Local and Geographic Variation in the Pheromone Blend of the Spruce Beetle, Dendroctonus Rufipennis Kirby (Coleoptera: Curculionidae)

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by British Columbia's network of post-secondary digital repositories

LOCAL AND GEOGRAPHIC VARIATION IN THE PHEROMONE BLEND OF THE SPRUCE BEETLE, DENDROCTONUS RUFIPENNIS KIRBY (COLEOPTERA: CURCULIONIDAE)

Rylee Isitt

B.Sc., University of Northern British Columbia, 2012

THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN NATURAL RESOURCES AND ENVIRONMENTAL STUDIES (BIOLOGY)

UNIVERSITY OF NORTHERN BRITISH COLUMBIA

August 2016

Abstract

The use of aggregation and anti-aggregation pheromones by spruce beetles has enabled the development of synthetic lures and repellants for monitoring and management purposes. However, the successful application of these tools across the spruce beetle’s large range may be stymied by geographic variation in the beetle’s response to and production of pheromone blends. Furthermore, a relative lack of published data on spruce beetle pheromone dynamics and regional pheromone variation may impede further research and the development of improved lures. Here I provide quantitative measurements of pheromone blends from spruce beetles obtained from numerous sites across Canada. I provide new evidence of geographic variation between the pheromone blends of beetles from eastern and western Canada, as well as within British Columbia and Alberta. I also show that feeding appears to be a prerequisite for pheromone production by spruce beetles, and that females transition from producing an aggregation pheromone to an anti-aggregation pheromone as they feed. These findings contribute towards a better understanding of spruce beetle chemical ecology, and may aid in the development of new, regionally-specific lure formulations.

Table of Contents

Abstract ...... ii Table of Contents ...... iii List of Tables ...... v List of Figures ...... vi Acknowledgements...... vii Chapter One. Literature Review and Objectives ...... 8 Spruce Beetle Ecology and Economic Importance ...... 8 Spruce Beetle Chemical Ecology ...... 9 Management of the Spruce Beetle ...... 10 Spruce Beetle Pheromone Variation ...... 11 Factors Influencing Pheromone Production ...... 11 Objectives ...... 12 Chapter Two. Spruce beetle pheromone production dynamics ...... 14 Abstract ...... 14 Introduction ...... 15 Methods ...... 16 Results ...... 20 Discussion ...... 22 Chapter Three. Variation in spruce beetle pheromone blends ...... 25 Abstract ...... 25 Introduction ...... 26 Methods ...... 30 Results ...... 40 Discussion ...... 53 Chapter Four. Effects of MCH amendments to a commercial spruce beetle lure on trap catches of spruce beetles and heterospecifics ...... 62 Abstract ...... 62 Introduction ...... 63 Methods ...... 64 iii

Results and Discussion ...... 66 Chapter Five. Conclusion ...... 69 References...... 70

List of Tables

Table 3.1. Sites from which spruce beetle infested trees were harvested...... 31

Table 3.2. GC-MS columns and methods used for analyses of spruce beetle hindgut and pooled aeration extracts of pheromone blends...... 38

Table 3.3. Enantiomeric ratios for chiral pheromone components extracted from the hindguts of spruce beetles ...... 50

Table 3.4. Enantiomeric ratios for chiral pheromone components obtained from pooled aerations of spruce beetles ...... 51

Table 3.5. Kendall’s tau-b correlations between absolute amounts (ng) of pheromone components extracted from beetle hindguts...... 52

Table 4.1. Mean counts of bark beetles and clerids caught in multiple funnel traps using three different semiochemical combinations ...... 67

List of Figures

Figure 2.1. Hindgut pheromone contents of Nova Scotia spruce beetles following different feeding durations and pairing treatments...... 21

Figure 3.1. Pooled aeration apparatus for obtaining extracts of volatiles from spruce beetle . 35

Figure 3.2. Representative tracings of galleries made over 48 hours by beetles from Rocky Mountain House (AB) tree B...... 41

Figure 3.3. Mean amounts of verbenene from hindgut extractions and pooled aerations of spruce beetles after 48 hours of feeding...... 44

Figure 3.4. Mean amounts of frontalin from hindgut extractions and pooled aerations of spruce beetles after 48 hours of feeding...... 45

Figure 3.5. Mean amounts of MCOL from hindgut extractions and pooled aerations of spruce beetles after 48 hours of feeding...... 46

Figure 3.6. Mean amounts of seudenol from hindgut extractions and pooled aerations of spruce beetles after 48 hours of feeding...... 47

Figure 3.7. Mean amounts of MCH from hindgut extractions and pooled aerations of spruce beetles after 48 hours of feeding...... 48

Acknowledgements

I would like to thank Drs. Dezene Huber, Katherine Bleiker, and Deepa Pureswaran for a great deal of support and resources committed to this project. Also thanks to Dr. Kirk Hillier and his lab for their generous assistance leading up to and during my time in Nova Scotia. A huge thanks to the many people at the Pacific Forestry Centre (Canadian Forest Service) in Victoria, BC who helped make this project possible. David Dunn and Rebecca Dixon at the Chemical Services Laboratory performed the GC-MS analyses of pheromone extracts. Eleanor Stewart and Charlene Lloyd sorted and counted trap catches from Rocky Mountain House, and Dr. Katherine Bleiker helped to locate and arrange for transport of infested spruce and feeding bolts from sites in BC and Alberta. I am very grateful to everyone who located, felled, bucked, and transported infested and uninfested spruce and provided field sites for trapping experiments. Those not already mentioned above include Dan Lavigne and Gary Holloway (Newfoundland and Labrador Department of Natural Resources), Pam Melnick, Margriet Berkhout, and Darcy Evanochko (Alberta Ministry of Agriculture and Forestry), Colin Chisholm and Doug Thompson (UNBC, Aleza Lake Research Forest), Gina Penny, Jacqueline Gordon, Jeff Ogden, Justin Smith, Michael LeBlanc, Morgan Oikle, and Tanya Borgal (Nova Scotia Department of Natural Resources), and Peter Romkey (Acadia University). I thank the K.C. Irving Environmental Science Centre at Acadia University for providing facilities for the storage and handling of infested spruce bolts. I also thank Dr. Lisa Poirier as well as the Enhanced Forestry Lab at UNBC for the same. Additional assistance was provided by Kathryn Berry, Eilish Engelhart, Heather Crozier, Ian Higgins, Loay Jabre, John Orlowsky, Timothy Owen, Victoria Rezendes, and the UNBC Forest Insect Research Group. Funding for this project was provided by the University of Northern British Columbia, the Natural Sciences and Engineering Research Council of Canada, the Canada Research Chairs program, the Canada Foundation for Innovation, the British Columbia Knowledge Development Fund, the provinces of Alberta and Nova Scotia through SERG International and Forest Protection Limited, and the Canadian Forest Service.

vii

Chapter One. Literature Review and Objectives

Spruce Beetle Ecology and Economic Importance

The spruce beetle (Dendroctonus rufipennis Kirby) is a native coniferophagous bark beetle that infests all species of spruce (Picea spp.) in North America. It spends most of its life within the phloem (inner bark) of the host tree, where it reproduces, feeds, and develops.

The duration of the beetle’s life cycle depends on temperature, with two years being typical.

Dispersal and colonization of new host trees usually occurs between May and July.

Oviposition occurs less than a week later, and the eggs incubate for up to several weeks before hatching. Both adults and larvae feed on the phloem tissues of the host tree. Under a two-year life cycle, beetles overwinter as larvae, pupate and eclose as callow adults the following spring or summer, then overwinter as adults before emerging and flying to new host trees two years after the initial colonization (Furniss and Carolin 1977, S. Wood 1982).

The spruce beetle is an eruptive species which typically infests stressed or downed trees, but at high population densities will outbreak into healthy spruce, particularly those of large diameter (Furniss and Carolin 1977). By preferring large-diameter, mature spruce during outbreaks, the spruce beetle may facilitate succession and growth release of understorey trees

(Lindgren and Lewis 1997, Veblen et al. 1991). Nevertheless, the economic damage caused by outbreaks has led to the spruce beetle being regarded as a pest species. The spruce beetle can cause high mortality of mature spruce (Werner and Holsten 1983), and may contribute to fire risk by increasing woody fuel loads (Garbutt et al. 2006).

Spruce Beetle Chemical Ecology

Live trees under attack by bark beetles are able to respond through the production and release of defensive chemicals. These consist largely of oleoresin terpenes. The physical properties of the oleoresin allow it to flush away or entrap attacking beetles, while its chemical constituents have fungicidal and insecticidal properties (Franceschi et al. 2005,

Theis and Lerdau 2003). However, the tree’s chemical defences may also aid attacking beetles. There is strong evidence that bark beetles are able to locate potential host trees by detecting and moving towards emissions of volatile terpenes (Moeck and Simmons 1991). In response, conifers may have evolved towards producing less constitutive (constantly present) oleoresin, while relying more on inducible oleoresin production in response to attack (Clark et al. 2010). Once bark beetles locate a host tree, they must still deal with the physical and chemical hazards posed by the oleoresin. When attacking a healthy tree, bark beetles must attack quickly and in large numbers in order to damage the tree’s vascular tissues before it can release copious amounts of oleoresin. This “mass attack” is coordinated through the use of aggregation pheromones, which attract conspecifics of both sexes (Raffa and Berryman

1983). Some of these aggregation pheromone components are likely derived from oleoresin terpenes, such as the conversion of α-pinene into verbenene (Blomquist et al. 2010, Hunt et al. 1989). For the spruce beetle, the known aggregation pheromone components are frontalin

(Dyer 1973, Gries et al. 1988), MCOL (Borden et al. 1996), verbenene (Gries et al. 1992a), and seudenol (Furniss et al. 1976, Vité et al. 1972). All four of these pheromone components are chiral, thus having two enantiomers, e.g. (+)-frontalin and (−)-frontalin. Later in the attack, the beetles produce an anti-aggregation pheromone, which deters further arrivals and prevents intraspecific competition due to overcrowding (Werner and Holsten 1995). The

spruce beetle’s achiral anti-aggregation pheromone component is MCH (Rudinsky et al.

1974).

Management of the Spruce Beetle

The spruce beetle’s use of pheromones and host kairomones has enabled the exploitation of these chemicals for spruce beetle management, monitoring, and research.

Spruce beetle management has been traditionally accomplished by clearcutting overmature stands, removing infested and windthrown trees, applying insecticides, and using trap trees to kill or remove large numbers of beetles. However, these methods require ground access to affected sites, may be stymied by an inability to quickly detect new outbreaks, and have the potential to be controversial in the cases of clearcutting and insecticide use (Furniss and

Carolin 1977, Holsten 1994). Some of these difficulties can be addressed through the use of pheromone-based lures and repellants. Synthetic anti-aggregation pheromone can be used to exclude spruce beetles from an area, and may be deployed as aerial applications in remote locations. Aggregation pheromone can be used to attract beetles to traps in order to monitor or reduce local populations (Borden 1989, Holsten 1994). Such techniques are effective for many bark beetle species, including Ips duplicatus Sahlberg (Schlyter et al. 2001),

Dendroctonus Brevicomis LeConte (DeMars et al. 1980), and Dendroctonus pseudotsugae

Hopkins (Ross et al. 1996).

Spruce Beetle Pheromone Variation

The use of synthetic lures for studying, monitoring, and managing spruce beetle populations may be hindered by regional variation in the response of the beetles to the lures.

For example, the addition of racemic MCOL to a spruce beetle lure consisting of α-pinene and frontalin significantly increased trap catches in Alaska (Werner 1994, Borden et al.

1996), northern British Columbia (Setter and Borden 1999), and Utah (Ross et al. 2005), but decreased trap catches in southeastern British Columbia and northwestern Alberta (Borden et al. 1996). There is also evidence of regional variation in the pheromone blends produced by beetles. Ryall et al. (2013) detected seudenol in beetles from Newfoundland, but not frontalin or MCOL, components which have been identified in beetles from BC (Borden et al. 1996, Gries et al. 1988). Finally, Borden et al. (1996) obtained inconsistent results from repeated trapping experiments within the same field site, suggesting that the response of spruce beetles to a given semiochemical blend can vary considerably within populations.

Factors Influencing Pheromone Production

Pheromone production and response are different mechanisms. Regardless, as with any communication system, there must be an association between production and response so that the sender can produce a signal which is understood by the receiver (Alexander 1962).

Therefore, it may be possible to understand variation in the response of spruce beetles to synthetic semiochemical blends by studying variation in the pheromone blends that the beetles themselves produce. Pheromone production is likely governed in part by genetics

(Miller et al. 1989, Roelofs et al. 2002), so spruce beetles from different lineages – both within and between populations – may produce different pheromone blends. Pheromone 11

production may also be influenced by the chemical environment of the host tree (Taft et al.

2015). One of the components of the spruce beetle’s aggregation pheromone blend, verbenene, appears to be derived from α-pinene obtained from the host tree (Blomquist et al.

2010, Hunt et al. 1989). The concentration of α-pinene in interior spruce varies strongly both within and between spruce populations, and its enantiomeric ratio also varies between populations (Pureswaran et al. 2004b). This may lead to variation in the amounts and enantiomeric ratios of verbenene produced by spruce beetles. Spatial variation in the chemical environment of host trees may also contribute to within-population pheromone variation (Birgersson et al. 1988).

Objectives

My objectives were to obtain quantitative measurements of the pheromone blends produced by spruce beetles in eastern and western Canada, and to document variation both within and between populations. In addition, I investigated pheromone production dynamics at a coarse scale, measuring how the pheromone blends produced by spruce beetles changed with the duration of feeding and the presence or absence of a mate. To date, publications on the pheromone blends produced by spruce beetles have been mostly qualitative and lacking in comparisons between different spruce beetle populations. And, to my knowledge, the only prior data on the enantiomeric ratios of pheromone components produced by spruce beetles were presented as unpublished data by Gries (1992), from a site near Gold Bridge, BC (G.

Gries, personal communication). A quantitative understanding of the pheromone blends produced by different spruce beetle populations will help to fill in some of these gaps. The

results of my study may contribute to a better understanding of the spruce beetle’s chemical ecology, facilitate future research, and inform the development of new synthetic lures.

Chapter Two. Spruce beetle pheromone production dynamics

Abstract

Little is known about the pheromone dynamics of the spruce beetle, and publications seem to disagree on the cues which prompt the transition from an aggregation to an anti- aggregation pheromone blend. Pheromone production may change in response to factors such as feeding duration and the presence of a mate. Some bark beetle species, such as

Dendroctonus frontalis, appear to produce large quantities of pheromone prior to dispersal.

However, initial failures to detect pheromone in extracts from emergent spruce beetles suggested that they required some additional cue before producing pheromone. These failures underscored the need for a better understanding of spruce beetle pheromone dynamics. My goals were to determine if feeding in a new host was a requirement for pheromone production by spruce beetles, and to quantify variation in the pheromone blend in response to different feeding durations as well as the presence or absence of a mate. Using spruce beetles from Nova Scotia, I collected hindgut pheromone extracts from unfed male and female beetles, as well as from beetles which had fed in spruce bolts for 24 and 48 hours, with and without a mate. Samples were analyzed using GC-MS. My results show that fed spruce beetles produce much more pheromone than unfed beetles, suggesting that feeding is a prerequisite for pheromone production. Additionally, females appeared to transition from producing an aggregation pheromone blend to producing an anti-aggregation pheromone blend in response to feeding duration, regardless of the presence or absence of a mate. In the context of prior literature, this suggests that there are multiple cues which prompt spruce beetles to produce anti-aggregation pheromone. These findings may contribute to a better understanding of how spruce beetle attacks progress through the aggregation and 14

anti-aggregation phases, and may help researchers develop reliable pheromone extraction protocols for the spruce beetle.

Introduction

The spruce beetle (Dendroctonus rufipennis Kirby) relies on the production of and response to several different pheromone components in order to attract mates and to coordinate mass attacks on healthy trees while discouraging intraspecific competition due to overcrowding. To accomplish this, spruce beetles first produce an aggregation pheromone in order to attract conspecifics of both sexes, and later produce an anti-aggregation pheromone that deters further arrivals (Werner and Holsten 1995). The known aggregation pheromone components of the spruce beetle are frontalin (Dyer 1973, Gries et al. 1988), MCOL (Borden et al. 1996), verbenene (Gries et al. 1992a), and seudenol (Furniss et al. 1976, Vité et al.

1972), and the known anti-aggregation pheromone component is MCH (Rudinsky et al.

1974). However, the cues for production of aggregation and anti-aggregation pheromones by spruce beetles appear to be poorly understood. For example, Holsten (1994) suggests that the production of MCH depends on reaching a threshold beetle density within the host tree, while

Rudinsky et al. (1974) show that MCH production can occur immediately following the pairing of males and females. An improved understanding of spruce beetle pheromone dynamics may not only offer new insights into the beetle’s chemical ecology, but can also help researchers optimize beetle rearing and handling methods in order to improve the quality of extracted pheromone samples.

During a project designed to quantify geographic pheromone variation in spruce beetles (chapter 3), I had initially assumed that emergent, unfed spruce beetles would have

sequestered aggregation pheromone, much like Dendroctonus frontalis (Zimmermann)

(Coster and Vité 1972). However, I failed to isolate any detectable pheromone from newly emerged D. rufipennis. The concentration and composition of the pheromone blends produced by bark beetles can vary in response to numerous factors, such as exposure to host tree volatiles, duration of feeding, and the presence of a mate (Coster and Vité 1972, Hughes

1973, Pureswaran et al. 2000, Rudinsky et al. 1974). Suspecting that a lack of understanding of these pheromone dynamics for the spruce beetle led me to make an erroneous assumption,

I decided to conduct a small pheromone dynamics experiment on spruce beetles collected from Nova Scotia.

My objectives were to test whether feeding in a new host spruce would increase pheromone production by spruce beetles, and to quantify any differences that might exist in the pheromone blends produced by beetles that fed for 24 hours or 48 hours, with and without a mate.

Methods

An infested white spruce was located near Stanley, Nova Scotia in November 2013. It was felled, cut into bolts, and transported to Acadia University. The bolts spent the winter outdoors on a pallet covered by a tarp, and the beetles within overwintered as larvae. In the spring, I sealed the bolts on both ends with molten paraffin wax in order to reduce dehydration. To speed up development of the beetles, I brought them into a 21°C phytotron pod at the K.C. Irving Environmental Science Centre at Acadia University in mid March,

2014. By late April, they contained mature adults and required an overwintering period before they would emerge. I used an artificial overwintering treatment in which I placed the

bolts into a 4°C cold room for at least 60 days. This temperature and duration was previously determined to be optimal by Dr. Katherine Bleiker at the Pacific Forestry Centre, resulting in a synchronized emergence of beetles once the bolts were brought back to room temperature.

To keep daily emergence rates manageable, I emerged the beetles in two groups: one starting in mid July, and another in mid August. I took the bolts containing beetles to be emerged into room temperature, placed into large plastic bins with mesh-lined ventilation holes. Beetles which emerged were retained in the emergence bins, and I collected them daily.

To reduce the possibility of unintended mating, I sexed the beetles within hours of collection using the characteristics described by Lyon (1958). Due to concerns over the potential impacts of post-emergence cold treatments on pheromone production, I did not refrigerate beetles for later use, and instead used beetles on the same day as emergence. After sexing, beetles were assigned to one of three feeding treatments (unfed, fed for 24 hours, and fed for 48 hours) and one of two grouping treatments (feeding alone or “solo”, and feeding with a mate of the opposite sex or “paired”). Since the paired grouping treatment was only applicable to feeding beetles, all unfed beetles were assigned to the “solo” treatment.

Additionally, males almost always chewed their way out of solo feeding treatments, so all fed males were paired with a female. This resulted in five overall treatments: unfed, fed for 24 hours solo (females only), fed for 48 hours solo (females only), fed for 24 hours paired, and fed for 48 hours paired.

The feeding treatments used uninfested spruce bolts (feeding bolts) that were obtained from a felled white spruce near Acadia University. I introduced beetles into feeding bolts by coaxing them into shallow holes (galleries) drilled under the bark, plugged the holes with shredded phloem, and then stapled mesh over the holes. I inserted solo females alone, one per

gallery, while pairs were inserted female first, and followed by a male. Beetles were left to excavate galleries and feed within the bolts for either 24 or 48 hours, after which I removed them by peeling away the bark and phloem. I separated the fed beetles by sex and treatment, and used all beetles that were in good health for that day’s pheromone extractions. Beetles assigned to the unfed treatment were not placed into feeding bolts, and were used for pheromone extractions on the same day that they emerged.

I obtained pheromone extracts for each individual beetle by removing the hindgut using sharp forceps, and soaking it in a vial of solvent. The solvent was a 4:1 mixture of pentane (≥ 99.0%, Fluka 76869) and hexane (≥ 99.0%, Fluka 52767), respectively, spiked with 5 ng/µL of heptyl acetate (≥ 98%, Sigma-Aldrich W254703) as an internal standard. I placed each beetle’s hindgut separately into 50 µL of the solvent mixture in a 200 µL GC vial insert nested inside of a 2 mL GC vial with a Teflon-lined septa cap. I let these vials sit at room temperature for 24 hours in order to mix gut contents with the solvent, and then moved them into a -30°C freezer for short term storage. I also made a number of blanks, which followed the same procedure but lacked beetle hindguts. Prior to being shipped off for chemical analyses, I removed the vial inserts and transferred the samples minus the gut tissue into the outer vials using a Hamilton syringe.

The hindgut extracts were analyzed using GC-MS by the Chemical Services

Laboratory at the Pacific Forestry Centre in Victoria, BC, using an Innowax 30 m x 250 µm x 0.25 µm column (Agilent 19091N133). The inlet was maintained at 200°C with a 1.0 mL/min flow rate and 5:1 split. The oven was held at 40°C for 1 minute, ramped to 130°C at

4°C/min, then to 240°C at 30°C/min, and held at 240°C for 5 minutes. SIM was used for m/z values of 43, 70, 82, 91, 97, and 97.1. Analytical standards for frontalin, verbenene, MCOL,

seudenol, and MCH were purchased from Contech, Inc. These were used to verify peak identities against known spectra as well as to create standard curves for converting peak areas into amounts (ng/beetle) for each of the pheromone components. The heptyl acetate internal standard was used to normalize samples in order to account for variation in injection volume or ionization. While this worked very well for hindgut extracts from male beetles, those from female beetles had degraded heptyl acetate peaks and the formation of a 1-heptanol peak.

Brian Sullivan (personal communication) confirmed that this occurs in female Dendroctonus spp., presumably due to an esterase in the gut. This issue was addressed by normalizing samples against the summed areas of the heptyl acetate and 1-heptanol peaks.

For statistical analyses, I first split the data into male and female groups. I then used

Shapiro-Wilk tests to check whether the amounts of each pheromone component were normally distributed within the five treatment groups. The data were significantly non- normal in most cases. I used Kruskal-Wallis tests to determine if there were significant differences across the treatment groups in the amounts of any of the pheromone components.

If so, I used Holm-Bonferroni corrected Dunn’s post-hoc tests to determine whether pairs of treatment groups were significantly different at α = 0.05. These tests were done separately for males and females, and for each of the four detected pheromone components (frontalin,

MCOL, seudenol, and MCH). I performed all statistical analyses using R v.3.2.3 (R Core

Team 2015) with the plyr (Wickham 2011), reshape2 (Wickham 2007), and FSA (Ogle 2015) libraries. I used the ggplot2 library (Wickham 2009) to create all charts.

Results

Unfed beetles of both sexes produced very little of any pheromone component (Fig.

2.1). Frontalin was a minor pheromone component in the Nova Scotia beetles, and was produced almost entirely by fed males. Seudenol was the majority aggregation pheromone component, produced predominantly by females. MCOL was only detected in the hindguts of fed females.

Solo and paired females that fed for 48 hours produced the most MCH of all treatment groups, significantly more than unfed females and paired females that fed for 24 hours. Although not significant, unfed females as well as females which fed for 48 hours produced noticeably less seudenol than females which fed for 24 hours. Paired males that fed for either 24 hours or 48 hours produced noticeably more frontalin and seudenol than unfed males, although these relationships were not statistically significant in all cases.

Frontalin MCOL a a 0.15 n=15 3 n=11 Female Female 2 0.10 a a n=17 0.05 a a a a 1 a n=14 a n=15 n=14 n=11 n=17 n=15 n=15 0.00 0 2.0 b 0.5 ab n=11 n=15 0.4 1.5 Male Male 0.3 1.0 0.2 0.5 a 0.1 a a a eetle) n=15 n=15 n=15 n=11 0.0 0.0 Unfed 24h 48h 24h paired 48h paired Unfed 24h 48h 24h paired 48h paired

Seudenol MCH 10.0 a a 100 b n=14 n=15 n=11 Female 75 Female 7.5 b 50 n=17 Hindgut Content (ng/b Content Hindgut 5.0 a a a ab a n=11 n=17 n=15 n=14 a 2.5 n=15 25 n=15 0.0 0 b a 0.8 n=15 n=15 a 75 0.6 n=11 Male Male 0.4 50 0.2 a 25 a a n=15 n=15 n=11 0.0 0 Unfed 24h 48h 24h paired 48h paired Unfed 24h 48h 24h paired 48h paired Feeding Treatment

Figure 2.1. Hindgut pheromone contents of Nova Scotia spruce beetles following different feeding durations and pairing treatments. Paired beetles fed with the opposite sex, and solo beetles fed alone. Means which share a common letter for a given pheromone component and sex are not significantly different in Kruskal-Wallis tests using Bonferroni corrected Dunn’s post-hoc tests (α = 0.05). 21

Discussion

The near total absence of detectable pheromone in the hindguts of unfed spruce beetles shows that they mostly lacked sequestered pheromone, and that pheromone production began once introduced into a new host. Coster and Vité (1972) found the opposite to be true for Dendroctonus frontalis (Zimmermann); the amount of frontalin and trans- verbenol in the hindguts of D. frontalis decreased after feeding. The natal bolts used in my experiment were highly infested and contained very little remaining phloem tissue by the time the beetles had matured to adults. The hindguts of unfed individuals were clear, while those of fed beetles contained orange-colored material. This suggests that ingested phloem is a prerequisite for pheromone production. Verbenene, though not detected in the Nova Scotia beetles, is likely produced from an α-pinene precursor (Blomquist et al. 2010, Hunt et al.

1989), which is a component of the host tree’s oleoresin. This provides one mechanism by which pheromone production may depend on feeding. The metabolic pathways responsible for the production of MCOL, seudenol, and MCH are not yet understood (Seybold et al.

2000), but may also depend on precursor chemicals obtained from the host tree. However, frontalin appears to be produced de-novo in the mevalonate pathway (Barkawi et al. 2003), suggesting that its production does not depend on specific host tree precursors.

The nature of the pheromone blend produced by females was noticeably different between beetles that had fed for 24 hours and those that had fed for 48 hours. The blend at 24 hours had consistently more seudenol and less MCH than the blend at 48 hours, regardless of whether females were solo or paired. Since MCH is an anti-aggregation pheromone component (Rudinsky et al. 1974), and seudenol is an aggregation pheromone component

(Furniss et al. 1976), this suggests that feeding females transition over time from producing

an aggregation pheromone to an anti-aggregation pheromone. This may seem to contradict prior suggestions that the production of MCH depends on attack density (Holsten 1994) or the pairing of males and females (Rudinsky et al. 1974). However, these are not mutually exclusive possibilities; spruce beetles may regulate the production of pheromone components such as MCH in response to numerous stimuli.

Although there were no significant differences between the amounts of pheromone components in the hindguts of solo and paired females, solo fed females tended to contain a greater quantity of MCH and MCOL when compared to paired fed females. This trend has proven consistent for spruce beetles from numerous sample sites across Canada, and is more pronounced outside of Nova Scotia (chapter 3). Since the two compounds are very similar, these results may be due to a single phenomenon. In the context of the findings of Rudinsky et al. (1974), who documented that unfed spruce beetles only produced MCH when paired, there appear to be multiple interacting effects influencing the amount of MCH produced by beetles. A decrease in hindgut aggregation pheromone content for paired (and presumably mated) females is consistent with Coster and Vité (1972), who documented a similar effect for D. frontalis.

Males produced the majority of the frontalin. This same trend is also seen in beetles from many other sample sites across Canada (chapter 3). In trapping experiments, frontalin is often more attractive to females than to males (Dyer 1973, Dyer and Chapman 1971, Dyer and Lawko 1978, Borden et al. 1996, Setter and Borden 1999). These qualities are similar to those of a male-produced sex pheromone, consistent with the hypothesis that many bark beetle aggregation pheromone components evolved from sex pheromones (Raffa et al. 1993).

My results suggest that feeding upregulates pheromone production by spruce beetles, and that the pheromone blend produced by feeding females shifts over time towards anti-aggregation. This shift, coupled with an increase in MCH concentration as more beetles arrive, may help to explain how spruce beetle attacks transition from the aggregation to anti-aggregation stages. However, Rudinsky et al. (1974) showed that galleries containing females become unattractive to other males once a male enters, suggesting that numerous interacting factors influence MCH production. In addition to contributing towards a better understanding of the spruce beetle’s chemical ecology, my findings should assist in the improvement of pheromone extraction protocols. Researchers who wish to extract pheromones from spruce beetles will likely obtain better results by feeding their beetles for carefully controlled durations. Future research should repeat this experiment using smaller time intervals in order to resolve non-linear effects and to facilitate the development of statistical models.

Chapter Three. Variation in spruce beetle pheromone blends

Abstract

Semiochemical lures and repellants are promising tools for managing and monitoring populations of the spruce beetle (Dendroctonus rufipennis Kirby). However, evidence of geographic variation in the response of spruce beetles to semiochemical lures and inconsistent results between trapping experiments suggest a need for a better understanding of how the spruce beetle’s natural pheromone blend varies across its range. My goals were to quantify the amounts and enantiomeric ratios of pheromone components produced by spruce beetles from numerous sites across Canada. I collected beetles from eastern and western

Canada and obtained pheromone extracts using a combination of pooled aeration and hindgut extraction techniques. Samples were analyzed by GC-MS using achiral and chiral columns.

My results show significant local and geographical variation in the pheromone blend of the spruce beetle, and high variation among individuals. In general, frontalin was produced in greater amounts in western Canada than in eastern Canada, while the pattern was reversed for

MCOL and seudenol. The enantiomeric ratios of frontalin produced by females from Rocky

Mountain House, Alberta and possibly Newfoundland differed significantly from those produced by all other groups. Verbenene was commonly detected in hindgut extractions of beetles from Rocky Mountain House, but not in any other sites in Alberta or BC, suggesting that regional variation may occur at fairly small scales. My data offer new insights into the chemical ecology of the spruce beetle, and provide quantitative measurements of regional spruce beetle pheromone blends that can serve as starting points for trapping experiments and the development of new, regionally-specific lure formulations.

Introduction

The spruce beetle (Dendroctonus rufipennis Kirby) is a native bark beetle with a wide distribution across North America. The spruce beetle plays an important role in the ecology of spruce forests. At low population densities, it infests large diameter downed trees or severely stressed trees and may facilitate the succession and release of understorey trees

(Lindgren and Lewis 1997, Veblen et al. 1991). At high population densities, the spruce beetle outbreaks into healthy spruce, aggregating to trees in large numbers in coordinated mass attacks (Furniss and Carolin 1977). These mass attacks allow the beetles to quickly kill host trees, which increases beetle survival by reducing the amount of defensive oleoresin produced by the trees (Raffa and Berryman 1983). Prior spruce beetle outbreaks have killed considerable numbers of trees throughout North America (Garbutt et al. 2006, Hodgkinson

1986, Kruse and Pelz 1991, Schmid and Frye 1977), and the spruce beetle is considered one of the most destructive agents affecting mature spruce (Hodgkinson 1986, Werner and

Holsten 1995). Spruce beetles use aggregation and anti-aggregation pheromones to coordinate and regulate mass attacks, respectively (Werner and Holsten 1995). The use of synthetic semiochemicals which mimic bark beetle pheromones has shown great promise as a management strategy for several species, including Ips duplicatus Sahlberg (Schlyter et al.

2001), Dendroctonus Brevicomis LeConte (DeMars et al. 1980), and Dendroctonus pseudotsugae Hopkins (Ross et al. 1996). Anti -aggregation pheromone can be used to reduce the influx of beetles into an area, and aggregation pheromone (lures) can be used to attract beetles to trap trees or artificial traps, either for monitoring or reduction of local beetle populations (Borden 1989, Holsten 1994). However, evidence of variation in the response to and production of pheromone blends both within and between spruce beetle populations

(Borden et al. 1996, Ross et al. 2005, Ryall et al. 2013, Setter and Borden 1999, Werner

1994) may preclude the possibility of a universal spruce beetle lure.

The spruce beetle infests all species of spruce (Picea spp.) across North America

(Schmid and Frye 1977). Following dispersal from natal trees, spruce beetles must be able to locate new host trees, overcome the physical and chemical defenses of healthy trees, and avoid overcrowding within the host to reduce intraspecific competition. This is achieved in part through behavioral responses to semiochemicals produced by beetles and complimented by host kairomones (Raffa et al. 1993, Raffa and Berryman 1983, D. Wood 1982).

Pioneering female spruce beetles locate suitable hosts by sensing volatile terpene kairomones released from host trees (Pureswaran and Borden 2005), which are components of the tree’s oleoresin (Langenheim 1994). Once the pioneering female has located a host tree, she produces an aggregation pheromone that attracts numerous conspecifics of both sexes. In this manner, large numbers of spruce beetles are able to aggregate to a host and initiate a mass attack. This mass attack is important for increasing beetle survival by preventing the host tree from producing copious amounts of defensive oleoresin (Raffa and Berryman 1983). Once the spruce beetle density within the host reaches a threshold, an anti-aggregation pheromone is produced which deters overcrowding (Werner and Holsten 1995).

Identified spruce beetle aggregation pheromone components include frontalin (Dyer

1973, Dyer and Chapman 1971, Gries et al. 1988), seudenol (3-methyl-2-cyclohexen-1-ol)

(Furniss et al. 1976, Gries et al. 1988, Vité et al. 1972), MCOL (1-methyl-2-cyclohexen-1- ol) (Borden et al. 1996), and verbenene (Gries et al. 1992a). A host tree volatile, α-pinene, enhances attraction (Dyer and Chapman 1971, Dyer and Lawko 1978). Identified anti- aggregation pheromones include MCH or seudenone (3-methyl-2-cyclohexen-1-one) (Furniss

et al. 1976, Kline et al. 1974, Rudinsky et al. 1974) and possibly 1-octen-3-ol (Pureswaran and Borden 2004, Pureswaran et al. 2004a). MCH is produced when male and female spruce beetles are paired (Rudinsky et al. 1974). Field trapping experiments have demonstrated that spruce beetles are attracted to various blends of the above aggregation pheromone components as well as host tree volatiles such as α-pinene (Borden et al. 1996, Dyer 1973,

Dyer and Chapman 1971, Dyer and Lawko 1978, Furniss et al. 1976, Ross et al. 2005, Setter and Borden 1999), and that MCH inhibits attraction (Furniss et al. 1976, Holsten et al. 2003,

Kline et al. 1974, Lindgren et al. 1989, Rudinsky et al. 1974).

There is disagreement among field experiments as to which synthetic spruce beetle aggregation blends work best, or even whether specific pheromone components synergize or inhibit attraction. Much of this variation appears to be regional. For example, the response of spruce beetles to racemic MCOL, (+)-MCOL, and (−)-MCOL varies between trapping experiments conducted in different regions (Borden et al. 1996, Ross et al. 2005, Setter and

Borden 1999, Werner 1994). In Newfoundland, Ryall et al. (2013) found that the addition of seudenol to frontalin and α-pinene lures in multiple funnel traps significantly increased trap catches, whereas Ross et al. (2005) found that the same seudenol amendment in Utah had no significant effect on the number of beetles captured. These studies suggest that the behavioral response of spruce beetles to specific pheromone blends varies between regional populations.

Additionally, existing literature hints at the possibility of regional variation between pheromone blends produced by beetles. For example, Ryall et al. (2013) were only able to detect seudenol in beetles from Newfoundland. They did not detect either frontalin or

MCOL, components which have been identified in beetles from BC (Borden et al. 1996,

Gries et al. 1988). Together, these studies demonstrate that the same pheromone blend may

elicit different responses from spruce beetles in different regions, and that naturally produced pheromone blends may vary among regions. This variation may reduce the efficacy of a universal lure for spruce beetle management, requiring the development of regionally- specific lures. Quantitative data on how the spruce beetle pheromone blend varies within and between regional populations may inform the development of new lure formulations, and will contribute towards a better understanding of the spruce beetle’s chemical ecology and evolution.

My objectives were to quantify the amounts and enantiomeric ratios of pheromone components produced by spruce beetles from numerous sites across Canada, and to use this data to describe pheromone blend variation within and between regional populations.

Methods

Nine infested spruce trees were located in six sites across Canada (Table 3.1), and five uninfested spruce from nearby sites were felled to act as food and temporary hosts for emerging spruce beetles. The uninfested spruce were chosen so that emergent beetles could be fed in bolts of the same species as the natal tree, and from a nearby site. Both infested and uninfested spruce were cut into bolts and transported either to Acadia University (for the

Nova Scotia bolts) or to the University of Northern British Columbia (all others). Bolts were sealed on both ends with molten paraffin wax in order to prevent desiccation. I stored uninfested spruce bolts in 4°C cold rooms until needed.

In most cases, infested bolts contained beetles in the last year of their life cycle. Bolts collected in the fall or winter usually contained adult beetles that would be ready to emerge after overwintering, and bolts collected in the spring or summer contained mature adults after leaving them at ambient outdoor temperatures until the fall. The bolts collected from near

Stanley, Nova Scotia were an exception. By winter, these bolts still contained late instar larvae. To expedite their development, I brought them into a 21°C phytotron pod at the K.C.

Irving Environment Science Centre at Acadia University in mid March, 2014. By late April, they contained mature adults and were ready for a final overwintering.

Table 3.1. Sites from which spruce beetle infested trees were harvested.

Site (ID) Harvest Date Coordinates No. of trees Tree Species Condition

Aleza Lake Research Forest, BC (ALRF) Sept 19, 2013 54°04'18.2"N 1 Interior hybrid spruce Windthrown 122°07'43.4"W Valemount, BC (VM) Jun 3, 2013 52°46'12.7"N 1 Interior hybrid spruce Windthrown 119°15'32.8"W Grande Prairie, AB (GP) May 31, 2013 54°51'38.9"N 2 Interior hybrid spruce Windthrown 118°42'48.8"W Rocky Mountain House, AB (RMH) Apr 15, 2015 52°27'43.6"N 3 White spruce Standing 115°24'23.7"W Stanley, NS (STA) Nov 28, 2013 45°43'59.3"N 1 White spruce Standing 64°03'56.5"W Gallants, NL (GAL) Feb 13, 2015 48°40'22.3"N 1 White spruce Standing 58°14'47.8"W

In order to obtain beetles on a flexible schedule throughout the year, I used artificial overwintering to control the timing of emergence. Using a method developed by Dr.

Katherine Bleiker at the Pacific Forestry Centre, I placed bolts containing adult beetles requiring a final overwintering into a 4°C cold room for at least 60 days. After this duration, the bolts could be warmed to room temperature, prompting a synchronized emergence of nearly 100% of beetles. When possible, I kept the duration of the overwintering treatment to between 60 and 90 days. However, due to problems with the initial pheromone extraction protocol, samples had to be recollected for BC and Grande Prairie (AB) beetles using infested bolts that had remained in the cold room for an extended period of time.

After the overwintering treatments were complete, I took selected groups of infested bolts to room temperature in large plastic emergence bins with mesh-lined ventilation holes.

At the same time, I warmed up one or more uninfested bolts of a matching species from the same geographic area. Only beetles from one site and one tree were emerged at a time. For heavily infested trees, I emerged beetles in multiple stages by warming up groups of bolts at different times. This kept emergence rates moderate, so that I could process a greater proportion of emerged beetles.

I collected emerged beetles daily. To reduce the occurrence of mating, I sexed the beetles within hours of collection using the characteristics described by Lyon (1958).

Preliminary experiments demonstrated that it was necessary to place spruce beetles into spruce bolts and allow them to feed on fresh phloem, otherwise they did not produce detectable quantities of pheromone (chapter 2). Additionally, due to concerns over the potential impacts of post-emergence cold treatments on pheromone production, I did not refrigerate beetles for later use. Only beetles which emerged on a given day were placed into

that day’s feeding treatments, as soon as possible after collecting and sexing them. I introduced beetles into feeding bolts by coaxing them into shallow holes drilled under the bark, plugged the holes with shredded phloem, and then stapled mesh over the holes. In some cases, I placed a lone female into a gallery, resulting in a “solo female” feeding treatment. In other cases, the female was followed by a male, resulting in “paired female” and “paired male” feeding treatments. Beetles were left to excavate galleries and feed within the bolts for

48 hours. Although a 24 hour feeding duration would likely have provided pheromone extracts containing larger amounts of the aggregation pheromone components (chapter 2), I did not discover this effect until well after pheromone samples had been collected. Once the feeding treatment was finished, I removed the spruce beetles by peeling away the bark and phloem. I again sexed the spruce beetles and separated them into the three treatment groups – solo females, paired females, and paired males. If in good health, these beetles were used as soon as possible (always on the day of removal) for pheromone extractions.

I extracted pheromones from beetles using two different methods – hindgut extractions and pooled aerations. Beetles were used either for hindgut extractions or pooled aerations, but not both. Both methods used the same solvent, a 4:1 mixture of pentane (≥

99.0%, Fluka 76869) and hexane (≥ 99.0%, Fluka 52767), respectively, spiked with 5 ng/µL of heptyl acetate (≥ 98%, Sigma-Aldrich W254703) as an internal standard. Pooled aerations required a minimum of about 15 beetles of a given feeding treatment group, so could not be performed when emergence rates were low (at the early or late stages of emergence, or for lightly infested trees).

I performed hindgut extractions of pheromone by removing the hindgut and other abdominal tissue from an individual beetle using sharp forceps. I placed this tissue into a 200

µL GC vial insert, along with 50 µL of the solvent mixture, and placed the insert into a 2 mL

GC vial with a Teflon-lined septa cap. I repeated this procedure for numerous beetles of each feeding treatment and natal tree. I let these vials sit at room temperature for 24 hours in order to mix gut contents with the solvent, and then moved them into a -30°C freezer for short term storage. I treated blanks identically, except that they did not contain tissue. Prior to being shipped off for chemical analyses, I removed the vial inserts and transferred the samples minus the gut tissue into the outer vials using a Hamilton syringe.

The pooled aeration apparatus (Fig. 3.1) was modified from Gries et al. (1992b), and methods were based on Gries et al. (1988), Gries et al. (1992b), and Pureswaran et al.

(2004). The main modification I made to these published methods was to split the aeration into two phases. For the first phase, I sealed the aeration chambers and allowed volatiles to accumulate within for 24 hours. For the second phase, I unsealed the aeration chambers and aspirated the chamber contents through the adsorbent columns using 6 L of air (0.6 LPM for

10 minutes). This was done to solve issues with sample contamination caused by the air flow, despite the use of activated charcoal filters. I assembled six aeration apparatus so that beetles from multiple feeding treatments as well as blanks could be aerated simultaneously. Each consisted of an adsorbent column, glass aeration chamber (ARS RodaViss® horizontal volatile collection chamber), medical stopcock, humidifying bubbler, flow meter, and activated charcoal air filter, in that order. The adsorbent columns consisted of 4” lengths of

6mm OD glass tubing, into which I secured 160 mg of HayeSep-Q 80-100 mesh (Supelco) between wads of glass wool.

Figure 3.1. Pooled aeration apparatus for obtaining extracts of volatiles from spruce beetles.

For non-blank aerations, I placed approximately 15-30 live beetles from a single feeding treatment group into an aeration chamber fitted with an adsorbent column. I closed the stopcock and sealed the free end of the adsorbent column with Telfon tape. I then wrapped the aeration chamber and adsorbent column in aluminum foil to prevent photodegradation of volatiles and behavioral effects of light sources on the beetles. After 24 hours, I opened the stopcock, removed the Teflon tape from the adsorbent column, and attached the free end of the column to a vacuum pump (Gast DOA-P704-AA). I turned on the pump and regulated the air flow rate to 0.6 LPM. After 10 minutes, I turned off the pump, disconnected the adsorbent column, and transported it to a different lab to be rinsed with solvent. I flushed the adsorbent columns with two 500 µl aliquots of the solvent mixture, the first which was allowed to percolate through the column for 15 minutes, the second which was forced through the column with pressurized ultrapure nitrogen (Praxair, Inc.). I collected the sample as it dripped from the adsorbent column into a glass culture tube, and transferred it into a GC vial using a Pasteur pipette. I stored these samples at -30°C until shipped off for chemical analyses. To clean the adsorbent columns for re-use, I placed them in an oven at

150°C and forced hot air through them at a low flow rate for approximately 2 hours.

Both hindgut extracts and pooled aeration samples were analyzed by GC-MS at the

Chemical Services Laboratory of the Pacific Forestry Centre in Victoria, BC. They used the same methods for both types of samples (Table 3.2). In addition to running all samples on an achiral column, a subset of samples with strong analyte peaks were run a second time on a chiral column in order to measure the enantiomeric ratios of verbenene, frontalin, MCOL, and seudenol. In early 2015 the Innowax column wore out and was replaced, and the method was changed slightly. To ensure that this did not introduce a significant source of variation, I

had a subset of prior samples re-run on the new column and compared them to the older dataset. Analytical standards for frontalin, verbenene, MCOL, seudenol, and MCH were purchased from Contech, Inc., in both racemic and optically active mixtures (where applicable). The heptyl acetate internal standard was used to normalize samples in order to account for variation in injection volume or ionization. While this worked very well for pooled aerations and hindgut extracts from male beetles, hindgut extractions from female beetles had degraded heptyl acetate peaks and the formation of a 1-heptanol peak. Brian

Sullivan (personal communication) confirmed that this occurs in female Dendroctonus spp., presumably due to an esterase in the gut. This issue was addressed by normalizing samples against the summed areas of the heptyl acetate and 1-heptanol peaks.

Table 3.2. GC-MS columns and methods used for analyses of spruce beetle hindgut and pooled aeration extracts of pheromone blends.

Type of analyses Column Flow Inlet Oven SIM ions rate Achiral 30 m x 250 µm x 1.0 200°C, 5:1 40°C for 1 min., ramp to 130°C at 4°C/min., ramp to 43.00, 70.00, 0.25 µm ml/min split 240°C at 30°C/min., held at 240°C for 5 mins. Total 82.00, 91.00, Innowax time: 32.167 mins. 97.00, 97.10 (Agilent 19091n133)

Achiral (Feb. 30 m x 250 µm x 1.0 200°C, 5:1 40°C for 2 min., ramp to 130°C at 3°C/min., ramp to 43.00, 70.00, 2015 and 0.25 µm ml/min split 240°C at 30°C/min., held at 240°C for 5 mins. Total 82.00, 91.00, onwards) WAX-plus (Zebron) time: 40.667 mins. 97.00, 97.10

Chiral 60 m Cyclodex-β 1.0 200°C, 100°C for 15 min., ramp to 230°C at 7°C/min. 6-17 mins: 43.00, (J&W 112-2562) ml/min 20:1 split 91.00, 97.10 17+ mins: 82.00, 97.00

The majority of achiral GC-MS data were non-normal, skewed to the right, and leptokurtic. Additionally, beetles tended to produce very little of some pheromone components, leading to ties at 0 ng/beetle for the least prevalent components. For these reasons, I used non-parametric statistical analyses for the achiral data. I assessed correlations using Kendall’s tau-b rank correlation coefficients (corrected for the presence of ties). To assess differences in the absolute amounts of pheromone components produced by beetles from different groupings, I ran Kruskal-Wallis tests against each of the five pheromone components, with the site/natal tree as the independent variable. I did this separately for solo females, paired females, and paired males. If the Kruskal-Wallis tests indicated significant differences, I followed up with Dunn’s post-hoc tests using Holm-Bonferroni corrections for multiple comparisons with α = 0.05. Enantiomeric ratio data were reasonably normally distributed. I used ANOVA to determine if significant variation existed in the proportion of

(+)-enantiomer for each chiral pheromone component between beetles grouped by sex and natal tree. If so, I followed up with a Tukey’s HSD post-hoc test to determine which pairwise combinations exhibited significant differences. I used R v.3.2.3 (R Core Team 2015) to perform statistical analyses, with the plyr (Wickham 2011), reshape2 (Wickham 2007), and

FSA (Ogle 2015) libraries. I used the ggplot2 library (Wickham 2009) to create all charts.

Results

Galleries created by solo females differed noticeably from those created by male and female pairs (Fig. 3.2). Approximately 40% of solo females did not create galleries, and remained in the drilled tunnel for the 48 hour feeding duration. In contrast, nearly all male- female pairs created galleries. Of the solo females which did create galleries, the mean gallery length was 1.1 cm. Male and female pairs created significantly longer galleries that averaged 1.9 cm in length (t = -4.76; d.f. = 19.8; p = 1.2 x 10-4). For both solo females and pairs, approximately 40% of galleries were oriented downwards from the drill hole, with the remainder oriented upwards.

Approximately 25% of galleries that were populated with pairs were later found to contain only the female, with exit holes indicating that the male had chewed his way out of the gallery. When solo female and paired feeding treatments were run concurrently, galleries that originally contained solo females were occasionally entered by males who had escaped their starting galleries.

Solo Females

Pairs

1 cm

Figure 3.2. Representative tracings of galleries made over 48 hours by beetles from Rocky Mountain House (AB) tree B. Dotted lines indicate drill holes.

Kruskal-Wallis tests indicated that all five pheromone components differed significantly in the amounts extracted from the hindguts of females from different trees. For males, there were significant differences in the amounts of verbenene, frontalin, and seudenol extracted from beetles from different natal trees.

Beetles from Nova Scotia and Newfoundland produced less frontalin, more MCOL, and more seudenol when compared to beetles from western Canada (with a few exceptions seen mostly in hindgut extractions) (Figs. 3.4 – 3.6). These trends were particularly clear in pooled aerations. Verbenene was found in relatively large quantities in the hindguts of beetles from Rocky Mountain House, but only one pooled aeration contained a detectable

(and very small) quantity of verbenene (Fig. 3.3).

Solo females produced different relative ratios of pheromone components than did paired females. Solo females almost always produced more MCOL, seudenol, and MCH, and less frontalin when compared to paired females (Figs. 3.4 – 3.7). MCH was the majority pheromone component produced by solo females. In contrast, paired females usually produced blends dominated by frontalin. However, paired females from Nova Scotia and

Newfoundland were noticeable exceptions. They produced significantly greater amounts of

MCH than paired females from all other sites, and their pheromone blends were dominated by MCH.

Differences in pheromone production between males and females were evident.

Males produced the majority of frontalin, females produced the majority of seudenol and

MCH, and MCOL was produced solely by females (Figs. 3.4 – 3.7).

In addition to geographic variation, variation within sites was apparent in Alberta, where multiple infested trees were obtained per site. The hindguts of solo females from RMH

tree C contained significantly less frontalin and verbenene than the hindguts of solo females from RMH trees A and B (Figs. 3.3, 3.4). The hindguts of solo and paired females from

RMH tree C also contained significantly more seudenol than those of females from trees A and B (Fig. 3.6 ). In all sites, variation between individual beetles from the same natal tree was quite high, with standard errors that frequently approached the means.

50 b 0.5 n=9 40 Solo Females 0.4 Solo Females 30 0.3 20 0.2 b 10 a a a a n=19 a a a 0.1 n=21 n=17 n=17 n=19 n=23 n=11 n=28 0 0.0 n=3 n=3 n=3 n=3 n=3 bc 0.5

n=8 Paired Females Paired Females 30 0.4 20 0.3 0.2 b 10 n=20 ac a a a a n=23 a a 0.1 n=18 n=16 n=15 n=15 n=17 n=31 0 0.0 n=2 n=2 n=2 n=2 n=2 n=3 10.0 b n=2 n=17 Paired Males Paired Males 7.5 b 0.4 n=20 ab 5.0 n=8 0.2 Mean hindgut verbenene content (ng/beetle S.E.) ± 2.5

a a a a a a Mean aeration verbenene content (ng/beetle/day ± S.D.) n=2 n=2 n=2 n=2 n=3 n=16 n=14 n=14 n=14 n=11 n=22 0.0 0.0

GP (AB) AGP (AB) B GP (AB) A VM (BC) A STA (NS)GAL A (NL) A VM (BC) A STA (NS)GAL A (NL) A ALRF (BC) A RMH (AB)RMH A (AB)RMH B (AB) C ALRF (BC) A RMH (AB)RMH B (AB) C

Natal Tree Natal Tree Figure 3.3. Mean amounts of verbenene from hindgut extractions and pooled aerations of spruce beetles after 48 hours of feeding. Solo beetles fed alone, while paired beetles fed with the opposite sex. Sample size for pooled aerations is the number of aerations, each of which contained 18 beetles on average. Bars that do not share a common letter suffix within a row are significantly different in Holm-Bonferroni corrected Dunn’s post-hoc tests (α = 0.05). 44

c n=3 n=9 2.0 90 Solo Females Solo Females 1.5 60 abc bc abd abcd abc d ad d 1.0 n=21 n=17 n=17 n=19 n=19 n=23 n=11 n=28 30 1.3 1.1 0.3 0.5 1.9 0 0 0.1 0.5 ± 0.4 ± 0.3 ± 0.2 ± 0.3 ± 0.6 ± 0 ± 0 ± 0.1 n=3 n=3 n=3 n=3 0 0.0 abcd 6 n=2

60 n=8 Paired Females Paired Females 4 40 abd n=15 n=2 2 20 a abcd ad n=20 bcd abcd n=18 n=16 bc n=23 c n=2 n=2 n=15 n=17 n=31 n=2 n=3 0 0 250 a n=2 n=16 90 Paired Males 200 Paired Males ab ab ab n=17 150 60 n=14 n=8 ab bc ab 100 n=2 n=2 Mean hindgut frontalin content (ng/beetle S.E.) ± n=20 n=22 30 n=14 n=2 bc Mean aeration frontalin content (ng/beetle/day ± S.D.) n=14 c 50 n=3 n=11 n=2 0 0

GP (AB) AGP (AB) B GP (AB) A VM (BC) A STA (NS)GAL A (NL) A VM (BC) A STA (NS)GAL A (NL) A ALRF (BC) A RMH (AB)RMH A (AB)RMH B (AB) C ALRF (BC) A RMH (AB)RMH B (AB) C

Natal Tree Natal Tree Figure 3.4. Mean amounts of frontalin from hindgut extractions and pooled aerations of spruce beetles after 48 hours of feeding. Solo beetles fed alone, while paired beetles fed with the opposite sex. Sample size for pooled aerations is the number of aerations, each of which contained 18 beetles on average. Bars that do not share a common letter suffix within a row are significantly different in Holm-Bonferroni corrected Dunn’s post-hoc tests (α = 0.05). The numbers after the sample size for solo female hindgut extracts indicate the mean ± S.E. for small bars. 45

c 20 n=3 n=23

20 Solo Females Solo Females 15 n=3 15 c n=28 10 10 ac n=19 abc abc abc 5 ab ab n=9 n=19 n=11 5 n=21 b n=3 n=3 n=17 n=17 n=3 0 0 0.6 ab n=3

n=31 Paired Females Paired Females 2 0.4

b n=2 1 n=17 0.2 a a a a ab a a n=18 n=16 n=15 n=15 n=8 n=20 n=23 0 0.0 n=2 n=2 n=2 n=2 0.5 0.5

0.4 Paired Males 0.4 Paired Males 0.3 0.3

Mean hindgut contentMCOL (ng/beetle S.E.) ± 0.2 0.2 Mean aeration contentMCOL (ng/beetle/day ± S.D.) 0.1 a a a a a a a a a 0.1 n=16 n=14 n=14 n=14 n=8 n=20 n=17 n=11 n=22 0.0 0.0 n=2 n=2 n=2 n=2 n=2 n=3

GP (AB) AGP (AB) B GP (AB) A VM (BC) A STA (NS)GAL A (NL) A VM (BC) A STA (NS)GAL A (NL) A ALRF (BC) A RMH (AB)RMH A (AB)RMH B (AB) C ALRF (BC) A RMH (AB)RMH B (AB) C

Natal Tree Natal Tree Figure 3.5. Mean amounts of MCOL from hindgut extractions and pooled aerations of spruce beetles after 48 hours of feeding. Solo beetles fed alone, while paired beetles fed with the opposite sex. Sample size for pooled aerations is the number of aerations, each of which contained 18 beetles on average. Bars that do not share a common letter suffix within a row are significantly different in Holm-Bonferroni corrected Dunn’s post-hoc tests (α = 0.05). 46

12.5 c n=3 n=23 10.0 Solo Females 15 Solo Females

7.5 10 n=3 5.0 bc n=28 ad ad abd bd abd 5 2.5 a a n=17 n=19 n=9 n=19 n=11 n=21 n=17 n=3 n=3 n=3 0.0 0 c 0.5

n=17 Paired Females Paired Females 2.0 c 0.4 n=23 1.5 0.3 bc 1.0 n=31 0.2 n=3 0.5 a a a a ab a 0.1 n=18 n=16 n=15 n=15 n=8 n=20 n=2 n=2 n=2 n=2 n=2 0.0 0.0 0.6 ab 0.5 n=11 Paired Males Paired Males b 0.4 0.4 n=22 0.3 0.2 Mean hindgut seudenol content (ng/beetle S.E.) ± 0.2

a a a a ab a a Mean aeration seudenol content (ng/beetle/day ± S.D.) 0.1 n=16 n=14 n=14 n=14 n=8 n=20 n=17 0.0 0.0 n=2 n=2 n=2 n=2 n=2 n=3

GP (AB) AGP (AB) B GP (AB) A VM (BC) A STA (NS)GAL A (NL) A VM (BC) A STA (NS)GAL A (NL) A ALRF (BC) A RMH (AB)RMH A (AB)RMH B (AB) C ALRF (BC) A RMH (AB)RMH B (AB) C

Natal Tree Natal Tree Figure 3.6. Mean amounts of seudenol from hindgut extractions and pooled aerations of spruce beetles after 48 hours of feeding. Solo beetles fed alone, while paired beetles fed with the opposite sex. Sample size for pooled aerations is the number of aerations, each of which contained 18 beetles on average. Bars that do not share a common letter suffix within a row are significantly different in Holm-Bonferroni corrected Dunn’s post-hoc tests (α = 0.05). 47

400 a 30 n=3 ab n=23 300 n=21 Solo Females Solo Females a 20 n=3 n=28 200 ab n=3 n=3 n=19 abc abc abc 10 n=9 n=19 n=11 n=3 100 c bc n=17 n=17 0 0 b b 4 n=3 40 n=17 n=31 Paired Females Paired Females 3 30 n=2 2 20 a a a n=23 1 10 n=18 a a a a n=20 n=16 n=15 n=15 n=8 0 0 n=2 n=2 n=2 n=2 1.00 a 0.5 n=14 a 0.75 n=11 Paired Males 0.4 Paired Males 0.3 0.50 Mean hindgut MCH content (ng/beetle S.E.) ± 0.2 0.25 Mean aeration MCH content (ng/beetle/day ± S.D.) a a a a a a a 0.1 n=16 n=14 n=14 n=8 n=20 n=17 n=22 0.00 0.0 n=2 n=2 n=2 n=2 n=2 n=3

GP (AB) AGP (AB) B GP (AB) A VM (BC) A STA (NS)GAL A (NL) A VM (BC) A STA (NS)GAL A (NL) A ALRF (BC) A RMH (AB)RMH A (AB)RMH B (AB) C ALRF (BC) A RMH (AB)RMH B (AB) C

Natal Tree Natal Tree Figure 3.7. Mean amounts of MCH from hindgut extractions and pooled aerations of spruce beetles after 48 hours of feeding. Solo beetles fed alone, while paired beetles fed with the opposite sex. Sample size for pooled aerations is the number of aerations, each of which contained 18 beetles on average. Bars that do not share a common letter suffix within a row are significantly different in Holm-Bonferroni corrected Dunn’s post-hoc tests (α = 0.05). 48

There were no significant differences between the enantiomeric ratios of verbenene and MCOL extracted from the hindguts of beetles, regardless of site/natal tree or sex (Table

3.3). Since verbenene was consistently produced only by beetles from Rocky Mountain

House, its enantiomeric ratio could not be characterized outside of RMH. The enantiomeric ratios of MCOL and verbenene obtained in pooled aerations generally agree with those from hindgut extractions (Table 3.4). The mean enantiomeric ratios combining all hindgut data regardless of natal tree or sex were 89:11 (+:−)-verbenene and 69:31 (+:−)-MCOL.

Frontalin was usually produced as nearly pure (−)-frontalin. A noticeable exception occurred among female beetles from the Rocky Mountain House (AB) and Newfoundland trees. In these cases, the female beetles produced a significantly greater proportion of

(+)-frontalin. Note, however, that the enantiomeric ratio for Newfoundland was based on a single individual, and this individual may not be representative. The mean frontalin enantiomeric ratio from all hindgut data excluding the females from RMH and NL was 5:95

(+:−)-frontalin, while that of females from RMH and NL was 38:62 (+:−)-frontalin.

The enantiomeric ratio of seudenol differed significantly between beetles from different natal trees, with no apparent geographic patterns. Additionally, the enantiomeric ratio of seudenol differed noticeably between the hindgut extraction and pooled aeration methods in several cases.

Table 3.3. Enantiomeric ratios for chiral pheromone components extracted from the hindguts of spruce beetles. Ratios that do not share a common letter suffix within the same columns (including between males and females) are significantly different in ANOVA with Tukey’s HSD post-hoc tests (α = 0.05).

Mean Enantiomeric Ratio (+:−) ± SE from Hindguts of Individual Beetles Site & Tree Verbenene Frontalin MCOL Seudenol

ALRF (BC) A 0:100 ± 0 (n=7) a 61:39 ± 2.2 (n=4) a VM (BC) A 0:100 ± 0 (n=4) a GP (AB) A 70:30 ± 3.7 (n=2) a 58:42 ± 6.8 (n=2) b GP (AB) B 3:97 ± 1.2 (n=5) a 70:30 ± 4.7 (n=8) a 0:100 ± 0 (n=2) c RMH (AB) A 87:13 ± 3.0 (n=7) a 40:60 ± 0.8 (n=7) b

Females RMH (AB) B 90:10 ± 1.4 (n=12) a 37:63 ± 1.0 (n=12) b 77:23 ± 9.6 (n=5) a 89:11 ± 5 (n=6) ab RMH (AB) C 91:9 ± 1.2 (n=2) a 39:61 ± 2.3 (n=2) b 74:26 ± 5.4 (n=12) a 91:9 ± 4.7 (n=10) a STA (NS) A 67:33 (n=1) a 58:42 (n=1) ab GAL (NL) A 41:59 (n=1) b 64:36 ± 2.1 (n=13) a 86:14 ± 4.3 (n=11) ab

ALRF (BC) A 4:96 ± 0.8 (n=15) a VM (BC) A 1:99 ± 0.9 (n=10) a GP (AB) A 2:98 ± 1.9 (n=3) a

GP (AB) B 3:97 ± 1.5 (n=8) a RMH (AB) A 88:12 ± 3.8 (n=3) a 10:90 ± 3.8 (n=8) a Males RMH (AB) B 91:9 ± 0.5 (n=3) a 9:91 ± 2.7 (n=15) a RMH (AB) C 86:14 ± 5.2 (n=2) a 7:93 ± 2.9 (n=13) a STA (NS) A 0:100 (n=1) a GAL (NL) A 9:91 ± 1.2 (n=17) a

Table 3.4. Enantiomeric ratios for chiral pheromone components obtained from pooled aerations of spruce beetles. N indicates the number of aerations, with each aeration containing 18 beetles on average. Standard deviations indicate the variation between aerations, and not between individual beetles.

Mean Enantiomeric Ratio (+:−) ± SD from Aerations of Pooled Beetles Sex Site & Tree Verbenene Frontalin MCOL Seudenol

ALRF (BC) A 23:77 ± 4.3 (n=4) 56:44 (n=1) 45:55 ± 13.3 (n=2) VM (BC) A 15:85 (n=1) 71:29 (n=1) GP (AB) A 54:46 (n=1)

Females STA (NS) A 68:32 ± 5.6 (n=3) 13:87 ± 3.9 (n=3) GAL (NL) A 61:39 ± 4.1 (n=2) 40:60 ± 9.6 (n=2)

ALRF (BC) A 3:97 ± 0.3 (n=2) VM (BC) A 4:96 (n=1)

GP (AB) A 4:96 ± 1.1 (n=2) RMH (AB) B 3:97 ± 0.4 (n=2) Males RMH (AB) C 73:27 (n=1) 4:96 ± 0.2 (n=2) STA (NS) A 25:75 (n=1) GAL (NL) A 9:91 ± 0.7 (n=3)

Statistically significant correlations existed between the amounts of various pheromone components detected in the hindguts of beetles (Table 3.5). Regardless of treatment group, there was a significant positive correlation between frontalin and verbenene.

This correlation was particularly significant for females. Highly significant positive correlations were found between all pairwise combinations of MCOL, seudenol, and MCH in females. These three compounds are very similar, with MCOL and seudenol being structural isomers, and MCH differing from seudenol only by the oxidization of the functional group.

All three of these pheromone components were also negatively correlated with frontalin, although this was not always statistically significant.

Table 3.5. Kendall’s tau-b correlations between absolute amounts (ng) of pheromone components extracted from beetle hindguts. Positive and negative values indicate positive and negative correlations, respectively.

verbenene frontalin MCOL seudenol Females, solo frontalin 0.47 **** MCOL 0.01 -0.12 seudenol 0.08 -0.21 ** 0.58 **** MCH -0.01 -0.15 * 0.71 **** 0.52 ****

Females, paired frontalin 0.39 **** MCOL -0.09 -0.19 ** seudenol -0.15 * -0.10 0.30 **** MCH -0.08 -0.13 * 0.41 **** 0.31 ****

Males, paired frontalin 0.16 * * P ≤ 0.05 ** P ≤ 0.01 *** P ≤ 0.001 **** P ≤ 0.0001

Discussion

My results provide new evidence of geographic variation in the spruce beetle pheromone blend. Beetles from Nova Scotia and Newfoundland produced less frontalin and more MCOL and seudenol when compared to beetles from Alberta and BC. This is congruent with the findings of Ryall et al. (2013), who detected only seudenol in beetles from

Newfoundland, and showed that seudenol amendments to frontalin and α-pinene lures improved trap catches in Newfoundland. Smaller scale geographic variation may also exist, such as the presence of verbenene in the hindguts of spruce beetles from Rocky Mountain

House, AB, and its general absence in other sites in AB and BC. Female spruce beetles from

Rocky Mountain House, AB produced significantly different enantiomeric ratios of frontalin versus females from Grande Prairie, AB or BC sites. Spruce beetles of both sexes from all sites across Canada consistently produced an enantiomeric excess of (−)-frontalin. However,

Gries (1992) measured nearly pure (+)-frontalin produced by spruce beetles from near Gold

Bridge, BC (G. Gries, personal communication). To confirm that my results were not in error

(e.g., due to a mislabeled analytical standard), I verified the order of elution for frontalin enantiomers on a cyclodex-β column against an independent source (Perez-Sanchez 1996).

Assuming that the data published by Gries (1992) are not in error, the enantiomeric ratio of frontalin produced by spruce beetles appears to vary strongly within BC. The existence of geographical variation in spruce beetle pheromone blends is not surprising given the variation seen in the response of spruce beetles to synthetic lures in trapping experiments

(Borden et al. 1996, Ross et al. 2005, Ryall et al. 2013, Setter and Borden 1999, Werner

1994), and the existence of geographic variation in the pheromone blends produced by other bark beetle species such as Dendroctonus brevicomis LeConte (Pureswaran et al. 2016),

Dendroctonus frontalis Zimmermann (Grosman et al. 1997) and Ips pini Say (Cognato et al.

1999, Miller et al. 1989, Teale et al. 1994).

Variation in the spruce beetle pheromone blend was also seen within sites, such as between female beetles from different natal trees taken from the same stands in Rocky

Mountain House and Grande Prairie, Alberta. Significant within-site variation is not apparent in the pheromone blends produced by males, but since multiple trees per site were only obtained in Alberta, I cannot say if the output of males is similarly consistent in other sites.

Assuming that this between-tree variation exists elsewhere and that production of and response to pheromone blends are correlated, it may explain why Borden et al. (1996) obtained inconsistent results from repeated trapping experiments conducted in the same site.

The mean enantiomeric ratio of MCOL from female beetles across Canada, 69:31

(+:−)-MCOL, agrees reasonably well with Gries (1992). The production by spruce beetles of an excess of (+)-MCOL is consistent with trapping experiments showing that spruce beetles either responded preferentially to (+)-MCOL, or exhibited no preference (Borden et al. 1996,

Setter and Borden 1999, Werner 1994). Although the enantiomeric ratio of seudenol extracted from beetle hindguts from different natal trees differs significantly within and between sites, these results should be interpreted with caution due to the low sample sizes for

Grande Prairie, AB and Nova Scotia. However, noticeable differences are also apparent between pooled aerations of beetles from different sites (Table 3.4), suggesting that the effect is not due merely to high individual variation and low sample size.

Male and female beetles produced very different pheromone blends. Females were the only sex which produced MCOL, and females also produced more MCH and seudenol than males. Males produced almost exclusively frontalin, and in much larger quantities when

compared to females. Additionally, in Rocky Mountain House, females and males produced significantly different enantiomeric ratios of frontalin. The presence of a mate also influenced pheromone production. Solo females produced more MCH, MCOL, and seudenol, and less frontalin when compared to paired females. Although MCOL, seudenol, and frontalin are considered aggregation pheromone components, the change in their relative ratios between solo and paired females may point towards more nuanced or multifunctional roles. These differences in pheromone production between the sexes may allow dispersing spruce beetles to determine the sex ratio of conspecifics in host trees based on the ratios of pheromone components and their enantiomers. I am not aware of any evidence for or against this possibility, and it may be a promising avenue for future research.

Although most of the trends discussed above can be seen in both hindgut extractions and pooled aerations, the two techniques often produced different results. Aerations frequently captured less of a given pheromone component per beetle than did hindgut extractions, but this was not always the case, and the magnitude of the difference was inconsistent between pheromone components. For example, while aerations of males collected more frontalin per beetle than did hindgut extractions, aerations of females collected far less MCH per beetle when compared to hindgut extractions. Disagreements between hindgut extractions and aeration techniques have previously been documented by

Pureswaran et al. (2000). Wood et al. (1966) determined that volatiles may adhere strongly to frass and release over a period of time. In contrast, a vigorous solvent extraction procedure may remove a large proportion of the volatiles. If the degree of adhesion to the frass differs between pheromone components, such as due to differences in volatility or polarity, aerations of living beetles or frass would likely provide extracts with different blends of volatiles when

compared to solvent extracts from hindgut contents. Aerations may be more representative of pheromone blends present in the field, while hindgut extractions remain a useful tool for measuring variation between individuals.

Variation in the pheromone blends produced by bark beetles can be caused by differences in the chemical environment of their host trees (Taft et al. 2015). Verbenene is likely produced by the hydroxylation and/or autoxidation of α-pinene (Blomquist et al. 2010,

Hunt et al. 1989). Pureswaran (2004) found that the concentration of (+)-α-pinene in white spruce varies significantly with geography, and documented high between-tree variation in the concentrations of both (−)- and (+)-α-pinene. Therefore, the verbenene produced by spruce beetles is likely influenced by the amount and/or enantiomeric ratio of α-pinene present in host trees. However, only Rocky Mountain House beetles consistently produced verbenene, and beetles from different natal trees within RMH produced significantly different amounts of verbenene despite being fed on bolts cut from the same uninfested spruce. This suggests that the α-pinene content of the host tree was not the predominant factor influencing the beetles’ production of verbenene. The possible effects of host tree chemistry on other spruce beetle pheromone components are less clear. Frontalin appears to be produced de novo from multiple precursors via the mevalonate pathway (Barkawi et al.

2003), and the biosyntheses of MCH, MCOL, and seudenol are not yet understood (Seybold et al. 2000). However, the strongly significant correlations I observed between these three pheromone components, their structural similarities, and the fact that MCOL will isomerize to racemic seudenol in the presence of a trace acid (J.P. LaFontaine, personal communication, April 2016) suggest a common origin. The possibility remains that MCH,

MCOL, and seudenol are influenced by host tree chemistry.

Genetic differentiation may be an important factor contributing to the variation observed in spruce beetle pheromone production. Maroja et al. 2007 found three distinct spruce beetle clades. The largely allopatric “Rocky Mountain” clade is found in Engelmann spruce in western North America, while two broadly sympatric “northern” clades are found associated with white spruce across North America. Small genetic changes can lead to large differences in pheromone blends. For example, the activation of a previously nonfunctional gene in two species of Ostrinia moths led to major shifts in their pheromone phenotypes

(Roelofs et al. 2002). Miller et al. (1989) found supporting evidence for a genetic basis for variation in the production of the ipsdienol aggregation pheromone in Ips pini bark beetles.

The genetic variation in spruce beetles demonstrated by Maroja et al. (2007) may be associated with significant differences in the pheromone blends both between and within regional populations.

Evidence based on pollen records, microfossils, and molecular data suggest that the

Pleistocene glaciations pushed spruce beetles and their hosts into several geographically isolated refugia in eastern and western North America (Maroja et al. 2007). According to refugia theory, this would encourage the genetic divergence of the isolated groups (Coyne and Orr 2004). Following the retreat of the glaciers, the beetles and their hosts were able to re-establish a contiguous distribution across North America. However, there is strong evidence of isolation by distance, suggesting that barriers to gene flow continue to exist between regional spruce beetle populations. Between the northern and Rocky Mountain clades, barriers to gene flow may be caused by a combination of host specificity and differences in aggregation pheromone blend (Maroja et al. 2007).

The two northern spruce beetle clades are as genetically different from one another as they are from the Rocky Mountain clade, yet they share a sympatric range and likely the same host species, P. glauca (Maroja et al. 2007). This suggests the existence of a barrier to gene flow that is based neither on geography nor host specificity. Bark beetle aggregation pheromones function partly as sex pheromones (Borden 1985), and variation creates an opportunity for assortative mating (Butlin 1995). Significant differences in the pheromone blend produced between spruce beetles which emerged from different trees in the Rocky

Mountain House site suggests that assortative mating is occurring. This may point to the existence of pheromone races, which are populations within a species that produce or respond to different chemical mating cues (Cognato et al. 1999). The existence of pheromone races and assortative mating has been demonstrated for Ips pini ( Cognato et al. 1999, Miller et al. 1989, Teale et al. 1994). If a similar situation exists for the spruce beetle, it would act as a barrier to gene flow between spruce beetle populations, possibly explaining the findings of Maroja et al. (2007).

The spruce beetle shares parts of its range with two other Dendroctonus spp. that use many of the same pheromone components. The Douglas-fir beetle, D. pseudotsugae Hopkins, inhabits Douglas-fir in western North America (Furniss and Kegley 2014), and the eastern larch beetle, D. simplex LeConte, inhabits tamarack east of the Rocky Mountains and in parts of Alaska (Seybold et al. 2002). The spruce beetle, Douglas-fir beetle, and eastern larch beetle are closely related species which share many of the same aggregation pheromone components (Symonds and Elgar 2004). All three species are attracted to traps containing commercial spruce beetle lures (chapter 4). However, the role of pheromones in mate finding and species recognition should lead to selective pressures that discourage the production of

and response to ambiguous pheromone blends (Cognato et al. 1997). The ranges of the spruce beetle, Douglas-fir beetle, and eastern larch beetle do not completely overlap, so these selective pressures would change across the landscape. Localized selective pressures may also arise from different communities of predators, parasites, and competitors. For example, some populations of Ips pini appear to have developed unique pheromone blends as a way of evading detection by local predators (Raffa and Dahlsten 1995, Raffa 2001). Coevolution within geographically distinct ecological communities may have driven the differentiation of the spruce beetle’s pheromone blend between regional populations.

Individual variation is very high. Spruce beetles from the same sex, treatment, and natal tree produced a range of pheromone component amounts that span two to three orders of magnitude. This level of individual variation in pheromone production appears to be common. Ips typographus (L.) individuals from within the same population, tree, and stage of attack exhibit orders-of-magnitude variation in the amounts of various pheromone components (Birgersson et al. 1988), and similarly high individual variation is seen in

Dendroctonus frontalis Zimmermann (Pureswaran et al. 2008). This degree of individual variation may seem paradoxical because of the role of bark beetle pheromones in species recognition. Stabilizing selection would be expected to reduce variability within species in order to maintain the recognition system (Linn and Roelofs 1995). However, for species such as the spruce beetle which aggregate in large numbers, the pheromone plume is comprised of the outputs from many individuals. This is expected to reduce the apparent variation that can be acted on by natural selection, allowing high individual variation to persist (Pureswaran et al. 2008). Assortative mating or sexual selection may also help to explain the persistence of this variation (Lande 1981).

Individual variation in pheromone production may arise from differences in natal environments which affect body condition and the ability to produce pheromone (Anderbrant et al. 1985), or by spatial variation in the chemical environment of the host tree (Birgersson et al. 1988). Additionally, hindgut extracts likely removed variable amounts of abdominal tissue from the beetles, possibly adding random error to subsequent measurements of pheromone component amounts. Another source of individual variation may come from the hypothesized existence of cheaters: individuals who take advantage of the pheromone production of others, but do not produce any of their own (Raffa 2001). Consistent with this hypothesis is the observation that individual variation in the amounts of pheromones is often highly skewed, with a few beetles producing substantial amounts of pheromone, and many producing little (Birgersson et al. 1988, Miller et al. 1989, Pureswaran et al. 2008). My results for spruce beetle show the same pattern; a small proportion of individuals produced a large proportion of the pheromone. Dugatkin et al. (2005) found that the existence of traits which benefit other individuals allows for the genesis and persistence of “cheaters” within populations. Although the existence of cheaters in bark beetle populations seems logical and consistent with the distributions of pheromone amounts produced by individuals, studies on

D. frontalis (Pureswaran et al. 2006) and I. typographus (Birgersson et al. 1988) were unable to support the cheater hypothesis.

The pheromone blends extracted from solo females after 48 hours of feeding were typically dominated by MCH, suggesting that they were producing an anti-aggregation pheromone blend. Nevertheless, males who escaped from concurrent paired feeding treatments often entered the galleries of solo females. This suggests that the females produced a pheromone blend that was attractive towards males at some point during the 48

hour feeding duration. Feeding females appear to transition from producing an aggregation pheromone blend to an anti-aggregation pheromone blend over a period of 24 to 48 hours

(chapter 2). This evidence shows that the feeding treatments were probably too long to capture aggregation pheromone blends from females. Future experiments should consider shorter feeding treatment durations.

My results provide evidence of significant geographical and local variation in the relative amounts of pheromone components and their enantiomeric ratios produced by spruce beetles. Beetles from eastern Canada produced different pheromone blends than those from western Canada. At smaller scales, spruce beetles from Rocky Mountain House, AB produced much more verbenene than beetles from all other sites, including nearby sites in

AB and BC. Comparisons to previously published data (Gries 1992) also suggest that different populations of spruce beetles produce very different enantiomeric ratios of frontalin.

More work must be done to better understand the causes and relevance of the variation I have observed, to fill in gaps in the data where sample sizes are low, and to collect data from additional sample sites. The quantitative measurements I have obtained for pheromone component amounts and proportions may be used as a starting point for future trapping experiments and the development of regionally-specific spruce beetle lures. I hope that these results will benefit future research into the population genetics, pheromone biosynthesis, chemical ecology, and management of the spruce beetle.

Chapter Four. Effects of MCH amendments to a commercial spruce beetle lure on trap catches of spruce beetles and heterospecifics

Abstract

During feeding treatments, escaped male spruce beetles were observed entering the galleries of solo females. However, the pheromone blends extracted from solo females suggested that they were producing large proportions of MCH, an anti-aggregation pheromone component. In the closely related species Dendroctonus pseudotsugae, MCH is a multifunctional pheromone component whose activity depends on concentration. At high doses, it is an anti-aggregation pheromone component, while at low doses, it synergizes attraction to other aggregation pheromone components. My goal was to determine whether a similar effect might occur for the spruce beetle. I conducted trapping experiments in BC and

Alberta, comparing catches of spruce beetles between treatments consisting of a commercial lure, and a commercial lure amended with an MCH bubble cap. The release rates of the lure and MCH bubble caps were such that the output roughly mimicked pheromone blends obtained in pooled aerations of solo female spruce beetles after 48 hours of feeding. The

MCH amendments significantly inhibited attraction to the commercial lure. This is consistent with many prior spruce beetle trapping experiments involving MCH, and shows that, at least at the release rates tested, MCH functions only as an anti-aggregation pheromone component.

The males observed entering the galleries of solo females were likely responding to an aggregation pheromone blend produced earlier in the feeding treatment, but which transitioned to an anti-aggregation pheromone blend by the end of the 48 hour duration.

Introduction

I had previously observed that female spruce beetles which fed alone in a spruce bolt for 48 hours produced a pheromone blend consisting mainly of MCH (chapter 3), an anti- aggregation pheromone component (Rudinsky et al. 1974). However, I observed that male spruce beetles, having escaped from other feeding treatments, were attracted to the galleries of the solo females. The pheromone blend produced by feeding female spruce beetles shifts towards anti-aggregation over time (chapter 2), so the males were likely responding to a blend produced earlier in the feeding treatment. Nevertheless, even 24 hours into feeding, females from Nova Scotia were already producing pheromone blends dominated by MCH

(chapter 2). This suggests that females begin to produce MCH shortly after colonizing a host.

The possibility remains that male spruce beetles were attracted to a pheromone blend consisting of a large proportion of MCH.

The behavioral activity of MCH may be dose-dependent, or vary based on its relative proportion in the overall pheromone blend. A multifunctional role of MCH is already known for D. pseudotsugae – it is anti-aggregative at high concentrations, but synergizes attraction at low concentrations (Rudinsky and Ryker 1980). Verbenone exhibits a similar dose- dependent role in Dendroctonus frontalis Zimmerman (Rudinsky 1973). Pureswaran et al.

(2000) suggest that, at least for Dendroctonus ponderosae Hopkins, the transition from an aggregation to an anti-aggregation pheromone blend may have more to do with the relative ratios of pheromone components than the presence or absence of any individual component.

In this experiment, I tested the hypothesis that MCH may be a multifunctional pheromone component for the spruce beetle. Specifically, I conducted trapping experiments to determine whether approximately equal ratios of frontalin, MCOL, and MCH were more

or less attractive to the spruce beetle than the combination of frontalin and MCOL alone.

These ratios roughly mimic those measured in pooled aerations of solo female beetles from western Canada (chapter 3), and could be achieved with commercial spruce beetle lures and

MCH bubble caps.

Methods

Synthetic spruce beetle lures (P/N 3123) and MCH bubble caps (P/N 3311) were purchased from Synergy Semiochemicals Corp. The lure consisted of a blend of racemic frontalin, racemic MCOL, and Engelmann spruce extract. The MCH bubble cap released at approximately 5 mg/day, while the frontalin and MCOL component of the lures released at approximately 5 mg/day and 6 mg/day, respectively (at 25°C).

Two experimental sites were chosen in spruce-leading stands. The Rocky Mountain

House (RMH), AB site (N 52.46296°, W 115.40068°) was experiencing a spruce beetle outbreak and had an endemic population of eastern larch beetle. Traps were set up in the

RMH site on May 7th, 2015, starting at the coordinates given and running north. The Aleza

Lake Research Forest (ALRF), BC site (N 54.05123°, W 122.04318°) contained an endemic population of spruce beetles and Douglas-fir beetles. Traps were set up in the ALRF site on

May 21, 2015, starting at the coordinates given and running west. Lindgren multiple-funnel traps (12 funnels), 45 per site, were placed along a transect running parallel to an access road in both sites. Traps were hung by wire from poles driven into the ground. The traps were separated from their neighbors by 15 meters on either side, and placed about 10 meters into the stand from the edge of the road. Wet cups were used for the traps, filled with

approximately 250 mL of propylene glycol-based plumbing antifreeze (Prestone Products

Corp.).

For each group of three adjacent traps, a “control”, “commercial lure”, and

“amended” treatment were assigned in a randomized block design. Commercial lure treatments consisted of the spruce beetle lure only (frontalin, MCOL, and Engelmann spruce extract), amended treatments consisted of the spruce beetle lure plus MCH bubble cap, and the control treatments lacked semiochemical loads. The semiochemical packets were hung inside the funnels by wire, approximately halfway up the funnel traps.

Trap catches were collected approximately weekly from the Aleza Lake Research

Forest (BC) site, until the traps were taken down on August 4th, 2015. Due to the distance from the University of Northern British Columbia, the RMH trap catches were only collected when the experiment was taken down on August 29th, 2015. All Dendroctunus spp. captured in the traps were identified, sexed, and counted. Additionally, clerids (Thanasimus spp.) were counted, but not sexed or identified to species.

Data were non-normal and were rank transformed prior to statistical analyses. The mean per-trap counts and sex ratios (when determined) of D. rufipennis, D. simplex, D. pseudotsugae, and clerid beetles were compared between treatment groups using ANOVA for a randomized block design, with Tukey’s HSD post-hoc tests. Statistical analyses were performed using R v.3.2.3 (R Core Team 2015).

Results and Discussion

Traps loaded with the commercial spruce beetle lures attracted a significantly greater number of spruce beetles, Douglas-fir beetles (in the ALRF site), eastern larch beetles (in the

RMH site), and Thanasimus sp. when compared to unbaited controls (Table 4.1). Adding

MCH bubble caps to the commercial lures completely inhibited their attractiveness to all three bark beetle species. These results are consistent with prior trapping experiments involving D. rufipennis (Holsten et al. 2003), D. pseudotsugae (Ross and Daterman 1995a,

Ross and Daterman 1995b), and D. simplex (Baker et al. 1977). The identical responses by these three bark beetle species are not surprising given that they are closely related and share some of the same aggregation pheromone components, including frontalin (all three species) and MCOL (D. rufipennis and D. pseudotsugae) (Symonds and Elgar 2004). Frontalin and

MCOL are the two pheromone components of the SB lure.

Amending the SB lures with MCH bubble caps did not affect trap catches of

Thanasimus sp. (Table 4.1). This is consistent with Ross and Daterman (1995a), who found that MCH treatments had no effect on the numbers of Thanasimus undatulus (Say) captured in traps. In both sites, Thanasimus sp. comprised a large proportion of the total catch. This may be a concern when using lethal traps, since Thanasimus sp. are predators of bark beetles.

Bycatch also included other predators and parasitoids, such as ichneumonids, cucujids, and lycids. Lethal traps using the SB lure may reduce local populations of predators and parasitoids that provide natural biological control of bark beetle populations. However, this also suggests a possible management strategy whereby stands are protected using a combination of commercial spruce beetle lure and MCH bubble caps. The MCH would inhibit the attractiveness of the lure to the bark beetles, but not to Thanasimus sp. Any bark

beetles which arrived would be faced with a greater risk of predation.

Table 4.1. Mean counts of bark beetles and clerids caught in multiple funnel traps using three different semiochemical combinations. Error is reported as standard deviation. Different letters for means within a column indicate statistically significant differences in ANOVA for randomized blocks (using rank-transformed data) with Tukey’s HSD post-hoc tests.

A. Aleza Lake Research Forest, BC:

Spruce Beetle Douglas-fir Beetle Thanasimus sp. Treatment Mean / trap ♀ Prop. Mean / trap ♀ Prop. Mean / trap Control 0.27 ± 0.59 a 0.67 ± 0.58 a 0.13 ± 0.52 a 0 ± 0 a 0.53 ± 0.92 b SB lure 11.1 ± 7.11 b 0.47 ± 0.22 a 55.6 ± 73.4 b 0.33 ± 0.14 a 46.9 ± 17.0 a SB lure + MCH 0.40 ± 0.63 a 0.30 ± 0.45 a 0.47 ± 0.74 a 0.60 + 0.55 a 53.2 ± 15.8 a

B. Rocky Mountain House, AB:

Spruce Beetle Eastern Larch Beetle Thanasimus sp. Treatment Mean / trap ♀ Prop. Mean / trap ♀ Prop. Mean / trap Control 1.36 ± 1.50 a 0.72 ± 0.31 a 0 ± 0 a N/A 0 ± 0 b SB lure 210 ± 159 b 0.58 ± 0.12 a 21.4 ± 52.2 b 0.35 ± 0.29 a 395 ± 237 a SB lure + MCH 1.86 ± 2.41 a 0.67 ± 0.30 a 1.79 ± 6.40 a 0.31 ± 0.44 a 339 ± 197 a

The lack of species specificity in the SB lure might allow for purposeful co-baiting of

D. pseudotsugae or D. simplex along with D. rufipennis. However, it may also be a symptom of an ambiguous and suboptimal lure formulation. Spruce beetles from several sites across

Canada produced predominantly (−)-frontalin and (+)-MCOL (chapter 3), in contrast to the

SB lure which uses racemic frontalin and MCOL. The racemic MCOL in the SB lure more closely matches the production by Douglas-fir beetles. Lindgren et al. (1992) determined that

Douglas-fir beetles from near Kamloops, BC produced 55:45 (+)-:(−)-MCOL. In the same paper, the results of trapping experiments showed that the beetles were attracted to either isomer of MCOL, and that a racemic blend caused an additive effect. On the other hand, 67

spruce beetles in some regions of western North America appear to prefer (+)-MCOL over

(−)-MCOL, even being repelled by (−)-MCOL in one site (Borden et al. 1996, Werner 1994).

The use of racemic frontalin and MCOL in the SB lure may reduce its attractiveness to spruce beetles.

My results do not support the hypothesis that roughly equal ratios of frontalin,

MCOL, and MCH may be attractive to spruce beetles. The presence of large proportions of

MCH in the pheromone blends of solo female spruce beetles which were observed to attract males (chapter 3 ) is likely due to changes in the nature of the pheromone blend later in the feeding treatment. This is supported by observations of feeding female spruce beetles from

Nova Scotia, which increased the amount and proportion of MCH in their pheromone blends over time (chapter 2).

However, the combination of SB lure and MCH bubble cap used in this trapping experiment is quite different from the naturally produced pheromone blend. Solo female beetles produced blends containing optically active frontalin and MCOL. In contrast, the spruce beetle lure used in this trapping experiment contained racemic frontalin and MCOL.

Additionally, the release rates of the lures and bubble caps represent hundreds of beetle- equivalents of each of these pheromone components (chapter 3). Given that dose-dependent effects are known to occur for other species of Dendroctonus (Rudinsky 1973, Rudinsky and

Ryker 1980), the high release rates used in this experiment may have contributed to an anti-aggregation effect. Using custom semiochemical blends and release devices with lower release rates was beyond the scope of this experiment, but would be a more conclusive test of the working hypothesis.

Chapter Five. Conclusion

 Feeding female spruce beetles appear to transition from producing an aggregation pheromone blend to producing an anti-aggregation pheromone blend over a period of 24 to 48 hours, regardless of the presence of a mate.  Both geographic (between site) and local (within site) variation appears in the pheromone blend of the spruce beetle.  Spruce beetles from Nova Scotia and Newfoundland produced less frontalin, and more MCOL and seudenol compared to spruce beetles from BC and Alberta.  Spruce beetles from Rocky Mountain House, Alberta produced significantly more verbenene than beetles from all other sites.  Females from Rocky Mountain House produced frontalin with a significantly greater proportion of the (+)- enantiomer when compared to both males and females from all other sites, with the exception of a single female from Newfoundland.  Although the spruce beetles in this study always produced an excess of (−)-frontalin, Gries (1992) has documented the opposite from beetles collected near Gold Bridge, BC.  The enantiomeric ratio of MCOL produced by beetles favors (+)-MCOL, which mirrors trapping experiments showing that beetles tend to respond preferentially to the (+)- enantiomer.  Individual variation in the amounts of pheromone components produced by spruce beetles is very high, with a few beetles producing a disproportionate amount.  Consistent differences between the hindgut extraction and pooled aeration methods suggest that aeration-based techniques may provide a more accurate measure of actual pheromone release.  Clerid beetles, common predators of the spruce beetle, were attracted in very large numbers to both commercial SB lures on their own, as well as those amended with MCH. A combination of the SB lure and MCH bubble caps may be useful for discouraging spruce beetle infestations while simultaneously attracting their predators.

References

Alexander, R. D. 1962. Evolutionary change in cricket acoustical communication. Evolution 16:443–467.

Anderbrant, O., F. Schlyter, and G. Birgersson. 1985. Intraspecific competition affecting parents and offspring in the bark beetle Ips typographus. Oikos 89–98.

Baker, B. H., B. B. Hostetler, and M.M. Furniss. 1977. Response of eastern larch beetle (Coleoptera: Scolytidae) in Alaska to its natural attractant and to Douglas-fir beetle pheromones. The Canadian Entomologist 109:289–294.

Barkawi, L. S., W. Francke, G. J. Blomquist, and S. J. Seybold. 2003. Frontalin: De novo biosynthesis of an aggregation pheromone component by Dendroctonus spp. bark beetles (Coleoptera: Scolytidae). Insect Biochem. Mol. Biol. 33:773–788.

Birgersson, G., F. Schlyter, G. Bergström, and J. Löfqvist. 1988. Individual variation in aggregation pheromone content of the bark beetle, Ips typographus. J. Chem. Ecol. 14:1737–1761.

Blomquist, G. J., R. Figueroa-Teran, M. Aw, M. Song, A. Gorzalski, N. L. Abbott, E. Chang, and C. Tittiger. 2010. Pheromone production in bark beetles. Insect Biochem. Mol. Biol. 40:699–712.

Borden, J., G. Gries, L. Chong, R. Werner, E. Holsten, H. Wieser, E. Dixon, and H. Cerezke. 1996. Regionally‐specific bioactivity of two new pheromones for Dendroctonus rufipennis (Kirby)(Col., Scolytidae). J. Appl. Entomol. 120:321–326.

Borden, J. 1985. Aggregation Pheromones, pp. 257–285. In G. Kerkut and L. Gilbert (eds.), Comprehensive insect physiology, biochemistry and pharmacology. Pergamon Press, Oxford.

Borden, J. 1989. Semiochemicals and bark beetle populations: exploitation of natural phenomena by pest management strategists. Ecography 12:501–510.

Butlin, R. 1995. Genetic variation in mating signals and responses, pp. 327–366. In D. Lambert and H. Spencer (eds.), Speciation and the Recognition Concept: Theory and Application. Johns Hopkins University Press, Baltimore, Maryland.

Clark, E. L., A. L. Carroll, and D. P. W. Huber. 2010. Differences in the constitutive terpene profile of lodgepole pine across a geographical range in British Columbia, and

correlation with historical attack by mountain pine beetle. Canadian Entomologist 142:557–573.

Cognato, A. I., S. J. Seybold, D. L. Wood, and S. A. Teale. 1997. A Cladisitic Analysis of Pheromone Evolution in Ips Bark Beetles (Coleoptera: Scolytidae). Evolution 51:313– 318.

Cognato, A. I., S. J. Seybold, and F. A. Sperling. 1999. Incomplete barriers to mitochondrial gene flow between pheromone races of the North American pine engraver, Ips pini (Say)(Coleoptera, Scolytidae). Proceedings of the Royal Society of London. Series B: Biological Sciences 266:1843–1850.

Coster, J. E., and J. P. Vité. 1972. Effects of feeding and mating on pheromone release in the southern pine beetle. Annals of the Entomological Society of America 65:263–266.

Coyne, J. A., and H. A. Orr. 2004. Speciation. Sinauer Associates Sunderland, MA.

DeMars, C. J., G. W. Slaughter, W. D. Bedard, N. X. Norick, and B. Roettgering. 1980. Estimating western pine beetle-caused tree mortality for evaluating an attractive pheromone treatment. J. Chem. Ecol. 6(5):853–866.

Dugatkin, L. A., M. Perlin, J. S. Lucas, and R. Atlas. 2005. Group-beneficial traits, frequency-dependent selection and genotypic diversity: an antibiotic resistance paradigm. Proceedings of the Royal Society B: Biological Sciences 272:79–83.

Dyer, E. 1973. Spruce beetle aggregated by the synthetic pheromone frontalin. Canadian Journal of Forest Research 3:486–494.

Dyer, E., and J. Chapman. 1971. Attack by the spruce beetle induced by frontalin or billets with burrowing females. Can. For. Serv. Bimonthly Res. Notes 27:10–11.

Dyer, E., and C. Lawko. 1978. Effect of seudenol on spruce beetle and Douglas-fir beetle aggregation. Bi-monthly Research Notes 34.

Franceschi, V. R., P. Krokene, E. Christiansen, and T. Krekling. 2005. Anatomical and chemical defences of conifer bark against bark beetles and other pests. New Phytologist 167: 353–376.

Furniss, M. M., B. H. Baker, and B. B. Hostetler. 1976. Aggregation of spruce beetles (Coleoptera) to seudenol and repression of attraction by methylcyclohexenone in Alaska. The Canadian Entomologist 108:1297–1302.

Furniss, R. L., and V. M. Carolin. 1977. Western Forest Insects. USDA Forest Service, Miscellaneous Publication N 1339. Washington, DC, US Government Printing Office.

Furniss, R. L., and S. J. Kegley. 2014. Douglas-fir Beetle. USDA Forest Service, Forest Insect and Disease Leaflet 5.

Garbutt, R., E. A. Allen, and B. C. Hawkes. 2006. Spruce beetle and the forests of the southwest Yukon, vol. BC-X-406. Pacific Forestry Centre.

Gries, G. 1992. Ratios of geometrical and optical isomers of pheromones: irrelevant or important in scolytids? J. Appl. Entomol. 114:240–243.

Gries, G., J. Borden, R. Gries, J. Lafontaine, E. Dixon, H. Wieser, and A. Whitehead. 1992a. 4-Methylene-6, 6-dimethylbicyclo [3.1. 1] hept-2-ene (verbenene): New aggregation pheromone of the scolytid beetle Dendroctonus rufipennis. Naturwissenschaften 79:367–368.

Gries, G., J. Borden, H. Pierce, B. Johnston, and A. Oehlschlager. 1992b. 3, 7, 7- trimethyl-1, 3, 5-cycloheptatriene in volatiles of female mountain pine beetles, Dendroctonus ponderosae. Naturwissenschaften 79:27–28.

Gries, G., H. Pierce Jr, B. Lindgren, and J. Borden. 1988. New techniques for capturing and analyzing semiochemicals for scolytid beetles (Coleoptera: Scolytidae). J. Econ. Entomol. 81:1715–1720.

Grosman, D. M., S. M. Salom, F. W. Ravlin, and R. W. Young. 1997. Geographic and gender differences in semiochemicals in emerging adult southern pine beetle (Coleoptera: Scolytidae). Annals of the Entomological Society of America 90(4):438– 446.

Hodgkinson, R. S. 1986. Use of trap trees for spruce beetle management in British Columbia, 1979–1984, vol. 5. Ministry of Forests.

Holsten, E. H. 1994. The role of spruce beetle pheromones as management strategies in Alaska. USDA Forest Service General Technical Report PSW-GTR-150:11–14.

Holsten, E. H., P. J. Shea, and R. R. Borys. 2003. MCH released in a novel pheromone dispenser prevents spruce beetle, Dendroctonus rufipennis (Coleoptera: Scolytidae), attacks in south-central Alaska. J. Econ. Entomol. 96:31–34.

Huber, D. P., and J. H. Borden. 2001. Angiosperm bark volatiles disrupt response of Douglas-fir beetle, Dendroctonus pseudotsugae, to attractant-baited traps. J. Chem. Ecol. 27:217–233.

Huber, D. P., R. Gries, J. H. Borden, and H. D. Pierce Jr. 2000. A survey of antennal responses by five species of coniferophagous bark beetles (Coleoptera: Scolytidae) to bark volatiles of six species of angiosperm trees. Chemoecology 10:103–113.

Hughes, P. R. 1973. Dendroctonus: Production of Pheromones and Related Compounds in Response to Host Monoterpenes1. Zeitschrift für angewandte Entomologie 73:294–312.

Hunt, D. W. A., J. H. Borden, B. S. Lindgren, and G. Gries. 1989. The role of autoxidation of α-pinene in the production of pheromones of Dendroctonus ponderosae (Coleoptera: Scolytidae). Can. J. For. Res. 19(10):1275–1282.

Kline, L., R. Schmitz, J. Rudinsky, and M. Furniss. 1974. Repression of spruce beetle (Coleoptera) attraction by methylcyclohexenone in Idaho. The Canadian Entomologist 106:485–491.

Kruse, J., and R. Pelz. 1991. Developing a public consensus on the management of spruce beetles on the Kenai Peninsula. Institute of Social and Economic Research, University of Alaska Anchorage.

Lande, R. 1981. Models of speciation by sexual selection on polygenic traits. Proceedings of the National Academy of Sciences 78:3721–3725.

Langenheim, J. H. 1994. Higher plant terpenoids: a phytocentric overview of their ecological roles. J. Chem. Ecol. 20:1223–1280.

Lindgren, B. S., and K. J. Lewis. 1997. The natural role of spruce beetle and root pathogens in a sub-boreal spruce forest in central British Columbia: a retrospective study. USDA Forest Service General Technical Report NE-236:122–130.

Lindgren, B. S., G. Gries, H. D. Pierce Jr, and K. Mori. 1992. Dendroctonus pseudotsugae Hopkins (Coleoptera: Scolytidae): production of and response to enantiomers of 1-methylcyclohex-2- en-1-ol. J. Chem. Ecol. 18(7):1201–1208.

Lindgren, B. S., M. D. McGregor, R. D. Oakes, and H. E. Meyer. 1989. Suppression of spruce beetle attacks by MCH released from bubble caps. Western Journal of Applied Forestry 4:49–52.

Linn Jr, C., and W. Roelofs. 1995. Pheromone communication in moths and its role in the speciation process, pp. 263–300. In D. Lambert and H. Spencer (eds.), Speciation and the Recognition Concept: Theory and Application. Johns Hopkins University Press, Baltimore and London.

Lyon, R. L. 1958. A useful secondary sex character in Dendroctonus bark beetles. The Canadian Entomologist 90(10):582–584.

Maroja, L. S., S. M. Bogdanowicz, K. F. Wallin, K. F. Raffa, and R. G. Harrison. 2007. Phylogeography of spruce beetles (Dendroctonus rufipennis Kirby)(Curculionidae: Scolytinae) in North America. Mol. Ecol. 16:2560–2573.

Miller, D., J. Borden, and K. N. Slessor. 1989. Inter-and intrapopulation variation of the pheromone, ipsdienol produced by male pine engravers, Ips pini (Say)(Coleoptera: Scolytidae). J. Chem. Ecol. 15:233–247.

Moeck, H. A., and C. S. Simmons. 1991. Primary attraction of mountain pine beetle, Dendroctonus ponderosae Hopk. (Coleoptera: Scolytidae), to bolts of lodgepole pine. Canadian Entomologist 123: 299–304.

Ogle, D. H. 2015. FSA: Fisheries Stock Analysis. R package version 0.8.4.

Pérez-Sánchez, A. L. 1996. Chemical ecology of Rhynchophorus palm weevils and Oryctes coconut beetles. Doctoral dissertation. Dept. of Chemistry, Simon Fraser University.

Poland, T. M., and J. Borden. 1998. Disruption of secondary attraction of the spruce beetle, Dendroctonus rufipennis, by pheromones of two sympatric species. J. Chem. Ecol. 24:151–166.

Poland, T. M., J. Borden, A. Stock, and L. Chong. 1998. Green leaf volatiles disrupt responses by the spruce beetle, Dendroctonus rufipennis, and the western pine beetle, Dendroctonus brevicomis (Coleoptera: Scolytidae) to attractant-baited traps. J. Entomol. Soc. BC 95.

Pureswaran, D. S., and J. H. Borden. 2004. New repellent semiochemicals for three species of Dendroctonus (Coleoptera: Scolytidae). Chemoecology 14:67–75.

Pureswaran, D. S., and J. H. Borden. 2005. Primary attraction and kairomonal host discrimination in three species of Dendroctonus (Coleoptera: Scolytidae). Agricult. For. Entomol. 7:219–230.

Pureswaran, D. S., R. Gries, J. H. Borden, and H. D. Pierce Jr. 2000. Dynamics of pheromone production and communication in the mountain pine beetle, Dendroctonus ponderosae Hopkins, and the pine engraver, Ips pini (Say)(Coleoptera: Scolytidae). Chemoecology 10:153–168.

Pureswaran, D. S., R. Gries, and J. H. Borden. 2004a. Antennal responses of four species of tree-killing bark beetles (Coleoptera: Scolytidae) to volatiles collected from beetles, and their host and nonhost conifers. Chemoecology 14:59–66.

Pureswaran, D. S., R. Gries, and J. H. Borden. 2004b. Quantitative variation in monoterpenes in four species of conifers. Biochem. Syst. Ecol. 32:1109–1136.

Pureswaran, D. S., R. W. Hofstetter, B. T. Sullivan, A. M. Grady, and C. Brownie. 2016. Western Pine Beetle Populations in Arizona and California Differ in the Composition of Their Aggregation Pheromones. J. Chem. Ecol. 42:404–413.

Pureswaran, D. S., B. T. Sullivan, and M. P. Ayres. 2006. Fitness consequences of pheromone production and host selection strategies in a tree-killing bark beetle (Coleoptera: Curculionidae: Scolytinae). Oecologia 148:720–728.

Pureswaran, D. S., B. T. Sullivan, and M. P. Ayres. 2008. High individual variation in pheromone production by tree-killing bark beetles (Coleoptera: Curculionidae: Scolytinae). Naturwissenschaften 95:33–44.

R Core Team. 2015. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria.

Raffa, K. F. 2001. Mixed messages across multiple trophic levels: the ecology of bark beetle chemical communication systems. Chemoecology 11:49–65.

Raffa, K., and D. Dahlsten. 1995. Differential responses among natural enemies and prey to bark beetle pheromones. Oecologia 102:17–23.

Raffa, K. F., T. W. Phillips, and S. M. Salom. 1993. Strategies and mechanisms of host colonization by bark beetles. Beetle-pathogen interactions in conifer forests 103–128.

Raffa, K., and A. Berryman. 1983. The role of host plant resistance in the colonization behavior and ecology of bark beetles (Coleoptera: Scolytidae). Ecol. Monogr. 27–49.

Roelofs, W. L., W. Liu, G. Hao, H. Jiao, A. P. Rooney, and C. E. Linn. 2002. Evolution of moth sex pheromones via ancestral genes. Proceedings of the National Academy of Sciences 99:13621–13626. 75

Ross, D. W., and G. E. Daterman. 1995a. Efficacy of an antiaggregation pheromone for reducing Douglas-fir beetle, Dendroctonus pseudotsugae Hopkins (Coleoptera: Scolytidae), infestation in high risk stands. The Canadian Entomologist 127:805–811.

Ross, D. W., and G. E. Daterman. 1995b. Response of Dendroctonus pseudotsugae (Coleoptera: Scolytidae) and Thanasimus undatulus (Coleoptera: Cleridae) to Traps with Different Semiochemicals. J. Econ. Entomol. 88:106–111.

Ross, D. W., G. E. Daterman, and A. S. Munson. 2005. Spruce beetle (Coleoptera: Scolytidae) response to traps baited with selected semiochemicals in Utah. West. North Am. Nat 65.

Ross, D. W., K. E. Gibson, R. W. Their, and S. A. Munson. 1996. Optimal Dose of an Antiaggregation Pheromone (3-Methylcyclohex-2-en-1-one) for Protecting Live Douglas-Fir from Attack by Dendroctonus pseudotsugae (Coleoptera: Scolytidae). J. Econ. Entomol. 89:1204–1207.

Rudinsky, J. A. 1973. Multiple functions of the southern pine beetle pheromone verbenone. Environ. Entomol. 2:511–514.

Rudinsky, J. A., and L. C. Ryker. 1980. Multifunctionality of Douglas-fir beetle pheromone 3,2-MCH confirmed with solvent dibutyl phthalate. J. Chem. Ecol. 6:193– 201.

Rudinsky, J., C. Sartwell, T. Graves, and M. Morgan. 1974. Granular formulation of methylcyclohexenone: an antiaggregative pheromone of the Douglas fir and spruce bark beetles (Col., Scolytidae). Zeitschrift für Angewandte Entomologie 75:254–263.

Ryall, K., P. Silk, G. Thurston, T. Scarr, and P. de Groot. 2013. Elucidating pheromone and host volatile components attractive to the spruce beetle, Dendroctonus rufipennis (Coleoptera: Curculionidae), in eastern Canada.

Schlyter, F., Q.-H. Zhang, G.-T. Liu, and L.-Z. Ji. 2001. A successful case of pheromone mass trapping of the bark beetle Ips duplicatus in a forest island, analysed by 20-year time-series data. Int. Pest Manag. Rev. 6:185–196.

Seybold, S. J., M. A. Albers, and S. A. Katovich. 2002. Eastern Larch Beetle. USDA Forest Service, Forest Insect Disease Leaflet 175.

Seybold, S. J., J. Bohlmann, and K. F. Raffa. 2000. Biosynthesis of coniferophagous bark beetle pheromones and conifer isoprenoids: evolutionary perspective and synthesis. The Canadian Entomologist 132:697–753.

Safranyik, L. 2011. Development and survival of the spruce beetle, Dendroctonus rufipennis, in stumps and windthrow, vol. BC-X-430. Pacific Forestry Centre.

Schmid, J., and R. H. Frye. 1977. Spruce beetle in the Rockies, vol. Rm-49. USDA Forest Service.

Setter, R. R., and J. H. Borden. 1999. Bioactivity and efficacy of MCOL and seudenol as potential attractive bait components for Dendroctonus rufipennis (Coleoptera: Scolytidae). The Canadian Entomologist 131:251–257.

Symonds, M. R., and M. A. Elgar. 2004. The mode of pheromone evolution: evidence from bark beetles. Proc. R. Soc. Lond. , Ser. B: Biol. Sci. 271:839–846.

Taft, S., A. Najar, and N. Erbilgin. 2015. Pheromone production by an invasive bark beetle varies with monoterpene composition of its naïve host. J. Chem. Ecol. 41:540–549.

Teale, S. A., B. J. Hager, and F. X. Webster. 1994. Pheromone-based assortative mating in a bark beetle. Anim. Behav. 48:569–578.

Theis, N., and M. Lerdau. 2003. The evolution of function in plant secondary metabolites. Int. J Plant Sci. 164(S3):S93–S102.

Veblen, T. T., K. S. Hadley, M. S. Reid, and A. J. Rebertus. 1991. The response of subalpine forests to spruce beetle outbreak in Colorado. Ecology 72:213–231.

Vité, J., G. Pitman, A. Fentiman, and G. Kinzer. 1972. 3-Methyl-2-cyclohexen-1-ol isolated from Dendroctonus. Naturwissenschaften 59:469–469.

Werner, R. A. 1994. Research on the use of semiochemicals to manage spruce beetles in Alaska, pp. 15-21 . In PJ Shea, technical coordinator, Proceedings of the symposium on management of western bark beetles with pheromones: research and development. General Technical Report PSW-GTR-150, USDA Forest Service, Pacific Southwest Research Station, Albany, CA.

Werner, R. A., and E. H. Holsten. 1983. Mortality of white spruce during a spruce beetle outbreak on the Kenai Peninsula in Alaska. Canadian Journal of Forest Research 13: 96– 101. 77

Werner, R. A., and E. H. Holsten. 1995. Current status of research with the spruce beetle, Dendroctonus rufipennis. USDA For. Serv. Gen. Tech. Rep. INT-GTR 318:23–29.

Wickham, H. 2007. Reshaping Data with the {reshape} Package. J Stat. Softw. 21:1–20

Wickham, H. 2009. Ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag, New York.

Wickham, H. 2011. The Split-Apply-Combine Strategy for Data Analysis. J Stat. Softw. 40:1–29

Wood, D. L. 1982. The role of pheromones, kairomones, and allomones in the host selection and colonization behavior of bark beetles. Annu. Rev. Entomol. 27:411–446.

Wood, D. L., L. E. Browne, R. M. Silverstein, and J. O. Rodin. 1966. Sex pheromones of bark beetles—I. Mass production, bio-assay, source, and isolation of the sex pheromone of Ips confusus (LeC.). J. Insect Physiol. 12(5):523–536.

Wood, S. 1982. Systematic Section, pp. 192-197. In S. Wood (ed.), Great Basin Naturalist Memoirs.