Investigation of beef brisket palatability from three USDA quality grades

Abstract

The objective of this study was to investigate differences in smoked beef brisket palatability from three USDA quality grades (USDA Prime, average Choice, and Select quality grades) and to investigate the ability of Rapid Evaporative Ionization Mass Spectrometry to classify brisket type and sensory characteristics. Briskets (n = 54; 18 per quality grade) were seasoned with a blend of 1:1 coarse salt/black pepper, and then cooked in a smoker to a final internal temperature of 93˚C. For sensory analysis briskets were separated into point and flat portions, and then sliced perpendicular to the muscle fiber. Consumer panelists (n=360) evaluated palatability traits, acceptability of each trait, and willingness to pay (WTP). An interaction between quality grade and muscle was observed (P ≤ 0.03) for all palatability traits, proportion of acceptable samples, and WTP. Consumers could not distinguish between quality grades of the point portions for tenderness, juiciness, flavor and overall liking (P > 0.05). Point samples, regardless of quality grade were scored greater (P < 0.05) than Prime flat samples. Choice and Select flats scored lower than all other treatment combinations for tenderness, flavor liking, and overall liking. In alignment with palatability traits, consumers were willing to pay the most for point portions, regardless of quality grade (P < 0.05), followed by Prime flat portions. Choice and Select flat portions had the lowest WTP (P < 0.05). Consumer acceptability generally followed similar trends as palatability scores. Quality grade had no effect on the eating quality of the point portions of smoked briskets, and point portions received superior palatability scores to flat portions. Prime flat portions had greater eating quality compared to that of Choice and Select flat portions, and consumers were willing to pay more for what they perceived as superior eating quality. A sample was taken from the brisket flat (pectoralis profundi) and point (pectoralis superficialis) for REIMS analysis. The samples were vacuum packaged, labeled, and frozen. Frozen samples were thawed, and a molecular fingerprint of each sample was collected via REIMS to build prediction models. When Prime, Choice, and Select, flat and point sections were combined, balanced prediction accuracy reached 98.2%. Sensory models were built to assign samples into High. Medium, and Low categories based on all sensory attributes. (i.e., tenderness, juiciness, flavor, and overall liking). Tenderness class was predicted with 90.7% balanced prediction accuracy, and juiciness class was predicted with 88.9% balanced prediction accuracy. Additionally, Flavor class was predicted with 85.2% balanced prediction accuracy, and overall liking class was predicted with 91.7% balanced prediction accuracy. The current data suggests that with fine tuning REIMS generates a metabolic fingerprint that can meaningfully predict beef brisket composition and numerous beef quality attributes.