Scholarly article on topic 'Physicochemical, textural, volatile, and sensory profiles of traditional Sepet cheese'

Physicochemical, textural, volatile, and sensory profiles of traditional Sepet cheese Academic research paper on "Animal and dairy science"

0
0
Share paper
Academic journal
Journal of Dairy Science
OECD Field of science
Keywords
{""}

Academic research paper on topic "Physicochemical, textural, volatile, and sensory profiles of traditional Sepet cheese"

WAV.

J. Dairy Sci. 94:4300-4312 doi:10.3168/jds.2010-3941

© American Dairy Science Association®, 2011.

Physicochemical, textural, volatile, and sensory profiles of traditional Sepet cheese

D. Ercan,* F. Korel,*1 Y. Karagül Yüceer,t and Ö. KinikJ

'Department of Food Engineering, Izmir Institute of Technology, Urla, Izmir, Turkey tDepartment of Food Engineering, Canakkale Onsekiz Mart University, Canakkale, Turkey ^Department of Dairy Technology, Ege University, Bornova, Izmir, Turkey

ABSTRACT

Characterization of traditional cheeses is important for the protection of diversity of tradition and contributing baseline data for further research and quality control. Sepet cheese is a traditional cheese and specific to the Aegean region of Turkey. In this study, 52 Sepet cheese samples were analyzed to characterize the physicochemical, textural, volatile compounds, and sensory profiles. The changes in the physicochemical and volatile compositions were investigated during production and ripening periods. The average dry matter (DM; 55.16%), fat-in-DM (45.80%), protein (29.18%), salt-in-DM (12.88%), water activity (0.83), pH (5.50), titratable acidity (1.69%), ripening and lipolysis indices (11.06 and 6.36), firmness (212.20 N), springiness (0.62), cohesiveness (0.57), adhesiveness (0.48 Nmm), and chewiness (66.87 N) values of Sepet cheese samples were determined. Hexanoic, octanoic, decanoic, and butyric acids, which were responsible for the cheesy, waxy, goaty odors, were the most abundant volatile compounds in these cheeses. Most of the volatile compounds increased significantly during production and ripening. Significant changes in most of the physico-chemical characteristics were observed up to the third month of ripening. As a result of the descriptive sensory analysis, Sepet cheeses were described with descriptors such as free fatty acid, animal like, sulfurous, creamy, cooked, and whey, and aromatics with high salty basic taste.

Key words: Sepet cheese, texture, odor, sensory profile

INTRODUCTION

Traditional cheeses are considered to be produced locally or regionally for many generations and they have an important place in rural region food culture (Weichselbaum et al., 2009). Many traditional cheese types are produced and consumed locally in Turkey

Received October 18, 2010. Accepted May 3, 2011.

Corresponding author: figenkorel@iyte.edu.tr

(Turkoglu et al., 2003). However, production of almost all traditional cheeses occurs to meet the local demand and those cheeses are not very well known in other regions (Tarakci and Temiz, 2009). Sepet cheese is one of the traditional cheeses of the Aegean region in Turkey. It is named Sepet (basket) cheese because the appearance of the surface of Sepet cheese has a basket-weave impression. Baskets, made from stalks collected near rivers and moist areas, are used in the production of this cheese (Kamber, 2008). A photograph of Sepet cheese and a flow diagram of its production are given in Figures 1 and 2, respectively. Mostly raw goat milk is used in production (Kinik et al., 1999). Faced with extinction of some traditional cheeses and changes in milk production and cheese manufacture, a real need exists for the knowledge of characterization of texture and flavor of traditional cheeses. The acceptability of cheese by consumers is mainly based on its texture and flavor. The texture is affected during production and ripening (Beuvier and Buchin, 2004). During manufacture, milk changes physically and rheologically and the cheese matrix is formed (Gunasekaran and Ak, 2003). Then, changes in the cheese matrix continue during ripening because the matrix is influenced by the loss of water and proteolysis (Beuvier and Buchin, 2004). One of the quality parameters of cheese is its flavor. Formation of flavor compounds depends on physicochemical parameters such as moisture, salt content, pH, and ripening conditions. Cheese flavor is derived from the breakdown of milk proteins, fats, and lactose by enzyme activities or chemical reactions (for example, via the Maillard and Strecker reactions between AA and various car-bonyls; Forde and Fitzgerald, 2000; Fox et al., 2000). To date, much research has been done to demonstrate the specific characteristics of traditional cheeses (Qian and Reineccius, 2002; Hayaloglu et al., 2008; Tarakci and Temiz, 2009; Picon et al., 2010; Sanchez-Macias et al., 2010). To the best of our knowledge, no study exists that evaluates all of the aspects of traditional Sepet cheeses. Therefore, the aims of this study were to investigate the physicochemical, textural, aroma, and organoleptic characteristics of Sepet cheeses that were collected from different towns of the Aegean Region

Figure 1. Photograph of Sepet cheese.

and to determine the changes of physicochemical and volatile composition of Sepet cheeses collected from 2 different farms during production and ripening for 6 mo.

MATERIALS AND METHODS Sample Collection

Between 2008 and 2009, 52 Sepet cheese samples, which were ripened approximately for 3 mo, were collected from different towns located near Izmir, Turkey. Samples were taken to the laboratory in an ice box on the same day and they were stored at 4°C until analysis. Samples were also vacuum packaged and stored at -20°C for volatile compound analysis.

In the second part of the study, 2 production runs were done using goat milk in dairy farms located in Zeytineli and Germiyan towns (Izmir). In Zeytineli, natural rennet was used as a coagulant agent, which was produced from the fourth stomach of a young goat. In Germiyan, microbial rennet (Mayasan A.S., Istanbul, Turkey) was used in the production. Two cheese samples (2 replicates) were produced in each production and ripened in brine in a cool and shady place. Milk, curd, first day, and first, third, and sixth month cheeses were analyzed.

Physicochemical Analysis

Cheeses were analyzed for DM by the oven-drying method (IDF, 1982), fat by the Gerber method (IDF, 1997), total nitrogen (TN) by the Kjeldahl method (IDF, 1993), salt by the Mohr method (IDF, 1988), and

pH by the method described by Shakeel-Ur-Rehman and Fox (2002). The water activity (aw) values of samples were measured using a water activity meter (HygroLab V3, Bassersdorf, Switzerland). Titratable acidity was determined using the method of AOAC (1995). Protein and fat contents of milk samples were measured using a Lactostar analyzer (Funke-Dr.N.Gerber Labortechnik GmbH, Berlin, Germany).

Procedures described by Bynum and Barbano (1985) and Metin (2006) were followed to determine pH 4.6-sol-uble nitrogen (pH 4.6-SN) and 12% trichloroacetic acid-soluble nitrogen (TCA-SN) fractions, which are the indications of the extent of proteolysis. The index of lipolysis (acid degree value) of samples was determined by using the method described by Renner (1986) and expressed as mEq of KOH/100 g of fat. All analyses were performed in duplicate.

Texture Analysis

Cheese samples were cut into 25-mm cubes and stored in a refrigerator before texture analysis. Texture profile analysis (TPA) test was performed in triplicate using a TA.XTPlus Texture Analyzer (Stable Microsystems Ltd., Godalming, UK). By using a 50-kg force load cell, a double bite compression cycle was carried out

Raw milk i

Heating to 40°C i

Remieting (1 g/5 L of milk) and coagulation (2 h) Cutting (particles 1 cm2)

Draining

Placing into baskets

Draining and dry salting i "

Ripening (in brine 140 g of NaCl/L) Marketing

Figure 2. Flow diagram of Sepet cheese production.

using the method described by Everard et al. (2007). Parameters of TPA were defined as given in Everard et al. (2007).

Volatile Compound Analysis

For the extraction of volatile compounds, the solidphase microextraction (SPME) method was used. For this purpose, a fiber, provided by Supelco (57348-U; Supelco Inc., Bellefonte, PA), coated with the following sorbent material: divinylbenzene/Carboxen/ polydimethylsiloxane, was used. Samples were defrosted at 4°C before the day of analysis. The outer surfaces of samples were removed and the samples were grated. Three grams of grated samples were weighed into a 20mL vial, and a PTFE/butyl septum was immediately sealed with an aluminum crimp seal. The sample was equilibrated at 60°C at 500 rpm for 30 min. Then, fiber was exposed into the headspace for 30 min at 60°C while the sample was agitated at 500 rpm. Volatile compounds adsorbed on the fiber desorbed in the injector port of the GC (Agilent Technologies, Wilmington, DE) for 5 min.

A PerkinElmer Clarus 600 GC/olfactometry (GC/O) system (PerkinElmer Inc., Waltham, MA), equipped with a flame ionization detector, was used for the volatile compound analysis of the 52 Sepet cheese samples. The temperature of the injector port was 250°C. The oven was temperature programmed as follows: 40°C (6 min), 5°C/min to 100°C (2 min), and 10°C/min to 250°C (4 min). The carrier gas was He with a 1 mL/min flow rate. A BP20 wax capillary column (SGE International Pty. Ltd., Ringwood, Victoria, Australia; 30 m x 0.25mm i.d. x 0.25-^m film thickness) was used. The GC/O analysis was performed by one highly trained sniffer, and aroma intensity was recorded using a 10-point scale, with 10 indicating extremely strong, 6 indicating medium intensity, 3 indicating weak intensity, and 0 indicating not detected (Qian et al., 2002). The sniffer had at least 60 h of experience (especially with Sepet cheese aroma) with the GC/O technique, scale using, and odor description. The analysis was performed in duplicate. Identification was done with comparison of odor and retention indices of volatile compounds and authentic standards analyzed at the same conditions. Retention indices were calculated using the n-alkane series (C6-C30; Van den Dool and Kratz, 1963).

For the Sepet cheeses produced in 2 different towns, the changes in volatile composition during production and ripening were investigated using GC (Agilent 6890 N Network GC Systems)-MS (5973N Network Mass Selective Detector; Agilent Technologies). An Elite-225 capillary column (PerkinElmer Instruments, Shelton, CT; 30 m x 0.25-mm i.d. x 0.25-^m film thickness)

was used. The inlet temperature was 220°C. The oven temperature was programmed as: 40°C (6 min), 5°C/ min to 100°C (2 min), and 10°C/min to 220°C (5 min). The carrier gas, He, was used at 1 mL/min. The transfer line temperature was 280°C. The mass scan range was 26 to 350 amu at 1.18 scans/s and ionization voltage was 70 eV. The analysis was performed in duplicate. Identification was done by comparing ion spectra and retention times with authentic standards and spectra from the mass spectral library [National Institute of Standards and Technology (NIST) 98, version 2.0; Ringoes, NJ; Carbonell et al., 2002].

Sensory Analysis

A roundtable discussion with a 5-member panel was conducted to identify the descriptive flavor terms for the Sepet cheeses. The panelists were staff and graduate students in the Department of Food Engineering of Canakkale Onsekiz Mart University (Canakkale, Turkey). The panelists identified and defined the flavor terms from representative cheeses (Table 1). The panelists received about 50 h of training during identification and definition of descriptive terms. The panelists quantified the descriptive terms using 20-point product-specific scales anchored on the left with "not" and on the right with "very" (Meilgaard et al., 1999). The panelists were provided with water, unsalted bread, and expectoration cups. The cheeses were presented in plastic plates and coded with 3-digit numbers. Five samples were evaluated in each session. Panelists evaluated each cheese twice. Duplicate samples were served in different sessions (Karagul-Yuceer et al., 2007).

Statistical Analysis

Pearson correlation coefficients (r) were calculated to determine linear relations between the quality characteristics of Sepet cheeses. Analysis of variance and the Student-Newman-Keuls range test were performed to investigate the differences during production and ripening.

A nonmetric multidimensional scaling (MDS) method was also applied to provide a visual representation of distances among Sepet cheese samples collected from different towns located near Izmir, based on physico-chemical, textural, sensory, and volatile characteristics. Multidimensional scaling is an ordination and numerical technique for finding a configuration of points in low-dimensional space that represents multivariate data. Multidimensional scaling plots the samples on a map where similar samples are placed near each other and different samples are located away from each other. The points close to each other on the map show the

Table 1. Language used to evaluate Sepet cheese flavor

Descriptor Definition Reference

Cooked Aromatics associated with cooked milk Milk heated to 85°C for 30 min1

Whey Aromatics associated with whey powder Solubilize 5 g of whey powder in 100 mL of water1

Creamy Aromatics associated with milk fat Cream or butter1

Free fatty acids Aromatics associated with butyric acid 10 |j,L of butyric acid in methanol1

Dirty moist cloth Aroma associated with wet cloth Dirty moist cloth

Storage Aromatics associated with warehouse Long time stored cheese, assignment by panel

Fruity Aromatics associated with fruits Fresh pineapple, ethyl hexanoate, 20 mg/kg2

Nutty The nut-like aromatics associated with nuts Lightly toasted unsalted nuts2

Metallic Aroma associated with tin cans Tin cans

Animal like Aromatics associated with barns and stock 5% Na-caseinate solution in water1

Sulfurous Aromatics associated with sulfurous compounds Boiled mashed egg1

Sour Taste sensation elicited by acids 0.08% citric acid solution in water3

Bitter Taste sensation elicited by caffeine 0.08% caffeine solution in water3

Salty Taste sensation elicited by salts 0.5% sodium chloride solution in water3

Sweet Taste sensation elicited by sugars 2% sucrose solution in water3

Umami Chemical feeling factor elicited by certain peptides and nucleotides 1% monosodium glutamate solution in water3

Astringent Puckering of oral epithelium as a result of exposure Tea solution, soak 6 tea bags in hot water for 1 h4

to alum or tannins

Bite Chemical feeling factor elicited by carbonation on the tongue Soda water1

Reference adapted from Karagul-Yuceer et al. (2007). 2Reference adapted from Drake et al. (2001). 3Reference adapted from Meilgaard et al. (1999). 4Reference adapted from Friedeck et al. (2003).

relationship between the samples, as well as similarity of behavior with respect to the remaining samples. The points located on the plots reproduce distances between each sample and the distances among samples are calculated with an appropriate distance measure, such as Euclidean distance or Manhattan distance. This distance matrix is used for conducting MDS ordination. A stress coefficient and the proportion of variance of the scaled data are used to measure how well any given configuration fits the data. The smaller the stress, the better the representation (Kruskal, 1964; Ba§pinar et al., 2000; Karagul-Yuceer et al., 2007, 2009; Holland,

2008). Nonmetric MDS procedures deal with ordinal data, which implies less restrictive criteria of fit. Therefore, nonmetric MDS usually achieve solutions in the same or lower dimensionalities than metric factor analysis or principal component analysis (Rabinowitz, 1975). MacFie and Thomson (1984), Bieber and Smith (1986), Heymann (1994), Lawless et al. (1995), Popper and Heymann (1996) and Karagul-Yuceer et al. (2007,

2009) covered the use of MDS in sensory analysis of foods. Because different scales were used to measure characteristics of Sepet cheeses, data were standardized before the MDS method. A 52 x 52 symmetrical distance matrix was created with the dissimilarity measure (Euclidean distance) from the data, including 52 rows (each row represents each Sepet cheese sample) with columns, which include all physicochemical, textural, sensory, and volatile characteristics. The MDS ordination was performed on this matrix. All statisti-

cal analysis was done using SPSS software (version 13; SPSS Institute Inc., Chicago, IL).

RESULTS AND DISCUSSION Physicochemical Parameters

Dry matter, fat-in-DM, protein, salt-in-DM, water activity, pH, and titratable acidity values of Sepet cheeses are given in Table 2. Except NaCl content, the remaining physicochemical parameters were similar to the results of previously reported studies about Sepet cheese by Kinik et al. (1999) and Karaka§ and Korukluoglu (2006).

Protein degradation is a part of the cheese aging process (Bynum and Barbano, 1985). Variations in ripening indices, and TCA-SN and TN ratios of Sepet cheeses were observed. Singh et al. (2003) reported that the final pH, moisture, salt-in-moisture content, temperature, and time of ripening could be used to control proteolysis in cheese. Changes in those control parameters could be the reason for variation in levels of protein degradation in Sepet cheeses. Hayaloglu et al. (2008) explained that the reason for lower levels of pH 4.6-SN and TCA-SN in Kuflu cheese than of other mold-ripened cheeses may be because of the higher salt-in-moisture contents of Kuflu cheese samples. A significant negative correlation was observed between the level of TCA-SN and salt-in-moisture content (r = -0.28, P < 0.05).

Table 2. Physicochemical characteristics of Sepet cheeses (n = 52)

Physicochemical

characteristics1 Mean Minimum Maximum SD CV

DM (%) 55.16 44.56 64.39 5.13 9.30

Fat-in-DM (%) 45.80 33.43 58.35 4.60 10.04

Water in fat-free constituent (%) 59.87 48.45 74.92 6.02 10.05

Protein (%) 29.18 24.40 33.69 2.25 7.72

Salt-in-DM (%) 12.88 5.01 26.98 4.64 35.99

Water activity 0.83 0.74 0.91 0.05 5.88

pH 5.50 5.03 6.47 0.37 6.81

Titratable acidity (%) 1.69 0.60 2.85 0.57 33.45

Lipolysis index (acid degree value) 6.36 1.64 41.63 7.23 113.77

pH 4.6-SN/TN (%; ripening index) 11.06 3.33 34.52 7.40 66.91

TCA-SN/TN (%) 8.56 1.89 28.02 6.18 72.22

1rTitratable acidity was expressed as g of lactic acid/100 g of cheese; SN/TN = soluble nitrogen/total nitrogen; TCA = trichloroacetic acid.

The lipolysis indices in Sepet cheeses varied considerably. The average lipolysis index was higher than the value reported by Kinik et al. (1999). Lipolysis is related to lipases in cheese, originating from milk, rennet paste, starter, adjunct starter, or nonstarter bacteria, which could be different during production of cheese samples (Fox et al., 2000). The variation in lipolysis indices of Sepet cheeses could be related to variations in microbial loads, milk species, ripening conditions, and periods. Samples that had higher ripening and lipolysis indices than others were placed on the right-hand side of Figure 3.

Textural Parameters

Results of the TPA are given in Table 3. Variations were observed in textural parameters of Sepet cheeses. Figure 3 shows the similarities among the cheeses in terms of physicochemical, textural, sensory, and volatile characteristics, which were produced by MDS. Except cheeses 7, 8, 11, 33, 42, 48, and 50; all other cheeses grouped together close to the center of the graph (Figure 3). Most of the cheeses were similar in terms of physicochemical, textural, sensory, and volatile characteristics. Sample 42 and 43, due to their higher adhe-

Euclidean distance model

- VAR25 vlÄll VAR42 О ft °VAR43 VAR33 Й? VAR4

- V/ < VAR11 RS О VAR12 VAR 7 О о ) VAR5Q О VAR48

t I I I I ( I

•3-2-10 1 2 3 4

Dimension 1

Figure 3. Geometrical representation of cheeses in terms of physicochemical, textural, sensory, and volatile characteristics by multidimensional scaling (MDS). Each Var represents a different cheese among 52 samples; stress coefficient = 0.19, R2 = 0.90.

siveness; samples 7, 8, 11, 48, and 50, due to their higher firmness and chewiness; and sample 33, due to its low springiness, were placed away from the other samples in Figure 3. Production technology, milk composition, moisture, pH, salt content, lipolysis, and proteolysis occurring during ripening have effects on cheese texture (Lawrence et al., 1987; Fox et al., 2000; Gunasekaran and Ak, 2003). Firmness was significantly correlated with DM (r = 0.66, P < 0.01) and water-in-fat-free constituents content (r = -0.67, P < 0.01). It was found that more than half of Sepet cheeses had 54 to 63% water-in-fat-free constituents content and these cheeses were considered as semi-hard cheese according to IDF (1981). Chewiness was also significantly correlated with DM (r = 0.48, P < 0.01) and water-in-fat free constituent contents (r = -0.49, P < 0.01). A significant negative correlation between fat in DM and firmness were observed (r = -0.35, P < 0.05). Similarly, Gwartney et al. (2002) and Gunasekaran and Ak (2003) stated that low-fat cheeses had a more compact protein matrix and harder texture than those of whole-fat cheeses. Firmness and salt contents were correlated significantly (r = 0.28, P < 0.01). The correlation was in agreement with the results of the study of Prasad and Alvarez (1999). Fox et al. (2000) explained that salting affects cheese texture, due to protein solubility and protein conformation, and cause hard cheese texture. Significant negative correlation was observed between aw and firmness (r = -0.49, P < 0.01). Furthermore, significant correlations were observed between textural parameters and indices of ripening and lipolysis. Springiness and ripening index were negatively correlated (r = -0.45, P < 0.01). Gu-nasekaran and Ak (2003) reported that the decrease in springiness occurred during ripening with the effects of proteolytic breakdown of the protein matrix. Firmness and the ripening index were negatively correlated (r = -0.36, P < 0.01). Likewise, Brown et al. (2003) observed that firmness of cheeses decreased as the cheeses aged. Significant correlation between the lipolysis index and springiness was also observed (r = -0.44, P < 0.01).

Volatile Compounds Profile

A total of 41 volatile compounds were identified in the headspace of Sepet cheeses (Table 4). Free fatty

acids (FFA), ketones, aldehydes, esters, and others were found in the volatile fraction of Sepet cheeses. Sample 7, 33, and 48 were placed away from the other samples in Figure 3, because the odor intensities of esters and ketones were higher in those samples than of other Sepet cheeses. Sample 11 was also placed away from the rest in Figure 3 due to its high odor intensity of ethanol. Variations in flavor are mainly controlled by intricate biochemical reactions that occur due to the activities of starter cultures and their enzymes. These biochemical reactions are affected by the physicochemi-cal properties of cheese and ripening conditions (Forde and Fitzgerald, 2000).

In all Sepet cheeses, FFA were the most abundant volatile compounds among all of the identified fractions. Hexanoic, octanoic, decanoic, and butyric acids had the highest percentages in the volatile fraction of Sepet cheese, in decreasing order. The GC/O results of FFA were also higher than of the other compounds. Hexanoic, octanoic, and butyric acids were perceived as a mild to strong goat-like, waxy, and cheesy odor. Ripening index was significantly correlated with odor intensities of butyric (r = 0.30, P < 0.05), hexanoic (r = 0.42, P < 0.01), heptanoic (r = 0.28, P < 0.05), octanoic (r = 0.46, P < 0.01), and 4-methyloctanoic (r = 0.29, P < 0.05) acids. Moreover, the lipolysis index was positively correlated with odor intensities of butyric (r = 0.47, P < 0.01), hexanoic (r = 0.58, P < 0.01), octanoic (r = 0.59, P < 0.01), 4-methyloctanoic (r = 0.44, P < 0.01), and decanoic (r = 0.33, P < 0.05) acids. McSweeney and Sousa (2000) reported that linear FFA are generally produced from lipolysis of milk fat. The source of FFA can also be contributed by metabolism of deamination of AA and lipid oxidation (McSweeney and Sousa, 2000; Nogueira et al., 2005). Similarities were found between FFA of Sepet, Maltese goat milk, Minas, goat milk Jack cheese, and blue cheeses (Chiofa-lo et al., 2004; Frank et al., 2004; Nogueira et al., 2005; Attaie 2009). The esters in the volatile fraction of Sepet cheeses were ethyl hexanoate, ethyl decanoate, ethyl octanoate, and ethyl butyrate. The odor intensities of esters changed from medium to weak. These esters were appeared to be responsible for the characteristic fruity, green odor perceived in Sepet cheeses. Pinho et al. (2003) stated that most esters have effects on cheese

Table 3. Textural characteristics of Sepet cheeses (n = 52)

Textural

characteristics Mean Minimum Maximum SD CV

Firmness (N) 212.20

Springiness (—) 0.62

Cohesiveness (—) 0.57

Adhesiveness (Nmm) 0.48

Chewiness (N) 66.87

65.55 504.11 93.08 43.86

0.42 0.84 0.12 18.60

0.21 1.56 0.23 40.59

0.00 14.34 2.43 511.59

29.28 116.30 25.40 37.98

Table 4. Volatile compounds of Sepet cheeses (n = 52)

GC/O3 GC

Data for odor Data for volatile

intensities fraction

Chemical RI1 Odor2 Mean SD Mean4 SD

Free fatty acid

Acetic acid 1,401 Vinegar, sour 1.27 1.81 1.97 4.55

Butyric acid 1,592 Cheesy, sharp 3.94 1.55 7.49 4.89

Hexanoic acid 1,777 Goaty 5.б9 1.89 29.92 8.56

Heptanoic acid 1,904 Goaty, cheesy 1.23 1.21 2.41 5.26

Octanoic acid 2,02б Waxy 5.58 1.90 25.85 5.59

4-Methyl octanoic acid 2,209 Goaty, sour 1.00 1.87 G.88 2.90

Decanoic acid 2,382 Sour/waxy 3.13 1.33 15.2б 8.05

Dodecanoic acid 2,487 Soapy G.56 0.96 G.47 G.58

Ketone

Acetone 7б4 Alcoholic G.G6 0.31 0.26 1.43

Diacetyl 1,020 Buttery G.63 G.84 1.66 3.18

Acetoin 1,224 Buttery G.54 0.85 0.26 G.62

2-Nonanone 1,3б3 Fatty, floral 0.81 1.12 G.48 0.95

2-Tridecanone 1,б27 Fatty G.79 1.13 G.69 1.40

Aldehyde

Hexanal 1,102 Woody G.G4 0.19 G.G9 G.37

Octanal 1,229 Sweet, citrus 0.12 G.32 G.10 G.19

Nonanal 1,374 Hay/sweet 0.13 G.40 0.11 0.31

Decanal 1,49б Floral, waxy 0.35 G.62 0.16 G.32

E-2 Nonenal 1,519 Fatty, cucumber G.60 0.96 G.27 0.61

E-2, Z-6 Nonadienal 1,553 Hay/fatty G.52 1.04 G.49 2.18

2,3-Butanedial 1,599 Cheesy 1.17 1.54 1.62 3.43

E-2 Decenal 1,б14 Hay/fatty g.60 G.89 0.31 G.57

Ethyl butyrate 1,G68 Fruity G.23 0.55 G.64 2.55

Ethyl hexanoate 1,189 Orange, sour 1.G2 1.18 G.97 2.37

Ethyl octanoate 1,50б Fruity G.48 1.00 0.41 1.17

Ethyl decanoate 1,ббб Fruity G.90 1.18 0.35 0.61

Ethanol 978 Alcoholic G.54 G.87 1.14 2.67

Isoamyl alcohol 1,177 Bitter G.44 G.70 G.57 1.06

D-Limonene 1,182 Sour 0.12 G.38 G.39 1.17

1-Hexanol 1,297 Not detected G.G0 G.G0 0.01 G.G3

5-Decalactone 2,251 Peach G.58 0.96 G.60 1.05

o-Aminoacetophenone 2,373 Grape G.G8 G.33 0.16 0.36

Y-Dodecalactone 2,401 Coconut G.67 1.04 1.14 2.22

Unknown 1 1,133 Sweet G.G8 G.33 0.06 G.27

Unknown 2 1,253 Sweet 0.21 G.64 0.16 G.60

Unknown 3 1,2б9 Sweet 0.17 G.58 G.12 G.37

Unknown 4 1,424 Sour G.37 G.79 G.17 0.35

Unknown 5 1,457 Fruity 1.31 1.53 1.23 1.82

Unknown 6 1,531 Fatty 0.19 G.69 G.10 G.48

Unknown 7 1,848 Fruity G.50 G.98 G.24 G.49

Unknown 8 2,253 Fruity 0.60 G.98 G.60 2.11

Unknown 9 2,б88 Not detected G.G0 G.G0 G.23 1.14

1Retention index on BP20 wax column.

2Odors were determined on olfactory port.

3GC/olfactometry system.

4Mean values of % areas (area of component/total identified area).

aroma by minimizing the sharpness and the bitterness, which stem from the odors of fatty acids and amines, respectively. Significant positive correlations were observed between the lipolysis index and odor intensities of ethyl butyrate (r = 0.30, P < 0.05), ethyl octanoate (r = 0.30, P < 0.05), and ethyl decanoate (r = 0.30, P < 0.05). Esters are produced by esterification of alco-

hols and carboxylic acids, or alcoholysis of alcohols and acylglycerols or from alcohols and fatty acyl-coenzyme A (CoA) derived from the metabolism of fatty acids, AA, and carbohydrates (Liu et al., 2004). Similar esters were also observed in Parmigiano-Reggiano and Minas cheeses (Qian and Reineccius, 2002; Nogueira et al., 2005). Octanal, nonanal, decanal, (E,Z)-2,6-nonadienal,

Figure 4. Flavor profile diagram of Sepet cheese (A = basic taste, B = aromatics).

(E)-2-decenal, (E)-2-nonenal, and 2-3 butanedial were aldehydes found in Sepet cheeses. Although their fractions were low, they affected the aroma of Sepet cheeses with green and fatty odors. Carunchia Whetstine et al. (2003) reported that aldehydes affect the overall aroma of the goat cheese. Curioni and Bosset (2002) mentioned that aldehydes result from transamination or by Strecker degradation. The odor intensity of 2-3 butanedial was significantly correlated with the ripening index (r = 0.28, P < 0.05). Ketones such as acetoin, 2-nonanone, and 2-tridecanone were found in volatile fractions of Sepet cheeses. The formation of methyl ketones is a result of enzymatic oxidation of FFA to P-ketoacids and decarboxylation to alkan-2-ones with

1 less carbon atom (McSweeney and Sousa, 2000). The odor intensities of 2-tridecanone (r = 0.42) and 2-nona-none (r = 0.35) were significantly correlated with the lipolysis index (P < 0.01). The aroma intensity values of ketones changed between medium to weak among Sepet cheese samples. Frank et al. (2004) also found that 2-nonanone was important in blue cheese aroma. Similarly, Gonzalez De Llano et al. (1990) reported that the odd carbon-numbered ketones, especially 2-heptanone and 2-nonanone, were the most abundant compounds in the volatile fraction of artisanal Gam-onedo blue cheeses. Similar ketones were also found in the volatile fraction of Tulum cheese by Hayaloglu et al. (2007). D-Limonene was detected in some Sepet cheeses

and its odor intensities were recorded as weak. The occurrence of D-limonene is related to animal feeding and increases with green grass feeding (Chiofalo et al., 2004). Weak odor intensities of ethanol, isoamyl alcohol, o-aminoacetophenone, 6-decalactone, Y-dodecalactone, and some unknown components were also observed in some Sepet cheeses.

Sensory Profile

Descriptive sensory analysis was performed to determine the basic tastes and aromatics of Sepet cheeses. The predominant basic taste was perceived as salty (16.60 ± 4.64) for all Sepet cheeses (not shown in Figure 4A). The other basic tastes that were perceived by panelists were sour (2.00 ± 0.49), bite (1.64 ± 0.73), umami (1.52 ± 0.26), sweet (1.18 ± 0.19), bitter (0.66 ± 0.28), and astringent (0.52 ± 0.61), in decreasing order (Figure 4A). Free fatty acid (3.50 ± 1.12), animal like (3.22 ± 1.26), sulfurous (2.59 ± 0.49), creamy (2.09 ± 0.27), cooked (2.09 ± 0.29), whey (1.96 ± 0.25), dirty moist cloth (1.38 ± 0.51), storage (0.67 ± 0.57), fruity (0.52 ± 0.22), nutty (0.23 ± 0.21), and metallic (0.19 ± 0.19) aromatics were observed as characteristic terms for Sepet cheeses, in decreasing order (Figure 4B). Some of these terms such as cooked, whey, fruity, sulfur, FFA, nutty, and FFA, creamy, and goaty were also used as descriptive terms for Cheddar and Ezine cheese, respectively (Drake et al., 2001; Karagul-Yuceer et al., 2007).

Changes During Production and Ripening

In the second part of the study, physicochemical parameters and volatile compounds were investigated during production and ripening. The changes in physi-cochemical parameters are given in Table 5. Milk used in dairy farms located in Germiyan town and Zeytineli town had 5.22 ± 0.00% fat content, 0.57 ± 0.00% TN, and 0.05 ± 0.00% titratable acidity. The DM contents of cheeses produced in Germiyan and Zeytineli increased significantly during production due to the concentration of milk constituents in cheese (P < 0.01). The DM contents increased significantly up to the third month in Germiyan and sixth month in Zeytineli (P < 0.05). During ripening, an increase in salt content can cause a slight increase in DM content (Fox et al., 2000). During cheese production, titratable acidity significantly increased (P < 0.01). Tarakci and Kucukoner (2006) stated that the initial increase in acidity is due to lactic acid formation. During production of Sepet cheeses, aw decreased significantly (P < 0.01) due to the increase in DM content and the presence of salt. During production and up to the first month, salt-in-DM increased

3 oc c et X Cft — c^ CD et c "d et « ^HfO^ CDCDCD

c o c CD CD CD CD

+ -H + + -H -H +

a oc CO — Cft O^ ^ CO oq^^c^

CO ic iO oc CDOOC^

a c^ et X o c^ c^ CD « iC u et « CDCDCD

c CD CD CD CD CD CD

+ -H + + -H -H +

c b- ^ CO CO CO -H a O^C^f^

CO iC iO OC ^ CDOOC^

X c^ co et X iO <— C^ CD « c^ <J -Q -c oot- CD'—i '—

c CD CD CD CD CD CD

+ -H + + -H -H +

CO co CO CO c^ ^ CO ^ t- o oot-cc

^ ^ oc ^ CDt-^C^

-Q tf C^ '—i c^ CD .c CO C^ -Que oe^co

c CD CD CD CD CD CD

+ -H -H +

c CO OC co OC iC CO ^ O CO

oc i-H OC CDCDCD

b CO CO CD CO ^ -d -c CD'—^CD

c CD CD CD CD CD CD

+ -H + + -H -H +

co iO iC ^ C^ CO CO t-- CO CO COOCO

c^ CO i-H OC c^ CD-^CD

c^ et X i-H CO C^ CD et CO CD i—i LO CO

c CD CD CD CD CD CD

+ -H + + -H -H +

CO iC CO t-- cyi co CDt-^CD

CO iC i-H OC iO ^ rHOO^i

X b- et X O ^ CO CD et OC c^ CO C^ CDCDCD

c CD CD CD CD CD CD

+ -H + + -H -H +

a c^ CO t"-^H iC CO 00 lO IC

c^ iC i-H OC iO CDOOC^

LC c et X iO CO lO CD C^ « XI XI CDCDCD

c CD CD CD CD CD CD

+ -H + + -H -H +

b- CO iC a ^ CO O OC C^CDC^

a CD OC iO CDt-^C^

"LO j: JD d ^—i C^ C^ CD .c CO CD X u u '—iC^^t CD'—^CD

cd CD CD CD CDCDCD

CO 00 CO to t-^ CO CO '—i CO O^C^O^

C<D OC CDCDCD

co u rt ^ C^ "ct CO rH CO C^C^'—

c CD CD CD CDCDCD

CO -H o CD C^ t-- CO CO

CO CO OC CD-^CD

5> 5> ^ ^ ^

TRADITIONAL SEPET CHEESE Table 6. Volatile compounds during production and ripening of Sepet cheeses1

Compound Rt (ions)2 Dairy Curd d 1 mo 1 mo 3 mo 6

Free fatty acid 0.00 ± 0.00b 0.00 ± 0.00b 6.43 ± 3.08b 13.39 ± 3.38ab

Acetic acid 10.13 Z 22.02 ± 4.54a

(43,45,60) G 0.00 ± 0.00a 0.00 ± 0.00a 1.81 ± 0.77a 1.67 ± 0.77a 1.93 ± 0.86a

Butyric acid 15.50 Z 0.00 ± 0.00c 4.71 ± 1.72c 60.96 ± 3.14b 72.17 ± 5.86b 84.16 ± 1.59a

(60,73,28) G 4.72 ± 1.13c 10.20 ± 2.65bc 24.13 ± 3.44ab 31.11 ± 1.16a 42.20 ± 7.81a

Hexanoic acid 21.26 Z 8.20 ± 1.50c 15.18 ± 3.41c 95.19 ± 1.94b 125.72 ± 11.60a 145.52 ± 1.40a

(60,73,87) G 5.58 ± 1.17c 16.53 ± 2.92c 63.52 ± 4.07b 63.15 ± 1.82b 80.88 ± 5.62a

Octanoic acid 24.97 Z 9.01 ± 1.17c 11.32 ± 0.45c 57.99 ± 0.18b 66.62 ± 3.21a 68.45 ± 2.33a

(60,73,43) G 6.46 ± 0.32b 18.11 ± 5.92b 44.61 ± 5.30a 41.55 ± 1.94a 49.39 ± 0.71a

Decanoic acid 27.46 Z 4.39 ± 0.13b 13.35 ± 2.58b 39.92 ± 5.44a 44.40 ± 9.46a 37.46 ± 4.19a

(73,60,41) G 3.76 ± 1.27b 9.85 ± 0.93b 21.77 ± 3.66a 32.50 ± 3.86a 32.60 ± 2.33a

Ester Ethyl acetate 2.56 Z 0.00 ± 0.00a 0.00 ± 0.00a 5.75 ± 0.35a 4.33 ± 2.24a 4.53 ± 0.29a

(43,28,29) G 0.00 ± 0.00a 1.78 ± 1.78a 1.94 ± 1.94a 1.11 ± 0.33a 1.58 ± 1.58a

Allyl acetate 7.03 Z 0.00 ± 0.00a 0.00 ± 0.00a 2.45 ± 0.32a 1.13 ± 1.13a 1.91 ± 0.18a

(43,56,73) G 0.00 ± 0.00a 0.50 ± 0.49a 0.34 ± 0.34a 1.20 ± 0.41a 2.07 ± 1.06a

Ethyl hexanoate 13.25 Z 0.94 ± 0.94d 2.89 ± 0.65c 6.08 ± 0.03b 6.28 ± 0.12b 8.75 ± 0.23a

(88,99,43) G 0.23 ± 0.23b 2.15 ± 0.78ab 4.39 ± 1.04ab 7.11 ± 3.00ab 11.82 ± 3.08a

Ethyl octanoate 19.09 Z 0.00 ± 0.00b 0.94 ± 0.59b 3.67 ± 0.03a 3.94 ± 0.20a 4.37 ± 0.18a

(88,101,127) G 0.12 ± 0.12c 1.07 ± 0.08bc 2.30 ± 0.45bc 3.47 ± 0.41ab 5.59 ± 1.19a

Ethyl decanoate 23.93 Z 0.83 ± 0.83b 2.93 ± 0.90ab 4.87 ± 0.41a 5.78 ± 0.11a 5.89 ± 0.41a

(88,101,157) G 0.75 ± 0.75b 2.24 ± 0.15ab 3.13 ± 0.22a 3.15 ± 0.23a 4.16 ± 0.53a

Ketone

Acetone 2.17 Z 12.04 ± 6.31a 10.81 ± 5.87a 3.95 ± 3.95a 15.22 ± 10.7a 0.00 ± 0.00a

(43,58) G 0.00 ± 0.00a 0.00 ± 0.00a 4.15 ± 4.15a 24.16 ± 7.72b 0.00 ± 0.00a

Acetoin 9.74 Z 0.00 ± 0.00b 0.00 ± 0.00b 3.15 ± 0.80ab 10.87 ± 4.10a 10.90 ± 0.97a

(45,43,88) G 0.26 ± 0.26a 2.00 ± 2.00a 1.67 ± 1.67a 1.04 ± 1.04a 1.03 ± 1.03a

2-Heptanone 11.79 Z 1.64 ± 0.16b 2.05 ± 0.15b 5.04 ± 0.52a 5.28 ± 0.59a 6.81 ± 0.48a

(43,58,71) G 0.80 ± 0.23b 3.33 ± 0.68b 6.19 ± 0.21ab 7.95 ± 0.08ab 12.30 ± 3.37a

2-Tridecanone 18.05 Z 0.45 ± 0.45b 1.43 ± 0.43b 3.90 ± 0.06a 4.66 ± 0.51a 4.53 ± 0.32a

(58,43,71) G 0.96 ± 0.23b 1.49 ± 0.44b 2.40 ± 0.29ab 4.56 ± 0.35a 5.46 ± 1.17a

Alcohol

Ethanol 2.10 Z 14.36 ± 5.41b 4.08 ± 4.08b 75.78 ± 11.10a 58.27 ± 20.1ab 57.69 ± 4.54ab

(45,46) G 12.68 ± 3.02a 51.43 ± 0.92a 48.62 ± 6.62a 28.77 ± 10.00a 33.43 ± 10.80a

1-Propanol 3.01 Z 1.17 ± 1.17b 0.47 ± 0.47b 9.46 ± 2.47a 13.28 ± 1.40a 11.72 ± 0.13a

(31,29,27) G 0.25 ± 0.25a 4.22 ± 1.31a 4.68 ± 0.63a 5.37 ± 0.54a 5.25 ± 1.56a

2-Butanol 3.41 Z 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a

(45,31,59) G 0.00 ± 0.00a 0.00 ± 0.00a 0.74 ± 0.73a 0.00 ± 0.00a 0.00 ± 0.00a

a-Pinene 6.10 Z 1.11 ± 1.11a 1.19 ± 1.19a 3.47 ± 0.41a 1.70 ± 1.70a 3.47 ± 0.92a

(93,71,91) G 1.21 ± 1.15a 1.08 ± 1.08a 1.14 ± 1.14a 5.85 ± 2.72a 8.26 ± 6.59a

Nonanal 6.40 Z 0.74 ± 0.73a 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a

(57,56,98) G 0.73 ± 0.73a 0.67 ± 0.67a 0.00 ± 0.00a 2.80 ± 2.80a 4.41 ± 4.41a

D-Limonene 11.33 Z 0.00 ± 0.00a 0.00 ± 0.00a 2.26 ± 0.42a 1.81 ± 0.79a 1.78 ± 0.54a

(68,93,67) G 0.00 ± 0.00a 0.19 ± 0.18a 0.44 ± 0.44a 2.01 ± 2.01a 3.19 ± 1.41a

Benzene, 1,2-dichlora 15.37 Z 0.00 ± 0.00a 0.39 ± 0.39a 1.31 ± 0.18ab 1.80 ± 0.45b 0.00 ± 0.00a

(146,148,117) G 0.00 ± 0.00a 0.96 ± 0.96a 0.89 ± 0.89a 1.37 ± 1.37a 1.41 ± 1.41a

Unknown 1 2.47 Z 0.00 ± 0.00a 1.67 ± 1.67a 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a

G 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a

Unknown 2 2.87 Z 0.00 ± 0.00a 0.00 ± 0.00a 3.29 ± 3.29a 5.43 ± 5.43a 1.81 ± 0.48a

G 0.00 ± 0.00a 1.94 ± 1.94a 1.83 ± 1.83a 0.00 ± 0.00a 0.00 ± 0.00a

Unknown 3 4.52 Z 0.00 ± 0.00b 0.00 ± 0.00b 0.00 ± 0.00b 0.00 ± 0.00b 1.11 ± 0.13a

G 0.00 ± 0.00b 0.36 ± 0.36b 3.41 ± 0.06a 1.72 ± 0.92ab 1.51 ± 0.01ab

Unknown 4 4.77 Z 0.00 ± 0.00b 0.68 ± 0.68b 13.09 ± 1.61a 16.09 ± 5.24a 5.08 ± 0.37ab

G 1.05 ± 1.05a 6.72 ± 1.54a 4.77 ± 1.14a 2.52 ± 0.12a 4.21 ± 1.76a

Unknown 5 7.54 Z 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a 2.09 ± 1.14a 1.01 ± 0.35a

G 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a

Unknown 6 12.12 Z 0.00 ± 0.00a 0.00 ± 0.00a 0.98 ± 0.97a 1.03 ± 1.03a 0.00 ± 0.00a

G 0.00 ± 0.00a 0.47 ± 0.47a 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a

Unknown 7 14.50 Z 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a 1.15 ± 1.15a 0.00 ± 0.00a

G 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a 0.00 ± 0.00a

a dMeans within rows are significantly different (P < 0.05) according to the Student-Newman-Keuls test.

1Data are expressed as average peak area (x 105) ± SE (n = 2).

2Rt = retention time; ions: used for detection; Z = dairy farm in Zeytineli town (Izmir, Turkey); G = dairy farm in Germiyan town (Izmir,

Turkey).

Figure 5. Gas chromatogram of volatile compounds of Sepet cheese produced in Zeytineli (Izmir, Turkey). (1 = ethanol, 2 = ethyl acetate, 3 = unknown 2, 4 = 1-propanol, 5 = unknown 3, 6 = unknown 4, 7 = a-pinene, 8 = allyl acetate, 9 = unknown 5, 10 = acetoin, 11 = acetic acid, 12 = D-limonene, 13 = 2-heptanone, 14 = ethyl hexanoate, 15 = butyric acid, 16 = 2-tridecanone, 17 = ethyl octanoate, 18 = hexanoic acid, 19 = ethyl decanoate, 20 = octanoic acid, 21 = decanoic acid).

significantly (P < 0.01). After the first month of ripening, no significant changes in salt, fat, and TN contents in DM of both cheeses were observed. The pH 4.6-SN and TCA-SN fractions increased during production and ripening up to the third month (P < 0.05). Then, a slight increase in fractions continued during ripening. No significant (P > 0.01) differences in fat-in-DM, DM, salt-in-DM, TN in DM, pH 4.6 SN/TN, or TCA SN/TN were observed among Zeytineli and Germiyan cheeses. Moatsou et al. (2004) investigated the effect of artisanal rennet on the characteristics of Feta cheeses and found that the use of artisanal rennet had no significant effect on the evolutions of TS, ash, NaCl, fat, total protein, and proteolysis. A significant difference in titratable acidity was observed between Zeytineli and Germiyan cheeses. Zeytineli cheese produced with natural rennet had a higher lactic acid bacteria population than did Germiyan cheese that was produced with commercial rennet (data not shown). Olarte et al. (2000) reported the relation of lactic acid bacteria and initial lactic acid formation. Hatzikamari et al. (1999) also observed higher initial levels of lactic acid bacteria in low-pH cheeses.

Most of the volatile compounds identified were present at all stages of the Sepet cheese production and ripening. Indeed, their peak areas varied (Table 6). A sample chromatogram is given in Figure 5. All peak areas for FFA increased significantly up to the first month (P < 0.05). Then, slight increasing trends in decanoic and octanoic acids were observed in volatile composition of the cheese produced in Germiyan. The significant increases in hexanoic and butyric acids continued during ripening of cheese produced in Germiyan (P < 0.05). Significant increases in acetic, hexanoic, and

octanoic acids were observed up to the sixth month in cheese produced in Zeytineli (P < 0.05). At the end of ripening, the peak areas of FFA in the volatile fraction of cheese that was produced in Zeytineli with natural rennet were greater than the peak areas of FFA in the volatile fraction of cheese that was produced with commercial rennet in Germiyan. The difference might have arisen from lipases in artisanal rennet that were not present in the commercial rennet. Horne et al. (2005) reported that artisanal cheese had greater frequency of hexanoic acid than the industrial cheese and a wild LAB strain from the raw milk or lipases in artisanal rennets might have caused this difference.

Most esters showed a significant increase until the first month. Alcohols such as ethanol and 1-propanol also significantly increased until the first month in volatile composition of the cheese produced in Zeytineli (P < 0.05). Methyl ketones; 2-heptanone, and 2-trideca-none increased significantly (P < 0.05) until the first month of ripening and continued to increase slightly up to the third month. Same trends for FFA and esters were reported by Ziino et al. (2005) for Sicilian cheeses, by Tavaria et al. (2004) for Serra da Estrela Cheese, and by Pinho et al. (2001) for ewe cheeses.

CONCLUSIONS

In this research, the physicochemical, textural, volatile, and sensory profiles of traditional Sepet cheese, and changes in physicochemical and volatile composition during production and ripening were studied. Significant correlations were found between physicochemi-cal and textural parameters and volatile compounds. It was concluded that the variability of odor intensities

and textural parameters were mainly based on variations in some physicochemical parameters and ripening and lipolysis indices.

ACKNOWLEDGMENTS

This research was supported by the Research Funds of ízmir Institute of Technology (Project no 2008-ÍYTE-17). The authors thank the Biotechnology and Bioengineering Research Center and Environmental Research Center of ízmir Institute of Technology (Turkey) for providing the GC and GC/MS for aroma analysis. The authors also thank panel members for their participation.

REFERENCES

AO AC. 1995. Official Methods of Analysis. 16th ed. AOAC No. 920.124. Association of Official Analytical Chemists, Arlington, VA.

Attaie, R. 2009. Quantification of volatile compounds in goat milk Jack cheese using static headspace gas chromatography. J. Dairy Sci. 92:2435-2443. Ba§pinar, E., M. Mende§, and H. Qamdeviren. 2000. Multidimensional scaling analysis and its usage. Biyoteknoloji (KUKEM) . Dergisi 24:89-98.

Beuvier, E., and S. Buchin. 2004. Raw milk cheeses. Pages 319-345 in Cheese: Chemistry, Physics and Microbiology. P. Fox, P. L. H. Mc-Sweeney, T. M. Cogan, and T. P. Guinee, ed. Elsevier Academic Press, London, UK. Bieber, S. L., and D. V. Smith. 1986. Multivariate analysis of sensory

data: A comparison of methods. Chem. Senses 11:19-47. Brown, J. A., E. A. Foegeding, C. R. Daubert, M. A. Drake, and M. Gumpertz. 2003. Relationships among rheological and sensorial properties of young cheeses. J. Dairy Sci. 86:3054-3067. Bynum, O. G., and M. D. Barbano. 1985. Whole milk reverse osmosis retentates for cheddar cheese manufacture: Chemical changes during aging. J. Dairy Sci. 68:1-10. Carbonell, M., M. Nuñez, and E. Fernández-García. 2002. Evolution of the volatile components of ewe raw milk La Serena cheese during ripening. Correlation with flavour characteristics. Lait 82:683698.

Carunchia Whetstine, M. E. C., Y. Karagul-Yuceer, Y. K. Avsar, and M. A. Drake. 2003. Identification and quantification of character aroma components in fresh Chevre-style goat cheese. J. Food Sci. 68:2441-2447.

Chiofalo, A. Z., R. Costa, L. Liotta, L. Mondello, P. Dugo, and V. Chiofalo. 2004. Characterization of Maltese goat milk cheese flavor using SPME-GC/MS. S. Afr. J. Anim. Sci. 34:176-180. Curioni, P. M. G., and J. O. Bosset. 2002. Key odorants in various cheese types as determined by gas chromatography-olfactometry. Int. Dairy J. 12:959-984. Drake, M. A., S. C. McIngvale, P. D. Gerard, K. R. Cadwallader, and G. V. Civille. 2001. Development of a descriptive language for cheddar cheese. J. Food Sci. 66:1422-1427. Everard, C. D., C. P. O'Donnell, D. J. O'Callaghan, E. M. Sheehan, C. M. Delahunty, B. T. O'Kennedy, and V. Howard. 2007. Prediction of sensory textural properties from rheological analysis for process cheeses varying in emulsifying salt, protein and moisture contents. J. Sci. Food Agric. 87:641-650. Forde, A., and G. F. Fitzgerald. 2000. Biotechnological approaches to the understanding and improvement of mature cheese flavor. Curr. Opin. Biotechnol. 11:484-489. Fox, P. F., T. P. Guinee, T. M. Cogan, and P. L. H. McSweeney. 2000. Fundamentals of Cheese Science. Aspen Publishers Inc. Frederick, MD.

Frank, D. C., C. M. Owen, and J. Patterson. 2004. Solid phase microextraction (SPME) combined with gas-chromatography and olfac-tometry-mass spectrometry for characterization of cheese aroma compounds. Lebensm. Wiss. Technol. 37:139-154.

Friedeck, K. G., Y. Karagul-Yuceer, and M. A. Drake. 2003. Soy protein fortification of a low-fat dairy-based ice cream. J. Food Sci. 68:2651-2657.

Gonzalez De Llano, D., M. Ramos, C. Polo, J. Sanz, and I. Martinez-Castro. 1990. Evolution of the volatile components of an artisanal blue cheese during ripening. J. Dairy Sci. 73:1676-1683.

Gunasekaran, S., and M. M. Ak. 2003. Cheese Rheology and Texture. CRC Press LLC, Boca Raton, FL.

Gwartney, E. A., E. A. Foegeding, and D. K. Larick. 2002. The texture of commercial full-fat and reduced-fat cheese. J. Food Sci. 67:812-816.

Hatzikamari, M., E. Litopoulou-Tzanetaki, and N. Tzanetakis. 1999. Microbiological characteristics of Anevato: A traditional Greek cheese. J. Appl. Microbiol. 87:595-601.

Hayaloglu, A. A., E. Y. Brechany, K. C. Deegan, and P. L. H. McSwee-ney. 2008. Characterization of the chemistry, biochemistry and volatile profile of Kuflu cheese, a mould-ripened variety. LWT-Food Sci. Technol. 41:1323-1334.

Hayaloglu, A. A., S. Cakmakci, E. Y. Brechany, K. C. Deegan, and P. L. H. McSweeney. 2007. Microbiology, biochemistry, and volatile composition of Tulum cheese ripened in goat's skin or plastic bags. J. Dairy Sci. 90:1102-1121.

Heymann, H. 1994. A comparison of free-choice profiling and multidimensional scaling of vanilla samples. J. Sens. Stud. 9:445-453.

Holland, S. M. 2008. Non-metric multidimensional scaling (MDS). University of Georgia, Athens, GA. Accessed June 6, 2010. http:// www.uga.edu/strata/software/pdf/mdsTutorial.pdf.

Horne, J., S. Carpino, L. Tuminello, T. Rapisarda, L. Corallo, and G. Licitra. 2005. Differences in volatiles, and chemical, microbial and sensory characteristics between artisanal and industrial Piacentinu Ennese cheeses. Int. Dairy J. 15:605-617.

IDF (International Dairy Federation). 1981. Catalogue of cheeses. International Dairy Federation Bull. Doc. 141. International Dairy Federation, Brussels, Belgium.

IDF (International Dairy Federation). 1982. Cheese and processed cheese, determination of the total solids content. International Dairy Federation IDF Standard 4A. International Dairy Federation, Brussels, Belgium.

IDF (International Dairy Federation). 1988. Determination of salt content (Mohr method). International Dairy Federation IDF Standard 12B. International Dairy Federation, Brussels, Belgium.

IDF (International Dairy Federation). 1993. Milk—Determination of nitrogen content. International Dairy Federation IDF Standard 20B. International Dairy Federation, Brussels, Belgium.

IDF (International Dairy Federation). 1997. Milk and milk products: Determination of fat content (general guidance on the use of bu-tyrometric methods). International Dairy Federation IDF Standard 152A. International Dairy Federation, Brussels, Belgium.

Kamber, U. 2008. The traditional cheeses of Turkey: The Aegean region. Food Rev. Int. 24:39-61.

Karagul-Yuceer, Y., M. Isleten, and C. U. Pala. 2007. Sensory characterization of Ezine cheese. J. Sens. Stud. 22:49-65.

Karagul Yuceer, Y., B. Tuncel, O. Guneser, B. Engin, M. Isleten, K. Yasar, and M. Mendes. 2009. Characterization of aroma-active compounds, sensory properties, and proteolysis in Ezine cheese. J. Dairy Sci. 92:4146-4157.

Karaka§, R., and M. Korukluoglu. 2006. A traditional cheese: Microbiological and chemical properties of Sepet cheese. Gida. 31:169172.

Kinik, O., E. Ergullu, and N. Akbulut. 1999. A study on the production and some characteristics of Sepet cheese. Gida. 24:151-161.

Kruskal, J. B. 1964. Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis. Psychometrika 29:1-27.

Lawless, H. T., N. Sheng, and S. S. C. P. Knoops. 1995. Multidimensional scaling of sorting data applied to cheese perception. Food Qual. Prefer. 6:91-98.

Lawrence, R. C., L. K. Creamer, and J. Gilles. 1987. Texture development during cheese ripening. J. Dairy Sci. 70:1748-1760.

Liu, S. Q., R. Holland, and V. L. Crow. 2004. Esters and their biosynthesis in fermented dairy products: A review. Int. Dairy J. 14:923-945.

MacFie, H. J. J., and D. M. H. Thomson. 1984. Multidimensional scaling. Pages 351-375 in Sensory Analysis of Foods. J. R. Piggott, ed. Elsevier Science & Technology, London, UK.

McSweeney, P. L. H., and M. J. Sousa. 2000. Biochemical pathways for the production of flavor compounds in cheese during ripening: A review. Lait 80:293-324.

Meilgaard, M., G. V. Civille, and B. T. Carr. 1999. Sensory Evaluation Techniques. CRC Press, Boca Raton, FL.

Metin, M. 2006. Süt ve Mamülleri Analiz Yontemleri. Ege Üniversitesi Ege Meslek Yüksek Okullari Yayinlari No: 24. Izmir.

Moatsou, G., E. Moschopoulou, A. Georgala, E. Zoidou, I. Kandara-kis, S. Kaminarides, and E. Anifantakis. 2004. Effect of artisanal liquid rennet from kids and lambs abomasa on the characteristics of Feta cheese. Food Chem. 88:517-525.

Nogueira, M. C. L., G. Lubachevsky, and S. A. Rankin. 2005. A study of the volatile composition of Minas cheese. LWT-Food Sci. Tech-nol. 38:555-563.

Olarte, C., S. Sanz, E. Gonzalez-Fandos, and P. Torre. 2000. The effect of a commercial starter culture addition on the ripening of an artisanal goat's cheese (Cameros cheese). J. Appl. Microbiol. 88:421-429.

Picon, A., P. Gaya, E. Fernández-García, A. Rivas-Cañedo, M. Ávila, and M. Nuñez. 2010. Proteolysis, lipolysis, volatile compounds, texture, and flavor of Hispánico cheese made using frozen ewe milk curds pressed for different times. J. Dairy Sci. 93:2896-2905.

Pinho, O., I. M. P. L. V. O. Ferreira, S. Casal, J. O. Fernandes, M. B. P. P. Oliveira, and M. A. Ferreira. 2001. Method optimization for analysis of the volatile fraction of ewe cheese by solid-phase microextraction. Chromatographia 53(Suppl.):S390-S393.

Pinho, O., C. Pérès, and I. M. P. L. V. O. Ferreira. 2003. Solid-phase microextraction of volatile compounds in "Terrincho" ewe cheese: Comparison of different fibers. J. Chromatogr. A 1011:1-9.

Popper, R., and H. Heymann. 1996. Analyzing differences among products and panelists by multidimensional scaling. Pages 1-35 in Multivariate Analysis of Data in Sensory Science. T. Naes and E. Risvik. ed. Elsevier, Amsterdam, the Netherlands.

Prasad, N., and V. B. Alvarez. 1999. Effect of salt and chymosin on the physico-chemical properties of feta cheese during ripening. J. Dairy Sci. 82:1061-1067.

Qian, M., C. Nelson, and S. Bloomer. 2002. Evaluation of fat-derived aroma compounds in blue cheese by dynamic headspace GC/Ol-factometry-MS. J. Am. Oil Chem. Soc. 79:663-667.

Qian, M., and G. Reineccius. 2002. Identification of aroma compounds in Parmigiano-Reggiano cheese by gas chromatography/olfactometry. J. Dairy Sci. 85:1362-1369.

Rabinowitz, G. B. 1975. An introduction to nonmetric multidimensional scaling. Am. J. Pol. Sci. 19:343-390.

Renner, E. 1986. Milchpraktikum. Fachgebiet Milchwissenschaft. Jus-tus-Liebig-Universität Giessen, Germany.

Sánchez-Macías, D., M. Fresno, I. Moreno-Indias, N. Castro, A. Mo-rales-delaNuez, S. Alvarez, and A. Argüello. 2010. Physicochemi-cal analysis of full-fat, reduced-fat and low-fat artisan-style goat cheese. J. Dairy Sci. 93:3950-3956.

Shakeel-Ur-Rehman, and F. P. Fox. 2002. Effect of added a-ketoglutaric acid, pyruvic acid or pyridoxal phosphate on proteolyis and quality of Cheddar cheese. Food Chem. 76:21-26.

Singh, T. K., M. A. Drake, and K. R. Cadwallader. 2003. Flavor of Cheddar cheese: A chemical and sensory perspective. Compr. Rev. Food Sci. Food Safety. 2:139-161.

Tarakci, Z., and E. Kucukoner. 2006. Changes on physicochemical, lipolysis and proteolysis of vacuum packed Turkish Kashar cheese during ripening. J. Centr. Eur. Agric. 7:459-464.

Tarakci, Z., and H. Temiz. 2009. A review of the chemical, biochemical and antimicrobial aspects of Turkish Otlu (herby) cheese. Int. J. Dairy Technol. 62:354-360.

Tavaria, F. K., A. C. Silva Ferreira, and F. Xavier Malcata. 2004. Volatile free fatty acids as ripening indicators for Serra da Estrela cheese. J. Dairy Sci. 87:4064-4072.

Turkoglu, H., Z. G. Ceylan, and K. S. Dayisoylu. 2003. The microbiological and chemical quality of Orgu cheese produced in Turkey. Pakistan J. Nutr. 2:92-94.

Van den Dool, H., and P. D. Kratz. 1963. A generalization of the retention index system including linear temperature programmed gasliquid partition chromatography. J. Chromatogr. A 11:436-471.

Weichselbaum, E., B. Benelam, and H. S. Costa. 2009. Synthesis report No. 6: Traditional foods in Europe. European Food Information Resource (EuroFIR) Project Management Office, Institute of Food Research, Norwich, UK.

Ziino, M., C. Condurso, V. Romeo, D. Giuffrida, and A. Verzera. 2005. Characterization of ''Provola dei Nebrodi'', a typical Sicilian cheese, by volatiles analysis using SPME-GC/MS. Int. Dairy J. 15:585-593.