts and L. wingfieldii was only occasionally re-isolated, the cultures typically did not grow and a few were contaminated with other blue stain species. On the 0-3 rating scale, mean adventitious blue stain rating was 2.0 in the A. areolatum treated bolts, 1.7 in control bolts and 1.1 in the

L. wingfieldii treated ones.

57

4 Discussion Experiments 1 and 2 showed similar results; there were fewer S. noctilio drill scars in sections of trees or bolts inoculated with L. wingfieldii than other sections. Probing and

drilling activity, though not always significant, followed similar patterns. These findings

corroborate Hanson (1939) and Spradbery & Kirk’s (1978) hypotheses that the woodwasp

detects and avoids trees colonized by ophiostomatoid fungi. This is of considerable

reproductive benefit to the wasp as its symbiont is a poor competitor with this particular

species of fungus (Chapter 5) and thus if the female selected these areas for oviposition it

could be expected to inhibit the wasp offspring’s development.

The lack of effect of O. minus on the number of wasp drill scars in Experiment 1 in

contrast to its avoidance of L. wingfieldii was surprising especially since there was a lack of

wasp drilling associated with the presence of adventitious fungi in Experiment 2. Similar to

L. wingfieldii , O. minus outcompetes the wasp’s symbiont A. areolatum (Chapter 5). The wasp is likely to have interacted with both fungal species in its native range in Eurasia since they are both regular associates of T. piniperda there (reviewed in Kirisits 2004), and so have had the opportunity to evolve the ability to detect both species. One explanation is that there may have been volatiles coming from the extensive defensive reaction zones in the sapwood in the O. minus inoculated sections that overwhelmed volatiles emitted by O. minus itself, making it difficult for the wasp to detect the volatiles from the fungus. There are a number of volatile metabolites associated with ophiostomatoid fungi (reviewed in Hanssen 1993) and they differ from those associated with pine response to injury (Cheniclet 1987). Wound response associated compounds include alpha-pinene and 3-carene (Cheniclet 1987;

Manninen et al. 2002) and both are known attractants to S. noctilio (Simpson 1976) so their effect may have interfered with volatile compounds produced by the fungus. In pines inoculated with fungal pathogens, associated with the wound accumulate in

58 especially large amounts and are persistent (Cheniclet 1987) so these terpenes could be expected to affect the wasp’s oviposition behaviour at the time of the experiment. Since

Experiment 2 was conducted on cut bolts and there was no associated resinous reaction, the adventitious fungi would be expected to be more readily detected by the wasps than the O. minus in Experiment 1.

The preference for treated trees over untreated trees for searching, probing and drilling, and the corresponding patterns in drill scars, was not expected. The untreated trees were included in the experiment to allow comparison to woodwasp activity in the absence of any fungi. There was no discernable difference in characteristics between the two sets of trees so it is not likely that the tree attributes influenced this outcome. It is more likely that attractive volatiles from the reaction zones in the treated trees (described previously) stimulated an increase in wasp searching and drilling activity.

Fungal growth was more limited than expected in Experiment 1. The density of inoculations was expected to overwhelm the tree’s defenses and result in growth of the inoculated fungi from most of the inoculation points. Trees are resistant to some fungi during favourable environmental conditions and this is true for many of the ophiostomatoid species

(Schoenweiss 1981; Smalley et al. 1993). Weather conditions during the two study periods were cooler than normal and precipitation did not appear to differ substantially from climate normals (Environment Canada 2010) so environmental stress should have been relatively low during the experiment and this would enhance the tree’s ability to resist fungal colonization.

The extensive reaction zones in the fungus-inoculated treatment sections suggest that the trees were able to respond to, and contain, the pathogen at most of the inoculation sites. Re- isolation of these fungi in 2008 suggests that they were present in a quiescent state and therefore could have been detectable by the wasps. Conifers are known to sequester viable fungal inoculum within reaction zones (Raffa & Smalley 1988). Although fungi were rarely

59 re-isolated in either experiment in 2009, overall results (reaction zones, insect behaviour and drilling) were similar to 2008 when the inoculated fungi were re-isolated so it is expected that the lack of re-isolation was not due to the absence of the fungi

This study is the first to demonstrate that S. noctilio avoids areas of trees colonized by

at least one species of bark beetle-vectored fungus and clearly shows that interactions

between S. noctilio and bark beetles could occur via interaction with fungal associates of beetles. These interactions could influence the selection of oviposition sites by the wasp and limit its colonization in areas of the tree previously colonized by beetles carrying these fungi.

Competition with beetles for un-colonized host resources could be a factor in helping to limit

S. noctilio populations in its native Eurasian range and, given the well established populations of phloem- and woodboring beetles in its introduced range in North America, could do so here as well.

A more extensive survey of the response of S. noctilio to other ophiostomatoid species (including O. minus ) is warranted. The investigation of the volatile profiles of these fungal species may also provide more insight into the interaction of the wasp with the subcortical insect community. The extent of the woodwasp’s response to these fungi would provide insight into the effect which this phenomenon may have on wasp populations, i.e. does the wasp abandon the whole tree if a repellant fungus is present?

60

Chapter 4

Pinus spp. colonization by Sirex noctilio and it co- habitants in southern Ontario, and evidence for interaction

Abstract

The woodwasp Sirex noctilio has been recently introduced into North America. Several

natural enemies are known to help limit populations in both its native and introduced ranges.

Other phloem- or woodboring insects could share the tree with the woodwasp and could also

influence its distribution or its life history characteristics, and affect populations of the

woodwasp over time. Sixty S. noctilio -infested pines were felled and all phloem- and

woodboring insects were collected and identified from each 1 m section of the tree. Sirex

noctilio was sometimes in a tree alone but commonly shared the tree with phloem- or

woodboring beetles including Tomicus piniperda , Pissodes nemorensis , Ips grandicollis ,

Monochamus carolinensis and Gnathotrichus materiarius . Sirex noctilio was distributed

throughout the tree stem and this distribution overlapped with that of the beetles. Signs of

interaction between the woodwasp and co-habiting beetles included less emergence of S.

noctilio from trees with beetles as compared to those without, and greater prothorax size in

woodwasps emerging from trees with beetles. This finding demonstrates that beetles may

affect S. noctilio population dynamics over time.

61

1 Introduction

The woodwasp Sirex noctilio was recently introduced into eastern North America (Hoebeke et al. 2005; de Groot et al. 2006). In its native range in Eurasia and North Africa, it is a secondary pest of pine species (Spradbery & Kirk 1978) and is of little economic or ecological concern. Prior to its introduction in North America, the wasp established in parts of the southern hemisphere (summarized by Ciesla 2003) where it has at times caused extensive mortality in plantations of introduced pines (Neumann & Minko 1981). The extent to which this species will become a problem in North America is yet to be determined

(Dodds et al. 2010, but see Yemshanov et al. 2009).

Models predict that most of North America is climatically suitable for S. noctilio to establish and persist (Carnegie et al. 2006). Natural enemies of the woodwasp include several parasitoids, a parasitic nematode and birds; much recent research is focussed on the effects of the first two factors on S. noctilio morality (reviewed in Neumann et al. 1987; Hurley et al.

2007). The woodwasp is expected to share its host tree with other phloem- and woodboring insects but the degree to which interactions with them could influence S. noctilio survival and development, and subsequently its population, is not well known. This could be a factor limiting its impact in Eurasia where there is a well developed community of insects co- habiting the tree with the wasp (Wermelinger et al. 2008) and may be important in the wasp’s introduced range in North America.

There are a number of Curculionidae and Cerambycidae that inhabit pine trees in eastern North America. Insects such as the introduced bark beetle Tomicus piniperda prefer pine hosts with similar characteristics to that favoured by S. noctilio (Schroeder 1987; Paine et al. 1997). A number of other bark beetles, as well as species of Monochamus and Pissodes , favour dying or recently dead pines (USDA 1985) and all could be expected to share host

62 trees with the woodwasp. Therefore there is great potential for interactions to occur between these beetles and the woodwasp.

Sirex noctilio has a complex biology and this may influence the way in which interactions with other subcortical insects could occur. The woodwasp has an obligate relationship with its fungal symbiont, Amylostereum areolatum (Gaut 1969). After the female woodwasp oviposits into the sapwood of her host tree (Madden 1974), she deposits the fungus as well as a phytotoxic mucus in an adjacent drill tunnel (Coutts & Dolezal 1969).

The mucus and the fungus are thought to act to act in concert to kill the tree thus improving conditions for fungal growth (Coutts 1969b). When the growth of A. areolatum is impeded, woodwasp eggs fail to hatch (Madden 1981) or larvae may starve (Coutts & Dolezal 1965;

King 1966).

Many subcortical beetle species that are expected to share the tree with S. noctilio have known association with ophiostomatoid (blue stain) fungi (e.g. Nevill & Alexander

1992; Hausner et al. 2005) that they introduce into the tree. These beetle-associated fungi are primary colonists of dying or recently dead trees (Kirisits 2004) and these first arrivers are characteristically fast-growing and effective at exploiting a new resource (Boddy 2000).

Amylostereum areolatum is a weak competitor in comparison to some of these fungi and cannot establish in sapwood already colonized by these species (Chapter 5 and see Hanson

1939; King 1966; Titze & Stahl 1970). Therefore, the woodwasp must compete with these beetles for an ephemeral resource for its symbiont and this exploitative competition for substrate could act as a bottom-up factor influencing S. noctilio populations.

It would be of reproductive benefit for the wasp to deposit its eggs and fungus into a competition-free substrate such as trees, or parts of trees, uncolonized by other phloem- or woodboring species. Sirex noctilio avoids at least one species of beetle-vectored fungus

(Chapter 3). We could postulate that the benefit of ovipositing in competition-free space was

63 an evolutionary driver for S. noctilio and that competition pushed it to use living trees not

colonized by ophiostomatoid fungi. Of all the woodwasps, S. noctilio has the largest supply of mucus and this contributes to its ability to successfully overcome the defences of living trees in contrast to its congenerics which do not (Spradbery 1977).

Evidence of interspecific interactions between S. noctilio and subcortical beetles could include inter-tree partitioning, altered within-tree distribution or different patterns in S. noctilio life history traits. Spatial partitioning of species in standing host trees is well documented in bark beetle communities where colonization of the tree by two or more species affects the distribution of one or more of them (e.g. Paine et al. 1981; Schlyter &

Anderbrant 1993; Ayres et al. 2001). Smaller female size and fewer progeny as a result of interspecific competition are also evident in these bark beetle communities (e.g. Rankin &

Borden 1991; Schlyter & Anderbrant 1993; Amezaga & Garbisu 2000). Though not described in the bark beetle community, sex ratio may also be affected by this type of competition (Andersen 1961).

The objectives of this study were threefold: i) to examine whether other phloem- and woodboring insects share the tree with S. noctilio and if so, determine their identity and the frequency with which they are present; ii) to examine within-tree distribution of S. noctilio and other phloem- and woodborers and investigate whether they overlap in distribution within the tree; and iii) if S. noctilio inhabits the same areas of the tree as subcortical beetles do, to investigate evidence of differing patterns in woodwasp life history traits (abundance, size and sex ratio) with and without these beetles.

2 Methods Sixty dead S. noctilio -infested Pinus spp. trees were selected from 17 pine forests in southern

and central Ontario. These forests were those in which S. noctilio females had been collected 64 in survey traps within the preceding two years. To select the trees, in each site trees that had recently died (those with red or brown foliage) were examined with binoculars along their entire stem length for symptoms of S. noctilio attack (resin beads or drips, described in

Chapter 2). Trees with a minimum of five resin beads and that did not have extensive signs of

woodpecker activity were selected. I aimed to collect a minimum of three trees meeting these

selection criteria per site, however, in some sites there were fewer than three trees meeting

the selection criteria. Between one and nine trees per site were collected depending on the

number of trees available. Twenty two trees were felled in May 2007 and 38 in May 2008.

Tree species, height, diameter at breast height (dbh), crown dominance class, and

canopy closure were recorded for each tree. Tree crown dominance was classified by canopy

position of the tree as follows: dominant trees had crowns extending above the main canopy,

co-dominant tree crowns were at the level of the main canopy, intermediate tree crowns

reached the main canopy but were below it, and suppressed trees were entirely overtopped by

the main canopy. Canopy closure was the number of quadrants of the tree crown touched or

overlapped by neighbouring trees, and provided an estimate of sun exposure. Tree location,

species and other attributes are shown in Table 4.1.

Trees were cut at the base, sectioned into 1 m lengths (bolts) and numbered

consecutively. The mid-bolt diameter, bark thickness and bark texture of each segment was

recorded. To calculate bark thickness I used the mean of eight bark thickness measurements,

four from each end of the bolt, taken at 90° from each other. Bark texture was categorized as

furrowed (thick ridged bark), scaly (having thick scales of bark), flakey (thin, peeling bark)

or smooth.

65

Table 4.1. Geographical location, tree attributes and year felled for all study trees in southern and central Ontario. Location & UTM co-ordinates (zone easting northing) Species Dbh Height Dominance Canopy Year (cm) (m) closure (no. felled quadrants overlapped)

Glen Cross, ON (17T 577708 4872360) Pinus banksiana 15.4 11.56 Suppressed 4 2007 Palgrave, ON (17T 595740 4864020) Pinus banksiana 10.5 10.41 Suppressed 3 2007 Palgrave, ON (17T 595740 4864020) Pinus banksiana 11.4 11.50 Suppressed 3 2007 Palgrave, ON (17T 595740 4864020) Pinus banksiana 14.0 15.44 Co-dominant 2 2007 Palgrave, ON (17T 595740 4864020) Pinus banksiana 11.6 12.49 Co-dominant 1 2007 Indian Point, ON (17T 378740 5069815) Pinus resinosa 16.9 11.70 Co-dominant 3 2008 Indian Point, ON (17T 378740 5069815) Pinus resinosa 12.7 10.80 Intermediate 3 2008 Indian Point, ON (17T 378740 5069815) Pinus resinosa 11.4 11.50 Intermediate 4 2008 Limerick Forest, Spencerville, ON (18T 450441 4966719) Pinus resinosa 16.7 13.30 Suppressed 4 2008 Brighton, ON (18T 272880 4880485) Pinus sylvestris 11.5 7.41 Suppressed 2 2007 Brighton, ON (18T 272880 4880485) Pinus sylvestris 13.8 13.22 Co-dominant 1 2007 Brighton, ON (18T 272880 4880485) Pinus sylvestris 16.7 12.93 Intermediate 3 2007 Cavan, ON (17T 697478 4895472) Pinus sylvestris 14.0 15.10 Suppressed 4 2008 Cavan, ON (17T 697478 4895472) Pinus sylvestris 8.8 12.50 Suppressed 4 2008 Douglastown, ON (17T 661596 4759839) Pinus sylvestris 17.7 14.47 Co-dominant 4 2007 Douglastown, ON (17T 661596 4759839) Pinus sylvestris 11.5 8.90 Intermediate 2 2007 Douglastown, ON (17T 661596 4759839) Pinus sylvestris 11.5 10.90 Intermediate 5 2007 Eden Mills, ON (17T 571153 4825963) Pinus sylvestris 11.0 16.41 Intermediate 3 2007 Eden Mills, ON (17T 571153 4825963) Pinus sylvestris 11.7 16.63 Intermediate 3 2007 Elmwood, ON (17T 502506 4898199) Pinus sylvestris 17.7 13.38 Co-dominant 3 2007 Elmwood, ON (17T 502506 4898199) Pinus sylvestris 15.5 15.92 Co-dominant 1 2007 Elmwood, ON (17T 502506 4898199) Pinus sylvestris 17.2 16.28 Co-dominant 2 2007 Elmwood, ON (17T 502506 4898199) Pinus sylvestris 13.0 14.10 Co-dominant 3 2008 Elmwood, ON (17T 502506 4898199) Pinus sylvestris 15.0 13.30 Co-dominant 3 2008 Elmwood, ON (17T 502506 4898199) Pinus sylvestris 16.0 12.80 Co-dominant 2 2008 Forestville, ON (17T 551101 4728315) Pinus sylvestris 19.6 15.20 Co-dominant 3 2008 Forestville, ON (17T 551101 4728315) Pinus sylvestris 15.7 11.20 Intermediate 3 2008 Goderich, ON (17T 443122 4839677) Pinus sylvestris 11.8 14.22 Intermediate 1 2007 Goderich, ON (17T 443122 4839677) Pinus sylvestris 10.9 9.03 Suppressed 3 2007 66

Table 4.1 cont

Location Species Dbh Height Dominance Canopy Year (cm) (m) closure (no. felled quadrants overlapped) Goderich, ON (17T 443122 4839677) Pinus sylvestris 12.3 10.03 Suppressed 3 2007 Limerick Forest, Spencerville, ON (18T 450441 4966719) Pinus sylvestris 21.0 18.40 Suppressed 4 2008 Orangeville, ON (17T 574289 4862585) Pinus sylvestris 13.0 6.80 Intermediate 3 2008 Orangeville, ON (17T 574289 4862585) Pinus sylvestris 11.0 6.30 Suppressed 4 2008 Orangeville, ON (17T 574289 4862585)` Pinus sylvestris 13.5 6.80 Intermediate 2 2008 Sandbanks Provincial Park, ON, Site 2 (18T 317667 4863693) Pinus sylvestris 9.9 10.43 Intermediate 3 2007 Sandbanks Provincial Park, ON, Site 2 (18T 317667 4863693) Pinus sylvestris 18.9 15.90 Co-dominant 4 2007 Sandbanks Provincial Park, ON, Site 2 (18T 317667 4863693) Pinus sylvestris 14.6 13.45 Suppressed 3 2007 Sandbanks Provincial Park, ON, Site 1 ( ca. 18T 317667 4863851) Pinus sylvestris 11.3 7.40 Intermediate 3 2008 Sandbanks Provincial Park, ON, Site 1 (ca. 18T 317667 863851) Pinus sylvestris 11.0 8.20 Intermediate 3 2008 Sandbanks Provincial Park, ON, Site 2 (18T 317667 4863693) Pinus sylvestris 13.8 11.80 Co-dominant 2 2008 Sandbanks Provincial Park, ON, Site 2 (18T 317667 4863693) Pinus sylvestris 19.0 11.30 Co-dominant 4 2008 Sandbanks Provincial Park, ON, Site 2 (18T 317667 4863693) Pinus sylvestris 17.5 11.80 Co-dominant 4 2008 Sandbanks Provincial Park, ON, Site 2 (18T 317667 4863693) Pinus sylvestris 14.5 10.60 Co-dominant 1 2008 Sandbanks Provincial Park, ON, Site 2 (18T 317667 4863693) Pinus sylvestris 18.5 12.60 Co-dominant 3 2008 Sandbanks Provincial Park, ON, Site 2 (18T 317667 4863693) Pinus sylvestris 14.0 10.70 Co-dominant 4 2008 Sandbanks Provincial Park, ON, Site 2 (18T 317667 4863693) Pinus sylvestris 18.0 12.60 Co-dominant 3 2008 Sandbanks Provincial Park, ON, Site 2 (18T 317667 4863693) Pinus sylvestris 17.0 11.30 Co-dominant 3 2008 Sandbanks Provincial Park, ON, Site 2 (18T 317667 4863693) Pinus sylvestris 21.5 12.80 Co-dominant 3 2008 Sauble Beach, ON (17T 479691 4944524) Pinus sylvestris 12.0 8.12 Suppressed 3 2008 Sauble Beach, ON (17T 479691 4944524) Pinus sylvestris 16.0 11.40 Suppressed 4 2008 Sauble Beach, ON (17T 479691 4944524) Pinus sylvestris 16.0 10.90 Suppressed 2 2008 Tottenham, ON (17T 592582 4873134) Pinus sylvestris 10.5 4.50 Suppressed 2 2008 Tottenham, ON (17T 592582 4873134) Pinus sylvestris 5.5 4.20 Suppressed 3 2008 Tottenham, ON (17T 592582 4873134) Pinus sylvestris 3.5 4.80 Suppressed 4 2008 Tottenham, ON (17T 592582 4873134) Pinus sylvestris 9.3 7.50 Intermediate 4 2008 Tottenham, ON (17T 592582 4873134) Pinus sylvestris 7.5 10.50 Co-dominant 4 2008 Tottenham, ON (17T 592582 4873134) Pinus sylvestris 12.3 10.00 Intermediate 4 2008 Tottenham, ON (17T 592582 4873134) Pinus sylvestris 8.5 6.70 Intermediate 4 2008 West Bay, ON (17T 409901 5075367) Pinus sylvestris 22.5 16.80 Intermediate 2 2008 West Bay, ON (17T 409901 5075367) Pinus sylvestris 18.6 15.70 Suppressed 3 2008

67

Bolts were stood apex side up in individual emergence containers which were 36 cm x 102 cm cardboard cylinders with tight-fitting plastic lids (Greif Lok-Rim® Fibre drum,

Greif Inc, Delaware). A hole of approximately 9 cm in diameter was cut in the centre of each lid and a 475 ml plastic cup, with a small exit hole cut in the bottom of it, was inverted and glued to the lid (Fig.4.1). A second, smaller, clear plastic, cup was set over top of the first to contain the insects until collected. With this arrangement emerging insects that were attracted to light would exit the cylinder into the collection cup. A string was stapled to the end of each bolt and the other end glued outside the exit hole of the cup to assist insects in exiting the cylinder. The rearing containers were stored upright in covered, translucent sided sheds at

Angus, Ontario.

Figure 4.1. Collection cup arrangement for collection of subcortical insects from rearing containers containing Pinus sylvestris bolts.

All emerging phloem- and woodboring insects were collected regularly from each rearing container between May and October. During the period of highest insect emergence

(late June to early September) insects were collected five times per week; otherwise they were collected three times per week. On each collection visit, the rearing containers were examined for insects that did not exit the cylinder into the collecting cup; these insects were 68 manually extracted. Each collection from each rearing container was stored individually in a vial of 70% ethanol until the insects could be identified in the laboratory. A number of female S. noctilio were kept alive to be used in a subsequent experiment (Chapter 3) so they

were individually contained and labelled, and processed after the completion of that

experiment. The prothorax width of each woodwasp was measured and its sex was recorded.

Prothorax width was selected as a surrogate for woodwasp size as it showed good test, re-test

reliability. Bolts from trees collected in 2008 were left in the rearing containers through the

2009 growing season in order to assess whether S. noctilio had a two-year lifecycle in

Ontario. Insects emerging in the second year were collected and processed as above.

Siricids, curculionids and cerambycids were identified to species using keys and

information available in Schiff et al. (2006), Bright (1976), Passoa and Cavey (updated 2006)

Downie and Arnett (1996) and Yanega (1996), along with supplementary material provided

by H. Goulet (unpublished). Insects that required specialized taxonomic expertise (weevils

and Tetropium spp.) were identified by, or had identifications verified by, specialists in these

taxa. Species authorities can be found in Appendix 1 and are not referred to in the following

sections. Voucher specimens are deposited in the Forest Health insect collection maintained

by the Canadian Forest Service, Natural Resources Canada, Sault Ste. Marie, Ontario.

2.1 Data Analysis To address the difference in size between the trees, insect abundance data were standardized by the surface area of the tree; this was calculated using the equation for the surface area of the sides of a tapered cylinder. Standardized emergence data from the first season after felling were used in all analyses.

Data were tested for normality and homoscedasticity before analysis, and were transformed using either log n (tree level analysis) or log n+1 (bolt level analysis). Data that

could not be adequately transformed were evaluated with the equivalent nonparametric test. 69

A two sample t-test or a Mann-Whitney U test was used to compare insect parameters at a whole tree level, and repeated measures ANOVA or a Wilcoxon test was used to compare insect parameters by bolt. All univariate analyses were conducted in SYSTAT 12.0

(SYSTAT Software Inc, Chicago) and the alpha value was set to p < 0.05. There was no statistical difference in parameters related to the year in which the tree was cut, so year effects were not addressed in the results.

Principal Components Analysis (PCA) and Redundancy Analysis (RDA) were selected for multivariate modelling as these two techniques provided the best representation of the raw data; analyses were performed with CANOCO 4.5 (ter Braak and Šmilauer,

Wageningen). Only species represented by ≥ 15 individuals were used in the analyses. For the RDA analyses Monte Carlo simulation was used (999 permutations) and only the significant environmental variables (p < 0.05) were included in the model. Environmental variables that might be expected to explicitly autocorrelate were not included in the model.

Analyses were based on a correlation matrix of log + 1 transformed species data.

3 Results A total of 15 986 subcortical insects representing 43 species emerged from the bolts in the first year (2007/2008). Sirex noctilio , which was the most common insect collected (7018

individuals), emerged from all 60 trees. The other major insect species included Tomicus

piniperda (6242), Pissodes nemorensis (1111), Ips grandicollis (521), Gnathotrichus materiarius (218) and Monochamus carolinensis (191). There were two other siricid species collected, Sirex nigricornis (204) and S. edwardsii (20). There were 496 individuals representing seven different species that emerged from the 38 trees in the second year. From

16 of these trees, 152 Sirex noctilio emerged (72 female, 80 male) in 2009. One species of

70 cerambycid, Xylotrechus sagittus sagittus (68 individuals) was primarily collected in year two. A full species list can be found in Appendix 2.

Both male and female S. noctilio emerged from all three tree species and in both years

(Table 4.2). Sirex noctilio emerged from each pine species between late June and the third week of September in each year. The number of S. noctilio males and females per tree was highly variable (Table 4.2) and ranged from 3 to 569 individuals, and on a per bolt basis ranged from 0 to 149. Sex ratios (M:F) over all 60 trees ranged widely, from 0.3 to 41.0.

Mean prothorax width ranged from 2.6 mm in Pinus banksiana to 4.6 mm in Pinus resinosa

(Table 4.2).

Table 4.2. Summary statistics for Sirex noctilio and most common co-habiting species emerging from Pinus spp. trees from southern and central Ontario. Per tree emergence averaged over the number of trees the insect species emerged from (indicated in parenthesis). 2008 2007 Pinus resinosa Pinus sylvestris Pinus sylvestris Pinus banksiana (n=4) (n=34) (n=17) (n=5) Sirex noctilio (F) Mean 63 (4) 22 (33) 39 (17) 15 (5) Range 15-26 5-76 5-127 1-42 S. noctilio (M) Mean 39 (4) 71 (34) 144 (17) 39 (5) Range 4-84 1-414 5-452 12-104 Sex ratio (M:F) Median 0.53 3.65 3.90 4.67 Range 0.27- 0.95 0.50-41 0.83-5.04 0.54-13.0 Prothorax width Mean 4.64 (4) 2.91 (34) 2.87 (17) 2.58 (5) Range 4.03-4.64 2.35-4.23 2.24-4.45 1.96-3.23 Tomicus piniperda Mean 247 (3) 208 (15) 339 (7) 2(2) Range 19-630 1-1730 1-2375 1-3 Pissodes Mean 163(4) 20 (17) 10 (11) 3 (2) nemorensis Range 6-377 1-161 1-29 2-3 Ips grandicollis Mean 37 (4) 19 (13) 10 (11) 3 (5) Range 2-104 1-141 1-24 1-6 Gnathotrichus Mean 0 5 (7) 31 (6) 0 materiarius Range - 1-12 3-120 - Monochamus Mean 0 11 (10) 16 (3) 0 carolinensis Range - 1-49 11-13 - n = number of trees, M=male, F=female

The three most commonly collected co-habiting insects, T. piniperda , P. nemorensis and I. grandicollis emerged from all three pine species, though few individuals emerged from

P. banksiana (Table 4.2). Gnathotrichus materiarius and M. carolinensis , the next two most

71 abundant species, emerged only from Pinus sylvestris . There was a high variability of most co-habitants between trees in each site but M. carolinensis was more common in two of the

17 sites.

3.1 Within tree distribution The distribution of S. noctilio emergence throughout the tree stem for each tree species is shown in Figs. 4.2a-c. The following analyses were restricted to the lower seven bolt levels within the tree as this allowed all P. banksiana , all P. resinosa and most P. sylvestris trees to be included in the analyses. The within-tree distribution of male and female woodwasp emergence was similar in P. sylvestris in both years so the two years were pooled. In P. sylvestris (n = 36), neither male nor female emergence was distributed evenly through the tree stem (males: F2,216 , = 5.52, p < 0.001, females F2,216 = 2.77, p = 0.01). The greatest number of wasps emerged in the mid-tree (Fig. 4.2a). Female wasp emergence in P. resinosa

did not differ significantly through the tree (F6, 18 = 2.06, p = 0.11); though females were uncommon in the first metre (Fig. 4.2b). Male emergence was insufficient to examine statistically. In P. banksiana , male emergence differed through the stem but female emergence did not (males: F 6,24 , = 3.53, p = 0.01, females F 6,24 = 1.39, p = 0.26) (Fig. 4.2 c).

In general, regardless of tree species and year T. piniperda emergence (Fig. 4.3a),

when present, was especially prevalent in the lower 4 m of the tree though this species was

found throughout the tree. Pissodes nemorensis (Fig. 4.3b) and Ips grandicollis (Fig. 4.3c) were distributed throughout the tree as was M. carolinensis (Fig.4.3d), though the latter was

uncommon in the first metre of the tree. Gnathotrichus materiarius (Fig.4.3e) was most

abundant in the first metre of the tree when it was present.

72

a) Male 9 Female

7 Males: p <0.001 5 Females: p = 0.01

3 Metres above ground 1

0 5 10 15 20 Mean std. S. noctilio (+SE)

b) Male 9 Female

7 Males: insufficient data to 5 analyse Females: p = 0.11

3

Metres above ground 1 0 5 10 15 20 Mean std. S. noctilio (+SE)

c) Male 9 Female

7

Males: p = 0.01 5 Females: p = 0.26

3 Metres above ground 1 0 5 10 15 20 Mean std. S. noctilio (+SE)

Figure 4.2. Mean male and female Sirex noctilio emergence, standardized by tree surface area, per one metre section of tree stem from a) Pinus sylvestris (n = 36 trees), b) Pinus resinosa (n = 4) and c) Pinus banksiana (n = 5) from southern and central Ontario 2007-2008. Significance results for repeated measures ANOVA for woodwasp emergence per first four 1m sections of tree trunk embedded.

73

a) 9

7

5

3

Metresabove ground 1

0 10 20 30 40 50 60 70 Mean std. T. pinipdera (+ SE)

b) 9 7

5

3 Metres above ground 1 0 10 20 30 40 50 60 70 Mean std. P. nemorensis (+SE)

c) 9

7

5

3

Metresabove ground 1

0 10 20 30 40 50 60 70 Mean std. I. grandicollis (+SE)

d) 9 7

5

3 Metresabove ground

1

0 10 20 30 40 50 60 70 Mean std. M . caroliniensis (+ SE)

e) 9 7

5

3

Metresabove ground 1 0 10 20 30 40 50 60 70 Mean std. G. materiarius (+ SE)

Figure 4.3. Mean beetle emergence, standardized by tree surface area, per one metre section of Pinus spp. tree stem (tree species pooled) a) Tomicus piniperda (n = 27 trees), b) Pissodes nemorensis (n =3 4), c) Ips grandicollis (n = 33), d) Monochamus carolinensis (n = 13), e) Gnathotrichus materiarius (n = 13), from southern and central Ontario positive in 2007-2008.

74

3.2 Effect of tree characteristics on the bark beetle and woodborer community

A PCA of all trees and both years pooled was used as both eigenvalues and the interpretation of the model were similar to that for the years and tree species examined separately. In this model, 35% of the variance was accounted for on the first two axes; 19.8% on axis one and functional groups (Fig. 4.4). Axis one was strongly influenced by cerambycid beetles (e.g. M. carolinensis , Tetropium species novo, Neacanthocinus pusillus , and Astylopsis sexgutta ) on the negative end of the axis and Curculionidae (e.g. P. nemorensis , T. piniperda , I. grandicollis ) 1.0 G. mat O. cael. X. sax

S. noct.

T. lin. Axis 1 1 Axis λ 0.198 0.198 H. opac. M. car. T. sp. n A. sex P. nem. H. pin. A. pus I. grand T. pin.

S. edw.

-0.4 S. nig.

-0.6 Axis 2 λ 0.155 1.0

Figure 4.4 Principle Components Analysis ordination scatterplot of bark and woodboring beetle species (represented by ≥ 15 individuals) from Sirex noctilio -positive Pinus spp. from southern and central Ontario. A. pus = Neacanthocinus pusillus , A. sex = Astylopsis sexgutta , G. mat = Gnathotrichus materiarius , H. opac = Hylastes opacus , H. pin = Hylurgops rugipennis pinifex , I. grand = Ips grandicollis , M. car = Monochamus carolinensis , O. cael = Orthotomicus caelatus , P. nem = Pissodes nemorensis , T. sp. n = Tetropium sp. nov., T. pin = Tomicus piniperda , T. lin = Trypodendron lineatum , S. edw = Sirex edwardsii , S. nig = Sirex nigricornis , S. noct = Sirex noctilio , X. sax = Xyleborinus saxesini . 75 and the other two siricids ( S. nigricornis and S. edwardsii ) influenced the positive end.

Ambrosia beetles (e.g. G. materiarius and Xyleborinus saxesini ) influenced axis two. Sirex noctilio was situated centrally in this plot and not associated with any particular species or feeding guild.

In the RDA, although both tree height and dominance class (classified as a nominal variable) were significant environmental characteristics in the model, only 11% of species variance was accounted for on the first two axes, 7.1 % on axis one and 3.9% on axis two.

Tree height had a strong influence on axis one, tree dominance influenced axis two. Both S. noctilio and T. piniperda were weakly associated with greater tree height, the cerambycid beetles were associated with co-dominant tree class. A further 16.8% of variance was accounted for by axis four which had no apparent association with environmental characteristics.

To assess the role of local characteristics, a second RDA was performed on bolt data, using tree as a covariate. In this model, bolt diameter, average bark thickness, position in tree and bark texture were all significant environmental variables, however, the first two axes in the model accounted for only 5.2% of the species variance (3.8% on axis one and 1.3% on axis two). Tomicus piniperda and other bark beetles were associated with bolts having larger diameters and thicker bark and S. noctilio was associated with the scaly textured bark common in the lower mid-tree.

3.3 Evidence for interactions Sirex noctilio was the sole subcortical inhabitant of 10% (6) of the trees (all P. sylvestris ). Of

the remaining trees two patterns were evident, distinguished by a natural break in the data

set; they either had few co-habiting individuals (≤9), or many ( ≥17). Using this approach,

58% of the trees were in the latter category (31 P. sylvestris , 3 P. resinosa and 1 P. banksiana ). Within this group of 35 trees, about half had two or more co-habiting species, the 76 remaining had one primary co-habiting species that exceeded 10 individuals per tree (Fig.

4.5); these primary co-habitants were most commonly T. piniperda (9 trees), M. carolinensis

(4) or Neacanthocinus pusillus (2). In trees with ≤9 co-habiting individuals (14 P. sylvestris ,

1 P. resinosa and 4 P. banksiana ), there were a variety of beetle species represented. Most

commonly these were P. nemorensis , I. grandicollis and M. carolinensis . In trees with >1

and ≤9 individuals there was always more than one beetle species present.

T. piniperda 3 or more co- 19% habitants 29%

M. carolinensis 15%

A. pusillus 7% 2 co-habitants P. nemorensis 22% 4%

G. materiarius 4%

Figure 4.5. Primary co-habitants (>10 individuals) with Sirex noctilio in beetle-positive ( ≥17 beetles per tree) Pinus spp. trees (all tree species pooled) from southern and central Ontario in 2007-2008.

To facilitate exploration of the data for evidence of possible interactions between the woodwasp and co-habiting phloem- and woodborers that would suggest future experimental direction, the data were organized and processed as follows. First, all beetles, regardless of species, were pooled. The effects of each beetle species were expected to be indirect, as a result of interactions between beetle-vectored ophiostomatoid fungi and either the woodwasp itself or its fungus (see Chapters 3 and 5 for examples). Most of the co-habiting beetle species have documented relationships with ophiostomatoid fungi and the rest of the species

77 could be expected to as well. Therefore, they could all be expected to have similar effects on the woodwasp. Second, trees with few beetles (≤9 individuals in total) were grouped with those with no co-habitants and categorized as “beetle-negative”. Trees with ≥ 17 beetles were termed beetle-positive. Third, all tree species were pooled together as initial analysis by species showed similar patterns. Fourth, a subset of trees of more similar heights was selected for analysis because there was a significant difference in height between beetle- positive and beetle-negative trees (e.g. tree height t = -3.31, df = 58, p = 0.002) and this overwhelmed any effect of co-habitants. Sirex noctilio was always alone in trees less than 7

m in height and always shared trees that were over 15.5 m with bark beetles, so trees 7-15.5

m in height (n = 45) were selected. Within this subset, beetle-negative (n = 18) and beetle-

positive (n = 27) trees were not significantly different in height (t = 0.29, df = 43, p = 0.77)

or dbh (t = -1.78, df = 43, p = 0.08). Although sample sizes were unequal, the variance was

slightly greater in the larger sample (beetle-positive trees), and a further sub-set of trees with

a more equal sample size elicited the same result.

Significantly fewer S. noctilio emerged from beetle-positive than beetle-negative trees

(t = 2.1, df = 43, p = 0.04) (Fig. 4.6). An ANCOVA model using dbh as a covariate improved

on this model (beetle effect: F 1,42 = 9.11, p = 0.004, dbh: F 1,42 = 9.32, p = 0.004). Results

were similar when males and females were examined separately (males: beetle effect: F 1,42 =

9.63, p = 0.003, dbh: F 1,42 = 7.98, p = 0.007; females: beetle effect: F 1,42 = 5.13, p = 0.03, dbh: F 1,42 = 7.67, p = 0.008). A repeated measures ANOVA by bolt showed that this distribution pattern was a result of greater S. noctilio abundance (pooled, male and female) through the tree (Table 4.3). There was no evidence of interaction between bolt level and beetle presence. In this model, only the lower four bolts of each tree were analysed as sample sizes decreased at higher bolt positions; however the results were similar when more bolt levels were used.

78

Beetle-negative 9 Beetle-positive

7

5 Beetle presence: p = 0.004

3 Metres above ground 1

0 5 10 15 20 Mean std. S. noctilio (+SE)

Figure 4.6. Mean Sirex noctilio emergence, standardized by tree surface area, per one metre tree stem section in beetle-positive ( ≥ 17 beetles per tree) and “beetle-negative” ( ≤9 beetles per tree) Pinus spp. trees in southern and central Ontario (2007-2008). Effect of beetle presence results from ANCOVA embedded.

Table 4.3 Repeated measures ANOVA results for effect of beetle presence on Sirex noctilio emergence from Pinus spp. in southern and central Ontario in 2007-2008. Emergence from the lower 4 (1 metre length) sections of tree stem compared. Between subjects Within subjects Within subjects Beetle status Bolt Interaction

S. noctilio pooled F1,43 =7.46, p=0.009 F3,129 =9.82, p<0.001 F3,129 =0.18, p=0.83

Female F1,43 =9.73, p=0.003 F3,129 =4.55, p=0.005 F3,129 =0.25, p= 0.81

Male F1,43 =9.74, p=0.003 F3,129 =9.14, p<0.001 F3,129 =0.43, p= 0.67 n = 45 trees

The average S. noctilio prothorax was wider in beetle-positive trees (3.18 mm) than beetle-negative ones (2.82 mm) (Mann-Whitney U = 138.5, p = 0.015). Results differed between gender (male: Mann-Whitney U = 130, p = <0.001; female: Mann-Whitney U =

199, p = 0.43), however females were still slightly larger in beetle-positive trees than beetle- negative ones. The difference in prothorax width was not significant with the Wilcoxon test of bolts 1 to 4 (average prothorax: z = -1.83, p = 0.07, female: z = -1.46, p = 0.14, male: z = -

1.83, p = 0.07).

79

Sex ratio did not differ significantly between beetle-negative and positive trees

(Mann-Whitney U = 267.5, p = 0.42).

Because larval or pupal mortality in the presence of ophiostomatoid fungi is thought to occur (Morgan & Stewart 1966b) and this could affect the number of S. noctilio emerging from beetle-positive trees, a subset of bolts was dissected to look for evidence of this interaction. Eighteen S. noctilio -positive bolts, nine each beetle-positive and beetle-negative, were selected. These bolts were sawed lengthwise into 1 cm wide planks and then were examined for dead larvae or pupae. There were no dead larvae or pupae found in beetle- positive bolts, but seven dead larvae were found in beetle-negative bolts. Because there was no evidence of fungus-mediated larval or pupal mortality in this subset of beetle-positive bolts, the remaining bolts were not dissected.

4 Discussion More than half of the 60 trees selected had 17 or more individuals of co-habiting beetles, and there were a number of different species and guilds of co-habitants. However, in 10% of the trees S. noctilio was found alone in the tree. None of the most common species of co-habiting

beetles was unexpected; similarly to S. noctilio , T. piniperda favours low vigour pines as brood habitat (Schroeder 1987) and the other common co-habiting species favour weakened, dying or recently dead pines (e.g. USDA 1985). Although the presence of these co-habitants was not unexpected, this is the first published description of the phloem- and woodboring insect community sharing the tree with S. noctilio in North America.

The within-tree distribution of S. noctilio is similar to that found elsewhere in the insect’s known range. In P. sylvestris and P. resinosa most of the woodwasp emergence was in the mid-tree and this is similar to that described in New Zealand, Australia and

(Morgan & Stewart 1966a; Neumann et al. 1982; Penteado et al. 1998). The mid-tree peak in 80 emergence may relate to the insect’s behaviour. Morgan and Stewart (1966a) suggest that a peaked distribution pattern is a result of an adjustment of the female’s oviposition circuit.

The female woodwasp typically works her way up the trunk of the tree and then flies to the bottom to begin another circuit but she may only fly part way down the trunk to begin the next circuit thus concentrating oviposition in the mid-tree region.

As a group, phloem- and woodboring beetles were distributed throughout the entire stem of the tree and the within-tree distribution of the commonly collected subcortical beetles in this study was generally similar to that described in the literature. For example, T. piniperda is known to favour the rough-barked lower stem of the tree but is also found in thinner barked sections (Schroeder 1997); G. materiarius is commonly found in the lower bole (USDA 1985); P. nemorensis is usually found throughout the trunk (Atkinson et al.

1988); and Monochamus species are typically found in regions of the trunk at least 3 m above ground (Walsh & Linit 1985). These patterns are all similar to my own findings. The literature suggests that the patterns described in this study are not atypical and therefore, we can expect that S. noctilio will regularly interact with phloem- and woodboring beetles throughout its distribution within the tree.

Tree characteristics that were associated with emergence patterns of subcortical insect species in the multivariate analyses can be expected to play a role in the distribution of these insects. Although both T. piniperda and S. noctilio were associated with larger trees in

the first RDA, T. piniperda was associated with sections of the tree that had furrowed bark

(as described by Schroeder 1997), and S. noctilio with sections of the tree stem with scaly bark common to the lower-mid trunk. The latter association may be a result of the above- described oviposition behaviour of the wasp as described by Morgan and Stewart (1966a), but it also corresponds to the observations of Jackson (1955) and Clark (1927) who suggest

81 that S. noctilio is restricted from ovipositing in lower, thicker barked areas of the stem due to its short ovipositor (but see Morgan & Stewart 1966a).

This study was not designed to investigate differences in S. noctilio characteristics

and subcortical insect communities between pine species, however, some patterns warrant

further investigation as these factors may affect population dynamics in different stand types.

The trend to larger S. noctilio and lower male:female sex ratio in P. resinosa as compared to

the other tree species, suggests that populations could increase more quickly in these

plantations. Conversely, the trend to fewer and smaller woodwasps in P. banksiana , relative

to other pine species, could limit population build up in these stands. The scarcity of beetle

co-habitants in P. banksiana could have a contrasting effect if they are significant

contributors to the wasp’s population dynamics.

There was no evidence that S. noctilio was excluded from regions of the tree when

phloem- and woodboring beetles were present. However, the reduction in, but identical

distribution of, S. noctilio emergence in beetle-positive trees compared to beetle-negative ones is a clear sign of interaction between these two groups. The most abundant co-habitants were Curculionidae that feed in the phloem and were therefore unlikely to encounter the woodwasp directly. Cerambycidae may consume S. noctilio larvae during the beetle’s sub- cortical phase, however, this is more likely to be facultative and depend on chance encounters (Dodds et al. 2001), which are expected to be few given the density of these two guilds. Interactions between S. noctilio and beetles are more likely to be indirect, and either

caused by effects of the beetle on the host tree itself or mediated by the fungal associates of

the beetles. Associations between the common beetles and ophiostomatoid fungi are

numerous, e.g. T. piniperda (Kirisits 2004; Jankowiak 2006; Ben Jamaa et al. 2007); P. nemorensis (Nevill & Alexander 1992; Gebeyehu & Wingfield 2003 and reviewed in Viri

2004); I. grandicollis (Yearian et al. 1972); and G. materiarius (reviewed in Kirisits 2004).

82

Mechanisms of fungal mediation include outcompeting the symbiont of S. noctilio

and thus impairing larval development (Chapter 5; Hanson 1939; King 1966; Titze & Stahl

1970), by causing larval (Morgan & Stewart 1966b) or pupal (Hanson 1939) mortality, or by

inhibiting oviposition by the woodwasp (Chapter 3). The mechanisms that would explain the

interaction between these beetle-associates and the woodwasp or its symbiont would depend

on the relative order in which they enter the tree. Colonization of the tree by subcortical

insects that have extended or bimodal flight seasons and that are able to reproduce

throughout the flight season of S. noctilio (July to September) such as Ips grandicollis (Ayres et al. 2001), P. nemorensis (Atkinson et al. 1988) and Monochamus carolinensis (Walsh &

Linit 1985; Yanega 1996), could enter the tree before or after the woodwasp and could

plausibly interact with it through any of the suggested mechanisms. However, the primary

co-habitant, T. piniperda , is primarily reproductively active in the early spring (Kennedy &

McCullough 2002) and appears to follow the woodwasp into a tree more often than the

reverse (Chapter 2). Given this, the extent of the reduced S. noctilio emergence in beetle-

positive trees in the absence of signs of increased larval mortality was surprising, although it

is possible that dissection of more bolts would reveal greater woodwasp larvae mortality. Egg

mortality induced by ophiostomatoid fungi does not seem likely since egg eclosion generally

occurs in 16-28 days (Morgan & Stewart 1966a) and is likely to have occurred before most

T. piniperda enter the tree, however, egg eclosion may be delayed by temperature or

sapwood conditions (Madden 1981) so this mechanism is possible especially in cases when

the woodwasp oviposits later in the season. It is also possible that there were subtle

differences in tree condition that affected relative colonization by the different insect species.

The larger body size in beetle-positive trees was unexpected given that the S. noctilio

symbiont is considered to be a poor competitor against beetle-associated fungi (King 1966)

and that the woodwasp larvae depend on it for nutrition. Amylostereum areolatum grows

83 more quickly in drier wood (Coutts & Dolezal 1965), so it is possible that accelerated wood drying in the tree as a result of the introduction of ophiostomatoid fungi (Chow & Obermajer

2007) improved S. noctilio nutrition by creating more favourable conditions throughout the

tree for the growth of the woodwasp’s symbiotic fungus compared to conditions in beetle-

negative trees. Nutrition in later larval instars of the woodwasp is not well known (discussed

in Chapter 1) so it is possible that in later stages of larval development the larvae could

derive nutrients from ophiostomatoid fungi. Although this pattern of greater prothorax size

and fewer wasps suggests an ecological trade-off between insect abundance and size, the

bolts were typically not full of larvae so this is unlikely.

This study describes the rich community of phloem- and woodboring insects sharing host trees with S. noctilio in southern Ontario. The woodwasp frequently shared the tree with these beetles and there was no evidence of altered woodwasp distribution within the tree.

However, there was a reduction in S. noctilio emergence in trees with beetles and this shows

evidence of competition between these two groups. These signs of interactions are the first

described for this system, and they suggest that further, experimental, investigations are

needed into the interactions between these feeding guilds and the effect that they have on S.

noctilio characteristics and populations.

84

Chapter 5

Interactions between Amylostereum areolatum and two bark beetle-vectored fungi, Leptographium wingfieldii and Ophiostoma minus

Abstract The woodwasp Sirex noctilio is invading North American forests where it will interact with a large guild of pine-inhabiting beetles and their associated fungi. The woodwasp’s fungal symbiont, Amylostereum areolatum plays an essential role in the wasp’s larval development; however, it is expected to be a poor competitor in the presence of fungi vectored by co- occurring insects. Here, I examine the effect of temperature and substrate on the competitive interaction between A. areolatum and two fungi vectored by a common co-habitant of the wasp, Leptographium wingfieldii and Ophiostoma minus . Beetle-associated fungi were usually able to capture more uncolonized resource than A. areolatum regardless of substrate or temperature, however, A. areolatum was able to colonize more space at some temperatures. Amylostereum areolatum could not gain substrate already colonized by the ophiostomatoid competitor. These findings suggest that some bark beetle-vectored fungi have the capacity to impact S. noctilio fitness by impeding growth of its fungal associate, but

environmental factors could affect the magnitude of these interactions. Bark and woodboring

beetles could influence woodwasp population dynamics and thereby influence the wasp’s

pest status in its newly introduced range in North America.

85

1 Introduction The woodwasp, Sirex noctilio , newly introduced in North America, has an important symbiotic relationship with the basidiomycete fungus, Amylostereum areolatum (Talbot

1977). This relationship may influence whether the wasp becomes a significant pest in its

newly introduced range. During oviposition, the female wasp inoculates the tree with the

fungus and a phytoxic mucus that are thought to act together to impair the physiological and

defensive responses of the tree and eventually result in its death (Fong & Crowden 1973).

The fungal symbiont is also integral to the woodwasp’s development. Egg eclosion is

delayed when conditions in the tree impede fungal growth (Madden 1981), and the fungus

provides an essential nutritional resource for at least the early larval instars of the wasp

(Madden & Coutts 1979). Larvae may starve if A. areolatum growth is inhibited (Coutts &

Dolezal 1965; King 1966), but when conditions for fungal growth are optimal, larger adults

are produced and their reproductive potential is greater (Madden 1981). Since the

reproduction and development of this insect is highly dependent upon its fungal symbiont,

inhibition of the symbiont’s growth could have considerable consequences to the population

dynamics of the woodwasp. This in turn may affect its status as a pest in its introduced range.

Some authors postulate that A. areolatum is a weak competitor against the fungi

vectored by bark or woodboring beetles leading to an antagonistic interaction that may

subsequently impede the development and survival of S. noctilio larvae (Hanson 1939;

Morgan & Stewart 1966b; King 1966; Titze & Stahl 1970). These suggestions are however

based mostly on anecdotal evidence, with the exception of King (1966) who noted that

Trichoderma sp. and Diplodia pinea were strongly antagonistic to A. areolatum but did not

quantify her findings.

In North America where S. noctilio was discovered in 2005 (Hoebeke et al. 2005; de

Groot et al. 2006) potentially significant interactions may occur between S. noctilio and a 86

large guild of bark and woodboring beetles and their associated fungi. In Canada, the known distribution of S. noctilio as of 2010 includes all of southern Ontario and a few locations in

Quebec (Canadian Food Inspection Agency 2010). A common bark beetle that may be

associated with S. noctilio in this range in Canada is the pine shoot beetle Tomicus piniperda

(Morgan et al. 2004) which vectors several ophiostomatoid fungi, most commonly

Leptographium wingfieldii and Ophiostoma minus (Jacobs et al. 2004; Kirisits 2004; Hausner

et al. 2005; Ben Jamaa et al. 2007). These fungi do not have an exclusive association with

this species and are vectored by a number of other beetle species (Jacobs et al. 2004; Kirisits

2004). Competitive interactions between beetle-vectored fungal species and A. areolatum

could influence the reproductive fitness and distribution of S. noctilio within individual trees and also across a wide geographic area.

The success of a fungus in colonizing wood substrate depends on its competitive ability to capture the substrate either primarily or secondarily. Primary resource capture strategies are those in which the fungus rapidly exploits the uncolonized resource and gains control over it (Rayner & Webber 1984). Secondary resource capture occurs when one fungal species is able to gain access or to colonize area already occupied by another fungal species.

Antagonistic interactions between fungal species may occur when two or more species meet during primary resource capture and this may result in either a deadlock between the two species, or the partial or total replacement of one species by another. Inhibition of fungal competitors can occur at a distance (antibiosis) or via direct contact; sensitivity to antibiosis is both species and strain dependant (Rayner & Webber 1984; Boddy 2000).

The purpose of this study was to assess the outcome of in vitro competition between the woodwasp fungal symbiont, A. areolatum and selected bark beetle-vectored ophiostomatoid fungi. I sought: i) to evaluate the outcomes of competition for uncolonized substrate between

87

A. areolatum and L. wingfieldii and O. minus ; ii) to examine how temperature modifies the outcomes of these interactions; iii) to evaluate the ability of A. areolatum to establish in already occupied substrates (secondary resource capture); and iv) to evaluate the effect of substrate on the outcomes of experiments between wood-inhabiting fungi.

2 Materials and methods

2.1 Fungal isolates Both of the known, recently identified, strains of Amylostereum areolatum in Canada were used in these experiments (Bergeron et al . submitted). Three strains each of two frequent

fungal associates of T. piniperda , L. wingfieldii and O. minus , were used as the potential

competitors. Strains of the same fungal species may have different growth rates (e.g. Lieutier

et al. 2004) and thus will modify the results of competition; therefore, examination of both

different species and strains is necessary. Collection information for these strains is shown in

Table 5.1. Throughout this chapter each strain is referred to by its isolate number. Before use

in the following experiments, each strain was inoculated onto, and re-isolated from sterile

Pinus sylvestris wood chips to help mitigate the effects of storage in culture. All three fungal species were then grown on potato dextrose agar (PDA) for use in all of the following experiments.

Table 5.1. Provenance information for the strains of Amylostereum areolatum , Leptographium wingfieldii and Ophiostoma minus used in all Chapter 5 experiments. Species Collection no. Host Location Collector A. areolatum SSM 075 7011 Pinus sylvestris Sauble Beach, ON C. Davis SSM 075 7013 Pinus sylvestris Halton, ON C. Davis L. wingfieldii SSM 025 7010 Pinus sylvestris Listowel, ON C. Davis SSM 025 7011 Pinus sylvestris Barrie, ON C. Davis SSM 025 7012 Tomicus piniperda Listowel, ON C. Davis O. minus SSM 075 7007 Pinus sylvestris Bracebridge, ON C. Davis WIN(M) 861 Pinus sylvestris Toronto, ON L. Baumel WIN(M) 1275 Pinus sylvestris Barrie, ON C. Davis SSM, Forestry Centre Culture Collection, WIN(M), University of Manitoba Culture Collection

88

2.2 Interactions between A. areolatum and ophiostomatoid species on artificial media Each strain of A. areolatum was tested against each of the three strains of each of the

ophiostomatoid species. For each of the pairs, a 5-mm plug of each contender was inoculated

onto potato dextrose agar (PDA) on opposite sides of a 9-cm Petri plate. The A. areolatum plug was inoculated four days prior to the contender to address the growth lag of this species.

Five replicates of each combination were inoculated, and each strain was also inoculated on its own as a control. All plates were sealed and stored in the dark in a 25°C (±1°C) growth chamber. The outer extent of each of the contenders was traced on the bottom of the plate every other day for two weeks, starting from day two after ophiostomatoid species inoculation. Colony boundaries were confirmed under a dissecting scope. Sequential and final colony extent markings for each species were traced onto paper, scanned and the areas were measured using Scion Image software (Scion Corporation, Frederick Maryland).

Contact zones were examined under a dissecting scope and signs of hyphal or mycelial interaction (e.g. persistent uncolonized zones, hyphal necrosis or morphological changes) were noted. A subset of each combination was observed for a further four weeks to monitor for encroachment of one species into the others space.

2.3 Interactions on wood substrate and effect of temperature The interaction experiment was repeated on autoclaved Pinus sylvestris wood chips (approx. dimensions: 3cm x 4cm x 2mm) over a range of temperatures. This was done because the type of substrate and the temperature used can affect fungal growth rate, and can do so differently for each species (e.g. Uzunovi ć and Webber 1998; Rice et al. 2007) and these factors could affect the outcomes interactions. Ten replicates of each of the treatment combinations and corresponding control chips were inoculated at opposite edges of the top

89

surface of the woodchip, in a similar manner as the previous experiment. Wood chips were suspended on glass triangles above filter paper moistened with sterile water. Plates were stored in the dark in growth chambers set at 10, 15, 20, 25 and 30°C (±1°C). Hyphae of each fungal species were distinctive in appearance and in a preliminary trial the colony boundaries of each of the solo-growing strains identified under magnification were tested by taking wood samples from immediately outside of the outer colony limit. No fungi grew from these test isolations, so colony boundaries were determined visually, under a dissecting scope, for this experiment. At seven days, half of the replicates of each pair of strains were measured; colony boundary edges were marked and each chip was photographed. The surface area colonized by each fungal species was measured from the photo with Scion Image software.

At 14 days, fungal colony boundaries on each of the remaining chips were similarly marked and measured. On a subset of chips, subsurface cultures were taken from each side of the contact zone and subsequently examined for evidence of the competing species. The set of interactions tests conducted at 25°C were compared to those conducted on PDA to look at substrate-related differences in interaction outcomes.

2.4 Ability to grow on pre-colonized resource This experiment used the same species and strain combinations as in the previous experiment. Surface-sterilized P. sylvestris woodchips were inoculated with each of the previously-used strains and stored, as previously described, until approximately 75% of the wood chip was colonized. Half of the chips of each species were autoclaved. The competing species was inoculated on top of the colonized area of both autoclaved and live fungal colonies. There were five replicates of each pair of competitors on each of the two substrate treatments (autoclaved and non-autoclaved), and five control replicates of a single strain

90

growing on un-colonized wood chips. At two weeks post-inoculation, the growth, or lack of growth of the second species added was recorded.

2.5 Data analysis To test for evidence of A. areolatum growth inhibition in the presence of contenders, the sequential area of growth on PDA of each A. areolatum strain in the presence of each ophiostomatoid species (three strains and control) was analyzed by repeated measures

ANOVA. Woodchip surface area colonized at seven days by each A. areolatum strain vs. the strains of each ophiostomatoid species (at 25°C) was compared using ANOVA. To test differences in primary resource capture between strains, colony extent of each contending species at 14 days was compared using t-tests. The effect of temperature, ophiostomatoid strain and the interaction of the two on resource capture at 14 days were analyzed by

ANOVA for each A. areolatum strain. Data were tested for normality and variance by inspecting graphs of the residuals from regression. The surface area of the woodchip colonized by each fungus was log n transformed and analysed in this form. For all analyses,

the alpha value was set at 0.05. Analyses were conducted with SYSTAT 12.0 (SYSTAT

Software Inc, Chicago) (ANOVA) or calculated in Microsoft Excel (2003) (t-test).

3 Results

3.1 Interactions on artificial media Some strains of the ophiostomatoid fungi grew quickly resulting in the entire surface area of the Petri plates being colonized with the two competing fungal species within six to eight days. Therefore, the repeated measures analysis was restricted to the first three measurements

(days two, four and six) to allow comparison between all species and strains of fungi. There was a significant effect of L. wingfieldii strain on the growth of both A. areolatum strains 91

(strain 075 7011: F 3,16 = 4.78, p = 0.02; 075 7013: F 3,16 = 7.91, p = 0.002) over the first six days of growth. A. areolatum colonized more surface area in the presence of some of the competing strains than it did growing alone (Fig. 5.1a-b). There was a similar effect of

O. minus on the growth of one of the A. areolatum strains but not the other (075 7011:

F3,16 =1.97, p = 0.16; 075 7013: F 3,16 = 5.59, p = 0.008) (Fig. 5.1c-d).

When the plates were entirely colonized, each strain of O. minus and L. wingfieldii

captured substantially more area than each A. areolatum strain (compare columns 4 with 5 in

Table 5.2). At two weeks, A. areolatum strains growing without competitors colonized

significantly more area than the same strains growing with L. wingfieldii or O. minus (Table

5.2; p<0.001 for Tukeys HSD of control vs. each contending species/strain). Ophiostoma

minus (all three strains pooled) colonized more space than L. wingfieldii (pooled) did when

competing with one A. areolatum strain (075 7011: t = -2.41, df = 28, p = 0.02) but there was

no difference in the amount of substrate colonized by each of the ophiostomatoid species

when contending with the other A. areolatum strain (075 7013: t = 0.14, d f= 28, p = 0.89).

Inspection of the zone between A. areolatum and L. wingfieldii strains under a dissecting

scope revealed a thin uncolonized zone between the two species in all cases. The A.

areolatum hyphae at the leading edge looked similar with and without contenders and were

never necrotic. Compared to solo-growing controls of the same species, the L. wingfieldii

hyphae were sometimes thicker (025 7010), sometimes thinner (075 7011) and sometimes

necrotic (025 7012) at the contact zone. At contact zones between A. areolatum and O.

minus , hyphae met but did not intermingle. There was no difference in A. areolatum

appearance in this area. Ophiostoma minus tended to show denser growth (barrage) at the

contact zones (all strains). The ophiostomatoid contenders never overgrew the A. areolatum

92

12 a) 10 Control 8 vs 025 7010 vs 025 7011

area (sq. cm) 6 vs 025 7012 4

2 0 A. areolatum 2 4 6 Day

12 b) 10 Control vs 025 7010 8 vs 025 7011 (sq. cm) 6 vs 025 7012 4 2 A. areolatum 0 2 4 6 Day

12 Control c) 10 vs 075 7007 8 vs 861 vs 1275 area (sq. cm) 6 4

2 0 A. areolatum 2 4 6

Day

12 d) Control 10 vs 075 7007

8 vs 861

ara (sq. cm) vs 1275 6 4

2

A. areolatum 0 2 4 6 Day

Figure 5.1. Amylostereum areolatum surface area (+SE) colonized on potato dextrose agar in the presence of ophiostomatoid competitors over 6 days a) 075 7011 growth in presence of Leptographium wingfieldii strains, b) 075 7013 growth in presence of L. wingfieldii strains, c) 075 7011 growth in presence of Ophiostoma minus strains, d) 075 7013 growth in presence of O. minus strains.

93

strains, though perithecia of two of the O. minus strains (075 7007 and 861) were found on A. areolatum colonized space at four weeks.

3.2 Interactions on wood substrate at 25°C At 25°C, the wood chip surface area colonized by each A. areolatum strain at seven days was significantly affected by the presence of L. wingfieldii (075 7011: F 3,16 = 14.9, p <0.001; 075

7013: F3,16 = 85.5, p <0.001). Both strains of A. areolatum colonized significantly less area in the presence of each of the L. wingfieldii strains than they did growing alone (Fig.5.2a).

Amylostereum areolatum area at seven days was also significantly affected by the presence of O. minus (075 7011: F 3,16 = 5.2, p = 0.01; 075 7013: F3,16 = 3.44, p = 0.04) which was a result of reduced A. areolatum growth in the presence of strain 861 compared to the control growth in both cases (Fig. 5.2 b).

7 a)b) 7 b) 075 7011 A 075 7011 A a) 6 6 075 7013 075 7013 5 5 a B a AB B AB 4 4

area (sq. cm) area (sq. cm) ab b 3 c 3 ab 2 B B 2 b b 1 1 0 0 A. areolatum A. areolatum Control 025 7010 025 7011 075 7012 Control 075 7007 861 1275 L. wingfieldii strain O. minus strain

Figure 5.2 Surface area (+SE) of sterilized Pinus sylvestris woodchip colonized by each of two Amylostereum areolatum strains (075 7011 and 075 7011) in the presence of ophiostomatoid competitors at 7 days a) Both A. areolatum strains growing with 3 Ophiostoma minus strains compared to control growth, b) Both A. areolatum strains growing with 3 Leptographium wingfieldii strains compared to control growth. Results of Tukey’s HSD test indicated in caps for A. areolatum strain 075 7011 and in lower case for strain 075 7013.

94

At two weeks, both ophiostomatoid species (all strains) colonized substantially more space than the A. areolatum strains, and the solo-growing A. areolatum colonized substantially more space than colonies growing in the presence of competing species (see columns 6 and 7 in Table 5.2 and see Table 5.3 25°C results). Ophiostoma minus colonized significantly more space compared to L. wingfieldii when competing against A. areolatum strain 075 7011 (t = -3.07, df = 28, p = 0.005) but not strain 075 7013 (t = -1.41, df = 28, p =

0.17). Inhibition zones were frequently present on chips inoculated with A. areolatum and L.

Table 5.2. Mean per cent (±SE) surface area colonized by contending fungal strain at two weeks on potato dextrose agar and on Pinus sylvestris wood chips, all stored at 25°C. ANOVA results for difference in amount of substrate colonized on control substrate vs. that with each contending species. Amylostereum Ophiostomatoid species and % space % space % space % space areolatum strain strain colonized colonized colonized colonized by A. by by A. by areolatum contender areolatum contender on PDA on PDA on wood on wood chip chip 075 7011 Leptographium 025 7010 14.8 (0.43) 85.3 (0.43) 12.0 (2.13) 48.8 (7.30) wingfieldii 025 7011 21.8 (0.69) 78.2 (0.69) 6.0 (0.63) 93.4 (2.25) 025 7012 24.8 (1.98) 75.2 (1.98) 15.9 (2.20) 65.9 (2.55) Control 46.4 (1.44) - 82.7 (2.40) - ANOVA F3,16 =112.5, - F3,16 =54.5, - results p<0.001 p<0.001 075 7013 025 7010 13.2 (0.70) 86.8 (0.70) 13.9 (2.13) 73.7 (5.66) 025 7011 9.8 (0.56) 90.2 (0.56) 9.0 (1.47) 70.0 (7.70) 025 7012 24.0 (0.83) 76.0 (0.83) 15.2 (3.87) 69.8 (4.63) Control 39.2 (1.71) - 76.6 (6.06) - ANOVA F3,16 =62.0, - F3,16 =101.9, - results p<0.001 p<0.001 075 7011 Ophiostoma 075 7007 20.9 (0.56) 79.1 (0.56) 20.7 (1.17) 79.3 (1.17) minus 861 18.9 (0.98) 81.1 (0.98) 18.5 (0.90) 80.2 (1.98) 1275 19.7 (2.53) 80.3 (2.53) 15.3 (1.80) 86.0 (2.20) Control 46.4 (1.44) - 82.7 (2.40) - ANOVA F3,16 =189.6, - F3,16 =51.1, - results p<0.001 p<0.001 075 7013 075 7007 16.4 (0.93) 83.6 (0.93) 17.6 (1.54) 82.4 (1.54) 861 16.5 (0.29) 83.5 (0.29) 18.1 (3.00) 80.9 (2.43) 1275 13.4 (0.50) 86.6 (0.50) 11.0 (1.28) 87.8 (0.58) Control 39.2 (1.71) - 76.6 (6.006) - ANOVA F3,16 =169.0, - F3,16 =58.19, - results p<0.001 p<0.001

95

wingfieldii contenders and this resulted in less than 100% colonization of the chip explaining why columns 6 and 7 in Table 5.2 did not always equal 100%; a subset of these left to grow beyond two weeks showed no further colonization of this zone. Neither L. wingfieldii nor A. areolatum were isolated from the territory of the other in subcultures taken at four weeks.

Ophiostoma minus was isolated from A. areolatum space; A. areolatum was not found in space colonized by O. minus .

3.3 Effect of temperature on interactions on wood substrate At one week, A. areolatum strains colonized the greatest amount of substrate when growing

at 25°C and strain 075 7011 typically colonized more substrate than 075 7013 at each

temperature (Fig. 5.3). All three fungal species grew slowly at 10°C, the chips incubated at

this temperature were never fully colonized and the contenders did not compete for substrate.

Leaving this set a further week did not result in competition for substrate. Two L. wingfieldii

strains (025 7011 and 025 7012) grew very poorly at 30°C and many replicates of these

strains did not grow at all. Several replicates of the A. areolatum strain 075 7013 did not

grow at all when pitted against O. minus strain 1275 at 15°C. These data from poorly

growing competition sets were eliminated from the analyses.

6

5 075 7011 4 075 7013

area (sq. cm.) 3

2

1 A. areolatum 0 10 15 20 25 30 Temperature (°C)

Figure 5.3. Surface area (±SE) of sterilized Pinus sylvestris woodchip colonized at 7 days by each of two Ontario Amylostereum areolatum strains at 10-30°C.

96

With the exception of one strain combination ( A. areolatum 075 7013 and L. wingfieldii 025 7011 at 20°C), in each of the tested and analyzed strain combinations the

ophiostomatoid strain colonized significantly more of the chip than the A. areolatum strain did at temperatures between 15-25°C (Table 5.3). There was a significant effect of temperature, L. wingfieldii strain, and the interaction of the two on the amount of substrate colonized by A. areolatum strain 075 7011 at temperatures between 15-25°C (temperature:

F2,36 = 3.4, p = 0.04, strain F 2,36 = 11.83, p <0.001, interaction: F 4,36 = 3.48, p = 0.02, log n transformed). The woodwasp symbiont was able to colonize more substrate at 20°C when competing with L. wingfieldii strain 025 7012 than it could with the other temperature and strain combinations (Fig. 5.4). There was a significant effect of L. wingfieldii strain but not

Table 5.3. T-test values and significance results of colonization of Pinus sylvestris woodchip surface area at 2 weeks by Amylostereum areolatum vs. three strains of each of two ophiostomatoid species (log n transformed) at temperatures between 15 and 30°C. In each case the ophiostomatoid species colonized more area than the A. areolatum strain. Amylostereum Ophiostomatoid species 15°C 20°C 25°C 30°C areolatum strain and strain number 075 7011 Leptographium 025 7010 -3.88** -12.63*** -4.84** - wingfieldii 025 7011 -17.91*** -13.65*** -37.08*** - 025 7012 -7.67*** -3.68** -14.87*** - 075 7013 025 7010 -3.08** -11.24*** -9.88*** - 025 7011 -2.36* -1.68 -7.79*** - 025 7012 -9.78*** -5.90*** -10.11*** - 075 7011 Ophiostoma 075 7007 -6.37*** -6.48*** -38.52*** -19.16*** minus 861 -5.25*** -14.92*** -28.36*** -43.18*** 1275 -5.50*** -9.10*** -24.93*** -43.29*** 075 7013 075 7007 -19.17*** -12.13*** -29.76*** -85.87*** 861 -12.28*** -5.62*** -16.27*** -44.44*** 1275 -4.25** -4.19** - -26.12*** * p ≤ 0.05, ** * p ≤ 0.01, * **p ≤ 0.001

temperature or interaction on A. areolatum strain 075 7013 (temperature: F 2,36 = 1.61, p=0.23, strain F 2,36 = 13.77, p <0.001, interaction: F 4,36 2.09, p = 0.10)(Fig. 5.4b). At temperatures between 15-30°C, there was a significant effect of temperature (but not of the O. minus or the interaction of the two) on the amount of substrate colonized by A. areolatum strain 075 7011

97

control 16 a) vs 025 7010 14 vs 025 7011 12 vs 025 7012 10

area (sq. cm) 8 6 4 2 0 A. areolatum 15 20 25 Temperature (°C) 16 b) control 14 vs 025 7010 12 vs 025 7011 10 vs 025 7012 area (sq. cm) 8 6 4 2 0 A. areolatum 15 20 25 Temperature (°C)

16 c) control 14 vs 075 7007 12 vs 1275 10 vs 861

area (sq. cm) 8 6 4 2 0 A. areolatum 15 20 25 30

Temperature (°C) d)

16 control 14 vs 075 7007 12 vs 1275 10 vs 861

area (sq. cm) 8 6 4 2

0 A. areolatum 15 20 25 30

Temperature (°C)

Figure 5.4. Amylostereum areolatum colony area (±SE) in the presence of contending species at different temperatures at 2 weeks on wood chip substrate a) 075 7011 growth in presence of Leptographium wingfieldii strains, b) 075 7013 growth in presence of L. wingfieldii strains, c) 075 7011 growth in presence of Ophiostoma minus strains, d) 075 7013 growth in presence of O. minus strain. 98

(temperature: F 3,48 = 40.19, p<0.001, strain F 2,48 = 2.19, p = 0.12, interaction: F 6,48 =1.50, p =

0.21). This strain was able to capture more substrate at 20°C than at other temperatures (Fig.

5.4c). There was a significant effect of temperature, O. minus strain and the interaction of the

two on area colonized by A. areolatum 075 7013 at temperatures between 20-30°

(temperature: F 2,36 = 54.09, p <0.001, strain F 2,36 = 7.59, p = 0.002, interaction: F 4,36 = 2.97, p

= 0.03) and results were similar for the 15-30° temperature range when the incomplete data set was included in the analysis (Fig. 5.4d).

3.4 Growth on pre-colonized substrate Neither A. areolatum strain grew on any of the live (un-autoclaved) strains of either L. wingfieldii or O. minus , but both strains grew on all of the autoclaved ophiostomatoid colonized substrate. One strain of L. wingfieldii (025 7010) did not grow on either strain of live A. areolatum , neither did it grow on one of the autoclaved A. areolatum strains (075

7011). Two strains of O. minus (1275 and 861) did not grow on live A. areolatum strain 075

7011.

4 Discussion The outcomes of experimental interactions on both artificial and wood substrates and at all temperatures were generally similar; that is, both L. wingfieldii and O. minus were able to colonize more of the substrate than A. areolatum and both prevented the woodwasp symbiont from colonizing this space. Therefore, if a beetle introduced one or both of these fungi near to an oviposition site made by S. noctilio , or vice versa, the growth of the wasp’s symbiont

would be limited and thus its larval nutritional resources as well. This could be especially

deleterious at early larval stages when wasp offspring are thought to depend on A. areolatum

as their sole food source (e.g. Madden & Coutts 1979). Given that S. noctilio size is related to 99

conditions in the tree that influence growth of its fungal symbiont (Madden 1981), interactions with other fungi that limit the growth of A. areolatum can be expected to reduce the size of adult woodwasps. As S. noctilio size is related to both fecundity and to fertility

(Madden 1974) factors negatively affecting insect size will ultimately reduce insect populations. Therefore, fungal interaction has the potential to exert a degree of biological control in this system.

Although the ophiostomatoid competitor typically occupied more substrate than A.

areolatum at temperatures between 15 and 25/30°C, the relative amount of substrate that A.

areolatum colonized was influenced by temperature in interactions with some of the

ophiostomatoid strains. From these results, it appears that sapwood temperatures of around

20°C (Fig. 5.4) could be expected to make the outcome of some inter-species interactions

more favorable, relatively, for the woodwasp. Within a tree, sapwood temperatures could be

expected to range up to 30°C (Stockfors 2000) and at temperatures of 30°C the A. areolatum

strains were able to colonize little substrate when competing with the O. minus strains (Fig.

5.4c,d); therefore these conditions would be more problematic for the woodwasp symbiont.

Trees infested by S. noctilio are thought to have higher sapwood temperatures than un-

infested trees (Jamieson 1957), so this may put added competition pressure on the woodwasp

symbiont after its initial establishment in the tree.

Though results on PDA and woodchips do not entirely correspond, it is evident that

some strains of the ophiostomatoid competitors ( L. wingfieldii (025 7012) and O. minus (075

7007)) facilitated the growth of one strain of the A. areolatum (075 7013) at certain

temperatures. Neutralistic interactions between fungi favouring one of the contenders is

described by (Rayner & Webber 1984) in their review of fungal interactions though they do

not discuss mediation of this type of interaction from a distance. In contrast, other

100

ophiostomatoid competitors inhibited A. areolatum growth on the wood substrate. This is better described in the literature where long-range inhibition can occur through a variety of volatile and non-volatile channels (Rayner & Webber 1984). These phenomena may influence the relative outcome of the interactions, for example L. wingfieldii strain 025 7012

tended to stimulate A. areolatum growth while L. wingfieldii strains 025 7010 and 025 7011

tended to inhibit it, and this fits with the trend for the A. areolatum strains to colonize more

substrate when in competition with this strain than they did with other L. wingfieldii strains

(Fig. 5.4a,b).

Amylostereum areolatum could not establish on substrate already occupied by active

colonies of either of the ophiostomatoid species. It could establish however on previously

colonized but dead colonies of the same competitor suggesting that this pattern was as a

result of antibiosis rather than competition for nutrients. Given this outcome, we could expect

that offspring from S. noctilio ovipositing in sapwood already colonized by one of these

ophiostomatoid fungi would not succeed. There was inter-strain variability in the ability of

each ophiostomatoid species to establish on A. areolatum colonized substrate, so only some

strains would have the potential to capture already colonized substrate from A. areolatum and

potentially influence the woodwasp larval development by arriving after it. That half of the

strains were not able to establish in one of the two A. areolatum strains is surprising in light

of King’s (1966) comment that the woodwasp symbiont was always overtaken by

competitors; however, it illustrates the importance of examining inter-strain differences when

evaluating fungal interactions.

It was not unexpected that the ophiostomatoid contenders were able to colonize a

large proportion of the available substrate, as these pathogens are typically the primary

colonists of a dying or recently-dead trees (Kirisits 2004) . First arrivers are characteristically

101

fast-growing and are effective at exploiting a new resource (Boddy 2000). Considering successional progression, it is somewhat surprising that A. areolatum was not more effective at secondary resource capture, whereas some of the ophiostomatoid strains were.

Amylostereum areolatum is a wood-decaying species and could be expected to arrive later in fungal succession of wood substrates (Kirisits 2004). Primary colonizers are typically easily replaced fungal species (Rayner & Webber 1984). However, antibiosis from living fungal competitors can affect succession (Stephen et al. 1993) and ophiostomatoid species have variable saprotrophic abilities (Kirisits 2004).

The degree to which these ophiostomatoid species are an issue for S. noctilio’s

development will depend on a number of factors including the timing of entry of the insects

(and therefore the fungal associates), the frequency with which they are introduced into the

tree, and the role that A. areolatum plays in nutrition during later larval instars of the insect.

Tomicus piniperda , the bark beetle expected to be the most likely vector of these tested fungi,

is more likely to enter the tree in the late winter or early spring after S. noctilio colonization, rather than prior to it (Chapter 2), so the two ophiostomatoid species tested are more likely to be introduced later. Therefore offspring of S. noctilio that oviposit late in the summer may be

more affected by fungal competition as their offspring would be at an earlier larval stage

when the fungus is likely to enter the tree. Not every individual of T. piniperda carries these

fungal associates (Jacobs et al. 2004), so the frequency with which these fungal interactions

could occur is unknown. Other phloem- and woodboring insects do enter the tree shortly

after S. noctilio (Chapter 2), and they could introduce similar ophiostomatoid species either

directly, or indirectly from the bark, and this could set up fungal interactions. The role of A.

areolatum in nutrition of later larval instars of S. noctilio is not well described, so entry of

102

these ophiostomatoid fungi at a later stage in insect development may have limited effect on the woodwasp.

The interactions I observed followed general patterns, however, there were differences between strains. In particular, one of the A. areolatum strains (075 7011) tended to colonize more space than the other strain (Table 5.2, Fig. 5.2), it showed different responses to temperature and contending strain (Fig. 5.4), and was more resistant to secondary resource capture. Therefore, offspring of woodwasps carrying strain 075 7011 may benefit from their symbiont having a greater ability to procure and maintain nutritional resources. There were also inter-strain differences in the ophiostomatoid competitors; one strain of L. wingfieldii (025 7012) tended to facilitate A. areolatum growth at times, allowed

relatively more A. areolatum colonization and could not establish on living colonies of A.

areolatum in contrast to the other two L. wingfieldii strains tested. These findings illustrate

the importance of using more than one strain of a fungal species in interaction experiments,

rather than making conclusions based on one strain alone.

Though the overall results were similar, there were some differences in results on the

different media used in these experiments. For example, O. minus strain 861 tended to enhance the growth of A. areolatum when tested on PDA, but inhibited it on the wood

substrate. Differences in facilitation and inhibition tendencies between some of the strains on

different media could affect results of relative substrate capture. Depending on the goals of

the experiment, one must be cautious about using the results on PDA to predict patterns in

this system.

The phytotoxic mucus that S. noctilio deposits with its fungus has direct effects on its fungal symbiont and in addition to releasing A. areolatum from the wax packet that surrounds it while stored in the mycangia, it is thought to stimulate fungal growth (Boros

103

1968; Titze & Turnbull 1970). We could expect then that this mucus may have some effect on the amount of substrate that the fungus is able to colonize and therefore may influence the outcomes of the interactions between competing fungi under natural conditions. The mucus is very difficult to work with and to incorporate into interaction experiments; however, this would be an important step to further research in this area.

Wood moisture is another key factor affecting the interactions between A. areolatum and ophiostomatoid species. The woodwasp symbiont grows more quickly in wood having less than 70% moisture content (Coutts & Dolezal 1965), while L. wingfieldii grows more quickly in wood between 120-150% moisture (Hironori et al. 2003). At lower wood moisture, the S. noctilio symbiont may fare better in competition against ophiostomatoid fungi.

My work is the first to show quantitative evidence of the outcomes of interactions

between the S. noctilio symbiont, A. areolatum , and species of beetle-vectored fungi .

Amylostereum areolatum was clearly a poor competitor against L. wingfieldii and O. minus and though it fared better at certain temperatures it was consistently outcompeted by these fungi and excluded from substrate already colonized by them regardless of substrate or temperature conditions. This will have reproductive consequences for the woodwasp if it is vying for space with phloem- and woodboring beetles in the tree and could ultimately affect its population dynamics as it establishes in North American forests.

A revised version of this paper has been accepted by The Canadian Entomologist

104

Chapter 6

Synthesis and general discussion

In this project, I set out to describe the community of phloem- and woodboring insects that inhabit pines with the woodwasp Sirex noctilio , to investigate for evidence of potential

effects of these insects on woodwasp traits, and to explore possible mechanisms of

interactions between the wasp and other subcortical insects.

Although the woodwasp was sometimes alone in a host tree, S. noctilio often shared trees with phloem- and woodboring beetles, most commonly Tomicus piniperda , Pissodes nemorensis , Ips grandicollis and Monochamus carolinensis (Chapter 4). The presence of resin drips and beads on the trees indicated that the woodwasp was reproductively active between mid-July and late August (Chapter 2). This is concurrent with three of the common co-habitants, P. nemorensis (Atkinson et al. 1988), I. grandicollis (Ayres et al. 2001) and M.

carolinensis (Walsh & Linit 1985; Yanega 1996), all of whom have extended or bi-modal flight seasons and can be expected to enter the tree around the same time as the woodwasp.

In contrast, Tomicus piniperda is primarily reproductively active in the early spring

(Kennedy & McCullough 2002) so is more likely to enter the tree the following spring. This fits with observations of evidence of insect activity; longhorn beetles arrived shortly after the woodwasp and bark beetles most commonly arrived the following spring (Chapter 2).

There was no evidence that the woodwasp was excluded from any area of the tree in the presence of beetles, and within-tree distributions of both the woodwasp and of the common beetles were generally similar in this study to that described by other authors

(Chapter 4). Although the woodwasp and the beetles tended to have differing abundance distributions through the stem of the tree, these distributions overlapped and wasps and

105

beetles co-habited sections of the tree to some extent. This overlap in distribution provides opportunity for interactions to occur between the woodwasp and beetles, and my findings demonstrate that they do; S. noctilio was less abundant, though larger, in trees where beetles

were present in comparison to those where they were infrequent or absent (Chapter 4). The

two potential mechanisms of interactions between the wasp and other subcortical that I

examined in this thesis can help to explain these patterns.

Interactions between the woodwasp’s symbiont Amylostereum areolatum and

ophiostomatoid fungi carried by the beetles could have affected the wasp’s abundance in

beetle-positive trees by impeding A. areolatum growth and subsequent offspring

development (Madden 1981). In the laboratory trials, the final outcomes of interactions

between A. areolatum and the two ophiostomatoid fungi on both substrates and at most temperatures were similar; both Leptographium wingfieldii and Ophiostoma minus were able

to colonize more of the substrate than A. areolatum and both prevented the woodwasp symbiont from colonizing this space (Chapter 5). This type of interaction is most likely to occur when beetles colonizing the tree around the same time as the woodwasp introduce these fungi. Since wasp egg eclosion usually occurs within 16-28 days (Morgan & Stewart

1966), it is within this window of time that we could expect that the inhibition of A. areolatum growth by ophiostomatoid fungi would hinder the hatching of woodwasp eggs.

For example M. carolinensis is thought to be associated with O. minus (Wingfield 1987) and

enters the tree shortly after the woodwasp. Relationships between ophiostomatoid fungi and

phloem- and woodboring beetles are not fully described (Six 2003) so we could expect that

other beetles with similar reproductive periods to the wasp could carry these fungi as well.

Alternatively, if individual S. noctilio oviposit later in the season (Chapter 2) and eggs do not

eclose before temperatures decrease below the wasp’s developmental threshold, woodwasp

106

egg eclosion could also be affected by fungi vectored by other insects, such as T. piniperda ,

that enter the tree early in the following spring. Although I only tested two ophiostomatoid

fungi in the experiments, species in this fungal group are typically the primary colonists

(Kirisits 2004) and first arrivers are characteristically fast-growing and effective at exploiting

a new resource (Boddy 2000), so we could expect that other ophiostomatoid species and

strains will also outcompete A. areolatum .

Another likely reason for the differences in S. noctilio abundance between beetle-

positive and beetle-negative trees found in Chapter 4 is that the woodwasp avoids ovipositing

in ophiostomatoid-colonized trees or areas of trees habited by beetles. Sirex noctilio detected

the presence of L. wingfieldii ; there were fewer S. noctilio drill scars in sections of trees or

bolts inoculated with L. wingfieldii than those not inoculated with the fungus, and probing

and drilling activity followed a similar pattern (Chapter 3). This finding fits with my

observations that there were rarely symptoms of the wasp activity in parts of the tree with

evidence of prior phloem- and woodborer activity (Chapter 2). The woodwasp’s apparent

lack of ability to detect O. minus suggests that this mechanism of interaction may not be common (Chapter 3). However, that the presence of adventitious blue stain fungi (those not intentionally introduced) decreased the wasp’s drilling activity suggests that it may not be rare either. Though the wasp’s ability to detect ophiostomatoid fungi may indeed be inconsistent between fungal species, it is also plausible that cues from O. minus were masked by tree volatiles (as discussed in Chapter 3) and wasp avoidance of ophiostomatoid-colonized trees could be more prevalent than suggested by this study. Further research in this area is warranted.

The trend to larger woodwasps in beetle-positive trees compared to beetle-negative ones does not fit with the outcomes of the fungal interaction trials. I expected wasps in

107

beetle-positive trees to be smaller as a result of the limited food resources (Madden 1981) resulting from the fungal competition evident in Chapter 5. This finding could be a result of the spatial distribution of insects within the tree though, as the tree will not be evenly colonized with these fungi at the time the woodwasp is developing. Therefore, competition between the two fungal species may not occur for all wasp offspring, or may not occur when

A. areolatum is critical for the wasp’s development. As I suggest in Chapter 4, the introduction of the ophiostomatoid species into the tree could enhance growth of the woodwasp symbiont by causing the drying sapwood throughout the tree (Chow & Obermajer

2007) thus improving conditions for A. areolatum growth.

Though the effects of beetle-presence on the woodwasp evident in Chapter 4 (reduced wasp abundance but larger body size) would have contrasting effects on woodwasp populations, the presence of beetles would still have a considerable overall negative effect on woodwasp population dynamics. Using the average abundance and prothorax width of female wasps from Chapter 4, the average number of woodwasp eggs produced per tree in beetle-negative and beetle-positive trees can be calculated with Madden’s (1974) equation

(see Chapter 1). Based on these calculations, the average number of eggs produced per tree in beetle-negative trees exceeded that of beetle-positive ones by 810 eggs per tree (avg. beetle negative = 3111 eggs, beetle-positive = 2301). Although this would be offset to some degree as smaller wasps tend to lay a smaller proportion of their eggs (Madden 1974), over time beetles will clearly have a considerable effect on woodwasp population dynamics.

Tomicus piniperda is the most abundant co-habitant with the woodwasp, however, one must ask if it is the most important since there are temporal and spatial differences that may limit its interactions with the woodwasp. It appears that T. piniperda is more likely to enter the tree the following spring after S. noctilio colonization (Chapter 2). At this time,

108

most egg eclosion should have occurred so larval or pupal mortality should be the most plausible mechanism of interaction but there was no evidence of this (Chapter 4). Tomicus piniperda favours the lower sections of the tree, that region with thick, furrowed bark

(Chapter 4 and Schroeder 1997), and the woodwasp more typically favours the mid-stem, thinner barked areas (Chapter 4). So, although there is overlap in distribution between these two species, there may be limited interaction because of brood habitat choices. Other co- habitants may be more important for the wasp. For example, M. carolinensis oviposits in the tree shortly after the woodwasp and tends to oviposit in similar areas of the tree as S. noctilio does (Chapter 2). This beetle could introduce fungi close to the wasp, thereby initiating fungal-fungal competition. Monochamus carolinensis is thought to be associated with O. minus (Wingfield 1987) and could be expected to vector other species as well since the fungal associates of longhorn beetles are not well described so the list of associates is likely to be incomplete. In addition, the beetle excavates a relatively large oviposition pit which could be expected to allow the introduction of fungal spores from the bark. Because this beetle reproduces at a similar time to the wasp, the wasp may also avoid following M. carolinensis into areas of the tree where the beetle has oviposited. Similar relationships could exist with I. grandicollis and P. nemorensis as both insects can reproduce throughout the summer, and thus enter the tree around the same time as the woodwasp.

Site and tree conditions may be an important factor in these interactions. One site,

E6B, was planted on a sandy, well drained site and there were a number of pests and pathogens evident in the stand. There was also a great deal of woodwasp activity in this site and trees died very quickly after attack (Chapter 2). From this, we should expect a large number of woodwasp offspring to emerge from the infested trees, however average woodwasp emergence per tree from this site was well below the mean of all sites (mean

109

emergence per tree E6B = 70, all sites = 131 derived from Chapter 4 data). In this site, there were a number of Monochamus sp. entering the trees shortly after the wasp (Chapter 2), perhaps because of the rapidly declining condition of the tree. We could expect then that these tree conditions would contribute to the occurrence of fungal competition as these trees died very quickly and therefore were not able to defend themselves against new pests or pathogens. This would allow immediate growth of beetle-vectored fungi and expedite competition between species of fungi. Therefore the role of phloem- and woodboring beetles in control of S. noctilio populations could be variable across the landscape.

My research has demonstrated that Sirex noctilio often shares trees with other

phloem- and woodborers in its introduced range in North America and that these beetles can

help to control woodwasp populations in this region, at least in some sites. These co-habiting

beetles, along with native parasitoids and predators, and climate factors will function together

to help limit the growth of S. noctilio populations. Knowledge of the function of Deladenus

siricidicola as a natural enemy of the wasp in this new introduced range, as well as factors

such as overwinter mortality will help to improve our predictions of the woodwasp’s

significance as a pest. Evidence to date suggests that, in its current range in northeastern

North America, S. noctilio is more likely to function in similar manner to that in its native

range, as a secondary pest, than it does elsewhere in its introduced range. It is conceivable

that S. noctilio will not require management other than good silvicultural management of

Pinus spp. stands.

110

References

Adams, A.S., Currie, C.R., Cardoza, Y., Klepzig, K.D. and Raffa, K.F. 2009. Effects of symbiotic bacteria and tree chemistry on the growth and reproduction of bark beetle fungal symbionts. Canadian Journal of Forest Research. 39: 1133-1147.

Amezaga, I. and Garbisu, C. 2000. Effect of intraspecific competition on progeny production of Tomicus piniperda (Coleoptera; Scolytidae). Environmental Entomology 29: 1011-1017.

Andersen, F.S. 1961. Effect of density on sex ratio. Oikos 12: 1-16.

Atkinson, T.H., Foltz, J.L., and Connor, M.D. 1988. Bionomics of Pissodes nemorensis Germar (Coleoptera: Curculionidae) in northern Florida. Annals of the Entomological Society of America 81: 255-261.

Ayres, M.P., Wilkens, R.T., Ruel, J.J., and Vallery, E. 2000. Fungal relationships and the nitrogen budget of phloem-feeding bark beetles (Coleoptera: Scolytidae). Ecology 81: 2198- 2210.

Ayres, B.D., Ayres, M.P., Abrahamson, M.D., and Teale, S.A. 2001. Resource partitioning and overlap in three sympatric species of Ips bark beetle (Coleoptera: Scolytidae). Oecologia 128: 443-453.

Barras, S.J. 1970. Antagonism between Dendroctonus frontalis and the fungus Ceratocystis minor . Annals of the Entomological Society of America 63: 1187-1190.

Barras, S.J. 1973. Reduction of progeny and development in the southern pine beetle following removal of symbiotic fungi. Canadian Entomologist 105: 1295-1299.

Bedding, R.A. and Akhurst, R.J. 1974. Use of the nematode Deladenus siricidicola in the biolgical control of Sirex noctilio in Australia. Journal of the Australian Entomological Society 13: 129-135.

Bedding, R.A. and Akhurst, R.J. 1978. Geographical distribution and host preferences of Deladenus species (Nematoda: Neotylenchidae) parasitic in siricid woodwasps and associated hymenopterous parasitoids. Nematologica 24: 286-294.

Bedding, R.A. and Iede, E.T. 2005. Application of Beddingia siricidicola for Sirex Woodwasp Control. In Grewel, P.S.,Ehlers, R.-U., Shapiro-Ilan, D.I. (Eds) as Biocontrol Agents. CAB International, pp. 385-399.

Ben Jamaa, M.L., Lieutier, F., Yart, A., Jerraya, A., and Khouja, M.L. 2007. The virulence of phytopathogenic fungi associated with the bark beetles Tomicus piniperda and Orthotomicus erosus in Tunisia. 37: 51-63.

Bergeron, M.J., Leal. I., Foord, B., Ross, G., Davis, C., Slippers, B., de Groot,P. and Hamelin, R. 2010. Putative origin of clonal lineages of Amylostereum areolatum , the fungal symbiont associated with Sirex noctilio, in eastern Canada. Fungal Genetics and Biology 92: 1138. 111

Bleiker, K.P. and Six, D.L. 2007. Dietary benefits of fungal associates to an eruptive herbivore: potential implications of multiple associates on host population dynamics. Environmental Entomology 36: 1384-1396.

Boddy, L. 2000. Interspecific combative interactions between wood-decaying basidiomycetes. FEMS Microbiology Ecology 31: 185-194.

Boros, C.B. 1968. The relationship between the woodwasp, Sirex noctilio F. and the wood- rot fungus Amylostereum sp. M. Ag. Sc. Thesis University of Adelaide.

Bridges, J.R. 1983. Mycangial fungi of Dendroctonus frontalis and their relationship to beetle population trends. Environmental Entomology 12: 858-861.

Bridges, J.R. and Moser, J.C. 1983. Role of two phoretic mites in transmission of bluestain fungus, Ceratocystis minor. Ecological Entomology 8: 9-12.

Bright, D.E. Jr. 1976. The insects and arachnids of Canada Part 2: The bark beetles of Canada and Alaska Coleoptera:Scolytidae. Canada Department of Agriculture, Ottawa, Ontario.

Bruzzone, O.A., Villacide, J.M., Bernstein, C., and Corley, J.C. 2009. Flight variability in the woodwasp Sirex noctilio (Hymenoptera: Siricidae): and analysis of flight data using wavelets. Journal of Experimental Biology 212: 731-737.

Büchner, P. 1965. Siricids. In Endosymbiosis of with Plant Microorganisms. John Wiley , . pp. 83-92.

Büchner, P. 1928. Holznahrung und Symbiose. J. Springer, Berlin.

Canadian Food Inspection Agency [updated 2010] Summary of plant quarantine pest and disease situations in Canada 2006-2008 Sirex wasp [online]. Available from: http:www.inspection.gc.ca/ english/plaveg/pestrava/surv/obje.shtml [accessed 17 June 2010].

Cameron, E.A. 1965. The Siricinae (Hymenoptera: Siricidae) and their parasites. C.I.B.C. Technical Bulletin. Rep. No. 5.

Cardoza, Y.J., Klepzig, K.D, and Raffa, K.F. 2006. Bacteria in oral secretions of an endophytic insect inhibit antagonistic fungi. Ecological Entomology. 31: 6.6-645.

Carnegie, A.J., Matsuki, M., Haugen, D.A., Hurley, B.P., Ahumada, R., Klasmer, P., Sun, J., and Iede, E.T. 2006. Predicting the potential distribution of Sirex noctilio (Hymenoptera: Siricidae), a significant exotic pest of Pinus plantations. Annals of Forest Science 63: 119- 128.

Cartwright, K.St.G. 1929. Notes on the fungus associated with Sirex cyaneus . Annals of Applied Biology 16: 182-187.

Cartwright, K.S.G. 1938. A further note on fungus association in the Siricidae. Annals of Applied Biology 25: 430-432.

112

Cheniclet, C. 1987. Effects of wounding and fungus inoculation on producing systems of maritime pine. Journal of Experimental Botany 38: 1557-1572.

Chow, S. and Obermajer, A. 2007. Moisture and blue stain distribution in mountain pine beetle infested lodgepole pine trees and industrial implications. Wood Science Technology 41: 3-16.

Ciesla, W.M. 2003. European woodwasp: A potential threat to North America's conifer forests. Journal of Forestry 101: 18-23.

Clark, A.F. 1927. The infestation of in Canterbury. Te Kura Ngahere 2:10-16.

Cooke, R.C. and Rayner , A.D.M. 1984. Ecology of Saprotrophic Fungi . Longman, London and New York .

Coutts, M.P. 1965. Sirex noctilio and the physiology of . Some studies of interactions between the insect, the fungus, and the tree in Tasmania. Forest Timber Bureau of Australia Bulletin 41.

Coutts, M.P. 1969a. The mechanism of pathogenicity of Sirex noctilio on Pinus radiata I. Effects of the symbiotic fungus Amylostereum sp. (Thelophoraceae). Australian Journal of Biological Sciences 22: 915-924.

Coutts, M.P. 1969b. The mechanism of pathogenicity of Sirex noctilio on Pinus radiata II. Effects of S. noctilio mucus. Australian Journal of Biological Sciences 22: 1153-1161.

Coutts, M.P. and Dolezal, J.E. 1965. Sirex noctilio , its associated fungus and some aspects of wood moisture content. Australian Forest Research 1: 3-13.

Coutts, M.P. and Dolezal, J.E. 1969. Emplacement of fungal spores by the woodwasp, Sirex noctilio , during oviposition. Forest Science 4: 412-416.

Dajoz, R. 2000. Insects and Forests: the Role and Diversity of Insects in the Forest Environment. Intercept Ltd., London, Paris.

de Groot, P., Nystrom, K., and Scarr, T. 2006. Discovery of Sirex noctilio (Hymenoptera: Siricidae) in Ontario, Canada. The Great Lakes Entomologist 39: 49-53.

Dodds, K.J., de Groot, P., and Orwig, D.A. 2010. The impact of Sirex noctilio in Pinus resinosa and Pinus sylvestris stands in New York and Ontario. Canadian Journal of Forest Research 40: 212-223.

Dodds, K.J., Graber, C., and Stephen, F.M. 2001. Facultative intraguild predation by larval cerambycidae (Coleoptera) on bark beetle larvae (Coleoptera: Scolytidae). Environmental Entomology 30: 17-22.

Dolezal, J.E. 1967. Some observations on behaviour of Sirex noctilio when used for artificial attack on Pinus radiata . Australian Forest Research 2: 26-30.

113

Downie, N.M. and Arnett, R.H. 1996. The Beetles of Northeastern North America, Vol. 2. The Sandhill Crane Press, Gainesville, FL.

Environment Canada. 2010. National Climate Data and Information Archive. [online] Available from: http://www.climate.weatheroffice.gc.ca/climateData/canada_e.html [Accessed April 2010].

Flamm, R.O., Wagner, T.L., Cook, S.P., Pulley, P.E., Coulson, R.N., and AcArdle, T.M. 1987. Host colonization by cohabiting Dendroctonus frontalis , Ips avulsus , and I. calligraphus (Coleoptera: Scolytidae). Environmental Entomology 16: 390-399.

Fong, L.K. and Crowden, R.K. 1973. Physiological effects of mucus from the wood wasp, Sirex noctilio F., on the foliage of Pinus radiata D. don. Australian Journal of Biological Sciences 2: 365-378.

Francke-Grosman, H. 1939. On the symbiosis of woodwasps (Siricinae) with fungi (Uber das Zusammenleben von Holzwespen (Siricinae) mit Pilzen). Zeitschrift Fuèr Angewandte Entomologie 25: 647-680. [Rutland., D translation]

Gaut, I.P.C. 1969. Identity of the fungal symbiont of Sirex noctilio . Australian Journal of Biological Science 4: 905-914.

Gebeyehu, S. and Wingfield, M.J. 2003. Pine weevil Pissodes nemorensis : threat to South African pine plantations and options for control. South African Journal of Science 99: 531- 536.

Gilbertson, R.L. 1984. Relationships between insects and wood-rotting basidiomycetes. In Wheeler, Q., Blackwell, M.(Eds) Fungus-insect Relationships, Perspectives in Ecology and Evolution. Columbia University Press, New York.

Gilmour, J.W. 1965. The life cycle of the fungal symbiont of Sirex noctilio . New Zealand Journal of Forestry 10: 80-89.

Goldhammer, D.S., Stephen, F.M., and Paine, T.D. 1990. The effect of the fungi Ceratocystis minor (Hedgecock) Hunt, Ceratocystis minor (Hedgecock) Hunt var barrasii Taylor, and SJB 122 on reproduction of the southern pine beetle, Dendroctonus frontalis Zimmerman (Coleoptera; Scolytidae). Canadian Entomologist 122: 407-418.

Hall, M.J. 1968. A survey of siricid attack on radiata pine in Europe. Australian Forestry 32: 155-162.

Hanson, H.S. 1939. Ecological notes on the sirex woodwasps and their parasites. Bulletin of Entomological Research 30: 27-65.

Hanssen, H.P. 1993. Volatile metabolites produced by species of Ophiostoma and Ceratocystis . In Wheeler, Q., Blackwell, M. (Eds) Ceratocystis and Ophiostoma , Ecology and Pathogenicity. APS Press, St. Paul, Minnesota.

Haugen, D.A. Sirex noctilio. 2006. Provisional Record, Database of the Exotic Forest Pest Information System for North America, North American Forest Commission. Available 114

from: www.spfnic.fs.fed.us/exfor/

Haugen, D.A., Bedding, R.A., Underdown, M.G., and Neumann, F.G. 1990. National strategy for control of Sirex noctilio in Australia. Australian Forest Grower 13: Special Liftout Section 13.

Hausner, G., Iranpour, M., Kim, J.J., Breuil, C., Davis, C.N., Gibb, E.A., Reid, J., Loewen, P.C., and Hopkin, A.A. 2005. Fungi vectored by the introduced bark beetle Tomicus piniperda in Ontario, Canada, and comments on the taxonomy of Leptographium lundbergii , Leptographium terebrantis , Leptographium truncatum , and Leptographium wingfieldii . Canadian Journal of Botany 83: 1222-1237.

Hironori, T., Toshiyuki, K., Hayato, M., and Shuichi, D. 2003. Effects of temperature, water activity and oxygen concentration of extension of blue-stain caused by Leptographium wingfieldii in Japanese red pine ( Pinus densiflora ) wood. Journal of the Japan Wood Research Society 49: 446-451.

Hoebeke, E.R., Haugen, D.A., and Haack, R.A. 2005. Sirex noctilio : discovery of a palearctic siricid woodwasp in New York. Newsletter of the Entomological Society 50: 24- 25.

Hofstetter, R.W., Cronin, J.T., Klepzig, K.D., Moser, J.C., and Ayres, M.P. 2006. Antagonisms, mutualisms and commensalisms affect outbreak dynamics of the southern pine beetle. Oecologia 147: 679-691.

Hofstetter, R.W., Dempsey, T.D., Klepzig, K.D., and Ayres, M.P. 2007. Temperature- dependant effects on mutualistic, antagonistic, and commensalistic interactions among insects, fungi and mites. Community Ecology 8: 47-56.

Hofstetter, R.W., Mahfouz, J.B., Klepzig, K.D., and Ayres, M.P. 2005. Effects of tree phytochemistry on the interactions among endophloedic fungi associated with the southern pine beetle. Journal of Chemical Ecology 31: 539-560.

Holmer, L. and Stenlid, J. 1993. The importance of inoculum size for the competitive ability of wood decomposing fungi. FEMS Microbiology Ecology 12: 169-176.

Hui, Y. and Xue-Song, D. 1999. Impacts of Tomicus minor on distribution and reproduction of Tomicus piniperda (Col., Scolytidae) on the trunk of the living Pinus yunnanensis trees. Journal of Applied Entomology 123: 329-333.

Hurley, B.P., Slippers, B., Croft, P.K., Hatting, H.J., van der Linde, M., Morris, A.R., Dyer, C., and Wingfield, M.J. 2008. Factors influencing parasitism of Sirex noctilio (Hymenoptera: Siricidae) by the nematode Deladenus siricidcola (Nematoda: Neotylenchidae) in summer rainfall areas of . Biological Control 45: 450-459.

Hurley, B.P., Slippers, B., and Wingfield, M.J. 2007. A comparison of control results for the alien invasive woodwasp, Sirex noctilio , in the southern hemisphere. Agricultural and Forest Entomology 9: 159-171.

Iede, E.T., Penteado, S.R.C., and Schaitza, E.G. 1998. Sirex noctilio problem in Brazil: 115

detection, evaluation, and control. In Iede, E.T., Schaitza, E., Penteado, S., Reardon, R.C., Murphy, S.T. (Eds) Proceedings of a Conference: Training in the Control of Sirex noctilio by the use of Natural Enemies. USDA Forest Service, Morgantown WV.

Jackson, D.S. 1955. The Pinus radiata /Sirex noctilio relationship at Rotoehu Forest. New Zealand Journal of Forestry 7: 26-40.

Jacobs, K., Bergdahl, D.R., Wingfield, M.J., Halik, S., Seifert, K.A., Bright, D.E., and Wingfield, B.D. 2004. Leptographium wingfieldii introduced into North America and found associated with exotic Tomicus piniperda and native bark beetles. Mycological Research 108: 411-418.

Jamieson, J.P. 1957. Some observations on the relation of the internal temperature of P. radiata to infestation by Sirex . New Zealand Journal of Forestry 7: 32-34.

Jankowiak, R. 2006. Fungi associated with Tomicus piniperda in Poland and assessment of their virulence using Scots pine seedlings. Annals of Forest Science 63: 801-808.

Jankowiak, R. and Kola řik, M. 2010. Diversity and pathogenicity of ophiostomoid fungi associated with Tetropium species colonizing Picea albies in Poland. Folia Microbiologica. 55: 145-154.

Jankowiak, R. and Rossa, R. 2007. Filamentous fungi associated with Monochamus galloprovincialis and Acanthocinus aedilis (Coleoptera: Cerambycidae) in scots pine. Polish Botanical Journal 52: 143-149.

Kennedy, A.A. and McCullough, D.G. 2002. Phenology of the Larger European Pine Shoot Beetle Tomicus piniperda (L.) (Coleoptera: Scolytidae) in relation to native bark beetles and natural enemies in pine stands. Environmental Entomology 31: 261-272.

King, J.M. 1966. Some aspects of the biology of the fungal symbiont of Sirex noctilio (F.). Australian Journal of Botany 14: 25-30.

Kirisits, T. 2004. Fungal associates of European bark beetles with special emphasis on the Ophiostomatoid fungi. Lieutier, F.,Day, K.R.,Battisti, A.,Grégoire, J.-C., Evans, H.F. (Eds) In Bark and Wood Boring Insects in Living Trees in Europe, a Synthesis. Kluwer Academic Publishers, Dordecht/Boston/London.

Klasmer, P., Fritz, G., Corely, J., and Botto, E. 1998. Current status of research on Sirex noctilio F. in the Andean-Patagonian region in . In Iede, E.T., Schaitza, E., Penteado, S., Reardon, R.C., Murphy, S.T. (Eds) Proceedings of a Conference: Training in the Control of Sirex noctilio by the use of Natural Enemies. USDA Forest Service, Morgantown WV.

Klepzig, K.D. 1998. Competition between a biological control fungus, Ophiostoma piliferum , and symbionts of the southern pine beetle. Mycologia 90: 69-75.

Klepzig, K.D., Flores-Otero, J., Hofstetter, R.W., and Ayres, M.P. 2004. Effects of available water on growth and competition of southern pine beetle associated fungi. Mycological Research 108: 183-188. 116

Klepzig, K.D. and Wilkens, R.T. 1997. Competitive interactions among symbiotic fungi of the southern pine beetle. Applied and environmental microbiology 63: 621-627.

Kopper, B.J., Klepzig, K.D., and Raffa, K.F. 2004. Components of antagonism and mutualism in Ips pini -fungal interactions: relationship to a life history of colonizing highly stressed trees. Environmental Entomology 33: 28-34.

Kukor, J.J. and Martin, M.M. 1983. Acquisition of digestive enzymes by siricid woodwasps from their fungal symbiont. Science 220: 1161-1163.

Lam, K., Tsang, M., Labire, A., Gries, R., and Gries, G. 2010. Semiochemical-mediated oviposition avoidance by female house flies, Musca domestica , on animal feces colonized with harmful bacteria. Journal of Chemical Ecology 36: 141-147.

Lieutier, F., Yart, A., Ye, H., Sauvard, D., and Gallois, V. 2004. Variations in growth and virulence of Leptographium wingfieldii Morelet, a fungus associated with the bark beetle Tomicus piniperda L. Annals of Forest Science 61: 45-53.

Long, S.J., Williams, D.W., and Hajek, A.E. 2009. Sirex species (Hymentoptera: Siricidae) and their parasitoids in Pinus sylvestris in eastern North America. Canadian Entomologist 141: 153-157.

Madden, J.L. 1968. Physiological aspects of host tree favourability for the woodwasp, Sirex noctilio F. Proceedings of the Ecological Society of Australia 3: 147-149.

Madden, J.L. 1974. Oviposition behaviour of the woodwasp, Sirex noctilio F. Australian Journal of Zoology 3: 341-351.

Madden, J.L. 1977. Physiological reactions of Pinus radiata to attack by woodwasp, Sirex noctilio F. (Hymenoptera: Siricidae). Bulletin of Entomological Research 3: 405-426.

Madden, J.L. 1981. Egg and larval development in the woodwasp, Sirex noctilio F. Australian Journal of Zoology 29: 493-506.

Madden, J.L. 1982. Avian predation of the woodwasp, Sirex noctilio F., and its parasitoid complex in Tasmania. Australian Wildlife Research 9: 135-144.

Madden, J.L. 1988. Sirex in Australasia. In Berryman, A.A. (Ed) Dynamics of Forest Insect Populations: Patterns, Causes, Implications. Plenum Press, New York & London.

Madden, J.L. and Coutts, M.P. 1979. The role of fungi in the biology and ecology of wood wasps (Hymenoptera: Siricidae). In Batra, L.R. (Ed) Insect-fungus Symbiosis. Allanheld, Osmun & Co, New Jersey.

Madden, J.L. and Irvine, C.J. 1971. The use of lure trees for the detection of Sirex noctilio in the field. Australian Forestry 3: 164-166.

Manninen, A.-M., Tarhanen, S., Vuorinen, M., and Kainulainen, P. 2002. Comparing the variation of needle and wood terpenoids in Scots pine provenances. Journal of Chemical Ecology 28: 211-228. 117

Morgan, F.D. 1968. Bionomics of Siricidae. Annual Review of Entomology 13: 239-256.

Morgan, F.D. and Stewart, N.C. 1966a. The biology and behaviour of the woodwasp Sirex noctilio (F.) in New Zealand. Transactions of the Royal Society of New Zealand 7: 195-204.

Morgan, F.D. and Stewart, N.C. 1966b. The effect of persuasoria (L.) on a population of Sirex noctilio (F.). Transactions of the Royal Society of New Zealand 8: 31-38.

Morgan, R.E., de Groot, P., and Smith, S.M. 2004. Susceptibility of pine plantations to attack by the pine shoot beetle ( Tomicus piniperda ) in southern Ontario. Canadian Journal of Forest Research 34: 2528-2540.

Murphy, S.T. 1998. Indigenous Siricid spp. parasitoid communities and principal biological control agents of Sirex noctilio in Autralasia: a review. In Iede, E.T., Schaitza, E., Penteado, S., Reardon, R.C., Murphy, S.T. (Eds) Proceedings of a Conference: Training in the Control of Sirex noctilio by the use of Natural Enemies. USDA Forest Service, Morgantown WV.

Neumann, F.G. and Minko, G. 1981. The sirex wood wasp in Australian radiata pine plantations . Australian Forestry 44: 46-63.

Neumann, F.G., Harris, J.A., Kassaby, F.Y., and Minko, G. 1982. An improved technique for early detection and control of the Sirex wood wasp in radiata pine plantations. Australian Forestry 45: 117-124.

Neumann, F.G., Morey, J.L., and McKimm, R.J. 1987. The sirex wasp in Victoria. epartment of Conservation, Forests and Lands, Victoria, Bulletin No. 29.

Nevill, R.J. and Alexander, S.A. 1992. Transmission of Leptographium procerum to eastern white pine by Hylobius pales and Pissodes nemorensis (Coleoptera: Curculionidae). Plant Disease 76: 307-310.

Nuttall, M.J. 1989. Sirex noctilio F., sirex wood wasp (Hymenoptera: Siricidae). In P. J. Cameron, P.J., Hill, R.L., Bain, J., Thomas, W.P. (Eds), A Review of Biological Control of Pests and Weeds in New Zealand 1974-1987. Technical Communication No. 10, C.A.B. International & DSIR, Wallingford, UK.

Paine, T.D., Raffa, K.F., and Harrington, T.C. 1997. Interactions among Scolytid bark beetles, their associated fungi, and live host conifers. Annual Review of Entomology 42: 179-206.

Paine, T.K., Birch, M.C., and Svira, P. 1981. Niche breadth and resource partitioning by four sympatric species of bark beetles (Coleoptera: Scolytidae). Oecologia 48: 1-6.

Parkin, W.A. 1941. Symbiosis in larval Siricidae (Hymenoptera). Nature 147: 329.

Parkin, E.A. 1942. Symbiosis and siricid woodwasps. Annals of Applied Biology 29: 268- 274.

Passoa, S., Cavey, J. [last updated June 2006]. Key to help screen Tomicus piniperda (L.) USDA Forest Service. [online] Available at: 118

http://www.barkbeetles.org/exotic/tmcspnpe.html [Accessed June 2008]

Penteado, S.R.C., de Oliveira, E.B., and Iede, E.T. 1998. Sampling methods for evaluating Sirex noctilio attack levels in stands and for monitoring the efficiency of natural enemies. In Iede, E.T., Schaitza, E., Penteado, S., Reardon, R.C., Murphy, S.T. (Eds) Proceedings of a Conference: Training in the Control of Sirex noctilio by the use of Natural Enemies. USDA Forest Service, Morgantown WV.

Raffa, K.F. and Smalley, E.B. 1988. Seasonal and long-term responses of host trees to microbial associates of the pine engraver Ips pini . Canadian Journal of Forest Research 18: 1624-1634.

Rankin, L.J. and Borden, J.H. 1991. Competitive interactions between the mountain pine beetle and the pine engraver in lodgepole pine. Canadian Journal of Forest Research 21: 1029-1036.

Rawlings, G.B. 1948. Recent observations on the Sirex noctilio population in Pinus radiata forest in New Zealand. New Zealand Journal of Forestry 5: 411-421.

Rawlings, G.B. and Wilson, N.M. 1949. Sirex noctilio as a beneficial and destructive insect to Pinus radiata in New Zealand. New Zealand Journal of Forest Science 6: 20-29.

Rayner, A.D.M. and Boddy, L. 1988. Fungal Decomposition of Wood: its Biology and Ecology. John Wiley and Sons, Chichester, New York.

Rayner, A.D.M. and Webber, J.F. 1984. Interspecific mycelial interactions - an overview. In Jennings, D.H., Rayner, A.D.M. (Eds) The ecology and Physiology of the Fungal Mycelium. Cambridge University Press, Cambridge, United Kingdom.

Rice, A.V., Thormann, M.N., and Langor, D.W. 2007. Mountain pine beetle-associated blue- stain fungi are differentially adapted to boreal temperatures. Forest Pathology 28: 113-123.

Röder, G., Rahier, M., and Naisbit 2007. Coping with an antagonist: the impact of a phytopathogenic fungus on the development and behaviour of two species of alpine leaf beetle. Oikos 116: 1514-1523.

Schenk, J.A. and Benjamin, D.M. 1969. Notes on the biology of Ips pini in central Wisconsin jack pine forests. Annals of the Entomological Society of America 62: 480-485.

Schiff, NM, Valley, SA, LaBonte, JR, and Smith, DR. 2006. Guide to the Siricid Woodwasps of North America. Morgantown, West Virginia, United States Department of Agriculture Forest Service Forest Health Technology Enterprise Team.

Schlyter, F. and Anderbrant, O.1993. Competition and niche separation between two bark beetles: existence and mechanisms. Oikos 68: 437-447.

Schoenweiss, D.F. 1981. The role of environmental stress in diseases of wood plants. Plant Disease 65: 308-314.

Schroeder, L.M. 1987. Attraction of the bark beetle Tomicus piniperda to Scots pine trees in 119

relation to tree vigor and attack density. Entomologia Experimentalis et Applicata 44: 53-58.

Schroeder, L.M. 1997. Oviposition behavior and reproductive success of the cerambycid Acanthocinus aedilis in the presence and absence of the bark beetle Tomicus piniperda . Entomologia Experimentalis et Applicata 82: 9-17.

Simpson, R.F. 1976. Bioassay of pine oil components as attractants for Sirex noctilio (Hymepoptera: siricidae) using electroantennogram techniques. Entomologia Experimentalis et Applicata 19: 11-18.

Six, D.L. 2003. Bark beetle-fungus symbiosis. In Bourtzis, K., Miller, T.A.(Eds) Insect Symbiosis. CRC Press, Boca Raton.

Six, D.L. and Paine, T.D. 1998. Effects of mycangial fungi and host tree species on progeny survival and emergence of Dendroctonus ponderosae (Coleoptera: Scolytidae). Environmental Entomology 27: 1393-1401.

Slippers, B., Wingfield, M.J., Coutinho, T.A., and Wingfield, B.D. 2001. Population structure and possible origin of Amylostereum areolatum in South Africa. Plant Pathology 50: 206-210.

Smalley, E.B., Raffa, K.F., Proctor, R.H., and Klepzig, K.D. 1993. Tree responses to infection by species of Ophiostoma and Ceratocystis . In Wheeler, Q., Blackwell, M. (Eds) Ceratocystis and Ophiostoma : Taxonomy, ecology, and pathogenicity. APS Press, St. Paul, Minnesota. pp. 207-217.

Spradbery, J.P. 1973. A comparative study of the phytotoxic effects of siricid woodwasps on conifers. Annals of Applied Biology 75: 309-320.

Spradbery, J.P. 1977. The oviposition biology of siricid woodwasps in Europe. Ecological Entomology 2: 225-230.

Spradbery, J.P. 1990. Predation of larval siricid woodwasps (Hymenoptera: Siricidae) by woodpeckers in Europe. The Entomologist 109: 67-71.

Spradbery, J.P. and Kirk, A.A. 1978. Aspects of the ecology of siricid woodwasps (Hymenoptera: Siricidae) in Europe, North Africa and Turkey with special reference to the biological control of Sirex noctilio F. in Australia. Bulletin of Entomological Research 68: 341-359.

Spradbery, J.P. and Kirk, A.A. 1981. Experimental studies on the responses of European siricid woodwasps to host trees. Annals of Applied Biology 98: 179-185.

Stephen, F.M., Berisford, C.W., Dahlsten, D.L., Fenn, P., and Moser, J.C. 1993. Invertebrate and microbial associates. In Schowalter, T.D., Filip, G.M. (Eds) Beetle-pathogen Interactions in Conifer Forests. Academic Press, London, San Diego.

Stockfors, J. 2000. Temperature variations and distribution of living cells within tree stems: implications for stem respiration modeling and scale-up. Tree Physiology 20: 1057-1062.

120

Talbot, P.H.B. 1977. The Sirex-Amylostereum-Pinus association. Annual Review of Phytopathology 15: 41-54.

Taylor, K.L. 1978. Evaluation of the insect parasitoids of Sirex noctilio (Hymenoptera: Siricidae) in Tasmania. Oecologia (Berl.) 32: 1-10.

Taylor, K.L. 1981. The Sirex woodwasp: ecology and control of an introduced forest insect. In Kitching, R.L., Jones, R.E. (Eds) The Ecology of Pests. CSIRO, Australia.

Thomsen I.M. and Koch, J. 1999. Somatic compatibility in Amyostereum areolatum and A. chailletii as a consequence of symbiosis with siricid woodwasps. Mycological Research 103: 817-823.

Titze, J.F. and Stahl, W. 1970. Inoculations of Pinus radiata trees with Amylostereum areolatum , the symbiotic fungus of Sirex noctilio . Australian Forestry Research 4: 3-14.

Titze, J.F. and Turnbull, C.R.A. 1970. The effect of club gland secretions of Sirex noctilio on the growth of the symbiotic fungus Amylostereum areolatum . Australian Forest Research 4: 27-29.

United States Department of Agriculture, Forest Service. 1985. Insects of Eastern Forests. Misc. Publ. 1426. Washington DC, United States Department of Agriculture, Forest Service.

Uzunovi ć, A. and Webber, J.F. 1998. Comparison of bluestain fungi grown in vitro and in freshly cut pine billets. European Journal of Forest Pathology, 28: 323-334.

Vaartaga. O. and King, J.M. 1964. Inoculation experiments with Amylostereum sp. from a woodwasp. Plant Disease Report 48: 438-440.

Valiev, A., Ogel, Z.B. and Klepzig, K.D. 2009. Analysis of cellulose and oxidase production by southern pine beetle associated fungi. Symbiosis. 49: 37-42.

Vasiliauskas, R. and Stenlid, J. 1999. Vegetative compatibility groups of Amylostereum areolatum and A. chailletii from Sweden and Lithuania. Mycological Research 103: 824-829.

Vasiliauskas, R., Stenlid, J., and Thomsen, I.M. 1998. Clonality and genetic variation in Amylostereum areolatum and A. chailletii from northern Europe. New Phytologist 139: 751- 758.

Villicide, J.M. and Corley, J.C. 2008. Parasitism and dispersal potential of Sirex noctilio : implications for biological control. Agricultural and Forest Entomology 10: 341-345.

Viri, H. 2004. Fungal associates of European bark beetles with special emphasis on the Ophiostomatoid fungi. In Lieutier, F., Day, K.R., Battisti, A., Grégoire, J.-C., Evans, H.F. (Eds) Bark and Wood boring Insects in Living Trees in Europe, a Synthesis. Kluwer Academic Publishers, Dordecht/Boston/London.

Wainhouse, D. 2004. Ecological Methods in Forest Insect Management. Oxford University Press, United Kingdom.

121

Walsh, K.D. and Linit, M.J. 1985. Oviposition biology of the pine sawyer, Monochamus carolinensis (Coleoptera; Cerambycidae). Annals of the Entomological Society of America 78: 81-85.

Waterhouse, D. F and Sands, D. P. A. 2001. Classical Biological Control of in Australia., ACIAR (Australian Centre for International Agricultural Research), Canberra, Australia.

Wermelinger, B., Rigling, A., Schneider Mathis, D., and Dobbertin, M. 2008. Assessing the role of bark- and woodboring insects in the decline of Scots pine ( Pinus sylvestris ) in the Swiss Rhone valley. Ecological Entomology 24: 103-110.

Whitney, H.S. 1982. Relationships between bark beetles and symbiotic organisms. In Mitton, J.B., Sturgeon, K. (Eds) Bark Beetles in North American Conifers. A System for the study of Evolutionary Biology. University of Texas Press, Austin.

Williams, D.W. and Mastro, V. 2008. Evaluation of Beddingia siricidicola as a biological control agent of Sirex noctilio in North America. McManus, D.,Gottschalk, K.W (Eds). Proceedings 19th U.S. Department of Agriculture Interagency Research Forum on Invasive Species 2008. United States Department of Agriculture Forest Service, Newtown, .

Wingfield, M.J. 1987. Fungi associated with the pine wood nematode Bursaphelenchus xylophilus , and cerambycid beetles in Wisconsin. Mycologia 79: 325-328.

Wong, L.K. and Crowden, R.K. 1976. Preliminary studies on the mucus secretion of the wood wasp, Sirex noctilio F. I. Physicochemical and biochemical properties. Australian Journal of Biological Sciences 29: 21-32.

Yanega, D. 1996. Field Guide to Northeastern Longhorned Beetles (Coleoptera: Cerambycidae). Illinois Natural History Survey, Illinois.

Yearian, W.C., Gouger, R.J., and Wilkinson, R.C. 1972. Effects of the bluestain fungus, Ceratocystis ips , on the development of Ips bark beetles in pine bolts. Annals of the Entomological Society of America 65: 481-487.

Yemshanov, D., McKenney, D.W., de Groot, P., Haugen, D., Sidders, D., and Joss, B. 2009. A bioeconomic approach to assess the impact of an alien invasive insect on timber supply and harvesting: a case study with Sirex noctilio in eastern Canada. Canadian Journal of Forest Research 39: 154-168.

Yu Q., P. de Groot, I. Leal, C. Davis, W. Ye, & B. Foord. 2009. Characterization of Deladenus siricidicola (Tylenchida: Neotylenchidae) associated with Sirex noctilio (Hymenoptera: Siricidae) in Canada. International Journal of Nematology 19:23-32.

Zondag, R. 1979. Control of Sirex noctilio F. with Deladenus siricidicola bedding. II. Introductions and establishments in the South Island 1968-75. New Zealand Journal of Forestry Science 9: 68-76.

122

Appendix

Appendix 1. Phloem- and woodboring insects collected from each tree species from pine forests in southern and central Ontario.

Tree species Insect species 2007 (Year 2008 (Year 2009 (Year one only) 1 2 emergence) emergence) Pinus sylvestris Astylopsis sexgutta (Say) 1 17 21 Apion sp . 1 1 0 Buprestidae 0 0 20 Cerambycidae 3 0 0 Cryphalus ruficollis Hopkins 1 0 0 Curculionidae 15 17 2 Gnathotrichus materiarius (Fitch) 184 34 3 Hylastes opacus Erichson* 5 11 0 Hylastes porculus Erichson 1 3 0 Hylastes tenuis (Eichoff) 0 8 1 Hylesinus criddlei Swaine 1 0 0 Hylobius congener Dalla Torre, Schenkling and Marshall 0 1 0 Hylurgops rugipennis pinifex (Fitch) 0 11 13 Ips calligraphus (Germar) 1 0 0 Ips grandicollis (Eichoff) 107 253 17 Ips pini (Say) 0 3 0 Ips sp. 2 0 0 Monochamus carolinensis (Olivier) 32 159 53 Monochamus notatus (Drury) 0 8 0 Monochamus scutellatus scutellatus (Say) 0 0 3 Monochamus sp. 6 2 0 Monochamus titillator (Fabricius) 5 0 0 Neacanthocinus obsoletus (Olivier) 1 0 0 Neacanthocinus pusillus Kirby 10 72 14 Orthotomicus caelatus (Eichoff) 38 2 0 Pissodes affinis Randall 0 1 0 Pissodes nemorensis Germar 112 344 0 Pissodes nemorensis or strobi 1 0 0 Pissodes sp. 0 2 3 Pissodes strobi (Peck) 0 3 0 Pityophthorus sp. 1 1 0 Pogoncherus mixtus Haldeman 0 0 1 Polygraphus rufipennis (Kirby) 0 0 6 Polydrusus sericeus (Schaller)* 1 0 0 Sirex edwardsii Brullé 9 11 1 Sirex nigricornis Fabricius 165 36 4 Sirex noctilio Fabricius* 3133 3201 105 Strophosoma melanogrammum (Forster)* 1 0 0 Tetropium cinnamopterum Kirby 2 2 0 Tetropium schwarzianum Casey 3 0 9 Tetropium sp. 1 1 0 Tetropium sp. nov. 0 20 2 continued 123

Appendix 1 cont. Tree species Insect species 2007 (Year 2008 (Year 2009 (Year one only) 1 2 emergence) emergence) Pinus sylvestris Tomicus piniperda (Linnaeus)* 2375 3122 19 Trypodendron betulae Swaine 6 0 0 Trypodendron lineatum (Olivier) 9 69 9 Trypodendron sp. 0 7 0 Xyleborinus saxesenii (Ratzeburg)* 23 0 16 Xyleborus dispar (Fabricius) 1 0 0 Xyleborus obesus LeConte 2 0 0 Xyleborus sayi (Hopkins) 6 0 0 Xyleborus sp. 1 0 0 Xyloterinus politus (Say) 6 0 0 Xylotrechus sagittatus sagittus (Germar) 0 2 37

Pinus banksiana Astylopsis sexgutta 1 Curculionidae 1 Ips grandicollis 13 Neacanthocinus pusillus 11 Orthotomicus caelatus 1 Pissodes nemorensis 5 Sirex nigricornis 3 Sirex noctilio* 275 Tomicus piniperda* 4

Pinus resinosa Curculionidae 8 0 Dendroctonus valens LeConte 1 0 Hylastes opacus* 3 0 Hylastes tenuis 0 4 Hylurgops pinifex 14 52 Ips grandicollis 148 0 Pissodes nemorensis 650 0 Pissodes sp. 2 3 Pissodes strobi 1 0 Sirex noctilio* 409 47 Strophosoma melanogrammum* 1 0 Tomicus piniperda* 741 2 Xylotrechus s. sagittus 0 29 * Insects known to be introduced into North America

124