Experimental treatments and sample preparation

Akaushi-cross cattle (n = 20) were fed standard finisher diets that either included a 27g/d GCE top dress (proprietary supplement; Mootral SA, Rolle, Switzerland) or did not (CON), and fed for either 9 or 12 mos (9MO; 12MO) at the West Texas A&M University Research Feedlot 12.87 km east of Canyon, TX. Because this was not an animal performance study, cattle of similar diets (GCE or CON) were fed in groups. The individual animal was considered the experimental unit, from which the subprimals for steaks and patties were derived. At the end of each feeding duration, cattle were harvested (n = 19) at the West Texas A&M University Meat Laboratory (Est. 7124) in Canyon, Texas. One animal (CON diet; 12MO duration-fed) was removed from the study before the harvest date due to illness. Carcasses were fabricated into strip loins from one side (left) of each carcass (Institutional Meat Purchase Specification (IMPS) #180; USDA, 2014; mean marbling score = 514 [USDA Moderate]; n = 19; 4–5 per treatment combinations), and clods and outside rounds (IMPS #114 and #171B, respectively; USDA, 2014) constituted ground patty samples (mean = 11.98% fat), a common commercial practice for formulating fresh ground beef as described by Gredell et al. (2018).

Strip loins were held at 2–4°C and aged 14 d before slicing (Treif, Model Lion F, Shelton, CT, USA) into six 2.54-cm-thick steaks. Steaks were fabricated and allocated for analysis in order from anterior to posterior (wedge = proximate and fatty acid analysis; followed by one steak for color analysis, the next for slice shear force (SSF) and volatile analysis, 2 steaks for consumer panel evaluation, and the last 2 steaks for descriptive trained sensory. Steaks were individually identified, vacuum packaged, and stored frozen (−20°C) until subsequent analysis, with the exception of the steak designated for color analysis. Vacuum packaged steaks from each strip loin used for color assessment were transported to Lubbock, TX, overwrapped (oxygen-permeable polyvinyl chloride fresh meat film; 15,500 to 16,275 cm³ O₂ m⁻² 24 h⁻¹ at 23°C, E-Z Wrap Crystal Clear Polyvinyl Chloride Wrapping Film, Koch Supplies, Kansas City, MO, USA), and placed in a retail display with continuous fluorescent lighting, conditions comparable with Dobbins et al. (2024), within 48 h. The most anterior “wedge” steak from the anterior end of the strip loin was used for proximate analysis. Steaks that were frozen until further analysis were transported to Texas Tech University and thawed for 24 h at 2–4°C for consumer, sensory, tenderness, proximate, fatty acid, and volatile analysis.

No subprimals (clods and outside rounds) were frozen before blending and grinding for ground patty formation. At day 15 postmortem, beef subprimals were trimmed of excessive fat and were cut into cubes less than or equal to 12.9 cm² for grinding, adapted from methods of Gredell et al. (2018). Each batch was ground and blended using a mixer-grinder (Thompson Meat Machinery, Model 840-M2-1, Crestmead, QLD, Australia) equipped with a coarse grinding plate (1.27 cm). Following mixing, batches were ground a second time using the same grinder equipped with a fine grinding plate (3.175 mm). Each batch was then formed into 151-g patties (Patty-O-Matic Inc., Model 330A, Farmingdale, NJ, USA). Each patty was individually identified, vacuum packaged, and stored frozen (−20°C) until subsequent analysis, except for the patties destined for color analysis. Fresh patties from each carcass used for color assessment were vacuum packaged, transported to Lubbock, TX, overwrapped, and placed in a retail display within 24 h. Frozen patties were transported to Texas Tech University for consumer, sensory, tenderness, proximate, fatty acid, and volatile analysis.

Instrumental retail color evaluation

One steak and one patty sample from each diet × duration combination designated for retail color display were removed from their vacuum package at Texas Tech University and individually placed on black expanded polystyrene trays and overwrapped with polyvinyl chloride film using a Minipack-torre, Minispenser (Dalmine, Italy). Steaks and patties were placed under continuous fluorescent lighting in 2 coffin-style cases (M1-GEA, Hussmann, Birdgeton, MO, USA) at temperatures of 2 to 4°C for 0–7 d (steaks) and 0–5 d (patties), respectively. Retail color evaluation parameters and perceptible color calculations were determined in reference to the Display Guidelines for Meat Color Research in the American Meat Science Association Guidelines for Meat Color Measurements (King et al., 2023). Samples were allowed 30 min of bloom time before day 0 measurement. Packages were randomly rotated between rows every 24 h within their original cases. The L*, a*, and b* values were determined at 3 locations per steak or patty and averaged every 24 h during display, using a Hunter MiniScan EZ 4500 (Hunter Associates Laboratory, Inc. Reston, VA, USA) using illuminant A with a 10° observer angle and a 45°/0° directional viewing geometry, 31.8-mm port, and 25-mm viewed area.

Trained color evaluation

Trained color evaluators, consisting of Texas Tech University Department of Animal and Food Sciences personnel, participated in evaluating samples for percentage discoloration utilizing the American Meat Science Association Guidelines for Meat Color Measurements (King et al., 2023). Panelists were screened based on normal acuity and ability to discriminate color differences using the Farnsworth-Munsell 100-Hue test. Successful panelists demonstrated scores of 50 or less as per the guidelines of AMSA (King et al., 2023). Panelists were further trained and tested on samples for approximately 14 h to objectively evaluate samples using AMSA’s scale for meat display color stability of whole muscle and surface discoloration as reference. Ballots were provided to panelists via electronic tablets (iPad, Apple Inc., Cupertino, CA, USA) using electronic surveys (Qualtrics, Provo, UT, USA). Color was evaluated when steaks were placed in the case (0 d = allowed 30 min bloom) and subsequently every 24 h for not more than 7 d. Panelists were asked to evaluate samples every 24 h (± 1 h) for the entirety of the display period. At each evaluation time point, 6 panelists evaluated the predominant lean color (at least 12.9 cm² of contiguous lean) of each steak for percentage lean discoloration, on an anchored, continuous 100-point line scale, with “0” representing no discoloration, and “100” representing completely discolored samples. Responses were averaged between panelists for each sample.

Proximate analysis

Steaks and patties for proximate analysis were thawed for 24 h at 2°C to 4°C. Steaks were trimmed of external fat and connective tissue, cut into 2.54 cm × 2.54 cm pieces, and ground through a 4-mm plate (#12 2/3 HP Electric Meat Grinder, Model: MG-204182-13, Gander Mountain, Saint Paul, MN, USA). Proximate analysis of raw steaks was conducted according to AOAC Official Method 2007.04 (Anderson, 2007) using a near-infrared spectrophotometer (FoodScan, FOSS NIR systems, Inc., Laurel, MD, USA). Percentages of collagen, fat, moisture, and protein were determined for each strip loin steak and patty.

Cooked sample preparation

Before cooking, samples were thawed at 2°C to 4°C for 24 h, and steaks were trimmed to remove external fat. Cooked sample preparation and respective analyses were completed in accordance with the AMSA Sensory Guidelines (AMSA, 2016). No cooking traits, beyond peak internal temperature, were obtained for analysis. Groups of steaks or patties (6 per sensory analysis or shear force analysis; groups of steaks and patties, cooked independently) were singly layered and cooked on a centrally located grill grate in a combi-oven (Model SCC WE 61 E; Rational, Landberg am Lech, Germany); individual steak/patty temperature was monitored during the cooking process using the built-in combi-oven thermocouple (3mm thick probe) placed in the geometric center of one centrally located sample. Samples were removed from the oven to target a peak internal temperature of 71°C for steaks and 77°C for patties. Immediately following cooking, samples were allowed to rest before recording a peak internal temperature using a calibrated type K thermocouple thermometer (AccuTuff 340, model 34040, Cooper-Atkins Corporation, Middlefield, CT, USA) placed in the geometric center of each sample. Before consumer, sensory, and tenderness evaluation, cooked steaks were trimmed of any remaining external fat and connective tissue.

Consumer panel evaluation

The West Texas A&M University institutional review board approved procedures for the use of human subjects for consumer sensory evaluation (#2022.04.019). Consumer panelists (n = 100) were recruited by email list and paid for their participation in Lubbock, Texas. Consumers were recruited based on the criteria that they were beef consumers at least once per week and were at least 18 years of age. Five panel sessions were conducted with 20 consumers seated in individual sensory booths/areas for approximately 1 h duration. After cooking, the panel steaks were then portioned into 1.27-cm × 1.27-cm pieces each, whereas patties were cut into 8 wedge-shaped pieces, and each of the portions was immediately served in a random order to consumers on a Styrofoam plate. Consumers were served 8 samples (4 steaks and 4 patties) immediately post-cooking in a predetermined, random order, approximately 6 min apart. Each consumer evaluated the steak samples first, then the patty samples, one from each treatment. Panelists were provided with a ballot, plastic fork and knife, toothpick, napkin, expectorant cup, cup of water, palate cleansers (unsalted crackers), and palate refresher (diluted sugar-free apple juice: 90% water and 10% sugar-free apple juice) to use between samples.

Each ballot was provided electronically on individual tablets (Qualtrics Surveys, Provo, UT, USA; iPad, Apple Inc., Cupertino, CA, USA) and contained an information sheet, demographic questionnaire, beef steak purchasing behavior sheet, and 8 sample ballots. Before the start of each panel, panelists were given verbal instructions about the ballot and the use of the palate cleansers/refreshers. No consumer was sitting next to another consumer sampling the same sample.

Attributes for all 8 samples were ranked on an electronic ballot with unstructured, numeric line scales (0–100) for juiciness, tenderness, flavor liking, off-flavor intensity, and overall liking. The “0” anchors were labelled as not juicy, not tender, dislike extremely, no off-flavor, and dislike extremely, whereas the “100” anchors were labelled as very juicy, very tender, like extremely, extreme off-flavor, and like extremely. Also, each consumer rated each sample as either acceptable or unacceptable (yes/no) for each palatability trait. Finally, consumers were asked to designate each sample as unsatisfactory, everyday quality, better than everyday quality, or premium quality.

Descriptive trained sensory analysis

Sensory analysis was conducted at Texas Tech University. Panelists were trained via methods described by Ponce et al. (2019) to detect multiple beef flavor characteristics using the lexicon developed by Adhikari et al. (2011) and objectively quantify the intensity of each attribute using an unstructured 100-point line scale. Samples designated for sensory analysis were randomly assigned to sensory sessions so that all treatments were represented in each panel. One panel session was conducted each day with 4 samples per session for a total of 10 panel sessions (5 steak sessions and 5 patty sessions). One sample representing each of the 4 treatments per panel session was served to 6 panelists. Before statistical analysis, panelist responses for each steak or patty were averaged.

Cooked steaks were trimmed of any remaining external fat and connective tissue, then portioned into six 1.27-cm × 1.27-cm × 2.54-cm cube-pieces each, whereas patties were cut into 6 wedge-shaped, equally sized portions and held at 50–55°C in a warming box (Cambro Manufacturing, Huntington Beach, CA, USA) for no more than 15 min before being served to panelists. Samples were served under red incandescent light to mask color variation among samples during 30-min panel sessions. Panelists were supplied with distilled water, unsalted saltine crackers, and apple juice to cleanse and refresh their palates between samples. According to the beef flavor lexicon, panelists evaluated each sample for tenderness, juiciness, beef flavor identity, browned, roasted, fat-like, buttery, umami, metallic, oxidized, liver-like, sour, and bitter on an unstructured 100-point line scale (0 = not present; 100 = very intense). Additionally, panelists collaborated and defined a garlic intensity attribute (0 = not present; 100 = roasted garlic). Upon discovery of novel sensory attribute (garlic intensity; not previously defined in Adhikari et al., 2011), group collaboration among trained panelists, a single panel value was produced. For all other traits, after each panel session, individual panelist ratings were averaged to obtain a single panel rating for each sensory attribute of each sample.

SSF

Samples designated for SSF were thawed and cooked as previously described. Tenderness was evaluated by SSF as described by Shackelford et al. (1999). In brief, following the endpoint temperature reading, a 1–2 cm slice was removed across the width of the steak from the lateral end to square off the steak and expose the muscle fibers. Using a cutting guide, a 5-cm long × 1-cm thick section was obtained from the lateral end by cutting at a 45° angle, parallel to the muscle fiber orientation. The sample was center sheared perpendicular to the muscle fiber using a G-R Shear Machine (Model GR-152 [Slice Shear Machine], G-R Electric Manufacturing Company LLC, Manhattan, KS, USA) equipped with a load cell of 50 kg operating at a cross-head speed of 500 mm/min.

Fatty acid analysis

For the determination of fatty acids, total lipid was extracted using the methods of O’Fallon (2007) from 1 g of homogenized raw sample. Samples were flash frozen using liquid nitrogen and transferred to a blender (NutriBullet LEAN, Pacoima, CA, USA) to produce a fine-ground, homogenous powder. Samples were stored in an individual bag at −80°C until further analysis. Vials were stored at −80°C until fatty acid methyl extraction (FAME) separation and quantification with gas chromatography (GC). From an injection of 1 μL of sample at a 50:1 split ratio with hydrogen as a carrier gas (2.4 mL/min), FAME were separated on an HP-88 capillary column (100 m × 250 μm × 0.2 μm; Agilent Technologies) using a GC system equipped with a flame ionization detector (Agilent Technologies 7890A). For each sample, oven temperature remained 35°C for 2 min, then increased 4°C/min to 170°C and remained at 170°C for 4 min, and then increased 3.5°C/min and remained at 240°C for 2 min. The injector and detector were operated at 250°C. Fatty acids were identified based on similarity of their retention times with reference standards (Nu-Chek Prep, GLC 463). Fatty acid concentrations were calculated relative to initial wet sample weight (mg/g) and expressed as a percentage of the total lipid fraction in the sample. The total percentage of saturated (SFA), monounsaturated (MUFA), and polyunsaturated fatty acids (PUFA) was calculated for each sample.

Volatile compound analysis

Volatile compounds were measured similarly to Gardner and Legako (2018) with modifications described in Hernandez (2022). One steak and patty from each treatment was thawed and cooked according to the methods previously described. Five grams of cooked, homogenized (as previously described), frozen sample were weighed into 20-mL glass vials. An internal standard solution (1,2-dichlorobenzene, 2.5 μg/μL) was dispensed in each sample. Vials were sealed and loaded into a −20°C dry air-cooling block (MeCour Temperature Control, LLC, Groveland, MA, USA). A Gerstel autosampler (Multipurpose Sampler, Gerstel, Inc.) removed samples from the cooling block and placed them in a 65°C agitator for extraction. An 85 μm film thickness carboxen/polydimethylsiloxane fiber (Supelco Inc., Bellefonte, PA, USA) was exposed in the headspace of the vial for solid phase microextraction (SPME) of volatile compounds. Separation and detection of volatile flavor compounds were carried out using gas chromatography-mass spectrometry (GC-MS). After extraction, the SPME fiber was injected into the GC (7890B series, Agilent, Santa Clara, CA, USA) and desorbed onto a VF-5ms capillary column (30 m × 0.25 mm × 1 μm; Agilent J&W GC columns, the Netherlands). Separated compounds were introduced to the single quadrupole mass spectrometer (5977A, Agilent, Santa Clara, CA, USA) through electron ionization at 70 eV. Mass range was determined at 45–500 m/z. Authentic standards (Sigma-Aldrich, St. Louis, MO, USA) were used to confirm compound identities through retention time and fragmentation pattern of 1 target ion and 2 qualifying ions. A calibration curve and the internal standard were used for absolute quantitation of volatile compounds (ng per g of sample).

Statistical analyses

Statistical analyses were conducted using the procedures of SAS (version 9.4; SAS Institute Inc., Cary, NC, USA) using α = 0.05. Treatment comparisons were tested for significance using the GLIMMIX procedure. Fixed effects included diet (GCE or CON) and duration-fed (9MO or 12MO), utilizing a 2 × 2 factorial arrangement (GCE-9; GCE-12; CON-9; CON-12). Random effects included the animal, peak temperature in cooked sample analysis, and panel order designation for consumer and descriptive sensory. Proximate fat percentages were included in the model as a covariate, but only accounted for minimal variation and were therefore removed. Color data (instrumental and subjective) were analyzed as repeated measures using a compound symmetry covariance structure, and day of display was included as a fixed effect. Sensory panels were fed in a balanced complete block design with panel (5 consumers) as the block. For all analyses, the Kenward-Roger approximation was used for estimating denominator degrees of freedom, and the PDIFF option was used to separate treatment means when the F-test on the main effect or effect interaction was significant (P ≤ 0.05).

Instrumental and panelist retail color assessment

Instrumental and panelist discoloration assessments of uncooked ground beef patties are reported in Table 1. No diet × day interaction occurred (P ≥ 0.106). For all other ground beef patty color assessments, a significant day effect occurred (P ≤ 0.042) for all attributes, except for 9MO-patty L* values (P = 0.169). Values for 9MO-patties, a* and b* declined (P ≤ 0.011), whereas discoloration percentage increased (P < 0.001) over time for 9MO-patties. Notably, upon transportation, 9MO-patties produced considerable discoloration (mean = 33.36% discolored), abnormal for day-0 display. Regarding 12MO-patties, all instrumental color values declined (P ≤ 0.042) over time, whereas discoloration percentages increased (P < 0.001).

Instrumental and panelist discoloration assessments of uncooked steaks are reported in Table 2. Interactions among the steak 9MO-diet treatments by day included instrumental L* values (P = 0.008) in which, generally, GCE-9 steaks sustained higher brightness (L*) values through day 5 of display when compared with CON-9 steaks. Moreover, 9MO steaks declined in instrumental redness (a*) values over time (P < 0.001) and differed (P = 0.011) among diets. Within 9MO steaks, redness values were higher (P = 0.011) for GCE-9 than CON-9 steaks.

A day effect (P ≤ 0.027) occurred for all remaining color assessments for both 9MO and 12MO steaks. Both a* and b* values declined (P < 0.001) for 9MO steaks d 1 through the duration of retail display. Percentage discoloration for 9MO and 12MO steaks increased (P ≤ 0.010), most notably by days 5 and 7, respectively, and did not differ (P ≥ 0.173) from days 0 to 3 and days 0 to 5, respectively. Diet effects occurred for both a* and b* values of 9MO steaks, suggesting GCE steaks were redder and more yellow via instrumental values than CON steaks. All other remaining day effects were not significant (P ≥ 0.515).

Choi et al. (2010) reported the impact of garlic powder inclusion in the diets of broilers; as garlic level increased, broiler thighs were lighter (L*), redder (a*), and more yellow (b*) in color. This similarly aligns with the findings of the current study among diet treatments for 9MO steaks, though the remainder of the observations (12MO and patties) dissimilarly suggest no difference in color attributes among diet treatments. Similar to the L* observations of the current study, Janz et al. (2007) also reported no instrumental lightness differences among pork chops from pigs fed garlic versus a control diet. The antioxidant properties promoted by the inclusion of garlic or a garlic supplement in animal diets likely influence muscle metabolism, and consequently, affect the inherent oxidative stability and antioxidant capacity of fresh meat during retail display (Janz et al., 2007; Choi et al., 2010).

Objective tenderness and proximate analysis

Table 3 includes the means for SSF of steaks and proximate analyses of both steaks and patties. No difference (P ≥ 0.303) existed for SSF or proximate analysis of steaks and patties among diet treatments, suggesting that feeding a GCE supplement will not impact tenderness or product composition. Overall, steaks qualified as Certified Very Tender based on the SSF values (less than 15.4 kg) indicated by ASTM International (2011). Mean SSF and percentage collagen of steaks, as well as percentage protein of patties, were not affected (P ≥ 0.196) by duration of feeding. A considerably higher (P ≤ 0.034) percentage of fat (20.88% versus 9.40%) and a lower (P ≤ 0.024) percentage of moisture were recorded for 9MO steaks and patties than 12MO samples. The percentage of collagen was detected at higher (P = 0.032) values for 9MO patties than 12MO patties.

Consumer evaluation & acceptability

Consumer demographics and consumption behavior information from the 100 consumers are reported in Table 4. Diet and duration-fed interactions, as well as main effect means for consumer rating, acceptability, and perceived quality levels, are reported in Tables 5 and 6 for patties and steaks, respectively. Consumers rated GCE-12 patties higher (12.68; P = 0.020) for off-flavor intensity than GCE-9 (5.93); whereas CON patties (both CON-9 and CON-12) remained intermediate for off-flavor ratings. This suggests, the longer cattle were fed with the GCE supplement, the greater the perceivable off-flavor. This similarly aligns with the findings of a study conducted on dairy cows, in which researchers detected milk and cheese produced from cows supplemented with either diallyl sulfide (common volatile in garlic) or whole cloves of garlic had stronger (P < 0.001) garlic-like aroma and flavor, compared to milk and cheese produced from cows on a commercial diet (Rossi et al., 2017).

Moreover, consumers deemed CON-9 patties the highest percentage (P = 0.017) as a premium quality level (17.38%), when compared to all other treatment combinations (≤ 7.75%). No other interactions (P ≥ 0.189) were detected from consumer evaluations.

Juiciness, tenderness, off-flavor intensity, and flavor liking of strip loin steaks were not impacted by diet or duration-fed (P ≥ 0.188). Analysis of consumer evaluations suggests diet and duration-fed impacted consumer ratings, perceived quality levels, and the percentage of steaks and patties rated acceptable. Steaks from GCE-fed cattle rated lower (P = 0.054) for overall liking than CON steaks. In a study evaluating pork loin chops, Janz et al. (2007) collected samples derived from pigs fed oils/oleoresins of common herbs and spices (including garlic) and fed them to unbiased consumers to evaluate detectable flavor/aroma differences. Similar to the current study, consumers were unable to distinguish treated whole muscle pork samples from the control using flavor as the evaluation factor (Janz et al., 2007). Moreover, patties were not impacted by diet or duration-fed (P ≥ 0.230) for juiciness, tenderness, flavor liking, and overall liking, except 9MO patties depicted higher tenderness ratings (P = 0.039) than 12MO patties.

Percentages of acceptability for consumer evaluations (juiciness, tenderness, flavor, and overall liking), as well as perceived quality level percentages (unsatisfactory, everyday quality, better than everyday quality, and premium quality) were rated similarly (P ≥ 0.209) among diet and duration-fed treatments for steaks. Consumers regarded 9MO patties at a higher percentage of tenderness acceptability (91.54%; P = 0.032) than 12MO patties (84.47%). All other percentages of consumer acceptability and perceived quality levels of patties for diet and duration-fed evaluations were rated similarly (P ≥ 0.101). In general, the majority of steaks and patties were rated as everyday quality or better.

Descriptive trained panels

Table 7 contains the interaction and main effect means of descriptive trained panelist attributes affected by diet and duration-fed for steaks and patties. Descriptive trained panelists’ ratings depicted an interaction (P = 0.028) among patties, rating CON-12 patties higher for the oxidized attribute than CON-9 patties, whereas GCE treatments were rated intermediately.

Of note, a key ingredient of the manufacturer’s GCE supplement encompasses garlic substrates; therefore, garlic intensity was measured. Odor has been noted as a key factor influencing food flavor (Shankaranarayana et al., 1975). Therefore, upon olfactory recognition of garlic in cooked steaks and patties, panelists evaluated samples for garlic intensity. Panelists determined an effect (P = 0.013) among steaks, and a tendency (P = 0.064) among patties, in which GCE samples rated higher for garlic intensity when compared to CON samples. No difference (P ≥ 0.335) for garlic intensity existed between duration-fed treatments. Similarly, Leong et al. (2010) fed swine increasing levels of garlic oil in an attempt to mask other undesirable flavors or aromas. Researchers reported that as garlic oil increased in the swine’s diets, both garlic aroma and flavor were detected and intensified (Leong et al., 2010). Amongst other ruminant studies, Fraser et al. (2007) supplemented sheep with levels of garlic powder to quantify residuals within tissues, in addition to sensory effects. While successfully masking other natural off-flavors within the sheep meat, researchers reported an obvious detection of garlic flavor within meat samples. Garlic intensity was especially noted within subcutaneous fat portions, further producing a less desirable flavor profile (Fraser et al., 2007).

Panelists detected a stronger umami flavor in GCE steaks (P = 0.001) and 9MO steaks (P = 0.005) when compared with CON and 12MO steaks, respectively. Presumably, as similarly reported by Corbin et al. (2015) and Motoyama et al. (2016) referencing beef with high percentages of intramuscular fat, some umami difference may correspond with the higher fat percentage among 9MO steaks (20.88% fat) versus the 12MO steaks (9.40% fat). Previous literature indicates that as intramuscular fat percentage increases, umami flavor also increases (Corbin et al., 2015; Motoyama et al., 2016). Ratings for browned and oxidized attributes were also rated higher (P ≤ 0.035) for 9MO steaks than 12MO steaks. Regarding patties, panelists rated GCE patties lower for browned (P = 0.018) and higher for bitter (P = 0.020) attributes, when compared to CON patties. Though statistically different, most ratings were numerically similar and would likely go undetected by the average consumer, the exception being garlic intensity. All other descriptive panel attributes of steaks and patties among varying diets and durations fed did not differ (P ≥ 0.069). Among duration-fed treatments, fat percentage differences (P ≤ 0.034) produced higher percentages in 9MO versus 12MO steaks, though surprisingly few descriptive sensory differences occurred. Corbin et al. (2015) also noted, once tenderness acceptability is met, the palatable expectation then relies on flavor. As flavor has previously been attributed to fat type and content, Corbin et al. (2015) reported the assessment of numerous quality levels of beef steaks, where few descriptive sensory attributes differed among quality levels once exceeding approximately 9% chemical fat, much like what was observed in the current study.

Fatty acid profile

Interaction and main effect means for the percentage of fatty acids are located in Tables 8 and 9 for steaks and patties, respectively. An interaction (P = 0.002) among treatment combinations of steaks was detected for total percentage of PUFA, suggesting CON-12 steaks have a higher percentage (7.36%) of PUFA than all other treatment combinations (≤ 4.81%). Among patties, the percentage of α-linolenic (C18:3 n-3) was reported at a higher (P = 0.039) percentage rate for GCE-9 patties than the remaining treatments. Eicosadienoic (C20:2) percentage within CON-12 patties was detected higher than both GCE treatment combinations, whereas CON-9 eicosadienoic percentages were determined intermediate.

All fatty acid percentages were determined similarly (P ≥ 0.107) for steaks among diet treatments, with the exception of eicosadienoic (C20:2; P < 0.001), which was detected at a higher percentage for CON steaks than GCE steaks, which mimics the interaction found in patties. Among duration-fed effects, pentadecylic (C15:0), palmitic (C16:0), margaric (C17:0), stearic (C18:0), myristoleic (C14:1 n-5), palmitoleic (C16:1 n-7), oleic (C18:2 n-6), and eicosadienoic (C20:2) were detected at higher (P ≤ 0.006) percentages for 9MO steaks than 12MO steaks, and tended (P ≤ 0.068) to contain higher percentages of myristic (14:0), heptadecenoic (C17:1), and α-linoleic (C18:3 n-3).

Total percentage of PUFA was detected at higher (P = 0.047) rates for GCE patties, and tended (P = 0.100) to contain a lower percentage of total MUFA, when compared to CON patties. Among duration-fed effects, 9MO patties contained a higher (P = 0.028) percentage of myristoleic (C14:1 n-5) and a lower (P = 0.005) total percentage of PUFA. All other fatty acid percentages among treatments for patties did not differ (P ≥ 0.081).

In summary, similar to observations among steaks, but in contrast to patty results, unsaturated fatty acid percentages were detected at higher levels in meat from sheep fed garlic substrates (Redoy et al., 2020). Additionally, Redoy et al. (2020) reported a decrease in SFA content, whereas no difference (P ≥ 0.124) occurred in the present study, regardless of sample type. Chu et al. (2003), on the other hand, detected higher levels of linoleic and linolenic acids (C18:2 n-6 and C18:3 n-3) in steaks from steers supplemented with garlic components.

Volatile compound analysis

Volatile compounds of steaks and patties affected by diets and duration fed are located in Tables 10 and 11. With the exception of ethanol (P ≥ 0.058) concentration, no differences existed among volatile compounds of steaks and patties, regardless of treatment (P ≥ 0.107). Controversially, ethanol concentration was lower (P = 0.038) for GCE steaks and higher (P = 0.044) for GCE patties when compared to the CON samples. Additionally, 12MO steaks tended (P = 0.058) to present a higher ethanol concentration than 9MO steaks. Shankaranarayana et al. (1975) in Volatile Sulfur Compounds in Food Flavors specifically listed garlic among foods attributed to sulfur-containing volatile compounds, including diallyl sulfide, dimethyl sulfide, and dimethyl disulfide, as well as methanethiol (Shankaranarayana et al., 1975). Interestingly, sulfur-containing compounds were not different (P ≥ 0.107), but all were detected numerically higher in GCE patties compared to CON patties, though the relation was reversed within steak samples. Flavor perception of a specific aroma or note is comprised of 100s to 1,000s of volatile compounds. The current method is limited to 5 sulfur-containing compounds; therefore, it is likely that many of the volatile compounds associated with garlic flavor were not captured. Moreover, sensory perception of garlic is also driven by various non-volatile components (Liu et al., 2019) that were not accounted for.