A G R I C U LT U R A L A N D F O O D S C I E N C E Agricultural and Food Science (2022) 31: 229–240 229 https://doi.org/10.23986/afsci.121501 Variations of early postmortem pH in carcasses of grass-fed steers and its relationship with glycolytic potential and meat quality traits Nancy Jerez-Timaure1, Gaspar Brinckmann2, Maureen Berkhoff1, Alfredo Ramírez-Reveco1 and Pablo Strobel1 1 Instituto de Ciencia Animal, Facultad de Ciencias Veterinarias, Universidad Austral de Chile, Valdivia, 5090000, Chile 2Escuela de Graduados, Facultad de Ciencias Veterinarias, Universidad Austral de Chile, Valdivia, 5090000, Chile e-mail: nancy.jerez@uach.cl This study aimed to investigate the multivariate relationship between the glycolytic potential, and meat quality traits from grass-fed steers carcasses with variations in early postmortem pH. From a contemporary group of steers (n=70) from the same production unit and slaughtered under similar conditions, thirty carcasses (10/group) were selected based on pH values measured at 3h (pH 3h ): Low (<6.2), Intermediate pH 3h (6.2–6.5), and High (>6.5). Car- casses segregated by pH 3h groups were different (p < 0.05) in muscular glycogen (MCG), glucose (G+G6P) content, glycolytic potential (GP) and GP strict (GP without lactate content). The interaction pH 3h groups × sampling time was significant only for lactate content (LC) (p < 0.05). Quality traits, except redness did not vary (p > 0.05) among pH 3h groups. Color variables had a positive and moderate correlation with MGC, G+G6P, LC, and GP. Results does not allow to recommend early carcass segregation by very early pH (3h postmortem); however, important bivariate and multivariate relationships between G+G6P, lactate content and instrumental color parameters in LL muscle from grass-fed cattle were demonstrated. Key words: beef, color, tenderness, lactate, glycogen, glucose Introduction The rate and extent of pH fall significantly influences meat quality traits, including color development, shelf life, water holding capacity, and tenderness (Young et al. 2004, Wu et al. 2014, Wang et al. 2021). These parameters are important for consumers and the meat processing industry. Postmortem glycolytic potential allows to measure the potential lactate formation and it has been recognized to be highly correlated with ultimate pH (Wulf et al. 2002, England et al. 2016). The pH decline is correlated with the enzymatic degradation of glycogen to lactate and this process includes the metabolic coupling between glycogen- olysis and glycolysis, and that are strongly upregulated by the AMPK (Strobel et al. 2021). The activities of various enzymes involved in glycogen synthesis and depletion play an important role in the rate and extent of glycolysis (Apaoblaza et al. 2017). Normal concentrations of muscle glycogen immediately at the time of slaughter and an adequate cooling managing of carcasses are relevant to let an efficient glycogenolytic/glycolytic flow required for lactate accumulation and pH decline (Strobel et al. 2021). Also, final meat quality is affected by the course of tem- perature in the postmortem period (Hamoen et al. 2013), specially at the onset of rigor (Hannula and Puolanne 2004). Immonen et al. (2000) studying the relationship between residual carbohydrate and ultimate pH, found that at low pH values (< 5.8), glycogen and glucose concentrations varied considerably, from 83 to 10 mmol kg-1. Skeletal muscle is composed of different types of fibers that affect the rate of energy metabolism in particular glycolysis (Hopkins et al. 2014). Each muscle fiber type has different glycolytic and oxidative capacities, contrac- tion speed, fiber size, myoglobin, and glycogen content (Schiaffino and Reggiani 1994). Muscle fiber type I has slow-twitch, oxidative metabolic characteristics, and a low glycogen content. Type IIA is a fast oxidative–glycolytic fiber. On the other hand, type IIB has fast-twitch, glycolytic metabolic characteristics, and a high glycogen content (Schiaffino and Reggiani 1994, Karlsson et al. 1999). Type I (fast-twitch) fibers have low ATPase activity and are more prone to inactivation by a rapid pH decline (Bowker et al. 2004). Many researchers have investigated the potential of using pH as a predictor variable for meat quality. Jones and Tatum (1994) studied the relationship between tenderness and the muscular pH measured at 3-hour postmortem (pH 3 ), finding that pH 3 values lower than 6.2 were strongly associated with shear force or tenderness. Smulders et al. (1990) also reported that pH 3h values below 6.3 were associated with a lower incidence of undesirable ratings for tenderness. Lomiwes et al. (2014) reported and important relationship between meat tenderness and ulti- mate pH (pH u ); these authors found that meat with high (≥ 6.2) and low (≤ 5.79) pH u were less tender after ageing. Received 31 August 2022 / Accepted 9 December 2022 The Scientific Agricultural Society of Finland ©This is an open access article under the CC BY 4.012 N. Jerez-Timaure et al. 230 Wu et al. (2014) studied the effect of pH u in beef on the proteolysis of large structural proteins related to tender- ness and contributed with evidence of the strong relationship between pH u and meat quality. Hamoen et al. (2013) compared experimental data on pH decline in carcasses to develop a prediction model for ultimate pH and meat quality; and Young et al. (2004) patented a method for early determination of meat ultimate pH. More recently, Shulte et al. (2022) reported a proteomic and metabolome study in beef longissimus thoracis muscles from carcasses classified by pH values at 6 h postmortem, resulting in apoptotic, calcium homeostasis, and proteasome systems influencing pH classifications that could explain the observed pH, proteolysis, and beef tenderness differences. Few studies (Smulders et al. 1990, Jones and Tatum 1994) had included the association between very early postmortem (3 h postmortem) pH and meat quality traits; however, none of these had evaluated the glycolytic potential during early postmortem. The objectives of this study were to investigate the multivariate relationship between glycolytic potential and meat quality traits from grass-fed steers carcasses with variations in early post- mortem pH (pH 3h ). Materials and methods Animals, slaughter, carcass handling, and sampling From a contemporary group (age approximately 24 months) of Angus and Friesian crossbred steers (n=70) raised on the same farm and fattened system (pasture, and supplementation with silage and minerals). Animals were transported in a single group in the same truck to the slaughterhouse. Total lairage time was 16 h. Animals had free access to drinking water, except during transportation. Animals were not mixed before or after transport. They were slaughtered on the same day, according to national regulations (INN 2002) (slaughter line speed is ap- proximately 50 cattle/h). All carcasses were kept in the same cold chamber (temperature ranged between +0.4 to 2.8 °C) during 48 h postmortem. Carcass average weights were 275.4 ± 21.4 kg. The muscular pH was measured with a using a pH meter (MPI pH meter, Meat probes Inc. USA) (at the level of the 12th rib) at 3 h (pH 3h ) and a sample of approximately 1 g were taken from the same location and time using a Bergstrom biopsy needle. Samples were placed in Eppendorf tubes, identified, and immediately frozen in liquid nitrogen (–196 °C) and the maintained in ultra-freezer (–80 °C). Carcasses were segregated according to Jones and Tatum (1994) as follows: Low pH 3h (<6.29), Intermediate pH 3h (6.3 to 6.49), and High pH 3h (>6.5). A total of 10 carcasses per each pH 3h -group were selected for the study and muscular pH and sampling were accomplished (for selected carcasses) at 6, and 24 h postmortem. At 24 h post- mortem, the entire LL muscle was removed from each left carcass side. Steaks of 2.5 cm-thick of each muscle were taken, and vacuum packed individually, one steak was used immediately to evaluate color and drip loss and two steaks were frozen at –20 °C for 30 days for further analysis. Glycogen, glucose, lactate contents and glycolytic potential Muscular glycogen (MGC), glucose+glucose-6 phosphate (G+G6P), and lactate contents (LC) were determined as described in Apaoblaza et al. (2015). Briefly, muscle samples (100 mg) were homogenized in ice-cold phosphate buffer (pH 7). Ten µl of homogenate was hydrolyzed in 200 µl of 0.1M HCl at 100 °C for 2 h, after which pH was adjusted to 6.5–7.5. MCG and glucose (G+G6P) were determined via NADP reduction with a linked assay involving hexokinase and glucose-6-phosphate dehydrogenase (Glucose HK 16-50 Sigma). The LC was determined from the same homogenate via NAD reduction with a linked assay involving lactate dehydrogenase and glutamate pyruvate transaminase (Boehringer Mannheim). Glycolytic potential (GP) was calculated with the following formula GP = (LC) + 2([MGC] + [G+G6P]) (Monin and Sellier 1985). Also, the GP was calculated using the same equation exclud- ing the LC (called GP strict ). Values were expressed as mmol kg-1 of muscle. Meat quality traits Instrumental color and drip loss (DL) were measured in fresh samples immediately (after 24 h postmortem). DL was measured according to Honikel (1998), briefly, samples ranging 80 and 100 g were used, each portion was identified and separately placed in an air-tight plastic bag and stored at a refrigeration temperature of 2 °C for 48 h; after which, the weight of the meat sample was recorded. The DL was expressed in the percentage of water re- leased from the original sample. A Hunter Lab Mini Scan XE Plus (Hunter Associates, Reston, VA) was used with a 2.5-cm open port, Illuminant D65, and 10° standard observer to evaluate color objectively. Three readings will be Agricultural and Food Science (2022) 31: 229–240 231 obtained from the muscle surface and the mean will be calculated. Readings were obtained after exposing mus- cles to air for 30 min (bloom). The Hunter L, a, and b values were taken as indicators of lightness, redness, and yellowness, respectively. For Warner-Bratzler Shear force (WBSF) and cooking loss (CL) evaluation, steaks were cooked in a convection oven (Albin Trotter model E-EMB Digital) to a final internal temperature of 70 °C following the guidelines of the American Meat Science Association (AMSA 2015). The temperature was monitored using an oven thermometer (–10 – +110 °C ± 1 °C) inserted into the geometric center of each steak. The cooked steaks were chilled for 2 h at 2 °C, and then between eight to ten cores (1.27 cm in diameter) were removed parallel to the muscle fiber orientation. Cores were sheared once each on the Warner-Bratzler Meat Shear apparatus (GR Manufacturing Co., Manhattan, USA) to get WBSF values. Cooking loss (CL) was determined by weight loss during exposure to high temperature expressed as percentage compared to the original weight of the sample. Statistical analysis Analysis of variance was used with a linear mixed model consisting of pH 3h groups as a fixed effect, sampling time as a repeated measure, and fat thickness as a covariate. The statistical significance threshold will be set at p < 0.05. Values were expressed as the mean ± standard error (SE). Tukey-Kramer multiple comparison test was used for the comparison of means at a 5% level of significance. The Bonferroni correction was also performed to adjust the probability of p values. To analyze the relationship among the subgroup of the variables: meat quality traits with respect to the subgroups of glycolytic components traits and pH, a canonical correlation analysis (CCA) was carried out with meat quality data and glycolytic related data measured at 24 h postmortem. Wilk´s Lambda and Bartlett tests were used to de- termine the significance of canonical correlations. Additionally, bivariate correlations analyses were performed among all traits evaluated at 3, 6, and 24 h postmortem. Results Glycolytic potential and its components in LL muscles from carcasses grouped by pH 3h The interaction pH 3h group × sampling time was significantly different for LC (p < 0.05). The interaction pH 3h × sam- pling time was not significant (p > 0.05) for the rest of the variable, therefore means were presented by main fac- tors (pH 3h group and sampling time) for MCG, G+G6P, GP and GP strict . Carcasses that were segregated by pH 3h as Low (< 6.29), Intermediate (6.3 to 6.49), and High (> 6.5) were different in the MCG (p = 0.03), and G+G6P (p = 0.01). Carcasses classified as Intermediate pH 3h had less MGC and G+G6P than those classified as High or Low pH 3h ; consequently, Intermediate pH 3h carcasses also showed the lowest value of Gpot and GP strict (p = 0.02; Table 1). At 3 h postmortem, carcasses from the Intermediate and High groups exhibited the lowest content of lactate com- pared to carcasses from the Low pH 3h group (Table 2; p < 0.05). A similar trend was observed at 24 h, but without being statistically different (p > 0.05). The LC increased with time, resulting in exhibiting the highest values of LC, the carcasses categorized as Low pH 3h . Sampling times were statistically different (p < 0.05) for all traits except G+G6P. The MGC, and GP strict significantly decreased with time (p < 0.05), and LC increased (p = 0.01) with time (Table 3). On the other hand, G+G6P and GP were not affected by sampling time (p > 0.05). MGC = Muscular glycogen content; G+G6P = Glucose + Glucose 6-phosphate; LC = Lactate content; GP = Glycolytic Potential. GP strict = GP without LC; * mmoles kg-1; n = 30 Table 1. Means and standard errors of the glycolytic potential and its components by pH 3h groups Variables* pH 3h groups p-value High (>6.5) Intermediate (6.2 to 6.49) Low (< 6.2) MGC 12.32 ± 1.37 b 5.97 ± 1.44 a 11.77 ± 1.14 b 0.03 G+G6P 4.93 ± 0.52 ab 3.69 ± 0.54 a 5.86 ± 0.58 b 0.01 LC 38.34 ± 1.76 38.34 ± 1.42 41.82 ± 2.05 0.20 GP 72.84 ± 6.02 ab 57.67 ± 4.70 a 77.08 ± 4.07 b 0.02 GP strict 34.50 ± 4.38 b 19.32 ± 3.64 a 35.27 ± 3.78 0.02 N. Jerez-Timaure et al. 232 Figure 1 shows the dynamic of decreasing pH values for Low (< 6.2), Intermediate (6.2–6.5), and High (> 6.5) pH3h from 3 h to 24 h postmortem. The pH3h groups were significantly (p < 0.05) different at 3 h and 6 h. The highest ultimate pH values were consistently from the High pH3h group, also noticeable that the drastic decline occurred between 3 and 6 h postmortem. The highest decreased variation of pH from 24 to 3 h was observed in the High pH3h carcasses (–10.41%) compared to the Intermediate pH3h (–6.69%) and the Low pH3h ones (–5.58%), respectively. Effects of early pH on meat quality traits Means for WBSF, CL, DL, pH, L (lightness), b (yellowness) evaluated after 24 h postmortem were not statistically different (p > 0.05) when comparing pH 3h groups (Table 4). However, early pH affected (p = 0.01) the redness (a values); meat samples from carcasses classified as Intermediate pH 3h resulted in fewer redness intensities than the other pH 3h groups. Table 2. Effect of the interaction pH 3h groups × sampling time on lactate content Sampling time pH 3h groups High (>6.5) Intermediate (6.2 to 6.49) Low (< 6.2) 3 h 31.22 ± 0.86 a/d 33.87 ± 1.15 a/d 40.55 ± 1.01 b/d 6 h 42.24 ± 0.85 a/e 39.53± 0.94 a/e 42.01 ± 0.95 a/de 24 h 41.54 ± 0.98 a/e 39.88 ± 0.05 ab/e 45.34 ± 0.09 a/e a,b,c: different letters within the same row indicate p < 0.05; d,e,f : different letters within the same column indicate p < 0.05. Values are presented as means ± SE in mmoles kg-1. n = 10/group MGC = Muscular glycogen conten; G+G6P = Glucose + Glucose 6-phosphate; LC = Lactate content; GP = Glycolytic Potential; GP strict: = GP without LC; n = 30/sampling time. * expressed as mmoles kg-1 Table 3. Means and SE of the glycolytic potential and its components by sampling time Variables* Sampling time p-value 3 h 6 h 24 h MGC 15.00 ± 1.19 b 9.17 ± 1.47 ab 5.61 ± 1.47 a 0.001 G+G6P 4.94 ± 0.56 4.46 ± 0.49 5.13 ± 0.63 0.85 LC 35.21 ± 1.38 a 41.21 ± 1.87 b 42.26 ± 1.89 b 0.01 GP 75.11± 4.85 69.54 ± 4.48 63.73 ± 5.87 0.48 GP strict 39.90 ± 4.38 b 28.31 ± 3.72 ab 21.48 ± 3.62 a 0.01 5.4 5.6 5.8 6.0 6.2 6.4 6.6 6.8 3 h 6 h 24 h H i g h I n t e r m e d i a t e L o w Fig. 1. Dynamic decreasing of pH values in carcasses by pH 3h categories (Low:< 6.2; Intermediate: 6.2 to 6.5; and High:> 6.5). n = 10 per group Agricultural and Food Science (2022) 31: 229–240 233 The LL muscle from carcasses categorized as High and Low pH 3h tended to have lower WBSF values than the Inter- mediate group (Table 4). Carcasses from the Intermediate pH 3h group showed the highest DL values (0.92 ± 0.16), cooking loss (16.21 ± 0.45), ultimate pH (5.93 ± 0.05,) and WBSF (2.20 ± 0.09), without being statistically different from the other two groups (p > 0.05). Canonical correlation and bivariate correlation The canonical correlation analysis between the meat quality traits (U) and glycolytic related variables (V) revealed four canonical correlations resulted to be significant (p < 0.05). The first two canonical correlation resulted with an eigenvalue >1. Canonical redundancy analysis revealed that the first canonical correlation represents 76.80% of the explained variance, with a high degree of association between U and V (0.897). The second one represents 20.79% of the explained variance with lower degree of correlation (0.726). The total explained variance of two canonical correlation was 97.59%. The standardized canonical coefficient or canonical weight of original variables represent their relative contribution to the corresponding canonical variates. The largest canonical weights for the canonical variate U1 were detected in color related traits (a and L; Table 5) and the variable yellowness resulted with the largest correlation coefficient (0.959), followed by redness (0.942) and lightness values (0.812). These correlation coefficients also indicates that these variables exert an important contribution to the constitution of the canonical variate U1. Canonical variate U2 is represented mainly by CL (0.827), followed by WBSF (0.493) and L (–0.407); with a moderate correlation coefficient 0.623, and 0.729 for CL and WBSF, respectively. The variance of the canonical variate U1 associated with the variables of its own group represented 49.6% of the total data varia- tion which could be attributed to the high loading values of the variables related to color. However, the variance of the canonical variate U2 only represented 17% both canonical variates sum up 66.6%. Table 6 shows the standardized coefficients and correlations coefficients between the original glycolytic com- ponents and their canonical variates (V1 and V2). The variables G+G6P and LC were the largest contributors to the formation of the canonical variate V1, mainly represented by G+G6P (0.542) and LC (0.384) with a high correlation coefficient (r= 0.92 and 0.90, respectively; Table 6). Regarding canonical variate V2, LC showed the highest canonical coefficient (1.775); however, only the variable pH u resulted with high correlation (r= 0.72); Table 4. Means and standard errors of beef traits according to pH 3h groups evaluated after 24 h postmortem Variable* pH 3h groups p-value High (>6.5) Intermediate (6.2–6.49) Low (<6.2) L 24.06 ± 1.20 23.21 ± 0.69 23.84 ± 0.69 0.80 a 12.57 ± 0.35 a 11.69 ± 0.64b 13.56 ± 0.62 a 0.01 b 7.84 ± 0.35 7.37 ± 0.40 8.18 ± 0.37 0.88 CL, % 14.81 ± 0.41 16.21 ± 0.45 14.91 ± 0.55 0.24 DL, % 0.77 ± 0.18 0.92 ± 0.16 0.75 ± 0.10 0.68 WBSF, kg. 1.82 ± 0.11 2.20 ± 0.09 1.80 ± 0.07 0.35 beef pH 5.90 ± 0.06 5.93 ± 0.05 5.82 ± 0.04 0.08 L = lightness; a = redness; b: yellowness; DL = drip loss (%); CL = cooking loss (%); WBSF = Warner Bratzler shear force (kg). n = 10 per group Table 5. Standardized canonical coefficient (canonical weights) and canonical correlation coefficient (CCC) between the original variables and its canonical variate “Meat Quality” (U) Original variables Canonical variate U 1 Canonical variate U 2 Canonical weight CCC Canonical weight CCC L –0.515 0.812 –0.407 –0.275 a –0.864 0.942 –0.162 0.022 b 0.286 0.959 –0.178 –0.107 DL 0.032 –0.116 0.033 0.127 CL –0.015 0.660 0.817 0.623 WBSF –0.120 0.241 0.493 0.729 L = lightness; a = redness; b = yellowness; DL = drip loss (%); CL = Cooking loss (%); WBSF = Warner Bratzler shear force (kg); n = 30 N. Jerez-Timaure et al. 234 therefore, this canonical variate (U2) showed an irregular conformation and canonical variate V1 is represented mainly by G+G6P and LC. The variance of the canonical variate V1 associated with the variables of its own group represented 54.3% of the total data variation which could be attributed to the high loading values of the variables G+G6P and LC; the variance of the canonical variate V2 is the lowest one (8.5%). Both canonical variates covered 62.8% of the total variation. Canonical cross correlation describes the correlation between variables and the opposite canonical variate (Tables 7 and 8). Color instrumental parameters (L, a, and b) showed the highest canonical cross loading with the canoni- cal variate V1; meanwhile, for the V2, the cross correlations were significant for CL and WBSF with low to medium coefficients (Table 7). This result indicates a strong and significant linear correlation between these variables. In Table 8 is presented the canonical cross-correlation coefficients between original glycolytic related traits and ca- nonical variates U1 and U2. As expected, G+G6P and LC showed the highest canonical cross correlation with U1 and all cross-correlation coefficient with U2 were very low, verifying that this canonical variate is not applicable. The cross variance between meat quality traits (canonical variate U) and glycolytic related traits (V) accounted for 48.2% (Table 7) and the cross variance between V and U accounted for 48.9% (Table 8). The cross variance for the first canonical variate (U1 = 43.7% and V1 = 39.9%) confirmed that only the first canonical variates (U1 and V1) were representative of each group of variables based also of the fact that high correlation coefficients were detected between some original variables and the opposite canonical variate. The canonical correlation between meat quality traits and glycolytic related parameters (G+G6P, and LC) aligns with the results obtained from the linear correlation between these variables (Table 9). A bivariate correlation analysis was performed including data from all sampling times (3, 6, and 24 h) for glycolytic related traits. Results Table 6. Standardized canonical coefficient (canonical weight), and canonical correlation coefficient (CCC) between the original variables and the canonical variate “Glycolytic components” (V) Original variables Canonical variate V 1 Canonical variate V 2 Canonical weight CCC Canonical weight CCC MGC –0.091 0.413 –0.039 –0.436 G+G6P 0.542 0.920 –0.184 –0.291 LC 0.384 0.900 1.775 0.245 pH –0.329 –0.587 –0.190 –0.720 MGC = Muscular glycogen content; G+G6P = Glucose +Glucose 6-phosphate; LC = Lactate content; n = 30 Table 7. Cross correlations (canonical cross-loadings) between the meat quality traits and canonical variates V1 and V2 Original variables Canonical Variate V1 Canonical Variate V2 L 0.729 –0.200 a 0.845 0.016 b 0.861 –0.078 Drip loss –0.104 0.092 Cooking loss 0.592 0.452 WBSF 0.217 0.529 Total variance, % 43.7 4.5 WBSF = Warner Bratzler shear force; n = 30 Table 8. Cross correlations (canonical cross-loadings) between the glycogenic related traits and canonical variates U1 and U2 (n = 30) Original variables Canonical variate U1 Canonical variate U2 MGC 0.371 –0.317 G+G6P 0.826 –0.211 LC 0.808 0.178 pH u –0.527 –0.052 Total variance, % 39.9 9.0 Agricultural and Food Science (2022) 31: 229–240 235 showed that pH 3h was not correlated (p > 0.05) with any of the meat quality characteristics; however, pH u showed a significant and negative (p < 0.05) correlation with L (r= –0.55), a (r= –0.59) and b (r= –0.59) values. Correlation coefficients indicate that, the L, a, and b values decrease as the pH increases. At all sampling times, MGC, G+G6P, LC, and GP showed significant (p < 0.05) and moderate correlation with colored-related variables: L, a, and b val- ues (Table 8). The CL was correlated (P < 0.05) with G+G6P (at 3, 6, and 34 h postmortem), with LC and GP (at 6 and 24 h postmortem), and with pH at 6 h postmortem. DL and WBSF were not significantly correlated with any variable measured. Discussion Postmortem factors, like temperature, pH rate of decline and aging time have a great impact on meat quality traits (Pogorzelski et al. 2022). At the slaughterhouse the ultimate pH (measured between 24 and 48 h postmortem) is a widely-used indicator of beef carcass quality (Hopkins et al. 2014). Few studies (Smulders et al. 1990, Jones and Tatum 1994, Shulte et al. 2022) had investigated the association between early postmortem pH and meat quality traits, but none of those had included the glycolytic potential and their compounds during the early postmortem. The results from this study showed that carcasses classified by pH measured at 3h postmortem significantly var- ied in glycolytic potential, MGC and G+G6P content; however, there were not variation in meat quality traits except in redness (a value). The carcasses with an intermediate pH 3h (6.2 to 6.49) exhibited less amount of mus- cular glycogen and G+G6P resulting with the lowest glycolytic potential when comparing to High or Low pH 3h car- casses. Interaction pH 3h category × sampling time was only significant for LC; which increased with postmortem time; however, lower values were found in Intermediate pH 3h carcasses. Immonen et al. (2000) investigating the relationship between glycogen and ultimate pH, found that in muscle with low pH values, residual muscle glycogen varied from 83 to 10 mmol kg -1. In this study, MGC were below 45 mmol kg-1 at 3h postmortem, and after 24 h MGC values ranged between 16.19 and 4.14 mmol kg-1. Steers selected for this study were grass-fed without energy supplementation during the fattening phase; also they were fasted 16 h Table 9. Significant bivariate correlations between pH, glycolytic potential and its components, and meat quality traits at different sampling times Variable Sampling times L a b DL CL WBSF MGC 3 h 0.433 0.597 0.586 – – – 6 h 0.636 0.622 0.700 – – – 24 h 0.383 – 0.396 – – – G+G6P 3 h 0.619 0.725 0.764 – 0.514 – 6 h 0.685 0870 0.852 – 0.529 – 24 h 0.721 0.785 0.829 – 0.417 – LC 3 h 0.529 0.426 0.525 – – – 6 h 0.395 0.590 0.508 – 0.665 – 24 h 0.597 0.781 0.776 – 0.648 – GP 3 h 0.617 0. 89 0.310 – – – 6 h 0.708 0.781 0.817 – 0.459 – 24 h 0.661 0.760 0.786 – 0.467 – GP strict 3 h 0.515 0.666 0.667 – – – 6 h 0.674 0.701 0.763 – – – 24 h 0.608 0.625 0.772 – – – pH 3 h – – – – – – 6 h – -0.351 -0.351 – -0.413 – 24 h -0.550 -0.591 -0.591 – – – Only significant r correlation coefficients (p < 0.05) are shown. MGC = Muscular glycogen content; G+G6P = Glucose +Glucose 6-phosphate; LC = Lactate content; GP = Glycolytic Potential; GP strict = GP without LC; L = lightness; a = redness; b = yellowness; DL: drip loss (%); CL = Cooking loss (%); WBSF = Warner Bratzler shear force (kg); n = 30 N. Jerez-Timaure et al. 236 before slaughtering, this condition could explain the very low levels of MGC detected in this study. Previous studies performed in the same region also reported very low levels of postmortem MGC. Amtmann et al. (2006) found that samples from longissimus thoracis muscles in grass-fed carcasses with pH >5.8 contained 14.9 ± 13.0 mmol kg-1 of glycogen and in carcass with pH <5.8 found 35.5 ± 15.7 mmol kg-1. Also, Apaoblaza et al. (2015) compared carcasses with high (> 5.9) and normal (<5.8) muscle pH at 24 h postmortem and reported MGC at 30 min post-slaughter of 65.5 ± 5.6 mmol kg-1 in normal pH carcasses and 29.5 ± 7.22 mmol kg-1 in high pH carcasses leading to 5.71 ± 0.01 and 6.34 ± 0.05 at 24 h, respectively. In this study, we did not take samples at 30 min postmortem. Postmortem pH is influenced by the rates of ATP hydrolysis and glycolysis, phosphofructokinase activity, AMPK activity and to a much lesser extent, pH buffering capacity (Apaoblaza et al. 2015, Strobel et al. 2021, Wang et al. 2021). Chauhan and England (2018) indicated that the role of MGC in elevated pH u is quite clearly defined; how- ever, the role in normal pH u is less recognized when MGC is not reduced. This discrepancy in variation in pH u in the presence of residual glycogen has been partially explained in an in vitro study utilizing a reconstituted mus- cle glycolytic buffer system (England et al. 2016). These authors proved that in the presence of glycogen excess, oxidative muscle still produces high ultimate pH, suggesting that the glycolytic capacity of muscle dictates ulti- mate pH instead of glycogen content. A rapid early-postmortem glycolytic metabolism in carcasses categorized as Intermediate pH 3h suggests that initial MGC is not directly associated with early pH and that there is a variation on the rate of glycolysis which is not directly related to the initial pH. In addition, pre-slaughter factors, such as feeding, transportation time, and environmental conditions affect both MGC and pH u . Apaoblaza et al. (2017) evaluating the effect of the time of year and supplementation with flacked corn on MGC, LC and GP; found that the energy supplementation four weeks before slaughter reduced the depletion of muscle glycogen concentration but did not affect pH u . Other factors, such as the type of muscle, its histology, and biochemical properties are also related to growth rate and the number and fiber types affect postmortem decline of pH (Onopiuk et al. 2016). Muscle fiber type I has a slow-twitch, with a higher oxidative metabolism and low glycogen content and Type II is a fast-twitch glycolytic metabolic manner with higher content of MGC (Schiaffino and Reggiani 1994). Variations in metabolite content, especially glycogen and lactate content can explain variations in meat quality (Choe et al. 2008). In this study, the pH u was similar (p > 0.05) for all pH 3h categories (Fig. 1) showing a more drastic decline between 3 and 6 h postmortem, especially in High pH 3h carcasses. This result indicates that most of the biochemical changes occurred early in the very early postmortem (during the first six hours postmortem). We could not measure pH and take muscle samples at 9, 12, and 18 h due to slaughter operative plant restrictions. Further studies should be focus on the first 6 h postmortem, since differences between 6 and 24 h postmortem were not significant (p > 0.05). In this study, GP and GP strict were significantly different (p = 0.02) when comparing pH 3h groups. Low and High pH 3h carcasses resulted with similar (p > 0.05) glycolytic potential and the Intermediate pH 3h group of carcasses resulted with 15.17 and 19.42 mmol kg-1 of lactate less that High and low pH 3h groups, respectively. For decades, GP has been used as a good approximation to the total compounds transformable to lactic acid present in the muscle at slaughter (Monin et al. 1981, Monin and Sellier 1985). A curvilinear relationship was found between GP and ultimate pH, which described a threshold of 100 µmol g-1 of GP, below which lower GP was associated with higher ultimate pH and above which GP had no effect on pH u (Wulf et al. 2002). However, more recently several studies (Holdstock et al. 2014, England et al. 2016) have reported that the rate of depletion of muscular pH is not only de- pendent on glycogen concentration and their metabolites. In this study, the GP strict represented 32% of the sum of all glycolytic compounds at 24 h postmortem, leaving the rest (68%) to LC; similar values were found at 6 h post- mortem. This result agreed with Maribo et al. (1999), who stated that LC represents between 70 and 80% of the GP measured at 24 h postmortem. However, at 3 h postmortem, the amount of MCG + G+G6P and LC were close to 50%. The GP strict was studied considering that under postmortem conditions, gluconeogenesis is not possible, and lactate has not the potential to convert into glucose (unlike in vivo condition). Postmortem glycolytic poten- tial expresses the potential of lactate formation (Monin and Sellier 1985), and it has been recognized to be highly correlated with ultimate pH (Wulf et al. 2002); however, results from another study (England et al. 2016) also had indicated that although glycolytic potential is correlated with the extent of metabolism, pH u is not necessarily sub- strate driven. Przbylski et al. (1994) reported a quadratic relationship between glycolytic potential and ultimate pH, considering only direct glycogenolysis products; highlighting that this relationship might be useful to predict the potential meat ultimate pH from muscle obtained from live animals. Other authors (Van Laack and Kauffman 1999, Immonen and Puolanne 2000, Copenhafer et al. 2006) stated that postmortem glycolysis may stop in the presence of residual glycogen. Immonen et al. (2000) demonstrated that muscle with pH < 5.8 could have a wide range of glycogen content, indicating that other factors influencing the rate and extent of glycolysis like fiber type and the activity of various enzymes related to glycogen metabolism affect postmortem pH decline. Our results Agricultural and Food Science (2022) 31: 229–240 237 suggest that for GP determination in postmortem muscles is not necessary to include LC in the formula proposed by Monin and Sellier (1985), and therefore GP could be determined using only MGC and residual glucose contents. The pH represents one of the most important characteristics of meat because it is highly related to meat quality (Young et al. 2004). In this study, meat quality traits like WBSF, lightness (L), yellowness (b) DL, and CL were not affected by early pH variation, but redness (a; p = 0.01). Lomiwes et al. (2014) reported a curvilinear relationship between tenderness and ultimate pH (24 h postmor- tem), explaining that the mechanisms that lead to meat tenderization were not the same for a high final pH (≥ 6.2) and a low pH (≤ 5.79) muscle. Some researchers reported that tenderness had a linear relationship with pH u in Friesan-Holstein calves (Guignot et al. 1994). Pulford et al. (2008) reported a curvilinear relationship between these traits with samples form 22-month-old Angus or Angus × Limousin bulls. On the other hand, Wu et al. (2014) in a study to investigate the effect of ultimate pH on meat tenderness using Angus bulls between 18 to 24 months of age, found that the lowest shear force was observed in the high pH u meat, followed by meat from the low pH u group, the meat with intermediate pH u showed the highest shear force values. Jones and Tatum (1994) used pH 3h as a predictor parameter from carcasses produced under commercial conditions (n=240), finding that values low- er than 6.2 were associated with a reduction in the variability of shear force. Also, Smulders et al. (1990) reported that pH 3 values below 6.3 were associated with a lower incidence of undesirable ratings for tenderness. In this study, color-related parameters exhibited a positive and moderate correlation with MGC, G+G6P, LC, GP, and GP strict at any sampling time. Jeremiah et al. (1991) reported that colorimeter readings have a higher coeffi- cient correlation with muscle pH u than color scores (evaluated by panelists). The results from this study confirm a significant relationship between MGC and meat quality traits such as pH and meat color. Also, Przybylski et al. (2016) reported evidence of a significant relationship between glucose and lactate measured in drip loss with muscle glycogen, lactate, glycolytic potential, and meat quality traits. There is a consensus that the use of a CCA as a multivariate approach is appropriate for evaluating the interrela- tions among complex traits such meat quality and carcass traits (Vargas et al. 2018, 2021). For this study, CCA was suitable because it measures the magnitude of interrelations between sets of multiple variables (Ventura et al. 2011). In this study, CCA stablished that glycolytic related traits were represented by LC and G+G6P; on the other hand, meat quality traits were mostly represented with color instrumental parameter. The cross-correlation coef- ficients confirmed a strong association between these variables. Canonical weights are important parameters for defining the contribution of original variables to the canonical variates. However, the understanding of canonical loadings and cross-loadings is critical because these values describe the correlation between original and canoni- cal variates. The CCA was useful in describing the relation between meat quality traits and glycolytic metabolites, avowing the multicollinearity bias providing insights that may contribute to the beef industry to get better early postmortem predictors for meat quality. Also, it would allow to characterize meat quality considering glycolytic metabolites measured at a few hours postmortem. Our findings represent the first evidence of a strong multivari- ate relationship of meat quality traits with glycolytic components in postmortem muscle. Ultimate pH is a very common parameter used to evaluate carcass quality and as a predictor of meat quality. This technique is simple, but it is measured 24 h after refrigeration and it is not possible to take a decision before that time. Also, in the industries with a high rate of carcass production, conventional pH testing in chillers is a difficult task (Young et al. 2004). This study did not offer strong evidence to recommend the measurement of early pH (measure at 3 h postmortem) in beef carcass instead of ultimate pH as an indicator of meat quality, since varia- tion in meat quality traits evaluated were not correlated with meat quality traits or glycolytic metabolites, but LC. However, the results from this study provided a better understanding of early postmortem pH and its relationship with postmortem glycolysis. Rather than using early pH to predict meat quality, it would be possible to predict the potential meat ultimate pH from glycogen muscle, residual glucose, or lactate content from early postmortem samples. Edwards et al. (2010) reported a significant relationship between blood lactate concentration and pH measured at 1 h postmortem. Przybylski et al. (2016) concluded that simple measurement of glucose and lactate by applications of biosensors can be a good predictor of glycolytic potential and pH. A patent (WO 001112844) online method of predicting ultimate pH from a rapid measurement of glycogen potential was reported by Young et al. (2004) suitable to be used in slaughterhouses. The results also suggest that to identify changes in metabolites and enzymes, the muscle samplings should be carried out more often than just at 0.5 and 24 h, at least within the first six hours postmortem, since we observed that most of the postmortem variation in MGC, LC, GP and pH occurred during the first six hours postmortem. N. Jerez-Timaure et al. 238 Apparently, a 24-hour period would be too long to find differences in some parameters of muscle metabolisms that could finally affect meat quality. The determination of glycolytic enzymes and AMPK activity will be improved knowledge of how pH and temperature affect the biochemical and physical conversion of muscle to meat in beef will provide an understanding of how to optimize control of postmortem metabolism to ensure uniformity of carcass quality and reduce meat quality defects like dark cutting meat. Conclusions Results from this study indicate that the measurement of early pH (at 3 h postmortem) does not allow to recom- mend to segregate carcasses by pH, since the evolution of pH and the glycolytic potential differ in carcasses with an intermediate pH compared to the others and meat quality traits were not affected by pH 3h . Canonical correlation analysis is an optimized multivariate technique for evaluating the existence or non- existence of relationships between complex groups of variables. In this study it proves to be a powerful tool to study the relationship between the selected set of meat quality traits and glycolytic potential components. An important relationship between glucose and lactate content and instrumental color parameters in beef LL muscle from grass fed cattle were demonstrated. Considering that under postmortem conditions, gluconeogenesis is not possible, and lactate has not the poten- tial to convert into glucose, the measurement of GP strict calculated with glycogen and glucose content resulted as a suggested way to measure the glycolytic potential in postmortem muscles, without considering the lactate con- tent. This recommendation deserves further validation. The results strongly suggest that to investigate early postmortem metabolism, the muscle samplings should be concentrated within the first six hours postmortem, since we observed that most of the postmortem variation gly- colytic related traits and pH occurred during the first six hours postmortem. Acknowledgement This work was supported by Vicerrectoría de Investigación, Desarrollo y Creación Artística (VIDCA) of Universidad Austral de Chile (Research Project No. S2017-26). References AMSA 2015. Research guidelines for cookery, sensory evaluation, and instrumental tenderness measurements of meat. American Meat Science Association, Champaign, Illinois, USA Amtmann, V.A., Gallo, C., Van Schaik, G. & Tadich, N. 2006. Relaciones entre el manejo antemortem, variables sanguíneas indica- doras de estrés y pH de la canal en novillos. Archivos de Medicina Veterinaria 38: 259–264. (in Spanish). https://doi.org/10.4067/ S0301-732X2006000300010 Apaoblaza, A., Galaz, A., Strobel, P., Ramírez-Reveco, A., Jerez-Timaure, N. & Gallo, C. 2015. Glycolytic potential and activity of adenosine monophosphate kinase (AMPK), glycogen phosphorylase (GP) and glycogen debranching enzyme (GDE) in steer carcasses with normal (pH <5.8) or high (pH >5.9) determined in M. longissimus dorsi. Meat Science 101: 83–89. https://doi. org/10.1016/j.meatsci.2014.11.008 Apaoblaza, A., Strobel, P., Ramírez-Reveco, A., Jerez-Timaure, N., Monti, G. & Gallo, C. 2017. Effect of season, supplementation and fasting on glycolytic potential and activity of AMP-activated protein kinase, glycogen phosphorylase and glycogen debranch- ing enzyme in grass-fed steers as determined in Longissimus lumborum muscle. Livestock Science 202: 101–108. https://doi. org/10.1016/j.livsci.2017.05.028 Bowker, B.C., Botrel, C., Swartz, D.R., Grant, A.L. & Gerrard, D.E. 2004. Influence of myosin heavy chain isoform expression and portmortem metabolism on the ATPase activity of muscle fibers. Meat Science 68: 587–594. https://doi.org/10.1016/j.meats- ci.2004.05.010 Chauhan, S.S. & England, E.M. 2018. Postmortem glycolysis and glycogenolysis: insights from species comparisons. Meat Science 144: 118–126. https://doi: 10.1016/j.meatsci.2018.06.021 Choe, J.H., Choi, Y.M., Lee, S.H., Shin, H.G., Ryu, Y.C., Hong, K.C. & Kim, B.C. 2008. The relation between glycogen, lactate content and muscle fiber type composition and their influence on postmortem glycolytic rate and pork quality. Meat Science 80: 355–362. https://doi:10.1016/j.meatsci.2007.12.019 Copenhafer, T.L., Richert, B.T., Schinckel, A.P., Grant, A.L. & Gerrard, D.E. 2006. Augmented postmortem glycolysis does not oc- cur early postmortem in AMPKγ3-mutated porcine muscle of halothane positive pigs. Meat Science 73: 590–599. https://doi. org/10.1016/j.meatsci.2006.02.015 Agricultural and Food Science (2022) 31: 229–240 239 Edwards, N.N., Engle, T.E., Correa, J.A., Paradis, M.A., Grandin, T. & Anderson, D.B. 2010. The relationship between exsanguina- tion blood lactate concentrations and carcass quality in slaughter pigs. Meat Science 85: 435–440. https://doi.org/10.1016/j. meatsci.2010.02.012 England, E.M., Matarneh, S.K., Oliver, E.M., Apaoblaza, A., Scheffler, T.L., Shi, H. & Gerrard, D.E. 2016. Excess glycogen does not resolve high ultimate pH of oxidative muscle. Meat Science114: 95–102. https://doi.org/10.1016/j.meatsci.2015.10.010 Guignot, F., Touraille, C., Ouali, A., Renerre, M. & Monin, G. 1994. Relationships between postmortem pH changes and some traits of sensory quality in veal. Meat Science 3: 315–325. https://doi.org/10.1016/0309-1740(94)90049-3 Hannula, T. & Puolanne, E. 2004. The effect of cooling rate on beef tenderness: The significance of pH at 7 °C. Meat Science 67: 403–408. https://doi.org/10.1016/jmeatsci.2003.101.012 Hamoen, J.R., Vollebregt, H.M. & van der Sman, R.G.M. 2013. Prediction of the time evalution of pH in meat. Meat Science 141: 2363–2372. https://doi.org/10.1016/j.foodchem.2013.04.127 Holdstock, J., Aalhus, J.L., Uttaro, B.A., López-Campos, Ó., Larsen, I.L. & Bruce, H.L. 2014. The impact of ultimate pH on muscle characteristics and sensory attributes of the longissimus thoracis within the dark cutting (Canada B4) beef carcass grade. Meat Science 98: 842–849. https://doi: 10.1016/j.meatsci.2014.07.029 Honikel, K.O. 1998. Reference methods for the assessment of physical characteristics of meat. Meat Science 49: 447–457. https:// doi.org/10.1016/S0309-1740(98)00034-5 Hopkins, D.L., Ponnampalam, E.N., Van de Ven, R.J. & Warner, R.D. 2014. The effect of pH decline rate on the meat and eating quality of beef carcasses. Animal Production Science 54: 407–413. https://doi.org 10.1071/AN12314 Immonen, K. & Puolanne, E. 2000. Variation of residual glycogen-glucose concentration at ultimate pH values below 5.75. Meat Science 55: 279–283. https://doi: 10.1016/s0309-1740(99)00152-7 Immonen, K., Ruusunen, M., Hissa, K. & Puolanne, E. 2000. Bovine muscle glycogen concentration in relation to finishing diet, slaughter, and ultimate pH. Meat Science 55: 25–31. https://doi: 10.1016/s0309-1740(99)00121-7 INN (Instituto Nacional de Normalización) 2002. Santiago, Chile. Norma Chilena Oficial Nch. 1306 p. (in Spanish). Jeremiah, L.E., Tong, A.K.W. & Gibson, L.L. 1991. The usefulness of muscle color and pH for segregating beef carcasses into ten- derness groups. Meat Science 30: 97–114. https://doi.org/10.1016/0309-1740(91)90001-7 Jones, B.K. & Tatum, J.D. 1994. Predictors of beef tenderness among carcasses produced under commercial conditions. Journal of Animal Science 72: 1492–1501. https://doi.org/10.2527/1994.7261492x Karlsson, A.H., Klont, R.E., & Fernendez, X. 1999. Skeletal muscle fibres as factors for pork quality. Livestock Production Science 60: 255–269. https://doi.org/10.1016/S0301-6226(99)00098-6 Lomiwes, D., Farouk, M.M., Wu, G. & Young, O.A. 2014. The development of meat tenderness is likely to be compartmentalised by ultimate pH. Meat Science 96: 646–651. https://doi: 10.1016/j.meatsci.2013.08.022 Maribo, H., Støier, S. & Jørgensen, P.F. 1999. Procedure for determination of glycolytic potential in porcine m. longissimus dorsi. Meat Science 51: 191–193. https://doi: 10.1016/s0309-1740(98)00130-2 Monin, G., Sellier, P., Ollivier, L., Goutefongea, R. & Girard, J.P. 1981. Carcass characteristics and meat quality of halothane nega- tive and halothane positive Pietrain pigs. Meat Science 5: 413–423. https://doi: 10.1016/0309-1740(81)90040-1 Monin, G. & Sellier, P. 1985. Pork of low technological quality with a normal rate of muscle pH fall in the immediate post-mortem period: The case of the Hampshire breed. Meat Science 13: 49–63. https://doi.org/10.1016/S0309-1740(85)80004-8 Onopiuk, A., Półtorak, A. & Wierzbicka, A. 2016. Influence of post-mortem muscle glycogen content on the quality of beef during aging. Journal of Veterinary Research 60: 301–307. https://doi10.1515/jvetres-2016-0046 Przybylski, W., Sionek, B., Jaworska, D. & Santé-Lhoutellier, V. 2016. The application of biosensors for drip loss analysis and glyco- lytic potential evaluation. Meat Science 117: 7–11. https://doi: 10.1016/j.meatsci.2016.02.025 Pogorzelski, G., Pogorzelska-Nowicka, E., Pogorzelski, P., Półtorak, P., François Hocquette, J. & Wierzbicka, A. 2022. Towards an integration of pre- and post. Slaughter factors affecting the eating quality of bee. Livestock Science 255: 104795. https://doi. org/10.1016/j.livsci.2021.104795 Pulford, D.J, Fraga-Vazquez, S., Frost, D.F., Fraser-Smith, E., Dobbie, P. & Rosenvold, K. 2008. The intracellular distribution of small heat shock proteins in post-mortem beef is determined by ultimate pH. Meat Science 79: 623–630. https://doi.org/10.1016/j. meatsci.2007.10.027 Schiaffino, S. & Reggiani, C. 1994. Myosin isoforms in mammalian skeletal muscle. Journal of Applied physiology 77: 493–501. https://doi.org/10.1152/jappl.1994.77.2.493 Shulte, M.D., Hochmuth, K.G., Steadhan E.M., Lonergan, S.M., Hansen, S.l. & Huff-Lonergan, E.J. 2022. Early postmortem muscle proteome and metabolome of beef longissimus thoracis muscle classified by pH at 6 h postmortem. Journal of Proteomics 271: 104756. https://doi.org/10.1016/j.jprot.2022.104756 Smulders, F.J.M., Marsh, B.B., Swartz, D.R., Russell, R.L. & Hoenecke, M.E. 1990. Beef tenderness and sarcomere length. Meat Science 28: 349–363. https://doi.org/10.1016/0309-1740(90)90048-B Strobel, P., Galaz, A., Villaroel-Espíndola, F., Apaoblaza, A., Slebe, J.C., Jerez-Timaure, N., Gallo, C. & Ramírez-Reveco, A. 2021. Temperature, but not excess of glycogen, regulates “in vitro” AMPK activity in muscle samples of steer carcasses. PLoS ONE 16: e0229480. https://doi.org/10.1371/journal.pone.0229480 van Laack, R.L.J.M. & Kauffman, R.G. 1999. Glycolytic potential of red, soft, exudative pork longissimus muscle. Journal of Animal Science 77: 2971–2973. https://doi.org/10.2527/1999.77112971x N. Jerez-Timaure et al. 240 Vargas, J.A.C., Coutinho, J.E.S., Gomes, D.I., Alves, K.S. & Maciel, R.P. 2021. Multivariate relationship among pH, subcutaneous fat thickness, and color in bovine meat using canonical correlation analysis. Revista Colombiana de Ciencias Pecuarias 34: 117–125. (in Spanish). https://doi.org/10.17533/udea.rccp.v34n2a06 Vargas, J.A.C., Almeida, A.K., Härter, C.J., Souza, A.P., da Rocha Fernandes, M.H.M., de Resende, K.T. & de Almeida Teixeira, I.A.M. 2018. Multivariate relationship among body protein, fat, and macro minerals of male and female Saanen goats using canonical correlation analysis. Revista Brasileira de Zootecnia 47: e20170289. https://doi.org/10.1590/rbz4720170289 Ventura, H.T., Lopez, P.S., Peloso, J.V., Guimarães, S.E.F., Carneiro, A.P.S. & Carneiro, P.L.S. 2011. A canonical correlation analysis of the association between carcass and ham traits in pigs used to produce dry-cured ham. Genetic Molecular Biology Journal 34: 451–455. https://doi: 10.1590/S1415-47572011000300014 Wang, C., Matarneh, S.K., Gerrard, D. & Tan, J. 2021. Modelling of energy metabolism and analysis of pH variations in postmor- tem muscle. Meat Science 182: 108634. https://doi.org/10.1016/j.meatsci.2021.108634 Wu, G., Farouk, M.M., Clerens, S. & Rosenvold, K. 2014. Effect of beef ultimate pH and large structural protein changes with ag- ing on meat tenderness. Meat Science 98: 637–645. https://doi.org/10.1016/j.meatsci.2014.06.010 Wulf, D.M., Emnett, R.S., Leheska, J.M. & Moeller, S.J. 2002. Relationships among glycolytic potential, dark cutting (dark, firm, and dry) beef, and cooked beef palatability. Journal of Animal Science 80: 1895–1903. https://doi: 10.2527/2002.8071895x Young, O.A., West, J., Hart, A.L. & Van-Otterdijk, F.F.H. 2004. A method for early determination of meat ultimate pH. Meat Science 66: 493–498. https://doi:10.1016/S0309-1740(03)00140-2 Variations of early postmortem pH in carcasses of grass-fed steersand its relationship with glycolytic potential and meat qualitytraits Introduction Materials and methods Animals, slaughter, carcass handling, and sampling Glycogen, glucose, lactate contents and glycolytic potential Meat quality traits Statistical analysis Results Glycolytic potential and its components in LL muscles from carcasses grouped by pH3h Effects of early pH on meat quality traits Canonical correlation and bivariate correlation Discussion Conclusions Acknowledgement References