HS-GC-IMS reveals VOC differences between tree and herbaceous peony petals

Headspace-gas chromatography-ion mobility spectrometry (HS-GC-IMS) was employed for untargeted profiling of volatile organic compounds (VOCs) in full-bloom petals of six representative cultivars collected from Caojiaping Peony Garden, Lintao, Gansu. The panel comprised one tree-peony cultivar (Paeonia × suffruticosa ‘DIANMUDAN’), four herbaceous-peony cultivars (Paeonia lactiflora ‘MINI’, ‘LANFUSHI’, ‘DAOGELASI’ and ‘YILINIQIUZHANG’) and one Itoh hybrid (‘BACILA’). Petals stored at −80°C were analyzed by single headspace injection within 30 min. In total, 64 VOCs corresponding to 113 signals were identified, comprising 18 aldehydes, 11 alcohols, 11 ketones, 7 esters, 3 acids, 5 terpenes and 4 aromatics, including monomers, dimers and isomer peaks. Fingerprint spectra revealed aldehydes as the dominant class, accounting for 48–60% of total signal intensity. Principal component analysis (PCA) and partial least-squares discriminant analysis (PLS-DA) (three principal components, cumulative variance 70.5%, 200-permutation Q² = 0.68, p < 0.01) resolved the six cultivars into four chemically distinct phenotypic clusters. Nine discriminant markers (VIP ≥ 1, variable importance in projection) were obtained; among them, epoxylinalool-M and (Z)-linalool oxide-D were uniquely elevated in ‘DIANMUDAN’, whereas 2-hydroxybenzaldehyde-M and butanal were significantly enriched in ‘YILINIQIUZHANG’, and 1,8-cineole-M determined the cooling scent note of ‘MINI’. This study establishes an HS-GC-IMS-based VOC fingerprinting framework that links floral pigmentation and chemosensory phenotype in tree peony, herbaceous peony and their interspecific hybrids, enabling early selection of superior genotypes bearing targeted scent profiles and facilitating directed transference of fragrance traits across species together with colour co-selection.