2017 Volume 23 Issue 3 Pages 375-383
The aim of this study was to evaluate the physicochemical and textural characteristics and volatile compounds of Kazakh dry-cured beef made in China. Two types of Kazakh dry-cured beef were investigated: Kazakh dry-cured beef made with smoking and spices (T1) and without smoking and spices (T2). There were significant (P < 0.05) differences in values of aw, moisture, L*, cohesiveness and chewiness between the two types. A total of 86 volatile compounds were isolated by solid phase microextraction-gas chromatography/mass spectrometry (SPME-GC/MS). Hydrocarbons were the most abundant in T1 products and aldehydes in T2 products. Principal component analysis showed that the first principal component (PC1) was highly related to smoke derivatives—naphthalene, 2-cyclopenten-1-one derivatives, 4-methyl-4-hepten-3-one, acetophenone, 2,3-dihydro-1H-Inden-1-one, 2-furanmethanol, methoxy-phenyl-oxime, furfural, 1-(2-furanyl)-ethanone and phenols—and the second principal component (PC2) to lipid derivatives—straight-chain aliphatic aldehydes, methyl ketone, straight-chain alcohols and 2-pentyl-furan.
Dry-cured beef, a type of product made from beef cuts, is processed by curing and drying. Many types of dry-cured beef, produced in different countries and regions including Italian bresaola (Paleari et al., 2000), Brazilian Charqui (Youssef et al., 2007), Turkish pastirma (Kaban, 2009), and Spanish cecina (Molinero et al., 2008), are favored by consumers. The type of muscle used, the processing technology as well as its geographical origin can lead to different quality characteristics of these dry-cured meat products (Sánchez-Peña et al., 2005; Toldrá, 2006; Wȩsierska et al., 2014).
Kazakh dry-cured beef is a popular traditional meat product from the Xinjiang Uygur Autonomous Region of Northwestern China with its distinctive nomadic features. It is an important part of Kazakh food culture connecting the nation's history to the present day. Manufacturing traditional Kazakh dry-cured beef is simple and consists of three stages: cutting up the meat followed by salting and air-drying. Once dried, the meat is sometimes smoked. Unlike other dry-cured beef or ham products with longer ripening times (Molinero et al., 2008; Pugliese et al., 2015), Kazakh dry-cured beef is generally manufactured without any ripening stage. Kazakh dry-cured beef is traditionally made with a variable processing time depending on the size of the cut of meat used and the climatic conditions prevailing during the drying stage. In Xinjiang, it is made mainly during winter in a climate of low temperature and low humidity. Special geographical conditions and unique production methods combine to form the unique flavor and chewy texture of Kazakh dry-cured beef, which has become a characteristic regional food important for tourism in Xinjiang. At present, most Kazakh dry-cured beef is produced in family workshops or on a small scale in plants with no standard production methods. Scientific research on Kazakh dry-cured beef has just begun. Sha et al. (2016) have investigated changes in lipid oxidation, fatty acid profile and volatile compounds of traditional Kazakh dry-cured beef during its processing and storage. Information about quality characteristics of different types of Kazakh dry-cured beef is scarce, and is required for its industrial production.
The objectives of the present study are therefore to investigate the physicochemical and textural characteristics and volatile compounds of 30 samples of Kazakh dry-cured beef produced in China's Xinjiang region. Principal component analysis will be used to distinguish the differences in volatile flavor compounds between the samples.
Samples 30 Kazakh dry-cured beef production were purchased from 10 different producers for sale in three main production regions of Xinjiang, China: Altay City, Ili Kazakh Autonomous Prefecture and Tower City. Three individually wrapped products were purchased from the same producer. Detailed information on the raw meat characteristics and the processing conditions for Kazakh dry-cured beef production were obtained from each producer at the time of purchase. Based on the production process, 30 samples of Kazakh dry-cured beef were divided into two types: those made by smoking with spices (T1) and those made without smoking or spices (T2). Grass-fed cow or steer meat has been considered most suitable for making Kazakh dry-cured beef because of its unique flavor and chewy texture. The production process was generally as follows: the raw meat was cut into strips (thickness ∼ 3 cm, width ∼ 5 cm) along the direction of the smooth muscle fibers. For the T1 products, the meat strips were treated with salt and a mixture of spices (pepper, fennel, parsley and geraniol) for 24 – 48 h at 0 – 4°C, while only salt was used for the T2 products. After salting, the meat strips were hung in a clean ventilated environment to dry for 2 – 4 d at 0 – 12°C and 35% – 55% relative humidity. For making T1 products only, the meat strips were transferred to the smoking room at 60 – 80°C until the meat surface had hardened and become dark red. For T2 products, this step was omitted. All samples were purchased in December, then vacuum-packed and stored at −80°C until further analysis.
Physicochemical analysis Water activity (aw) was measured using a portable water activity meter (HygroPalm AW1, Rotronic AG, Bassersdorf, Switzerland). The meat color was assessed using a portable colorimeter (CR-400, Konica-Minolta Sensing Inc., Osaka, Japan) using CIELAB space (CIE, 1976). Lightness (L*), redness (a*) and yellowness (b*) were measured after exposing the meat samples to air for 30 min. The pH was determined using a pH-meter (IQ160, Spectrum Technologies Inc., Fort Worth, Tex., USA). The protein, fat, moisture and NaCl contents were analyzed according to ISO methods 937:1978 (ISO, 1978), 1443:1973 (ISO, 1973), 1442:1997 (ISO, 1997) and 1841:1996 (ISO, 1996). Each test sample was measured three times and the results were averaged.
Texture profile analysis (TPA) Meat samples were placed in a retort pouch and heated in a boiling water bath until a core-temperature of 70°C was attained. After cooling to room temperature (22°C), four meat cubes, 1 × 1 × 1 cm, were taken from each sample then used for instrumental texture analysis. TheTPA was performed using a texture analyzer (TA-XT plus, Stable Micro Systems, Godalming, UK). A two-cycle compression test was carried out using a cylindrical probe (35 mm diameter) at 75% compression using a pre-test speed of 2 mm/s, a test speed of 1 mm/s and a post-test speed of 2 mm/s. Six texture parameters, obtained from the force-time curve: hardness (kg), springiness, cohesiveness, gumminess, chewiness (kg) and resilience, were calculated using the Texture Expert software (Stable Micro Systems).
Volatile compounds analysis SPME-GC/MS was used to analyze the volatile compounds in the meat sample. Six g of ground meat samples were placed in a 20-mL sample vial and sealed with a septum. The vial was placed in a water bath at 50°C to equilibrate for 20 min then an SPME fiber (50/30 µm carboxen/divinylbenzene/polydimethylsiloxane, Supelco, Bellefonte, Pa., USA) was inserted into the headspace for extraction for 40 min at 50°C. The fiber was then injected into a 7890A-7000B GC-MS (Agilent Technologies Inc., Santa Clara, Calif., USA) for desorption for 7 min at 250°C. The volatile compounds were separated using a DB-WAX capillary column (30 m × 0.25 mm × 0.25 µm, J & W Scientific, Folsom, Calif., USA). The GC was operated in split mode (5:1) with a column flow rate of 6 mL/min. The temperature program was: 40°C for 3 min, heating to 200°C at 5°C/min, heating to 230°C at 10°C/min and finally kept at 230°C for 3 min. The GC/MS interface was set at 250°C. The MS operating conditions were: electron energy, 70 eV; interface temperature, 250°C; transfer line temperature, 280°C; ion source temperature, 230°C; quadrupole temperature, 150°C; and scan range, m/z 55 – 500.
The volatiles were identified by matching their mass spectra with those in the NIST mass spectra libraries and by comparing the retention indices (IR) with those reported in the literature. The IR was calculated in relation to the retention time of n-alkane standards (C7 – C23) (Supelco, Sigma-Aldrich, St. Louis, Mo., USA). The internal standard method was used for quantifying the volatiles. 2-methyl-3-heptanone (1 µL , 0.41 mg/mL) was added to the samples at the same time as the internal standards. The results were expressed as ng/g of dry matter. The concentration for each compound was calculated as follows:
 |
where Concn (i) was the concentration of a compound; Area (i) the area of a compound on the chromatogram; and Area (j) the area of the internal standard on the chromatogram.
Statistical analysis The physicochemical, textural and volatile compounds data were analyzed by one-way analysis of variance using version 19.0 of the SPSS statistics software (SPSS, Chicago, Ill., USA). The mean values were compared using Duncan's multiple range test at a significance level of P < 0.05. The Pearson correlation analysis was used to evaluate the relationships between the variables. Data obtained by SPME-GC/MS were analyzed by principal component analysis (PCA) to evaluate differences in the volatile profiles of the two types of Kazakh dry-cured beef.
Physico-chemical and texture properties Table 1 shows the physicochemical and textural characteristics of the two types of Kazakh dry-cured beef. The mean pH values were 5.72 (T2) and 5.85 (T1) with low coefficients of variation (CV = 5.15%) which were close to those reported for other dry-cured beef products (pH 5.72 – 5.95) (Paleari et al., 2000; Molinero et al., 2008; Ceylan and Aksu, 2011).
| T1 (n=12) | T2 (n=18) | SEM | P-level | CV% (n=30) | |
|---|---|---|---|---|---|
| pH | 5.85 | 5.72 | 0.05 | NS | 5.15 |
| aw | 0.92 | 0.94 | 0.00 | * | 2.56 |
| Moisture (g kg−1) | 564.76 | 615.21 | 0.01 | * | 13.53 |
| Fat (g kg−1) | 43.10 | 36.53 | 0.01 | NS | 84.41 |
| Protein (g kg−1) | 308.58 | 290.33 | 1.12 | NS | 20.68 |
| NaCl (g kg−1) | 47.91 | 39.82 | 0.31 | NS | 39.51 |
| L* | 36.66 | 32.33 | 0.96 | * | 15.37 |
| a* | 12.11 | 11.18 | 0.67 | NS | 31.61 |
| b* | 7.47 | 7.45 | 0.39 | NS | 28.29 |
| Hardness (kg) | 10.39 | 6.42 | 1.20 | NS | 82.21 |
| Springiness | 0.41 | 0.39 | 0.01 | NS | 15.96 |
| Cohesiveness | 0.46 | 0.42 | 0.01 | * | 11.68 |
| Gumminess (kg) | 5.16 | 2.77 | 0.68 | NS | 99.88 |
| Chewiness (kg) | 2.45 | 1.11 | 0.38 | * | 125.41 |
| Resilience | 0.19 | 0.18 | 0.01 | NS | 26.25 |
P-level: Level of significance found by analysis of variance. NS, not significant; *, P < 0.05.
The mean aw values were 0.92 (T1) and 0.94 (T2) and a low overall CV of 2.56%. These values were higher than the critical aw value of 0.75 which confers microbial stability on meat products stored at room temperature (Leistner, 1987). Therefore, Kazakh dry-cured beef can be considered as a “perishable” product with aw values ranging of 0.95 – 0.91, so must be stored at or below 10°C (Leistner and Roedel, 1975). The mean moisture contents were in the range of 564.76 – 615.21 g kg−1 (CV = 13.53%), higher than those of other dry-cured beef products: Turkish pastirma, 466.1 – 479.6 g kg−1 (Ceylan and Aksu, 2011); and Spanish Cecina de León, 501.1 – 601.4 g kg−1 (Molinero et al., 2008). The mean aw and moisture values in T1 products were significantly lower (P < 0.05) than in T2 products, because of differences in the extent of drying during the smoking process.
The mean protein contents were 290.33 (T2) and 308.58 (T2) g kg−1 (CV = 20.68%). This value was similar to that reported for bresaola (319.6 g kg−1) (Paleari et al., 2000), but higher than that for Charqui (263 g kg−1) (Youssef et al., 2007). The mean fat contents of 36.53 (T2) and 43.10 (T1) g kg−1 were higher than those of bresaola (17.425.0 g kg−1) (Paleari et al., 2000; Youssef et al., 2007). A higher CV of 84.41% was observed for fat content overall. The mean NaCl contents were 39.82 (T2) and 47.91 (T1) g kg−1 with a CV of 39.51%. There were no significant differences between the two types of samples regarding protein, fat and NaCl content (P > 0.05). A Pearson correlation analysis showed that the protein and NaCl contents were significantly negatively correlated with moisture content (r = −0.785** and r = −0.453*, respectively, P < 0.05) and aw (r = −0.746** and r = −0.426*, respectively, P < 0.05).
Regarding color parameters, the mean L*, a*and b* values for T2 and T1 were 32.33 and 36.66, 11.18 and 12.11 and 7.45 and 7.47, respectively. The L* values in T1 products were significantly higher than in T2 products (P < 0.05). The mean a* value had a higher CV (31.61%) than L* (CV = 15.37%) and b* (CV = 28.29%). This indicated a visually inconsistent red color among the samples. This may have been related to the high variability in fat content, which was significantly negatively correlated (r = −0.487, P < 0.05) with a*.
The texture parameters from TPA showed extremely high coefficients of variation (11.68 – 125.41%). These results illustrated the variability in the processing technology and conditions used by the producers of Kazakh dry-cured beef products on the market which has led to very different texture profiles. There were significant differences (P < 0.05) in cohesiveness and chewiness between the two types of samples, with no significant differences (P > 0.05) found in hardness, springiness, gumminess and resilience. The T1 products showed higher cohesiveness and chewiness values than the T2 products.
Volatile compounds A total of 86 volatile compounds were identified and quantified in the two types of Kazakh dry-cured beef (Table 2). The volatile classes comprised 21 hydrocarbons, 10 aldehydes, 15 ketones, 15 alcohols, five nitrogenous, four furans, two esters, two ethers, and 12 phenols. Seventy-three volatile compounds were detected in the T1 products and 36 in the T1 products. Only 23 volatile compounds were detected in both T1 and T2 products: hydrocarbons (10), aldehydes (6), alcohols (6), and furans (1).
| Code | Compound | RIa | T1(n=12) | T2 (n=18) | SEM | P-level | Identification |
|---|---|---|---|---|---|---|---|
| Hydrocarbons | |||||||
| A1 | Hexane | 806 | 29.61 | 41.78 | 4.17 | NS | MS |
| A2 | Trichloromethane | 924 | ND | 2.66 | 0.77 | - | MS |
| A3 | α-Pinene | 1019 | 35.41 | ND | 8.09 | - | MS; RI |
| A4 | α-Phellandrene | 1024 | 10.32 | ND | 1.87 | - | MS |
| A5 | Toluene | 1037 | 48.90 | 29.42 | 5.93 | NS | MS; RI |
| A6 | Undecane | 1088 | 69.16 | 24.31 | 10.44 | * | MS; RI |
| A7 | β-Pinene | 1100 | 47.69 | ND | 10.71 | - | MS; RI |
| A8 | β-Phellandrene | 1115 | 17.17 | ND | 2.94 | - | MS; RI |
| A9 | Ethylbenzene | 1123 | 29.29 | 11.74 | 4.53 | NS | MS; RI |
| A10 | o-Xylene | 1131 | 22.83 | 15.17 | 2.33 | NS | MS; RI |
| A11 | p-Xylene | 1137 | 32.25 | 30.15 | 6.37 | NS | MS; RI |
| A12 | 3-Carene | 1147 | 119.11 | ND | 30.86 | - | MS; RI |
| A13 | Ocimene | 1176 | 12.05 | ND | 2.47 | - | MS |
| A14 | Dodecane | 1196 | 134.27 | 49.18 | 21.10 | * | MS; RI |
| A15 | D-Limonene | 1197 | 159.98 | 21.93 | 21.12 | ** | MS; RI |
| A16 | Eucalyptol | 1210 | 205.09 | ND | 32.50 | - | MS |
| A17 | Styrene | 1257 | 31.15 | 22.36 | 4.06 | NS | MS; RI |
| A18 | Tridecane | 1295 | 92.78 | 40.93 | 18.32 | NS | MS; RI |
| A19 | Caryophyllene | 1605 | 469.66 | ND | 146.65 | - | MS; RI |
| A20 | Phenylethyne | 1607 | ND | 2.53 | 0.51 | - | MS |
| A21 | Naphthalene | 1750 | 79.25 | ND | 13.95 | - | MS; RI |
| Subtotal | 1645.99 | 292.17 | 236.65 | ** | |||
| Aldehydes | |||||||
| B1 | Butanal, 3-methyl- | 911 | 14.74 | 3.71 | 2.34 | * | MS; RI |
| B2 | Hexanal | 1078 | 25.68 | 126.47 | 29.14 | NS | MS; RI |
| B3 | Heptanal | 1183 | 47.77 | 65.84 | 10.84 | NS | MS; RI |
| B4 | Octanal | 1289 | 39.83 | 77.14 | 19.08 | NS | MS; RI |
| B5 | 2-Heptenal, (Z)- | 1322 | ND | 6.11 | 2.62 | - | MS; RI |
| B6 | Nonanal | 1395 | 157.76 | 253.13 | 54.29 | NS | MS; RI |
| B7 | (E)-2-Octenal | 1431 | 3.93 | 13.29 | 3.48 | NS | MS; RI |
| B8 | Decanal | 1496 | ND | 11.86 | 2.45 | - | MS; RI |
| B9 | 2-Nonenal, (E)- | 1536 | ND | 15.68 | 3.87 | - | MS; RI |
| B10 | Benzaldehyde, 4-(1-methylethyl)- | 1789 | 18.10 | ND | 2.57 | - | MS |
| Subtotal | 307.81 | 573.22 | 120.58 | NS | |||
| Ketones | |||||||
| C1 | 2-Pentanone | 973 | ND | 9.95 | 2.17 | - | MS; RI |
| C2 | 2-Octanone | 1282 | ND | 9.54 | 2.68 | - | MS; RI |
| C3 | 2,3-Octanedione | 1319 | ND | 53.36 | 12.49 | - | MS; RI |
| C4 | 2-Cyclopenten-1-one, 2-methyl- | 1374 | 30.21 | ND | 4.24 | - | MS; RI |
| C5 | 2-Nonanone | 1387 | ND | 2.14 | 0.62 | - | MS; RI |
| C6 | 2-Cyclopenten-1-one, 2,3-dimethyl- | 1454 | 19.52 | ND | 2.65 | - | MS |
| C7 | 2-Cyclopenten-1-one, 3,4-dimethyl- | 1489 | 31.84 | ND | 4.19 | - | MS |
| C8 | 2-Cyclopenten-1-one, 2,3,4-trimethyl- | 1504 | 17.42 | ND | 2.85 | - | MS |
| C9 | 2-Cyclopenten-1-one, 3-methyl- | 1527 | 68.40 | ND | 8.80 | - | MS |
| C10 | 2-Cyclopenten-1-one, 3,4,4-trimethyl- | 1563 | 14.44 | ND | 2.14 | - | MS |
| C11 | 2-Cyclopenten-1-one, 3,5,5-trimethyl- | 1620 | 10.39 | ND | 1.46 | - | MS |
| C12 | 2-Cyclopenten-1-one, 3-ethyl- | 1639 | 12.53 | ND | 1.96 | - | MS |
| C13 | Acetophenone | 1657 | 31.67 | ND | 3.32 | - | MS; RI |
| C14 | 4-Hepten-3-one, 4-methyl- | 1731 | 25.16 | ND | 3.37 | - | MS |
| C15 | 1H-Inden-1-one, 2,3-dihydro- | 2027 | 23.36 | ND | 2.83 | - | MS |
| Subtotal | 284.94 | 74.98 | 32.96 | ** | |||
| Alcohols | |||||||
| D1 | Ethanol | 936 | 35.13 | ND | 4.65 | - | MS; RI |
| D2 | 1-Penten-3-ol | 1159 | ND | 5.88 | 1.58 | - | MS; RI |
| D3 | 3-Methyl-1-butanol | 1209 | 18.09 | 17.19 | 6.17 | NS | MS; RI |
| D4 | 1-Pentanol | 1239 | 2.68 | 16.95 | 3.18 | * | MS; RI |
| D5 | 1-Hexanol | 1355 | 16.80 | 68.34 | 18.76 | NS | MS; RI |
| D6 | 1-Octen-3-ol | 1451 | 21.41 | 83.85 | 15.77 | NS | MS; RI |
| D7 | 1-Heptanol | 1458 | 16.59 | 53.03 | 10.66 | NS | MS; RI |
| D8 | (E)-2-Decen-1-ol | 1466 | ND | 2.02 | 0.59 | - | MS |
| D9 | 1,6-Octadien-3-ol, 3,7-dimethyl- | 1548 | 132.09 | ND | 14.72 | - | MS |
| D10 | 1-Octanol | 1560 | 20.13 | 51.94 | 10.36 | NS | MS; RI |
| D11 | 3-Cyclohexen-1-ol, 4-methyl-1-(1-methylethyl)- | 1608 | 66.87 | ND | 10.78 | - | MS |
| D12 | 2-Octen-1-ol, (E)- | 1615 | ND | 14.96 | 3.21 | - | MS |
| D13 | 1-Nonanol | 1659 | ND | 9.17 | 1.88 | - | MS; RI |
| D14 | 2-Furanmethanol | 1662 | 129.69 | ND | 16.03 | - | MS |
| D15 | Phenylethyl alcohol | 1918 | 15.40 | ND | 2.37 | - | MS |
| Subtotal | 474.89 | 323.32 | 64.76 | NS | |||
| Nitrogenous | |||||||
| E1 | Pyridine, 3-methyl- | 1222 | 20.30 | ND | 2.25 | - | MS; RI |
| E2 | Pyridine, 4-methyl- | 1299 | 14.45 | ND | 1.94 | - | MS; RI |
| E3 | Pyridine, 2,4-dimethyl- | 1333 | 26.18 | ND | 7.34 | - | MS; RI |
| E4 | Pyridine, 3,5-dimethyl- | 1344 | 14.49 | ND | 2.15 | - | MS; RI |
| E5 | Oxime-, methoxy-phenyl- | 1753 | 60.32 | ND | 6.50 | - | MS |
| Subtotal | 135.75 | ND | 17.39 | - | |||
| Furans | |||||||
| F1 | Furan, 2-pentyl- | 1230 | 9.70 | 29.32 | 6.84 | NS | MS; RI |
| F2 | Furfural | 1465 | 48.49 | ND | 6.21 | - | MS; RI |
| F3 | Ethanone, 1-(2-furanyl)- | 1508 | 65.66 | ND | 7.69 | - | MS |
| F4 | 2-Furancarboxaldehyde, 5-methyl- | 1577 | 11.65 | ND | 2.43 | - | MS; RI |
| Subtotal | 135.51 | 29.32 | 14.96 | ** | |||
| Esters | |||||||
| G1 | Ethyl acetate | 884 | 9.74 | ND | 1.83 | - | MS; RI |
| G2 | Hexanoic acid, ethyl ester | 1233 | 9.04 | ND | 1.45 | - | MS; RI |
| Subtotal | 18.78 | ND | 2.51 | - | |||
| Ethers | |||||||
| H1 | Estragole | 1675 | 32.84 | ND | 5.63 | - | MS; RI |
| H2 | Benzene, 1-methoxy-4-(1-propenyl)- | 1832 | 419.40 | ND | 94.98 | - | MS |
| Subtotal | 452.24 | ND | 100.30 | - | |||
| Phenols | |||||||
| I1 | Phenol, 2-methoxy- | 1863 | 244.17 | ND | 31.12 | - | MS; RI |
| I2 | Phenol, 4-methoxy-3-methyl- | 1876 | 20.98 | ND | 2.67 | - | MS |
| I3 | Phenol, 2,6-dimethyl- | 1912 | 19.61 | ND | 2.73 | - | MS |
| I4 | Phenol, 2-methoxy-4-methyl- | 1961 | 97.45 | ND | 12.48 | - | MS; RI |
| I5 | Phenol, 2-methyl- | 2005 | 105.39 | ND | 14.93 | - | MS |
| I6 | Phenol | 2008 | 257.57 | ND | 31.20 | - | MS; RI |
| I7 | Phenol, 4-ethyl-2-methoxy- | 2035 | 62.93 | ND | 8.14 | - | MS |
| I8 | Phenol, 2,5-dimethyl- | 2079 | 22.93 | ND | 3.51 | - | MS |
| I9 | Phenol, 4-methyl- | 2085 | 123.81 | ND | 15.49 | - | MS |
| I10 | Phenol, 3-methyl- | 2093 | 98.79 | ND | 13.58 | - | MS |
| I11 | Phenol, 3-ethyl- | 2180 | 25.15 | ND | 2.53 | - | MS |
| I12 | Phenol, 2,6-dimethoxy- | 2272 | 25.89 | ND | 3.20 | - | MS |
| Subtotal | 1104.68 | ND | 138.81 | - |
RI: Retention index.a: RI in agreement with literature values for a DB-WAX capillary column.SEM: Standard error of the mean. P-level: Level of significance found by analysis of variance. NS, not significant; *, P < 0.05; **, P < 0.01.ND: Not detected.
The profiles of volatile compounds were very different between the two types of samples. The volatile composition in the T1 product was: hydrocarbons (1645.99 ng/g), phenols (1104.68 ng/g), alcohols (474.89 ng/g), ethers (452.24 ng/g), aldehydes (307.81 ng/g), ketones (284.94 ng/g), nitrogenous (135.75 ng/g), furans (135.51 ng/g), and esters (18.78 ng/g). The major compounds present were: caryophyllene, phenol, 2-methoxy-phenol, eucalyptol, D-limonene, and nonanal. Similarly, hydrocarbons were reported as the dominant volatiles in Turkish pastirma (Kaban, 2009). In contrast, the volatile composition of the T2 product was: aldehydes (573.22 ng/g), alcohols (323.32 ng/g), hydrocarbons (292.17 ng/g), ketones (74.98 ng/g), and furans (29.32 ng/g).The major compounds present were: nonanal, hexanal, 1-octen-3-ol, octanal, 1-hexanol, and heptanal. The results were in agreement with those reported by Sha et al. (2016) for Kazakh dry-cured beef. In other dry-cured hams (Purriños et al., 2011; Marušić et al., 2014; Pugliese et al., 2015), aldehydes were also found to be the most abundant compound, whereas esters were the majorcompounds found in dry-cured foal (Lorenzo, 2014) and alcohols in San Daniele ham (Gaspardo et al., 2008).
Hydrocarbons were the most numerous of the chemical classes among the volatile components detected. In general, most aromatic hydrocarbons (toluene, ethylbenzene, o-xylene, p-xylene, and styrene), and long-chain aliphatic hydrocarbons (> 10 carbon atoms, undecane, dodecane, and tridecane) have been thought to originate from animal feedstuffs (Ruiz et al., 1999; Muriel et al., 2004). Numerous terpenes (α-pinen, β-phellandrene, β-pinene, β-phellandrene, 3-carene, ocimene, eucalyptol, and caryophyllene) were detected only in the T1 products, perhaps originating from added spices such as pepper and paprika (Ansorena et al., 2001; Kaban, 2009; Marušić et al., 2014). Naphthalene, an aromatic hydrocarbon found only in the T1 products, has been identified as being related to the smoking process (Muriel et al., 2004). Although D-limonene has been associated with animal feedstuffs, it has mainly been attributed to the added spices used in the preparation of dry-cured products (Ansorena et al., 2001). Therefore, the higher levels of hydrocarbons found in T1 products were mainly related to the use of spices with a higher content of terpenes. Compared with aliphatic hydrocarbons, aromatic hydrocarbons and terpenes have lower threshold values and make an important contribution to the flavor of dry-cured meat products (Ramírez and Cava, 2007; Kaban, 2009).
Aldehydes have very low flavor threshold values and play an important role in the formation of the flavor of dry-cured products (Purriños et al., 2011; Marušić et al., 2014; Pugliese et al., 2015). The straight-chain aliphatic aldehydes (hexanal, heptanal, octanal, (Z)-2-heptenal, nonanal, (E)-2-octenal, decanal, and (E)-2-nonenal) detected are typical products of lipid oxidation (Purriños et al., 2011), while branched aldehydes such as 3-methyl-butanal and 4-(1-methylethyl)-benzaldehyde could derive from Strecker degradation reactions of amino acids (Muriel et al., 2004). Aldehydes were present at relatively high levels in T2 products compared with T1 products, the difference in quantity not being significant (P > 0.05).
In the present study, ketones showed larger differences between the two types of samples both in kind and quantity. 4-Methyl ketones (2-pentanone, 2-octanone, 2, 3-octanedione, and 2-nonanone) were detected only in T2 products. They have also been found in other dried meats (Hierro et al., 2004; Marušić et al., 2014), and could originate from the oxidation of lipids (auto-oxidation or mould metabolism) (Muriel et al., 2004). 2-Pentanone, 2-octanone, and 2-nonanone characterize the typical aroma of blue cheeses and fruity aromas (García-González et al., 2008). Their absence in T1 products may be related to the smoking process, where antioxidants such as phenols can be formed which prevent the auto-oxidation of lipids (Jerković, 2010). Eight 2-cyclopenten- 1-one derivatives (2-methyl-2-cyclopenten-1-one, 2,3-dimethyl-2- cyclopenten-1-one, 3,4-dimethyl-2-cyclopenten-1-one, 2,3,4-trimethyl-2-cyclopenten-1-one, 3-methyl-2-cyclopenten-1- one, 3,4,4-trimethyl-2-cyclopenten-1-one, 3,5,5-trimethyl-2- cyclopenten-1-one, 3-ethyl-2-cyclopenten-1-one), 4-methyl-4- hepten-3-one, acetophenone, and 2,3-dihydro-1H-inden-1-one were detected only in T1 products. In previous studies (Ansorena et al., 2001; Hierro et al., 2004; Jónsdóttir et al., 2008; Stojković et al., 2015), some enones such as 2-methyl-2-cyclopenten-1-one, 2,3-dimethyl-2-cyclopenten-1-one, 3,4-dimethyl-2-cyclopenten-1- one, 3-methyl-2-cyclopenten-1-one, 3,5,5-trimethyl-2-cyclopenten- 1-one have also been reported in other smoked meat products, possibly formed during the smoking process. The ketones content was significantly (P < 0.05) higher in the T1 products than in the T2 products.
Regarding alcohols, most straight-chain alcohols (1-pentanol, 1-hexanol, 1-heptanol, 1-octanol, 1-nonanol, 1-penten-3-ol, 1-octen-3-ol, (E)-2-decen-1-ol, (E)-2-octen-1-ol) detected could have been produced by lipid oxidation, while branched alcohols, such as 3-methyl-1-butanol, could have arisen from the Strecker degradation of amino acids (Pugliese et al., 2015). Other alcohols such as ethanol, 3,7-dimethyl-1,6-octadien-3-ol, 4-methyl-1-(1- methylethyl)-3-cyclohexen-1-ol, 2-furanmethanol, phenylethyl alcohol were detected only in T1 products. They were most likely to have been derived from the smoking process or the use of spices in the T1 products. The alcohol content was not significantly (P > 0.05) different between the two types of sample. Because of their high sensory threshold values, most alcohols typically contribute little to the flavor of dry-cured meat products, but 1-penten-3-ol and 1-octen-3-ol have been considered as important flavor compounds (Marušić et al., 2014).
The nitrogenous compounds detected included 3-methyl-pyridine, 4-methyl-pyridine, 2, 4-dimethyl-pyridine, 3, 5-dimethyl-pyridine, and methoxy-phenyl-oxime. These compounds, found only in T1 products, were formed by Maillard compounds or nitrogen-containing proteins and amino acids pyrolysis during the smoking process (Stojković et al., 2015). 3-methyl-pyridine was also isolated from smoked dried meats (Hierro et al., 2004; Stojković et al., 2015).
Two esters (ethyl acetate and hexanoic acid ethyl ester) were detected only in T1 products. These esters are generated by esterification between ethanol and carboxylic acids in the presence of certain microorganisms (Montel et al., 1998). Since the processing time was relatively short, lower amounts of esters, an aroma of typical aged-meat products, were found in Kazakh dry-cured beef.
Two ethers (estragole and 1-methoxy-4-(1-propenyl)-benzene) were detected only in T1 products. They are most likely to have come from the added spices.
The furans detected included 2-pentyl-furan, furfural, 1-(2-furanyl)-ethanone, and 5-methyl-2-furancarboxaldehyde. 2-Pentyl-furan, a typical product of lipid oxidation, has been found in other dry-cured meats (Purriños et al., 2011; Lorenzo, 2014). Furfural, 1-(2-Furanyl)-ethanone, and 5-methyl-2-furancarboxaldehyde can also be generated during the smoking process (Jerković, 2010). 1-(2-Furanyl)-ethanone and 5-methyl-2- furancarboxaldehyde contribute greatly to the characteristic smoked aroma compared with furfural, a weak odorant (Jónsdóttir et al., 2008).
Phenols, typical flavor compounds in smoked meat products, have the characteristics of wood smoke derivatives from the pyrolysis of cellulose, hemicelluloses and lignin (Jerković, 2010). In the present study, 12 phenol derivatives were detected only in the T1 products. Among the phenols detected, phenol and 2-methoxy-phenol, present in very high quantities, have been identified as the main compound contributing to the smokehouse odor (Jónsdóttir et al., 2008).
Principal component analysis The results of the principal component analysis (Fig. 1 and 2) describe the relationships between the volatile compounds detected and illustrate the differences in the volatile flavors describing the samples. The first two principal components (PC1 and PC2) explained 37.83% and 20.61% of the total variance, respectively, accounting for 58.44% of the variability.
Score plot of principal component analysis of volatile componds from two types of Kazakh dry-cured beef.
The location of samples of two types of Kazakh dry-cured beef in the first principal components (PC1 and PC2). (●, T1 products; ▴, T2 products).
Fig. 1 shows the location of the variables in PC1 and PC2. PC1 was highly related to the volatile compounds generated during the smoking process: naphthalene, 2-cyclopenten-1-one derivatives, 4-methyl-4-hepten-3-one, acetophenone, 2, 3-dihydro-1H-Inden-1-one, 2-furanmethanol, methoxy-phenyl-oxime, furfural, 1-(2-furanyl)-ethanone, and phenols. The lipid-derived compounds, straight-chain aliphatic aldehydes, methyl ketone, straight-chain alcohols, and 2-pentyl-furan, were mainly correlated with PC2.
The location of the samples in PC1 and PC2 is shown in Fig. 2. The two types of Kazakh dry-cured beef are clearly separated by the PC1 axis. The T1 products were located on the positive axis on PC1, while the T2 products were located on the negative axis. These results indicate marked differences in volatile flavor between the two types of sample.
Overall, many factors are responsible for the flavor characteristics of Kazakh dry-cured beef: the added spices, smoking treatment, degree of lipid oxidation, compounds originating from animal feedstuff, the Strecker degradation of amino acids and microbial metabolism. Of these factors, the added spices, smoking treatment and lipid oxidation are the most important pathways for generating the volatile components in Kazakh dry-cured beef. The differences in the type or amount of volatile components between the two types of Kazakh dry-cured beef are related to the characteristics of the raw meat, the presence of spices, and the conditions during salting, drying, and smoking. Therefore, further studies on the factors influencing the flavor profile of Kazakh dry-cured beef are necessary for improving the quality control of this product.
The results of this study have provided a wider knowledge of the physicochemical and textural characteristics and volatile compounds of Kazakh dry-cured beef. This is important for controlling the quality of products and establishing product standards. The two types of Kazakh dry-cured beef exhibited significant (P < 0.05) differences in the mean values of aw, moisture, L*, cohesiveness, and chewiness. A total of 86 volatile compounds were identified and quantified by GC/MS. In Kazakh dry-cured beef, hydrocarbons were the most abundant compound in the T1 type and aldehydes in the T2 type. Smoke treatment, added spices and lipid oxidation were the main pathways for generating volatile components in Kazakh dry-cured beef. Principal component analysis showed that there were marked differences in volatile flavors between the two types of samples. The main differences were caused by the presence or absence of smoke derivatives and spice ingredients.
Acknowledgements This research was funded by the National Natural Science Funds of China (Grant No.31460403). The authors acknowledge the participation of the Animal Science Academy of Xinjiang Uygur Autonomous Region, China who collected the samples.
