Insect-Associated Bacteria Boost VOC Emissions, Including Benzenoids, Esters, Hydrocarbons, and Lactone, during Cereal Leaf Beetle Feeding on Wheat Authors: Beata Wielkopolan, Patryk Frąckowiak, Małgorzata Majcher, Aleksandra Obrępalska-Stęplowska Published: 2025 ABSTRACT AbstractPlant volatile organic compounds (VOCs) constitute part of the signal transmitted between plants, serving as defence molecules that repel plant pests or attract their natural enemies. Insect-associated bacteria are important factors influencing many aspects of plant-insect interactions. In this study, we have undertaken to evaluate the impact of the insect-associated bacteria on the abundance and the content of VOCs emitted from wheat plants damaged by important cereal pest – cereal leaf beetle (CLB, Oulema melanopus), in both larval and adult stages. We assumed that CLB-associated bacteria, as an additional factor, may cause different and higher VOC emissions from plants attacked by insects. VOCs were collected by solid-phase microextraction and analyzed with comprehensive two-dimensional gas chromatography-mass spectrometry. We revealed that CLB-associated bacteria contribute to both the increased number and total peak area of VOCs emitted by plants exposed to CLB, particularly those classified as hydrocarbons, benzenoids, esters, and lactones. This effect is especially pronounced in plants damaged by larvae. Our findings demonstrate that physical damage alone may be sufficient to trigger the release of certain volatiles from plant tissues, but CLB feeding in the presence of its associated bacteria leads to a higher quantity and greater diversity of VOCs emissions.Supplementary InformationThe online version contains supplementary material available at 10.1007/s10886-025-01660-9.Keywords: Cereal leaf beetle, Volatile organic compounds, Benzenoids, Bacteria, Wheat Introduction Wheat is the primary food for billions of people worldwide, providing the main source of energy and fiber for humans (Qiao et al. 2019). Hence reducing losses of this crop caused by herbivores is a priority. Among the pests causing high losses in wheat production, cereal leaf beetle (CLB, Oulema melanopus, Chrysomelidae, Coleoptera) is an important one. The CLB feeds on a wide range of cereals. The larvae are considered the main harmful stage and are the target of insecticide treatments. Their feeding, especially on the flag and sub-flag leaves, causes a decrease in the quality and quantity of the yield. The latter can be reduced by 30–50% if chemical treatment is not applied against this pest (Evans 2023). CLB population is mainly controlled by chemical protection (insecticides), thus new, more environmentally friendly methods are being sought. Plants synthesize numerous secondary metabolites, among them volatile organic compounds (VOCs), which are gaining interest, because of their potential use in improving plant protection against herbivorous pests (Junker et al. 2018; Murali-Baskaran et al. 2022). It was reported that plant VOCs can discourage herbivores from feeding and oviposition (Unsicker et al. 2009), attract natural enemies (predators, parasitoids) which contribute to reducing the population of the insect pest (Clavijo McCormick et al. 2012), and transmit information about the attack to distant parts of the same plant and neighboring plants (Kalske et al. 2019). A wide variety of VOCs are released by plants constitutively or being induced in response to biotic and abiotic stresses. Some compounds can be released from plants immediately after damage, including green leaf volatiles (GLVs), or within a few days. The quantity and quality of VOCs varies among others by plant species, genotype, and developmental stage (Maffei 2010; Piesik et al. 2011; Grunseich et al. 2020). Many factors influence insect behavior, including gustatory, tactile, visual cues, as well as volatile chemical signals received by insects through complex chemosensory systems. Many insects are particularly sensitive to odors that indicate the availability of food resources, the presence of competitors, predators or potential mates (Price et al. 2011; Davis et al. 2013). Although the influence of odor on insects behavior has long been known, relatively little attention has been paid to how microbes associated with insects can influence the emission of VOCs. The saliva and oral secretion of insects applied to damaged plant tissue during insect feeding may contain microbes, that interfere with plant responses to insects (Chung et al. 2013; Wang et al. 2016) and shape interactions between plants and insects (Wielkopolan et al. 2022). Previous research indicated, that microbes can play important roles in modifying plant traits that indirectly influence host-plant selection by insect herbivores (Biere and Bennett 2013; Wielkopolan and Obrępalska-Stęplowska 2016; Hassani et al. 2018; Grunseich et al. 2020). For instance, microbial pathogens can alter the production of plant VOCs to attract insect vectors to increase their chances of spreading (Tasin et al. 2012; Mauck et al. 2016; Eigenbrode et al. 2018; Grunseich et al. 2020; Wrzesińska-Krupa and Obrępalska-Stęplowska 2025). There are numerous examples where microbial (microbes considered as bacteria and fungi) volatile organic compounds (MVOCs) play a role in the interaction between insects and plants, including attracting and repelling insects, mimicking plant hormones, or inducing plant resistance (Davis et al. 2011, 2013; Piesik et al. 2015, 2020; Wielkopolan and Obrępalska-Stęplowska 2016). For example, yeast (Kuraishia capsulata Wickerham and Pichia pini (Holst) Phaff) – associated with Dendroctonus ponderosae Hopkins (Curculionidae, Coleoptera) are able to produce anti-aggregation pheromone verbenone (Hunt and Borden 1990; Davis et al. 2013). In turn, Sanchez-Cruz et al. (2020) indicated that bacteria volatiles isolated from genital chambers of Cyclocephala lunulata Burmsister and Cyclocephala barrerai Martinez (Scarabaeidae, Coleoptera) act as aggregation attractants. VOCs function as aggregation attractants may be used in mass trapping for management of herbivorous insects. Semiochemicals (chemicals produced by one organism that elicit a behavioral response in the recipient organisms) are important for integrated pest management, because they have a wide range of applications, as sex pheromones, aggregation pheromones or plant VOCs using as attractants or repellents (Gaffke et al. 2021; Sendel et al. 2022). The main objective of the study was to evaluate the VOCs released by wheat plants in the first four hours of CLB larvae and adults feeding and to assess whether CLB-associated bacteria affect the blend of VOCs emitted by wheat plants. The plants were therefore exposed to the feeding of CLB adults, larvae with a natural microbiome, as well as bacteria-depleted insects. We assumed that the CLB-associated bacteria could modulate the emission of VOCs by wheat plants during CLB feeding. We hypothesized firstly, that in wheat plants that are exposed to more than one factor, i.e. CLB and CLB-associated bacteria, the plant response in the form of VOCs emission will also be stronger. Secondly, differences in composition, and bacterial biodiversity between larval and adult stages (which was found to be higher in the larval stage of CLB) (Wielkopolan et al. 2021) will be reflected in the composition of VOCs. Thirdly, the blend of VOCs emissions caused by feeding by the larvae and adults will differ due to the different manner of wheat leaf damage caused by these two insect developmental stages. Methods and Materials Insect Material CLB larvae (3rd instar) and adults were collected from a field of winter wheat in Słupia Wielka (52°13′02″N 17°13′04″E) in 2022. All collected insects were starved for 24 h. To obtain bacteria-depleted larvae and adults some of the collected insects were feed for 24 h on wheat leaves coated with a mixture of antibiotics (AB+) according to Chung et al. (2013) with some modifications. Three antibiotics (0.01 g of neomycin sulfate, 0.05 g of chlortetracycline and 0.003 g of streptomycin) (Merck, Poznań, Poland) were suspended together in 50 mL of ultrapure Milli-Q water. The control group consisted of insects with a natural microbiome, that were fed on leaves not treated with AB (AB-). Wheat leaves on which insects fed were previously disinfected in 70% ethanol, rinsed twice in sterile Milli-Q water, and placed on a 1% agarose medium (Petri dishes) to retain moisture. Plant Material Plants of winter wheat (Triticum aestivum L.) variety Arkadia (Danko Hodowla Roślin Sp. z o.o z/s in Choryń) were maintained at the Research Centre of Quarantine, Invasive and Genetically Modified Organisms, at the Institute of Plant Protection-National Research Institute, in Poznań, in 2022. In each pot, one seed was sown. The plants were kept at a temperature of 18 °C during the day and 16 °C at night, with 35% humidity, without additional lighting (daylight only). After less than 7 weeks of cultivation, by which time the wheat plants were in their 31 BBCH (Biologische Bundesanstalt, Bundessortenamt und Chemische Industrie) growth stage, they were damaged mechanically or by CLB feeding. Insects obtained according to the procedure described in the section “Insect material” were applied to wheat plants. Two larvae or adults with a natural (L(AB-) and A(AB-), respectively) or depleted (L(AB+) and A(AB+), respectively) bacterial microbiome were applied to wheat plants, per leaf. Mechanical damage to leaves corresponding to CLB feeding was done with a sterile needle, scraping tissue into the lower cuticle of the leaf to mimic larval feeding (md/L) and making a hole throughout the thickness of the leaf to mimic the damage of an adult specimen (md/A). Two mechanical injuries were made on 1 leaf. The control (C) group consisted of undamaged plants. Each plant treatment has been performed with 3 repetitions (a total of 21 plants for 7 treatments). Collection of Volatile Compounds Each plant was placed individually in a glass chamber under the same temperature and photoperiod conditions. Immediately after applying insects to plants or mechanical damage, the collection of VOCs began for a period of 4 h. Volatiles were extracted by solid-phase microextraction with a 2 cm carboxene/PDMS/DVB fiber (Supelco). All fibers were conditioned prior to analysis as recommended by the producer. For each plant treatment, SPME fibers were exposed in the headspace of olfactory chambers for 4 h, during this time insects were left on the plant. To serve as blanks, SPME fibers were also collected from empty chambers. After extraction, all fibers were sealed and immediately transported for gas chromatography analysis. Analysis of Volatile Compounds by GCxGC-ToFMS GCxGC-ToFMS analysis was performed on an Agilent 6890 N gas chromatograph equipped with a cryogenic (N2) modulator coupled to a PEGASUS 4D time of flight mass spectrometer (LECO, St. Joseph, MI). SPME injections were performed manually in a splitless mode. Compounds were resolved on a nonpolar polar column system: SLB-5 (30 m x 0.25 mm x 0.25 μm) as a first dimension column and SPB-50 (1 m x 0.25 mm x 0.25 μm) as a second dimension column. The carrier gas was helium at a constant flow rate of 0.8 ml/min. The operating conditions for the first column were the following: initial oven temperature 45 °C (1 min), 6 °C/min to 220 °C and 25 °C/min to 240 °C (5 min). The second column was programmed at the same rates as the first one, but with a 30 °C offset (i.e. initial temperature was 75 °C). The injector temperature was 250 °C, transfer line temp was 270 °C. The modulation time was 4 s. The ToF spectrometer was operating at a range of 33–383 m/z and at 150 full spectra/sec speed. Data were processed by ChromaTOF v.4.50 supplied with the instrument. The quantities of the major components were estimated based on the peak areas of the compounds. Volatiles were identified by a comparison of mass spectra and retention indices (RI) with the NIST 2.0 (The National Institute of Standards and Technology) library (Gaithersburg, MD, USA). A mixture of n-alkanes (C7-C20) dissolved in hexane was supplied by Supelco (Bellefonte, PA, USA) for retention index determination. The data was pre-processed by LECO Chroma TOF (4.50.8.0, Leco Co., St. Joseph, MI). For the analyte match criteria, the following parameters were used: signal-to-noise (S/N) = 100, minimum similarity spectral match = 800, mass threshold = 10. Statistical Analysis All statistical analyses were conducted using the R programming environment (version 4.1.3) (Team 2022). Principal Component Analysis (PCA) has been conducted using ’factoextra’ (Kassambara and Mundt 2020) and ‘corrplot’ (Wei and Simko 2021). The 3D plot of PCA results has been prepared using ‘ggplot2’(Wickham 2008) and ‘plotly’(Inc 2015) packages. Statistical differences between experimental groups of plants were analyzed using the Kruskal-Wallis test with a p-value cut-off set at p < 0.05. To determine the specific differences, a post-hoc Dunn’s test corrected with the Holm method was conducted. The analysis of differences in the plant treatments, graphs, and Spearman’s correlation analysis was performed using the ‘ggstatsplot’ (Patil 2021) package in R. The heatmap has been created using the ‘ComplexHeatmap’(Gu et al. 2016) package. The experimental conditions were grouped by hierarchical clustering (Euclidean distance; ‘ComplexHeatmap’ package defaults settings), and the compounds were partitioned with k‑means clustering. The bubble plot representing both the quantity and sum of the peak area of emitted VOCs, categorized based on compound groups, has been conducted using the ‘ggplo2’ package. Comparison of unique and shared VOCs for plants exposed to insect herbivory with a natural and depleted bacterial microbiome, as well as undamaged controls, was conducted using an online Venn diagram generator program (https://bioinformatics.psb.ugent.be/webtools/Venn/). Results The chromatographic analysis of VOCs allowed for the identification of 498 VOCs from all tested plant treatments (Supplementary Material S1). The hundred twenty five statistically significant VOCs out of 498 (with a Fisher ratio higher than 0) were used for statistical analysis and are shown in the Supplementary Material S2. Bacteria Associated with CLB Increase the Emission of VOCs from Wheat Plants during CLB Feeding In the first stage of the study, we evaluated whether the presence of the CLB-associated bacteria, in addition to the damage caused by the insect itself, increases the emission of volatiles from plants. Indeed, for wheat plants exposed to CLB (both in the larval and adult stages with a natural microbiome, [L(AB-)) and A(AB-)), respectively], the highest sum of peak areas and the highest number of VOCs were recorded, in comparison to plants damaged by bacteria-depleted larvae (L(AB+)) and bacteria-depleted adults (A(AB+)), as well as to other plant treatments (Fig. 1a). The total sum of peak area was 8.23 times higher in L(AB-) than in L(AB+) and 4.11 times higher in A(AB-) than in A(AB+). The differences in the amount of emitted VOCs between plants damaged by insects with a natural microbiome and the other tested plant treatments were statistically significant (Fig. 1b). The composition of VOCs emitted by plants infested by insects (especially for the larval stage) with a natural microbiome differed from those emitted by plants damaged by bacteria-depleted insects, as evidenced by clustering of treatments in the 3D PCA analysis (Fig. 1c) and the low level of correlation (r coefficient of 0.61 for larval stage and 0.80 for adult stage) (Fig. 1d). Thus, these results confirmed our hypothesis that the wheat plants that were exposed to more than one factor, i.e. CLB and CLB-associated bacteria, the defense response in the form of VOCs emissions was stronger. (a) The total sum of volatile organic compounds (VOCs) peak areas (left scale) and number of identified VOCs (number above each plot) after considering the Fischer test for each wheat plant treatment. (b) Violin and box plots of the distribution of VOCs in studied plant treatments. Results are given on a log2 scale. The result of the Kruskal-Wallis test was presented in the top-left corner. Different letters above the charts indicate the presence of statistical significance (p < 0.05) between the studied plant treatments, calculated using the Dunn’s test adjusted with the Holm method using the “ggstatsplot” package in R. (c) The 3D Principal component analysis (PCA) (indicated results for dimension 1, dimension 2 and dimension 3) of plant treatment variants. (d) Spearman’s correlation coefficient between VOCs in each treatment condition. C-control (undamaged plants); plants damaged by CLB larvae (L(AB-), adults (A(AB-)) with a natural microbiome; bacteria-depleted larvae (L(AB+)) and adults (A(AB+)); mechanically damaged wheat plants imitating the feeding of CLB larvae (md/L) and adults (md/A) Since the larval microbiome was found to be more diverse and composed of more bacterial genera than that of adult CLB (Wielkopolan et al. 2021), we decided to check if this difference in abundance and variability of CLB-associated bacteria was reflected in the quantity and quality of emitted VOCs from plants damaged by an insect in larval and adult stages. Based on the comparison of data for plants damaged by CLB larvae and damaged by adults with a natural microbiome [L(AB-) vs. A(AB-)], we showed that plants damaged by larvae with a natural microbiome emitted a higher total sum of peak area of VOCs (which was 7.08E + 08), in comparison to plants exposed to adult insects with a natural microbiome (5.86E + 08) (Fig. 1a), but the difference in the amount of emitted VOCs was not statistically significant (Fig. 1b). The composition of VOCs emitted by plants damaged by larvae and adults with a natural microbiome was basically comparable, as demonstrated by their clustering in the three-dimensional PCA analysis (Fig. 1c) and a high level of correlation (with an r coefficient of 0.95) (Fig. 1d). CLB-Associated Bacteria Contribute to the Increased Emission of VOCs from the Groups of Benzenoids, Esters, Hydrocarbons, and Lactones After demonstrating that CLB-associated bacteria and their composition affect VOC emissions from plants attacked by this pest, we asked which chemical groups of VOCs were emitted in the greatest amount in the presence of CLB-associated bacteria. We found out that one hundred twenty five statistically significant VOCs (Supplementary Material S2) from all studied groups of wheat plants were assigned to 16 chemical groups: acids (2), alcohols (11), aldehydes (7), aromatic compounds (7), aromatic hydrocarbons (1), benzenoids (23), esters (8), furanones (2), furans (3), hydrocarbons (25), ketones (10), lactones (1), nitrogen-containing volatiles (8), sulfur compounds (2), terpenes (8), and others (7). The highest sum of peak areas for almost all above-mentioned chemical groups of VOCs was noted in the case of plants infested by insects with a natural microbiome (especially in the larval stage), in comparison to plants damaged by bacteria-depleted insects [L(AB-) vs. L(AB+) and A(AB-) vs. A(AB+)], as well as to other plant treatments. This was especially noticeable in the case of benzenoids, esters, hydrocarbons, and lactones groups. Interestingly, plants damaged by insects with a natural microbiome released the highest number of VOCs belonging to benzenoids (22) compared to other plant treatments (≤ 17) (Fig. 2). A bubble plot representing the differences in the sum of volatile organic compounds (VOCs) peak areas (categorized based on chemical group membership) for the following studied plant treatments: C-control (undamaged plants); plants damaged by CLB larvae (L(AB-)), adults (A(AB-)) with a natural microbiome; bacteria-depleted larvae (L(AB+)) and adults (A(AB+)); mechanically damaged plants imitating the feeding of CLB larvae (md/L) and adults (md/A). The size of each bubble indicates the number of VOCs identified within a specific chemical group, while the placement on the plot signifies the total sum of peak areas of VOCs attributed to that particular analytical group. The highest sum of VOC peak area value changes are highlighted in bold font The higher emission of the aforementioned 16 chemical groups of VOCs from plants damaged by larvae with a natural microbiome compared to those exposed to adult insects with a natural microbiome [L(AB-) vs. A(AB-)], except for the chemical groups of furans and nitrogen-containing volatile may be due to differences in the composition of the bacterial microbiome between the two CLB developmental stages studied, or from another way of leaf tissue damage (Fig. 2). Interestingly, acids were detected only in untreated plants (controls) (Z-11-hexadecenoic acid) and mechanically damaged plants simulating larval feeding (md/L) (Z-11-hexadecenoic acid and aminoacetic acid), while two furanones (dihydro-3,3-dimethyl-2(3 H)-furanone and 5,5-dimethyl-2(5 H)-furanone) were identified solely in plants exposed to bacteria-depleted larvae (L(AB+)) (Fig. 3 and Supplementary Material S2). (a) The heatmap illustrates the categorization of 125 statistically significant volatile organic compounds (VOCs) based on the plant treatment (bottom of the heatmap) and categories to which a specific compound belongs (right side of the plot), and the overall amount. Results are given on a log2 scale (green leaf volatiles (GLVs) are highlighted in green). The highest VOC emissions are shown in red and the lowest in blue color. (b) VENN diagrams indicate the number of shared and unique VOCs for all analyzed plant treatments. (c) A table describing the category and average sum of peak area of 24 most differentiating VOCs among all treatment conditions. The 10 volatile compounds with the highest average peak area for larvae or adults-damaged plants containing a natural microbiome were marked with an asterisk. C-control (undamaged plants); plants damaged by CLB larvae (L(AB-)), adults (A(AB-)) with a natural microbiome; bacteria-depleted larvae (L(AB+)) and adults (A(AB+)); mechanically damaged wheat plants imitating the feeding of CLB larvae (md/L) and adults (md/A). Green leaf volatiles were marked with an asterisk (*). The highest average peak area value changes are highlighted in bold font CLB-Associated Bacteria (Especially of Larval Stage) Contribute to the Increased Emission of Butyrolactone, Butyl Ester Acetic Acid and Pentane Next, we wanted to see which specific compounds belonging to the aforementioned 16 chemical groups are emitted in greater amounts due to the CLB-associated bacteria. To determine the qualitative and quantitative distribution of individual VOCs identified across all plant treatments, the results were presented on a heatmap (Fig. 3a). The unique and shared VOCs for each plant treatment were identified using a Venn diagram (Fig. 3b). Among the identified 125 VOCs, 59 compounds were common to all tested plant treatments (Fig. 3b), of which 24 were identified as the most differentiating across all studied plant treatments, as demonstrated by the PCA analysis (Fig. 3c and Supplementary Material S3). For the following 10 out of the 24 most differentiating VOCs we observed the highest average peak area for plants on which insects with a natural microbiome (both, larval and imago stage) fed in comparison to those damaged by bacteria-depleted insects [L(AB-) vs. L(AB+) and A(AB-) vs. A(AB+)]: 2-ethyl-1-hexanol (alcohol), tetrahydro-4a,8a-butano[1.4]dioxino[2.3-b]−1.4-dioxin and 1.3.5-cycloheptatriene (aromatic compounds), benzene (benzenoid), butyl acetic acid and 2.2.4-trimethyl-1.3-pentanediol diisobutyrate (esters), hexane, pentane, and heptane (hydrocarbons), and butyrolactone (lactone) (Fig. 3c). For plants infested by CLB larvae and adults with a natural microbiome, the average peak area for butyrolactone was, respectively, at least 26 and 10 times higher, for butyl ester acetic acid 57 times and 3 times higher, for pentane 27 times and 9 times higher, compared to plants on which bacteria-depleted CLB larvae and adults were feeding, respectively. Among the 24 most differentiating VOCs, two belonged to GLVs, namely benzene (with 9 times higher average peak area for (L(AB-) vs. L(AB+)) and hexane (detected with ~ 6 times higher amounts for A(AB-) vs. A(AB+)). Wheat Volatile Compounds Elicited by CLB-Associated Bacteria Were Partly Distinct from those Induced by CLB Alone The final stage of the study was to identify unique VOCs to each tested plant treatment and determine VOCs whose emissions were induced exclusively by CLB-associated bacteria (potential microbial-induced plant volatiles). Unique VOCs were determined for plant treatments as indicated previously by the Venn diagram (Fig. 4). All these unique VOCs assigned to the appropriate chemical group and with an indication of their average peak areas are listed in Table 1. A Venn diagram comparing the VOCs identified in plants fed on by larvae or adults with a natural microbiome [L(AB-) ∪ A(AB-)] versus those identified from plants damaged by bacteria-depleted larvae or adults [L(AB+) ∪ A(AB+)]. A list of unique VOCs induced by CLB-associated bacteria (15) and those related to insect feeding (11) was compiled, with the exact assignment of each compound to the respective plant treatment (L - associated with larvae feeding, A - associated with adults feeding). The table additionally shows the chemical group assignment and the average peak area of each VOCs induced by CLB-associated bacteria (15) and those related to insect feeding (11). The highest average peak area value changes are highlighted in bold font Table 1The unique VOCs identified in each plant treatment. C-control (undamaged plants); plants damaged by CLB larvae (L(AB-)), adults (A(AB-)) with a natural microbiome; bacteria-depleted larvae (L(AB+)); mechanically damaged plants imitating the feeding of CLB larvae (md/L). Green leaf volatiles are marked with asterisk (*)Plant treatmentAnalyteChemical groupAverage peak areaL(AB-)2,4-dimethylHexanehydrocarbon 6.04E + 05 10,13-dioxo-Tricyclo[6.3.3.0]tetradec-4-eneother 1.60E + 05 A(AB-)(9Z)−9-Heptadecene-4,6-diyn-8-olalcohol 4.01E + 05 Chlorobenzene * benzenoid 7.43E + 05 L(AB+)dihydro-3,3-dimethyl-2(3 H)-furanone *furanone 2.72E + 05 5,5-dimethyl-2(5 H)-furanone *furanone 2.20E + 05 3,3-dimethyl-1-butenehydrocarbon 2.14E + 05 2,6,6-Trimethyl-2-cyclohexene-1,4-dioneketone 2.09E + 05 N, N’-Bis(2-methyl-2-nitrosopentan-4-one)nitrogen-containing volatile 2.95E + 05 1-(1-butenyl)Pyrrolidine nitrogen-containing volatile 2.74E + 05 md/L Aminoacetic acid acid 2.11E + 04 methoxy-phenylOxime nitrogen-containing volatile 1.31E + 04 2-Amino-1-(o-methoxyphenyl)propane nitrogen-containing volatile 6.13E + 03 á-Myrcene terpene 1.12E + 05 3,7,7-trimethyl-Bicyclo[4.1.0]heptan-2-oneterpene 4.81E + 04 4,6,6-trimethyl-Bicyclo[3.1.1]hept-3-en-2-oneterpene 3.25E + 04 C 1-chloroHeptane hydrocarbon 4.42E + 04 Argon other 6.79E + 04 Sulfurous acid, dodecyl 2-propyl estersulfur compound 2.08E + 04 The unique VOCs identified in each plant treatment. C-control (undamaged plants); plants damaged by CLB larvae (L(AB-)), adults (A(AB-)) with a natural microbiome; bacteria-depleted larvae (L(AB+)); mechanically damaged plants imitating the feeding of CLB larvae (md/L). Green leaf volatiles are marked with asterisk (*) Two unique VOCs were recorded for plants damaged by larvae with a natural microbiome (L(AB-)), namely 2,4-dimethylhexane (6.04E + 05), 10,13-dioxo-tricyclo[6.3.3.0]tetradec-4-ene (1.60E + 05), and two for plants infested by adults with a natural microbiome (A(AB-)): (9Z)−9-heptadecene-4,6-diyn-8-ol (4.01E + 05), chlorobenzene (7.43E + 05, classified as GLV) (Fig. 3b; Table 1). Interestingly, no unique VOCs were detected in plants damaged by bacteria-depleted adults. In contrast, six unique VOCs were identified in plants infested by bacteria-depleted larvae (L(AB+)). These compounds belonged to various chemical groups, including to furanones (dihydro-3,3-dimethyl-2(3 H)-furanone (2.72E + 05) and 5,5-dimethyl-2(5 H)-furanone (2.20E + 05)), nitrogen-containing volatiles (N, N’-Bis(2-methyl-2-nitrosopentan-4-one) (2.95E + 05) and 1-(1-butenyl)pyrrolidine (2.74E + 05)), a hydrocarbon (3,3-dimethyl-1-butene (2.14E + 05)) and ketone (2,6,6-trimethyl-2-cyclohexene-1,4-dione (2.09E + 05)) (Fig. 3b; Table 1). For plants mechanically damaged in a manner similar to larval feeding (md/L), six unique VOCs were noted. Interestingly three of them belonged to terpenes: (α-myrcene, 3,7,7-trimethyl-bicyclo[4.1.0]heptan-2-one, and 4,6,6-trimethyl-bicyclo[3.1.1]hept-3-en-2-one). Control (C, undamaged) plants emitted 3 unique VOCs: 1-chloroheptane, argon, and sulfurous acid, dodecyl 2-propyl ester (Fig. 3b; Table 1). Next, we sought to determine whether CLB-associated bacteria themselves might be responsible for the emission of unique, specific VOCs. To this end, VOCs emitted by plants damaged by insects with a natural microbiome were compared to those emitted by plants damaged by bacteria-depleted insects [L(AB-) ∪ A(AB-) vs. L(AB+) ∪ A(AB+)] to identify potential microbial-induced plant volatiles (Fig. 4). Based on the results, it was found that CLB-associated bacteria may induce the synthesis of 15 VOCs, 10 of which were common to both developmental stages [L(AB-) ∪ A(AB-)] (Fig. 4). Notably, a large portion of these common VOCs belonged to the benzenoid group. In the case of 8 of the 10 shared VOCs [L(AB-) ∪ A(AB-)], higher average peak area values were recorded for plants damaged by larvae with a natural microbiome in comparison to those damaged by adults with a natural microbiome (Fig. 4). At least 1.5 fold increases in emissions were recorded for the following compounds: 1,1’-bicyclohexyl (~ 1.6 times higher), 3,5-di-tert-butyl-4-hydroxybenzaldehyde (~ 1.9 times higher), (2-methyl-1-butenyl)-benzene (2 times higher). Only in the case of 2,2-dimethyl-1-hexanol and nitromethane, a higher average peak area was observed to be emitted from plants damaged by adults with a natural microbiome, compared to plants infested by larvae with a natural microbiome (Fig. 4). Additionally, 11 VOCs associated with feeding by bacteria-depleted insects [L(AB+) ∪ A(AB+)] were determined (Fig. 4), of which only three were common to both CLB developmental stages. The highest average peak area was observed for 4-methyl-1-pentanol (belonging to GLV), with a higher average value recorded from plants on which adult insects fed, compared to those damaged by larvae (~ 1.64 times higher). Among the other two common VOCs, one was a hydrocarbon (4,4-dimethyl-(E)−2-pentene) and the other was a terpene (3,7-dimethyl-1,6-octadien-3-ol). The remaining eight VOCs were classified as specific to bacteria-depleted larvae feeding, including the two previously mentioned furanones (Fig. 4). Discussion The main method of reducing losses caused by CLB is currently chemical protection, but new, environmentally friendly strategies for controlling this pest are being sought. Among natural enemies that may limit CLB population are hymenopteran parasitoids, ground beetle (Carabidae), rove beetle (Staphylinidae), several ladybird species (Coccinellidae) (Kheirodin et al. 2019; Samu et al. 2024). Plants have evolved many defense mechanisms against herbivorous insects (Wielkopolan and Obrępalska-Stęplowska 2016), including the emission of a specific blend of volatiles (Dicke and Baldwin 2010; Frago et al. 2017; Raglin et al. 2022). Many reviews have summarized the function of gut bacteria in insects (Wielkopolan and Obrępalska-Stęplowska 2016; Coolen et al. 2022; Mondal et al. 2023), but there is little literature data on the effect of insect-associated bacteria on emissions of VOCs by insect-damaged plants. In previous research, we have shown that CLB-associated bacteria can alter the defense response of wheat plants. For instance, we noted lower expression of gene encoding serine protease inhibitor in plants damaged by larvae with a natural microbiome, compared to plants damaged by bacteria-depleted larvae which is beneficial for insect (Wielkopolan et al. 2018, 2022). Therefore, we assumed that CLB-associated bacteria could also alter the emission of VOCs during the first four hours of CLB feeding, which can have a significant impact on communication between plants, deterring the pest or attracting its parasitoids, and predators. We collected VOCs within four hours of the CLB feeding, because it is known that the plant in response to a stress factor can quickly release VOCs (depending on the pathway of synthesis of the compound) (Smith and Beck 2015), in order to identify those compounds that should be the subject of further studies aimed at indicating their function in wheat-CLB interactions. As we hypothesized, CLB bacteria being an additional factor influenced the defense response of wheat plants, which is reflected in the higher number and higher peak area value of VOCs (especially those belonging to hydrocarbons, benzenoids, esters, and lactone chemical group) released by plants damaged by insects with a natural microbiome, compared to plants on which bacteria-depleted insects were feeding. For individual volatile compounds, higher peak areas for butyrolactone, butyl ester acetic, and pentane were recorded from plants damaged by insects with a natural microbiome (especially for larval stage), compared to plants infested by bacteria-depleted insects. Our results are consistent with those obtained by Wari et al. (2019) where honeydew secretions of the rice brown planthopper (Nilaparvata lugens, BPH) were found to strongly elicit the accumulation of phytoalexins and release of VOCs from rice leaves (Oryza sativa), that served to attract natural enemies of herbivores. It was suggested that rice plants perceived BPH via their honeydew-associated bacteria, because bacteria-depleted BPH caused much less defense in rice (phytoalexin accumulation) compared to BPH with a natural microbiome. In the literature, opposite observations can also be found, such as the one reported by Frago et al. (2017) regarding the interaction between Vicia faba and pea aphids. They indicated a reduction in plant VOC emissions due to pea aphid-associated bacteria (Hamiltonella defensa) consequently reducing the recruitment of the parasitic wasp (Aphidius ervi). This highlights the existence of species-specific responses in the analyzed systems. To the best of our knowledge, studies on the impact of beetle-associated bacteria on the emission of VOCs are not available, therefore we decided to analyse it on example of CLB. The bacterial microbiome of the larva differs from that of the adult, namely, it is more biodiverse and contains more different genera of bacteria (Wielkopolan et al. 2021). We therefore assumed that the difference in the bacterial microbiome between the two CLB developmental stages would be reflected in the composition of the emitted VOCs. This was confirmed by a higher peak area value for larvae-damaged plants than for adults with a natural bacterial microbiome. The statistically insignificant results for this comparison may be due to some differences in the microbiome composition between individuals. It is known, that the manner of tissue damage helps the plant recognize the herbivore pest, but it is not sufficient to trigger a full defense response (Acevedo et al. 2015). Importantly, mechanical damage along with CLB insect contact (with or without the microbiome) makes the plant’s defense response stronger than in the case of mechanical damage alone - which has also been observed in the case of VOC emissions from wheat plants. Similar results were obtained by Piesik et al. (2010) who indicated that CLB-damaged plants emitted more VOCs than these wounded mechanically. Although in our study, no statistically significant differences were found in the emission of VOCs between plants damaged mechanically similarly to larvae and similarly to adults feeding, respectively, the blend of VOCs between the two treatments differed in the number of VOCs, value of peak areas, and in the emission of unique compounds (we recorded 6 unique compounds for plants damaged similarly to larvae feeding and none for those damaged similarly to adult feeding). Perhaps some of the unique VOCs synthesized under a specific stressor (insect or bacteria) could be used in practice to reduce crop losses and diagnose the presence of herbivorous pests on wheat (Liu et al. 2024; Tian et al. 2024). In our study, unique VOCs were determined for each plant treatment, except for plants damaged by bacteria-depleted adult CLB, and for plants damaged mechanically similar to adult CLB feeding. Interestingly, for plants infested by bacteria-depleted larvae, six unique VOCs (belonging to furanone, nitrogen-containing volatiles, hydrocarbon, and ketone chemical groups) were detected. Also for plants mechanically damaged in a manner similar to larval feeding, six unique VOCs were noted, three of them belonged to terpenes. For plants damaged by CLB with a natural microbiome, two compounds with anti-bacterial and anti-fungal properties (2,4-dimethylhexane and 10,13-dioxo-tricyclo[6.3.3.0]tetradec-4-ene) (Udobre et al. 2015; Gambo et al. 2022) were found for plants damaged by larvae and two for plants damaged by adults ((9Z)−9-heptadecene-4,6-diyn-8-ol and chlorobenzene classified as GLV). Further research may provide information on what role these compounds could play in wheat plant-CLB-predator, parasitoid interactions, which is of great importance for practice because insect herbivores rely on plant-produced cues that provide information about plant location, nutritional quality, and defensive status (Bruce et al. 2005; Grunseich et al. 2020). For almost all of the 16 chemical groups to which the identified VOCs were assigned, the highest value of peak area was recorded for plants infested by insects with a natural microbiome, especially in the larval stage. This was especially noticeable in the case of benzenoids, esters, hydrocarbons, and lactones chemical groups. Interestingly, for plants damaged by insects with a natural microbiome, the chemical group of benzenoids included 22 compounds. Benzenoids were also detected among the unique compounds emitted by plants damaged by CLB with a natural microbiome. For instance, for plants on which larvae with a natural microbiome fed, the unique benzenoid was 2,6-bis(1,1-dimethylethyl)−4-(1-oxopropyl)phenol, which is known for its antibacterial activity (Meshram et al. 2013), while for plants infested by CLB adults with a natural microbiome was chlorobenzene (classified as GLV). It was reported that benzenoids can act as attractants for pollinators (e.g., benzaldehyde), while also serving as compounds associated with defense against herbivores (e.g., benzothiazole) (Slavković and Bendahmane 2023). Interestingly, only one VOC belonging to lactones, namely butyrolactone, was identified for all analyzed plant treatments and exhibited the highest peak area in plants damaged by insects with a natural microbiome (especially in the larval stage). It was indicated that butyrolactone might act against pathogenic fungi of the Fusarium genus (Wang et al. 2022), and has been identified as a VOC induced by the pathogen, following Phytophthora sp. infection (Favaro et al. 2024). Among identified compounds, we identified 15 VOCs emitted due to the presence of the CLB-associated bacteria (6 of them belong to benzenoids) and 11 VOCs released by plants on which bacteria-depleted insects fed. Among 11 VOCs, three were found to be common to the feeding activity of both bacteria-depleted larvae and adults, with the highest average peak area observed for 4-methyl-1-pentanol (an alcohol belonging to GLVs). It was indicated that 4-methyl-1-pentanol exhibits activity against a broad spectrum of pathogens, including the fungus Valsa pyri (Shi et al. 2024). Additionally, its potentially repellent effect on the behavior of adult Chlorophorus diadema (Motschulsky) was indicated (Zhang Yan et al. 2015). It is considered that GLVs, C6 aldehydes, alcohols, and esters are enzymatically produced from membrane lipids upon disruption of membrane integrity in leaves as a result of mechanical wounding, or insect feeding (Ameye et al. 2018). In the presented results, CLB-associated bacteria stimulated plants to produce 6 aldehydes, 10 alcohols and 7 esters. GLVs and terpenoids are considered to be common compounds of herbivore-induced plant volatiles (Turlings and Erb 2018), especially for chewing insects (Ameye et al. 2018), that are implicated in the attraction of natural enemies (Turlings and Erb 2018). In the presented results 8 terpenes were identified, two of which were observed on plants where CLB (larvae and adults) with a natural microbiome fed (limonene and 2,6,6-trimethyl-, (ñ)-bicyclo[3.1.1]hept-2-ene). GLVs (six-carbon volatile oxylipins) are produced rapidly (sometimes it is named GLV burst) in response to stresses, such as insects herbivory that damage plant tissue. GLVs are the first compounds emitted by plants, and therefore should be considered important mediators between plants and the organisms that interact with them (Matsui and Engelberth 2022). GLVs can, among others, deter herbivores, and attract their natural enemies (Dicke et al. 2009; Scala et al. 2013), as well as activate, and promote plant defences (Engelberth et al. 2004; Matsui and Engelberth 2022). Importantly, other plants can use GLVs to set their defenses in a process called priming (Matsui and Engelberth 2022). In our study, twelve GLVs were identified, nine of which were emitted from plants damaged by CLB with natural microbiomes regardless of the developmental stage (heptanal, cyclohexanone, benzene, 1-methoxy-2-propyl acetate, 2-ethylfuran, hexane, 2-methylpenatne and 2-propyl-1,3-dioxolane), while one was unique to plants on which insect adults with a natural microbiome were feeding (chlorobenzene). Due to the fact that GLVs are rapidly emitted by plants in response to insect feeding and due to their functions, their role in the wheat-CLB-and CLB associated bacteria system should be further determined. Conclusions Summarizing, according to numerous studies, including those using the CLB larvae, insect herbivores often rely on microbial symbionts to overcome host-plant defenses (Wielkopolan et al. 2022). In this study, we found out that damage alone may be sufficient to provoke the release of certain volatiles from plant tissues, but it is clear that the feeding of CLB with CLB-associated bacteria contributes to higher and more diverse emissions of VOCs, with specific categories of emitted volatiles, including unique ones. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Author Contributions B.W., A.O.S. and P.F. designed the experiments. B.W., M.M., and P.F. performed the experiments. B.W., A.O.S., P.F. and M.M. analyzed the data and wrote the paper. P.F. and A.O.S. designed the Figures, A.O.S supervised the study. All the authors reviewed and approved the final manuscript. All authors equally contributed to this work. Funding This research received no external funding. Data Availability Data is provided within the manuscript or supplementary information files. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Beata Wielkopolan and Patryk Frąckowiak are first co-authors. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Data Availability Statement Data is provided within the manuscript or supplementary information files.