Integrated volatile and non-volatile metabolomics reveal aroma chemotypes in Capsicum cultivars

flavor necessitates advanced profiling strategies beyond conventional methods. Here, SPME-GC × GC-MS and UHPLC-QTOF-MS combined with feature-based molecular networking (FBMN) were employed to comprehensively profile the flavor-related metabolomes of four representative Chinese chili cultivars (XML, CTJ, ZDT, and DLJ). A total of 452 volatile compounds were tentatively identified and 53 non-volatile metabolites annotated. Multivariate analysis demonstrated clear cultivar-specific metabolic separation. Odor threshold-based screening identified 13 potent odorants; notably, 2-methoxy-3-(2-methylpropyl)pyrazine was absent only in XML, while ultra-low-threshold trans-β-ionone was most abundant in this cultivar. These volatile signatures delineated three distinct aroma chemotypes: norisoprenoid-sesquiterpene (XML), monoterpenoid-floral (CTJ/ZDT), and C₉ aldehyde-green (DLJ). At the non-volatile level, FBMN revealed cultivar-specific enrichment patterns among capsaicinoid, lipid, and flavonoid molecular families. These integrated chemical fingerprints provide candidates for flavor-oriented breeding strategies.