An essential-oil blend delivers dual-action control of grey mold in pomegranate by suppressing Botrytis cinerea and reshaping the husk volatilome

Abstract

Pomegranate (Punica granatum L.) fruit is highly susceptible to postharvest grey mold caused by Botrytis cinerea, which develops during cold storage and causes substantial losses, while concern over fungicide residues and resistance has intensified the search for natural alternatives. The present study evaluated an essential-oil blend (EOB) composed of thymol, eugenol and 1,8-cineole, applied at three doses, for the control of B. cinerea on pomegranate husk during 28 days of cold storage, and combined decay assessment with headspace solid-phase microextraction–gas chromatography–mass spectrometry (HS-SPME–GC–MS) profiling of the husk volatilome to elucidate the relationship between disease suppression and volatile change. Decay was effectively controlled in a dose-dependent manner: incidence at day 28 decreased from 72.4% in the inoculated control to 11.7% at the highest dose, corresponding to a disease-control efficacy of up to 83.8%, and the onset of decay was progressively delayed from day 7 in the control to days 21 and 28 in the lower- and higher-dose treatments, respectively. Volatile profiling identified 67 compounds across seven chemical classes, of which 49 differed significantly among treatments. The higher EOB doses elicited a marked, threshold-type accumulation of defense-related aldehydes and C13-norisoprenoids, including the bioactive apocarotenoid β-ionone, whereas esters and acids increased in parallel with decay and terpenes declined over storage. Multivariate, hierarchical-clustering and co-occurrence-network analyses resolved the volatilome into two distinct branches: an ester–acid module associated with disease progression and a separate aldehyde–C13-norisoprenoid module associated with the treatment response. To our knowledge, this is the first report to characterize the pomegranate husk volatilome in relation to essential-oil-mediated control of B. cinerea. These observations, together with the known antifungal activity of the oil constituents and the recognized role of C6 and apocarotenoid volatiles in plant defense, suggest that the essential-oil blend controls pomegranate grey mold through a combination of direct antifungal activity and an induced host response, and support its use as a natural, residue-free component of integrated postharvest disease management.

Graphical abstract

Highlights

• •(i) An EOB of thymol + eugenol+1,8-cineole controlled pomegranate grey mold dose-dependently (up to 83.8%).
• •(ii) Higher doses triggered a threshold-type accumulation of defence aldehydes and C13-norisoprenoids.
• •(iii) β-ionone was a key induced apocarotenoid.
• •(iv) Network analysis resolved the volatilome into separate decay- and defense-associated modules.

1. Introduction

Pomegranate (Punica granatum L.) is an economically important fruit crop of arid and semi-arid regions, valued for the nutritional and antioxidant properties of its arils and for the protective role of its thick husk (pericarp), which encloses and shields the edible tissue during the postharvest period (Ozrenk et al., 2022). Despite this natural barrier, the fruit is highly perishable in storage, and postharvest decay remains one of the principal constraints on its marketing and long-distance shipment (Caleb et al., 2013; Palou et al., 2013; Ugolini et al., 2014). Among the diseases responsible for these losses, grey mold caused by B. cinerea is one of the most damaging, since the fungus infects flowers and the fruit crown in the field, remains latent until maturation, and develops during cold storage to cause extensive rotting (Kanetis et al., 2015; Palou et al., 2013). B. cinerea ranks among the most economically important necrotrophic plant pathogens worldwide and causes major postharvest losses across a wide range of fruits, including pomegranate (Dean et al., 2012; Romanazzi and Feliziani, 2014). The conditions that prevail throughout the pomegranate storage chain, non-chilling temperatures of 5–10 °C and high relative humidity, are particularly conducive to the disease, and losses can exceed 30% of harvested fruit when no postharvest treatment is applied (Palou et al., 2013). Control has traditionally relied on synthetic fungicides such as fludioxonil; however, the repeated use of these compounds has favoured the emergence of resistant Botrytis populations and has raised growing concern over chemical residues, human health and environmental impact (Kanetis et al., 2015; Romanazzi and Feliziani, 2014). These concerns have intensified the search for safe, natural and residue-free alternatives for managing postharvest grey mold.

Plant essential oils have attracted increasing attention as natural antifungal agents because of their volatile nature, broad-spectrum activity and generally-recognized-as-safe status, and their efficacy against B. cinerea has been demonstrated across several fruit crops (Ayu et al., 2025). In apple, post-harvest treatments with essential oils and their volatile constituents have reduced grey-mold decay while modulating fruit quality, sugar, vitamin and antioxidant dynamics, biogenic amines and the volatile profile (Karakus et al., 2023; Karakus et al., 2024b; Karakus et al., 2024a; Karakus et al., 2024c). In grape, essential-oil treatments have likewise suppressed B. cinerea and influenced berry metabolism, nutritional quality and stress responses in several cultivars, including 'Narince' and 'Karaerik' (Kaya et al., 2024a; Kaya et al., 2024b; Kaya et al., 2024c). The activity of these oils is attributed largely to phenolic and terpenoid constituents such as thymol, eugenol, carvacrol and 1,8-cineole, which disrupt fungal membranes, inhibit spore germination and impair mycelial growth (Hou et al., 2020; Marei et al., 2012). Beyond their direct fungitoxic action, essential oils can also act indirectly by eliciting host defense responses; treatment of apple fruit with thyme oil, for example, controlled grey mold not only through direct antifungal activity but also through the induction of pathogenesis-related gene expression, indicating a priming of host defenses (Sivakumar and Bautista-Baños, 2014; Banani et al., 2018). Such dual modes of action make essential-oil blends particularly attractive for postharvest disease control. The three constituents of the present blend were selected to combine complementary antifungal chemistries: thymol (a phenolic monoterpenoid) and eugenol (a phenylpropanoid) are membrane-active compounds with strong, well-documented fungitoxicity against B. cinerea, whereas 1,8-cineole is a volatile oxygenated monoterpene with marked vapour-phase activity (Hou et al., 2020; Marei et al., 2012; Hassani et al., 2012). Blends of such phenolic and terpenoid constituents have repeatedly been reported to act additively or synergistically and to lower the effective dose required relative to single components (Sivakumar and Bautista-Baños, 2014), which provided the rationale for combining them here.

The fruit response to such treatments is frequently accompanied by changes in the volatile profile of the tissue, and headspace solid-phase microextraction coupled with gas chromatography–mass spectrometry (HS-SPME–GC–MS) has become a standard tool for characterizing pomegranate volatiles and their changes during storage (Beaulieu and Stein-Chisholm, 2016; Caleb et al., 2013). However, most studies on pomegranate volatiles have focused on the arils and juice in relation to flavour and sensory quality, whereas the volatile composition of the husk, the tissue that is in direct contact with the pathogen and with an applied treatment, has received comparatively little attention. To the best of our knowledge, the relationship between essential-oil-mediated control of grey mold and the accompanying changes in the husk volatilome has not been characterized, and it remains unclear whether the volatile changes reflect mainly direct antifungal action, an induced host response, or both. In many fruit–pathogen systems, the tissue volatile pool is not merely a by-product of ripening but participates actively in the outcome of infection: constitutive and induced volatiles such as C6 green-leaf compounds and terpenoids can restrict fungal growth, while the progress of decay itself releases characteristic ester and alcohol volatiles, so that the volatile profile simultaneously reflects host resistance and disease development (Hammerbacher et al., 2019; Neri et al., 2015; Banani et al., 2018).

The present study was, therefore, conducted to evaluate an essential-oil blend composed of thymol, eugenol and 1,8-cineole, applied at three doses, for the control of B. cinerea on pomegranate husk during cold storage. To our knowledge, this is the first study to characterize the husk volatilome of pomegranate in relation to essential-oil-mediated control of B. cinerea. Combining decay assessment with HS-SPME–GC–MS profiling of the husk, the study addressed the following questions: (i) to what extent, and at which doses, does the essential-oil blend suppress grey-mold decay over a 28-day storage period?; (ii) how does the treatment alter the concentration and composition of the seven major husk volatile classes during storage?; and (iii) which volatile compounds and compound classes are associated with disease progression versus with the treatment-induced response? Decay development, the dynamics of individual and grouped volatile compounds, and their interrelationships were analyzed to elucidate the relationship between essential-oil treatment, the husk volatilome, and grey-mold suppression in pomegranate.

2. Materials and methods

2.1. Plant material and treatments

Pomegranate (Punica granatum L. cv Hicaz) fruit of uniform size and free from visible defects were used in this study. Fruit was surface-disinfected by immersion in 1% (v/v) sodium hypochlorite for 2 min, rinsed twice with sterile distilled water and air-dried at room temperature before treatment. Thymol (Aldrich, C121452), eugenol (Fluka, 45980) and 1,8-cineole (Aldrich, 183164) were purchased from Sigma-Aldrich (Shanghai, China) and were used as received without further purification. The fruit were assigned to eight treatments (T1–T8), each comprising three replicates of ten fruit (10 fruit per treatment), in a completely randomized design. The treatments were: a negative control (T1, sterile distilled water); a positive control inoculated with B. cinerea (T2); the essential-oil blend (EOB) applied at three doses without the pathogen (T3, T5 and T7 for doses 1, 2 and 3, respectively); and the EOB applied at the same three doses together with B. cinerea (T4, T6 and T8). The same carrier (2 mL Tween 80, 10 mL oil and 88 mL distilled water) was used for all EOB doses, and the water control (T1) and pathogen control (T2) were prepared with the equivalent carrier solution without oil, so that the effect of the surfactant carrier was accounted for across treatments. The three working doses were obtained by adding equal volumes of each oil to 2500 mL of distilled water: 2.5 mL of each oil for dose 1, 5 mL for dose 2 and 10 mL for dose 3, corresponding to total oil contents of 750, 1500 and 3000 μL, respectively. These doses were established in preliminary range-finding trials on the same cultivar, in which lower oil contents gave incomplete control and the highest dose produced no visible phytotoxicity (rind bleaching, browning or off-odour) on non-inoculated fruit. The EOB was applied by uniform surface spraying of the whole fruit until run-off. After treatment, fruit were left at room temperature for 30 min and then inoculated at the calyx (crown) end of the fruit by depositing a B. cinerea spore suspension (1 × 10⁵ conidia mL⁻¹) onto two wound sites (each ∼2 mm deep) made with a sterile needle. All fruit were then held under cold storage at 5 ± 1 °C and 90 ± 5% relative humidity for 28 days. A schematic overview of the experimental design, treatments and sampling scheme, together with representative fruit images, is presented in Fig. 1.

Fig. 1.

Overview of the experimental design and representative pomegranate fruit. (A) Schematic of the experimental layout. Pomegranate (Punica granatum L.) fruit of uniform size were surface-sanitized and assigned to eight treatments (T1–T8), each with three replicates, in a completely randomized design: T1, water control; T2, B. cinerea control; T3, T5 and T7, essential-oil blend (EOB) at doses 1, 2 and 3 without the pathogen; and T4, T6 and T8, EOB at doses 1, 2 and 3 together with B. cinerea. The EOB (thymol + eugenol + 1,8-cineole) was applied at total oil contents of 750, 1500 and 3000 μL for doses 1–3, and inoculated treatments were challenged with a B. cinerea suspension (1 × 10⁵ conidia mL⁻¹). All fruit were held at 5 ± 1 °C and 90 ± 5% relative humidity for 28 days, with decay assessment and HS-SPME–GC–MS profiling of the husk performed at days 7, 14, 21 and 28. (B) Representative images of the fruit, showing inoculation with B. cinerea, healthy EOB-treated fruit, sound arils, early and advanced decay in the control, and fruit sampled for volatile analysis.

2.2. Assessment of decay

Decay development on the husk was recorded at 7, 14, 21, and 28 days of storage. Decay incidence was expressed as the percentage of the husk surface showing visible lesions, and disease-control efficacy was calculated relative to the positive control as (1 − decay_tᵣ_eatment/decay_positive control) × 100. At each sampling day, treated non-inoculated fruit (T1, T3, T5, T7) were additionally inspected for any visible signs of phytotoxicity or rind injury (bleaching, browning or off-odour). For volatile analysis, husk (pericarp) tissue was sampled from the equatorial region of each fruit, pooled per replicate, frozen in liquid nitrogen and ground to a fine powder prior to extraction.

2.3. Analysis of husk volatile compounds

Volatile compounds were extracted from the pomegranate husk by headspace solid-phase microextraction (HS-SPME) and analyzed by gas chromatography–mass spectrometry (GC–MS), following procedures established for pomegranate volatile profiling during storage with minor modifications (Caleb et al., 2013; Beaulieu and Stein-Chisholm, 2016). For each sample, 2 g of finely ground husk tissue was placed in a 20 mL headspace vial with 2 mL of saturated NaCl solution, and 10 μL of 2-octanol (10 mg L⁻¹ in water) was added as an internal standard. The vial was sealed and equilibrated at 40 °C for 15 min, after which the headspace was sampled with a DVB/CAR/PDMS SPME fibre (50/30 μm) for 50 min at 40 °C. The fibre was thermally desorbed in the GC injector at 250 °C for 5 min in splitless mode (splitless time 1 min). Separation was performed on a DB-Wax polar capillary column (60 m × 0.25 mm i.d., 0.25 μm film thickness; Agilent J&W) with helium as the carrier gas at a constant flow of 1 mL min⁻¹. The oven temperature was held at 40 °C for 5 min, raised to 220 °C at 4 °C min⁻¹ and held at 220 °C for 10 min. Mass spectra were acquired in electron-impact mode at 70 eV over m/z 35–350. Compounds were identified by comparison of their mass spectra with the NIST and Wiley spectral libraries and confirmed using linear retention indices calculated relative to a homologous series of n-alkanes (Van den Dool and Kratz, 1963). Each compound was semi-quantified relative to the internal standard and expressed as 2-octanol equivalents (μg L⁻¹); these values therefore represent relative, internal-standard-referenced concentrations rather than absolute quantities. The individual compounds were grouped into seven chemical classes: C6 compounds, alcohols, esters, acids, aldehydes, terpenes and C13-norisoprenoids. Volatile analysis was carried out on the same sampling days as the decay assessment (7, 14, 21 and 28 days), with three replicates per treatment. Within these classes, a subset of nine compounds was designated a priori as defense-associated volatile markers, and this subset constitutes the operational definition of the term “defense-associated volatiles” used throughout the present work. The subset comprised six aldehydes and three C13-norisoprenoids, listed individually in Fig. 4A, and was defined before any statistical comparison on two criteria: (i) the compound is derived from a biosynthetic route with an established role in plant defense, namely the lipoxygenase–hydroperoxide lyase branch of the oxylipin pathway for the aliphatic and aromatic aldehydes, or the oxidative cleavage of carotenoids for the C13-norisoprenoids; and (ii) the compound has a documented antifungal activity or a documented defense-signaling function against necrotrophic fungi, including B. cinerea (Kishimoto et al., 2006; Hammerbacher et al., 2019; Felemban et al., 2024). Compounds belonging to the ester, acid and alcohol classes, which accumulate as a consequence of tissue maceration and decay, were deliberately excluded from this subset. Total defense volatiles, therefore, denotes the summed concentration of these nine markers only. The designation is chemical and literature-based; it indicates the documented defense-related origin and activity of these compounds, and does not in itself constitute evidence that a defense response was activated in pomegranate husk.

Fig. 4.

Dose-dependent induction of defense-related volatile compounds in pomegranate husk. (A) Heatmap of log₂ fold-change (relative to the water control, T1) for nine defense-associated volatiles, six aldehydes and three C13-norisoprenoids, at day 28 across the eight treatments; cell values are log₂ fold-change. (B) Accumulation of total defense volatiles (sum of the nine markers) over 28 days of storage. (C) Day-28 dose–response relationship, with treatments grouped by essential-oil-blend (EOB) dose level and pathogen status; the shaded band marks the activation zone. Heatmap values are means of three replicates. All nine compounds differed significantly among treatments at day 28 by one-way ANOVA (P < 0.001 in every case; Fisher's LSD, P ≤ 0.05). In panels B and C, points are means and error bars the standard error of the mean (n = 3). Fold-changes were computed as log₂[(mean_t + 1)/(mean_T1 + 1)].

2.4. Statistical analysis

All experiments were arranged in a completely randomized design with eight treatments and three replicates (n = 3). Decay incidence (%) and the concentration of each volatile compound (μg L⁻¹) were analyzed separately for each sampling day (7, 14, 21, and 28 days) by one-way analysis of variance (ANOVA). When the treatment effect was significant, means were separated using Fisher's least significant difference (LSD) test at P ≤ 0.05. Results are presented as means ± standard error of the mean (SE), and the proportion of variance explained by treatment was reported as the effect size (η²). Pearson correlation coefficients were used to assess relationships between decay incidence and volatile compound classes. Multivariate structure of the volatile profile was examined by principal component analysis (PCA) on auto scaled (mean-centered, unit-variance) data and by two-way hierarchical clustering using Ward's minimum-variance linkage on Euclidean distances. A correlation-based co-occurrence network was constructed using Pearson coefficients (|r| ≥ 0.7, P < 0.001), with community structure identified by greedy modularity maximization. Significance was set at P ≤ 0.05 (*P < 0.05, **P < 0.01, ***P < 0.001). All statistical analyses were carried out in R (v4.3, R Foundation for Statistical Computing, Vienna, Austria). One-way ANOVA followed by Fisher's LSD test for mean separation was performed using the agricolae package; Pearson correlations were computed with the Hmisc package; principal component analysis was conducted with FactoMineR and factoextra; two-way hierarchical clustering and heatmaps were produced with pheatmap; and the correlation-based co-occurrence network was constructed and analyzed with igraph. All figures were generated using the ggplot2 package.

3. Results

3.1. Disease suppression and biocontrol efficacy

No decay symptoms developed on the husk of fruit treated with sterile water (T1) or with the essential-oil blend (EOB) in the absence of the pathogen (T3, T5 and T7) at any sampling point during the 28-day storage period, and these treatments remained at 0% incidence throughout. In the inoculated positive control (T2), decay was first detected at day 7 (6.2%) and increased progressively, reaching 18.5% at day 14, 38.4% at day 21 and 72.4% at day 28 (Fig. 2A). The EOB treatments delayed the onset of decay in a dose-related manner: symptoms first appeared at day 21 in T4 (dose 1), only from day 28 in T6 (dose 2), and T8 (dose 3) showed the lowest incidence among the inoculated treatments at every sampling day (Fig. 2A). At the final sampling day, decay incidence differed significantly among treatments (F = 346.49, P < 0.001; Fig. 2B). Mean incidence decreased from 72.4% in the positive control to 25.9% in T4, 20.3% in T6 and 11.7% in T8, and the four inoculated treatments fell into separate statistical groups (Fig. 2B). Across the inoculated series, decay declined steadily as the applied oil content increased (Fig. 2C). The disease-control efficacy, expressed relative to the positive control, increased from 64.2% at dose 1 to 72.0% at dose 2 and 83.8% at dose 3 (Fig. 2D). The complete day-by-day decay means, standard errors and LSD groupings are shown in Fig. 2A and B.

Fig. 2.

Dose-dependent suppression of grey mold and the accompanying defense-volatile response in pomegranate (Punica granatum) husk during 28 days of cold storage. (A) Time course of decay incidence (%) from day 7 to day 28. (B) Decay incidence at the final sampling day (day 28). (C) Inverse dose–response relationship across the inoculated series: decay incidence at day 28 (red, left axis) declined as the applied oil content increased, while the sum of the defense-related classes, aldehydes plus C13-norisoprenoids (blue, right axis), rose in parallel; the shaded area marks the defense-volatile level. (D) Disease-control efficacy of each dose, expressed as the percentage reduction in decay incidence relative to the inoculated control. T1, negative control (sterile water); T2, positive control (Botrytis cinerea alone); T4, T6 and T8, essential-oil blend (EOB; thymol + eugenol + 1,8-cineole) doses 1 (750 μL), 2 (1500 μL) and 3 (3000 μL) co-inoculated with B. cinerea. Decay appeared first in T2 (day 7), then in T4 (day 21), and only from day 28 in T6, while T8 remained lowest among the inoculated treatments, demonstrating dose-dependent disease suppression (T8 < T6 < T4 < T2). Increasing the dose from 750 to 3000 μL progressively suppressed disease (efficacy 64.2 → 72.0 → 83.8%) while amplifying the induced volatile response. Data are means of three replicates; bars and error bars indicate the standard error of the mean (n = 3). In panel B, different italic letters denote significant differences among treatments by Fisher's LSD test at P ≤ 0.05 (one-way ANOVA: F = 346.49, P < 0.001; LSD₀.₀₅ = 4.0). Control efficacy in panel D was calculated as (1 − decay_treatment/decay_positive-control) × 100 using day-28 means.

3.2. Changes in volatile compound classes

The total concentrations of the seven volatile classes changed with both treatment and storage time (Fig. 3). At day 28, the strongest treatment effects were recorded for aldehydes and C13-norisoprenoids (F = 374.6 and 92.8, respectively; both P < 0.001), which together accounted for almost all of the among-treatment variance (η² = 0.99 and 0.98). Both classes increased in all EOB-treated fruit, with a clear separation between the lower and higher doses; aldehydes reached 1841 μg L⁻¹ in T8 compared with 374 μg L⁻¹ in T1, and C13-norisoprenoids reached 790 μg L⁻¹ in T8 versus 115 μg L⁻¹ in T1 (Fig. 3A and B). C6 compounds and alcohols also differed significantly among treatments (F = 17.4 and 10.0, respectively; both P < 0.001; Fig. 3E and G), as did esters and acids, although with smaller effects (esters F = 6.8, P < 0.001; acids F = 3.3, P < 0.05; Fig. 3C and D). At the level of individual markers, the nine defense-associated volatile compounds defined in the Materials and methods (six aldehydes and three C13-norisoprenoids; individually listed in Fig. 4A) showed a dose-related response at day 28, expressed as the log₂ fold-change relative to the water control (Fig. 4A). The two higher EOB doses (T5–T8) increased these markers, whereas the lowest dose (T3, T4) showed little or no change (Fig. 4A). The largest increases were recorded for benzaldehyde, β-ionone and phenylacetaldehyde, with log₂ fold-changes in T8 of 3.8, 3.4 and 3.1, respectively (all P < 0.001; Fig. 4A). The total concentration of these defense volatiles remained low until day 21 and increased sharply by day 28 in the higher-dose treatments (Fig. 4B). When summarized by dose level, total defense volatiles in the inoculated series rose from 244 μg L⁻¹ at dose 1 to 2318 μg L⁻¹ at dose 2 and 2631 μg L⁻¹ at dose 3, with the main increase occurring between dose 1 and dose 2 (Fig. 4C). This rise in the defense marker pool occurred over the same dose range across which decay incidence declined (Fig. 2C). Esters and terpenes followed contrasting trajectories during storage (Fig. 5). Total esters increased over the storage period in most treatments and differed significantly among treatments at day 28 (F = 6.8, P < 0.001; Fig. 5A and B). In contrast, total terpenes declined steadily from day 7 to day 28, remained highest in the untreated controls at the final sampling, and differed significantly among treatments (F = 5.5, P < 0.01; Fig. 5C and D; Fig. 3F). The C6 “green-leaf” volatiles and alcohols were high in the freshly treated control fruit at day 7, decreased markedly by day 14, and were re-established in the EOB-treated fruit during later storage (Fig. 6A–C); at day 28 the C6 fraction consisted mainly of (E)-3-hexenol, (E)-2-hexenol and hexanol (Fig. 6B), and the alcohol fraction was dominated by octanol, phenylethyl alcohol and benzyl alcohol (Fig. 6D). Class means with standard errors and significance letters for each sampling day are shown in Fig. 3, while the concentrations of the individual compounds contributing to each class are given in Supplementary Table S1, with the corresponding per-compound ANOVA in Supplementary Table S2.

Fig. 3.

Concentrations of seven volatile compound classes in pomegranate husk across treatments. For each class, the upper bar panel shows the final concentration at day 28, and the lower line panel shows the full storage time course (days 7–28). (A) Aldehydes, (B) C13-norisoprenoids, (C) esters, (D) acids, (E) alcohols, (F) terpenes, (G) C6 compounds. T1, water control; T2, Botrytis cinerea control; T3/T5/T7, essential-oil blend (EOB) doses 1–3 without pathogen; T4/T6/T8, EOB doses 1–3 with B. cinerea. Aldehydes and C13-norisoprenoids showed the strongest treatment response, accumulating sharply by day 28 in all EOB-treated fruit, whereas terpenes declined over storage. Bars are means of three replicates; error bars indicate the standard error of the mean (n = 3). Different italic letters above bars denote significant differences among treatments at day 28 (Fisher's LSD, P ≤ 0.05). For each class the one-way ANOVA F-statistic, its significance (***P < 0.001, **P < 0.01, *P < 0.05, ns = not significant) and the effect size (η²) are given above each panel. Aldehydes and C13-norisoprenoids exhibited near-complete partitioning of variance among treatments (η² = 0.99 and 0.98, respectively).

Fig. 5.

Contrasting storage dynamics of two major volatile classes in pomegranate husk. (A) Ester accumulation over 28 days and (B) ester levels at day 28. (C) Terpene depletion over 28 days and (D) terpene levels at day 28. T1, water control; T2, Botrytis cinerea control; T3/T5/T7, essential-oil blend (EOB) doses 1–3 without pathogen; T4/T6/T8, EOB doses 1–3 with B. cinerea. Total esters rose progressively in most treatments, whereas terpenes declined steadily over storage; the highest EOB doses (T7, T8) retained comparatively higher terpene levels at day 28, consistent with carry-over of the applied oil constituents. Points and bars are means of three replicates; error bars indicate the standard error of the mean (n = 3). In panels B and D, different italic letters denote significant differences among treatments (Fisher's LSD, P ≤ 0.05). Day-28 one-way ANOVA: esters F = 6.8, P < 0.001, η² = 0.75; terpenes F = 5.5, P < 0.01, η² = 0.71.

Fig. 6.

Storage dynamics and day-28 composition of C6 “green-leaf” volatiles and alcohols in pomegranate husk. (A) Time course of total C6 compounds and (C) total alcohols over 28 days. (B, D) Stacked bar plots showing the contribution of individual compounds to the total C6 (B) and alcohol (D) pools at day 28 for each treatment. T1, water control; T2, Botrytis cinerea control; T3/T5/T7, essential-oil blend (EOB) doses 1–3 without pathogen; T4/T6/T8, EOB doses 1–3 with B. cinerea. In the time-course panels, points are means ± standard error (n = 3). Day-28 one-way ANOVA: C6 compounds F = 17.4, P < 0.001, η² = 0.88; alcohols F = 10.0, P < 0.001, η² = 0.81; treatment means were separated by Fisher's LSD (P ≤ 0.05). Stacked bars show mean compound concentrations (n = 3).

3.3. Multivariate structure and individual compounds

Principal component analysis of the complete 67-compound profile separated the treatments along the first two components, which explained 40.0% (PC1) and 15.7% (PC2) of the total variance (Fig. 7A–C). PC2, driven mainly by defense-related aldehydes and C13-norisoprenoids, separated the higher-dose EOB treatments (T5–T8) from the controls, whereas PC1 was associated with esters and acids (Fig. 7B). Consistent with this, two-way hierarchical clustering of the day-28 profile placed T7 and T8 in a distinct cluster characterized by elevated aldehydes and C13-norisoprenoids, while the untreated controls (T1, T2) retained the highest terpene levels (Fig. 8). At the level of individual compounds, the day-28 response of all 67 volatiles across the eight treatments is summarized in the radial map of the husk volatilome (Fig. 9), which shows, for each compound, the log₂ fold-change relative to the water control across the concentric treatment rings together with the per-compound ANOVA significance on the outer ring. Of the 67 compounds, 49 differed significantly among treatments (P < 0.05) and 29 were highly significant (P < 0.001) (Fig. 9). The aldehyde and C13-norisoprenoid sectors showed the most consistent and significant increases in the higher-dose treatments, whereas many esters were down-regulated relative to the control (Fig. 9). Concentrations of every individual compound for all treatments and sampling days are provided in Supplementary Table S1, and the full per-compound ANOVA results in Supplementary Table S2. Although the first two components together explained 55.7% of the total variance, leaving a substantial proportion distributed across higher components, this is expected for an untargeted 67-variable volatile dataset with marked compound-to-compound variability; the first two components nonetheless captured the principal, treatment-related axes of variation, and the treatment separation was independently confirmed by hierarchical clustering (Fig. 8) and by the per-compound ANOVA (Fig. 9; Supplementary Tables S1 and S2).

Fig. 7.

Principal component analysis (PCA) of the complete volatile profile (67 compounds) of pomegranate husk across all treatments and storage days. (A) Score plot of the first two principal components; symbol colour denotes treatment and symbol shape denotes storage day, with shaded 95% confidence ellipses per treatment. (B) Corresponding compound loadings, colored by chemical class. (C) Scree plot of the variance explained by the first three components (bars) and the cumulative variance (line). PCA was performed on auto scaled (mean-centered, unit-variance) concentrations of all 67 volatile compounds from 96 samples (8 treatments × 4 storage days × 3 replicates).

Fig. 8.

Two-way hierarchical clustering heatmap of the complete day-28 volatile profile of pomegranate husk. Rows represent the 67 individual volatile compounds (annotated by chemical class in the left colour strip) and columns the eight treatments. Cell colour encodes the row-standardized abundance (z-score), so that orange indicates relatively high and blue relatively low levels of a given compound across treatments. Dendrograms show the clustering of compounds (left) and treatments (top). Each cell is the mean of three replicates. Values were standardized within each compound (row z-score) before clustering. Hierarchical clustering used Ward's minimum-variance linkage on Euclidean distances, applied independently to compounds and to treatments.

Fig. 9.

Radial map of the complete day-28 volatilome of pomegranate husk. Each angular sector is one of the 67 volatile compounds, grouped by chemical class (inner colored band). The eight concentric rings represent the treatments, from T1 (innermost) to T8 (outermost); ring-cell colour encodes the log₂ fold-change relative to the water control (T1), with blue indicating a decrease and orange an increase. The dots on the outer ring summarize the per-compound one-way ANOVA significance. The radial layout makes the coordinated induction of aldehydes and C13-norisoprenoids in the higher-dose treatments, and the contrasting down-regulation of many esters, visible at a glance. Cell values are means of three replicates expressed as log₂[(mean_t + 1)/(mean_T1 + 1)]. Per-compound treatment effects were tested by one-way ANOVA (df = 7, 16); outer-ring dot size and colour denote significance (P < 0.001, < 0.01, < 0.05, or ns). Of the 67 compounds, 49 differed significantly among treatments (P < 0.05) and 29 were highly significant (P < 0.001).

3.4. Relationships between decay and volatile composition

Across all 96 samples, decay incidence was positively correlated with total acids (r = +0.42, P < 0.001) and esters (r = +0.30, P < 0.01) and negatively correlated with terpenes (r = −0.29, P < 0.01), whereas its correlations with aldehydes and C13-norisoprenoids were not significant (Fig. 10A and B). Total acids increased with decay incidence across the dataset (Fig. 10C). In the correlation-based co-occurrence network of the volatilome, the compounds resolved into several metabolic modules: a densely connected ester–alcohol–acid module was linked to decay incidence, whereas the aldehydes and C13-norisoprenoids formed a separate, internally cohesive module (Fig. 11). The strongest individual associations with decay were recorded for ethyl octanoate and hexanoic acid (r ≈ +0.47) (Fig. 11).

Fig. 10.

Relationships between husk volatile composition and grey-mold decay. (A) Pearson correlation matrix among the seven volatile classes and decay incidence; the decay row/column is outlined. (B) Correlation of each volatile class with decay incidence, ranked; orange bars indicate positive and blue bars negative associations. (C) Scatter plot of total acids against decay incidence across all samples, with the linear regression (dashed line); points are colored by decay level. Correlations were computed as Pearson coefficients across all 96 samples (8 treatments × 4 storage days × 3 replicates), using treatment- and day-specific decay values. Significance: ***P < 0.001, **P < 0.01, *P < 0.05, ns = not significant. Acids showed the strongest association with decay (r = +0.42, P < 0.001) and terpenes the strongest negative association (r = −0.29, P < 0.01).

Fig. 11.

Co-occurrence (correlation) network of the pomegranate husk volatilome across all samples. Each node is a volatile compound (or decay incidence, red), colored by chemical class; node size is proportional to its degree (hub centrality). Shaded clouds delineate the five metabolic modules recovered by community detection. Edges join variable pairs with strong correlations (orange, positive; blue, negative); red edges mark links to decay. Edges: |Pearson r| ≥ 0.7 (P < 0.001) for compound pairs and |r| ≥ 0.35 for links to decay, across all 96 samples (8 treatments × 4 storage days × 3 replicates).

4. Discussion

The combined assessment of decay development and husk volatile profiling in this study showed that an essential-oil blend (EOB) of thymol, eugenol, and 1,8-cineole controlled B. cinerea on pomegranate husk in a dose-dependent manner, and that this control was accompanied by a coordinated, dose-dependent reorganization of the husk volatilome. The volatile changes resolved into two separable branches, one associated with the advance of disease and one with the treatment-induced response, providing a metabolite-level framework for the dual mode of action of essential oils. These findings are discussed below in relation to the magnitude and timing of decay control, the induction of defense-related volatiles, the volatile signatures of disease progression, and the multivariate organization of the husk volatilome.

4.1. Dose-dependent suppression of grey mold

The EOB delayed the onset of decay and reduced its final incidence at day 28 from 72.4% in the inoculated control to 11.7% at the highest dose (Fig. 2A and B), corresponding to a disease-control efficacy of up to 83.8% (Fig. 2D). The strongly significant treatment effect at every sampling day (F = 346.49 at day 28, P < 0.001; Fig. 2B) and the orderly separation of the inoculated treatments into distinct statistical groups (Fig. 2B) indicate that the applied oil content governed both the magnitude and the timing of disease development, with decay onset progressively delayed from day 7 in the control to day 21 and day 28 in the lower- and higher-dose treatments, respectively (Fig. 2A). This direct antifungal effect is consistent with the well-documented activity of thymol, eugenol and 1,8-cineole against B. cinerea, which has been attributed to their lipophilic, phenolic and terpenoid character that disrupts the integrity of the fungal plasma membrane, causes leakage of cytoplasmic contents and inhibits spore germination and mycelial growth (Hou et al., 2020; Marei et al., 2012). Comparable dose-related control of grey mold by essential oils and their volatile constituents has been reported across a range of fruits, including apple, where treatment both reduced decay and modified fruit quality, sugar, vitamin, antioxidant and biogenic-amine profiles (Banani et al., 2018; Karakus et al., 2023, Karakus et al., 2024b, Karakus et al., 2024a, Karakus et al., 2024c), and grape, where essential-oil treatment suppressed B. cinerea while reshaping berry metabolism and stress responses in several cultivars (Kaya et al., 2024a; 2024b; 2024c). The efficacy values obtained here fall within the range reported for these systems and confirm that a thymol–eugenol–1,8-cineole blend is effective against pomegranate grey mold under cold-storage conditions. From a practical standpoint, the progressive gain in efficacy from 64.2% at the lowest dose to 83.8% at the highest (Fig. 2D) suggests that dose optimization is an important lever for maximizing protection, although the highest dose did not fully eliminate decay, indicating that the treatment is best regarded as one component of an integrated postharvest strategy rather than a stand-alone replacement for conventional control.

4.2. A coordinated induction of defense-related volatiles

A central finding of this study was that the higher EOB doses triggered a marked accumulation of defense-associated volatiles in the husk, dominated by aldehydes and C13-norisoprenoids, whereas the lowest dose did not (Fig. 3A and B; Fig. 4). These two classes showed by far the strongest treatment effects of all seven classes, accounting for almost all of the among-treatment variance (aldehydes F = 374.6, η² = 0.99; C13-norisoprenoids F = 92.8, η² = 0.98; both P < 0.001; Fig. 3), and individual markers such as benzaldehyde, β-ionone and phenylacetaldehyde increased several-fold in the highest-dose treatment. The accumulation was largely confined to the later storage period, rising sharply only after day 21 (Fig. 4B), and displayed a clear threshold-type response, with the major increase occurring between the first and second dose (Fig. 4C). It should be emphasized that, here and throughout, “defense-associated volatiles” refers specifically to the nine chemically defined markers introduced in Section 2.3, and not to the aldehyde and C13-norisoprenoid classes in their entirety. The two classes are discussed separately as bulk classes (Fig. 3A and B), whereas the marker subset is used wherever a defense-related interpretation is offered (Fig. 4). This distinction matters because a small number of aldehydes and C13-norisoprenoids without a documented defense role also contribute to the class totals, and the class-level and marker-level signals are therefore not interchangeable.

The induction of these particular volatile classes is consistent with their recognized roles in plant defense. C6 “green-leaf” volatiles (GLVs) and the related aldehydes are derived from the lipoxygenase–hydroperoxide lyase branch of the oxylipin pathway and are among the most rapidly formed wound- and defense-signaling volatiles in plants (Hammerbacher et al., 2019). Their relevance to B. cinerea in particular is well established at the genetic level: overexpression of hydroperoxide lyase in Arabidopsis enhanced resistance to B. cinerea, whereas its suppression increased susceptibility and lowered the accumulation of (Z)-3-hexenal and (E)-2-hexenal (Kishimoto et al., 2006). In the present study, the C6 fraction declined during the early storage period and was re-established in the EOB-treated fruit during later storage (Fig. 6A and B), paralleling the rise in aldehydes and suggesting renewed activity of this defense-related pathway in the treated husk. Equally notable was the strong accumulation of C13-norisoprenoids, especially β-ionone (Fig. 4A). These compounds are apocarotenoids formed by the oxidative cleavage of carotenoids, either enzymatically by carotenoid cleavage dioxygenases or non-enzymatically by reactive oxygen species (Felemban et al., 2024). Importantly, β-ionone has recently been shown to act as more than a flavour volatile: its application to Arabidopsis caused a global reprogramming of the transcriptome, affecting genes involved in stress tolerance and pathogen defense, and increased resistance against B. cinerea (Felemban et al., 2024). The pronounced, dose-dependent accumulation of β-ionone and related C13-norisoprenoids in the higher-dose treatments therefore suggests that, beyond their direct fungitoxic action, the higher EOB doses elicited a host response in the husk involving both oxylipin- and carotenoid-derived defense volatiles. Beyond Arabidopsis, comparable associations have been reported in fruit systems: aromatic aldehydes such as benzaldehyde and phenylacetaldehyde accumulate in fruit tissue in response to fungal challenge and have recognized antifungal properties, and C13-norisoprenoids of the type detected here have been linked to defense-related volatile responses in grape and other fruits following essential-oil treatment against B. cinerea (Kaya et al., 2024a; Karakus et al., 2024b). These parallels support the interpretation that the aldehyde and C13-norisoprenoid accumulation observed in the present study reflects a defense-associated, rather than a purely decay-associated, component of the volatile response.

Such a dual mode of action, combining direct antifungal activity with the priming or elicitation of host defenses, has been demonstrated for thyme oil in apple, where control of grey mold was accompanied by the induction of pathogenesis-related (PR) gene expression (Banani et al., 2018), and is consistent with the broader body of work showing that essential-oil treatments reshape fruit metabolism and stress responses in apple and grape (Karakus et al., 2024a, Karakus et al., 2024c; Kaya et al., 2024a; 2024c). The threshold-type dose response observed here (Fig. 4C) is particularly relevant, as it indicates that the induced component became prominent only once a sufficient dose was applied, which may help to explain why the lowest dose, while still reducing decay, did not mobilize the same volatile defense response as the higher doses. It should be emphasized, however, that the present study measured volatile composition rather than defense-gene expression or enzyme activity; the induction of host defenses is therefore inferred from the volatile signatures and would benefit from direct confirmation at the transcriptional or biochemical level in future work.

4.3. Volatile signatures of disease progression

In contrast to the defense-associated volatiles, the accumulation of esters and acids during storage was positively associated with decay, whereas terpenes declined and were negatively associated with it (Fig. 5; Fig. 10; Fig. 11). Total esters and acids increased over the storage period (Fig. 5A and B) and showed significant positive correlations with decay incidence across all samples (acids r = +0.42, P < 0.001; esters r = +0.30, P < 0.01; Fig. 10B), and the strongest individual associations with decay were recorded for ethyl octanoate and hexanoic acid (Fig. 11). This pattern is consistent with reports that fungal infection and the accompanying tissue breakdown increase the release of esters and fatty-acid-derived volatiles from decaying fruit. In B. cinerea-infected grapes, for example, ester acetates such as 2-phenylethyl acetate and isoamyl acetate increased markedly in infected berries, and the volatile changes were accompanied by up-regulation of an oxylipin-pathway gene, while certain fatty-acid esters changed in the opposite direction (Santos et al., 2022). More broadly, rotting fruit emits elevated volatile esters and acids as a consequence of microbial metabolism and senescence-related processes, and these compounds have been proposed as volatile markers of decay (Neri et al., 2015). The decline of terpenes over storage, with the highest residual levels in the untreated controls (Fig. 5C and D; Fig. 3F), is in keeping with the general decrease of these constituents during fruit storage and senescence, and indicates that the terpene pool was not a driver of the observed disease suppression. Crucially, the absence of a positive correlation between decay and the aldehyde or C13-norisoprenoid classes (Fig. 10B) reinforces the interpretation that these latter classes reflect the treatment-induced response rather than the decay process itself, and that the ester–acid increase is a consequence, rather than a cause, of disease.

4.4. Two distinct branches of the husk volatilome

The multivariate analyses integrated these observations into a coherent picture. Principal component analysis separated the higher-dose EOB treatments from the controls along the second component, which was driven mainly by the defense-related aldehydes and C13-norisoprenoids, whereas the first component reflected the ester- and acid-associated changes occurring during storage (Fig. 7). Hierarchical clustering likewise grouped the two highest-dose treatments (T7, T8) into a distinct cluster characterized by elevated aldehydes and C13-norisoprenoids, while the untreated controls retained the highest terpene levels (Fig. 8). Most strikingly, the correlation-based co-occurrence network resolved the volatilome into separate modules: a densely connected ester–alcohol–acid module that was linked to decay incidence, and a distinct, internally cohesive module formed by the aldehydes and C13-norisoprenoids (Fig. 11). The radial map of all 67 compounds confirmed that the most consistent and significant up-regulation occurred in the aldehyde and C13-norisoprenoid sectors of the higher-dose treatments, while many esters were down-regulated relative to the control (Fig. 9), and the per-compound analysis showed that 49 of the 67 compounds differed significantly among treatments (Fig. 9; Supplementary Tables S1 and S2). Taken together, these results indicate that the two processes operating in the treated husk, disease progression and the essential-oil-induced response, are reflected in distinct, separable branches of the volatile metabolic network. The decay-associated branch (esters, acids) tracks the advance of B. cinerea and the accompanying tissue breakdown, whereas the defense-associated branch (aldehydes, C13-norisoprenoids) accumulates in parallel with the suppression of disease in the higher-dose treatments (Fig. 2C and D; Fig. 11). This separation provides a volatile-level framework consistent with the proposed dual mode of action of essential oils, namely a combination of direct antifungal activity and the elicitation of host defense responses, and is in agreement with recent studies linking essential-oil treatment to both pathogen suppression and altered volatile and metabolic profiles in other fruit crops (Banani et al., 2018; Karakus et al., 2023, Karakus et al., 2024b; Kaya et al., 2024a; 2024b). The use of an untargeted, network-level analysis of the husk volatilome, rather than a focus on a small set of target compounds, was instrumental in revealing this organization, and may offer a useful approach for dissecting the modes of action of other natural antifungal treatments.

4.5. Limitations and future directions

Several limitations of the present study should be acknowledged and define clear priorities for future work. First, the antifungal effect of the blend was assessed only on fruit; in vitro assays of mycelial growth inhibition, spore-germination inhibition and minimum inhibitory and fungicidal concentrations (MIC/MFC) were not performed, and would be needed to quantify the intrinsic activity of the blend and of its individual constituents. Second, the study evaluated the three-component blend only; single- and two-component treatments were not included, so the present data establish the efficacy of the blend but do not formally distinguish additive from synergistic interactions among thymol, eugenol and 1,8-cineole. Third, the induced-defense component is inferred from the volatile signatures alone; defense-related gene expression (e.g., qRT-PCR of PR genes) and enzyme activities (e.g., phenylalanine ammonia-lyase, chitinase) were not measured, and direct molecular and biochemical confirmation is required before a host-defense mechanism can be firmly established. Fourth, fruit quality attributes were not quantified; although no visible phytotoxicity was observed at any dose, systematic quality assessment is needed to confirm that the higher doses are free of adverse effects. Addressing these points in future studies, ideally on naturally infected fruit and across cultivars, would strengthen the mechanistic interpretation and support practical application of the approach.

5. Conclusions

The thymol–eugenol–1,8-cineole blend provided effective, dose-dependent control of B. cinerea on pomegranate husk during cold storage, reducing decay incidence from 72.4% in the inoculated control to 11.7% at the highest dose and reaching a disease-control efficacy of up to 83.8%. This control was mirrored by a coordinated, dose-dependent reorganization of the husk volatilome, in which the higher doses elicited a marked accumulation of defense-related aldehydes and C13-norisoprenoids, including the bioactive apocarotenoid β-ionone that formed a volatile module distinct from the ester–acid module associated with disease progression. The clear separation of these two branches of the volatile network suggests that the blend acts through a combination of direct antifungal activity and an induced host response, although the latter is inferred from volatile signatures and would require confirmation at the transcriptional and biochemical levels. The untargeted, network-level analysis of the husk volatilome applied here proved a useful means of dissecting these complementary modes of action and may be readily extended to other natural antifungal treatments and fruit-pathogen systems. From an applied perspective, these findings support the use of essential-oil blends as a natural, residue-free component of integrated postharvest disease management for pomegranate, and the marked influence of dose indicates that dose optimization is a key consideration for maximizing efficacy. In the present study no visible phytotoxicity was observed on non-inoculated fruit at any dose; however, because fruit quality attributes were not quantified, the absence of adverse effects on quality cannot be confirmed, and higher doses should be evaluated for possible phytotoxic or quality effects before practical use. Future work combining volatile profiling with defense-gene expression, carotenoid metabolism, and enzyme activity, and extending the assessment to naturally infected fruit and to overall storage quality, would help to confirm the induced-defense component and to advance this approach towards commercial application.

Author contributions

S.K. and O.K. conducted the experiments. O.K. performed the statistical analyses. O.K. and S.K. contributed to data interpretation and manuscript preparation. O.K. coordinated the writing process, including preparation of the initial draft and subsequent revisions, and reviewed the final version of the manuscript. All authors contributed to the manuscript revision and approved the final version for submission.

Data publication consent

Not applicable.

Ethics approval

This study involved plant material only and did not include human participants or animal subjects; therefore, no ethical approval was required.

Consent for publication

All authors have read and approved the final manuscript and consent to its submission and publication.

Data availability

All data generated or analyzed during this study are contained within the article and its supplementary material. Decay incidence and the seven volatile classes are reported in Figs. 2, 3, 5 and 6; Supplementary Table S1 contains the concentrations (mean ± SE) and LSD groupings of all 67 individual volatile compounds for every treatment and sampling day, and Supplementary Table S2 the corresponding per-compound one-way ANOVA results. No further datasets were generated.

Declaration of generative AI and AI-assisted technologies in the writing process

During the preparation of this work the authors used AI-assisted tools solely for minor linguistic refinement. After using these tools, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article. All scientific content, analyses, figures and interpretations are the original work of the authors.

Funding

This study was conducted without financial support from any governmental, commercial or non-profit funding agency.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Handling Editor: Dr. Siyun Wang

Appendix A. Supplementary data

The following are the Supplementary data to this article: