SEARCH

SEARCH BY CITATION

Keywords:

  • arbuscular mycorrhiza;
  • herbivore-induced plant volatiles;
  • induced plant defence;
  • multi-trophic interaction;
  • predatory mites

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

1. Indirect induced plant defence via emission of herbivore-induced plant volatiles (HIPV) to recruit natural enemies of herbivores is a ubiquitous phenomenon, but whether and how emission of above-ground HIPVs is adaptively modulated by below-ground mutualistic micro-organisms is unknown.

2. We investigated the effects of the mycorrhizal fungus Glomus mosseae on chemical composition of HIPVs emitted by bean plants Phaseolus vulgaris attacked by spider mites, Tetranychus urticae, using proton-transfer mass spectrometry, and attraction of the spider mites’ natural enemy, the predatory mite Phytoseiulus persimilis, to these HIPVs using a Y-tube olfactometer.

3. Mycorrhiza significantly changed the HIPV composition. Most notably, it increased the emission of β-ocimene and β-caryophyllene, two compounds synthesized de novo upon spider mite attack. The constitutively emitted methyl salicylate was increased by spider mite infestation but decreased by mycorrhiza.

4. The predators responded strongly to HIPVs emitted by plants infested for 6 days and preferred HIPVs of mycorrhizal plants to those of non-mycorrhizal plants. In contrast, they were less responsive and indiscriminative to mycorrhization when exposed to volatiles emitted by non-infested plants and plants infested by spider mites for 1 or 3 days.

5. Our study provides a key example of an adaptive indirect HIPV-mediated interaction of a below-ground micro-organism with an above-ground carnivore.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Plants may either constitutively or after induction defend themselves against herbivores, either directly through chemical substances and/or morphological structures negatively affecting the herbivores or indirectly by favouring establishment and/or attraction of the herbivores’ natural enemies (e.g. Price et al. 1980; Karban & Baldwin 1997; Sabelis et al. 1999). One form of induced indirect defence of plants is the production and release of volatiles attracting third trophic-level natural enemies such as carnivorous predators and parasitoids. Systemic release of natural enemy attracting volatiles upon herbivore attack of above-ground plant parts is a ubiquitous phenomenon (for review, Dicke & Vet 1999; Paré & Tumlinson 1999; Hare 2011). Analogous phenomena may occur below-ground upon attack of the roots (e.g. Horiuchi et al. 2005; Rasmann et al. 2005; Wenke, Kai & Piechulla 2010). However, below-ground emission of herbivore-induced plant volatiles (HIPVs) is less well documented, which is not necessarily because it is less common than above-ground emission, but possibly also because the soil environment is more difficult to study. Based on observations that below- and above-ground plant-associated processes are mutually dependent (Van der Putten et al. 2001), recent research revealed that indirect defence mechanisms such as volatile emission induced by herbivory on below- and above-ground plant parts commonly interact (Bezemer & van Dam 2005; Rasmann & Turlings 2007; Soler et al. 2007; Erb et al. 2008). Similarly, below- and above-ground micro-organisms may affect each other, mediated by induced plant defence reactions (Walters & Heil 2007). While these examples document that processes induced by above- and below-ground herbivores or by above- and below-ground micro-organisms may interfere with each other, little is known about how below-ground micro-organisms affect herbivore-induced volatile production (Fontana et al. 2009; Leitner et al. 2010) and associated recruitment of natural enemies above-ground (Guerreri et al. 2004). This is an important field of research because on above-ground plant parts, the endogenous biochemical cascades (the salicylate and jasmonate/ethylene-mediated pathways) and defence reactions induced by pathogenic micro-organisms and herbivores often interfere with each other due to pleiotropic effects and resource allocation trade-offs (Thaler et al. 1999, 2002; Bostock et al. 2002).

The effects of mutualistic below-ground micro-organisms such as plant-growth-promoting rhizobacteria and arbuscular mycorrhizal fungi on induced indirect plant defences against above-ground herbivores are largely unknown. Van Oosten et al. (2008) did not find an effect of the plant-growth-promoting rhizobacteria Pseudomonas sp. on volatile emission of Arabidopsis thaliana attacked by caterpillars of the small white Pieris rapae or the beet armyworm Spodoptera exigua, measured in the level of attraction of the parasitoid Cotesia rubecula. Only a few studies looked into the effects of mycorrhizal symbiosis, the association between plant roots and mycorrhizal fungi, on volatiles emitted by above-ground plant parts, three of them dealt with HIPVs. Rapparini, Llusià & Peňuelas (2008) detected constitutive changes in volatile production following mycorrhization (by Glomus spp.) of Artemisia annua but did not test for changes in volatiles induced by herbivore attacks. Fontana et al. (2009) and Leitner et al. (2010) determined mycorrhiza (Glomus intraradices)-induced quantitative and qualitative changes in the volatile blends released by above-ground plant parts of Plantago lanceolata (Fontana et al. 2009) and Medicago truncatula (Leitner et al. 2010) induced by caterpillars of Spodoptera spp. Guerreri et al. (2004) observed stronger attraction of the parasitoid Aphidius ervi to constitutive volatiles of mycorrhizal (Glomus mosseae) tomato plants as compared to non-mycorrhizal plants. However, the mycorrhiza-induced behavioural changes of the parasitoids occurred independent of, and were not altered by, the presence of their hosts, the plant-attacking aphid Macrosiphum euphorbiae. This result is evolutionarily puzzling in a tri-trophic context: the parasitoids do not gain any benefits from the plants themselves and would therefore be betrayed by the mycorrhiza–plant mutualism due to enhanced attraction to plants without receiving a reward, that is, potential hosts. In any case, Guerreri et al. (2004) is the only study attempting to link mycorrhizal symbiosis and prey/host searching based on plant-emitted volatiles by above-ground third trophic-level natural enemies. No study to date pinpointed the chemical changes in the volatile blends and simultaneously linked them to behavioural changes of the herbivores’ natural enemies.

Depending on the species involved and the ecological context, arbuscular mycorrhizal symbiosis is commonly but not always beneficial to both partners and hence considered mutualistic (Smith & Read 2008; Hoeksema et al. 2010). Among other (ex)changes, the mutualism mainly involves plants trading carbon for nutrients from the fungus. In some systems, mycorrhiza increases the vegetative and/or reproductive growth of their host plants but can simultaneously favour plant-attacking, and thereby plant-fitness-decreasing, organisms such as herbivores. This is, for example, true for the interaction between the mycorrhizal fungus Glomus mosseae (Nicol. & Gerd.), common bean plants Phaseolus vulgaris L. (e.g. Isobe & Tsuboki 1999) and the herbivorous two-spotted spider mite Tetranychus urticae Koch (Hoffmann et al. 2009). Spider mites are highly detrimental to both, the bean plant and the mycorrhizal fungus (Hoffmann et al. 2011), and should thus be important drivers in the evolution of the bean–fungus interaction. We argue that for evolutionary optimization and stability of the fungus–plant mutualism, mycorrhizal symbiosis should compensate for herbivore enhancement by simultaneously enhancing the performance of the herbivores’ main natural enemies, predatory mites of the family Phytoseiidae.

The first chronological step in manipulating the above-ground tri-trophic plant–herbivore–carnivore interaction is to increase the attractiveness of herbivore-infested plants to recruit more third trophic-level natural enemies. We therefore hypothesized that mycorrhiza-induced changes in plant chemistry and associated volatile emission should lead to stronger attraction of third trophic-level natural enemies in multi-trophic systems where mycorrhiza enhances the performance of the herbivores of their host plants, as observed in the interaction between G. mosseae, P. vulgaris and T. urticae (Hoffmann et al. 2009). Accordingly, we compared (i) the composition of volatile blends released by common bean plants living in symbiosis or not with G. mosseae and attacked above-ground by T. urticae or not using proton-transfer reaction time-of-flight mass spectrometry (PTR-TOF-MS), and linked possible differences in the volatile blends to (ii) the response of a specialized natural enemy of the spider mites, the predatory mite Phytoseiulus persimilis Athias-Henriot (Fig. 1), to volatiles from mycorrhizal and non-mycorrhizal plants in binary-choice situations using a Y-tube olfactometer.

image

Figure 1.  Gravid female of the predatory mite Phytoseiulus persimilis (body length, approximately 0·5 mm) in a spider mite patch (© Peter Schausberger).

Download figure to PowerPoint

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Mites and plants

The tri-trophic system of P. vulgaris, T. urticae and predatory mites such as P. persimilis is a widely used, well-studied model system in research on HIPV (e.g. Sabelis & van de Baan 1983; Dicke et al. 1990; Sabelis et al. 1999). Recent investigations documented that this system is also perfectly suited for studying multi-trophic below- and above-ground interactions between plant-associated organisms such as the interaction between mycorrhizal fungi and herbivorous and carnivorous mites living on above-ground plant parts (Hoffmann et al. 2009, 2011; Hoffmann, Vierheilig & Schausberger 2011a,b).

Leaf samples used in olfactometer choice tests and mass spectrometry were derived from common bean plants, P. vulgaris var. ‘Taylor’s Horticultural’, colonized (hereafter termed +M or mycorrhizal plants) and not colonized (hereafter termed −M or non-mycorrhizal plants) by the arbuscular mycorrhizal fungus G. mosseae, and infested (+SM) or not infested (−SM) by the spider mite T. urticae. To generate mycorrhizal and non-mycorrhizal plants, we followed the protocol described by Hoffmann et al. (2009) with respect to plant growing, fungal inoculation, watering and fertilization. The G. mosseae inoculum (BEG 12) was originally obtained from the International Bank of Glomeromycota (http://www.kent.ac.uk/bio/beg). All plants used in the olfactometer tests and for mass spectrometry were randomly chosen and checked for the degree of mycorrhizal colonization. We used detached leaves instead of leaves attached to the potted plant or whole plants because this procedure better allows for standardization of age, functional part and biomass of the plant material (Choh & Takabayashi 2006). Additionally, confounding volatiles from other parts than the leaves such as the potting substrate or the roots can be more easily excluded. HIPVs of detached leaves and leaves attached to the plants are qualitatively similar but may differ quantitatively (e.g. for bean, Dicke et al. 1990; De Boer et al. 2008; for maize, Schmelz, Alborn & Tumlinson 2003; Williams et al. 2005). Leaves were immediately used in choice tests after detachment. After leaves used in choice tests had been detached, the remaining above-ground plant parts and the roots were removed from the planting pot. For plants used for mass spectrometry, 1 week before the experiment took place, a soil sample (approximately 22 cm3) was taken from each pot using a cork borer (Ø 2·2 cm). In either case, the planting substrate was rinsed off the roots with cold tap water. Roots were cleared by boiling them for 10 min in 10% KOH and stained by boiling for 5 min in a 5% black ink (Schaeffer, Ft. Madison, USA) and household vinegar (equal to 5% acetic acid) solution (Vierheilig et al. 1998). The percentage of root length colonized (RLC) by G. mosseae was estimated according to the modified gridline intersect method (Giovannetti & Mosse 1980). All 36 mycorrhizal plants (+M) used for the olfactometer tests and the eight mycorrhizal plants used for mass spectrometry had >10% RLC and on average 20·76 ± 1·16 (SE) and 29·75 ± 3·70 (SE) % RLC, respectively. All non-mycorrhizal plants (−M) had 0% RLC. Plants used for the olfactometer tests and mass spectrometry were 25–30 days post-inoculation with the mycorrhizal fungus.

The stock population of T. urticae was maintained on whole non-mycorrhizal bean plants at 25 ± 5 °C, 60–80% rH, 16:8 h L:D. Phytoseiulus persimilis subjected to olfactometer tests derived from a laboratory-reared population founded with specimens collected on clementine trees in Onda, Spain. In the laboratory, P. persimilis was maintained on T. urticae-infested bean leaves from non-mycorrhizal plants piled up on a plastic tile resting on water-saturated foam within an open plastic box half-filled with water. Prey was supplied by adding five to seven infested bean leaves three times a week (McMurtry & Scriven 1965). Predator rearing units were stored at 25 ± 1 °C, 60 ± 5% rH and 16:8 h L:D.

Olfactometer tests

The Y-tube olfactometer used was a modification of the olfactometer described by Sabelis & van de Baan (1983) and consisted of three glass tubes (40 mm inner diameter) of equal length (130 mm) melted together in a Y-shape (Fig. 2). The two upper arms (choice arms, Fig. 2B) joined at 75° at the intersection, leaving an angle of 142·5° between each choice arm and the base arm (Fig. 2C), at the bottom of which the predatory mites were released. Each choice arm was connected airtight to an acrylic jar consisting of three tubular chambers (total length, 55 mm; inner Ø, 35 mm). The middle chamber (Fig. 2A) contained the leaf (odour) sample and was connected airtight to the other chambers with female joints. Both the inner and the outer chamber were sealed at both ends with gauze. The inner chamber was connected to the choice arm by a male joint (reaching approximately 20 mm into the choice arm), and the gauze prevented the predatory mites from accessing the chamber with the leaf sample upon reaching the end of a choice arm. The outer chamber was filled with activated charcoal to purify the air sucked into the Y-tube. A Y-shaped stainless steel wire, starting 20 mm inside the bottom end of the base arm, branching at the centre of the intersection of the three arms and leading to the end of either choice arm, was placed inside the glass Y-tube in equidistance to the inner walls. The wire was fixed by a perpendicular 20-mm-long extension held in place by an inert plastic piece fitted into a small hole of the upper wall at the bottom end of the base arm. At the end of either choice arm, the wire was centrally inserted into the gauze separating the inner chamber of the acrylic jar from the middle chamber containing the leaf (odour) sample. During tests, the olfactometer was placed on a black table flush with the surface, and a cold light source was centred above the intersection of the Y-tube. Air was drawn through the Y-tube using a mini-diaphragm-vacuum pump (Laboport® N 86 KN.18; KNF Neuberger, Freiburg, Germany) connected to the bottom end of the base arm. Air was sucked in at the ends of the choice arms and was drawn through the chambers containing the charcoal and the leaf samples, down the choice arms (flow rate, 2·5 l min−1 per arm), and, after meeting at the intersection, down the base arm, and left the Y-tube at the bottom of the base arm (flow rate, 5·0 l min−1).

image

Figure 2.  Y-tube olfactometer used for the binary-choice tests, consisting of a Y-shaped cylindrical glass tube (inner diameter, 40 mm, each arm of Y 130 mm long) with a wire inside functioning as railroad for the predatory mites (© Peter Schausberger). Each choice arm (B) was distally connected to a chamber (A) containing a leaf sample of a mycorrhizal or non-mycorrhizal plant. The predatory mites were singly released on the wire at the bottom end of the base arm (C), which was thereafter connected to a suction pump (not shown here).

Download figure to PowerPoint

Each predatory mite was given a choice between the volatiles emanating from a leaf sample from a non-mycorrhizal and a mycorrhizal plant, both of which were either infested by T. urticae or non-infested. Infested plants were created by placing 30 gravid spider mite females onto each plant (randomly distributed among the trifoliate leaves). Since HIPV emission depends on the duration of infestation (e.g. Nachappa et al. 2006), we allowed the spider mites to feed and oviposit for 1, 3 or 6 days. Different plants were used for each of the four treatments (non-infested, 1, 3 and 6 days spider mite infestation). The spider mite eggs did not hatch within 6 days keeping the density of spider mites feeding on the plants constant between mycorrhizal and non-mycorrhizal treatments over time. The three youngest fully developed trifoliate leaves with their petioles were detached from the plants and immediately used for the olfactometer tests. Before tests, gravid P. persimilis females were randomly withdrawn from the rearing units, singly caged in closed acrylic cells (Schausberger 1997) and starved for approximately 20 h. Only females laying at least one egg during the starvation period were used for the choice test. Each predatory mite was released singly at the bottom end of the wire inside the glass tube, using a moistened fine brush, and then observed for 5 min at maximum whether it moved to an end of a choice arm. Predatory mites falling from the wire were discarded from analyses. Predatory mites not reaching the end of a choice arm within 5 min after release were judged as non-responsive. For each predatory mite, we recorded the responsiveness (reached the end of a choice arm or not) and, if responding, their decision (+M or −M odour source; left or right arm). Nine plant sample pairs, each consisting of a different set of a mycorrhizal and a non-mycorrhizal plant, were used for each binary-choice combination. Five to 10 predatory mites were tested per plant sample pair. After every five to 10 predatory mites, the chambers containing the leaf samples were disconnected from the Y-tube, the +M and −M leaf samples switched between arms, to avoid any inadvertent positional effect, and the wire cleaned with ethanol to avoid any influence of traces left by the predatory mites on the wire. All olfactometer tests were carried out within a couple of consecutive days in an air-conditioned room at 25 ± 1 °C.

Mass spectrometry

Corresponding to the olfactometer tests, we evaluated the volatiles emitted by non-mycorrhizal and mycorrhizal plants that were either infested by T. urticae or not, resulting in four treatments. Infested plants were created by placing 30 gravid spider mite females onto each plant (randomly distributed among the trifoliate leaves), which were then allowed to feed and oviposit for 6 days (corresponding to the 6 days treatment in the olfactometer tests). For analysis, three fully developed trifoliate leaves with their petioles were detached from the plant, weighed and thereafter immediately placed in a cylindrical glass container (length, 149 mm; inner Ø, 27 mm) closed at both ends with plastic screw caps. The sample container was connected airtight via tubes (length, approximately 100 mm) to a container with the charcoal filter (Supelco Supelpure™ HC; Sigma-Aldrich, Vienna, Austria) on the outer end (air in) and the mass spectrometer on the inner end (air out).

Real-time trace gas analysis of the leaves was performed using the high-sensitivity PTR-TOF-MS instrument PTR-TOF-8000 (Ionicon, Innsbruck, Austria) (for details, Jordan et al. 2009). The volatile blend of each leaf sample (four samples per treatment) was measured twice, using two modes of airflow. Airflow was either set at 130 sccm (standard cm3 min−1) for 15 min (high airflow) or set at 30 sccm for 10 min (reduced airflow), immediately following the high airflow measurement. For analyses, volatiles detected during high and reduced airflow were considered autocorrelated measurements. Instrument adjustments were 80 °C inlet system heat and 4 s dwell time. After the last sample of each treatment, the sample container was cleaned with acetone and allowed to air-dry for at least 1 h before being loaded with the first sample of another treatment. Additionally, we measured the air inside the empty sample container before loading with the first sample of each treatment at an airflow of 130 sccm for 13 min to determine the presence of background volatiles.

The PTR-TOF-MS instrument measures the mass spectrum every 4 s. Therefore, we calculated the mean mass spectrum for each measurement (lasting 15 and 10 min for the leaf samples and 13 min for the background) using the software TOF sampler 1.0.0 (Ionicon). We checked the mass spectra for presence of 19 compounds, known to be part of the volatile blends of common bean plants influenced by spider mite attack and to have major or minor importance for predatory mite recruitment (Dicke et al. 1990; De Boer, Posthumus & Dicke 2004; De Boer et al. 2008; Zhang et al. 2009), and determined the ion yield at the relevant atomic masses. To obtain the net ion yields, we subtracted the ion yields in the background volatiles from the ion yields in the leaf sample volatiles. With the method used, it is not possible to distinguish between two compounds (subsequently indicated by AND/OR) with the same atomic mass.

Statistical analyses

spss 15.0 (SPSS Inc., Chicago, IL, USA) was used for all statistical analyses. We used generalized linear models (GLM, binomial distribution, logit link) to assess the effect of plant sample within each choice situation (non-infested, infested for 1, 3 or 6 days) on responsiveness to and preference for the mycorrhizal or non-mycorrhizal volatile blend. Similarly, the effect of choice situation (non-infested, infested for 1, 3 or 6 days) on responsiveness of predatory mite females to volatile blends (reaching the end of an arm with the mycorrhizal or non-mycorrhizal odour source or not) was assessed by a GLM (binomial distribution, logit link) and Šidák post hoc comparisons using the choice situation with non-infested plants as the reference category. Within each choice situation, we used two-tailed binomial tests to assess whether the preference of the predatory mites for the mycorrhizal and non-mycorrhizal odour source and the left and right arm of the Y-tube, respectively, differed from random choice.

The influence of mycorrhiza and spider mite infestation on the weight of the leaf samples used in mass spectrometry was analysed by a GLM (normal distribution, identity link). To test whether the total composition of the volatiles was affected by mycorrhiza and the spider mites, we used generalized estimating equations (Hardin & Hilbe 2003) (GEE, normal distribution, identity link, exchangeable autocorrelation structure between the two modes of airflow) with compound nested within mycorrhiza, within spider mites, and within the interaction of mycorrhiza with the spider mites as between-subject variables. Subsequently, to test which compounds differed among treatments, separate GEEs (normal distribution, identity link) were used to assess the effects of mycorrhiza and spider mite infestation as between-subject factors, and mode of airflow (within-subject factor with an exchangeable autocorrelation structure) nested within mycorrhiza and spider mite infestation on emission of single volatile compounds by bean plants, P. vulgaris.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Olfactometer tests

In neither choice situation (non-infested, infested for 1, 3 or 6 days) did plant sample have an effect on responsiveness of P. persimilis (GLM, d.f. = 8 for each situation: Waldχ2 = 5·126, = 0·744 for non-infested; Waldχ2 = 1·773, = 0·987 for 1 day infested; Waldχ2 = 12·263, = 0·140 for 3 days infested; Waldχ2 = 1·818, = 0·986 for 6 days infested). Similarly, in neither choice situation did plant sample affect the preference of the predatory mites (GLM: Waldχ2 = 0·732, d.f. = 7, = 0·998 for non-infested; Waldχ2 = 2·601, d.f. = 6, = 0·857 for 1 day infested; Waldχ2 = 2·451, d.f. = 7, = 0·931 for 3 days infested; Waldχ2 = 5·231, d.f. = 8, = 0·733 for 6 days infested). Therefore, within each choice situation, the plant samples were pooled for subsequent analyses.

The responsiveness of P. persimilis differed among choice situations (GLM: Waldχ2 = 52·321, d.f. = 3, < 0·001; Fig. 3). Around 90% predators made a choice when exposed to volatile blends of plants infested for 6 days, which was significantly more (Šidák: < 0·001 for every pairwise comparison) than the ∼30% predators making a choice in situations with non-infested plants or plants infested for 1 or 3 days. Responsiveness to non-infested plants and plants infested for 1 or 3 days was similar (Šidák: > 0·990 for every pairwise comparison).

image

Figure 3.  Number of predatory mite females moving to the odour emanating from a leaf sample from a mycorrhizal or non-mycorrhizal plant, both of which were infested by spider mites for either 1, 3 or 6 days or non-infested, in a Y-tube olfactometer. Numbers inside parenthesis after spider mite treatment represent the number of females tested. Ns (non-significant) and *< 0·001 indicate the results of two-sided binomial tests assuming an equal distribution within spider mite treatments.

Download figure to PowerPoint

The number of P. persimilis females preferring the volatile blend from mycorrhizal plants was significantly higher than the number of females preferring the volatiles from non-mycorrhizal plants in the choice situation with plants infested for 6 days (two-sided binomial test: asymptotic < 0·001; Fig. 3). The females did not show a preference in the other choice situations (= 0·307 for non-infested, = 1·000 for 1 day infested, = 0·122 for 3 days infested). In neither choice situation was the response of the females influenced by side (left or right, two-sided binomial test: asymptotic = 0·839 for non-infested; = 1·000 for 1 day infested; = 0·701 for 3 days infested; = 1·000 for 6 days infested).

Mass spectrometry

Neither mycorrhiza nor spider mite infestation affected the weight of the leaf samples (gram, grand mean: 1·48 ± 0·09 SE) used for mass spectrometry (GLM; Waldχ2 = 0·423, = 0·515 for spider mite infestation, Waldχ2 = 0·482, = 0·488 for mycorrhiza, Waldχ2 = 0·911, = 0·340 for the interaction).

Compound nested within mycorrhiza (Waldχ2 = 323·434, d.f. = 19, < 0·001), nested within spider mites (Waldχ2 = 290·948, d.f. = 19, < 0·001) and nested within the interaction of mycorrhiza with spider mites (Waldχ2 = 316·272, d.f. = 19, < 0·001) had highly significant effects on the ion yields, indicating that the composition of the nineteen compounds differed among the four treatments (−M/−SM, −M/+SM, +M/−SM, +M/+SM; Fig. 4).

image

Figure 4.  Ions of single compounds detected in the volatile blend emitted by mycorrhizal (+M) and non-mycorrhizal (−M) bean plants infested by two-spotted spider mites for 6 days (+SM) or non-infested (−SM). N = 4 for each choice situation and mode of airflow: (H) high, (R) reduced. Compound numbers represent the following: 1, 2-butanone; 2, 1-penten-3-ol AND/OR 3-pentanone; 3, 2- AND/OR 3-methylbutanal nitrile; 4, indole; 5, 1-octen-3-ol AND/OR 3-octanone; 6, 2-methylpropanal-O-methyl oxime; 7, p-mentha-1,3,8-triene; 8, β-ocimene; 9, (Z)-3-hexen-1-ol-acetate; 10, nonanal; 11, 2- AND/OR 3-methylbutanal-O-methyl oxime; 12, hexyl acetate; 13, rose furan; 14, (E)-4,8-dimethyl-1,3,7-nonatriene; 15, methyl salicylate; 16, unknown; 17, linalool; 18, decanal; 19, β-caryophyllene.

Download figure to PowerPoint

GEEs for each single compound revealed that mycorrhiza and spider mite infestation significantly changed 13 and 14, respectively, of the measured 19 compounds of the volatile blend emitted by bean plants, P. vulgaris. Of 19 compounds, 13 were affected by the interaction between mycorrhiza and spider mite infestation (Table 1, Fig. 4).

Table 1.   Generalized estimating equations (normal distribution, identity link function) for the effects of mycorrhizal symbiosis (yes/no), spider mite infestation (yes/no) and mode of airflow (high/reduced) nested within the main effects on volatile compounds emitted by bean plants, P. vulgaris. Major compounds in spider mite-induced bean plant volatile blends (Dicke et al. 1990), and P values <0·05 are highlighted in bold
CompoundMycorrhizaSpider mitesMycorrhiza* spider mites
Wald χ2PWald χ2PWald χ2P
  1. Airflow (mycorrhiza) had a significant effect on (E)-β-ocimene AND/OR (Z)-β-ocimene, methyl salicylate, linalool and decanal (Wald χ2 > 4·5, < 0·034 for all); airflow (spider mites) had a significant effect on 2-OR 3-methylbutanal nitrile, methyl salicylate and decanal (Wald χ2 > 5·1, < 0·025 for all); airflow (mycorrhiza) and airflow (spider mites) did not have an effect on any other compounds (> 0·05).

2-Butanone1·1710·2794·2960·03892·593<0·001
1-Penten-3-ol AND/OR 3-pentanone4·6830·0300·0390·8432·2800·131
2- OR 3-Methylbutanal nitrile168·777<0·00141·833<0·00167·028<0·001
Indole8·7690·0033·8220·0512·5050·113
1-Octen-3-ol AND/OR 3-octanone3·8280·0500·4420·5060·3390·561
2-Methylpropanal-O-methyl oxime18·339<0·0017·1010·00818·922<0·001
P-Mentha-1,3,8-triene21·929<0·00115·951<0·00116·783<0·001
(E)-β-Ocimene AND/OR (Z)-β-ocimene1·3000·2549·2360·0026·0660·014
(Z)-3-Hexen-1-ol-acetate1·4650·2267·7950·00545·945<0·001
Nonanal9·5050·00217·087<0·00124·845<0·001
2- OR 3-Methylbutanal-O-methyl oxime9·8570·002109·620<0·0010·0800·777
Hexyl acetate13·634<0·0010·0010·97213·679<0·001
Rose furan13·203<0·00118·738<0·00112·561<0·001
(E)-4,8-Dimethyl-1,3,7-nonatriene50·437<0·00116·530<0·0011·6280·202
Methyl salicylate61·087<0·00159·359<0·00141·482<0·001
Unknown3·1090·0788·9220·00324·351<0·001
Linalool0·0000·9983·7720·0522·0220·155
Decanal14·073<0·0018·1300·00416·232<0·001
β-Caryophyllene0·5500·45832·285<0·0017·3910·007

Three of the five compounds known to play a major role in predatory mite recruitment, β-ocimene, β-caryophyllene and methyl salicylate, were affected by the interaction between mycorrhiza and the spider mites (Table 1, Fig. 4). The only compound, of all major and minor compounds, totally unaffected by mycorrhiza, spider mites and/or their interaction was linalool. (E)-4,8-Dimethyl-1,3,7-nonatriene was increased by spider mite infestation and decreased by mycorrhiza, but there was no interaction between the main effects. β-Ocimene and β-caryophyllene were only released by spider mite-infested plants, and their amounts were clearly higher on mycorrhizal than non-mycorrhizal plants (Table 1, Fig. 4). Both non-infested and spider mite-infested plants constitutively emitted methyl salicylate. Emission of methyl salicylate was increased by spider mite infestation but decreased by mycorrhiza (Table 1, Fig. 4). The difference between non-mycorrhizal plants with and without spider mite infestation was larger with high airflow than reduced airflow, whereas the opposite was true for mycorrhizal plants. Mode of airflow nested within spider mite infestation had an effect on linalool and methyl salicylate, that is, more ions were detected in spider mite-infested plants tested with reduced flow, and β-ocimene, which was detected in non-mycorrhizal spider mite-infested plants tested with reduced but not high flow. Mode of airflow nested within mycorrhiza had an effect on methyl salicylate, that is, more ions were detected with reduced flow.

Every compound of minor or unknown importance for recruitment of predatory mites from the distance was affected by mycorrhiza and/or the spider mites and/or their interaction (Table 1, Fig. 4). For example, mycorrhiza increased the amount of 1-penten-3-ol AND/OR 3-pentanone and reduced the amount of indole and 1-octen-3-ol AND/OR 3-octanone. Spider mite infestation increased the amount of 2-butanone emitted by mycorrhizal plants but decreased its emission on non-mycorrhizal plants. Spider mite infestation decreased the amount of 2- AND/OR 3-methylbutanal nitrile and 2-methylpropanal-O-methyl oxime on mycorrhizal plants but increased the amount on non-mycorrhizal plants. Mycorrhizal plants without spider mites did not emit p-mentha-1,3,8-triene, but plants of all other treatments did. Mode of airflow nested within mycorrhiza had an effect on the emission of 2- AND/OR 3-methylbutanal nitrile and decanal: both compounds were present with more ions in the blends of mycorrhizal plants tested with reduced flow. Mode of airflow nested within spider mites had an effect on decanal, that is, more ions were detected with reduced flow (Table 1, Fig. 4).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Mycorrhizal symbiosis quantitatively and qualitatively changed the emission of HIPVs. HIPVs of mycorrhizal plants were more attractive to the predatory mite P. persimilis than HIPVs of non-mycorrhizal plants. Whether stronger attraction of P. persimilis was due to changed overall composition of the volatile blend or to changes in single compounds needs further investigation. However, of the major compounds known to attract P. persimilis (Dicke et al. 1990), mycorrhizal symbiosis most notably increased the difference in β-ocimene emission between spider mite-infested and non-infested plants, whereas it decreased this difference in methyl salicylate emission. β-Ocimene on mycorrhizal plants and 2-butanone on non-mycorrhizal plants were the most abundant among the compounds influenced by herbivory.

Three compounds were synthesized de novo upon herbivore attack: 2- AND/OR 3-methylbutanal-O-methyl oxime was reduced by mycorrhiza, β-caryophyllene was not changed by mycorrhiza and only β-ocimene was affected by the interaction between mycorrhiza and the spider mites. These changes were similarly detected by both modes of airflow. Our study did not allow us to distinguish between the Z and E isomers of β-ocimene. However, the finding by Dicke et al. (1990) that (E)-β-ocimene is attractive whereas (Z)-β-ocimene is repellent for P. persimilis suggests that mycorrhiza mainly increased the E isomer. No compound was synthesized de novo by mycorrhization. Methyl salicylate and (E)-4,8-dimethyl-1,3,7-nonatriene were constitutively released by all plants but quantitatively changed by mycorrhiza, by the spider mites or by both. In comparison with our study, Leitner et al. (2010) did not detect mycorrhiza-induced changes in the emission of major HIPV compounds by M. truncatula but observed that mycorrhizal plants emitted higher amounts of minor compounds such as α-gurjunene. Fontana et al. (2009) observed that mycorrhizal herbivore-infested plants emitted lower amounts of terpenoids such as β-ocimene and β-caryophyllene than non-mycorrhizal herbivore-infested plants did. Whether mycorrhiza-induced changes in HIPVs of P. lanceolata (Fontana et al. 2009) and M. truncatula (Leitner et al. 2010) would affect third trophic-level natural enemies was not assessed. All three studies, Fontana et al. (2009), Leitner et al. (2010) and our study, suggest that the mycorrhiza-induced changes in volatile emission are highly sophisticated. In any case, these changes are more than a mere increase in the amount of emitted volatiles due to improved nutrient uptake. The large qualitative differences in the manipulation of single HIPV compounds by mycorrhiza observed in our study, and the studies by Leitner et al. (2010) and Fontana et al. (2009) point at strong dependency of the outcome on the involved fungus, plant and herbivore species.

As in previous studies showing that mycorrhizal symbiosis amends attractiveness and prey quality of the spider mites for the predatory mites (Hoffmann et al. 2009; Hoffmann, Vierheilig & Schausberger 2011a,b) and increases plant tolerance to herbivory by spider mites (Hoffmann et al. 2011a), this study suggests yet another compensating mechanism for herbivore enhancement, that is, change in HIPVs, leading to stronger attraction of the natural enemy of the spider mites, the predatory mite P. persimilis, to mycorrhizal than non-mycorrhizal plants. The change in HIPVs is adaptive for the predator, the plant and the fungus because it guides the predators to plants with more nutritious prey, consequently increasing predator fitness (Hoffmann, Vierheilig & Schausberger 2011b), and increases fitness of the mycorrhizal fungus due to the predators’ negative effect on the spider mites, relaxing herbivore pressure on the plant and in turn keeping up the fungal root colonization levels (Hoffmann et al. 2011a).

Our study is a key example of a sophisticated compensating mechanism via enhanced recruitment of third trophic-level natural enemies in multi-trophic systems where mycorrhizal symbiosis promotes herbivory. Assuming that the quality (positive, neutral or negative) of the effect of mycorrhizal symbiosis on the herbivores is decisive, this should also apply to systems with herbivores having different feeding modes (Gehring & Whitham 2002) and where roots are colonized by multiple AM fungi, which is common in natural settings (e.g. Smith & Read 2008). However, the ubiquity of this mechanism regarding plant, fungus and herbivore species and context (in)dependency remains to be shown. In general, the effects of mycorrhizal symbiosis on above-ground herbivores are highly variable, ranging from positive to neutral to negative (e.g. Gehring & Whitham 2002; Bennett, Alers-Garcia & Bever 2006). For inter-trophic-level signalling reliability and honesty, enhanced third trophic-level natural enemy recruitment through mycorrhiza-induced changes in volatile emission is not expected in systems where mycorrhizal symbiosis negatively affects the herbivores.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

We thank Philipp Sulzer (Ionicon GmbH) for help with mass spectrometry and Markus Strodl and Andreas Walzer (both University of Natural Resources and Life Sciences) for comments on an earlier version of this manuscript.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Bennett, A.E., Alers-Garcia, J. & Bever, J.D. (2006) Three-way interactions among mutualistic mycorrhizal fungi, plants, and plant enemies: hypotheses and synthesis. American Naturalist, 167, 141152.
  • Bezemer, M. & van Dam, N.M. (2005) Linking aboveground and belowground interactions via induced plant defenses. Trends in Ecology and Evolution, 20, 617624.
  • Bostock, R.M., Karban, R., Thaler, J.S., Weyman, P.D. & Gilchrist, D. (2002) Signal interactions in induced resistance to pathogens and insect herbivores. European Journal of Plant Pathology, 107, 103111.
  • Choh, Y. & Takabayashi, J. (2006) Intact lima bean plants exposed to herbivore-induced plant volatiles attract predatory mites and spider mites at different levels according to plant parts. Applied Entomology and Zoology, 42, 537543.
  • De Boer, J.G., Posthumus, M.A. & Dicke, M. (2004) Identification of volatiles that are used in discrimination between plants infested with prey or non-prey herbivores by a predatory mite. Journal of Chemical Ecology, 30, 22152230.
  • De Boer, J.G., Hordijk, C.A., Posthumus, M.A. & Dicke, M. (2008) Prey and non-prey arthropods sharing a host plant: effects on induced volatile emission and predator attraction. Journal of Chemical Ecology, 34, 281290.
  • Dicke, M. & Vet, L.E.M. (1999) Plant-carnivore interactions: evolutionary and ecological consequences for plant, herbivore and carnivore. Herbivores. Between Plants and Predators (eds H. Olff, V.K. Brown & R.H. Drent), pp. 483520. Blackwell Science, Oxford.
  • Dicke, M., van Beek, T.A., Posthumus, M.A., Ben Dom, N., van Bokhoven, H. & de Groot, A.E. (1990) Isolation and identification of volatile kairomone that affects acarine predator–prey interactions. Journal of Chemical Ecology, 16, 381396.
  • Erb, M., Ton, J., Degenhardt, J. & Turlings, T.C.J. (2008) Interactions between arthropod-induced aboveground and belowground defenses in plants. Plant Physiology, 146, 867874.
  • Fontana, A., Reichelt, M., Hempel, S., Gershenzon, J. & Unsicker, S.B. (2009) The effects of arbuscular myccorhizal fungi on direct and indirect defense metabolites of Plantago lanceolata L. Journal of Chemical Ecology, 35, 833843.
  • Gehring, C.A. & Whitham, T.G. (2002) Mycorrhizae-herbivore interactions: population and community consequences. Mycorrhizal Ecology (eds M.G.A. van der Heijden & I.R. Sanders), pp. 295320. Springer, Berlin.
  • Giovannetti, M. & Mosse, B. (1980) An evaluation of techniques for measuring vesicular arbuscular infection in roots. New Phytologist, 84, 489500.
  • Guerreri, E., Lingua, G., Digilio, M.C., Massa, N. & Berta, G. (2004) Do interactions between plant roots and the rhizosphere affect parasitoid behaviour? Ecological Entomology, 29, 753756.
  • Hardin, J.W. & Hilbe, J.M. (2003) Generalized Estimating Equations. Chapman & Hall/CRC, Boca Raton, FL, USA.
  • Hare, J.D. (2011) Ecological role of volatiles produced by plants in response to damage by herbivorous insects. Annual Review of Entomology, 56, 161180.
  • Hoeksema, J.D., Chaudhary, V.B., Gehring, C.A., Collins Johnson, N., Karst, J., Koide, R.T., Pringle, A., Zabinski, C., Bever, J.D., Moore, J.C., Wilson, G.W.T., Klironomos, J.N. & Umbanhowar, J. (2010) A meta-analysis of context-dependency in plant response to inoculation with mycorrhizal fungi. Ecology Letters, 13, 394407.
  • Hoffmann, D., Vierheilig, H. & Schausberger, P. (2011a) Mycorrhiza-induced trophic cascade enhances fitness and population growth of an acarine predator. Oecologia, 166, 141149.
  • Hoffmann, D., Vierheilig, H. & Schausberger, P. (2011b) Arbuscular mycorrhiza enhances preference of ovipositing predatory mites for direct prey-related cues. Physiological Entomology, 36, 9095.
  • Hoffmann, D., Vierheilig, H., Riegler, P. & Schausberger, P. (2009) Arbuscular mycorrhizal symbiosis increases host plant acceptance and population growth rates of the two-spotted spider mite Tetranychus urticae. Oecologia, 158, 663671.
  • Hoffmann, D., Vierheilig, H., Peneder, S. & Schausberger, P. (2011) Mycorrhiza modulates aboveground tri-trophic interactions to the fitness benefit of its host plant. Ecological Entomology, 36, 574581.
  • Horiuchi, J., Prithiviraj, B., Bais, H.P., Kimball, B.A. & Vivanco, J.M. (2005) Soil nematodes mediate positive interactions between legume plants and rhizobium bacteria. Planta, 222, 848857.
  • Isobe, K. & Tsuboki, Y. (1999) Study on the utilization of arbuscular mycorrhizal fungi in kidney bean cultivation – differences in kidney bean growth resulting from the infection of six species of arbuscular mycorrhizal fungi. Japanese Journal of Crop Science, 68, 112117.
  • Jordan, A., Haidacher, S., Hanel, G., Hartungen, E., Märk, L., Seehauser, H., Schottkowsky, R., Sulzer, P. & Märk, T.D. (2009) A high resolution and high sensitivity proton-transfer-reaction time-of-flight mass spectrometer (PTR-TOF-MS). International Journal of Mass Spectrometry, 286, 122128.
  • Karban, R. & Baldwin, I. (1997) Induced Responses to Herbivory. Chicago University Press, Chicago.
  • Leitner, M., Kaiser, R., Hause, B., Boland, W. & Mithöfer, A. (2010) Does mycorrhization influence herbivore-induced volatile emission in Medicago truncatula? Mycorrhiza, 20, 89101.
  • McMurtry, J.A. & Scriven, G.T. (1965) Insectary production of phytoseiid mites. Journal of Economic Entomology, 58, 282283.
  • Nachappa, P., Margolies, D.C., Nechols, J.R. & Loughin, T. (2006) Phytoseiulus persimilis response to herbivore-induced plant volatiles as a function of mite-days. Experimental and Applied Acarology, 40, 231239.
  • Paré, P.W. & Tumlinson, J.H. (1999) Plant volatiles as a defense against herbivores. Plant Physiology, 121, 325331.
  • Price, P.W., Bouton, C.E., Gross, P., McPheron, B.A., Thompson, J.N. & Weis, A.E. (1980) Interactions among three trophic levels: influence of plants on interactions between insect herbivores and natural enemies. Annual Review of Ecology and Systematics, 11, 4165.
  • Rapparini, F., Llusià, J. & Peňuelas, J. (2008) Effect of arbuscular mycorrhizal (AM) colonization on terpene emission and content of Artemisia annua L. Plant Biology, 10, 108122.
  • Rasmann, S. & Turlings, T.C.J. (2007) Simultaneous feeding by aboveground and belowground herbivores attenuates plant-mediated attraction of their respective natural enemies. Ecology Letters, 10, 926936.
  • Rasmann, S., Kollner, T.G., Degenhardt, J., Hiltpolt, I., Toepfer, S., Kuhlmann, U., Gershenzon, J. & Turlings, T.C.J. (2005) Recruitment of entomopathogenic nematodes by insect-damaged maize roots. Nature, 434, 732737.
  • Sabelis, M.W. & van de Baan, H.E. (1983) Location of distant spider mite colonies by phytoseiid predators: demonstration of specific kairomones emitted by Tetranychus urticae and Panonychus ulmi. Entomologia Experimentalis et Applicata, 33, 303314.
  • Sabelis, M.W., van Baalen, M., Bakker, F.M., Bruin, J., Drukker, B., Egas, M., Janssen, A.R.M., Lesna, I.K., Pels, B., van Rijn, P.C.J. & Scutareanu, P. (1999) The evolution of direct and indirect plant defence against herbivorous arthropods. Herbivores. Between Plants and Predators (eds H. Olff, V.K. Brown & R.H. Drent), pp. 109166. Blackwell Science, Oxford.
  • Schausberger, P. (1997) Inter- and intraspecific predation on immatures by adult females in Euseius finlandicus, Typhlodromus pyri and Kampimodromus aberrans. Experimental and Applied Acarology, 21, 131150.
  • Schmelz, E.A., Alborn, H.T. & Tumlinson, J.H. (2003) Synergistic interactions between volicitin, jasmonic acid and ethylene mediate insect-induced volatile emission in Zea mays. Physiologia Plantarum, 117, 403412.
  • Smith, S.E. & Read, D.J. (2008) Mycorrhizal Symbiosis, 3rd edn. Elsevier, New York.
  • Soler, R., Harvey, J.A., Kamp, A.F.D., Vet, L.E.M., van der Putten, W.H., van Dam, N., Stuefer, J.F., Gols, R., Hordijk, C.A. & Bezemer, T.M. (2007) Root herbivores influence the behaviour of an aboveground parasitoid through changes in plant-volatile signals. Oikos, 116, 367376.
  • Thaler, J.S., Fidantsef, A.L., Duffey, S.S. & Bostock, R.M. (1999) Trade-offs in plant defense against pathogens and herbivores: a field demonstration of chemical elicitors of induced resistance. Journal of Chemical Ecology, 25, 15971609.
  • Thaler, J.S., Karban, R., Ullman, D.E., Boege, K. & Bostock, R.M. (2002) Cross-talk between jasmonate and salicylate plant defense pathways: effects on several plant parasites. Oecologia, 131, 227235.
  • Van der Putten, W.H., Vet, L.E.M., Harvey, J.A. & Wäckers, F. (2001) Linking above- and belowground multitrophic interactions of plants, herbivores, pathogens and their antagonists. Trends in Ecology and Evolution, 16, 547554.
  • Van Oosten, V.R., Bodenhausen, N., Reymond, P., van Pelt, J.A., van Loon, J.C., Dicke, M. & Pieterse, C.M.J. (2008) Differential effectiveness of microbially induced resistance against herbivorous insects in Arabidopsis thaliana. Molecular Plant-Microbe Interactions, 21, 919930.
  • Vierheilig, H., Coughlan, A.P., Wyss, U. & Piché, Y. (1998) Ink and vinegar, a simple staining technique for arbuscular-mycorrhizal fungi. Applied and Environmental Microbiology, 64, 50045007.
  • Walters, D. & Heil, M. (2007) Costs and trade-offs associated with induced resistance. Physiological and Molecular Plant Pathology, 71, 317.
  • Wenke, K., Kai, M. & Piechulla, B. (2010) Belowground volatiles facilitate interactions between plant roots and soil organisms. Planta, 231, 499506.
  • Williams, L. III, Rodriguez-Saona, C., Paré, P.W. & Crafts-Brandner, S.J. (2005) The piercing-sucking herbivores Lygus hesperus and Nezara viridula induce volatile emissions in plants. Archives of Insect Biochemistry and Physiology, 58, 8496.
  • Zhang, P.-J., Zheng, S.-J., van Loon, J.J.A., Boland, W., David, A., Mumm, R. & Dicke, M. (2009) Whiteflies interfere with indirect plant defense against spider mites in Lima bean. Proceedings of the National Academy of Sciences of the United States of America, 106, 2120221207.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Lay Summary

FilenameFormatSizeDescription
FEC_1947_sm_LaySummary.pdf40KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.