Relationship between flavour deterioration and the volatile compound profile of semi-ripened sausage

Meat Science 93 (2013) 614–620

Contents lists available at SciVerse ScienceDirect

Meat Science journal homepage: www.elsevier.com/locate/meatsci

Relationship between flavour deterioration and the volatile compound profile of semi-ripened sausage José Manuel Lorenzo a, Mario Bedia b, Sancho Bañón b,⁎ a b

Centro Tecnológico de la Carne de Galicia, Rúa Galicia N° 4, Parque Tecnológico de Galicia, San Cibrán das Viñas, Ourense 32900, Spain Department of Food Technology, Nutrition and Hygiene, Veterinary Faculty, University of Murcia, Espinardo, Murcia 30071, Spain

a r t i c l e

i n f o

Article history: Received 10 July 2012 Received in revised form 18 October 2012 Accepted 10 November 2012 Keywords: Salami Flavour Odour Acceptability Volatile compounds Shelf life

a b s t r a c t This study provides data on the relationship between flavour deterioration and the volatile compound profile of semi-ripened pork salami kept under retail conditions for up to 150 days. The flavour of salami deteriorated for 120 days, resulting in rancidity and a loss of acceptability. TBARS increased from 0.16 to 0.57 MDA/kg. The flavour changes during the shelf life of salami were monitored from changes in the volatile profile. The retailing time influenced (pb 0.05) the level of 27 of the 30 headspace volatiles determined by SPME–GC/MS. Flavour deterioration was associated with the loss and/or degradation of volatiles resulting from spices and microbial activities, and the formation of volatiles from lipid oxidation. The levels of 2-heptenal and methyl esters of heptanoic, pentanoic and hexanoic acids were the best discriminators of storage time, and therefore seem to be promising as marker compounds of flavour deterioration and acceptability. © 2012 Elsevier Ltd. All rights reserved.

1. Introduction Dry-cured fermented sausages, such as salami, are stable for long periods of time at retail due to their low water activity and the use of preservatives, colourings, flavourings, antioxidants and acidifying cultures. The shelf life of salami is mainly limited by the sensory deterioration that accompanies oxidation phenomena, since pathogenic or spoilage bacteria have difficulties to proliferate in dry-cured sausages (Ordóñez, Hierro, Bruna, & de la Hoz, 1999). Previous studies reported that flavour deterioration precedes other signs of sensory spoilage in whole semi-ripened salami pieces kept at retail. As storage time increases, a gradual alteration of flavour takes place in the salami, causing loss of acceptability (Lee, Lee, Son, Choi, & Lee, 2009; Rubio et al., 2008; Summo, Caponio, & Pasqualone, 2006). Rancidity has been identified as the main cause of flavour deterioration in dry-cured sausage (Morrissey, Sheehy, Galvin, Kerry, & Buckley, 1998), while the possible formation of other off-flavours, such as mouldy, acid, putrid or pungent traits, seems to play a minor role in the loss of eating quality (Bedia, 2012). In contrast, the red-purplish colour of dry-cured meat due to nitrous-myoglobin formation is very stable, although a certain discolouration has been reported in vacuum-packed salami slices at retail (Rubio et al., 2008; Zanardi, Dorigoni, Badiani, & Chizzolini, 2002).

⁎ Corresponding author. Tel.: +34 868 888265. E-mail address: [email protected] (S. Bañón). 0309-1740/$ – see front matter © 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.meatsci.2012.11.006

The chemical pathways of flavour deterioration are complex and involve, on the one hand, the formation of compounds responsible for off-flavours and, on the other hand, the degradation of flavouring compounds from the mature meat seasoned with curing agents, pepper and other minor spices. Flavour will therefore correspond to the interaction between these two antagonist groups of aromatic compounds. Flavour changes during the storage of salami may be monitored through the volatile profile (Marco, Navarro, & Flores, 2008; Misharina, Andreenkov, & Vashchuk, 2001; Summo, Caponio, Pasqualone, & Gomes, 2011). The volatile compounds of sausage give an indication of the chemical and metabolic process that occurs during the display period. Several groups of volatile compounds have been reported in salami, such as aldehydes, ketones, alcohols, esters, alkanes and terpenes, among others. These volatiles can be grouped according to their possible origin, into volatiles from spices (terpenes and aliphatic sulphur compounds), lipid autooxidation (aldehydes, hydrocarbons, alcohols and ketones), microbial esterification (e.g. propyl acetate and ethyl propanoate), carbohydrate fermentation (e.g. 1,3-butanediol and phenyl acetaldehyde), amino acid catabolism (e.g. 2-methylbutanal, 2-methyl1-butanol and 2,3-butanediol) and other origins (Andrade, Cordoba, Casado, Córdoba, & Rodríguez, 2010; Lorenzo, Montes, Purriños, & Franco, 2012; Meynier, Novelli, Chizzolini, Zanardi, & Gandemer, 1999; Olesen, Meyer, & Stahnke, 2004; Summo et al., 2011), although several of these volatiles may have more than one origin. Rancidity in dry-cured sausage has been associated with the formation of aldehydes, ketones and carboxylic acids through lipid autooxidation, and, to a lesser extent, with the production

J.M. Lorenzo et al. / Meat Science 93 (2013) 614–620

of free amino acids such as lysine, tyrosine or aspartic acid (Morrissey et al., 1998; Ordóñez et al., 1999). Lipid oxidation largely takes place by reaction with the oxygen occluded in the sausage, depending on fat composition and the balance between pro-oxidant and antioxidant agents. The oxidation reactions of salami can be inhibited by means of several strategies, such as using antioxidants or vacuum or protective gases during mincing, stuffing or packing; however, meat oxidation continues during storage at a rate that is dependent on the temperature and the presence of fluorescent lighting in the display case. Some types of low-size salami are retailed as aerobically packed pieces to preserve their typical mould covering. The prooxidizing conditions responsible for flavour deterioration will be not be the same in whole pieces, which are gradually dehydrating and which are less exposed to lighting than sausage slices. The objective of this study was to determine the relationships between flavour deterioration and the volatile profile in a semi-ripened salami retailed as whole pieces. The possibility of monitoring flavour deterioration and shelf life through changes in the volatile profile during storage was also evaluated. 2. Material and methods 2.1. Sausage manufacture and sampling Three different batches of salami were manufactured by a local company (Elaborados Cárnicos de Lorca, Murcia, Spain) using the following recipe (g/kg): boned pork shoulder (880), water (44), sodium chloride (22), black and white pepper (10), dextrose, lactose and sucrose (20), dextrin (20), potassium nitrate (0.15) and sodium nitrite (0.15), sodium isoascorbate (0.5), sodium citrate (0.3) sodium glutamate (2.5) and Ponceau 4R red (0.2). The commercial mixture of additives and spices was provided by Cargill Texturizing Solutions (Barcelona, Spain). The meat was minced in an atmospheric mincer using a 6 mm plate (Laska GMBH, WW1302, Nu-Meat Technology, Girona, Spain). A commercial starter culture composed of (g per 100 g culture) Pediococcus pentosaceus (50), Staphylococcus carnosus (25) and Staphylococcus xylosus (25) was provided by Degussa Ferment's Aromatization SAS (La Ferté sous Jouarre, France). The lyophilised cultures were rehydrated (15 g in 200 mL chlorine-free water) for 4 h and then sown in the mass at a rate of 6 × 10 7 CFU/g. The meat was then mixed with the starter cultures, additives and spices in an AMU102 vacuum mixer (Maquinaria Vall, Miralcamp, Lleida, Spain). The paste was stuffed into the casing on a WF-612 automatic line (Albert Handtmann Mahcinefabrik, Biberach an der Riss, Germany). Swine casing, slightly curved, 40–43 mm calibre and 300–320 mm in length, was used. The casing was previously desalted and washed with chlorinated-free water. The recently stuffed pieces were bathed in mould (Penicilium chrysogenum PS5.1, Cargill Texturizing Solutions Cultures SAS, La Ferté Sous Jouarre, France) solution (0.8 g/L water) and hung from steel racks 1.2 m wide, 1.2 m deep and 2.2 m high. The loading density on the trolleys was 19 kg/m 3. The trolleys were placed in an air-drying room (Sabroe S.A., Barcelona, Spain) set at 15 ± 1 °C during the whole drying stage. The relative humidity (RH) was gradually reduced after the first day at ambient humidity to eliminate water through dripping, 6 days at 80±5% RH and 5 days at 70± 5% RH. The temperature and RH were verified using a P 650 thermohygrometer (Dostmann Electronic GmbH, WertheimReicholzheim, Germany) with a precision of 0.2 °C and 0.5% RH. The average proximal composition (g per 100 g) of fresh-made sausage (day 0) was: moisture (33.3 ± 1.3), total protein (22.4 ±0.5), total lipids (32.0 ± 0.6), and ash (5.1±0.2). The average values of pH and water activity of fresh-made sausage were 5.1 ± 0.0 and 0.87 ± 0.02, respectively (Bedia, 2012). To study shelf life, the salamis were packaged in perforated (6 mm diameter) polypropylene BA-85 bags (Plásticos

615

Sierra del Oro, Abarán, Murcia, Spain) and then stored at 10 ± 1 °C and 65± 0.2% R.H. for 0, 30, 60, 90, 120 and 150 days in an open display cabinet (Booster Group, Santiago, Chile) continuously illuminated with white fluorescent light (1000 lx), simulating retail display conditions. 2.2. Chemical analysis Volatile compounds were extracted using solid-phase microextraction (SPME). A SPME device (Supelco, Bellefonte, PA, USA) containing a fused-silica fibre (10 mm length) coated with a 50/30 μm layer of DVD/CAR/PDMS was used. The fibre was conditioned prior to analysis by heating it in a gas chromatograph injection port at 270 °C for 60 min. Extraction was performed at 35 °C for 30 min. Before extraction, samples were equilibrated for 15 min at the temperature used for extraction. Once sampling was finished, the fibre was withdrawn into the needle and transferred to the injection port of the gas chromatograph–mass spectrometer (GC–MS) system. Volatiles were analyzed in duplicate in all the dry-cured sausages. Analyses were performed on a Hewlett-Packard 6890N Series GC gas chromatograph fitted with a HP 5973N mass spectrometer and a MSD Chemstation (Hewlett-Packard, Palo Alto, CA, USA). A split/splitless injection port, held at 260 °C, was used to thermally desorb the volatiles from the SPME fibre onto the front of the DB-624 capillary column (J&W scientific: 30 m×0.25 mm id, 1.4 μm film thickness). The injection port was in splitless mode, the split valve opening after 2 min. Helium was used as a carrier gas with a linear velocity of 36 cm/s. The temperature programme was as follows: 40 °C, maintained for 2 min and then raised to 100 °C at 3 °C/min; then from 100 to 180 °C at 5 °C/min, and from 180 to 250 °C at 9 °C/min, with a final holding time of 5 min; total run time 50.8 min. The mass spectra were obtained using a mass selective detector working in electronic impact at 70 eV, with a multiplier voltage of 1953 V and collecting data at a rate of 6.34 scans/s over the range m/z 40–300. Compounds were identified by comparing their mass spectra with those contained in the NIST05 (National Institute of Standards and Technology, Gaithersburg) library and/or by calculation of the retention index relative to a series of standard alkanes (C5–C19) (for calculating Kovats indexes, Supelco 44585-U, Bellefonte, PA, USA) and matching them with data reported in the literature. The results are reported as relative abundance expressed as total area counts (AU×106). Lipid oxidation was determined as thiobarbituric acid reactive substances (TBARS) according to Botsoglou et al. (1994). To prepare the standard used, 73.2 mg 1,1,3,3-tetraethoxypropane (TEP) was diluted with 10 mL of 0.1 N HCl, immersed in a boiling water bath for 5 min, and quickly cooled under tap water. A stock solution of malondialdehyde (239 μg MDA/mL) was prepared by transferring the hydrolyzed 1,1,3,3-tetraethoxypropane (TEP) solution into a 100 mL volumetric flask and diluting with water. Working MDA solutions (2.39 μg/mL) were prepared by pipetting a 1 mL aliquot of the stock solution into another 100 mL volumetric flask and diluting to volume with water. Samples (2 g) were transferred to a 50 mL centrifuge tube, and volumes of 5% w/w aqueous trichloroacetic acid (TCA) (8 mL) and 0.8% butylated hydroxytoluene in hexane (5 mL) were successively added. The content of the tube was homogenized with a Silent Crusher for 1 min at high speed (9500 rpm) and centrifuged for 10 min at 3000 rpm. The top hexane layer was discarded. The bottom aqueous layer was made to a 10 mL volume with 5% TCA, and a 2.5 mL aliquot was pipetted into a screw-capped tube to which a volume 1.5 mL of 0.8% aqueous TBA was also added. Following incubation for 30 min at 70 °C, the tube was cooled under tap water; the reaction mixture was submitted to third-derivative spectrophotometry against blank reaction mixture. Aliquots of standard solutions were pipetted into screw-capped tubes and diluted to 2.5 mL volume with 5% TCA. A 1.5 mL volume of 0.8% TBA was added to each tube, and the reaction

616

J.M. Lorenzo et al. / Meat Science 93 (2013) 614–620

was carried out as prescribed. The MDA content (expressed as mg “MDA”/kg) was calculated as follows: MDA ¼

3D−0:0788 10  4:5451 W

3. Results

where: 3D W

from the previous FtA, a canonical discriminant analysis (CDA) was developed to assess the salami units according to storage time.

is the value of the third derivate in the sample and is the weight (g) of the sample

Colour measurements were made using a UV2 spectrophotometer (Pye Unicam, Cambridge, United Kingdom). The spectrophotometer wavelength used for the colour measurements was 532 nm. All the chemicals were provided by Sigma Chemical Co. Aldrich (St. Louis, USA). 2.3. Sensory analysis A quantitative descriptive sensory analysis (QDA) was carried out according to ISO 4121 (2003). Slices (8 mm thick) were cut with a MAS9101 slicer (Rober Bosch Hawgeräte GmbH, Ljubljana, Slovenia). Ten panellists were selected and trained according to ISO 8586-1 (1993). Most of the panellists had previous experience in QDA of dry-cured fermented sausage. There were six training sessions. In the first two sessions, the main sensory descriptors of freshly-made salami were identified and quantified in different samples; the next two sessions were concerned with identifying and selecting the sensory attributes related with off-odours and off-flavours. Samples of spoiled salami previously stored for six months at retail was used as references. The final two sessions were concerned with quantifying odour and flavour using an intensity scale from 1 (minimum) to 5 (maximum). Finally, the panel scored the acceptability using a five-point numerical scale. The maximum score corresponded to freshly-made salami with all its odour and flavour intact, while the minimum score corresponded to the above mentioned spoiled salami. 2.4. Statistical analysis The statistical model design was completely random, and the time of storage was considered as the main treatment. Nine salami units per retailing time were analyzed in duplicate (n = 54). The statistical analysis was made with the SPSS package (SPSS 19.0, Chicago, IL, USA). One way analysis of variance was used to analyze the effect of retailing time on the dependent variables studied. The least squares mean were separated using Duncan's t-test. All LSM statistical tests were performed with a significance level b0.05. Linear correlations between variables were determined as Pearson's coefficients. A factor analysis (FtA) was made to obtain a reduced number of principal components to explain the variability of the selected compounds. The relationship among the variables was represented by a principal component analysis (PCA). Using a new matrix of data integrated by the standardised reduced original variables, which were selected

Table 1 shows the effect of the retailing time on the odour, flavour, acceptability and lipid oxidation (expressed as TBARS) of pork salami kept for up to 150 days. The odour and flavour scoring from the sensory evaluation of salami provided similar information. The odour and flavour of salami gradually deteriorated as the storage time increased, compared with the freshly-ripened salami (day 0), while significant (p b 0.05) decreases in both odour and flavour intensity were noted from day 120. Coinciding with flavour deterioration, rancid odour and flavour were scored as imperceptible (day 0), detectable (p b 0.05) (day 60) and finally moderate (day 150). TBARS significantly increased from 0.16 (day 0) to 0.6 MDA/kg (days 120 and 120). These MDA levels suggest a modest degree of lipid oxidation in the salami kept at retail for up to 5 months, although rancidity was detected earlier by the sensory panel. Flavour deterioration associated with moderate rancidity resulted in a clear loss of acceptability of the salami, which decreased from acceptable (day 0) to unacceptable (day 150). The salami lost more than half of its initial acceptability after 120 days of storage, indicating the end of its shelf life (Bedia, 2012). Table 2 shows the effect of retailing time on the volatile compounds of pork salami kept for up to 150 days. A total of 30 volatile compounds were extracted by SPME and identified by GC/MS in the headspace during the storage time. The retailing time influenced (p b 0.05) the level of 27 volatile compounds present in the headspace at the end of storage, which proceeded from spices (54% of integrated area), lipid autooxidation (32.5%), carbohydrate fermentation (less than 10%) and benzene compounds (around 5%). Great variability in the volatile profile was observed in different salami units analyzed at the same control time, although it should be noted that three different industrial batches of salami were being sampled. The volatiles resulting from spices found in the headspace were: 1-methyl-4-cyclohene, copaene, limonene, 3-carene, β-caryophyllene and α-pinene. Except the last one, all these compounds decreased (pb 0.05) during retailing; for example, the initial relative areas (day 0) of both 1-methyl-4cyclohene and copaene were reduced by 75% at day 120, when flavour deterioration was clearly detected by the panel. This suggests that a noticeable loss and/or degradation of aromatic compounds from the pepper and garlic used takes place in the salami under retail display conditions. The volatiles found from microbial esterification were methyl esters of butanoic, hexanoic, heptanoic, pentanoic and decanoic acids. All these methyl esters also decreased (pb 0.05) during storage; for example, the initial relative area of heptanoic acid methyl ester fell by two thirds by day 120. The relative area of 2,3-butanediol derived from carbohydrate fermentation increased from day 0 to day 60 and then decreased to the same levels of day 0. These decreases in the levels of volatile compounds as a result of microbial activities may be explained by the gradual reduction of fermentative activity in the salami.

Table 1 Effect of storage time on the odour, flavour, acceptability and lipid oxidation (TBARS) of whole salami pieces kept at retail up to 150 days. Storage days

Odour Flavour Rancid odour Rancid flavour Acceptability TBARS

0

30

60

90

120

150

2.93 ± 0.78c 3.15 ± 0.73c 1.05 ± 0.28a 1.07 ± 0.28a 4.75 ± 0.35c 0.16 ± 0.02a

2.93 ± 0.69c 3.14 ± 0.59c 1.25 ± 0.46a 1.41 ± 0.55a 4.39 ± 0.24c 0.17 ± 0.07a

2.91 ± 0.71c 2.88 ± 0.77c 1.79 ± 0.75b 2.16 ± 0.93b 3.38 ± 0.89b 0.34 ± 0.19ab

2.86 ± 0.86bc 2.89 ± 0.99c 2.01 ± 0.94b 2.33 ± 1.07b 3.41 ± 0.30b 0.36 ± 0.39ab

2.55 ± 0.79b 2.39 ± 0.76b 2.55 ± 0.91c 2.95 ± 0.94c 2.36 ± 0.19a 0.61 ± 0.31b

1.95 ± 0.68a 1.86 ± 0.71a 3.11 ± 0.92d 3.43 ± 0.92c 2.14 ± 0.37a 0.57 ± 0.30b

Results are expressed as mean ± standard deviation. Means with different superscripts in the same row are different at p b 0.05.

J.M. Lorenzo et al. / Meat Science 93 (2013) 614–620

617

Table 2 Effect of storage time on the volatile profile determined by SPME–GC–MS of whole salami pieces kept at retail up to 150 days. Volatile compounds

KI

R

Storage days 0

30

60

90

120

150

Spices α-Pinene 3-Carene Limonene Copaene β-Caryophyllene 1-Methyl-4-cyclohexene

971 1058 1078 1349 1380 1133

m m m m m m

0.00 ± 0.00a 30.06 ± 18.08b 30.69 ± 4.97d 103.64 ± 17.02c 1.39 ± 0.36c 193.43 ± 40.99c

5.57 ± 8.46b 16.63 ± 5.98a 26.30 ± 5.61c 84.10 ± 10.70b 1.03 ± 0.27ab 159.17 ± 15.92b

0.00 ± 0.00a 19.64 ± 10.67a 25.26 ± 4.34c 87.00 ± 14.51b 1.16 ± 0.16bc 171.04 ± 32.85bc

0.00 ± 0.00a 15.95 ± 7.46a 22.26 ± 4.06bc 85.89 ± 12.36b 1.09 ± 0.19ab 165.56 ± 31.32bc

0.00 ± 0.00a 18.78 ± 6.25a 18.31 ± 3.42ab 77.49 ± 8.52b 0.98 ± 0.20ab 145.26 ± 20.96ab

0.00 ± 0.00a 15.22 ± 2.17a 14.73 ± 3.13a 64.71 ± 12.57a 0.89 ± 0.21a 124.97 ± 32.17a

Microbial esterification Butanoic acid, methyl ester Hexanoic acid, methyl ester Heptanoic acid, methyl ester Pentanoic acid, methyl ester Decanoic acid, methyl ester 2,3-Butanediol

715 1001 1092 846 1315 926

m k, m k, m m m k, m

15.36 ± 6.36a 42.55 ± 15.47e 140.96 ± 36.28c 48.66 ± 18.59c 27.49 ± 10.77abc 6.61 ± 6.01a

40.96 ± 25.88b 33.09 ± 4.51d 140.53 ± 15.75c 41.33 ± 6.84c 44.88 ± 11.26bc 10.82 ± 6.48ab

61.89 ± 51.24b 15.82 ± 3.73c 125.19 ± 54.09c 13.15 ± 4.20b 48.35 ± 39.87c 17.74 ± 9.34c

7.26 ± 8.71a 7.53 ± 2.66ab 89.74 ± 51.13b 3.87 ± 1.59a 55.88 ± 67.77c 13.40 ± 4.60bc

0.00 ± 00a 10.06 ± 7.24bc 46.74 ± 20.26a 0.82 ± 1.40a 12.92 ± 8.58ab 8.87 ± 5.94ab

0.00 ± 00a 1.26 ± 2.00a 22.91 ± 20.97a 0.00 ± 0.00a 4.07 ± 6.27a 8.69 ± 3.57ab

Lipid autoxidation Pentanal Hexanal Heptanal Decanal Octane Nonane 2-Heptenal (Z) Decane Undecane Tetradecanoic acid Cyclohexadecane Dodecane 2-Pentylfuran 2-Nonanone

740 842 991 1245 800 900 1057 1000 1100 1538 1571 1200 1055 1162

k, m k, m k, m k, m k, m k, m k, m k, m k, m k, m m k, m k, m k, m

0.00 ± 0.00 35.88 ± 5.87a 4.92 ± 2.29a 4.82 ± 1.54c 0.00 ± 00a 1.62 ± 1.66b 0.00 ± 0.00a 4.42 ± 3.80 15.70 ± 6.25d 2.92 ± 1.71a 44.26 ± 10.55c 5.05 ± 2.70c 12.13 ± 6.71b 2.63 ± 1.03a

1.28 ± 1.76 54.02 ± 17.37ab 2.88 ± 2.20a 3.88 ± 2.98c 5.17 ± 1.25a 0.00 ± 0.00a 1.22 ± 1.86a 5.45 ± 1.98 10.72 ± 3.96bc 4.66 ± 2.04ab 34.58 ± 5.62b 3.57 ± 2.48abc 6.36 ± 1.31a 0.94 ± 0.09b

1.62 ± 2.43 94.08 ± 36.27d 6.13 ± 4.81b 0.00 ± 0.00a 2.50 ± 01.96 0.00 ± 0.00a 4.91 ± 3.95b 3.19 ± 1.39 5.41 ± 2.06a 5.58 ± 4.44ab 36.32 ± 4.97b 1.85 ± 0.80a 8.61 ± 4.66a 1.20 ± 0.75b

1.21 ± 1.83 86.83 ± 19.54cd 7.10 ± 2.12b 0.00 ± 0.00a 4.15 ± 0.67a 0.44 ± 0.67a 6.05 ± 2.01b 4.40 ± 1.55 9.73 ± 2.94bc 10.41 ± 14.23c 34.01 ± 5.33b 2.76 ± 1.43ab 6.99 ± 3.28a 1.28 ± 0.67b

0.00 ± 0.00 71.60 ± 19.89bc 6.68 ± 1.78b 1.64 ± 1.25b 11.30 ± 13.11b 0.39 ± 0.59a 4.71 ± 1.02b 5.88 ± 2.33 12.14 ± 3.90cd 1.28 ± 1.24a 30.68 ± 5.09ab 4.28 ± 1.99bc 6.14 ± 1.33a 0.85 ± 0.67b

0.64 ± 0.98 63.76 ± 23.57b 5.78 ± 2.22b 0.58 ± 0.89ab 5.31 ± 4.94a 0.00 ± 0.00a 8.65 ± 1.64c 3.52 ± 2.21 7.63 ± 4.48ab 0.35 ± 0.54a 27.18 ± 4.67a 2.85 ± 2.03ab 5.56 ± 1.14a 0.72 ± 0.56b

Benzene compounds p-Xylene Methylbenzene o-Xylene Ethylbenzene

941 1144 969 932

k, m m k, m k, m

5.28 ± 0.99c 0.00 ± 0.00a 6.62 ± 8.27 27.07 ± 4.21c

3.72 ± 0.78b 0.97 ± 1.51ab 4.12 ± 1.97 20.84 ± 3.05b

3.39 ± 0.99b 1.84 ± 1.27b 2.29 ± 2.40 20.76 ± 5.73b

3.01 ± 0.60ab 4.42 ± 2.42c 2.86 ± 2.08 20.61 ± 4.70b

3.03 ± 0.54ab 3.66 ± 1.79c 2.34 ± 0.69 17.45 ± 3.44ab

2.60 ± 0.31a 1.90 ± 0.88b 2.39 ± 1.02 13.69 ± 3.21a

Results as expressed in arbitrary area units (×106). Results are expressed as mean ± standard deviation. Means with different superscripts in the same row are different at p b 0.05. KI: Kovats index calculated for DB-624 capillary column (J&W scientific: 30 m × 0.25 mm id, 1.4 μm film thickness) installed on a gas chromatograph equipped with a mass selective detector. R: Reliability of identification: k: Kovats index in agreement with; m: mass spectrum agreed with mass database (NIST05).

A total of 14 volatiles resulting from lipid autooxidation were found, representing the most numerous group of volatiles present in the headspace. In general, most of these volatile increased as storage time increased in the salami kept at retail. As regards aliphatic aldehydes, the relative area of hexanal increased from day 0 to day 90 and then slightly decreased until the end of the display period. A similar pattern was found for pentanal and decanal, while the relative areas of heptanal and 2-heptenal increased throughout storage. The effect of retailing time on the level of aliphatic alkanes was not clear, since no relevant (p b 0.05) differences in the relative area of octane, decane, undecane and dodecane were found between samples at days 0 and 120, while the relative area of nonane was lower at day 120 than at day 0. The relative area of both 2-pentylfuran and nonanone was decreased by day 30 after which remained constant during the rest of the storage period. Surprisingly, no volatile alcohols were found in the headspace of salami at any control time, while tetradecanoic acid, an acid derived from the oxidation of the corresponding aldehyde, reached its maximum relative area after 90 days and then decreased. Finally, four volatile compounds from benzene were also found in the headspace. The relative area of p-xylene and ethylbenzene decreased throughout storage, while the decrease in the relative area of o-xylene was not significant (p > 0.05), probably due to its great variability.

Table 3 shows the Pearson's correlation coefficients between the retailing time and sensory traits vs. volatile profile and TBARS. The values of 23 of the 30 volatile relative areas were correlated (pb 0.05) with the retailing time. As the display time was strongly correlated with sensory scoring, most of these 23 volatile compounds were also correlated with odour and flavour scores, rancidity and acceptability, although with some differences. The correlation coefficients for odour and flavour provided similar information. Odour and flavour were positively correlated with the volatiles resulting from microbial esterification and, to a lesser extent, from spices, ethylbenzene and other minor compounds, while flavour deterioration was correlated with increases in TBARS, 2-heptenal, heptanal, nonane and hexanal. Rancidity was clearly associated with the reduction of volatiles resulting from microbial esterification, spices, benzene compounds and decanal, together with increases of TBARS, 2-heptenal, heptanal, hexanal, octane, nonane and undecane. Increases in heptanoic acid methyl ester, pentanoic acid methyl ester and hexanoic acid methyl ester and 2-heptenal were the best indicators of the flavour deterioration and rancidity of salami. The acceptability score was closely dependent on odour and flavour, since it showed similar correlations to those observed for these variables. A factor analysis was made to obtain a reduced number of principal components explaining the variability of selected volatiles. When this data set was used, the first three principal components were

618

J.M. Lorenzo et al. / Meat Science 93 (2013) 614–620

Table 3 Pearson's correlations between the storage time, sensory attributes, volatile compounds and lipid oxidation (TBARS). Rancid flavour

Odour

R

p

R

p

R

p

R

0.35 −0.34 −0.04 −0.31 −0.18 −0.03 0.31 0.27 0.76 −0.57 −0.52 0.00 −0.39 −0.86 −0.85 −0.77 −0.27 −0.13 −0.21 −0.33 −0.41 −0.78 −0.53 −0.62 −0.47 −0.61 −0.67 −0.69 −0.33 0.47 −0.60 −0.81 0.00 0.07 −0.55 0.49

** *

0.36 −0.33 −0.05 −0.27 −0.15 0.02 0.32 0.34 0.67 −0.48 −0.40 0.03 −0.36 −0.80 −0.81 −0.70 −0.18 −0.03 −0.19 −0.26 −0.30 −0.68 −0.44 −0.54 −0.42 −0.53 −0.59 −0.60 −0.25 0.46 −0.51 −0.71 0.03 0.14 −0.44 0.46

** *

0.33 −0.37 −0.05 −0.30 −0.18 0.19 0.42 0.39 0.66 −0.51 −0.33 0.02 −0.37 −0.80 −0.79 −0.61 −0.03 0.10 −0.23 −0.16 −0.21 −0.63 −0.37 −0.49 −0.40 −0.54 −0.52 −0.59 −0.21 0.52 −0.43 −0.64 0.02 0.23 −0.36 0.51

** **

−0.14 0.26 0.03 0.12 0.05 −0.15 −0.17 −0.33 −0.49 0.28 0.14 0.13 0.41 0.48 0.53 0.46 0.04 −0.05 −0.15 −0.00 0.02 0.46 0.28 0.38 0.28 0.40 0.37 0.37 0.00 −0.19 0.31 0.49 0.13 −0.01 0.23 −0.44

Storage time

Octane Nonane Decane Undecane Dodecane Pentanal Hexanal Heptanal 2-Heptenal (Z) Decanal 2-Nonanone Butanediol Butanoic acid methyl ester Pentanoic acid methyl ester Hexanoic acid methyl ester Heptanoic acid methyl ester Decanoic acid methyl ester Tetradecanoic acid Pinene Caren Pentylfuran Limonene Cyclohexene Copaene Caryophyllene Cyclohexadecane Ethylbenzene p-Xylene o-Xylene Benzene 1 methyl Spice Microbial esterification Carbohydrate Lipid oxidation Benzene compounds TBARS

Rancid odour

*

* * *** *** *** ** *** *** *** *

* ** *** *** *** *** *** *** *** * *** *** ***

*** ***

*

* * *** *** ** ** *** *** ***

* *** *** *** ** *** *** *** *** *** ***

*** ***

*

** ** *** *** * ** *** *** ***

*** ** *** ** *** *** *** *** *** ***

** ***

Flavour p

* ***

** *** *** ***

*** * ** * ** ** **

***

**

R −0.23 0.31 0.06 0.22 0.12 −0.20 −0.27 −0.37 −0.58 0.35 0.19 −0.10 0.39 0.62 0.62 0.51 0.03 −0.06 0.23 0.02 0.07 0.53 0.30 0.41 0.32 0.46 0.43 0.44 0.06 −0.31 0.34 0.54 0.10 −0.15 0.28 −0.50

Acceptability p *

* ** *** **

** *** *** ***

** ** * *** ** ** * * ***

* ***

R

p

−0.36 0.35 0.04 0.30 0.19 −0.23 −0.44 −0.44 −0.66 0.49 0.31 0.01 0.36 0.80 0.76 0.56 0.01 0.10 0.24 0.07 0.13 0.57 0.31 0.43 0.36 0.52 0.47 0.57 0.14 −0.49 0.37 0.60 −0.01 −0.26 0.31 −0.55

** ** *

*** *** *** *** * ** *** *** ***

*** * *** ** *** *** *** *** ** ***

* ***

Correlation coefficients are significant for: ***p b 0.001; **p b 0.01; or *p b 0.05.

chosen (77.34% of the total variance). All the variables presented a communality higher than 0.790, which indicated that they are well represented by the three factors. A Varimax rotation was carried out to minimise the number of variables that influence each factor, and so facilitate the interpretation of the results. The first principal component (PC1), which explained 40.11% of total variance, was positively correlated with microbial esterification (pentanoic acid, methyl ester and hexanoic acid, methyl ester) and sensory characteristics (acceptability, odour and flavour) and negatively correlated to hexanal and rancidity. The PC2, which explained 23.85% of the total variance, was positively related with heptanoic acid, methyl ester, 1methyl-4-cyclohexene, copaene and ethylbenzene. PC3, with 25.69% of the total variance explained, was related with undecane and dodecane. Fig. 1 shows the projection of the variables in the rotated space defined by the three first principal components. Finally, a canonical discriminant analysis (CDA) was developed to assess the salami units. The data set was subjected to the CDA according to the storage time (T0, T30, T60, T90, T120, and T150). From the data set (15 parameters chosen after FtA) subjected to discriminant analysis, 6 variables (undecane, pentanoic acid methyl ester, ethylbenzene, 2-heptenal, acceptability and flavour) were retained at the end of the stepwise discriminant analysis and were linearly combined to form canonical discriminant functions. The two first canonical discriminant functions obtained were: (1) CAN1 = 0.262 [Decane] − 0.916 [Undecane] + 0.378 [Butanal] + 0.714 [Pentanoic acid Methyl ester]+1.105 [Ethylbenzene]−0.348 [2-Heptenal]+0.758 [Acceptability]+0.188 [Flavour]; and (2) CAN2=−0.745 [Decane] −0.508 [Undecane]+1.014 [Butanal]−0.958 [Pentanal]+1.259 [Ethylbenzene]−0.244 [2-Heptenal]−0.082 [Acceptability]+0.840 [Flavour].

Table 4 shows the main statistics of the two first discriminant functions. The two first CAN together accounted for 91.9% of the total variance, with CAN1 explaining 72.3% of total variability and CAN2 explaining 19.6%. The Wilks' lambda, which is a measurement of how well each function separates individuals (storage time) into groups, had low values, indicating the great discriminatory ability of the CANs. When the results

Fig. 1. Projection of the dependent variables (sensory attributes and volatile compounds) in the three first principal components calculated by means of factor analysis.

J.M. Lorenzo et al. / Meat Science 93 (2013) 614–620

619

Table 4 Discriminant analysis. Main statistics of the canonical discriminant functions calculated from 15 variables chosen after factor analysis. Canonical function

Eigenvalue

% variance

Canonical correlation

Wilks' lambda

Chi-square

Degree of freedom

p

CAN 1 CAN 2

22.5 6.1

72.3 19.6

0.98 0.93

0.001 0.027

312.0 166.8

40 28

*** ***

Level of significance: ***p b 0.001.

obtained from CAN1 were plotted against the results obtained from CAN2 on the coordinate for each display time, there was a good discrimination among groups according to the storage time (see Fig. 2). CAN1 allowed segregation of the freshly-made T0, T30 and T60 salamis from T90, T120 and T150 salamis, while CAN2 allowed segregation of T60 and T90 salamis from the rest of the salami groups. Finally, the discriminant analysis correctly attributed each storage time to its original group with an accuracy of 98% (53 of 54 salami units correctly classified). A sole T30 salami unit was assessed as T0 by CDA. 4. Discussion The flavour deterioration that occurs during the storage of salami has also been discussed by other authors (Lee et al., 2009; Misharina et al., 2001; Rubio et al., 2008; Summo et al., 2006) and has been seen to result in a gradual loss of acceptability. Both sensory and volatile data suggest that the flavour of freshly-made salami was not fully developed. A sausage ripened for only 12 days would present a weak flavour of ripened meat, although the development of flavour from lactic acidification is fast, lipid and protein oxidation and hydrolysis require more time for the salami to develop its flavour of ripened meat (Martínez, Bedia, Méndez, & Bañón, 2009). This would perhaps explain the low number of volatile compounds identified in the headspace of semi-ripened salami compared to those reported in other studies of dry-cured sausages (Andrade et al., 2010; Hierro et al., 2005; Lorenzo et al., 2012; Marco, Navarro, & Flores, 2007; Meynier et al., 1999; Misharina et al., 2001; Montel, Reitz, Talon, Berdagué, & Rousset, 1996; Olesen & Stahnke, 2000; Spaziani, del Torre, & Stecchini, 2009; Summo et al., 2011). However, many aromatic compounds often related with dry-cured fermented meat were present in the semi-ripened salami, although other common volatiles, such as octenal (dry-cured sausage), 1-octe-3-nol (mushroom), 2-nonenal (rancid, woody), alcohols and others were not found (Marco et al., 2007; Summo et al., 2011). The flavour of salami remained stable for around 100 days and then strongly deteriorated when incipient rancidity indicated that oxidation phenomena had taken place inside

Fig. 2. Discriminant analysis. Projection of the salami units grouped according to their storage time in the two first canonical discriminant functions.

the whole pieces of salami. No flavour deterioration was observed in a semi-ripened (for 15 days) smoked salami which was kept unpacked for 40 days (Summo et al., 2006). Flavour deterioration seems to be due to the loss and/or degradation of volatiles resulting from spices and microbial activity, along with the increase in some compounds from lipid oxidation, since benzene compounds are contaminants of SPME–GC materials (Meynier et al., 1999). Summo et al. (2011) reported decreases in the level of 3-carene, limonene, β-pinene, α-pinene and other minor volatile terpenes during the storage of Italian salami; however, Marco et al. (2008) and Misharina et al. (2001) found the opposite results. Terpenes can result from animal feedstuffs, but they mainly come from the spices used in dry-cured sausage production, such as black pepper (α-pinene, β-caryophyllene, 3-carene, limonene, and β-pinene) and garlic (Guadayol, Caixach, Cabañas, & Rivera, 1997). The volatiles derived from microbial activities (involving carbohydrate fermentation and ethyl esters) were often detected in fermented sausages (Ansorena, Gimeno, Astiasarán, & Bello, 2001; Lorenzo et al., 2012; Marco et al., 2008; Summo et al., 2011). The level of 2,3-butanediol (butter) decreased when the flavour had clearly deteriorated, while the gradual reduction of volatile methyl esters might have contributed to flavour deterioration during storage. This could be attributed to the low stability of 2,3-butanediol, as well as to the decrease of available sugars. The presence of ethyl esters in fermented meat products could be explained by the esterification activity of yeast, moulds and bacteria (Jelén & Wasowicz, 1998) contributing to flavour of fermented sausages because of their characteristic fruity notes, their low odour threshold value (Meynier et al., 1999) and their contribution to mask rancid odours (Stahnke, 1994). However, Olesen and Stahnke (2000) reported that yeast promoted the ester generation from the esterification of carboxylic acids and alcohols. Rancidity has been identified as the main cause of flavour deterioration in dry-cured sausage being mainly associated with lipid oxidation (Morrissey et al., 1998; Ordóñez et al., 1999; Zanardi et al., 2002). The level of MDA found in the semi-ripened salami was low compared with levels found in other studies of dry-cured sausages (Baka, Papavergou, Pragalaki, Bloukas, & Kotzekidou, 2011; Liaros, Katsanidis, & Bloukas, 2009; Rubio et al., 2008); however, other authors agree that MDA values slowly increase (Lee et al., 2009; Zanardi et al., 2002) or can even decrease (Marco et al., 2008) in retailed salami, depending on the storage conditions. Lipid oxidation largely takes place by reaction with the oxygen occluded in the mass and generates volatile and non-volatile flavouring compounds, such as aldehydes, ketones and acids, among others (Nawar, 1996). The intensity of lipid oxidation depends on lipid unsaturation and the balance between prooxidant and antioxidant agents in the sausage (Morrissey et al., 1998). Due to their low perception thresholds (Montel et al., 1996), the volatiles derived from lipid autoxidation play an important role in the flavour of dry-cured sausages (Sun et al., 2009). Aliphatic aldehydes such as hexanal (green grass and rancid), octanal (geranium herbal and floral) and nonanal (plastic and soap) (Marco et al., 2007) are considered as markers of lipid oxidation because they are derived from hydroperoxide degradation (Frankel, Selke, Neff, & Miyashita; 1992). Carboxylic acids resulting from the oxidation of aldehydes and ketones may provide rancid and pungent off-flavours. Increased levels of tetradecanoic and other carboxylic acids were found in salami stored at 10–12 °C for 120 days (Misharina et al., 2001). Marco et al. (2007) found no carboxylic acids in freshly-ripened salami. The results of both MDA and volatile

620

J.M. Lorenzo et al. / Meat Science 93 (2013) 614–620

compounds from lipid oxidation indicate that the semi-ripened salami of our study suffered modest oxidation at retail, probably needing larger storage times for the flavour to deteriorate. Two factors might have contributed to protect the sausage against oxidation: the addition of sodium ascorbate, a common antioxidant additive, and the retailing format, whole pieces, which are less exposed to fluorescent lighting during storage than sliced sausage. The low recovery of lipid oxidation products observed during the whole storage time might also be attributed to the anti-oxidative activity of the spices (Aguirrezábal, Mateo, Dominguez, & Zumalacárregui, 2000). Chemical phenomena responsible for the generation of aromatic volatiles in the mature meat are very complex and there are multiple reactions between a large number of compounds from meat, spices and sugars that contribute to flavour. The fermentative activity of lactic acid bacteria and Micrococcaceae further contributes to increase the biochemical complexity of flavour. Hence, it is difficult to determine the true impact on the flavour of individual volatile compounds present in the headspace, although, as can been seen, changes in the volatile profile during storage can be used to monitor the shelf life of salami, since the level of some volatile compounds would indicate flavour deterioration, and therefore, loss of acceptability. Using different statistical techniques, it has been established that there is a clear correspondence between the storage time, flavour scoring and the level of some volatile compounds from spices, microbial esterification and lipid autooxidation. Similar correlations between the storage time and the level of volatiles resulting from lipid oxidation and carbohydrate fermentation have also been reported in smoked salami (Summo et al., 2011). Volatile aldehydes seem to be promising for use as rancidity markers in salami. The sole alkenal detected in the headspace, 2-heptenal, was the aldehyde that best correlated with the storage time. 2-Heptenal is formed by autoxidation of polyunsaturated triacylglycerols containing linoleate (Frankel et al., 1992), and is present at low level compared with hexanal and other aldehydes in salami (Marco et al., 2007; Spaziani et al., 2009), where it is associated with rancid and dirty off-flavours (Marco et al., 2007). Misharina et al. (2001) reported strong increases in hexanal and some alkenals (2-heptenal, 2-nonenal, 2-decenal, 1,4-decadienal and undecenal) in the headspace of salami stored under strong oxidizing conditions. Ansorena et al. (2001) found increases in some alkanals (hexanal, heptanal and nonanal) and alkenals (octenal and 2,4-decadienal) during the storage of a Spanish dry-cured sausage stored for 2 and 5 months. In contrast, Summo et al. (2011) did not find alkenals, while hexanal, heptanal, octanal and nonanal increased in vacuum-packed smoked salami stored for 5 months. However, Marco et al. (2008) reported a reduction in most volatile aldehydes, including hexanal and 2-heptenal, in vacuum-packed salami stored for 60 days. 5. Conclusions The flavour of semi-ripened salami is stable for long periods of time under retail display conditions, although it finally deteriorates due to the development of incipient rancidity, resulting in loss of acceptability. The flavour deterioration during the shelf life of salami can be monitored through changes in the volatile profile. The loss and/or degradation of volatiles resulting from spices and microbial activities, together with the formation of volatiles as a result of lipid oxidation, seem to contribute to flavour deterioration. The levels of 2-heptenal and the methyl esters of heptanoic, pentanoic and hexanoic acids were the best discriminators of storage time, and therefore, they offer good possibilities for use as marker compounds of flavour deterioration and acceptability. References Aguirrezábal, M. M., Mateo, J., Dominguez, M. C., & Zumalacárregui, J. M. (2000). The effect of paprika, garlic and salt on rancidity in dry sausages. Meat Science, 54, 77–81. Andrade, M. J., Cordoba, J. J., Casado, E. M., Córdoba, M. G., & Rodríguez, M. (2010). Effect of selected strains of Debaryomyces hansenii on the volatile compound production of dry fermented sausage “salchichón”. Meat Science, 85, 256–264.

Ansorena, D., Gimeno, O., Astiasarán, I., & Bello, J. (2001). Analysis of volatile compounds by GC–MS of a dry fermented sausage: Chorizo de Pamplona. Food Research International, 34, 67–75. Baka, A. M., Papavergou, E. J., Pragalaki, T., Bloukas, J. G., & Kotzekidou, P. (2011). Effect of selected autochthonous starter cultures on processing and quality characteristics of Greek fermented sausages. Food Science and Technology, 44, 54–61. Bedia, M. (2012). Technological improvement of pork semi-ripened fermented sausage. Ph. Dr. Thesis. University of Murcia, Spain. Botsoglou, N. A., Fletouris, D. J., Papageorgiou, G. E., Vassilopoulos, V. N., Mantis, A. J., & Trakatellis, A. G. (1994). Rapid, sensitive, and specific thiobarbituric acid method for measuring lipid peroxidation in animal tissue, food, and feedstuff samples. Journal of Agricultural and Food Chemistry, 42, 1931–1937. Frankel, E. N., Selke, E., Neff, W. E., & Miyashita, K. (1992). Autoxidation of polyunsaturated triacylglycerols. IV. Volatile decomposition products from triacylglycerols containing linoleate and linolenate. Lipids, 27(6), 442–446. Guadayol, J. M., Caixach, J., Cabañas, J., & Rivera, J. (1997). Extraction, separation and identification of volatile organic compounds from paprika oleoresin (Spanish type). Journal of Agricultural and Food Chemistry, 43, 1868–1872. Hierro, E., Ordóñez, J., Bruna, J. M., Pin, C., Fernández, M., & de la Hoz, L. (2005). Volatile compound generation in dry fermented sausages by the surface inoculation of selected mould species. European Food Research and Technology, 220, 494–501. ISO 4121 (2003). International organization for standardization publications. Sensory analysis methodology. Evaluation of food products by methods using scales, www.iso.org ISO 8586–1 (1993). International organization for standardization publications. Sensory analysis methodology. General guidance for the selection and training and monitoring of assessors. Part 1. Selected assessors, www.iso.org Jelén, H., & Wasowicz, E. (1998). Volatile fungal metabolites and their relation to the spoilage of agricultural commodities. Food Reviews International, 14, 391–426. Lee, K. T., Lee, Y. K., Son, S. K., Choi, S. H., & Lee, S. B. (2009). Changes in various quality characteristics of short-ripened salami during storage at chilled or room temperatures. Korean Journal for Food Science of Animal Resources, 29, 24–33. Liaros, N. G., Katsanidis, E., & Bloukas, J. G. (2009). Effect of the ripening time under vacuum and packaging film permeability on processing and quality characteristics of low fat fermented sausages. Meat Science, 83, 589–598. Lorenzo, J. M., Montes, R., Purriños, L., & Franco, D. (2012). Effect of pork fat addition on the volatile compounds of foal dry-cured sausage. Meat Science, 91, 506–512. Marco, A., Navarro, J. L., & Flores, M. (2007). Quantification of selected odor-active constituents in dry fermented sausages prepared with different curing salts. Journal of Agricultural and Food Chemistry, 55, 3058–3065. Marco, A., Navarro, J. L., & Flores, M. (2008). The sensory quality of dry fermented sausages as affected by fermentation stage and curing agents. European Food Research and Technology, 226, 449–458. Martínez, P. J., Bedia, M., Méndez, L., & Bañón, S. (2009). Contribution of drying stage to the ripening of dry-cured fermented longaniza. Anales de Veterinaria de Murcia, 25, 121–132. Meynier, A., Novelli, E., Chizzolini, R., Zanardi, E., & Gandemer, G. (1999). Volatile compounds of commercial Milano salami. Meat Science, 51, 175–183. Misharina, T. A., Andreenkov, V. A., & Vashchuk, E. A. (2001). Changes in the composition of volatile compounds during aging of dry-cured sausages. Applied Biochemistry and Microbiology, 37, 413–418. Montel, M. C., Reitz, J., Talon, R., Berdagué, J. L., & Rousset, S. (1996). Biochemical activities of Micrococcaceae and their effects on the aromatic profiles and odours of dry sausages model. Food Microbiology, 13, 489–499. Morrissey, P. A., Sheehy, P. J. A., Galvin, K., Kerry, J. P., & Buckley, D. J. (1998). Lipid stability in meat and meat products. Meat Science, 49, 73–86. Nawar, W. W. (1996). Lipids. In O. R. Fennema (Ed.), Food chemistry (pp. 225–319). New York: Marcel Dekker. Olesen, P. T., Meyer, A. B. S., & Stahnke, L. H. (2004). Generation of flavour compounds in fermented sausages-the influence of curing ingredients, Staphylococcus starter culture and ripening time. Meat Science, 66, 675–687. Olesen, P. T., & Stahnke, L. H. (2000). The influence of Debarymyces hansenii and Candida utilis on the aroma formation in garlic spiced fermented sausages and model minces. Meat Science, 56, 357–368. Ordóñez, J. A., Hierro, E. M., Bruna, J. M., & de la Hoz, L. (1999). Changes in the components of dry-fermented sausages during ripening. Critical Reviews in Food Science and Nutrition, 39, 329–367. Rubio, B., Martínez, B., Sánchez, M. J., García-Cachán, M. D., Rovira, J., & Jaime, I. (2008). Effect of the packaging method and the storage time on lipid oxidation and colour stability on dry fermented sausage salchichón manufactured with raw material with a high level of mono and polyunsaturated fatty acids. Meat Science, 80, 1182–1187. Spaziani, M., del Torre, M., & Stecchini, M. L. (2009). Changes of physicochemical, microbiological, and textural properties during ripening of Italian low-acid sausages. Meat Science, 81, 77–85. Stahnke, L. H. (1994). Aroma components from dried sausages fermented with Staphylococcus xylosus. Meat Science, 38, 39–53. Summo, C., Caponio, F., & Pasqualone, A. (2006). Effect of vacuum-packaging storage on the quality level of ripened sausages. Meat Science, 74, 249–254. Summo, C., Caponio, F., Pasqualone, A., & Gomes, T. (2011). Vacuum-packed ripened sausages: Evolution of volatile compounds during storage. Journal of the Science of Food and Agriculture, 91, 950–955. Sun, W., Zhao, H., Zhao, Q., Zhao, M., Yang, B., Wu, N., et al. (2009). Structural characteristics of peptides extracted from Cantonese sausage during drying and their antioxidant activities. Innovative Food Science and Emerging Technologies, 10, 558–563. Zanardi, E., Dorigoni, V., Badiani, A., & Chizzolini, R. (2002). Lipid and colour stability of Milano-type sausages: Effect of packing conditions. Meat Science, 61, 7–14.