Peizi Sun, Songyi Lin, Yuting Tan, Yajie Qin, Xiaobin Bi, Dongmei Li
Off-odor development during retort sterilization limits the sensory quality of canned Antarctic krill, but measurements in whole krill cannot distinguish intrinsic fatty acid oxidation from matrix-dependent odor expression. To separate these effects, six major fatty acids (C14:0, C16:0, C16:1, C18:1, eicosapentaenoic acid [EPA], and docosahexaenoic acid [DHA]) were examined in a three-tier thermal oxidation model comprising fatty acid-only, lipid-depleted krill matrix, and complete krill meat systems heated at 115 °C for 36 min. Volatile profiles were characterized by gas chromatography-ion mobility spectrometry (GC-IMS) and headspace solid-phase microextraction-gas chromatography-mass spectrometry (HS-SPME-GC-MS), together with peroxide value (POV), thiobarbituric acid-reactive substances (TBARS), odor activity values (OAVs), sensory evaluation, and exploratory cross-tier correlation analysis. In the fatty acid-only model, EPA and DHA showed the highest POVs (5.26 ± 0.53 and 4.96 ± 0.57 meq/kg), whereas C16:1 and C18:1 produced broader volatile profiles. After matrix incorporation, these patterns changed: C18:1 supplementation was associated with a pronounced aldehydic profile, whereas DHA- and EPA-supplemented complete krill meat showed the highest TBARS values (3.37 ± 0.40 and 3.12 ± 0.37 mg MDA/kg) and high fishy odor scores (6.21 and 5.95). Total OAV was not significantly associated with fishy or oxidized odor, whereas (Z)-4-heptenal, (Z)-1,5-octadien-3-ol, 1-penten-3-ol, and trimethylamine showed closer sensory associations. Thus, intrinsic oxidation potential did not directly predict off-odor intensity; sensory deterioration was more closely related to the matrix-dependent expression of selected potentially odor-active compounds than to fatty acid unsaturation or total OAV.