Treemediated Interactions Between the Jack Pine

Ecological Entomology (2011), 36, 425–434 DOI: 10.1111/j.1365-2311.2011.01283.x

Tree-mediated interactions between the jack pine budworm and a mountain pine beetle fungal associate

LINDSAY J. COLGAN andNADIR ERBILGIN Department of Renewable Resources, University of Alberta, Edmonton, Canada

Abstract. 1. Coniferous trees deploy a combination of constitutive (pre-existing) and induced (post-invasion), structural and biochemical defences against invaders. Induced responses can also alter host suitability for other organisms sharing the same host, which may result in indirect, plant-mediated interactions between different species of attacking organisms. 2. Current range and host expansion of the mountain pine beetle (Dendroctonus ponderosae Hopkins; MPB) from lodgepole pine-dominated forests to the jack pine- dominated boreal forests provides a unique opportunity to investigate whether the colonisation of jack pine (Pinus banksiana Lamb.) by MPB will be affected by induced responses of jack pine to a native herbaceous insect species: the jack pine budworm (Choristoneura pinus pinus Freeman; JPBW). 3. We simulated MPB attacks with one of its fungal associates, Grosmannia clavigera Robinson-Jeffrey & Davidson, and tested induction of either herbivory by JPBW or inoculation with the fungus followed by a challenge treatment with the other organism on jack pine seedlings and measured and compared monoterpene responses in needles. 4. There was clear evidence of an increase in jack pine resistance to G. clavigera with previous herbivory, indicated by smaller lesions in response to fungal inoculations. In contrast, although needle monoterpenes greatly increased after G. clavigera inoculation and continued to increase during the herbivory challenge, JPBW growth was not affected, but JPBW increased the feeding rate to possibly compensate for altered host quality. 5. Jack pine responses varied greatly and depended on whether seedlings were treated with single or multiple organisms, and their order of damage.

Key words. Boreal forest, Grosmannia clavigera, induced responses, jack pine budworm, jack pine, monoterpenes, mountain pine beetle, tree-mediated interactions.

Introduction (Franceschi et al., 2005). Examples of these include resin ducts, stone cells, and a broad spectrum of stored secondary Coniferous trees have evolved both constitutive and inducible metabolites such as terpenoids, phenolics, and alkaloids. defence systems that deter or kill insects and inhibit or Induced responses are triggered by the threat of tissue exclude pathogens physically and/or chemically (Johnson damage and are deployed to increase tree defence (Franceschi & Croteau, 1987; Brignolas et al., 1998; Hudgins et al., et al., 2005). They include numerous anatomical changes, 2003; Lieutier et al., 2003, 2009; Franceschi et al., 2005; such as traumatic resin duct formation and the activation Eyles et al., 2007, 2010). Constitutive defences are general of polyphenolic parenchyma cells, as well as quantitative defences maintained by a tree that are present at the time and qualitative changes of secondary metabolites (Franceschi of attack to provide immediate resistance to an invasion et al., 2005). Induced responses may alter a host’s suitability for sub- Correspondence: Nadir Erbilgin, Department of Renewable Re- sequent organisms attacking the same host plant (Bonello sources, 442 Earth Sciences Building, University of Alberta, Edmonton et al., 2006). Consequently, host tree chemistry may mediate AB, T6G 2E3 Canada. E-mail: [email protected] the interactions among attacking organisms. There are many

© 2011 The Authors Ecological Entomology © 2011 The Royal Entomological Society 425 426 Lindsay J. Colgan and Nadir Erbilgin examples of changes in conifer resistance (Raffa et al., 1998; Materials and methods Wallin & Raffa, 2001) and susceptibility (Annila et al., 1999; Krokene et al., 1999; Kosaka et al., 2001) after initial damage. Study system and justification These studies have documented that the induction of resistance MPB is the most destructive pest of pine forests in or susceptibility of plants depends on the organisms involved western North America and attacks mature trees in several (e.g. herbivores vs. pathogens), the organ of induction (e.g. species of Pinaceae, including lodgepole pine (Pinus contorta stem vs. foliage), and resource availability (Herms & Mattson, Douglas var. latifolia Engelman), (Wood, 1982). MPB carries 1992; Eyles et al., 2007). a variety of staining fungi that facilitate beetle establishment by Plant-induced responses also mediate interactions among disrupting transpiration and weakening tree defences, as well multiple organisms, such as herbivores and pathogens. Plant- as providing nutrition for larvae and teneral adults (Bleiker mediated indirect interactions between pathogens and herbi- & Six, 2007). For additional details of its lifecycle refer to vores can occur when the first coloniser alters the host plant Safranyik et al. (2010). in a way that affects the second coloniser that is often spa- Over the past decade, warmer winters and an abundance tially or temporally separated from the first. Previous stud- of hosts have allowed MPB to greatly expand its range into ies have documented that pathogen infection interferes with higher elevations and move eastward from lodgepole pine hormonal signalling defined as ‘cross-talk’ between defence forests west of the Rocky Mountains to lodgepole x jack pine pathways against pathogens (associated with salicylic acid) hybrid forests of northern Alberta (Carroll et al., 2006). It is and phytophagous insects (associated with jasmonic acid) expected that MPB will invade jack pine forests in the boreal (Hatcher et al., 2004; Beckers & Spoel, 2006). Cross-talk (Logan et al., 2003) as previous studies have established that can have different impacts on interacting organisms; pathogen the beetle (Cerezke, 1995) and its fungal associates can thrive infection can make a plant more or less susceptible to an in jack pine (Rice et al., 2007, 2008). This host and range insect. expansion is of great concern because once it establishes in The evolutionary history between trees and attacking jack pine forests, MPB has the potential to spread across organisms influences the outcome of host-mediated interactions Canada and invade eastern pine forests (Colgan & Erbilgin, (Tollrain & Harvell, 1999). If an exotic organism enters an 2010; Safranyik et al., 2010). ecosystem, both the exotic and native organisms could directly Winter temperatures in Alberta’s jack pine forests are interact through host availability and indirectly through host expected to initially limit MPB populations to low or endemic tree chemistry. Therefore, native organisms could influence levels (Regni´ ere` & Bentz, 2007; CFS, 2008). In its historical the survival and spread of the exotic organism, and thus its range, endemic beetle populations are constrained to attack invasion success. As a result, relationships between exotic trees with lowered defences such as those weakened by other and native organisms can be studied under the framework of insects and pathogens (Safranyik & Carroll, 2006). plant-mediated interactions. These investigations will become The JPBW is a defoliator that outbreaks periodically and increasingly important as more species expand their ranges is sympatrically distributed with its host jack pine. For owing to climate change. details of its lifecycle refer to McCullough (2000). Severe In the present study, we focused on the potential invasion defoliation of jack pine can result in growth reduction, top- of the North American boreal forest by an insect species, kill, and tree mortality (Kulman et al., 1963). As JPBW the mountain pine beetle, Dendroctonus ponderosae Hopkins herbivory does not usually result in tree mortality, jack pine (hereafter MPB), and its potential indirect interaction with defoliation may provide opportunities for MPB attack (Colgan a native herbivore, the jack pine budworm, Choristoneura & Erbilgin, 2010). Consequently, both insects could affect pinus pinus Freeman (hereafter JPBW) on jack pine, Pinus jack pine suitability for one another by inducing responses banksiana Lamb. We used one of the fungal associates of in host trees. As plant-mediated interactions depend on the mountain pine beetle, Grosmania clavigera Robinson-Jeffrey organisms and organs involved (Bonello, 2010), studying the & Davidson, as a proxy for MPB-induced changes on jack pine indirect interaction between JPBW and MPB on jack pine is because MPB has not yet established in jack pine stands. This important in understanding tree-mediated interactions between fungus was chosen for our experiments as it is well adapted a range-expanding exotic and a native species and potentially to cold temperatures and is likely to become more prevalent forecasting the risk of invasion of the boreal forest. in the fungal community of MPB in the boreal forest (Rice et al., 2008). Experiment design and data collection Our objectives were to (i) examine if the growth of JPBW or G. clavigera is affected by prior induction with the other A greenhouse experiment was conducted at the University organism; (ii) compare needle monoterpene responses to the of Alberta (Edmonton, AB, Canada) using jack pine seedlings, induction, challenge, and control treatments; (iii) explore the JPBW, and G. clavigera. In March 2009, 2-year-old jack pine relationship among host monoterpenes, JPBW growth, and seedlings were obtained from Boreal Horticultural Services G. clavigera lesions; and (iv) investigate if changes in host Ltd (Bonnyville, AB, Canada) and planted in 1-gallon pots monoterpenes as a result of prior induction can explain changes in a greenhouse. After 1 month, individuals of similar root in jack pine resistance or susceptibility to a challenge with collar diameter (8.2 ± 1.0 mm) and height (43.8 ± 0.5cm) JPBW or G. clavigera. were selected and randomly assigned to a treatment group.

Table 1. The two stage design of the experiment: an induction stage with one organism and a challenge stage with the other.

Herbivory – Pathogen Untreated pathogen infection Herbivore control infection – herbivory Pathogen control Wounded control control

Induction stage 2L 4L 2L 4L 3F 6F 3F 6F 3W 6W Untreated Challenge stage 3F 3F Con Con 2L 2L Con Con Con Con Con n 12 12 12 11 12 11 6 6 6 6 6

Seedlings in control groups were not treated in the challenge stage but were left to grow for that period. Con, Control; F, Grosmannia clavigera inoculations; L, Jack pine budworm larvae; W, inoculation wound.

Seedlings were watered three times a week and pots were Second and third instar budworm larvae were collected rotated twice a week to exclude any variation in greenhouse in pollen cones from northeastern Ontario and couriered to conditions. Fertilisation was stopped 2 weeks before the the University of Alberta. Larvae were reared on a modified experiment began to exclude confounding effects of nutrient- McMorran diet (CFS Insect Production Services, Sault Ste. defence interactions (e.g. Zhao et al., 2007). Seedlings had Marie, ON, Canada) until they reached fifth instar. When JPBW herbivory (induction) followed by inoculation with enough fifth instar larvae were available they were placed on G. clavigera (challenge), or were inoculated with G. clavigera the seedlings and screen bags were secured around the crown. (induction) followed by JPBW herbivory (challenge). There Seedlings that did not have larvae feeding were also bagged were two levels of each induction type followed by one level to avoid any confounding effects. Larvae ceased feeding in of challenge as well as induction-only control treatments. direct sunlight and tried to escape from the bags. Consequently, Control treatments were designed for each induction type during the herbivory stage all seedlings were kept under shade and level: herbivore (only) controls for herbivory–pathogen cloth that allowed 60% of the light to penetrate. Pupae were treatment and pathogen (only) controls for pathogen–herbivory collected every other day, weighed to the nearest 0.001 g, and treatments. Untreated control trees were also included as a left individually in containers at room temperature (∼23 oC) to blank control for a total of 11 treatments (Table 1). emerge. Adults that emerged were collected, weighed within In the herbivory–pathogen treatments, seedlings received 18 h, and then killed in a freezer. Forewings were removed either two or four fifth instar larvae in the induction stage from frozen moths, scanned, and their area was measured using that fed until pupation (∼3 weeks). One week after pupation, ImageJ 1.43 (Rasband, 2009). JPBW herbivory was quantified 4 weeks from the start of herbivory, seedlings received a by counting the number of needle bundles damaged in the challenge of inoculation with three plugs of G. clavigera.These crown. seedlings were harvested 9 weeks later. Induction controls According to methods from previous studies (Krokene et al., for JPBW herbivory (herbivore controls) had two or four 1999; Rice et al., 2007, 2008), trees were inoculated with a larvae feed but were not inoculated in the challenge stage. In pathogen associate of MPB, G. clavigera, to induce responses the pathogen–herbivory treatments, seedlings were inoculated in the tree. A 3-mm cork borer was used to remove the with three or six plugs of G. clavigera in the induction outer bark and phloem for inoculations and an agar plug with stage. After 9 weeks, two fifth instar larvae were placed on G. clavigera mycelia was placed in the hole. For wounded seedlings and allowed to feed until pupation (∼3 weeks). controls, sterile agar was placed in the hole. Parafilm R was These seedlings were harvested 4 weeks after herbivory began. wrapped around the stem to secure the inoculum, which Induction controls were inoculated with three or six plugs of also minimised the possibility of contamination. Inoculation G. clavigera (pathogen controls) but did not receive JPBW points were spread evenly over the stem below the crown, ◦ larvae. To test the effect of the pathogen compared with and adjacent holes were placed 180 from one another. the physical damage of inoculation, a wounded control with Tree defence to G. clavigera or wounding was quantified an induction of three or six mechanical wounds ‘inoculated’ by measuring lesion length, which is commonly used as an with sterile agar was included. These seedlings did not have indicator of the induced response to pathogens (Bonello & JPBW herbivory during the challenge stage. The induction and Blodgett, 2003; Krokene et al., 2008). Better defended trees challenge stages of the experiment were staggered so JPBW generally have smaller lesions. The bark was removed around would feed simultaneously for all treatments. The experiment each inoculation point using a scalpel and the length of the was designed with equal time periods for the fungal and lesion was measured to the nearest 0.01 mm using callipers. defoliation stages; however, inclement weather delayed larval Small xylem samples inside the lesion were taken to re-culture collection until after pathogen induction seedlings had been G. clavigera on Petri dishes of malt extract agar. inoculated. As a result, the pathogen stages of the experiment Monoterpenes were measured as an indicator of a response were lengthened from 4 to 9 weeks and lesions were measured mechanism of jack pine to these treatments because they are after the induction and challenge stages. Herbivore–pathogen well studied in bark beetle-conifer systems (Raffa et al., 2005), treatments and controls ran from the beginning of July to are part of the inducible defence responses of jack pine to late September 2009 and pathogen–herbivore treatments and JPBW (Wallin & Raffa, 1999) and bark beetles (Franceschi controls ran from late April to late July 2009. For all et al., 2005), and play a major role in inter- and intra-specific treatments, JPBW herbivory occurred in July. communication in conifer-bark beetle systems (Erbilgin &

Raffa, 2000, 2001; Raffa et al., 2005). Needles were sampled examine differences in pupal mass among pathogen–herbivore, for monoterpene analysis before the induction stage (T0),after herbivore–pathogen, and herbivore control seedlings for all the induction but before the challenge stage (T1), and after the pupae and by sex. The herbivory rate of JPBW was calculated challenge stage (T2) for all seedlings, including the untreated by dividing the number of needle bundles damaged by the total control. As monoterpenes can be highly variable (Latta et al., number of feeding days for all the larvae on a seedling. This 2000), undamaged, current year growth was taken from differ- standardised among seedlings for crown size, larval mortality, ent branches in the crown to obtain a bulk sample. Samples and number of larval feeding days, and created a rate of were stored on dry ice after collection until they were trans- bundles damaged per larva per feeding day. For missing larvae ferred to a −40 oC freezer and stored until chemical analysis. or those whose date of death was unknown, we assumed larvae fed for 2 weeks to underestimate budworm feeding damage on Chemical analysis a seedling. Herbivory rates were transformed (ln) for pooled comparisons, which were made using contrast statements. Needle samples were ground in liquid nitrogen using a Since sex was not significant in the analysis of pupal mass, mortar and pestle. Monoterpenes were extracted using a it was not tested in the analysis of herbivory rate. method modified from Wallis et al. (2008). In a 1.5-ml The length of G. clavigera lesions could not be easily trans- microcentrifuge tube, 500 μl of dichloromethane (CH2Cl2) formed to meet anova assumptions. Instead, Proc Npar1way with 0.1% tridecane as an internal standard was added to was used to conduct non-parametric Wilcoxon’s two-sample 100 mg of ground tissue. Samples were vortexed for 30 s, tests for pooled comparisons listed above. placed in an ultrasonic bath for 10 min, and then centrifuged at Total monoterpene concentrations of needles over the exper- o 0 C and 13.2 thousand rpm for 15 min. Extract was pipetted iment (T0, T1,andT2) were examined by repeated measures into an amber glass gas chromatograph vial and the tissue was anova using Proc Mixed with a compound symmetric covari- extracted for a second time using the same method. The first ance structure. An lsmeans statement was used to examine the and second extracts were combined and stored at −40 oC until changes within each treatment. When examined by time period, gas chromatography analysis. the initial concentrations of total monoterpenes significantly Samples (1 μl) were injected in an Agilent 7890A Gas Chro- differed before the induction stage (T0) (see Results section matograph (Agilent Technologies, Palo Alto, California) with below). As a result, changes () in monoterpene concentra- an HP Innowax column (I.D. 0.32 mm, length 30 m) with a tions were examined using Proc GLM over the induction stage helium carrier gas flow of 1.8 ml/min. The temperature pro- (T1 = [T1 − T0]/T0) and the challenge stage (T2 = [T2 − ◦ gramme began at 50 C for 2 min, which was increased to T1]/T1). These values were ln(1 + x) transformed and pooled ◦ ◦ ◦ ◦ 160 Cby5 C/min and then to 250 Cby20 C/min. Peaks comparisons (listed above) were made between challenge- were identified using the following standards: Borneol, pule- treated trees and their induction control for T1 and T2. gone, α-terpinene, γ -terpinene, α-terpineol (Sigma-Aldrich, The change in the concentrations of α-pinene, β-pinene, St. Louis, Missouri), camphor, 3-carene, α-humulene, terpino- 3-carene, limonene, and myrcene over the experiment were lene, α-thujone and β-thujone, (−)-α-pinene, (−)-β-pinene, also examined. Together these monoterpenes composed on (S)-(−)-limonene, sabinene hydrate, myrcene, (−)-camphene, average over 80% of total monoterpene concentrations, and p-cymene (Fluka, Sigma-Aldrich, Buchs, Switzerland), bornyl were also measured in an earlier study of JPBW and mature acetate, ocimene (SAFC Supply Solutions, St. Louis, Mis- jack pine (Wallin & Raffa, 1999). Proc Mixed was used to souri), β-phellandrene (Glidco Inc., Jacksonville, Florida), and perform repeated measures anovas to examine the change by comparison of the retention index. Peaks were integrated in the concentration of these monoterpenes within each and sample amount was compared between samples for all treatment over the experiments. Similar to the analysis of total seedlings at different time points based on calibration with the monoterpene concentrations, Proc GLM was used to examine above-mentioned compounds. the change in individual compounds over the induction (T1) and challenge stage (T2). Contrast statements were used to Statistical analysis examine the pooled comparisons listed above. Canonical correlation analysis (CCA) was used to explore The statistical program SAS v 9.1 (SAS Institute, Cary, the relationship between the biological and chemical datasets. North Carolina) was used for all analyses. For all variables, CCA is a multivariate extension of correlation analysis pre-planned pooled comparisons were made between: herbi- and is used to study the relationship between multivariate vore controls (2L Con + 4L Con) and herbivore–pathogen explanatory and response datasets, although dependency can treatments (2L 3F + 4L 3F); pathogen controls (3F Con + 6F be interchangeable (Johnson & Wichern, 1988). A variate is Con) and pathogen–herbivore treatments (3F 2L + 6F 2L); constructed from each dataset, using a linear combination of herbivore–pathogen and pathogen–herbivore treatments; her- variables, and the correlation between them is maximised. The bivore controls and pathogen controls; and pathogen controls strength of the relationship between the variates is quantified by and wounded controls (3W Con + 6W Con). the canonical correlation coefficient. Each canonical function, The JPBW variables pupal mass, adult mass, and wing area composed of an explanatory and response variate, is orthogonal were correlated so pupal mass was chosen for the analyses. to the next. The standardised partial correlation of each variable All the pupae from a treatment were lumped together and with their canonical variate is a canonical score and can be analysed. Proc GLM (General Linear Model) was used to interpreted in a similar manner to a regression coefficient.

Proc Cancorr was used to produce canonical correlations, 1.6 with the biological variate consisting of mean pupal mass (per a 1.4 a tree), mean lesion length (per tree), and the natural logarithm of herbivory rate (per tree). The chemical variate consisted 1.2 of the change in the concentration of α-pinene, 3-carene, and ab 1 myrcene over the experiment (T = [T2 − T0]/T0). We chose ab to examine these compounds as they were highly correlated 0.8 with other important monoterpenes and were present in most samples as CCA omits observations with missing data. 0.6 Herbivory Rate c 0.4 c Results 0.2 (Bundles damaged/Larva/Feeding day) Jack pine budworm 0 2L3F 4L3F 3F2L 6F2L 2LCon 4LCon Pupal mass was used in analyses because it was highly ± correlated with adult mass (r = 0.93,P <0.001) and wing Fig. 1. The herbivory rate of jack pine budworm (mean SE). area (r = 0.65,P <0.001) and had the highest number of Herbivory rate was transformed (ln) for the analysis. Larvae on pathogen infection then herbivory (3F 2L + 6F 2L) fed at significantly observations. Although some larvae went missing (11%) or higher rates than the other treatments (2L 3F + 4L 3F + 2L Con + 4L died before pupation (19%) (a total of 130 out of 186 Con) (F = 50.18, d.f. = 1, P < 0.0001). Con, Control; F, Grosmannia larvae were included in the statistical analysis), these were clavigera inoculations; L, Jack pine budworm larvae. not disproportionally related to any particular treatment (F = 0.88, d.f. = 5, P = 0.501 and F = 1.36, d.f. = 5, P = 0.252, 4.01,P <0.0001) and the pathogen control (Z = 3.71,P = respectively). Missing or pupal mortality figures reflect the 0.0002) seedlings had significantly longer lesions than the her- sum of values from all trees within a treatment and not bivore–pathogen seedlings (7.9 ± 0.2 mm) (Fig. 2). Further, the mean values per treatment. Overall, pupal mass did not lesion lengths were significantly longer in the pathogen control vary among treatments with a mean mass ± SE of 0.04 ± than the pathogen–herbivore seedlings (Z = 1.07,P = 0.286). 0.001 g (F = 0.97, d.f. = 5,P = 0.443). Likewise, it did not differ when treatments were grouped into herbivore–pathogen, Jack pine monoterpenes pathogen–herbivore, and herbivore control seedlings (F = 1.84, d.f. = 2,P = 0.167). When pupae were assessed by In the repeated measures anova of total monoterpene con- sex (female : male sex ratio: 1.58), mass did not differ centrations in needles over the experiment, treatment (F10,88.9 = = = = by treatment for females (F 0.76, d.f. 5,P 0.567) 2.15, P = 0.028), time (F2,178 = 36.28, P<0.0001), and = = = nor males (F 0.97, d.f. 5,P 0.443). Similarly, when their interaction (F20,178 = 20.35,P <0.0001) were significant grouped by herbivore–pathogen (n = 50), pathogen–herbivore (Fig. 3). When examined by treatment, monoterpene concen- (n = 48), and herbivore control (n = 32), mass did not differ trations in all treatments, except the untreated control, changed for females (F = 1.15, d.f. = 2,P = 0.330) nor males (F = 30 1.84, d.f. = 2,P = 0.168). Defoliation of jack pine crowns averaged 22.5 ± 2.8% 25 in herbivore–pathogen, 17.4 ± 1.4% in pathogen–herbivore, and 16.8 ± 1.6% in herbivore control seedlings. Defoliation 20 was standardised by calculating the herbivory rate. Herbivory rates significantly differed among treatments (F = 16.65, 15 d.f. = 5, P<0.0001; Fig. 1) and was significantly higher on ± pathogen–herbivore seedlings (1.2 0.1 bundles/larva/day) (mm) Length 10 than the herbivore–pathogen (0.6 ± 0.07) and herbivore control (0.7 ± 0.1) seedlings (F = 50.18, d.f. = 1,P <0.0001). 5 There was no difference in damage rates between larvae feeding on herbivore–pathogen and herbivore control seedlings 0 (F = 0.35, d.f. = 1,P = 0.558). Herb–Path Path–Herb Path Con Wound Con

Fig. 2. Box plot of lesion lengths grouped according to induction- Grosmannia clavigera lesion lengths challenge type. Box indicates the 25th and 75th percentile, the line through the box is the median, the capped lines indicate As expected, the pathogen control had signiﬁcantly longer the 10th and 90th percentile, and the dots are the 5th and 95th lesions (12.4 ± 1.1 mm) than wounded control seedlings percentile. Pathogen (only) control (Path Con) had longer lesions (4.9 ± 0.2 mm) (Z = 3.97,P <0.0001). Lesion lengths did than Wound Con (χ 2 = 67.24, P < 0.0001). Pathogen (only) control not differ between three and six inoculation treatments and pathogen infection–herbivory (Path-Herb) had longer lesions than in pathogen–herbivore seedlings (Z =−0.34,P = 0.735). the herbivory–pathogen infection (Herb-Path) seedlings (χ 2 = 37.27, However, the pathogen–herbivore (11.1 ± 0.7 mm) (Z = P < 0.0001 and χ 2 = 18.91, P < 0.0001, respectively).

4000 2L 3F (Fig. 3b). Whereas, over T2 monoterpene concentrations con- (a) 4L 3F tinued to increase in pathogen–herbivore seedlings (54.0 ±

g/g) 3000 2L Con

μ 4.1%) but remained relatively constant in the pathogen con- 4L Con trols (−8.8 ± 3.9%) (F = 63.33, d.f. = 1,P <0.0001). The 2000 Untreated pathogen controls did not differ from the untreated control in = = = 1000 T2 (F 0.61, d.f. 1,P 0.437). When monoterpene concentrations at the end of study 0 were compared between pathogen–herbivore and herbi- 4000 vore–pathogen seedlings, we detected a surprising outcome 3F 2L (b) 6F 2L with herbivore–pathogen seedlings having lower monoter- g/g) Concentration ( μ 3000 3F Con pene concentrations (1374.0 ± 124.9 μg/g) than the pathogen– 6F Con herbivore seedlings (2359.9 ± 155.4 μg/g) (Fig. 3c). In T1, Untreated 2000 monoterpene concentrations remained constant in the herbivore–pathogen seedlings but increased in pathogen–herbivore

Concentration ( 1000 seedlings (F = 22.87, d.f. = 1,P <0.0001). In T2, concentrations continued to increase in the pathogen–herbivore 0 4000 seedlings but decreased in the herbivore–pathogen seedlings 2L 3F (c) (F = 134.72, d.f. = 1,P <0.0001).

g/g) 4L 3F

μ 3000 3F 2L Similarly, monoterpene responses differed between herbi- 6F 2L vore control and pathogen control seedlings. Monoterpene 2000 Untreated concentrations increased in both the herbivore and pathogen controls over the study, but in different stages. In T1, 1000 Concentration ( monoterpene concentrations remained relatively constant in the = 0 herbivore control but increased in the pathogen control (F Initial Induction Challenge 22.65, d.f. = 1P<0.0001). In T2, concentrations greatly ± increased in the herbivore controls but slightly decreased in Fig. 3. Total monoterpene concentrations (mean SE) before the = = induction stage (initial), after the induction stage, and after the the pathogen controls (F 158.54, d.f. 1,P <0.0001). At challenge stage. Seedlings that received (a) jack pine budworm the end of the experiment, the concentration of monoterpenes defoliation induction and the untreated control, (b) Grosmannia in the herbivore controls was higher than the pathogen controls. clavigera inoculation induction and the untreated (blank) control, and Monoterpene responses in seedlings differed between (c) an induction and challenge treatment and the untreated (blank) G. clavigera pathogen and wounded controls. In T1, monoter- control. pene concentrations increased more in pathogen (94.2 ± 19.10%) than wound control (42.1 ± 15.1%) seedlings (F = significantly over the experiment (Table S1). The constancy of 5.71, d.f. = 1,P = 0.019). Wound control seedlings also the untreated control indicates that although the study occurred increased compared to the untreated control (F = 7.26, d.f. = over the course of the summer, there was no significant change 1,P = 0.008). In T2, concentrations in the wound controls in monoterpene concentration in jack pine seedlings during this decreased (−47.4 ± 2.5%), whereas the pathogen controls did time. As initial monoterpene concentrations in needles differed not decrease significantly (−8.8 ± 3.9%) (F = 53.23, d.f. = among treatments (F10,150 = 4.42,P <0.0001), we analysed 1,P <0.0001). Again, wounded controls differed from the the change in concentration over the induction (T1) and the untreated control as their concentrations decreased and the challenge (T2) stages. untreated control remained constant (F = 45.41, d.f. = 1,P < For JPBW herbivory induction, over T1 the total monoter- 0.0001). pene concentrations of herbivore–pathogen (2L 3F + 4L 3F) The concentrations of α-pinene, β-pinene, 3-carene, my- (−2.9 ± 3.6%) seedlings and herbivore controls (2L Con + 4L rcene, and limonene were examined over T0, T1,andT2 within Con) (4.5 ± 2.9%) were not different (F = 0.64, d.f. = 1,P = each treatment (Table S2). The general trend is a decrease 0.427) nor were they different from the untreated control (F = in concentrations for herbivore–pathogen seedlings and large 0.41, d.f. = 1, P = 0.521 and F = 1.97, d.f. = 1, P = 0.164, increases forpathogen–herbivore, especially for α-pinene, β- respectively) (Fig. 3a). In contrast, over T2 concentrations pinene, and limonene. In T1, the change in concentration decreased in herbivore–pathogen seedlings (−17.1 ± 3.6%) of α-pinene and β-pinene differed significantly in most of the but greatly increased in the herbivore controls (111.6 ± 9.3%) pooled comparisons (Table S3a). In T2, almost all compounds (F = 297.90, d.f. = 1,P <0.0001). differed significantly in the pooled comparisons (Table S3b). For pathogen infection induction, over T1 both the pathogen–herbivore (3F 2L + 6F 2L) (86.7 ± 36.1%) and the pathogen control (3F Con + 6F Con) (94.2 ± 19.1%) seedlings Relationship among jack pine monoterpenes, Grosmannia had higher total monoterpene concentrations compared with the clavigera, and jack pine budworm untreated control (F = 10.76, d.f. = 1, P = 0.002 and F = 21.58, d.f. = 1, P<0.0001, respectively), but they were not Canonical correlation analysis was used to explore the different from one another (F = 2.10, d.f. = 1,P = 0.540) relationship between the proportional change in the major

3 seedlings except the untreated control. Systemic induction was observed in seedlings inoculated with G. clavigera on 2 the stem and needle monoterpene concentration increased. In seedlings that received a challenge treatment, monoterpene responses differed between trees with herbivory then pathogen 1 infection and trees with pathogen infection then herbivory. 2L 3F These results conﬁrm earlier studies that monoterpenes are 0 4L 3F important components of induced responses in jack pine (Raffa

Biological 1 3F 2L & Smalley, 1995; Wallin & Raffa, 1999) as well as in other -1 6F 2L pine species (see review by Eyles et al., 2010). JPBW herbivory induced systemic resistance in jack pine -2 against G. clavigera, demonstrated by shorter lesions on herbivore–pathogen seedlings compared with their pathogen (only) controls. A similar response between an herbivore and -3 -2 -1 0 1 23 a stem pathogen was observed in Austrian pine (Pinus nigra Chemical 1 Arnold) where previous defoliation by Neodiprion sertifer Geoffory resulted in smaller lesions in response to infection by Fig. 4. Plot of the results of the first canonical function from the the fungus Sphaeropsis sapinea (Eyles et al., 2007). Likewise, canonical correlation analysis. Biological 1 is a variate composed of Tomicus piniperda mean jack pine budworm pupal mass, mean Grosmannia clavigera Scots pine was more resistant to the beetle lesion length, and ln of herbivory rate. Chemical 1 is a variate L. after defoliation by Diprion pini L., even when only 10% composed of the change in α-pinene, 3-carene, and myrcene from foliage remained (Annila et al., 1999). In contrast, Wallin and before the induction stage (T0) to after the challenge stage (T2). Raffa (2001) reported that when more than 26% of mature jack pines were defoliated by JPBW, trees were more susceptible monoterpenes from T0 to T2 (T ) and the biological data to colonisation by a bark beetle Ips pini Say and a woodboring that were observed. Changes in α-pinene, β-pinene, limonene, beetle Monochamus carolinensis Olivier. The discrepancy myrcene, and 3-carene over the experiment were correlated between Wallin and Raffa (2001) and our study may be because (Table S4). Myrcene, α-pinene, and 3-carene were chosen we had lower levels of defoliation compared with that of Wallin for the chemical variate as they were present in almost all and Raffa (2001) or as a result of differences in ontogeny, trees and data were sufficient to produce a good model. Mean as Karban and Niiho (1995) suggested that similar types of pupal mass, mean lesion length, and the natural logarithm of damage could have different effects owing to plant ontogeny. herbivory rate were included as the biological variables. Furthermore, G. clavigera inoculations on jack pine seedlings The correlation for the most significant canonical function were used to mimic MPB colonisation in the present study and was 0.640 (P = 0.0005). The results of the canonical corre- the results may have been different if live beetles were used lation analysis are listed in Table S5 and the canonical scores on mature jack pine trees as in addition to inoculating a tree for the significant canonical functions are listed in Table S6. with fungi, live beetles cause physical damage to the phloem. When the two significant canonical functions were combined, In spite of our use of fungal inoculations, the outcome of our the chemical variates explained 70% of the variability in the research provides a starting point for future research into the chemical variables and 26% of the variability in the biologi- relationship between MPB and JPBW. cal variables. The two biological variates explained 63% and In contrast to JPBW, inoculation of jack pine trees by 23% of the variability in the biological and chemical variables, G. clavigera did not induce the similar systemic plant response respectively. When the first canonical function was graphed, on development of JPBW but increased larval feeding damage. seedlings grouped according to their induction-challenge treat- This result has important implications. First, an increase ments, although there appeared to be more variance in the in the defoliation rate in parallel to the increased tree inoculated-defoliated seedlings (Fig. 4). Examining the canon- defences possibly suggests compensatory feeding by JPBW. ical scores revealed that herbivore–pathogen seedlings gen- Compensatory feeding has been reported for some herbivorous erally had smaller lesions, lower JPBW herbivory rates, and insects in response to low nitrogen (Berner et al., 2005), larger proportional changes in concentration of α-pinene and including the JPBW (McCullough & Kulman, 1991), and 3-carene. In contrast, pathogen-herbivore seedlings had longer high levels of allelochemicals (Barbehenn et al., 2009a,b). lesions, higher herbivory rates, and larger proportional changes Although the possible mechanism(s) of compensatory feeding in the concentration of myrcene. was not explored, it can be partially explained by increased monoterpene concentrations in needles in both the inoculation Discussion and defoliation stages. Additional changes in other groups of chemicals in foliage, such as phenolics (Wallis et al., 2008) The results of the present study indicate that the indirect and alkaloids (Mumm & Hilker, 2006), may also have affected interaction between JPBW and G. clavigera is mediated, at jack pine suitability for JPBW. Further studies are needed to least in part, by host monoterpenes. Needle monoterpene confirm if the differences observed in larval herbivory are in concentrations changed over the experiment in all jack pine fact as a result of an increase in plant secondary metabolites,

© 2011 The Authors Ecological Entomology © 2011 The Royal Entomological Society, Ecological Entomology, 36, 425–434 432 Lindsay J. Colgan and Nadir Erbilgin if so which chemical group is responsible for the increased virgifera virgifera Le Conte in teosinte and cultivated maize. feeding, a reduction in nutrients, or both. No negative effects on larval performance were detected when Alternatively, inspite of the increases in needle monoter- D. virgifera arrived prior to S. frugiperda. To fully understand pene concentrations after G. clavigera inoculation, jack pine jack pine-mediated interactions between JPWB and G. clavig- trees were not better defended against JPBW as larvae sim- era (proxy to MPB), the sequence of attack by either organisms ply responded to these changes by increasing feeding and final should be taken into account in future studies in this system. JPBW pupal mass was unaffected by induction. This suggests In the present study, we used one of the fungal associates that increases in total monoterpene concentrations do not nec- of MPB, G. clavigera as a proxy for MPB-induced changes essarily provide better protection for plants. Some individual on jack pine and provided the first evidence that JPBW compounds may be more important in defence against different herbivory increased jack pine resistance to G. clavigera, enemies (Berenbaum, 1995). For example, large changes in the whereas G. clavigera inoculation had no apparent adverse concentration of myrcene in jack pine as a result of inoculation effect on JPBW growth. Further, the present study suggested may have affected JPBW feeding as it prefers tissues with a that the timing and sequence in which these organisms attack low myrcene content (Wallin & Raffa, 1998). jack pine can be important in determining tree-mediated Another possible implication is that the damage caused by interactions between JPWB and G. clavigera and should be either organism may also have different impacts on plant- taken into account in future studies. In addition, the methods induced chemistry; plants have developed different defence reported here may be a useful proxy for MPB attacks, as mechanisms to protect themselves against pathogens and bark beetle-associated fungi are commonly used to stimulate herbivores (Smith et al., 2009). Although we do not have tree defences (see Lieutier et al., 2009). Finally, although we a complete understanding of defensive pathways induced in provided evidence for the importance of the plant-mediated jack pine by either JPBW or G. clavigera, plant responses interaction between JPBW and G. clavigera, we are very to herbivores and pathogens are mainly regulated by two cautious about inferring that fungal inoculations may reflect hormones. In most species, salicylic acid controls plant the actual field conditions. We do not yet know if the impact of defensive responses to biotrophic pathogens, whereas jasmonic fungal inoculation on jack pine tree defences is similar to those acid-mediated responses are usually directed against herbivores caused by MPB or the role of other fungal associates of MPB and necrotrophic pathogens (Stout et al., 1999; Bostock, 2005; in their effects on jack pine tree defences. Nevertheless, when Heil & Ton, 2008) although both pathways can also be MPB establishes in a jack pine forest, further research on the activated by herbivores alone (e.g. Yang et al., 2011). Further MPB–JPBW interaction on mature jack pine trees might help studies in the jack pine system should identify these pathways answer these questions as well as identify the precise defensive and determine whether induction of one pathway prevents changes induced by MPB or JPBW and consequences of the synthesis of the other; therefore, leading to an impaired these changes on one another. Future research would provide capacity of a plant to respond to either pathogen infection or other important new insights about tree-mediated interactions insect damage (Bostock, 2005). between MPB and JPBW. An initial attack can change the host quality for organisms that follow on the same plant (Hatcher et al., 2004). In the Acknowledgements present study, jack pine trees with pathogen infection then herbivory greatly increased needle monoterpene concentrations We are grateful to Dan Rowlinson and Taylor Scarr (Ontario from their initial levels, whereas trees with herbivory then Ministry of Natural Resources, Sault Ste. Marie, ON) for pathogen infection decreased these concentrations. Monoter- providing JPBW larvae. Inka Lusebrink (University of Alberta) pene responses differed in magnitude and direction depend- provided help for GC analysis. This paper was supported ing on the sequence of attacks by JPBW or G. clavigera in by NSERC Discovery, NSERC CGS-M, Alberta Ingenuity- the induction and challenge stages. The canonical correlation New Faculty Award, and the Faculty of Agriculture, Life analysis demonstrated that the biological characteristics of the and Environmental Sciences, University of Alberta, Edmonton, organisms and the chemical changes in the seedlings separated AB. The critical examinations by two anonymous reviewers are according to pathogen–herbivore and herbivore–pathogen greatly appreciated as their comments substantially improved seedlings. These results suggest that the sequence of attack the quality of this manuscript. could be important in jack pine-mediated interactions between arriving organisms. Similar to our results, Trewhella et al. Supporting Information (1997) reported that the pine beauty moth (Panolis flammea Denis & Schiffermuller)¨ larvae preferred lodgepole pine trees Additional Supporting Information may be found in the online with no previous defoliation by the European sawfly (N. ser- version of this article under the DOI reference: tifer). In contrast, N. sertifer larvae significantly preferred 10.1111/j.1365-2311.2011.01283.x foliage from previously defoliated trees, suggesting that the Table S1. Results of repeated measures anovasoftotal sequence of attack can be important for both species for dif- needle monoterpene concentrations of jack pine sliced by ferent reasons. Likewise, in an annual plant system, Erb et al. treatment over the initial (T0), induction (T1), and challenge (2011) recently reported that defoliation by the fall armyworm, (T2) measurements. Spodoptera frugiperda J. E. Smith reduced colonisation by Table S2. The change in the concentration of individual root-feeding larvae of the western cotton rootworm, Diabrotica needle monoterpenes, α-pinene, β-pinene, 3-carene, myrcene,

© 2011 The Authors Ecological Entomology © 2011 The Royal Entomological Society, Ecological Entomology, 36, 425–434 The interaction of JPBW, G. clavigera and jack pine 433 and limonene in jack pine needles from the initial level (T0) viewpoint. New Zealand Journal of Forest Science (Suppl.), 40, to challenge (T2) (T = [T2 − T0]/T0). Values are means and S15–S24. with standard errors below in parenthesis. Bonello, P. & Blodgett, J.T. (2003) Pinus nigra –Sphaeropsis sapinea Table S3. (a) Pooled comparisons of the change in indi- as a model pathosystem to investigate local and systemic effects vidual needle monoterpene concentrations from initial (T ) of fungal infection of pines. Physiological and Molecular Plant 0 Pathology, 63, 249–261. to induction stage (T ) (T = [T − T ]/T ). Concentra- 1 1 1 0 0 Bonello, P., Gordon, T.R., Herms, D.A., Wood, D.L. & Erbilgin, N. tions were transformed (ln) for analysis. Con, Control; Herb, (2006) Nature and ecological implications of pathogen-induced Jack pine budworm herbivory; Path, Grosmannia clavigera systemic resistance in conifers: a novel hypothesis. Physiological pathogen infection. (b) Pooled comparisons of the change in and Molecular Plant Pathology, 68, 95–104. individual needle monoterpene concentrations from induction Bostock, R.M. (2005) Signal crosstalk and induced resistance: strad- (T1) to challenge stage (T2) (T2 = [T2 − T1]/T1). Concentra- dling the line between cost and benefit. Annual Review of Phy- tions were transformed (ln) for analysis. topathology, 43, 545–580. Table S4. Spearman’s correlation of the change in monoter- Brignolas, F., Lieutier, F., Sauvard, D., Christiansen, E. & Berry- pene concentrations from the initial to challenge stage (T = man, A.A. (1998) Phenolic predictors for Norway spruce resistance Ips typographus [T − T ]/T ) in the induction-challenge treated seedlings (n = to the bark beetle (Coleoptera: Scolytidae) and an 2 0 0 associated fungus, Ceratocystis polonica. Canadian Journal of For- 48). Listed as Spearman’s correlation coefficient and P -value est Research, 28, 720–728. below in italics. (CFS) Canadian Forest Service (2008) Risk Assessment of the Threat of Table S5. Results of canonical correlation analysis of Mountain Pine Beetle to Canada’s Boreal and Eastern Pine Forests. biological (mean pupal mass, lesion length, and ln of damage Canadian Forest Service, Pacific Forestry Center, Victoria, Canada. rate) and chemical variables (change in α-pinene, myrcene, Carroll, A.L., Regni´ ere,` J., Logan, J.A., Taylor, S.W., Bentz, B.J. & and 3-carene over the experiment (T = [T2 − T0]/T0). Powell, J.A. (2006) Impacts of Climate Change on Range Expansion Table S6. Canonical scores (standardised coefficients) for by the Mountain Pine Beetle. Canadian Forest Service, Pacific the significant canonical functions from the canonical correla- Forestry Centre, Victoria, Canada. tion analysis. Monoterpenes listed are the change in monoter- Cerezke, H.F. (1995) Egg gallery, brood production, and adult Dendroctonus ponderosae pene concentrations from before the induction stage (T ) to characteristics of mountain pine beetle, 0 Hopkins (Coleoptera: Scolytidae), in three pine hosts. Canadian after the challenge stage (T ) (T = [T − T ]/T ). 2 2 0 0 Entomologist, 127, 955–965. Please note: Neither the Editors nor Wiley-Blackwell Colgan, L.J. & Erbilgin, N. (2010) The ecological interaction of the are responsible for the content or functionality of any mountain pine beetle and jack pine budworm in the boreal forest. supplementary material supplied by the authors. Any queries Journal of Forestry Chronicle, 86, 766–764. (other than missing material) should be directed to the Erb, M., Robert, C.A.M., Hibbard, B.E. & Turlings, T.C.J. (2011) corresponding author for the article. Sequence of arrival determines plant-mediated interactions between herbivores. Journal of Ecology, 99, 7–15. Erbilgin, N. & Raffa, K.F. (2000) Opposing effects of host monoterpenes on responses by two sympatric species of bark beetles to References their aggregation pheromones. Journal of Chemical Ecology, 26, 2527–2548. Annila, E., Langstr˚ om,¨ B., Varama, N., Hiukka, R. & Niemela,¨ P. Erbilgin, N. & Raffa, K.F. (2001) Modulation of predator attraction (1999) Susceptibility of defoliated Scots pine to spontaneous and to pheromones of two prey species by stereochemistry of plant induced attack by Tomicus piniperda and Tomicus minor. Silva volatiles. Oecologia, 127, 444–453. Fennica, 33, 93–106. Eyles, A., Chorbadjian, R., Wallis, C., Hansen, R., Cipollini, D., Barbehenn, R.V., Jaros, A., Lee, G., Mozola, C., Weir, Q. & Salmi- Herms, D. et al. (2007) Cross-induction of systemic induced resis- nen, J.P. (2009a) Tree resistance to Lymantria dispar caterpillars: tance between an insect and a fungal pathogen in Austrian pine over important and limitations of foliar tannin composition. Oecologia, a fertility gradient. Oecologia, 153, 365–374. 159, 777–788. Eyles, A., Bonello, P., Ganley, R. & Mohammed, C. (2010) Induced Barbehenn, R.V., Jaros, A., Lee, G., Mozola, C., Weir, Q. & Salmi- resistance to pests and pathogens in trees. New Phytologist, 185, nen, J.P. (2009b) Hydrolyzable tannins as “quantitative defenses”: 893–908. limited impact against Lymantria dispar caterpillars on hybrid Franceschi, V.R., Krokene, P., Christiansen, E. & Krekling, T. (2005) poplar. Journal of Insect Physiology, 55, 297–304. Anatomical and chemical defences of conifer bark against bark Beckers, G.J.M. & Spoel, S.H. (2006) Fine-tuning plant defence beetles and other pests. New Phytologist, 167, 353–376. signalling: salicylate versus jasmonate. Plant Biology, 8, 1–10. Hatcher, P.E., Moore, J., Taylor, J.E., Tinney, G.W. & Paul, N.D. Berenbaum, M.R. (1995) The chemistry of defense: theory and (2004) Phytohormones and plant-herbivore-pathogen interactions: practice. Proceedings of the National Academy of Science, 92, 2–8. integrating the molecular with the ecological. Ecology, 85, 59–69. Berner, D., Blanckenhorn, W.U. & Korner, C. (2005) Grasshoppers Heil, M. & Ton, J. (2008) Long-distance signalling in plant defence. cope with low host plant quality by compensatory feeding and food Trends in Plant Science, 13, 264–272. selection: N limitation challenged. Oikos, 111, 525–533. Herms, D.A. & Mattson, M.J. (1992) The dilemma of plants: to grow Bleiker, K.P. & Six, D.L. (2007) Dietary benefits of fungal associates or defend. Quarterly Review of Biology, 67, 283–335. to an eruptive herbivore: potential implications of multiple associates Hudgins, J.W., Christiansen, E. & Franceschi, V.R. (2003) Methyl on host population dynamics. Environmental Entomology, 36, jasmonate induces changes mimicking anatomical defences in 1384–1396. diverse member of Pinaceae. Tree Physiology, 23, 361–371. Bonello, P. (2010) Potential of induced resistance as a tool for the Johnson, M.A. & Croteau, R.A. (1987) Biochemistry of conifer management of pathogens and insects in trees – an ecological resistance to bark beetles and their fungal symbionts. Ecology and

Metabolism of Plant Lipids, A. C. S. Symposium Series, Vol. 325 Rice, A.V., Thormann, M.N. & Langor, D.W. (2007) Virulence of, (ed. by G. Fuller and W. D. Nes), pp. 76–92. American Chemical and interactions among, mountain pine beetle associated blue-stain Society, Washington, District of Columbia. fungi and two species and their hybrids in Alberta. Canadian Johnson, R.A. & Wichern, D.W. (1988) Applied Multivariate Statisti- Journal of Botany, 85, 316–323. cal Analysis, 2nd edn. Prentice Hall, Englewood Cliffs, New Jersey. Rice, A.V., Thormann, M.N. & Langor, D.W. (2008) Mountain pine Karban, R. & Niiho, C. (1995) Induced resistance and susceptibility beetle associated blue-stain fungi are differentially adapted to boreal to herbivory: plant memory and altered plant development. Ecology, temperatures. Forest Pathology, 38, 113–123. 76, 1220–1225. Safranyik, L. & Carroll, A.L. (2006) The biology and epidemiology Kosaka, H., Aikawa, T., Ogura, N., Tabata, K. & Kiyohara, T. (2001) of the mountain pine beetle in lodgepole pine forests. The Mountain Pine wilt disease caused by the pine wood nematode: the induced Pine Beetle a Synthesis of Biology, Management, and Impacts on resistance of pine trees by the avirulent isolates of nematode. Lodgepole Pine (ed. by L. Safranyik and B. Wilson), pp. 3–66. European Journal of Plant Pathology, 107, 667–675. Canadian Forest Service, Pacific Forestry Center, Victoria, Canada. Krokene, P., Christiansen, E., Solheim, H., Franceschi, V.R. & Berry- Safranyik, L., Carroll, A.L., Regni´ ere,` J., Langor, D.W., Riel, W.G., man, A.A. (1999) Induced resistance to pathogenic fungi in Norway Shore, T.L. et al. (2010) Potential for range expansion of mountain spruce. Plant Physiology, 121, 565–569. pine beetle into the boreal forest of North America. Canadian Krokene, P., Nagy, N.E. & Solheim, H. (2008) Methyl jasmonate and Entomologist, 142, 415–442. oxalic acid treatment of Norway spruce: anatomically based defense Smith, J.L., De Moraes, C.M. & Mescher, M.C. (2009) Jasmonate- responses and increased resistance against fungal infection. Tree and salicylate-mediated plant defense responses to insect herbivores, Physiology, 28, 29–35. pathogens and parasitic plants. Pest Management Science, 65, Kulman, H.M., Hodson, A.C. & Duncan, D.P. (1963) Distribution 497–503. and effects of jack pine budworm defoliation. Forest Science, 9, Stout, M.J., Fidanstef, A.L., Duffey, S.S. & Bostock, R.M. (1999) 146–157. Signal interaction in pathogen and insect attack: systemic plant- Latta, R.G., Linhart, Y.B., Lundquist, L. & Snyder, M.A. (2000) mediated interactions between pathogens and herbivores of the Patterns of monoterpene variation within individual trees in tomato, Lycopersicon esculentum. Physiological and Molecular Ponderosa pine. Journal of Chemical Ecology, 26, 1341–1357. Plant Pathology, 54, 115–130. Lieutier, F., Brignolas, F., Sauvard, D., Yart, A., Galet, C., Brunet, M. Tollrain, R. & Harvell, C.D. (1999) The evolution of inducible et al. (2003) Intra- and inter-provenance variability in phloem defenses: current ideas. The Ecology and Evolution of Inducible phenols of Picea abies and relationship to a bark beetle-associated Defenses (ed. by R. Tollrain and C. D. Harvell), pp. 306–321. fungus. Tree Physiology, 23, 247–256. University Press, Princeton, New Jersey. Lieutier, F., Yart, A. & Salle, A. (2009) Stimulation of tree defenses Trewhella, K.E., Leather, S.R. & Day, K.R. (1997) Insect induced by Ophiostomatoid fungi can explain attack success of bark beetles resistance in lodgepole pine: effects on two pine feedings insects. on conifers. Annals of Forest Science, 801, 1–22. Journal of Applied Entomology, 121, 129–136. Logan, J.A., Jacques, R. & Powell, J.A. (2003) Assessing the impacts Wallin, K.F. & Raffa, K.F. (1998) Association of within-tree jack of global warming on forest pest dynamics. Frontiers in Ecology pine budworm feeding patterns with canopy level and within- and the Environment, 1, 130–137. needle variation of water, nutrient, and monoterpene concentration. McCullough, D.G. (2000) A review of factors affecting the population Canadian Journal of Forest Research, 28, 228–233. dynamics of jack pine budworm (Choristoneura pinus pinus Wallin, K.F. & Raffa, K.F. (1999) Altered constitutive and inducible Freeman). Population Ecology, 42, 243–256. phloem monoterpenes following natural defoliation of jack pine: McCullough, D.G. & Kulman, H.M. (1991) Effects of nitrogen implications to host mediated interguild interactions and plant fertilization on young jack pine (Pinus banksiana) and on its defence theories. Journal of Chemical Ecology, 25, 861–880. suitability of jack pine budworm (Choristoneura pinus pinus) Wallin, K.F. & Raffa, K.F. (2001) Effects of folivory on subcortical (Lepidoptera: Torticidae). Canadian Journal of Forest Research, 21, plant defences: can defence theories predict interguild processes? 1447–1458. Ecology, 82, 1387–1400. Mumm, R. & Hilker, M. (2006) Direct and indirect chemical defence Wallis, C., Eyles, A., Chorbadjian, R., McSpadden Gardener, B., of pine against folivorous insects. Trends in Plant Science, 11, Hansen, R., Cipollini, D. et al. (2008) Systemic induction of phloem 351–358. secondary metabolism and its relationship to resistance to a canker Raffa, K.F. & Smalley, E.B. (1995) Interaction of pre-attack and pathogen in Austrian pine. New Phytologist, 177, 767–778. induced monoterpene concentrations in host conifer defense against Wood, S.L. (1982) The Bark and Ambrosia Beetles of North and bark beetle-fungal complexes. Oecologia, 102, 285–295. Central America (Coleoptera: Scolytidae), a Taxonomic Monograph. Raffa, K.F., Krause, S.C. & Reich, P.B. (1998) Long-term effects of Brigham Young University, Provo, Utah. defoliation on red pine suitability to insects feeding on diverse plant Yang, J.W., Yi, H.-S., Kim, H., Lee, B., Lee, S., Ghim, S.-Y. et al. tissues. Ecology, 79, 2353–2364. (2011) Whitefly infestation of pepper plants elicits defence responses Raffa, K.F., Aukema, B.H., Erbilgin, N., Klepzig, K.D. & Wallin, K.F. against bacterial pathogens in leaves and roots and changes the (2005) Interactions among conifer terpenoids and bark beetles across below-ground microflora. Journal of Ecology, 99, 46–56. multiple levels of scale: an attempt to understand links between Zhao, W., Chen, S.P. & Lin, G.H. (2007) Compensatory growth population patterns and physiological processes. Recent Advances responses to clipping defoliation in Leymus chinensis (Poaceae) in Phytochemistry, 39, 80–118. under nutrient addition and water deficiency conditions. Plant Rasband, W.S. (2009) ImageJ: Image Processing in Java. National Ecology, 196, 85–99. Institute of Health, Besthesda, Maryland. Accepted 26 March 2011 Regni´ ere,` J. & Bentz, B. (2007) Modeling cold tolerance in the mountain pine beetle, Dendroctonus ponderosae. Journal of Insect First published online 1 June 2011 Physiology, 53, 559–572.