Oecologia DOI 10.1007/s00442-014-3129-x
PLANT-MICROBE-ANIMAL INTERACTIONS - ORIGINAL RESEARCH
Drought stress affects plant metabolites and herbivore preference but not host location by its parasitoids Berhane T. Weldegergis · Feng Zhu · Erik H. Poelman · Marcel Dicke
Received: 14 August 2014 / Accepted: 18 October 2014 © Springer-Verlag Berlin Heidelberg 2014
Abstract One of the main abiotic stresses that strongly affects plant survival and the primary cause of crop loss around the world is drought. Drought stress leads to sequential morphological, physiological, biochemical and molecular changes that can have severe effects on plant growth, development and productivity. As a consequence of these changes, the interaction between plants and insects can be altered. Using cultivated Brassica oleracea plants, the parasitoid Microplitis mediator and its herbivorous host Mamestra brassicae, we studied the effect of drought stress on (1) the emission of plant volatile organic compounds (VOCs), (2) plant hormone titres, (3) preference and performance of the herbivore, and (4) preference of the parasitoid. Higher levels of jasmonic acid (JA) and abscisic acid (ABA) were recorded in response to herbivory, but no significant differences were observed for salicylic acid (SA) and indole-3-acetic acid (IAA). Drought significantly impacted SA level and showed a significant interactive effect with herbivory for IAA levels. A total of 55 VOCs were recorded and the difference among the treatments was Communicated by Richard Karban. B. T. Weldegergis and F. Zhu made equal contributions to the article. B. T. Weldegergis (*) · F. Zhu · E. H. Poelman · M. Dicke Laboratory of Entomology, Wageningen University, PO Box 8031, 6700 EH Wageningen, The Netherlands e-mail: [email protected] F. Zhu e-mail: [email protected] E. H. Poelman e-mail: [email protected] M. Dicke e-mail: [email protected]
influenced largely by herbivory, where the emission rate of fatty acid-derived volatiles, nitriles and (E)-4,8-dimethylnona-1,3,7-triene [(E)-DMNT] was enhanced. Mamestra brassicae moths preferred to lay eggs on drought-stressed over control plants; their offspring performed similarly on plants of both treatments. VOCs due to drought did not affect the choice of M. mediator parasitoids. Overall, our study reveals an influence of drought on plant chemistry and insect-plant interactions. Keywords Cabbage · Herbivory · Parasitoids · Drought · Metabolites
Introduction In nature, plants are frequently experiencing simultaneous or sequential exposure to two or more (a)biotic stresses (Dicke et al. 2009; Holopainen and Gershenzon 2010; Stam et al. 2014). Importantly, because of climate change, plants have to deal more frequently with abiotic threats such as heat (Hasanuzzaman et al. 2013; Kotak et al. 2007), salt (Allakhverdiev et al. 2000) and drought (Shinozaki et al. 2003) stresses. The responses of plants to these abiotic stresses include physiological and molecular as well as chemical and biochemical changes (Copolovici et al. 2014; Hasanuzzaman et al. 2013; Tariq et al. 2013) that allow plants to maintain their growth, development and reproduction (Atkinson and Urwin 2012). Because plants are at the base of food webs, their responses to abiotic stress may alter their quality as a food plant and thus their interactions with organisms at higher trophic levels. For example, under drought stress, the level of constituents that determine the quality of the plant, such as plant nutrients, inorganic ions, amino acids, sugar-related
13
Oecologia
compounds and organic acids, may be accumulated in the plant tissues to limit negative osmotic effects. In response to drought stress, plants activate the expression of genes that regulate the formation of phytohormones (Iuchi et al. 2000; Liu and Baird 2003). These phytohormones are central to the regulation of biosynthetic pathways underlying growth and reproduction as well as defence or stressrelease mechanisms (Avanci et al. 2010; Kleczkowski and Schell 1995). The plant phenotypic changes in response to abiotic stress cascade into effects on growth, health and feeding behaviour of other interacting organisms (Copolovici et al. 2014; Gutbrodt et al. 2012; Mattson and Haack 1987). Drought stress has been reported to increase the populations of herbivorous insects on plant vegetation, thereby promoting insect outbreaks in natural ecosystems (Mattson and Haack 1987). This can be mediated by various mechanisms. During drought, plant temperature increases compared to well-watered plants due to stomata closure, while cooling due to transpiration is reduced (Mattson and Haack 1987). The increase in temperature changes the metabolite content of the drought-stressed plants and this might make plants more preferred for oviposition by herbivorous insects (Showler and Moran 2003). Host-plant selection by flying herbivores such as moths is based on successive behavioural processes including olfactory as well as visual and physical contacts (Schoonhoven et al. 2005). It has been demonstrated that plant volatile organic compounds (VOCs) play an important role in guiding moths in locating potential host plants (Kessler and Baldwin 2001; ReayJones et al. 2007; Renwick and Chew 1994; Wang et al. 2008). Acceptance of drought-stressed plants by herbivores may not correspond with the performance of the herbivore. Depending on their feeding guild, insects may perform differently on drought-stressed plants, and the effect can be positive, negative or neutral compared to their performance on plants with sufficient access to water (Huberty and Denno 2004). The shifts in plant resistance or susceptibility to herbivory may be explained by changes in plant chemical constituents under drought stress, where some plant primary metabolites such as sugar-related compounds may stimulate feeding, whereas secondary metabolites can act as both feeding attractant and/or deterrent (Gutbrodt et al. 2012). In their turn, the enemies of herbivores such as predators and parasitoids as well as their own enemies at the fourth trophic level use plant cues to locate their host or prey (Dicke and Baldwin 2010; Poelman et al. 2012). When plants are under attack by herbivores, blends of VOCs are released that provide natural enemies with information on the presence of their host (Dicke and Baldwin 2010; Dicke et al. 2009). The blend of VOCs emitted by plants due to biotic stress can be influenced when an abiotic stress is
13
added (Holopainen and Gershenzon 2010). In nature, plants not only face abiotic stresses but are also under constant threat of above- and below-ground organisms (Copolovici et al. 2014; Kessler and Baldwin 2001; Poelman et al. 2012). When multiple stresses appear simultaneously or sequentially, an overlap or/and crosstalk between signalling pathways may occur (Fraire-Velázquez et al. 2011; Mittler 2006; Stam et al. 2014). As a result, the VOCs emitted in response to double-stressed plants may have an effect on foraging behaviour of natural enemies of herbivores such as parasitoids compared to when there is only a single stress (de Rijk et al. 2013; Ponzio et al. 2013). For example, Tariq et al. (2013), have shown a decline in preference by Aphidius colemani—an aphid parasitoid—to plants exposed to double stresses: drought and root herbivory by Delia radicum. Moreover, plants under combined stress (drought and herbivory) emitted higher levels of monoterpenes compared to well-watered, herbivore-infested plants (Copolovici et al. 2014). As noted above, when a plant is exposed to drought stress, the level of primary metabolites may alter so as to protect the vital cellular part of the pant, which may increase or decrease the feeding behaviour of insects (Gutbrodt et al. 2012; Mattson and Haack 1987; Sicher and Barnaby 2012). Primary metabolites are important constituents in the formation of plant VOCs during their biosynthesis and, hence, the level of VOC emission may be dependent on the pools of these precursors of plant volatiles (Schwab et al. 2008). We studied how drought stress changes the interaction of plants with two members of their food web, using Brassica oleracea plants, the insect herbivore Mamestra brassicae and its parasitoid Microplitis mediator. We specifically addressed: (1) how drought affects plant physiology, (2) how herbivores respond to drought-stressed plants, and (3) how plant drought stress affects the host location abilities of the parasitoid enemies of this herbivore.
Materials and methods Plants and insects Five-week-old Brassica oleracea var. gemmifera L. cv. Cyrus plants (Brussels sprouts) were used in this study. The plants were grown in 1.45-l pots containing peat soil (Lentse potgrond, no. 4; Lent, The Netherlands) in a greenhouse (23° ± 2 °C, 70 % relative humidity, L16:D8). The cabbage moth Mamestra brassicae L. (Lepidoptera: Noctuidae) was reared on Brussels sprouts plants in a climate room (21 ± 1 °C, 50–70 % relative humidity, L16:D8). The parasitoid Microplitis mediator was reared on M. brassicae caterpillars (23 ± 2 °C, 70 % relative humidity, L16:D8) (Harvey et al. 2013).
Oecologia
Plant treatments All Brussels sprouts plants for the experiment were treated in a greenhouse (23 ± 2 °C, 70 % relative humidity, L16:D8). The plants were subjected to one of four treatments: (1) control: without drought or herbivory stress (C), (2) drought-stress only (D), (3) herbivore infestation only (H), (4) both drought-stress and herbivore infestation (HD). All plants were watered until saturation of the potting soil, allowing all treatments to start at the same moisture level. Half of the plants were provided with a regular watering regime (150 ml per day). The other half of the plants were deprived of further watering until they showed wilting symptoms, which was considered the first round of drought stress. Wilting symptoms were severe, and at that moment the drought-stressed plants were given 300 ml of water each in order to recover, and were left to wilt for a second round of drought stress. After inducing a third drought stress and full recovery, 15 first-instar caterpillars of M. brassicae were placed on both drought-stressed and well-watered plants and allowed to feed for 48 h. Drought stress in the natural environment is a long-term rather than a short-term process. Hence, we decided to induce successive drought stresses. For each treatment, 21 replicate plants were prepared and all C and H plants were watered on a daily basis at the same time of the day. Chemicals and reagents Ultra-grade solvents such as acetonitrile (>99.97 %), methanol (>99.95 %), isopropanol (>99.8 %), all GC-grade and water (ULC/MS grade) were purchased from Biosolve Chimie (Valkenswaard, The Netherlands). Formic acid (Ph Eur, 98–100 %), acetic acid (Ph Eur, 100 %) were purchased from Merck (Darmstadt, Germany). Commercialized standards and internal standards: Indole-3-acetic acid (IAA, >98 %), abscisic acid (ABA, >98 %), 2H6-ABA (>98 %), jasmonic acid (JA, >95 %) and 2H4-salicylic acid (2H4-SA), all HPLC grade were purchased from OlChemim (Olomouc, Czech Republic). In addition, other HPLC grade isotope labelled: 2H5-JA (97 %; CDN Isotopes, Quebec, Canada) and 13C6-IAA (Cambridge Isotope Laboratories, MA, USA) plus unlabelled: SA (>99 %; Sigma-Aldrich, St. Louis, MO, USA) standards were used. Sampling leaf tissue for phytohormone analysis To investigate the effect of drought stress on plant responses, we quantified phytohormone levels and volatile emission of plants that were subjected to the four treatments of drought and/or herbivory. We sampled tissue for phytohormone analysis from plants immediately after they had been used for volatile collection and
preference tests by parasitoids. The sample preparation was adopted with some modification from Muller and Munne-Bosch (2011), who describe optimum conditions for SA, IAA, JA and ABA extraction. Our study focused only on these four pre-selected phytohormones because they are associated with herbivory, droughtrelated plant stresses and/or growth (Pieterse et al. 2012). Three fully expanded leaves (second, fourth and sixth) from each plant, top (young leaf) to bottom (old leaf) were cut, wrapped in aluminium foil and immediately frozen in liquid nitrogen. The samples were ground to a fine powder using mortar and pestle, transferred into 15-ml Eppendorf tubes treated under liquid nitrogen and stored at −80 °C. Next, 150 mg of the powdered sample was weighed and transferred into a 2-ml Eppendorf tube and spiked immediately with 750 μl extracting mixed solvents—methanol:isopropanol:acetic acid at 20:79:1 ratio containing appropriate concentration (50 nmol final concentration) of isotope-labelled 2H4-SA, 13 C6-IAA, 2H5-JA and 2H6-ABA internal standards (I.S.). The sample was mixed using vortex (Genie™, NY, USA), sonicated (Branson Ultrasonics, Danbury, CT, USA) for 10 min, centrifuged (IEC Micromax, Buckinghamshire, UK) at 21,000g for 10 min and the supernatant was transferred into a clean 2-ml Eppendorf tube. Re-extraction was done by adding 750 μl of the solvent mix without I.S. and followed the same steps as the first extraction. The supernatants were combined, put to dryness in a SpeedVac (SAVANT SPD121P; Thermo Fisher Scientific, Waltham, USA) and re-dissolved in 300 μl of 10 % acetonitrile aqueous solution, mixed (vortex), sonicated for 5 min and centrifuged for 10 min. The sample was then filtered through 0.45 mm PTFE filter (Grace Alltech, Deerfield. IL, USA) into an HPLC vial (Grace Alltech) prior to analysis. We prepared 21 replicates for the H and D treatments as well as 20 replicates for C and HD treatment plants for phytohormone analysis. Analysis of phytohormones Analysis of phytohormones was performed using ultra performance liquid chromatography (UPLC) in an Aquity UPLC™ system with a quaternary pump furnished with an auto-sampler coupled to Xevo™ tandem quadrupole mass spectrometer (MS/MS) both from Waters (Milford, MA, USA). Separation of leaf extracts was performed using an Aquity UPLC™ BEH C18 (Waters) column with dimensions 2.1 × 100 mm, 2.7 μm operated at a temperature of 50 °C and a flow rate of 0.5 ml min−1. After injecting 10 μl of the sample volume, elution of analytes was carried out using a binary solvent system involving 0.1 % formic acid in water (solvent A) and 0.1 % formic acid in acetonitrile
13
Oecologia Table 1 Conditions applied for the analysis of the phytohormones a
Analyte
IAA 13
C6-IAAa SA 2 H4-SAa ABA 2 H6-ABAa JA 2
H5-JAa
Scan mode +
+ + + + + + +
Precursor ion (m/z)b 176
Product ions (m/z)c
Collision energy (V)d
Cone voltage (V)
182 139 143 265 271 211
103, 130 109, 136 65, 121 69, 97, 125 229, 247 235, 253 133, 151
30, 16 28, 16 24, 12 22, 18, 14 10, 6 10, 6 16, 12
18 18 18 20 10 10 16
216
135, 198
12, 10
14
a
Internal standards
b
Precursor ion (M + 1) of each analyte
c
Selected daughter ions for each analyte, underlined ions used for quantitation of each respective analyte d Collusion voltage to make the respective daughter ions in right
b
left to
(solvent B). A linear elution gradient with the following proportions (v/v) of solvent (t min, %A) was applied: (0, 95), (2, 70), (3, 50), (4, 10), and 6 min was used to equilibrate and bring solvent A to 95 %. MS/MS was performed using a Xevo™ tandem quadrupole mass spectrometer equipped with Turbo electrospray ionization source (Waters) and analyses of phytohormones was done in positive ion mode. Nitrogen was used as a disolvation, nebulizing and collision gas at flows of 1,000 and 50 l h−1, as well as 0.15 ml min−1, respectively. The following parameters were used for the source (ES+): temperature of 150 °C, disolvation temperature 650 °C and a capillary voltage of 3.0 kV. The cone voltage and collision energies (Table 1) used were dependent on the compound under investigation and were optimized using Waters IntelliStart MS Console. Ions used for characterizing and quantification were obtained through infusion experiments by direct injection of each standard solution of the phytohormone (labelled and non-labelled with isotope) into the mass spectrometer using the Waters IntelliStart MS Console. The mass spectrometer was operated in multiple reaction mode (MRM) due to its high selectivity using precursor-to-product ion transitions to avoid compounds that could present the same nominal molecular mass or peaks that could overlap. Using MRM conditions a specific precursor to two or more product ion transitions was monitored (Table 1). All phytohormone mass chromatograms from sample extracts showed identical retention times as the commercialized standards. Masslynx NT version 4.0 (Waters) software was used for instrument control, data acquisition and processing. A total of 21 replicates for H and D and 20 replicates for C and HD treatment plants
13
were analysed and the level of each phytohormone was quantified. Collection of plant volatiles Headspace collections of plant volatiles were carried out from 21 replicates of each C, D, H and HD treatment samples, while performing Y-tube olfactometer assays. Plant volatiles were collected for 2 h by sucking air out of the 30-l glass jars at 200 ml min−1 through a stainless steel cartridge filled with 200 mg Tenax TA (20/35 mesh; CAMSCO, Houston, TX, USA), while the rest of the air carrying the volatiles was led into the olfactometer for testing wasp behaviour (see below under ‘parasitoid preferences’). Periodically, volatiles from empty glass jars were collected and the compounds recorded, together with the volatiles originating from the Tenax TA adsorbent and the analytical instruments were excluded as artefacts from the data obtained for the plant samples as a correcting measure. Immediately after collection, the Tenax TA cartridges with sample volatiles were dry-purged under a stream of nitrogen (50 ml min−1) for 10 min at room temperature (21 ± 2 °C) to remove moisture and stored until analysis. Analysis of plant volatiles A Trace Ultra gas chromatograph (GC) coupled with Trace DSQ quadrupole mass spectrometer (MS) (both from Thermo Fisher Scientific) were used for separation and detection of plant volatiles. For details of the analytical conditions, refer to Poelman et al. (2012). We analysed 21 replicates for treatment samples—C, H and HD—plus 20 replicates for treatment D samples. Herbivore preference To study whether drought stress treatment of plants influences the oviposition preference of M. brassicae, we offered the moth a choice between plants with and without drought stress treatment (protocol as described above). After three rounds of drought stress, we detached three leaves (young, medium and old) of the plant, placed them in glass vials containing tap water, and matched them with leaves detached from similar positions of the control plants. The pairs of leaves were placed in a plastic cylinder (diameter 14 cm, height 22 cm) in which a male and a female moth were released. The females were allowed to oviposit overnight, and the numbers of egg clutches on each leaf were counted the next morning. In total, we tested 108 pairwise comparisons; in 41 out of these 108 tests no eggs were laid on leaves of either of the two treatments and these replicates were therefore excluded from data analysis.
Oecologia
Herbivore performance To test whether the drought stress treatment of the plant affects the growth of M. brassicae caterpillars, newly hatched first-instar (L1) larvae of the cabbage moth were inoculated on plants with and without drought treatment (protocol as described above). Fifteen plants were prepared for each treatment and each plant was inoculated with 10 newly hatched M. brassicae larvae. The caterpillars were allowed to feed on the plant for 7 consecutive days. After this incubation, all caterpillars that were recovered from the plants were weighed on a microbalance (accuracy = 1 μg; Sartorius, Göttingen, Germany). Parasitoid preference Shortly before M. mediator females were tested for their behavioural response to plant volatiles in a Y-tube olfactometer bio-assay (Takabayashi and Dicke 1992), caterpillars were removed from the plants. The plastic pots holding the plants were wrapped with aluminium foil to reduce the release of plastic- and soil-related volatiles and then the plants were placed in one of the two glass jars (30 l each). Pre-cleaned compressed air that had passed through charcoal (2.2 l min−1) was led through the glass jars that were connected to the two olfactometer arms, so that the two odour streams carrying plant volatiles entered the two arms. Mated female wasp were individually released in the Y-tube olfactometer and allowed to choose one of the two plant odours. The wasps that reached the end of one of the olfactometer arms within 10 min and stayed there for at least 10 s were considered making a final choice. Ten wasps were individually tested with each set of plants (in total, 7 pairs of each treatment combination were tested). The odour sources connected to each olfactometer arm were swapped after every five wasps in order to avoid any positional bias of the set-up. The Y-tube olfactometer setup was placed in a climatised room with artificial daylight and temperature of 21 ± 2 °C.
significance of the clusters obtained from PLS-DA and OPLS-DA analyses were evaluated using 2-tailed t test on the scores of the principal components (PCs). To evaluate the oviposition preference of cabbage moths for plants with or without drought stress, Wilcoxon matched-pair signedrank tests were used to test the differences in numbers of M. brassicae egg clutches on plants. To determine the performance of M. brassicae larvae on B. oleracea plants with or without drought stress, the differences in caterpillar fresh weight were analysed using Student’s t tests. Microplitis mediator preferences for plant volatiles, as tested in twochoice Y-tube olfactometer assays, were analysed using two-tailed binomial tests. All the statistical analyses were performed with the software package IBM SPSS Statistics 19 (SPSS, Chicago, IL, USA), except for the multivariate data analysis of plant volatiles, where SIMCA-P+ 12.0.1. statistical software (Umetrics, Umeå, Sweden) was used.
Results Phytohormones Plants that were subjected to drought stress, herbivory or the combination of drought and herbivory did not differ from control plants in fresh weight (one-way ANOVA, F = 0.259, df = 3, P = 0.855). Treatments that included herbivory induced higher levels of JA and ABA (oneway, ANOVA, Tukey–Kramer tests: JA: F = 12.042, df = 3, P
PLANT-MICROBE-ANIMAL INTERACTIONS - ORIGINAL RESEARCH
Drought stress affects plant metabolites and herbivore preference but not host location by its parasitoids Berhane T. Weldegergis · Feng Zhu · Erik H. Poelman · Marcel Dicke
Received: 14 August 2014 / Accepted: 18 October 2014 © Springer-Verlag Berlin Heidelberg 2014
Abstract One of the main abiotic stresses that strongly affects plant survival and the primary cause of crop loss around the world is drought. Drought stress leads to sequential morphological, physiological, biochemical and molecular changes that can have severe effects on plant growth, development and productivity. As a consequence of these changes, the interaction between plants and insects can be altered. Using cultivated Brassica oleracea plants, the parasitoid Microplitis mediator and its herbivorous host Mamestra brassicae, we studied the effect of drought stress on (1) the emission of plant volatile organic compounds (VOCs), (2) plant hormone titres, (3) preference and performance of the herbivore, and (4) preference of the parasitoid. Higher levels of jasmonic acid (JA) and abscisic acid (ABA) were recorded in response to herbivory, but no significant differences were observed for salicylic acid (SA) and indole-3-acetic acid (IAA). Drought significantly impacted SA level and showed a significant interactive effect with herbivory for IAA levels. A total of 55 VOCs were recorded and the difference among the treatments was Communicated by Richard Karban. B. T. Weldegergis and F. Zhu made equal contributions to the article. B. T. Weldegergis (*) · F. Zhu · E. H. Poelman · M. Dicke Laboratory of Entomology, Wageningen University, PO Box 8031, 6700 EH Wageningen, The Netherlands e-mail: [email protected] F. Zhu e-mail: [email protected] E. H. Poelman e-mail: [email protected] M. Dicke e-mail: [email protected]
influenced largely by herbivory, where the emission rate of fatty acid-derived volatiles, nitriles and (E)-4,8-dimethylnona-1,3,7-triene [(E)-DMNT] was enhanced. Mamestra brassicae moths preferred to lay eggs on drought-stressed over control plants; their offspring performed similarly on plants of both treatments. VOCs due to drought did not affect the choice of M. mediator parasitoids. Overall, our study reveals an influence of drought on plant chemistry and insect-plant interactions. Keywords Cabbage · Herbivory · Parasitoids · Drought · Metabolites
Introduction In nature, plants are frequently experiencing simultaneous or sequential exposure to two or more (a)biotic stresses (Dicke et al. 2009; Holopainen and Gershenzon 2010; Stam et al. 2014). Importantly, because of climate change, plants have to deal more frequently with abiotic threats such as heat (Hasanuzzaman et al. 2013; Kotak et al. 2007), salt (Allakhverdiev et al. 2000) and drought (Shinozaki et al. 2003) stresses. The responses of plants to these abiotic stresses include physiological and molecular as well as chemical and biochemical changes (Copolovici et al. 2014; Hasanuzzaman et al. 2013; Tariq et al. 2013) that allow plants to maintain their growth, development and reproduction (Atkinson and Urwin 2012). Because plants are at the base of food webs, their responses to abiotic stress may alter their quality as a food plant and thus their interactions with organisms at higher trophic levels. For example, under drought stress, the level of constituents that determine the quality of the plant, such as plant nutrients, inorganic ions, amino acids, sugar-related
13
Oecologia
compounds and organic acids, may be accumulated in the plant tissues to limit negative osmotic effects. In response to drought stress, plants activate the expression of genes that regulate the formation of phytohormones (Iuchi et al. 2000; Liu and Baird 2003). These phytohormones are central to the regulation of biosynthetic pathways underlying growth and reproduction as well as defence or stressrelease mechanisms (Avanci et al. 2010; Kleczkowski and Schell 1995). The plant phenotypic changes in response to abiotic stress cascade into effects on growth, health and feeding behaviour of other interacting organisms (Copolovici et al. 2014; Gutbrodt et al. 2012; Mattson and Haack 1987). Drought stress has been reported to increase the populations of herbivorous insects on plant vegetation, thereby promoting insect outbreaks in natural ecosystems (Mattson and Haack 1987). This can be mediated by various mechanisms. During drought, plant temperature increases compared to well-watered plants due to stomata closure, while cooling due to transpiration is reduced (Mattson and Haack 1987). The increase in temperature changes the metabolite content of the drought-stressed plants and this might make plants more preferred for oviposition by herbivorous insects (Showler and Moran 2003). Host-plant selection by flying herbivores such as moths is based on successive behavioural processes including olfactory as well as visual and physical contacts (Schoonhoven et al. 2005). It has been demonstrated that plant volatile organic compounds (VOCs) play an important role in guiding moths in locating potential host plants (Kessler and Baldwin 2001; ReayJones et al. 2007; Renwick and Chew 1994; Wang et al. 2008). Acceptance of drought-stressed plants by herbivores may not correspond with the performance of the herbivore. Depending on their feeding guild, insects may perform differently on drought-stressed plants, and the effect can be positive, negative or neutral compared to their performance on plants with sufficient access to water (Huberty and Denno 2004). The shifts in plant resistance or susceptibility to herbivory may be explained by changes in plant chemical constituents under drought stress, where some plant primary metabolites such as sugar-related compounds may stimulate feeding, whereas secondary metabolites can act as both feeding attractant and/or deterrent (Gutbrodt et al. 2012). In their turn, the enemies of herbivores such as predators and parasitoids as well as their own enemies at the fourth trophic level use plant cues to locate their host or prey (Dicke and Baldwin 2010; Poelman et al. 2012). When plants are under attack by herbivores, blends of VOCs are released that provide natural enemies with information on the presence of their host (Dicke and Baldwin 2010; Dicke et al. 2009). The blend of VOCs emitted by plants due to biotic stress can be influenced when an abiotic stress is
13
added (Holopainen and Gershenzon 2010). In nature, plants not only face abiotic stresses but are also under constant threat of above- and below-ground organisms (Copolovici et al. 2014; Kessler and Baldwin 2001; Poelman et al. 2012). When multiple stresses appear simultaneously or sequentially, an overlap or/and crosstalk between signalling pathways may occur (Fraire-Velázquez et al. 2011; Mittler 2006; Stam et al. 2014). As a result, the VOCs emitted in response to double-stressed plants may have an effect on foraging behaviour of natural enemies of herbivores such as parasitoids compared to when there is only a single stress (de Rijk et al. 2013; Ponzio et al. 2013). For example, Tariq et al. (2013), have shown a decline in preference by Aphidius colemani—an aphid parasitoid—to plants exposed to double stresses: drought and root herbivory by Delia radicum. Moreover, plants under combined stress (drought and herbivory) emitted higher levels of monoterpenes compared to well-watered, herbivore-infested plants (Copolovici et al. 2014). As noted above, when a plant is exposed to drought stress, the level of primary metabolites may alter so as to protect the vital cellular part of the pant, which may increase or decrease the feeding behaviour of insects (Gutbrodt et al. 2012; Mattson and Haack 1987; Sicher and Barnaby 2012). Primary metabolites are important constituents in the formation of plant VOCs during their biosynthesis and, hence, the level of VOC emission may be dependent on the pools of these precursors of plant volatiles (Schwab et al. 2008). We studied how drought stress changes the interaction of plants with two members of their food web, using Brassica oleracea plants, the insect herbivore Mamestra brassicae and its parasitoid Microplitis mediator. We specifically addressed: (1) how drought affects plant physiology, (2) how herbivores respond to drought-stressed plants, and (3) how plant drought stress affects the host location abilities of the parasitoid enemies of this herbivore.
Materials and methods Plants and insects Five-week-old Brassica oleracea var. gemmifera L. cv. Cyrus plants (Brussels sprouts) were used in this study. The plants were grown in 1.45-l pots containing peat soil (Lentse potgrond, no. 4; Lent, The Netherlands) in a greenhouse (23° ± 2 °C, 70 % relative humidity, L16:D8). The cabbage moth Mamestra brassicae L. (Lepidoptera: Noctuidae) was reared on Brussels sprouts plants in a climate room (21 ± 1 °C, 50–70 % relative humidity, L16:D8). The parasitoid Microplitis mediator was reared on M. brassicae caterpillars (23 ± 2 °C, 70 % relative humidity, L16:D8) (Harvey et al. 2013).
Oecologia
Plant treatments All Brussels sprouts plants for the experiment were treated in a greenhouse (23 ± 2 °C, 70 % relative humidity, L16:D8). The plants were subjected to one of four treatments: (1) control: without drought or herbivory stress (C), (2) drought-stress only (D), (3) herbivore infestation only (H), (4) both drought-stress and herbivore infestation (HD). All plants were watered until saturation of the potting soil, allowing all treatments to start at the same moisture level. Half of the plants were provided with a regular watering regime (150 ml per day). The other half of the plants were deprived of further watering until they showed wilting symptoms, which was considered the first round of drought stress. Wilting symptoms were severe, and at that moment the drought-stressed plants were given 300 ml of water each in order to recover, and were left to wilt for a second round of drought stress. After inducing a third drought stress and full recovery, 15 first-instar caterpillars of M. brassicae were placed on both drought-stressed and well-watered plants and allowed to feed for 48 h. Drought stress in the natural environment is a long-term rather than a short-term process. Hence, we decided to induce successive drought stresses. For each treatment, 21 replicate plants were prepared and all C and H plants were watered on a daily basis at the same time of the day. Chemicals and reagents Ultra-grade solvents such as acetonitrile (>99.97 %), methanol (>99.95 %), isopropanol (>99.8 %), all GC-grade and water (ULC/MS grade) were purchased from Biosolve Chimie (Valkenswaard, The Netherlands). Formic acid (Ph Eur, 98–100 %), acetic acid (Ph Eur, 100 %) were purchased from Merck (Darmstadt, Germany). Commercialized standards and internal standards: Indole-3-acetic acid (IAA, >98 %), abscisic acid (ABA, >98 %), 2H6-ABA (>98 %), jasmonic acid (JA, >95 %) and 2H4-salicylic acid (2H4-SA), all HPLC grade were purchased from OlChemim (Olomouc, Czech Republic). In addition, other HPLC grade isotope labelled: 2H5-JA (97 %; CDN Isotopes, Quebec, Canada) and 13C6-IAA (Cambridge Isotope Laboratories, MA, USA) plus unlabelled: SA (>99 %; Sigma-Aldrich, St. Louis, MO, USA) standards were used. Sampling leaf tissue for phytohormone analysis To investigate the effect of drought stress on plant responses, we quantified phytohormone levels and volatile emission of plants that were subjected to the four treatments of drought and/or herbivory. We sampled tissue for phytohormone analysis from plants immediately after they had been used for volatile collection and
preference tests by parasitoids. The sample preparation was adopted with some modification from Muller and Munne-Bosch (2011), who describe optimum conditions for SA, IAA, JA and ABA extraction. Our study focused only on these four pre-selected phytohormones because they are associated with herbivory, droughtrelated plant stresses and/or growth (Pieterse et al. 2012). Three fully expanded leaves (second, fourth and sixth) from each plant, top (young leaf) to bottom (old leaf) were cut, wrapped in aluminium foil and immediately frozen in liquid nitrogen. The samples were ground to a fine powder using mortar and pestle, transferred into 15-ml Eppendorf tubes treated under liquid nitrogen and stored at −80 °C. Next, 150 mg of the powdered sample was weighed and transferred into a 2-ml Eppendorf tube and spiked immediately with 750 μl extracting mixed solvents—methanol:isopropanol:acetic acid at 20:79:1 ratio containing appropriate concentration (50 nmol final concentration) of isotope-labelled 2H4-SA, 13 C6-IAA, 2H5-JA and 2H6-ABA internal standards (I.S.). The sample was mixed using vortex (Genie™, NY, USA), sonicated (Branson Ultrasonics, Danbury, CT, USA) for 10 min, centrifuged (IEC Micromax, Buckinghamshire, UK) at 21,000g for 10 min and the supernatant was transferred into a clean 2-ml Eppendorf tube. Re-extraction was done by adding 750 μl of the solvent mix without I.S. and followed the same steps as the first extraction. The supernatants were combined, put to dryness in a SpeedVac (SAVANT SPD121P; Thermo Fisher Scientific, Waltham, USA) and re-dissolved in 300 μl of 10 % acetonitrile aqueous solution, mixed (vortex), sonicated for 5 min and centrifuged for 10 min. The sample was then filtered through 0.45 mm PTFE filter (Grace Alltech, Deerfield. IL, USA) into an HPLC vial (Grace Alltech) prior to analysis. We prepared 21 replicates for the H and D treatments as well as 20 replicates for C and HD treatment plants for phytohormone analysis. Analysis of phytohormones Analysis of phytohormones was performed using ultra performance liquid chromatography (UPLC) in an Aquity UPLC™ system with a quaternary pump furnished with an auto-sampler coupled to Xevo™ tandem quadrupole mass spectrometer (MS/MS) both from Waters (Milford, MA, USA). Separation of leaf extracts was performed using an Aquity UPLC™ BEH C18 (Waters) column with dimensions 2.1 × 100 mm, 2.7 μm operated at a temperature of 50 °C and a flow rate of 0.5 ml min−1. After injecting 10 μl of the sample volume, elution of analytes was carried out using a binary solvent system involving 0.1 % formic acid in water (solvent A) and 0.1 % formic acid in acetonitrile
13
Oecologia Table 1 Conditions applied for the analysis of the phytohormones a
Analyte
IAA 13
C6-IAAa SA 2 H4-SAa ABA 2 H6-ABAa JA 2
H5-JAa
Scan mode +
+ + + + + + +
Precursor ion (m/z)b 176
Product ions (m/z)c
Collision energy (V)d
Cone voltage (V)
182 139 143 265 271 211
103, 130 109, 136 65, 121 69, 97, 125 229, 247 235, 253 133, 151
30, 16 28, 16 24, 12 22, 18, 14 10, 6 10, 6 16, 12
18 18 18 20 10 10 16
216
135, 198
12, 10
14
a
Internal standards
b
Precursor ion (M + 1) of each analyte
c
Selected daughter ions for each analyte, underlined ions used for quantitation of each respective analyte d Collusion voltage to make the respective daughter ions in right
b
left to
(solvent B). A linear elution gradient with the following proportions (v/v) of solvent (t min, %A) was applied: (0, 95), (2, 70), (3, 50), (4, 10), and 6 min was used to equilibrate and bring solvent A to 95 %. MS/MS was performed using a Xevo™ tandem quadrupole mass spectrometer equipped with Turbo electrospray ionization source (Waters) and analyses of phytohormones was done in positive ion mode. Nitrogen was used as a disolvation, nebulizing and collision gas at flows of 1,000 and 50 l h−1, as well as 0.15 ml min−1, respectively. The following parameters were used for the source (ES+): temperature of 150 °C, disolvation temperature 650 °C and a capillary voltage of 3.0 kV. The cone voltage and collision energies (Table 1) used were dependent on the compound under investigation and were optimized using Waters IntelliStart MS Console. Ions used for characterizing and quantification were obtained through infusion experiments by direct injection of each standard solution of the phytohormone (labelled and non-labelled with isotope) into the mass spectrometer using the Waters IntelliStart MS Console. The mass spectrometer was operated in multiple reaction mode (MRM) due to its high selectivity using precursor-to-product ion transitions to avoid compounds that could present the same nominal molecular mass or peaks that could overlap. Using MRM conditions a specific precursor to two or more product ion transitions was monitored (Table 1). All phytohormone mass chromatograms from sample extracts showed identical retention times as the commercialized standards. Masslynx NT version 4.0 (Waters) software was used for instrument control, data acquisition and processing. A total of 21 replicates for H and D and 20 replicates for C and HD treatment plants
13
were analysed and the level of each phytohormone was quantified. Collection of plant volatiles Headspace collections of plant volatiles were carried out from 21 replicates of each C, D, H and HD treatment samples, while performing Y-tube olfactometer assays. Plant volatiles were collected for 2 h by sucking air out of the 30-l glass jars at 200 ml min−1 through a stainless steel cartridge filled with 200 mg Tenax TA (20/35 mesh; CAMSCO, Houston, TX, USA), while the rest of the air carrying the volatiles was led into the olfactometer for testing wasp behaviour (see below under ‘parasitoid preferences’). Periodically, volatiles from empty glass jars were collected and the compounds recorded, together with the volatiles originating from the Tenax TA adsorbent and the analytical instruments were excluded as artefacts from the data obtained for the plant samples as a correcting measure. Immediately after collection, the Tenax TA cartridges with sample volatiles were dry-purged under a stream of nitrogen (50 ml min−1) for 10 min at room temperature (21 ± 2 °C) to remove moisture and stored until analysis. Analysis of plant volatiles A Trace Ultra gas chromatograph (GC) coupled with Trace DSQ quadrupole mass spectrometer (MS) (both from Thermo Fisher Scientific) were used for separation and detection of plant volatiles. For details of the analytical conditions, refer to Poelman et al. (2012). We analysed 21 replicates for treatment samples—C, H and HD—plus 20 replicates for treatment D samples. Herbivore preference To study whether drought stress treatment of plants influences the oviposition preference of M. brassicae, we offered the moth a choice between plants with and without drought stress treatment (protocol as described above). After three rounds of drought stress, we detached three leaves (young, medium and old) of the plant, placed them in glass vials containing tap water, and matched them with leaves detached from similar positions of the control plants. The pairs of leaves were placed in a plastic cylinder (diameter 14 cm, height 22 cm) in which a male and a female moth were released. The females were allowed to oviposit overnight, and the numbers of egg clutches on each leaf were counted the next morning. In total, we tested 108 pairwise comparisons; in 41 out of these 108 tests no eggs were laid on leaves of either of the two treatments and these replicates were therefore excluded from data analysis.
Oecologia
Herbivore performance To test whether the drought stress treatment of the plant affects the growth of M. brassicae caterpillars, newly hatched first-instar (L1) larvae of the cabbage moth were inoculated on plants with and without drought treatment (protocol as described above). Fifteen plants were prepared for each treatment and each plant was inoculated with 10 newly hatched M. brassicae larvae. The caterpillars were allowed to feed on the plant for 7 consecutive days. After this incubation, all caterpillars that were recovered from the plants were weighed on a microbalance (accuracy = 1 μg; Sartorius, Göttingen, Germany). Parasitoid preference Shortly before M. mediator females were tested for their behavioural response to plant volatiles in a Y-tube olfactometer bio-assay (Takabayashi and Dicke 1992), caterpillars were removed from the plants. The plastic pots holding the plants were wrapped with aluminium foil to reduce the release of plastic- and soil-related volatiles and then the plants were placed in one of the two glass jars (30 l each). Pre-cleaned compressed air that had passed through charcoal (2.2 l min−1) was led through the glass jars that were connected to the two olfactometer arms, so that the two odour streams carrying plant volatiles entered the two arms. Mated female wasp were individually released in the Y-tube olfactometer and allowed to choose one of the two plant odours. The wasps that reached the end of one of the olfactometer arms within 10 min and stayed there for at least 10 s were considered making a final choice. Ten wasps were individually tested with each set of plants (in total, 7 pairs of each treatment combination were tested). The odour sources connected to each olfactometer arm were swapped after every five wasps in order to avoid any positional bias of the set-up. The Y-tube olfactometer setup was placed in a climatised room with artificial daylight and temperature of 21 ± 2 °C.
significance of the clusters obtained from PLS-DA and OPLS-DA analyses were evaluated using 2-tailed t test on the scores of the principal components (PCs). To evaluate the oviposition preference of cabbage moths for plants with or without drought stress, Wilcoxon matched-pair signedrank tests were used to test the differences in numbers of M. brassicae egg clutches on plants. To determine the performance of M. brassicae larvae on B. oleracea plants with or without drought stress, the differences in caterpillar fresh weight were analysed using Student’s t tests. Microplitis mediator preferences for plant volatiles, as tested in twochoice Y-tube olfactometer assays, were analysed using two-tailed binomial tests. All the statistical analyses were performed with the software package IBM SPSS Statistics 19 (SPSS, Chicago, IL, USA), except for the multivariate data analysis of plant volatiles, where SIMCA-P+ 12.0.1. statistical software (Umetrics, Umeå, Sweden) was used.
Results Phytohormones Plants that were subjected to drought stress, herbivory or the combination of drought and herbivory did not differ from control plants in fresh weight (one-way ANOVA, F = 0.259, df = 3, P = 0.855). Treatments that included herbivory induced higher levels of JA and ABA (oneway, ANOVA, Tukey–Kramer tests: JA: F = 12.042, df = 3, P
