Integrated Volatilomics and Lipidomics Identify Lactones as Correlation Hubs Associated With Lipid Remodeling, Flavor, and Texture in Postharvest Nectarines

-aldehyde-to-lactone flavor transition. Lipidomic profiling identified 234 lipid species across five classes and 23 subclasses, dominated by glycerolipids (38.9%) and glycerophospholipids (37.2%). During storage, 192 species (82.1%) were differentially accumulated, featuring coupled glycerophospholipid degradation and triacylglycerol accumulation. Double bond index analysis further revealed class-specific unsaturation remodeling. Spearman correlation (|ρ| > 0.8, FDR < 0.05) yielded 454 strong volatile-lipid pairs (62.8% negative) and 70 texture-metabolite pairs. Mantel test, canonical correlation analysis, and Procrustes analysis confirmed robust inter-omics associations. In the correlation network, γ-octalactone and γ-decalactone emerged as hub nodes linking lipid metabolism to flavor dimensions. γ-Decalactone exhibited the strongest firmness correlation among all metabolites (ρ = -0.94), suggesting potential as a nondestructive softening indicator. Unsaturation-stratified analysis revealed that glycerophospholipid monounsaturated fatty acid (MUFA) species (double bond = 1) exhibited the strongest flavor associations, whereas class-level lipid totals were nonsignificant. This highlights molecular species specificity in lipid-flavor linkages. Machine learning identified four consensus markers (d-limonene, monogalactosyldiacylglycerol [MGDG] 36:4, MGDG 36:6, and phosphatidylethanolamine 43:2) that discriminate Control from 1-MCP-treated fruit, providing molecular targets for preservation optimization. PRACTICAL APPLICATIONS: The strong correlation between γ-decalactone and firmness (ρ = -0.94) suggests that gas-sensor or electronic-nose detection of this peach-aroma volatile could enable nondestructive assessment of nectarine softening. The correlation network further suggests that membrane lipid catabolism is closely associated with both lactone-based flavor development and texture loss, providing a mechanistic basis for optimizing 1-methylcyclopropene (1-MCP) dosage and timing to balance firmness retention with flavor preservation. The four consensus markers may additionally serve as molecular references for shelf-life prediction and quality grading. Regarding sensor-based implementation, the wide dynamic range of γ-decalactone observed in this study (<1 to ∼900 ng/g fresh weight) and its high concentration at marketable softening are favorable for electronic nose detection; however, practical deployment would require standardized headspace sampling protocols and cultivar-specific calibration.