Oecologia (2007) 153:365–374 DOI 10.1007/s00442-007-0741-z

PLANT ANI MAL INTE RA CTIONS

Cross-induction of systemic induced resistance between an insect and a fungal pathogen in Austrian pine over a fertility gradient Alieta Eyles · Rodrigo Chorbadjian · Chris Wallis · Robert Hansen · Don Cipollini · Dan Herms · Pierluigi Bonello

Received: 21 December 2006 / Accepted: 23 March 2007 / Published online: 24 April 2007 © Springer-Verlag 2007

Abstract Evidence for cross-induction of systemic resistance or susceptibility in plant–fungus–herbivore interactions is mostly derived from herbaceous model systems and not perennial woody plants. Furthermore, the eVects of environmental variables such as soil fertility on these tripartite interactions are generally unknown. This study examined cross-induction of systemic resistance in Pinus

Communicated by Judith Bronstein. Alieta Eyles and Rodrigo Chorbadjian contributed equally to the paper. A. Eyles (&) · C. Wallis · P. Bonello Department of Plant Pathology, Ohio State University, 201 Kottman Hall, 2021 CoVey Road, Columbus, OH 43210, USA e-mail: [email protected] R. Chorbadjian · D. Herms Department of Entomology, Ohio Agricultural Research and Development Center, Ohio State University, 1680 Madison Ave, Wooster, OH 44691, USA Present Address: R. Chorbadjian Departamento de Ciencias Vegetales, Facultad de Agronomía e Ingeniería Forestal, PontiWcia Universidad Católica de Chile, Casilla 306-22, Santiago, Chile R. Hansen Department of Food, Agriculture and Biological Engineering, Ohio State University, 108 Agricultural Engineering Building, 1680 Madison Avenue, Wooster, OH 44691, USA D. Cipollini Department of Biological Sciences, Wright State University, 208 Biological Sciences Building, 3640 Colonel Glenn Hwy, Dayton, OH 45435, USA

nigra (Austrian pine) to infection by Sphaeropsis sapinea (a fungal pathogen), or feeding by Neodiprion sertifer (European pine sawXy), by prior induction with either S. sapinea or N. sertifer, over a fertility gradient. In a replicated 3-year study, cross-induction of systemic induced resistance (SIR) was found to be both asymmetric within a single year and variable between years. Prior induction with insect defoliation induced SIR to subsequent fungal challenge in 2006 but not in 2005. In 2005, a fertility-independent negative systemic eVect of the fungal infection on herbivore growth was detected while herbivore survival was aVected by a signiWcant interaction between induction treatment and fertility level in 2006. Prior infection by the fungus induced SIR against the same fungus in both years regardless of fertility levels. This is the Wrst report of whole-plant SIR against a defoliating insect induced by a fungal pathogen and vice versa, under variable nutrient availability, in a conifer or any other tree. Keywords Conifer · Fungal induction · Insect defoliation · Defense induction · Resistance

Introduction Trees possess constitutive as well as inducible structural and chemical defenses against both insects and pathogens (Hammerschmidt and Schultz 1996; Pearce 1996; Larsson 2002). In several conifers, pathogenic infection and attack by bark beetles result in localized accumulation of pathogenesis-related proteins (Ekramoddoullah et al. 2000; Smith et al. 2006), phenolics and terpenoids (the latter are often associated with the formation of traumatic resin ducts) (Franceschi et al. 2005; Erbilgin et al. 2006; Zeneli et al. 2006).

123

366

In many host–pathogen systems, prior localized interactions with pathogens have been shown to induce pathogen resistance in previously non-infected parts of the plant, a phenomenon called systemic induced resistance (SIR) (Bonello et al. 2001; Kuc 2001; Bonello and Blodgett 2003). In model hosts (mostly herbaceous model species, particularly tomato, tobacco, and Arabidopsis), development of SIR begins with a hypersensitive response to the initial infection and is mediated by the accumulation of salicylic acid, jasmonic acid and/or ethylene (Stout et al. 1998, 1999; Durrant and Dong 2004). Other hormones, such as abscisic acid, can also play a role in SIR (Dammann et al. 1997; Bostock 1999). SpeciWcally, enhanced resistance to biotrophic pathogen infection appears to be mediated by salicylic acid while enhanced resistance to insect feeding, via the wound response pathway, is mediated by jasmonic acid. However, evidence suggests that this distinction may not be so well deWned and cross-talk may occur among signaling pathways (Bostock 1999; Paul et al. 2000; Durrant and Dong 2004; Hatcher et al. 2004; Beckers and Spoel 2006). SIR has been observed in pine in response to plantgrowth-promoting rhizobacteria (Enebak and Carey 2000) and pathogens (Bonello et al. 1991; Blodgett et al. 2007). However, limited evidence suggests that the signaling pathways in pine when attacked by necrotrophic pathogens might diVer from those of most herbaceous plants/pathogen systems studied to date. For example, local and systemic changes in phenolic composition of Scots pine (Pinus sylvestris L.) needles and ponderosa pine (Pinus ponderosa Dougl. ex Laws.) phloem in response to a root pathogens (Bonello et al. 1993; Bonello et al. 2003, respectively) and Austrian pine (P. nigra Arnold) needles and phloem in response to a canker pathogen, Sphaeropsis sapinea (Fr.:Fr.) Dyko and Sutton [syn. Diplodia pinea (Desm.) Kickx–Sutton 1980] (Bonello and Blodgett 2003; Blodgett et al. 2007) were never associated with the accumulation of salicylic acid, a rather common theme in herbaceous hosts. However, exogenous treatment of conifer tissues with either salicylic acid and its derivatives, or methyl jasmonate, can result in induction of expression of some defense-related genes (e.g., Davis et al. 2002) and enhanced localized resistance to pathogens and bark beetles (e.g., Reglinski et al. 1998; Franceschi et al. 2002; Hudgins et al. 2003; Schmidt et al. 2005; Erbilgin et al. 2006; Zeneli et al. 2006). Experiments by Bonello and Blodgett (2003), Luchi et al. (2005), and Blodgett et al. (2007) conWrmed that SIR occurs in the Austrian pine/S. sapinea model pathosystem. In these studies, prior infection by the pathogen resulted in SIR to S. sapinea, as evidenced by signiWcantly reduced challenge lesions in the main stem of inoculated trees compared with trees receiving mock inoculations (Blodgett et al. 2007). SIR in Austrian pine is associated with clear systemic eVects on phenolic and resin metabolism (Bonello

123

Oecologia (2007) 153:365–374

and Blodgett 2003; Luchi et al. 2005; Blodgett et al. 2007), the production of diVerentially expressed proteins (Wang et al. 2006) and is dependent upon the relative locations of the inducing and subsequent challenge inoculations (Blodgett et al. 2007). SpeciWcally, when the base of the main stem was induced with S. sapinea, SIR to the same pathogen occurred in the stem, while systemic induced susceptibility occurred in the shoot tips. Finally, the phenomenon was shown to be bidirectional, indicating that the long-distance transport of signal molecules is both acropetal and basipetal (Bonello and Blodgett 2003; Blodgett et al. 2007). Recent reviews reveal little evidence that fungal infection aVects either the behavior or performance of herbivorous insects on a systemic scale (Rostas et al. 2003). For example, elicitation of SIR by scab fungus infection in cucumber had no eVect on several arthropods (Ajlan and Potter 1991; Moran 1998). In contrast, the few studies of direct eVects of insect-induced SIR on resistance to pathogens frequently demonstrate systemic inhibition of fungal growth resulting from the feeding activity of herbivores (Rostas et al. 2003). It has been shown that in the intensively studied Rumex spp. attacked by the green dock beetle, Gastrophysa viridula (Col.: Chrysomelidae) and the biotrophic rust fungus Uromyces rumicis, beetle feeding signiWcantly reduces rust infection, both locally and systemically (Hatcher et al. 1994) but it is unclear what mechanisms are responsible for the inhibitory eVect of herbivore feeding on subsequent fungal infection. Very few tripartite studies investigating cross-induction of SIR have been documented for trees (Bonello et al. 2007). McNee et al. (2003) demonstrated negative systemic cross-eVects of infection of ponderosa pine with the root pathogen, Heterobasidion annosum (Fries) Brefeld on the feeding by the pine engraver beetle, Ips paraconfusus Lanier (Col: Scolytidae), but this study used logs of infected trees rather than standing infected trees. Simon and Hilker (2003) provided some evidence that feeding by the willow leaf beetle Plagiodera versicolora Laicharting (Col.: Chrysomelidae) increased its systemic susceptibility toward rust infection. The eVect of soil fertility on cross-eVects of systemic resistance has received little attention. The majority of studies have addressed eVects of soil fertility on constitutive secondary metabolism and insect resistance, and their results have varied (Kyto et al. 1996; Koricheva et al. 1998; Hatcher et al. 1997), possibly because nutrient availability may have non-linear eVects on these responses (Herms 1999, 2002; Herms and Mattson 1992). Relative to insect herbivores, even less is known about eVects of nutrient availability on tree resistance to pathogens, although increased N availability has been associated with increased susceptibility of plants to pathogens (Snoeijers et al. 2000; Herms 2002). Only a few studies have focused on foliar

Oecologia (2007) 153:365–374

and canker pathogens of conifers. For example, severity of infection of Monterey pine (P. radiata D. Don) by the Dothistroma needle cast fungus increased with N fertilization (Lambert 1986) while mortality of red pine (P. resinosa Ait.) attributed to S. sapinea was linked to higher foliar N (Stanosz et al. 2004). Finally, susceptibility of red pine to S. sapinea was shown to increase with increased nutrient availability (Blodgett et al. 2005). To our knowledge no studies exist on systemic crossresistance between plant pathogens and herbivores on Pinus spp. We used the Austrian pine/S. sapinea/European pine sawXy (Neodiprion sertifer; Hym. Diprionidae) model system to investigate these questions. Austrian pine is native to the Mediterranean basin where it is a dominant forest species. It is highly susceptible to Sphaeropsis shoot tip blight and canker (also known as Diplodia blight) and is readily defoliated by N. sertifer in its native range. The fungal pathogen, S. sapinea (Coelomycetes) aVects hosts in several coniferous genera and has caused extensive damage to pines and other conifers throughout the world (Blodgett et al. 2003). N. sertifer is an important outbreak defoliator of a number of pine species in Europe (Larsson et al. 2000) and North America (Drooz 1985). N. sertifer is univoltine and larvae feed during a short period of time in spring (typically from mid April to May in Ohio) exclusively on needles produced during the previous year. It is sensitive to host quality, and outbreaks are considered most likely when environmental factors weaken plant defenses (Larsson et al. 2000). This multi-component, highly replicated and repeated, outdoor 3-year study, aimed to link host physiological changes induced by the insect and fungal attack with eVects on the antagonists by simultaneously quantifying a suite of constitutive and systemically induced biochemical responses (i.e., lignin, several classes of phenolics, terpenes, and resistance proteins) in relation to whole plant physiology across a nutrient gradient. The present study reports rapid phenotypic responses to in planta induction and challenge treatments, while host physiological and defense responses will be reported elsewhere. We aimed to: (1) test whether inoculation of Austrian pine with the shoot and tip blight pathogen, S. sapinea induces systemic resistance to subsequent challenge with either the same pathogen or N. sertifer; (2) test whether defoliation by N. sertifer induces systemic resistance to subsequent challenge with either the same insect or S. sapinea; and (3) determine whether the strength of the constitutive or induced responses varies across a nutrient gradient.

Materials and methods The experiment was conducted at the Landscape Nursery Crop Engineering Research Laboratory of Ohio State

367

University’s, Food, Agricultural and Biological Engineering Department, Ohio Agricultural Research and Development Center (OARDC), Wooster, Ohio (40°81⬘N, 81°94⬘W). Plants Four-year-old Austrian pine saplings from unknown, open pollinated sources, were obtained from Ridge Manor Nursery (Madison, Ohio) and potted in 21-l containers using a commercial nursery substrate (KB container mix; Kurtz Bros., Ohio). In May 2004, two hundred and forty potted pines were assigned based on size [15.7% coeYcient of variation (CV) in initial diameter] to Wve blocks and positioned on a gravel bed. The saplings were exposed to ambient weather conditions. In June 2005, a new group of 120 six-year-old Austrian pines were assigned to each of the Wve blocks (10.7% CV in initial diameter) to receive the same fertility treatments. Experimental design The 3-year (2004–2006) experiment used a randomized complete block design with three levels of nutrient availability applied in a factorial combination with four induction treatments. Each fertility level was replicated 80 times, with each level divided into four induction treatments (see below). Each induction treatment (n = 20) was divided again into seven replicates for the insect challenge and ten for the pathogen challenge. The other three replicates were used for complementary phytochemical analysis, the results of which are to be published elsewhere. The 240 potted trees were preconditioned to the three fertility levels during 2004, then induced and challenged in 2005. The experiment was repeated with another group of 120 Austrian pines preconditioned in 2005 with the same three fertility levels, and in the following year, subjected to three induction treatments (mock inoculation was not included) and challenged. In this case, each fertility level was replicated 40 times, and each level divided into the three induction levels. Across fertility levels, fungal, and insect inductions were replicated 42 times each, and the control 36 times. Each combination was further divided in half for challenges with the insects (n = 60) and pathogen (n = 60). Fertility treatment During each growing season (bud expansion in early April until needle abscission in early October 2004–2006), three fertility levels were maintained via fertigation (i.e., nutrient application through irrigation) using 30, 75, or 150 p.p.m. N, with N, P2O5, and K2O supplied in solution at a ratio of 3:1:2, and other essential nutrients supplied in non-limiting quantities. These treatment levels were selected because

123

368

they correspond with low to high fertility regimes recommended for use in forest nursery production of containerized evergreen trees (Landis et al. 1989). Fertility treatments were applied by a computer-controlled nutrient delivery system via drip irrigation that was driven by calculated rates of evapotranspiration, allowing precise control of experimental conditions (Lee et al. 2000). Induction treatments Four induction treatments were applied in 2005 in factorial combination with the three fertility levels: (1) fungal inoculation, (2) mock inoculation, (3) insect defoliation, and (4) unwounded control. In 2005, all four inductions were applied on 10 May while in 2006, inductions 1, 3, and 4 were applied on 25 April. This timing was selected in relation to the natural N. sertifer feeding period in the Wooster area. 1. Fungal inoculation: at the beginning of bud expansion, trees were inoculated with S. sapinea on the stem 5 cm above the soil. BrieXy, the stem was wounded by removing the bark/phloem with a sterile cork borer (10 mm diameter). About 8-mm-diameter potato dextrose agar (PDA) plugs colonized with the pathogen (isolate 3AP) were placed mycelium-side down on the wounds. Plugs were obtained from margins of actively growing cultures incubated for at least 7 days in the dark at room temperature. Duct tape was wrapped around the stems at the inoculation site to retain the inoculum and limit desiccation and contamination. 2. Mock inoculation: to control for potential eVects of wounding alone on induced responses, trees were wounded as in the fungal induction but non-colonized PDA plugs were applied. Since mock induction did not have any signiWcant eVects on insect or pathogen response in 2005 (as found in previous studies, e.g., Bonello and Blodgett 2003), this treatment was omitted in 2006. 3. Insect defoliation: Neodiprion sertifer larvae were collected from mugo pines (Pinus mugo Turra) located at Wooster in 2005, and at Ohio State University, Columbus, Ohio in 2006. Austrian pines were defoliated naturally by introducing »150 third and fourth instar N. sertifer larvae to each treatment tree to achieve targeted defoliation levels (visually estimated to be around 75%). To introduce the larvae, small pine shoots (8 cm) detached from other plant material and bearing groups of larvae were tied to the trees and the larvae migrated readily to the foliage on the target tree. All trees were inspected on a daily basis to conWrm that the insects remained feeding gregariously on the assigned tree. Larvae fed on the trees for 8 days in 2005, and 13 days

123

Oecologia (2007) 153:365–374

in 2006, when the targeted defoliation level was achieved. Thereafter, larvae were removed from the trees by hand and the trees remained free of insects for the full length of the incubation period. Prior to introduction of larvae, three randomly selected branches on the same whorl on each tree were conWned in mesh bags to exclude herbivory. These branches were later used in challenge bioassays to monitor any systemic responses to defoliation. In addition, three branches of all other trees were bagged to control for any potential eVect of bagging even though previous studies that speciWcally tested for eVects of mesh sleeves on foliar chemistry or herbivore performance found no evidence of bagging eVects (Rossiter et al. 1988; Hanhimaki and Senn 1992). 4. Unwounded control: no induction treatment was applied. Challenge treatments Challenge bioassays were carried out in planta to test for rapidly induced systemic eVects of the induction treatments on subsequent fungal lesion length and subsequent larval growth and survival. In both years, pathogen and sawXy challenges were started concurrently and after a similar incubation period. This was based on growing degree days (GDD), or daily hours between a lower and an upper threshold of 5°C and 30°C, respectively. GDDs were calculated using the modiWed sine wave method (www. ipm.ucdavis.edu/WEATHER/ddretrieve.html). Threshold temperatures were chosen because they represent N. sertifer response to temperature (Regniere 1984), and were assumed to be appropriate for pines, which also begin growing early in the spring. The incubation period in 2005 was for 140 GDD, which corresponded to 16 calendar days, while in 2006 it was for 163 GDD (21 calendar days). Previous studies have shown that SIR to the pathogen is observable after 3 weeks of induction (Bonello and Blodgett 2003; Blodgett et al. 2007). 1. Pathogen challenge: one branch was inoculated with the same S. sapinea isolate used for the induction at a point 10 cm away from the main stem on a standardized whorl (using the same inoculation method as described above). After the incubation period, inoculated branches were pruned and lesion lengths resulting from the challenge inoculations measured. Lesion length is considered an appropriate estimate of relative host resistance in this and other canker systems (Blodgett et al. 2007 and references therein). 2. SawXy challenge: for each experimental tree unit, a group of ten early third instar N. sertifer larvae (the experimental unit since they feed gregariously) were

Oecologia (2007) 153:365–374

369

pre-weighed to determine initial mass, conWned to a branch within a mesh bag (to exclude natural enemies) and allowed to feed for 7 days from 27 May to 3 June 2005, and for 10 days from 16 to 26 May 2006. Although a longer feeding time was used in 2006, the length of the challenge bioassay was very similar for both years in terms of GDD and as a result, very similar insect growth responses in the control trees were obtained in both years (Fig. 1). The length of the feeding period was carefully chosen in order to measure growth of larvae before they entered a non-feeding stage (eonymph), hence larvae were actively feeding throughout the course of the bioassay. Following termination of the bioassays, any mortality was recorded and survivors re-weighed as a group. Treatment eVects on percent survival and growth of surviving larvae were determined. Mean larval weight was calculated by dividing total Wnal weight by number of surviving larvae. Larval growth (mg/larva) was calculated as the diVerence between mean Wnal and mean initial mass. Weather conditions Prevailing weather conditions for 2005 and 2006 were obtained from the OARDC weather station. Total precipitation in 2006 (15.9 mm) was more than double that in 2005 (7.7 mm). Trees were exposed to cooler average daily air temperatures in 2006 (14.1°C) than 2005 (17.9°C). Data analyses EVects of block, induction, fertilization, and their interactions on insect growth, survival and on challenge fungal 2005

35

a

2006

a a

Insect growth (mg/larva)

30

b

25 20 15 10 5 0 Control

Mock inoculation

Insect defoliation

Results EVects of fungal and herbivore induction on fungal challenge In 2005, basal stem inoculations with S. sapinea induced a 42% reduction in fungal challenge lesion lengths when compared to the control induction treatments (F3,102 = 9.41, P < 0.001) (Fig. 2). In 2006, S. sapinea induction resulted in a 33% decrease in the length of challenge lesion lengths compared to non-wounded trees (F2,59 = 8.10, P = 0.004) (Fig. 2). In 2006, insect defoliation induced a 19% reduction in challenge lesion lengths compared to corresponding non-wounded controls (Fig. 2). Fertilizer treatment and fertilizer £ induction interactions were not signiWcant in either year (fertilizer, F2,102 = 0.19, P = 0.828 for 2005; F2,59 = 1.33, P = 0.292 for 2006; interaction, F4,102 = 0.34, P = 0.913 for 2005; F4,59 = 0.84, P = 0.517 for 2006). No signiWcant block eVects or block interactions were seen for either year.

a

a

a

lesion lengths were assessed by ANOVA (PROC GLM, type III sums of squares; SAS Institute 2003 or SPSS 14.0 for Windows (SPSS, Chicago, Ill.). Fertilization and induction treatments were considered Wxed factors. All variables were tested for homoscedasticity and normal distribution of residuals. Only the percentage of insect survival required an arcsin square-root transformation to meet the assumption of homoscedasticity. Initial larval weight (mg/larva) was included as a covariate in the model when analyzing larval growth for 2005, because it added signiWcant explanation of the variability (F1,94 = 7.25, P = 0.008). SigniWcant diVerences among means were detected using Fisher’s protected LSD test.

Fungal inoculation

Induction treatment

Fig. 1 EVects of induction treatment on European pine sawXy (Neodiprion sertifer) larval growth for 2005 and 2006 (averaged over the three fertility levels). Within years, bars with the same letter are not signiWcantly diVerent (P > 0.05). Bars are mean § 1 SEM

EVects of fungal and herbivore induction on insect challenge In general, N. sertifer response to induction treatments was variable between treatments and years. In 2005, induction treatments had a statistically signiWcant eVect over N. sertifer larval growth (mg/larva) (F3,94 = 6.26, P < 0.001). Prior fungal induction reduced N. sertifer growth by 19% compared to the corresponding control, while larval growth on trees previously defoliated was not signiWcantly diVerent than on control trees (Fig. 1). Fertility level had a small positive eVect on N. sertifer growth (F2,94 = 2.83, P = 0.064). Across the induction treatments, the highest fertility level increased larval growth by 10% with respect to the lowest fertility level (data not shown). Larval growth was not aVected by the interaction between induction and fertility level (F6,94 = 1.18, P = 0.326). Neither the induction,

123

370

Oecologia (2007) 153:365–374

25

2005

a a

a

100

15

10

5

Control

Insect defoliation

Fungal Mock inoculation inoculation

a

b

60

40

2006

b

40

0 30

a

45

Lesion length (mm)

ab

20

0

50

Insect defoliation

80

b

Insect survival (%)

Lesion length (mm)

20

Control Fungal inoculation a a ab ab b b

35

b

30 25

75

150

Fertility level (ppm Nitrogen)

Fig. 3 EVects of induction treatment on European pine sawXy (N. sertifer) survival for 2006. Same letter above bars indicates no signiWcant diVerences between treatment levels (P > 0.05). Bars are mean § 1 SEM. Austrian pines were grown under the indicated rates of N delivery via fertigation, with N, P, and K applied in a ratio of 3:1:2

20 15 10 5 0

Control

Insect defoliation

Fungal inoculation

Induction treatment

Fig. 2 EVects of induction treatment on subsequent fungal challenge lesion length in Austrian pine branches. The same letter above bars indicates no signiWcant diVerence between induction treatments (P > 0.05). Bars are mean § 1 SEM

nor the fertility treatments or their interaction aVected N. sertifer survival in 2005 (F · 0.91, P ¸ 0.440) (data not shown). In contrast, N. sertifer growth in 2006 was not aVected by the induction treatments (F2,47 = 0.08, P = 0.926), fertility level (F2,47 = 2.06, P = 0.139), or their interaction (F4,47 = 0.64, P = 0.634). However, the eVect of the induction treatments over larval survival in 2006 varied with fertility level (F4,47 = 4.21, P = 0.005). SpeciWcally, in the lowest fertility level (i.e., 30 p.p.m. N), both fungal inoculation and insect defoliation decreased insect survival by at least 14% over the control, whereas at the highest fertility level (i.e., 150 p.p.m. N), insect survival increased by 21% over the control (Fig. 3).

Discussion This study showed that systemic cross-eVects of infection by S. sapinea and defoliation by N. sertifer in Austrian pine

123

were asymmetric within a single year and variable between years. SpeciWcally, induction with N. sertifer produced variable SIR responses to the fungal challenge, with no eVect in 2005 and a signiWcant induction of SIR to fungal challenge in 2006. Conversely, induction with S. sapinea resulted in increased resistance to both N. sertifer defoliation and pathogenic infection in 2005, but only to pathogenic infection in 2006. The study also showed that soil fertility exerted complex eVects on resistance to the insect, but not the pathogen. Such systemic cross-eVects between an insect and a fungal pathogen have not been previously examined in any coniferous species. Asymmetric plant-mediated responses have been found in other tripartite interactions involving hosts other than conifer species (Hatcher et al. 1994; Rostas et al. 2002; Rostas and Hilker 2003). In contrast to our Wndings, Simon and Hilker (2003) showed that prior insect herbivory on willow increased the host’s susceptibility toward rust infection whereas prior rust infection had a systemic negative eVect on larval performance in a willow–rust–leaf beetle tripartite interaction. This discrepancy may be explained, in part, by the very diVerent tripartite systems used in the two studies. We suggest, however, that the results found in the present study are not directly comparable to those reported in other tripartite studies. For example, one key diVerence between our study and many others studies is the use of a coniferous rather than an herbaceous host. There is increasing evidence to suggest that for conifers, the biochemical and molecular basis of SIR are distinct compared to angiospermous models (Hudgins et al. 2004; Piggott. et al. 2004; Luchi et al. 2005; Blodgett et al. 2007). For example, SIR in Austrian pine has never been linked with endogenous

Oecologia (2007) 153:365–374

accumulation of salicylic or jasmonic acid and has been associated with an eightfold increase in resin Xow as well as signiWcant accumulation of several phenolic compounds and lignin in response to infection by S. sapinea (Luchi et al. 2005). Another major diVerence between this study and others is the methodology used. In many of the reported experiments, systemic eVects were conducted using either detached leaves or punched-out leaf discs rather than whole attached organs (Ajlan and Potter 1991; Moran 1998; Rostas et al. 2002; Simon and Hilker 2003). In these experiments, it is unknown to what extent this type of bioassay may aVect the systemic responses when compared with in planta bioassays. Interestingly, the systemic eVects of fungal infection on the herbivore, and vice versa, were variable ranging from no eVects in Chinese cabbage leaves (Rostas and Hilker 2003) to reciprocal systemic eVects in willow (Simon and Hilker 2003). The data presented here suggest that when investigating cross-eVects over an extended period of time, the use of in planta bioassays may provide more meaningful data than the use of detached organs, particularly given the temporal and spatial nature of SIR. Furthermore, in Simon and Hilker’s paper (2003), eVects were measured on adjacent leaves in laboratory and greenhouse assays, whereas the eVects we observed in this study were: (1) at the whole-plant level, (2) on diVerent tissue types (needles and phloem), (3) in a Weld experiment that included variable nutrient availability, and that (4) was replicated over two seasons. Indeed, in this study, the asymmetric responses were observed to vary between years, demonstrating the need to conduct multi-year studies. Several factors could account for the diVerences observed between the 2 years of the study. In particular, diVerences in tree age and size may have played a role in modifying SIR expression. While every attempt was made to ensure that all variables were comparable between each year, for some factors this was not possible. SpeciWcally, older trees were used in 2006 and larvae were collected from diVerent locations each year. Also, prevailing weather diVered between years. In their review of cross-eVects of induced plant responses, Rostas et al. (2003) concluded that the lag time between an induction and challenge event might be critical to detect the expression of cross-eVects on a systemic scale. In the present study, the challenges were conducted after an induction period of 16 and 21 days in 2005 and 2006, respectively. Although the number of cumulative degree days remained fairly constant between years, other climatic factors such as the amount of rain might have played a role. In previous experiments using the same Austrian pine/ S. sapinea pathosystem, this induction period has resulted in the consistent expression of SIR in greenhouse experiments (Blodgett et al. 2007). Given that insect defoliation only

371

induced SIR to subsequent fungal challenge (and only in 2006), we suggest that the induction period required to induce SIR by insect damage may be diVerent from that for fungal infection and may in part explain the observed variability. In other studies, there is clear evidence to suggest a temporal factor in systemic responses. For example, Hatcher et al. (1994) demonstrated that in the dock–rust–chrysomelid tripartite interaction, maximum systemic inhibition of rust pustule density occurred 7 days after cessation of insect grazing and not 1 and 3 days. Future studies should include a time-course study to determine the appropriate lag time between herbivore damage and fungal infection that permits the cross-induction of SIR in Austrian pine. Finally, we suggest that the magnitude of systemic response to induction treatments was likely aVected by the route through which systemic signals moved from the induction to challenge locations and the connectivity among these organs (Herms and Mattson 1992; Orians 2005). For example, systemic signals from the fungal induction site to the foliage could have moved through either the phloem or the xylem, while systemic signals from the damaged foliage to either the stem phloem or foliage of other branches could have only moved through the phloem. In turn, the vascular connectivity between induced and challenge branches would inXuence the strength of induced responses. Thus, the generally stronger eVects of fungal induction of the phloem on foliage chemistry and challenge bioassays than the reverse may have resulted from constraints due to vascular architecture and signal transport routes, along with the identity of the inducing and challenge organisms. One notable response that was consistently observed was the elicitation of SIR by the fungal pathogen at the base of the stem against the same fungal pathogen on the branches. Fungal challenge lesions were approximately double the length in 2006 compared to 2005, and lesion lengths in the uninduced controls grew nearly 3 times faster in 2006 than in 2005 (0.20 and 0.07 cm/GDD in 2006 and 2005, respectively). However, this demonstrated that induction of SIR by fungi against fungi may be robust even under varying environmental conditions, which in our study also included variable nutrient availability. Induction of SIR by fungal pathogen to subsequent pathogen infection has been observed in other pathosystems (e.g., Moran 1998) but this response does not appear to be a consistent phenomenon across all pathosystems. For example prior infection resulted in increased susceptibility to subsequent fungal infection in other studies, including one on Austrian pine when the shoots were challenged instead of the stem (Hatcher and Paul 2000; Simon and Hilker 2003; Blodgett et al. 2007). Fertility treatment had a diVerent eVect on insect response each year. In 2005, higher fertility tended to

123

372

increase larval growth, which is consistent with other studies showing that the growth rates of many insects, including sawXies (Herms 2002; Saxon et al. 2004) are directly related to the N content of their food (Mattson 1980; Hogendorp et al. 2006; Kerpel et al. 2006). Although not presented here, larval growth rates showed the same response to the treatments as larval growth. On the other hand, fertility level in 2006 strongly aVected the percentage of insects surviving at each induction level. This showed that previous defoliation may rapidly enhance larval survival under environments without nutrient limitations, but have the opposite eVect under nutrient limited environments. Similarly, Glynn et al. (2003) showed that fertility level and previous defoliation may interact to aVect whitemarked tussock moth (Orgyia leucostigma: Lymantriidae) in poplar (Populus nigra L.). The mechanisms by which N. sertifer feeding inhibits subsequent infection by S. sapinea and vice versa remain unclear. Pathogens and herbivores are known to induce a great variety of local and systemic changes in plant primary (e.g., carbohydrate, N, protein, and sugars) and secondary (e.g., phenolics, terpenes, and defense enzymes) biochemistry (Hammerschmidt and Dann 1999; Bonello and Blodgett 2003; Cipollini et al. 2004; Saxon et al. 2004; Blodgett et al. 2007). Only a few studies on tripartite systems have attempted to couple physiological changes of plants caused by either herbivores or fungal pathogens with cross-eVects on the antagonists (Hatcher 1995; Siemens and MitchellOlds 1996; Moran 1998; Fidantsef et al. 1999; Stout et al. 1999; Rostas et al. 2002; Rostas and Hilker 2003). In those studies, there was conXicting evidence on whether systemic eVects induced by herbivore feeding or fungal infections on herbivore or fungal performance are mediated by an induced response or a change in nutritional quality of the plant. As suggested by Hatcher (1995) and Rostas et al. (2003), both nutrient allocation due to plant attack and cross-interactions between induced plant defensive mechanisms may be important factors for the outcome of crosseVects between a given herbivore and pathogen species. A lack of studies has made it diYcult to discern when, where, and how positive or negative reciprocal eVects might occur as a result of the induction of SIR by either party. Furthermore, predictability in the case of conifers is further hampered by a general lack of studies with woody plants. Our results indicate that the outcome of many tripartite interactions may be diYcult to predict, especially along temporal dimensions. This is complicated by our general lack of understanding of the mechanisms underlying these phenomena, particularly in trees. Thus, it is becoming increasingly clear that pathogens (and other symbiotic microbes) must become an explicit component of studies of plant defenses against herbivores. It is rather remarkable that, currently, despite their pervasiveness, microbial

123

Oecologia (2007) 153:365–374

interactions with plants are largely absent from the major hypotheses that comprise plant defense theory (Stamp 2003a). Such deWciency is now beginning to be redressed (e.g., Bonello et al. 2007) and a concerted eVort in this area will contribute signiWcantly to the maturation of existing plant defense theory (Stamp 2003b). Acknowledgements We thank Bryant Chambers, Diane Hartzler, Matt Solensky, Alejandro Chiriboga, Ilka Gomez, Duan Wang, Sheldon Steiner, and Kelly Hendricks for technical and Weld support. Thanks to Kurtz Bros. for nursery substrate. Thanks to two anonymous reviewers for useful suggestions on the manuscript. Salaries and research support provided in part by the USDA National Research Initiative Competitive Grants Program no. 2004-35302-14667 and by state and federal funds to the Ohio Agricultural Research and Development Center, Ohio State University. All experiments comply with the laws of the USA.

References Ajlan AM, Potter DA (1991) Does immunization of cucumber against anthracnose by Colletotrichum lagenarium aVect host suitability for arthropods. Entomol Exp Appl 58:83–91 Beckers GJM, Spoel SH (2006) Fine-tuning plant defence signalling: salicylate versus jasmonate. Plant Biol 8:1–10 Blodgett JT, Bonello P, Stanosz GR (2003) An eVective medium for isolating Sphaeropsis sapinea from asymptomatic pines. For Pathol 33:395–404 Blodgett JT, Eyles A, Bonello P (2007) Organ-dependent induction of systemic resistance and systemic susceptibility in Pinus nigra inoculated with Sphaeropsis sapinea and Diplodia scrobiculata. Tree Physiol 27:511–517 Blodgett JT, Herms DA, Bonello P (2005) EVects of fertilization on red pine defense chemistry and resistance to Sphaeropsis sapinea. For Ecol Manage 208:373–382 Bonello P, Blodgett JT (2003) Pinus nigra-Sphaeropsis sapinea as a model pathosystem to investigate local and systemic eVects of fungal infection of pines. Physiol Mol Plant P 63:249–261 Bonello P, Heller W, Sandermann H Jr (1993) Ozone eVects on rootdisease susceptibility and defence responses in mycorrhizal and non-mycorrhizal seedlings of Scots pine (Pinus sylvestris L.). New Phytol 124:653–663 Bonello P, Pearce RB, Watt F, Grime GW (1991) An induced papilla response in primary roots of Scots pine challenged in vitro with Cylindrocarpon destructans. Physiol Mol Plant P 39:213–228 Bonello P, Storer AJ, McNee WR, Gordon TR, Wood DL, Heller W (2001) Systemic eVects of Heterobasidion annosum (Basidiomycotina) infection on the phenolic metabolism of ponderosa pine and the feeding behavior of Ips paraconfusus (Coleoptera: Scolytidae). In: Proceedings of the Blodgett Forest Research Station 2001 Annual Symposium. Blodgett Forest Research Station, Georgetown, Calif. Bonello P, Storer AJ, Wood DL, Gordon TR (2003) Interactions among the root pathogen, Heterobasidion annosum, ponderosa pines, bark beetles and bark beetle associated fungi. In: Proceedings of the IUFRO working party on bark beetles—S7.03.05 (integrated control of scolytid bark beetles). Blodgett Forest Research Station, Georgetown, Calif. Bonello P, Gordon TR, Herms DA, Wood DL, Erbilgin N (2007) Nature and ecological implications of pathogen-induced systemic resistance in conifers: a novel hypothesis. Physiol Mol Plant Pathol (in press)

Oecologia (2007) 153:365–374 Bostock RM (1999) Signal conXicts and synergies in induced resistance to multiple attackers. Physiol Mol Plant Pathol 55:99–109 Cipollini D, Enright S, Traw MB, Bergelson J (2004) Salicylic acid inhibits jasmonic acid-induced resistance of Arabidopsis thaliana to Spodoptera exigua. Mol Ecol 13:1643–1653 Dammann C, Rojo E, SanchezSerrano JJ (1997) Abscisic acid and jasmonic acid activate wound-inducible genes in potato through separate, organ-speciWc signal transduction pathways. Plant J 11:773–782 Davis JM, Wu H, Cooke JEK, Reed JM, Luce KS, Michler CH (2002) Pathogen challenge, salicylic acid, and jasmonic acid regulate expression of chitinase gene homologs in pine. Mol Plant Microbe Interact 15:380–387 Drooz A (1985) Insects of eastern forests. USDA Forest Service, Washington, D.C. Durrant WE, Dong X (2004) Systemic acquired resistance. Annu Rev Phytopathol 42:185–209 Ekramoddoullah AKM, Yu XS, Sturrock R, Zamani A, Taylor D (2000) Detection and seasonal expression pattern of a pathogenesis- related protein (PR-10) in Douglas-Wr (Pseudotsuga menziesii) tissues. Physiol Plant 110:240–247 Enebak SA, Carey WA (2000) Evidence for induced systemic protection to fusiform rust in loblolly pine by plant growth-promoting rhizobacteria. Plant Dis 84:306–308 Erbilgin N, Krokene P, Christiansen E, Zeneli G, Gershenzon J (2006) Exogenous application of methyl jasmonate elicits defenses in Norway spruce (Picea abies) and reduces host colonization by the bark beetle Ips typographus. Oecologia 148:426–436 Fidantsef AL, Stout MJ, Thaler JS, DuVey SS, Bostock RM (1999) Signal interactions in pathogen and insect attack: expression of lipoxygenase, proteinase inhibitor II, and pathogenesis-related protein P4 in the tomato, Lycopersicon esculentum. Physiol Mol Plant Pathol 54:97–114 Franceschi VR, Krekling T, Christiansen E (2002) Application of methyl jasmonate on Picea abies (Pinaceae) stems induces defenserelated responses in phloem and xylem. Am J Bot 89:578–586 Franceschi VR, Krokene P, Christiansen E, Krekling T (2005) Anatomical and chemical defenses of conifer bark against bark beetles and other pests. New Phytol 167:353–375 Glynn C, Herms DA, Egawa M, Hansen R, Mattson WJ (2003) EVects of nutrient availability on biomass allocation as well as constitutive and rapid induced herbivore resistance in poplar. Oikos 101:385–397 Hammerschmidt R, Dann EK (1999) The role of phytoalexins in plant protection. In: Chadwick DJ, Goode JA (eds) Insect–plant interactions and induced plant defence. Wiley, New York, pp 175–190 Hammerschmidt R, Schultz JC (1996) Multiple defenses and signals in plant defense against pathogens and herbivores. In: Romeo JT, Saunders JA, Barbosa P (eds) Recent advances in phytochemistry: phytochemical diversity and redundancy in ecological interactions, vol 30. Plenum Press, New York, pp 121–154 Hanhimaki S, Senn J (1992) Sources of variation in rapidly inducible responses to leaf damage in the mountain birch–insect herbivore system. Oecologia 91:318–331 Hatcher PE (1995) Three-way interactions between plant pathogenic fungi, herbivorous insects and their host plants. Biol Rev 70:639– 694 Hatcher PE, Paul ND, Ayers PG, Whittaker JB (1997) Nitrogen fertilization aVects interactions between the components of an insect– fungus–plant tripartite interaction. Funct Ecol 11:537–544 Hatcher PE, Paul ND (2000) Beetle grazing reduces natural infection of Rumex obtusifolius by fungal pathogens. New Phytol 146:325– 333 Hatcher PE, Paul ND, Ayres PG, Whittaker JB (1994) Interactions between Rumex spp., herbivores and a rust fungus: Gastrophysa

373 viridula grazing reduces subsequent infection by Uromyces rumicis. Funct Ecol 8:265–272 Hatcher PE, Moore J, Taylor JE, Tinney GW, Paul ND (2004) Phytohormones and plant-herbivore-pathogen interactions: integrating the molecular with the ecological. Ecology 85:59–69 Herms DA (1999) Physiological and abiotic determinants of competitive ability and herbivore resistance. Phyton 39:53–64 Herms DA (2002) EVects of fertilization on insect resistance of woody ornamental plants: reassessing an entrenched paradigm. Environ Entomol 31:923–933 Herms DA, Mattson WJ (1992) The dilemma of plants—to grow or defend. Q Rev Biol 67:283–335 Hogendorp BK, Cloyd RA, Swiader JM (2006) EVect of nitrogen fertility on reproduction and development of citrus mealybug, Planococcus citri Risso (Homoptera : Pseudococcidae), feeding on two colors of coleus, Solenostemon scutellarioides L. Codd. Environ Entomol 35:201–211 Hudgins JW, Christiansen E, Franceschi VR (2003) Methyl jasmonate induces changes mimicking anatomical defenses in diverse members of the Pinaceae. Tree Physiol 23:361–371 Hudgins JW, Christiansen E, Franceschi VR (2004) Induction of anatomically based defense responses in stems of diverse conifers by methyl jasmonate: a phylogenetic perspective. Tree Physiol 24:251–264 Kerpel SM, Soprano E, Moreira GRP (2006) EVect of nitrogen on PassiXora suberosa L. (PassiXoraceae) and consequences for larval performance and oviposition in Heliconius erato phyllis (Fabricius) (Lepidoptera : nymphalidae). Neotrop Entomol 35:192–200 Koricheva J, Larsson S, Haukioja E, Keinanen M (1998) Regulation of woody plant secondary metabolism by resource availability: hypothesis testing by means of meta-analysis. Oikos 83:212–226 Kyto M, Niemela P, Annila E (1996) Vitality and bark beetle resistance of fertilized Norway spruce. For Ecol Manage 84:149–157 Kuc J (2001) Concepts and direction of induced systemic resistance in plants and its application. Eur J Plant Pathol 107:7–12 Lambert MJ (1986) Sulfur and nitrogen nutrition and their interactive eVects on Dothistroma infection in Pinus radiata. Can J For Res 16:1055–1062 Landis TD, Tinus RW, McDonald SE, Barnett JP (1989) Seedling nutrition and irrigation. The container nursery manual, vol 4. USDA Forest Service, Washington, D.C. Larsson S (2002) Resistance of trees to insects—an overview of mechanisms and interactions. In: Wagner MR, Clancy KM, Lieutier F, Paine TD (eds) Mechanisms and deployment of resistance in trees to insects. Kluwer, Dordrecht, pp 1–29 Larsson S, Ekbom B, Bjorkman C (2000) InXuence of plant quality on pine sawXy population dynamics. Oikos 89:440–450 Lee I, Fynn RP, Short TH (2000) Development and evaluation of a computer-controlled fertigation system. Appl Eng Agric 16:279– 284 Luchi N, Ma R, Capretti P, Bonello P (2005) Systemic induction of traumatic resin ducts and resin Xow in Austrian pine by wounding and inoculation with Sphaeropsis sapinea and Diplodia scrobiculata. Planta 221:75–84 Mattson WJ (1980) Herbivory in relation to plant nitrogen-content. Annu Rev Ecol Syst 11:119–161 McNee WR, Bonello P, Wood DL, Storer AJ, Gordon TR (2003) Feeding response of Ips paraconfusus to phloem and phloem metabolites of Heterobasidion annosum-inoculated ponderosa pine, Pinus ponderosa. J Chem Ecol 29:1183–1202 Moran PJ (1998) Plant-mediated interactions between insects and a fungal plant pathogen and the role of plant chemical responses to infection. Oecologia 115:523–530 Orians C (2005) Herbivores, vascular pathways, and systemic induction: facts and artifacts. J Chem Ecol 31:2231–2242

123

374 Paul ND, Hatcher PE, Taylor JE (2000) Coping with multiple enemies: an integration of molecular and ecological perspectives. Trends Plant Sci 5:220–225 Pearce RB (1996) EVects of exposure to high ozone concentrations on stilbenes in Sitka spruce [Picea sitchensis (Bong.) Carr.] bark and on its ligniWcation response to infection with Heterobasidion annosum. Physiol Mol Plant Pathol 48:117–129 Piggott N, Ekramoddoullah AKM, Liu JJ, Yu XS (2004) Gene cloning of a thaumatin-like (PR-5) protein of western white pine (Pinus monticola D. Don) and expression studies of members of the PR5 group. Physiol Mol Plant Pathol 64:1–8 Reglinski T, Stavely FJL, Taylor JT (1998) Induction of phenylalanine ammonia lyase activity and control of Sphaeropsis sapinea infection in Pinus radiata by 5-chlorosalicylic acid. Eur J For Pathol 28:153–158 Regniere J (1984) A method of describing and using variability in development rates for the simulation of insect phenology. Can Entomol 116:1367–1376 Rossiter M, Schultz JC, Baldwin IT (1988) Relationships among defoliation, red oak phenolics, and gypsy moth growth and reproduction. Ecology 69:267–277 Rostas M, Bennett R, Hilker M (2002) Comparative physiological responses in Chinese cabbage induced by herbivory and fungal infection. J Chem Ecol 28:2449–2463 Rostas M, Hilker M (2003) Indirect interactions between a phytopathogenic and an entomopathogenic fungus. Naturwissenschaften 90:63–67 Rostas M, Simon M, Hilker M (2003) Ecological cross-eVects of induced plant responses towards herbivores and phytopathogenic fungi. Basic Appl Ecol 4:43–62 SAS Institute (2003) SAS for personal computer, version 9.1. SAS, Carey, N.C. Saxon ME, et al (2004) InXuence of elevated CO2, nitrogen, and Pinus elliottii genotypes on performance of the redheaded pine sawXy, Neodiprion lecontei. Can J For Res 34:1007–1017 Schmidt A, Zeneli G, Hietala AM, Fossdal CG, Krokene P, Christiansen E, Gershenzon J (2005) Induced chemical defenses in conifers: biochemical and molecular approaches to studying their

123

Oecologia (2007) 153:365–374 function. In: Romeo J (ed) Chemical ecology and phytochemistry of forest ecosystems. Elsevier, Amsterdam, pp 1–28 Siemens DH, MitchellOlds T (1996) Glucosinolates and herbivory by specialists (Coleoptera: Chrysomelidae, Lepidoptera: Plutellidae): consequences of concentration and induced resistance. Environ Entomol 25:1344–1353 Simon M, Hilker M (2003) Herbivores and pathogens on willow: do they aVect each other? Agric For Entomol 5:275–284 Smith JA, et al (2006) Proteomic comparison of needles from blister rust-resistant and susceptible Pinus strobus seedlings reveals upregulation of putative disease resistance proteins. Mol Plant Microbe Interact 19:150–160 Snoeijers SS, Perez-Garcia A, Joosten M, De Wit P (2000) The eVect of nitrogen on disease development and gene expression in bacterial and fungal plant pathogens. Eur Plant Pathol 106:493–506 Stamp NE (2003a) Out of the quagmire of plant defense responses. Q Rev Biol 78:23–55 Stamp NE (2003b) Theory of plant defensive level: example of process and pitfalls in development of ecological theory. Oikos 102:672– 678 Stanosz GR, Trobaugh J, Guthmiller MA, Stanosz JC (2004) Sphaeropsis shoot blight and altered nutrition in red pine plantations treated with paper mill waste sludge. For Pathol 34:245–253 Stout MJ, Fidantsef AL, DuVey SS, Bostock RM (1999) Signal interactions in pathogen and insect attack: systemic plant-mediated interactions between pathogens and herbivores of the tomato, Lycopersicon esculentum. Physiol Mol Plant P 54:115–130 Stout MJ, Workman KV, Bostock RM, DuVey SS (1998) SpeciWcity of induced resistance in the tomato, Lycopersicon esculentum. Oecologia 113:74–81 Wang D, Eyles A, Mandich D, Bonello P (2006) Systemic aspects of host-pathogen interactions in Austrian pine (Pinus nigra): a proteomics approach. Physiol Mol Plant Pathol 68:149–157 Zeneli G, Krokene P, Christiansen E, Krekling T, Gershenzon J (2006) Methyl jasmonate treatment of mature Norway spruce (Picea abies) trees increases the accumulation of terpenoid resin components and protects against infection by Ceratocystis polonica, a bark beetle-associated fungus. Tree Physiol 26:977–988