Background

In beef, once tenderness reaches an acceptable level, flavor is the major factor driving consumer acceptability. This revelation refutes past claims stating that tenderness was considered to be the most important factor for customer satisfaction. The beef industry has invested resources into numerous studies surrounding the complexity of beef flavor. Previous research demonstrated that type of feed, marbling score, fabrication methods, and packaging can all play key roles in determining product flavor perception by consumers. However, a definite understanding of beef flavor components and their variation has not been achieved yet. Now research has shifted to learn what factors affect beef flavor and how to assess beef flavor, both positive and negative flavors. The development of the beef lexicon in 2011 helped define and identify flavors, aromas, and textures that were present in whole beef muscle (Adhikari et al., 2011). When expert trained panelists utilized the beef lexicon, understanding of factors that drive consumer liking and consumer flavor liking have been identified. Factors such as cooking method, internal degree of doneness, quality grade, cut and shelf‐life effects on beef flavor has been identified. Consumer panels such as central location test and home use test have given us greater insight into consumer preferences and how beef is cooked and consumed at home. However, one factor has not been known, what are the flavor attributes of beef in the retail meat case and what is the variation in major beef cuts. As of today, American consumers are bombarded with multiple types of cuts in the retail meat case. There are at least 50 different beef cuts available to consumers. Additionally, some consumers are demanding information on how   beef is grown, handled, fed and they want to know information about the use of growth promoting strategies and antibiotic usage. This has resulted in label claims that are verified by USDA for some beef cuts. This study was designed to examine beef flavor in four major beef cuts presented to consumers in retail beef cases across the US (five major cities that represent different geographical locations were selected). Beef top loin steaks, top sirloin steaks, chuck roasts and 80% lean ground beef were identified as beef cuts with the highest purchase percentage. These cuts would then be used to create a baseline of understanding beef flavors present in these cuts. From these results, the beef industry will be able to identify positive and negative beef flavors, volatile aromatic compounds associated with these flavor attributes and drivers of consumer liking.

The objective of this research was to attempt to provide a benchmark of beef flavor components and its variability across four beef products that consumers typically purchase by sampling the U.S. retail beef case in five cities for beef arm center roasts, top loin steaks, top sirloin steaks (cap off), and 20% lipid ground beef to determine a baseline for beef flavor as measured by trained descriptive flavor and texture attributes, consumer liking attributes, volatile aromatic compounds, and REIM volatile analysis. The specific objectives were to 1) understand and identify the baseline flavor of top loin steaks, top sirloin steaks, 80/20 ground beef, and chuck roasts by using a trained descriptive panel and gas chromatography and 2) compare descriptive panel data to consumer data by Colorado State University and Texas Tech University. Compare gas chromatography and REIMS data by Texas A&M University, Colorado State University, and Texas Tech University.

Methodology

Top loin steaks, top sirloin steaks, 80/20 ground beef, and chuck roasts were selected at five different cities in the United States; Denver, CO, Los Angeles, CA, Miami, FL, and New York, New York. Retail stores included major retail food chains such as major food retails (i.e.., Kroger’s, Farmer Johns, Whole Foods) and volume retailers (i.e.., Costco and Sam’s Club) in each city. The objective of sample selection was to have beef that was representative of the beef that consumers would have available for selection and consumption within each city. From each city, beef center chuck roasts, beef strip loin steaks and 80/20 ground beef were selected so that approximately 150 packages of each of the three beef types are sampled per city. Of the 150 packages, within a retail store and beef type, beef packages will be selected so that there are 3 representative samples for each cut type, defined as an experimental unit. One package was then assigned to trained panel evaluation and two packages were assigned to consumer evaluation in the two locations. Beef was packaged and shipped to the Texas Tech University Meat Science Laboratory for segmentation of product within a city to testing type. Sampling of retail stores occurred over a 3‐month period. After arrival at Texas Tech University, beef was labeled, vacuum‐packaged and frozen in oxygen‐barrier bag. A total of 200 samples were purchased with 50 samples per cut.

Top loin steaks, top sirloin steaks, and 80/20 ground beef were cooked on a stove top grill set at 350°F˚.  Ground beef was weighed out to make three 150g patties and a patty press was used to ensure constant thickness. Chuck roasts were cut into a 4X5 inch rectangle using a grid that was placed in the center of the chuck roast. The roast section was placed on a rack in a pan with 2 cups of water. Chuck roasts were cooked in an oven set at 350˚F. All cuts reaching an internal temperature of 158°F˚. All samples were evaluated using a trained expert descriptive panel that was trained prior to testing. Panelists used a 16‐point scale with (0 = none to 15 = extremely intensive). Samples that were given to panelists were also used in order to capture volatiles for gas chromatography using the Aroma‐Trax® method at Texas Tech University and Texas A&M University. Consumer samples were prepared as defined for trained panel and 120 beef consumers evaluated eight beef cuts (two of each of the four cuts) for overall, overall flavor, beefy, grilled, juiciness and texture liking.

Flavor attributes rated by panelists, consumer liking attributes, and volatile compounds identified by the gas chromatograph were used to understand differences in flavor between four beef cuts from five cities. Multivariate statistics were to further understand what flavor attributes were present within treatments, what volatile compounds attributes to flavors, and what the drivers of liking for consumers were.

Findings

Consumer evaluations provided no definitive evidence of specific flavor identification with strip loins, sirloins, roasts, and 80/20 ground beef samples. Evidence from consumer comments, along with past studies, suggests numerous factors in pre‐harvest and post‐harvest interventions may play critical roles in determining meat flavor. However, the REIMS analysis demonstrated a meaningful relationship between chemical profiles of four cuts of beef even with some technical challenges. Analysis of REIMS data from the additional 40 samples validated these conclusions.

The four beef cuts differed from one another in descriptive flavor attributes. Results indicated that chuck roasts and top sirloin steaks tended to be associated with negative flavor attributes compared to ground beef and top loin steaks. Cooking method could have had an impact on what flavors were expressed in beef due to the fact that chuck roasts were closely associated with bloody/serumy flavor aromatics. As chuck roasts were cooked in the oven, they had the lowest levels of major positive beef flavor attributes of beef flavor identity, brown, and roasted flavor aromatics; and salt and umami basic taste. Steaks and ground beef patties cooked using the stove top method produced higher amount of Maillard Reaction products and had higher levels of major beef flavors. Sirloin steaks had the lowest fat‐like flavor and more intense levels of burnt and cardboardy flavor; and bitter and sour basic tastes. Top sirloin steaks were more highly associated with off‐flavors such as liver, cardboardy, and sour flavor aromatics. Ground beef and top loin steaks were closely related to more typical positive flavors found in cooked meat. Volatile compounds help to understand why certain flavors were expressed in certain cuts. Chuck roasts volatiles were primarily made up of compounds that came from lipid thermal degradation due to its moist cooking environment. Top loin steaks had volatiles present that were associated with the Maillard reaction due to the samples having direct contact with the heating surface. Ground beef volatiles were closely related to both lipid thermal degradation and Maillard reaction products. Ground beef samples had the highest frequency of green and green hay‐like flavors and volatile aromatics associated with these flavors. 

Implications

Beef flavor is the most important factor in consumer acceptability. Beef flavor is very complex and majority of factors that impact flavor come from how the consumer chooses to cook beef. Different cooking methods, degree of doneness, and cooking temperatures have been shown to affect beef flavor. Additionally, there is variation in flavor across muscle cuts, quality grades, and production factors. This study was implemented to better understand flavor attributes, volatile aromatic compounds in beef cuts sold in the US and to provide a baseline for what beef flavors are present in four major beef cuts. Positive and negative beef flavors were found and these flavors were tied to consumer liking and volatile aromatic compounds. The results of this study demonstrate the potential of REIMS to provide rapid classification of products based on variation in chemical profiles. Furthermore, it suggests that the pre‐ and post‐harvest interventions could have an impact on meat quality and eating characteristics. Although, none of the pre‐ and post‐harvest interventions were known in this study, it is likely that it influenced different flavor attributes identified by the consumer as the products were collected from different cities, indicative of the beef products available in retail stores.