Essential-oil constituent blend and Botrytis cinerea exert opposing effects on the targeted peel metabolite profile of ‘Hicaz’ pomegranate during cold storage under controlled conditions

The peel of pomegranate ( Punica granatum L.) concentrates a large share of the fruit’s bioactive metabolites, and this fraction is at risk during cold storage from Botrytis cinerea , the necrotrophic agent of grey mould. The present study characterized how an essential-oil constituent blend (EOB) of thymol, eugenol and 1,8-cineole, applied at three doses with and without the pathogen, is associated with changes in a targeted metabolite panel of ‘Hicaz’ pomegranate peel over 28 days of cold storage at 5 ± 1 °C. Fifty-two metabolites spanning phenolics (12), anthocyanins (8), organic acids (11) and free amino acids (21) were quantified by HPLC at four samplings (days 7, 14, 21 and 28) and analyzed by three-way factorial ANOVA with Benjamini–Hochberg false-discovery-rate correction, followed by multivariate, partial-correlation network and discriminant analyses. EOB dose, pathogen inoculation, storage day and the dose × pathogen interaction each affected the majority of metabolites (FDR q < 0.05 for 45, 47, 48 and 44 of 52 metabolites, respectively). Applied alone, the blend increased total anthocyanins at every dose (+10% to +80%) and total organic acids at the intermediate dose (+32%), and the intermediate dose kept the summed pools within 6% of the control, whereas inoculation was accompanied by a broad depletion of the targeted panel reaching −55% relative to the control; the highest dose reduced this loss only marginally (−48%), a difference that was not statistically significant. The free amino acid pool, in which arginine was the most abundant amino acid, showed the largest reduction. Organic acids remained the most strongly intercorrelated class after statistically removing treatment and storage-day effects (mean partial |r| = 0.48 versus 0.17–0.20 for the other classes). A two-component PLS-DA model separated inoculated from pathogen-free samples (leave-one-storage-day-out accuracy 91.7%, AUC 0.98; permutation test P = 0.001), identifying candidate discriminant metabolites. The study was conducted on a single cultivar (‘Hicaz’) under controlled storage conditions. The highest dose was the only treatment that suppressed decay to the end of storage (lesion area 11.9% versus 26.8% in fruit inoculated without EOB at day 28; no decay occurred in any pathogen-free treatment; soluble solids and titratable acidity were unaffected by treatment, and peel colour varied without a consistent direction), whereas the intermediate dose best preserved the metabolite panel; decay suppression and metabolite preservation were thus maximized at different doses, indicating that the metabolite profile cannot be used as a proxy for disease control and that dose optimization must weigh both criteria.