PSVIII-12 Hexanal, as Identified Through Flash Gas Chromatography Electronic Nose, Strongly Correlates with Pv in Rendered Chicken Meal Samples, Although Additional Compounds Influence Aromatic Palatability

Abstract

Abstract Samples of rendered chicken meal (CM) and chicken by-product meal (CB) were obtained from a commercial plant with samples representing various storage time points and a range of peroxide values from < 10 (1), 20-50 (2), 50-100 (3), 100-200 (4). Samples were utilized in a novel aromatic palatability trial with 20 working Labrador Retrievers (n = 10 males and n = 10 females). Aggregated interaction time showed the most pronounced difference between the lowest PV sample and the highest PV for both CB and CM (P < 0.0001, 0.001). Both CB1 and CM1 were significantly different from all other samples (P = 0.05). Samples were also sent to a third party for HERACLES Neo flash gas chromatography electronic nose analysis (e-nose). Analysis of e-nose chromatograms showed good discrimination between samples (DI = 97) with the largest spatial differences between the low and high PV samples. Chromatogram peaks were tentatively identified according to their Kovats index and the AlphaChemBase database. Component 1 (X-axis) accounted for 99.4% of the difference while component 2 (Y-axis) showed 0.54% of the difference. The strongest odor, as determined by the highest area peak was hexanal which produces a fatty and fishy smell. When evaluated in a bivariate model hexanal vs PV was well correlated on a 2P curve (R2= 0.986, P < 0.0001). However, hexanal was less correlated with aromatic palatability (R2 = 0.616. P = 0.09). Therefore, a primary least squares model was created to better display the relationship between palatability and oxidation. Utilizing two factors, the PLSR model explained 72% of the variation in the e-nose peak area data (X) and 90% of the variation in the PV*palatability (Y). Of the 61 aromatic compounds identified, the model identified 45 compounds with a VIP over 0.8. Both PV and palatability were well-fitted by the model as indicated by their position between the outer two ellipses (R2 = 75% to 100%). In the model, Hexanal had a VIP of 1.1 suggesting it was an important component, although other compounds had significantly better VIP scores and are likely to better predict palatability. Therefore, these data show aromatic hexanal is an accurate predictor of the oxidation of a product, although it may not provide the full picture regarding palatability.