J Chem Ecol (2013) 39:516–524 DOI 10.1007/s10886-013-0266-3

Performance of an Herbivorous Leaf Beetle (Phratora vulgatissima)onSalix F2 Hybrids: the Importance of Phenolics

Mikaela Torp & Anna Lehrman & Johan A. Stenberg & Riitta Julkunen-Tiitto & Christer Björkman

Received: 18 September 2012 /Revised: 30 January 2013 /Accepted: 31 January 2013 /Published online: 1 March 2013 # Springer Science+Business Media New York 2013

Abstract The genotype of the plant determines, through the to Phratora vulgatissima. Synergistic effects of these com- expression of the phenotype, how well it is suited as food for pounds, as well as potential threshold concentrations, are herbivores. Since hybridization often results in profound ge- plausible. In addition, we found considerable variation in both nomic alterations with subsequent changes in phenotypic distributions and concentrations of different phenolics in the traits, it has the potential to significantly affect plant- F2 hybrids. The phenolic profiles of parental genotypes and herbivore interactions. In this study, we used a population of F2 hybrids differed significantly (e.g., novel compounds F2 hybrids that originated from a cross between a Salix appeared in the hybrids) suggesting genomic alterations with viminalis and a Salix dasyclados genotype, which differed in subsequent changes in biosynthetic pathways in the hybrids. both phenolic content and resistance to the herbivorous leaf beetle Phratora vulgatissima. We screened for plants that Keywords Hybridization . Phenolics . Luteolin . showed a great variability in leaf beetle performance (i.e., Salicylates . Phratora vulgatissima performance . oviposition and survival). By correlating leaf phenolics to Resistance . Salix the response of the herbivores, we evaluated the importance of different phenolic compounds for Salix resistance to the targeted insect species. The performance of P. vulgatissima Introduction varied among the F2 hybrids, and two patterns of resistance emerged: leaf beetle oviposition was intermediate on the F2 Hybridization between species is common in nature (Arnold, hybrids compared to the parental genotypes, whereas leaf 1997) and can result in substantial changes in various plant beetle survival demonstrated similarities to one of the parents. characteristics. Parental species and hybrids often differ in The findings indicate that these life history traits are controlled resistance to herbivores. This variation is probably mediated by different resistance mechanisms that are inherited differ- by modifications of plant chemistry, morphology, or phenology ently in the hybrids. Salicylates and a methylated luteolin in hybrid plants compared to their parents (Fritz et al., 1999; derivative seem to play major roles in hybrid resistance Cheng et al., 2011). Patterns of hybrid resistance are diverse: hybrids can express intermediate (Fritz et al., 1998;Czesaket al., 2004; Hochwender and Fritz, 2004), equal (Orians et al., M. Torp (*) : J. A. Stenberg : C. Björkman Department of Ecology, Swedish University of Agricultural 1997; Fritz et al., 1998;Hallgrenetal.,2003), increased (Fritz, Sciences, P.O. Box 7044, 750 07 Uppsala, Sweden 1999; Hjältén and Hallgren, 2002; Hochwender and Fritz, e-mail: [email protected] 2004), or decreased (Whitham, 1989; Orians et al., 1997; Hjältén, 1998; Czesak et al., 2004) resistance compared to A. Lehrman Department of Crop Production Ecology, Swedish University their parents. This is not surprising since different traits that of Agricultural Sciences, P.O. Box 7043, 750 07 Uppsala, Sweden affect herbivore responses, e.g., plant secondary chemistry (Tahvanainen et al., 1985) and nutrient content (Mattson, R. Julkunen-Tiitto 1980), exhibit varying and often complex inheritance patterns Natural Product Research Laboratory, Department of Biology, University of Eastern Finland, P.O. Box 111, (Rieseberg and Ellstrand, 1993; Cheng et al., 2011). The resis- 801 01 Joensuu, Finland tance of parental species and hybrids depends partially on J Chem Ecol (2013) 39:516–524 517 environmental conditions (Hochwender et al., 2000). In 2003). To understand the effect of hybridization from an addition, different herbivore species respond differently ecological point of view, it is important to include the more to variation in plant traits (Matsuki and MacLean, 1994; complex hybrid categories (Arnold, 1997). One way to Orians and Fritz, 1996). elucidate how hybridization influences plant-herbivore in- The importance of plant chemistry for the outcome of teractions is to study how plant characteristics that affect plant-herbivore interactions and plant resistance to herbi- herbivore responses (e.g., plant chemistry) are inherited vore attacks has been thoroughly addressed (e.g., Fraenkel, from parental species to hybrids (Orians et al., 2000). 1959; Haukioja, 1980; Harborne, 2001;Treutter,2006; Furthermore, by correlating variation in plant traits to the Barbehenn and Constabel, 2011). However, much is still response of herbivores, it might be possible to gain knowl- unknown about genetic or regulatory aspects of the biosyn- edge about the complex interactions among genes, traits, thesis of many of these defensive chemical compounds. and patterns on a broader ecological scale. Since small genetic changes can cause large shifts in the In this study, we used hybridization between Salix chemical properties of both plant individuals and species viminalis, which is susceptible to the chrysomelid leaf beetle (Nyman and Julkunen-Tiitto, 2005), hybridization should Phratora vulgatissima,andS. dasyclados, which exhibits a further increase the variability in plant chemical profiles. high degree of resistance to this insect pest. Crossing of two According to classic hybridization theory, a trait, whether it F1 hybrids created a F2 generation that gave us an opportunity is expressed by one or several genes, is expected to show to screen for plants that showed variability in the performance greater variance in the F2 hybrids compared to either of P. vulgatissima. Moreover, by analyzing the phenolic pro- the parental species or the F1 hybrids (Hochwender et al., files of the F2 hybrids, we aimed to find chemical clues that 2000). This is a result of recombination between chromo- explained the varied performance of the targeted insect spe- somes during the production of F2 hybrids (Hochwender cies. This study was done as a part of a Salix molecular and Fritz, 2004). Orians (2000)showedthatF2hybridplants breeding program with the long-term goal of developing demonstrated greater variance in secondary chemistry methods for selecting plants with increased resistance to insect than F1 and parental plants. The variability in plant pests and pathogens. chemical content expressed in the F2 generation can be used to evaluate the effect of plant chemistry on herbivore re- sponses and to identify the mechanisms behind such plant- Methods and Materials herbivore interactions. Willows (Salix spp.) are excellent for investigating the The Study System Willows (Salix spp.) are deciduous trees effects of hybridization on insect herbivore responses. Most or shrubs that are commonly cross-fertile and hybridize of the species of this genus are cross-fertile, and many easily. Cuttings root readily making cultivation and cloning hybrids occur naturally. There is a diverse insect community simple. The foliage of Salix species often is rich in carbon- that feeds on willows; in particular, chrysomelid beetles based phenolics, and generally there is a significant genetic (Coleoptera) may cause severe damage to the plants. effect on the phenolic metabolite content (Julkunen-Tiitto, Willows generally can be divided into two groups based 1986;Baumetal.,2009). Phenolic compounds play a major on their content of phenolics: some species or genotypes role in Salix resistance to herbivory (e.g., Tahvanainen et al., contain high levels of phenolic glycosides (e.g., salicylates) 1985). The two willow species used in this study vary in their and low levels of condensed tannins, whereas other species phenolic profiles; the foliage of S. viminalis is high in con- or genotypes demonstrate high levels of condensed tannins densed tannins but does not contain any salicylates, whereas S. but no phenolic glycosides (Julkunen-Tiitto, 1986, 1989; dasyclados leaves contain high levels of salicylates and small Baum et al., 2009). Insect herbivores vary in their ability amounts of condensed tannins (Julkunen-Tiitto, 1986). The to cope with these compounds; specialists often prefer to variation in phenolic content has been proposed to be respon- feed on willows with high levels of phenolic glycosides, sible for the different resistance of the two species for the while generalists usually avoid these plants (Rowell-Rahier, herbivorous leaf beetle P. vulgatissima; S. viminalis is suscep- 1984; Tahvanainen et al., 1985; Kelly and Curry, 1991; tible while S. dasyclados demonstrates a highdegreeofresis- Orians et al., 1997; Boeckler et al., 2011). tance (Glynn et al., 2004; Lehrman et al., 2012). Salix hybrids previously have been used as a model The blue willow leaf beetle (P. vulgatissima) is a common system in studies that try to correlate the response of her- herbivore on many species of Salix, including short rotation bivorous insects and plant chemical composition. Some of coppice crops such as S. viminalis. The beetles are univoltine these studies have focused on parental species and F1 hy- in Sweden and they overwinter as adults. In early spring, they brids only (Orians et al., 1997; Hallgren, 2003), whereas start feeding on their host plants and in late spring or early others have included F2 hybrids (Hallgren et al., 2003; summer, the beetles mate and start laying eggs. The eggs are Glynn et al., 2004) and back-crosses (Hallgren et al., laid in clusters on the leaves, and egg hatching takes place 518 J Chem Ecol (2013) 39:516–524 after approximately 2 wk (Kendall et al., 1996). After passing in Lehrman et al. (2012). In this F2 hybrid bioassay, 24 through three instar stages, the larvae pupate in the soil. The females and 16 males were caged together, in order to feed beetles, and especially their larvae, can cause severe damage and mate, for 20 d. Cages contained three potted plants of to Salix plants (Björkman et al., 2000). one of the twelve F2 hybrids (i.e. 12 cages, each containing 3 plants of respective hybrid genotype). After 20 d, the Plants A female S. viminalis (cv. ‘Jorunn’) was crossed with female beetles were transferred to separate jars and trans- amaleS. dasyclados (cv. ‘SW901290’), and two of the F1 ferred to a climate room as described above. Each jar progenies were further crossed to produce an F2 generation. contained one mature leaf from the mid-section of the ap- Twelve of these F2 hybrids were selected for testing of resis- propriate F2 hybrid, and as in the parental bioassay, beetles tance against P. vulgatissima. To enable as much genetic were provided with fresh leaves every secondday when variation as possible among the selected F2 hybrids, the eggs were counted and removed from the jars. The F2 selection was done based on differences in morphology, such hybrid bioassay was terminated after 14 d. About 20 females as leaf shape and branching, between the F2 hybrids. Winter- per F2 hybrid were included in the bioassay but the final dormant cuttings from annual shoots of the two parental number varied due to high mortality on some of the hybrids genotypes and the selected F2 hybrids were harvested and in the mating cages. stored at 4 °C. About six weeks prior to when the plants were to be used in the experiments, 10 cm cuttings were planted in Chemical Analyses Six F2 hybrids were selected for screen- pots (diam=11 cm) containing Hasselfors™ planting soil. ing of phenolics that might be related to beetle egg production Pots were transferred to a greenhouse where the cuttings were and survival (Fig. 1). These F2 hybrids were chosen due to grown in natural light with supplementary illumination at their variable effects on beetle performance. Five leaves from 300 μmolm2s-1 and temperatures ranging from 20 °C to 6-wk-old plants were sampled from the mid-section of 5 26 °C. Plants were watered regularly and fertilized (N:P:K= individual plants of each genotype. The leaves were dried in 51:10:43). New cuttings were continuously planted in order to agreenhouse(20–26 °C, natural light) for 3 d and then frozen provide fresh leaves from plants of the same age throughout at −20 °C in sealed plastic bags. Prior to extraction, composite the whole experimental period. leaf samples were milled using a ball mill. Phenolics were extracted from 6 to 8 mg leaf material with the Precellys 24- Bioassay In March, adult P. vulgatissima were collected in homogenizer (Betrin Technologies, France) for 30 s at the field from their overwintering sites in the vicinity of 4,780 × g with 600 μlof100%MeOHin2mlvials. Uppsala, central Sweden. To avoid disruption of their hiber- Samples were kept in an ice bath for 15 min with subsequent nation period, the beetles were stored at -4 °C until the start of centrifugation for 5 min at 11,300 × g at 4 °C. Supernatants the bioassay. At the start, P. vulgatissima beetles were trans- were transferred to 6 ml tubes, and the extraction procedure ferred from storage to room temperature, and their sexes were was repeated for three additional 5 min extractions. Methanol determined. from the combined extracts (2.4 ml) was evaporated to dryness In a first bioassay, where beetle performance on the parental under nitrogen. Dried samples were re-dissolved in 300 μl genotypes ‘Jorunn’ (S. viminalis)and‘SW901290’ (S. 100 % MeOH and 300 μldistilledH2O. Samples were ana- dasyclados) was tested (see Lehrman et al., 2012), individual lyzed with a high-performance liquid chromatography, HPLC beetles were kept isolated in 30 ml plastic jars (26×67 mm) with (Agilent, Series 1100, Germany), equipped with a diode array 3 perforated lids. Each jar contained 1 cm moistened oasis floral detector, DAD (G1315B). A reverse-phase (RP) C18 column foam and one mature leaf from the mid-section of the genotype (Agilent Technologies, USA) was used for the separation of to be tested. The jars were placed in a climate room at 20 °C the compounds using conditions described in Julkunen-Tiitto with a diurnal rhythm of 18:6, L:D. Beetles were provided with and Sorsa (2001). The compounds were detected at 270 nm fresh leaves every secondday and at the same time, the jars were and 320 nm based on their UV-spectra and retention times. cleaned, the oasis moistened, and beetle survival monitored. Concentrations were calculated using the following standards: After 10 d, females and males were paired in the jars for 24 h salicin for salicin; salicortin for salicortin and HCH-salicortin; to mate. After mating, males were removed and only females hyperin for hyperin and quercetin derivatives; luteolin 7- remained in the bioassay until they died. Every second day, eggs glucoside for luteolin derivatives; (+)-catechin for (+)-cate- were counted and removed from the jars. The bioassay was chin; kaempferol 3-glucoside for kaempferol-diglycoside and finished when the last beetle had laid no eggs for 4 d; the best dicoumaroyl-astragalin; chlorogenic acid for chlorogenic acid performing beetles continued their oviposition for about 3 mo. derivatives (i.e., compounds expressing chlorogenic acid spec- Results from the bioassay showed that the P. vulgatissima trum); and cinnamic acid for cinnamic acid derivatives (i.e., oviposition rate levelled out about 6 d after mating. compounds expressing p-OH-cinnamic acid spectrum). In the current study, we used the F2 hybrids in a bioassay Condensed tannins (methanol soluble and insoluble) were similar to the one described briefly above and in more detail quantified by oxidative depolymerization to anthocyanidins J Chem Ecol (2013) 39:516–524 519

Fig. 1 Phratora vulgatissima a oviposition (eggs/day; means± S.E.) and b survival (number of days: means±S.E.) during a 14 d bioassay in which the beetles were feeding on leaves from either one of the parental genotypes or one of twelve F2 hybrids originating from a crossing between Salix viminalis (‘Jorunn’) and S. dasyclados (‘SW901290’). The percentages given in b describe the number of beetles that survived during the 14 d bioassay (e.g., 100 % implies that all beetles survived the 14 d trial while 25 % denotes that only 1/4 of the beetles survived for 14 d). The highlighted hybrids were selected for phenolic analysis

in acid buthanol as described by Porter et al. (1986)using Lehrman et al. (2012)onP. vulgatissima oviposition and purified Betula nana leaf tannin as the standard. survival on ‘Jorunn’ and ‘SW901290’ during the first 14 days following mating (i.e., day 12 to 26 in their study). Since Statistics Data were analyzed using SAS 9.2. The non- beetle egg laying on the two parental genotypes was relatively parametric Kruskal-Wallis test was performed on beetle stable during the selected time period, we argue that beetle oviposition and survival on the parental generation since oviposition during those two weeks should be compatible and the data were not normally distributed. One-way ANOVAs comparable to the results of the current study. were performed on beetle oviposition and survival on the F2 generation, since these data met the criteria for adopting Leaf Beetle Performance on Parental Species and F2 parametric tests showing normal distribution. Regression Hybrids There was a significant difference between the two analyses were performed on the leaf concentration of differ- parental Salix genotypes in P. vulgatissima oviposition (H= ent phenolic compounds vs. the mean number of eggs 12.22, P<0.001; Fig. 1a) and survival (H=5.03, P=0.025; produced per day and beetle survival in order to find phe- Fig. 1b) with females feeding on the S. viminalis genotype nolic metabolites that were correlated to these key life ‘Jorunn’ performing better compared to females provided with history traits of the beetles. leaves from the S. dasyclados genotype ‘SW901290’.There was also a significant difference in oviposition (number of eggs/day) among the beetles fed on the F2 hybrid genotypes

Results (F11,230=44.16, P<0.001; Fig. 1a). The highest egg production was found for beetles on hybrid 312 while those on hybrid 310 In a previous study in which the host effect on P. vulgatissima exhibited the lowest egg production. Beetle oviposition on the life history traits was examined, Lehrman et al. (2012)showed F2 hybrids seemed to be intermediate between the two parental that a two week oviposition period is a good indicator of the genotypes, and none of the hybrids yielded egg production as suitability of the host plant. In the same study, P. vulgatissima high as that for beetles on the S. viminalis genotype ‘Jorunn’. oviposition and survival on the two parental genotypes On the other hand, beetle egg production was low on several of ‘Jorunn’ (S. viminalis)and‘SW901290’ (S. dasyclados)were the F2 hybrids, resembling the oviposition on the S. dasyclados evaluated. In our current study, we used the data from genotype ‘SW901290’. Beetle survival (number of days) on the 520 J Chem Ecol (2013) 39:516–524

F2 hybrids varied significantly between the F2 genotypes (Fig. 2), and the total concentration of these other phenolics

(F11,230=3.90, P<0.001; Fig. 1b). Survival rate generally was in the leaves varied among the hybrids (F5,29=6.00, P= high with the exception of the beetles feeding on hybrid 310, 0.001). The concentration and distribution of the individual which exhibited high mortality with only 25 % surviving the substances varied among the hybrids, e.g., chlorogenic acid, two week bioassay period. Beetle survival on hybrid 310 was hyperin, and luteolin derivatives were present in all hybrids even lower than on the parental S. dasyclados genotype whereas (+)-catechin, quercetin 3-glucoside, and some other ‘SW901290’. flavonoids showed a more scattered distribution (Table 1).

Phenolic Profiles of F2 Hybrids The phenolic profiles of the Relationships Between Leaf Beetle Performance and F2 F2 hybrids varied considerably (Table 1). Salicylates were Hybrid Plant Chemistry When beetle oviposition was com- present in all the studied F2 hybrids, but the total concen- pared to the individual phenolic compounds found in the F2 tration of salicylates varied significantly between the hy- hybrids, some noteworthy relationships appeared (Fig. 3). brids (F5,29=94.77, P<0.001) with hybrid 312 exhibiting Phratora vulgatissima oviposition was negatively related to the lowest concentrations and hybrid 310 the highest the leaf concentrations of both salicin and salicortin (R2=0.72, (Fig. 2). All hybrids also contained condensed tannins, and P=0.032 and R2=0.56, P=0.085, respectively) as well as to the the condensed tannin concentrations differed between the total concentration of leaf salicylates (R2=0.58, P=0.080; hybrids (F5,29=6.75, P<0.001). The hybrids with high Fig. 3a). Hybrid 416, however, can be seen as an outlier; levels of salicylates generally had low levels of condensed despite the relatively low salicylate concentration, oviposition tannins and vice versa (Fig. 2). In addition, a wide variety of on this hybrid was extremely low. Negative relationships were other phenolic compounds were found in the F2 hybrids found between beetle egg production and the leaf concentration

Table 1 Concentrations (means±S.E. mg/g d.w.) of different phenolic substances in the leaves of six F2 hybrids originating from a cross between Salix viminalis (‘Jorunn’) and Salix dasyclados (‘SW901290’)

Chemical compound F2 hybrids

Phenolics 312 169 386 396 416 310

Salicin 0.91±0.12 0.79±0.04 2.29±0.21 2.75±0.13 1.71±0.08 2.74±0.19 Salicortin 4.45±0.50 5.88±0.26 20.35±1.24 28.71±1.12 9.11±0.68 29.12±2.51 HCH-Salicortin - - 4.48±0.54 2.03±0.26 2.14±0.11 3.93±0.47 Chlorogenic acid 5.26±0.94 4.50±0.27 2.30±0.65 5.90±1.44 2.04±0.59 2.55±0.29 Neochlorogenic acid 0.22±0.02 0.96±0.10 - 0.83±0.07 1.04±0.08 - Hyperin 0.07±0.01 0.14±0.01 0.06±0.01 0.13±0.01 0.10±0.01 0.19±0.01 p-OH-cinnamic acid derivative 1 0.25±0.05 1.50±0.16 0.27±0.03 0.48±0.02 0.63±0.07 0.26±0.03 p-OH-cinnamic acid derivative 2 0.46±0.09 0.34±0.02 0.80±0.25 0.56±0.05 0.41±0.04 0.73±0.03 p-OH-cinnamic acid derivative 3 1.04±0.13 1.70±0.13 - 1.01±0.08 1.84±0.27 - p-OH-cinnamic acid derivative 4 0.87±0.11 1.36±0.09 - 0.78±0.07 1.52±0.24 - (+)-catechin - 0.22±0.14 1.42±0.37 1.28±0.18 2.36±0.43 1.33±0.33 Quercetin diglycoside 1 - 0.80±0.07 - 0.08±0.01 0.09±0.01 0.18±0.01 Quercetin diglycoside 2 0.09±0.01 0.19±0.01 - 0.15±0.01 0.40±0.02 0.19±0.01 Quercetin 3-glucoside 0.83±0.02 - 0.50±0.05 - 0.16±0.02 - Quercetin glycoside+Kaempferol 3-glucoside 0.50±0.02 0.35±0.03 0.39±0.02 0.35±0.03 0.39±0.04 0.37±0.06 Kaempferol diglycoside - 0.09±0.01 - - - - Quercitrin - 0.30±0.04 - - - - Methyl-apigenin-glycoside 0.19±0.01 0.15±0.01 - - - - Methyl-lureolin-glycoside 1 2.52±0.27 3.13±0.16 1.63±0.10 3.48±0.07 1.72±0.16 3.20±0.24 Methyl-lureolin-glycoside 2 0.93±0.05 1.20±0.07 1.30±0.06 1.60±0.07 2.01±0.13 1.87±0.12 Methyl-lureolin-glycoside 3 0.33±0.01 0.76±0.04 0.41±0.04 0.95±0.05 0.94±0.06 0.59±0.04 Dicoumaroyl-astragalin 2.18±0.20 1.88±0.24 2.10±0.30 1.99±0.13 1.73±0.13 0.56±0.04 Condensed tannins Methanol soluble 43.57±6.43 45.13±3.99 42.29±10.35 18.84±4.87 33.91±7.80 15.26±2.33 Methanol insoluble 40.50±2.33 43.85±0.50 28.90±1.08 25.26±1.87 28.65±1.80 25.50±2.57 J Chem Ecol (2013) 39:516–524 521

P=0.019; Fig. 3d). No significant relationships were found between beetle survival and leaf phenolic content.

Discussion

Host plant suitability is influenced by the genotype of the plant and accordingly, plant genetic variation can be projected upward onto higher trophic levels (Hochwender and Fritz, 2004). Two different patterns of Salix resistance to P. vulgatissima emerged from the results of this study. We found that leaf beetle oviposition was intermediate on the F2 hybrids compared to the parental genotypes, suggesting an additive inheritance of resistance traits (Fritz et al., 1998). On the other hand, beetle survival on the F2 hybrids demonstrated more similarities to one of the parents, the S. viminalis geno- type ‘Jorunn’, proposing dominant inheritance (Fritz et al., 1998). These findings indicate that there are differ- ent resistance mechanisms controlling oviposition and surviv- al, and that these resistance mechanisms are inherited differentlyinhybrids. When considering plant chemistry, Orians (2000)stated that while the most common pattern is that hybrids are inter- mediate compared to their parental species (e.g., Orians et al., 2000; Hallgren et al., 2003; Yarnes et al., 2008; Cheng et al., 2011), other patterns of expression have been observed; the concentration of chemical compounds in hybrids can be sim- ilar to one of the parents (Fahselt and Ownbey, 1968;Courtet al., 1992; Yarnes et al., 2008), higher than any of the parents (Spring and Schilling, 1989; Court et al., 1992; Hallgren et al., 2003; Nahrung et al., 2009;Kirketal.,2010; Cheng et al., 2011), or lower than any of the parents (Court et al., 1992; Yarnes et al., 2008; Cheng et al., 2011). Hybrids also can lack chemical substances that are present in the parental species or exhibit compounds that are not produced by the parents (Rieseberg and Ellstrand, 1993; Orians, 2000; Kirk et al., 2010; Cheng et al., 2011). When comparing the total concen- tration of all phenolic compounds found in the leaves of our parental genotypes and F2 hybrids, the hybrids showed phe- nolic profiles that were intermediate compared to the parents (for data on the chemical content of the S. viminalis genotype Fig. 2 The sum concentration (means±S.E.) of a salicylates (salicin, ‘ ’ ‘ ’ salicortin, and HCH-salicortin), b condensed tannins (methanol soluble Jorunn and the S. dasyclados genotype SW901290 ,see and methanol insoluble), and c other phenolic compounds (phenolic Lehrman et al., 2012). An additive inheritance for all phenolic acids, phenolic glycosides, and flavonoids) found in the leaves of the substances could thus be assumed. However, when evaluating F2 hybrids. The hybrids are ranked according to Phratora vulgatissima individual phenolic compounds or phenolic groups, other oviposition with beetles feeding on hybrid 312 exhibiting the highest egg production, while beetles feeding on hybrid 310 completely patterns appeared. Salix viminalis does not contain any salic- failed in their oviposition. Means capped with different letters are signif- ylates, while this substance group is present in S. dasyclados icantly different (Julkunen-Tiitto, 1986). According to the study of Lehrman et al. (2012), ‘SW901290’ contain about 4.5 mg salicylates per g of a luteolin derivative (R2=0.86, P=0.008; Fig. 3b) and (+)- d.w. leaf material. The salicylate concentration in the F2 catechin (R2=0.81, P=0.015; Fig. 3c). Moreover, beetle ovi- hybrids in this study varied between approximately 5 – position and the concentration of condensed tannins in the 35 mg/g d.w., which is higher than the salicylate concentration leaves demonstrated a positive relationship (R2 =0.79, of both parental genotypes. In the F2 hybrids, condensed 522 J Chem Ecol (2013) 39:516–524

Fig. 3 Relationships between Phratora vulgatissima oviposition (eggs/day) and leaf concentrations (mg/g d.w.) of a salicylates, b methyl-luteolin glycoside 2, c (+)-catechin and d condensed tannins in willow F2 hybrids. Numbers in the figure correspond to the different hybrids

tannins demonstrated concentrations that were intermediate or ‘Jorunn’. These findings suggest that, in the process of hy- lower than the parental genotypes, while the group of other bridization, 310 and 416 gained modifications that were not phenolic compounds showed equal or higher concentrations present in the parental genotypes, making them less suscepti- in the F2 hybrids compared to the parents. In addition, we ble to the leaf beetle P. vulgatissima. Examining the phenolic found some individual phenolics that were present in one of profiles of these two hybrids shows that 416 contains fairly the parents (‘Jorunn’) but were lacking in the hybrids (e.g., low amounts of salicylates, while the salicylate concentration some quercetin, myricetin, kaempferol, and rhamnetin deriv- in 310 is high. Salicylates previously have been reported as atives) and some new compounds that were present in the important compounds in willow resistance (Rowell-Rahier, hybrids but not in either of the parents (e.g., some quercetin, 1984;KellyandCurry,1991; Tahvanainen et al., 1985; apigenin, and luteolin derivatives). Loss of parental chemical Lehrman et al., 2012). The high levels of salicylates in 310 compounds is common in hybrids and the appearance of novel could, therefore, be a plausible explanation for the lack of hybrid chemicals increases in later generation hybrids oviposition and high mortality of beetles feeding on this (Rieseberg and Ellstrand, 1993). This is due primarily to i) hybrid. However, there was low egg production in 416, de- inhibition of biochemical pathways, which results in an accu- spite the rather low concentrations of salicylates found in this mulation of precursor chemicals and a simultaneous loss of hybrid, leading us to propose that there are resistance compo- end-products, ii) re-direction of biochemical pathways, which nents other than salicylates. A comparison of the concentra- results in new combinations of compound skeleton and side tion of a methyl-luteolin glycoside in 310 and 416 reveals chains where one part of the substance generally originates high amounts of this compound in both hybrids. In a previous from one parent and the other part is derived from the other study that included the parental genotypes ‘Jorunn’ and parent, iii) disruption of regulatory genes that alter gene ex- ‘SW901290’, we showed potential negative effects of both pression and lead to changes in where the chemicals are salicylates and a luteolin derivative on P. vulgatissima adult produced, and iv) further segregation of alleles at multiple and larval performance (Lehrman et al., 2012). The effect of loci (Orians, 2000). the luteolin derivative was, however, non-linear, and there was An intriguing result of this study is the characteristics of the a tendency towards a possible critical threshold concentration F2 hybrids 416 and 310. Beetle oviposition was extremely above which beetle performance was reduced. Noteworthy is low on these two hybrids; 416 demonstrated an egg produc- the fact that the luteolin derivative discussed in Lehrman et al. tion of 0.09 eggs/day, while there was no egg production at all (2012) was not present in any of the F2 hybrids in this study; on 310. As a comparison, oviposition on the resistant parent instead novel methylated luteolin derivatives were found. ‘SW901290’ was 0.11 eggs/day. In addition, beetles feeding Methylated flavonoids have higher insect and mammalian on 310 demonstrated high mortality with only 25 % surviving toxicity than the corresponding non-methylated ones the two week bioassay period (compared to a survival rate of (Seigler, 1998). We hypothesize that the performance of the 43 % on the resistant parent ‘SW901290’). Beetle survival on beetles feeding on the 310 and 416 hybrids probably is a 416 was, however, almost as high as on the susceptible parent consequence of synergistic effects between the concentrations J Chem Ecol (2013) 39:516–524 523 of salicylates and the methylated luteolin derivatives. The Arabidopsis, Wang et al., 2006;andSenecio, Hegarty et al., presence of potential threshold values, above which either 2008) have revealed significant changes to hybrid gene ex- compound group alone or a combination of both compound pression patterns compared to the expression patterns found in groups together results in toxic or even lethal effects, is the parental species. These findings, together with the results plausible. presented in our study, motivates further research that pene- Every hybridization event within a group of potentially trate deeper into the biochemical, genetic, and molecular interbreeding plant genotypes will result in a unique variation mechanisms behind the observed patterns. We would not find in the progeny. If the extreme variation in plant quality that it surprising if such research leads to the detection of hitherto was correlated to the performance of herbivorous beetles in unknown mechanisms that may explain anomalies in plant this study is representative for any cross between these two resistance studies. taxa, it could be argued that the results presented here are conservative estimates of the existing variation. The fact that Acknowledgements We thank Elena Ahlbin, Marie Melander, Sussi we found such large variation in beetle performance and plant Andersson Björkman, and Karin Eklund for assistance with insect chemistry highlights the potentially strong impact of plant bioassays and plant nursing, and Ann-Christin Rönnberg-Wästljung for advice and for providing plant material. This study was done as a part traits on beetle performance in willows. Although we were of the Salix Molecular Breeding Activities (SAMBA) project financed by able to capture a high degree of chemical variation within the the Swedish Energy Agency, the Faculty of Natural Resources and F2 population by analyzing a very limited number of plant Agricultural Sciences at the Swedish University of Agricultural Sciences individuals, the design did not allow us to determine which and Lantmännen Agroenergi AB. chemical traits actually were responsible for the observed bioassay patterns. We can only conclude that beetle oviposi- tion rate was negatively associated with salicylates, positively References associated with condensed tannins, and that high concentra- tions of a methylated luteolin derivative may contribute to the ARNOLD, M. L. 1997. Natural Hybridization and Evolution. Oxford poor beetle performance observed on some of the F2 hybrids. University Press, New York. Further studies including a larger set of plant individuals is BARBEHENN, R. V. and CONSTABEL, C. P. 2011. Tannins in plant- needed to identify the mechanistic connections in this study herbivore interactions. Phytochemistry 72:1551–1565. BAUM, C., TOLJANDER, Y. K., ECKHARDT, K.-U., and WEIH, M. 2009. system. The significance of host-fungus combinations in ectomycorrhizal A benefit of using F2 hybrids in experimental studies is that symbioses for the chemical quality of willow foliage. Plant Soil the plants have been subjected to recombination, which results 323:213–224. in a break-up of previous gene combinations and dissociation BJÖRKMAN, C., HÖGLUND, S., EKLUND, K., and LARSSON, S. 2000. Effects of leaf beetle damage on stem wood production in cop- of one trait from other traits, thus providing a greater ability to picing willow. Agr. Forest Entomol. 2:131–139. determine the independent contribution of different traits to BOECKLER,G.A.,GERSHENZON,J.,andUNSICKER,S.B.2011. overall plant resistance (Hochwender et al., 2000). Moreover, Phenolic glycosides of the Salicaceae and their role as anti- a broad range of chemical phenotypes enables screening for herbivore defences. Phytochemistry 72:1497–1509. CHENG, D., VRIELING, K., and KLINKHAMER, P. G. L. 2011. The effect favorable individuals, which can be used to further expand our of hybridization on secondary metabolites and herbivore resis- understanding of the intricate mechanisms behind plant- tance: implications for the evolution of chemical diversity in herbivore interactions. In this study, we found significant plants. Phytochem. Rev. 10:107–117. differences in the phenolic profiles of parental genotypes COURT, W. A., BRANDLE, J. E., POCS, R., and HENDEL, J. G. 1992. The chemical composition of somatic hybrids between Nicotiana and F2 hybrids; compounds present in the parents were absent tabacum and N. debneyi. Can. J. Plant Sci. 72:209–225. in the hybrids, novel compounds appeared in the hybrids, and CZESAK, M. E., KNEE, M. J., GALE, R. G., BODACH, S. D., and FRITZ, the ratios between different compounds changed. The diver- R. S. 2004. Genetic architecture of resistance to aphids and mites gence in chemical phenotypes of the F2 hybrids indicates in a willow hybrid system. Heredity 93:619–626. substantial genetic alterations in these plants and changes of FAHSELT, D. and OWNBEY, M. 1968. Chromatographic comparison of Dicentra species and hybrids. Am. J. Bot. 55:334–345. biosynthetic pathways. In addition to genomic alterations, FRAENKEL, G. S. 1959. The raison d’être of secondary plant sub- hybridization can result in profound changes to gene expres- stances. Science 129:1466–1470. sion (Hegarty et al., 2008). New patterns of gene expression FRITZ, R. S. 1999. Resistance of hybrid plants to herbivores: genes, – may arise by several mechanisms involving changes to up- environment, or both? Ecology 80:382 391. FRITZ, R. S., ROCHE, B. M., and BRUNSFELD, S. J. 1998. Genetic stream regulators and mutations of non-coding regulatory variation in resistance of hybrid willows to herbivores. Oikos DNA sequences (e.g., enhancers) of a gene (Rebeiz et al., 83:117–128. 2011). Genetic recombination during hybridization may in- FRITZ, R. S., MOULIA, C., and NEWCOMBE, G. 1999. Resistance of crease such alterations and result in activation of previously hybrid plants and animals to herbivores, pathogens and parasites. Annu. Rev. Ecol. Syst. 30:565–91. latent regulatory sequences. Recent transcriptomic studies in GLYNN,C.,RÖNNBERG-WÄSTLJUNG,A.-C.,JULKUNEN-TIITTO,R., several plant hybrid systems (e.g., maize, Stupar et al., 2007; and WEIH, M. 2004. Willow genotype, but not drought treatment, 524 J Chem Ecol (2013) 39:516–524

affects foliar phenolic concentrations and leaf-beetle resistance. MATSUKI, M. and MACLEAN JR., S. F. 1994. Effects of different leaf Entomol. Exp. Appl. 113:1–14. traits on growth rates of insect herbivores on willows. Oecologia HALLGREN, P. 2003. Effects of willow hybridization and simulated 100:141–152. browsing on the development and survival of the leaf beetle MATTSON, W. J. 1980. Herbivory in relation to plant nitrogen content. Phratora vitellinae. BMC Ecol. 3:5. Annu. Rev. Ecol. Syst. 11:119–161. HALLGREN, P., IKONEN,A.,HJÄLTÉN, J., and ROININEN, H. 2003. NAHRUNG, H. F., WAUGH, R., and HAYES, R. A. 2009. Corymbia Inheritance patterns of phenolics in F1, F2, and back-cross hy- species and hybrids: chemical and physical foliar attributes and brids of willows: implications for herbivore responses to hybrid implications for herbivory. J. Chem. Ecol. 35:1043–1053. plants. J. Chem. Ecol. 29:1143–1158. NYMAN,T.andJULKUNEN-TIITTO, R. 2005. Chemical variation within and HARBORNE, J. B. 2001. Twenty-five years of chemical ecology. Nat. among six northern willow species. Phytochemistry 66:2836–2843. Prod. Rep. 18:361–379. ORIANS, C. M. 2000. The effects of hybridization in plants on HAUKIOJA, E. 1980. On the role of plant defences in the fluctuation of secondary chemistry: implications for the ecology and evo- herbivore populations. Oikos 35:202–213. lution of plant-herbivore interactions. Am. J. Bot. 87:1749–1756. HEGARTY, M. J., BARKER, G. L., BRENNAN, A. C., EDWARDS, K. J., ORIANS,C.M.andFRITZ, R. S. 1996. Genetic and soil-nutrient effects on ABBOTT, R. J., and HISCOCK, S. J. 2008. Changes to gene expres- the abundance of herbivores on willows. Oecologia 105:388–396. sion associated with hybrid speciation in plants: further insights ORIANS, C. M., HUANG, C. H., WILD, A., DORFMAN, K. A., ZEE,P., from transcriptomic studies in Senecio. Philos. T. Roy. Soc. B DAO, M. T. T., and FRITZ, R. S. 1997. Willow hybridization 363:3055–3069. differentially affects preference and performance of herbivorous HJÄLTÉN, J. 1998. An experimental test of hybrid resistance to insects beetles. Entomol. Exp. Appl. 83:285–294. and pathogens using Salix caprea, S. repens and their F1 hybrids. ORIANS, C. M., GRIFFITHS, M., ROCHE, B. M., and FRITZ, R. S. 2000. Oecologia 117:127–132. Phenolic glycosides and condensed tannins in Salix sericea, S. HJÄLTÉN, J. and HALLGREN, P. 2002. The resistance of hybrid willows eriocephala and their F1 hybrids: not all hybrids are created to specialist and generalist herbivores and pathogens: the potential equal. Biochem. Syst. Ecol. 28:619–623. role of secondary chemistry and parent host plant status, pp. 153– PORTER, L. J., HRSTICH, L. N., and CHAN, B. G. 1986. The conversa- 168, in M. R. Wagner, K. M. Clancy, F. Lieutier, and T. D. Paine tion of proanthocyanidins and prodelohinidins to cyanidins and (eds.), Mechanisms and Deployment of Resistance in Trees to delphinidins. Phytochemistry 25:223–230. Insects. Kluwer Academic Publishers, New York. REBEIZ, M., JIKOMES, N., KASSNER, V. A., and CARROLL, S. B. 2011. HOCHWENDER, C. G. and FRITZ, R. S. 2004. Plant genetic differences Evolutionary origin of a novel gene expression pattern through influence herbivore community structure: evidence from a hybrid co-option of the latent activities of existing regulatory sequences. willow system. Oecologia 138:547–557. Proc. Natl. Acad. Sci. USA 108:10036–10043. HOCHWENDER, C. G., FRITZ, R. S., and ORIANS, C. M. 2000. Using RIESEBERG, L. H. and ELLSTRAND, N. C. 1993. What can molecular hybrid systems to explore the evolution of tolerance to damage. and morphological markers tell us about plant hybridization? Crit. Evol. Ecol. 14:509–521. Rev. Plant Sci. 12:213–241. JULKUNEN-TIITTO, R. 1986. A chemotaxonomic survey of phenolics in ROWELL-RAHIER, M. 1984. The presence or absence of phenolglycosides leaves of northern Salicaceae species. Phytochemistry 25:663– in Salix (Salicaceae) leaves and the level of dietary specialization 667. of some of their herbivorous insects. Oecologia 62:26–30. JULKUNEN-TIITTO, R. 1989. Phenolic constituents of Salix: a chemo- SEIGLER, D. S. 1998. Plant Secondary Metabolism. Kluwer Academic taxonomic survey of further Finnish species. Phytochemistry Publishers, USA. 28:2115–2125. SPRING, O. and SCHILLING, E. E. 1989. Chemosystematic investigation JULKUNEN-TIITTO, R. and SORSA, S. 2001. Testing the effects of drying of the annual species of Helianthus (Asteraceae). Biochem. Syst. methods on willow flavonoids, tannins, and salicylates. J. Chem. Ecol. 17:519–528. Ecol. 27:779–789. STUPAR,R.M.,HERMANSON,P.J.,andSPRINGER, N. M. 2007. KELLY,M.T.andCURRY, J. P. 1991. The influence of phenolic Nonadditive expression and parent-of-origin effects identified by compounds on the suitability of three Salix species as hosts microarray and allele-specific expression profiling of maize en- for the willow beetle Phratora vulgatissima. Entomol. Exp. dosperm. Plant Physiol. 145:411–425. Appl. 61:25–32. TAHVANAINEN,J.,JULKUNEN-TIITTO,R.,andKETTUNEN, J. 1985. KENDALL, D. A., HUNTER, T., ARNOLD, G. M., LIGGITT, J., MORRIS, Phenolic glucosides govern the food selection pattern of willow T., and WILTSHIRE, C. W. 1996. Susceptibility of willow clones feeding leaf beetles. Oecologia 67:52–56. (Salix spp.) to herbivory by Phyllodecta vulgatissima (L.) and TREUTTER, D. 2006. Significance of flavonoids in plant resistance: a Galerucella lineola (Fab.) (Coleoptera, Chrysomelidae). Ann. review. Environ. Chem. Lett. 4:147–157. Appl. Biol. 129:379–390. WANG, J., TIAN, L., LEE, H.-S., WEI, N. E., JIANG, H., WATSON, B., KIRK, H., VRIELING, K., VAN DER MEIJDEN, E., and KLINKHAMER,P. MADLUNG, A., OSBORN, T. C., DOERGE, R. W., COMAI, L., and G. L. 2010. Species by environment interactions affect CHEN, Z. J. 2006. Genomewide nonadditive gene regulation in pyrrolizidine alkaloid expression in Senecio jacobaea, Senecio Arabidopsis allotetraploids. Genetics 172:507–517. aquaticus, and their hybrids. J. Chem. Ecol. 36:378–387. WHITHAM, T. G. 1989. Plant hybrid zones as sinks for pests. Science LEHRMAN, A., TORP, M., STENBERG, J. A., JULKUNEN-TIITTO, R., and 244:1490–1493. BJÖRKMAN, C. 2012. Estimating direct resistance in willows YARNES, C. T., BOECKLEN,W.J.,AND SALMINEN, J. P. 2008. No against a major insect pest (Phratora vulgatissima) comparing simple sum: seasonal variation in tannin phenotypes and leaf- life history traits. Entomol. Exp. Appl. 144:93–100. miners in hybrid oaks. Chemoecology 18:39–51.