Behaviour of volatile compounds during the shelf life of yoghurt

doi: 10.1111/j.1471-0307.2009.00476.x ORIGINAL RESEARCH IDT International Blackwell Oxford, 1364-727X XXX 1364-0307 Society of UK Dairy Publishing Journal Technology of Ltd Dairy2009 Technology O RI G I NA L RESEARCH Behaviour of volatile compounds during the shelf life of yoghurt SANDR A MAR IA PIN TO,* M AR IA DAS GR AÇ AS C L E M E NT E and LUIZ RONALDO DE AB R E U Department of Food Science, Federal University of Lavras, Lavras, Minas Gerais, Brazil The production of acetaldehyde, diacetyl and ethanol was evaluated in whole plain yoghurts manufactured with commercial starter cultures, yoghurt acquired in a local market, and milk fermented by a single culture of either Streptococcus thermophilus or Lactobacillus delbrueckii ssp. bulgaricus. The headspace technique was used for sample preparation, following identification and quantification by gas chromatography. During an 8-h incubation period, mixed cultures were the most efficient in lowering the pH (from 6.30 to 4.8), followed by S. thermophilus ( from 6.30 to 5.18) and L. bulgaricus ( from 6.30 to 5.8). During the storage period, however, a single culture of L. bulgaricus decreased the pH more than S. thermophilus, but still less than the mixed commercial cultures. Plain yoghurts acquired in the market, those made with commercial starter cultures, and fermented milks obtained with single cultures showed, after 21 days of storage, concentrations of acetaldehyde from 11 to 35 mg/L, and of diacetyl from 0 to 0.85 mg/L. An increasing concentration of ethanol was observed during the storage period, and its production was observed even in the incubation stage of all products. It was also observed that the acetaldehyde concentration was inversely correlated to ethanol production in some products. The combination of headspace, identification and quantification techniques by gas chromatography in this work was efficient in the identification and quantification of the major aromatic compounds and ethanol content of yoghurt. Keywords Acetaldehyde production, Flavour, Diacetyl, Shelf life, Volatile compounds, Yoghurt. *Author for correspondence. E-mail: sandra@ufla.br I N T RO D U C T I O N *Author for correspondence. E-mail: sandra@ufla.br © 2009 Society of Dairy Technology Degradation processes of organic matter generally result in substances that lead to unpleasant texture and flavour and may be harmful to health. Nonetheless, modifications originating from micro-organisms utilized in food fermentation are beneficial, most often improving flavour and texture, and frequently increasing the contents of vitamins and organic acids, which contributes to the extension of the product’s shelf life. These effects are basically due to the partial action of fermentative micro-organisms on one or more food compounds: carbohydrates, proteins and fats (Tamime and Robinson 2007). The typical aroma of plain yoghurt is generated from metabolites produced by starter cultures, which secrete a wide range of compounds that may contribute to the formation of flavour. However, the most significant ones are lactic acid and the carbonyl compounds acetaldehyde and diacetyl (Gardini et al. 1999; Gallardo-Escamilla et al. 2005). Lactic acid is the main component of flavour and taste of fermented milks. The amount of this acid present in the product determines, to some extent, its acceptability by consumers. On the other hand, excessive quantities negatively modify aroma and taste. Lactobacillus delbrueckii ssp. bulgaricus is the major acetaldehyde-producing micro-organism in yoghurt, with the intensity of production dependent on the strain utilized in the fermentation. When associated with Streptococcus thermophilus, the rate of acetaldehyde production is considerably increased compared to production by L. bulgaricus alone (Sodini et al. 2000; Chaves et al. 2002; Gallardo-Escamilla et al. 2005). Therefore, the protosymbiotic association between those species favourably contributes to the production of this important aromatic compound throughout the preparation of yoghurt. M AT E R I A L S A N D M E T H O D S The research was conducted in the Laboratories of the Dairy Technology, Microbiology, and Sensory Analysis of the Department of Food Science of the Federal University of Lavras-Brazil in four distinct phases, as listed below. Vol 62, No 2 May 2009 International Journal of Dairy Technology 215 Vol 62, No 2 May 2009 Phase I: Evaluation of yoghurts manufactured with single and mixed cultures The work in this phase was done with the purpose of assessing the contributions of each micro-organism to the production of volatile compounds (acetaldehyde and diacetyl) and ethanol in yoghurt. Frozen cultures, the first composed of either micro-organism (S. thermophilus or L. bulgaricus) as a single culture, and the second by a mixed culture of both microorganisms, were inoculated in heat-treated milk. The generation of aromatic flavour compounds was surveyed every 7 days from the onset of incubation to the 42nd day of the storage period. Milks inoculated with cultures of L. bulgaricus, S. thermophilus and with a mixed culture of both were incubated at temperatures of 45°C, 40°C and 42°C, respectively, as indicated by the manufacturer’s laboratory. Phase II: Evaluation of yoghurts made with different commercial cultures In this phase, four different lyophilized commercial starter cultures were evaluated. All cultures were of high viscosity and had medium post-acidification properties. Yoghurts made with these cultures were evaluated from the second to the 35th day of the storage period. Yoghurt production Yoghurts were prepared according to traditional techniques described by Tamime and Robinson (2007), but with different commercial mixed cultures. The process was strictly the same for every culture, with only the sizes of the inoculums varying, as recommended by the manufacturers. Raw material Milk with a low somatic cell count was obtained from a cow that was not treated with any sort of medicine during the experimental period. Milk fat was standardized to 3%, producing so-called ‘whole’ yoghurt. Phase III: Evaluation of plain yoghurts In this phase, a trial was done every 7 days, in triplicate, with plain yoghurts (without either sugar, flavouring, colouring or fruit pulp) commercialized in Lavras and the surrounding areas, with a storage period of between 14 and 42 days. Collecting samples in the market Samples from different brands of plain yoghurts were collected at a regional market. At the time of sampling, the following data were collected: brand identification, manufacturing date, expiration date, display temperature, volume and type of packaging. Samples were properly conditioned in isothermic boxes containing ground ice and taken to the laboratory for analysis. 216 © 2009 Society of Dairy Technology Analyses during phases I, II and III All products in both Figures 6 and 9 were analysed starting from 14 days after manufacturing (based on the information in the labels—manufacturing date). They were from different brands, and produced with different technologies, including different starting cultures. pH The pH values of pre-heated (20°C) samples were measured with a calibrated pH-meter, model D474 (Micronal, São Paulo, Brazil). Chromatographic analysis Samples were prepared for chromatographic analyses according to the headspace method, slightly adapted from a technique kindly provided by the ‘Wiesby Starter Culture and Media Laboratory’. Sample preparation In a screw cap test tube with a central hole and septum were added: 2.5 mL of yoghurt, 2.5 mL of de-ionized distilled water, 2.5 g of sodium chloride and 1.0 mL butanol as an internal standard were added to a screw cap test tube with a central hole and a septum. The preparations were submitted to a freezing temperature (–18°C) for 24 h. After that, the tubes were inserted into a heating block. The temperature was gradually elevated to 80°C and kept at this temperature for 1 h with homogenization every 15 min with a vortex. Sample injection After sample homogenization, 500 µL of headspace was collected through the septum with a Gastight syringe (Hamilton Company, Reno, NV, USA) and injected into the chromatographer. Analysis of the main volatile flavour compounds (acetaldehyde and diacetyl) and ethanol was performed according to a technique proposed by Ott et al. (1997) and adapted by Abreu (1993) and Pinto (2001). A Varian 3800 gas chromatographer (Varian Inc., Palo Alto, CA, USA), connected to a Varian Star 4.5 work station and equipped with a Varian-Wax chromatographic column (30 m × 0.320 mm; 0.25 µm) and a flame ionization detector (FID), was used to separate and quantify the target compounds. The equipment conditions were: Injector: Model 1079; temperature: 200°C; split ratio: 5; sample injection: 500 µL of headspace. Column conditions: Initial temperature: 40°C (with 10 s of holing time); final temperature: 65°C (with an elevation rate of 4°C per minute, holding 1 min at the final temperature). Flow: 0.7 mL per min; total running time: 7.35 min. Detector: Type: FID; temperature: 200°C; sensitivity: 10–12 volts; attenuation: 1. Vol 62, No 2 May 2009 Peak identification Standards (acetaldehyde, diacetyl, ethanol and butanol) of known concentrations were injected for peak identification under the same conditions used for sample injection. Retention time and temperature for each peak were verified. Peak quantification Quantification of each compound was done by calibrating the reference standards (acetaldehyde, diacetyl and ethanol) and internal standard (butanol). All standards were of chromatographic purity grade (Sigma and Aldrich, St Louis, MO, USA). Since the viscosity and texture of the dilution medium for the standards had to be similar to yoghurt, it was necessary to perform a deodorization of a yoghurt, which served as a medium basis. Yoghurt deodorization Following the technique of Ott et al. (2000), skim yoghurt was kept in a water bath at 80°C for five hours with a rotavapor. After that, the sample was re-hydrated with deionized distilled water in the same amount that had evaporated. Figure 1 pH values from the incubation period (a) to 42 days of storage (b), in products obtained by fermentation with pure Streptococcus thermophilus culture, pure Lactobacillus delbrueckii ssp. bulgaricus culture, and mixed culture of S. thermophilus and L. bulgaricus. Standard sample preparation Samples were prepared by mixing 2.5 mL rehydrated deodorized yoghurt, 1.5 mL deionized distilled water, 0.5 mL butanol (20 mg/L), 0.5 mL diacetyl (20 mg/L), 0.5 mL acetaldehyde (20 mg/L), 0.5 mL ethanol (20 mg/L) and 2.5 mL sodium chloride (100 mg/L). Method calibration All sample peaks were identified as the listed standards and quantified against the internal standard. Statistical analysis The appropriate means were compared with SISVAR4.3 (System of Statistical Analysis) by analysing the variance and comparing the means of the treatments by the Scott–Knott test (Ferreira 1990). R E S U LT S A N D D I S C U S S I O N pH The data of the pH values of fermented milks produced by the mixed culture and by pure cultures of either S. thermophilus or L. bulgaricus are shown in Figure 1. Mixed cultures were more efficient in lowering the pH values, followed by S. thermophilus and L. bulgaricus. According to Ferreira (1993) and Gün and Issikli (2006), in the initial fermentation period, S. thermophilus decreases the pH down to values around 5.5, when it produces formic acid and consumes oxygen. These actions create better conditions for L. bulgaricus development, which in turn rapidly decreases pH values further, © 2009 Society of Dairy Technology Figure 2 pH values of yoghurts produced with commercial culture starters, measured throughout the storage period. subsequently generating appreciable amounts of acetaldehyde and diacetyl that contribute to the distinct flavour of the fermented product. Within 8 h of incubation, the mixed culture lowered the pH value from 6.30 to 4.8, whereas isolated cultures of S. thermophilus and L. bulgaricus with the same incubation period lowered the pH value to 5.18 and 5.87, respectively (Figure 1), reinforcing the well-known existence of a strong synergistic association between these species. Throughout the storage period, the pH value of milk fermented by the mixed culture was lower compared to those fermented by pure cultures. During the studied period, the pure culture of L. bulgaricus decreased the pH faster and to lower values in comparison to S. thermophilus. An accelerated decrease in pH values was observed from the onset to the fifth day of storage (Figure 2), 217 Vol 62, No 2 May 2009 Figure 3 pH values of yoghurts obtained in a market, measured throughout the storage period. except for one yoghurt, in which this was extended to the sixth day. After that, the pH values remained virtually constant. All yoghurts, regardless of whether they were produced in the laboratory or acquired in the market, showed pH values varying from 3.9 to 4.3, in agreement with those found in the literature and within the desirable range for a high quality product (Runge et al. 2003; Salvador and Fiszman 2004; Ferreira 1993). The data on the pH values of yoghurts acquired in the market are shown in Figure 3. All yoghurts were acquired after 14 days of production according to information on the label. The pH values varied within the range generally accepted for yoghurt. The causes of such variations cannot be precisely determined. The storage and delivery conditions are quite variable, mainly due to oscillating temperature values that remarkably influence microbial growth and enzymatic activity in yoghurt. Furthermore, different commercial brands may have been produced by different starter cultures, which may have different acidification properties. Acetaldehyde The results of acetaldehyde production during the incubation and storage periods are presented in Figure 4. From the onset up to 6 h of incubation, acetaldehyde production was higher in product 1 (S. thermophilus alone), followed by product 2 (mixed culture) and then product 3 (Lactobacillus bulgaricus alone) (Figure 4a). At this point, the production stopped in product 1. A slow but steadily increasing production of acetaldehyde was observed in product 2 until the end of the incubation period. In contrast, acetaldehyde production increased up to 7 h in product 3 and thereafter slightly decreased. There are a considerable number of works (Ferreira 1993; Runge et al. 2003; Tamime and Robinson 2007) demonstrating that the conditions at the beginning of incubation, such as a fairly high pH, the presence of a certain amount of oxygen, and the lack of a small amount of formic acid, are 218 © 2009 Society of Dairy Technology Figure 4 Acetaldehyde concentrations in fermented milks obtained with mixed and pure cultures, from inoculation to 42 days of storage (The standard deviations of the data in Figures 1 and 4 are shown in Tables 1 and 2). more favourable for S. thermophilus than for L. bulgaricus. However, with the development of the S. thermophilus, the conditions become more favourable for the L. bulgaricus. It is also well known that S. thermophilus has its growth interrupted early because it does not tolerate a pH lower than 5.5. During the storage period at 4°C (Figure 4b), as the conditions became favourable to Lactobacillus bulgaricus, product 2 generally produced higher concentrations of acetaldehyde compared to products 1 and 3. Such results reinforce the findings that demonstrate that L. bulgaricus is the major acetaldehyde-producing micro-organism in yoghurt, although its production is initiated later than that of S. thermophilus. The results do not allow for the accurate quantification of the actual production of acetaldehyde once the compound is highly volatile. The variability in the results obtained, especially for product 2, may be due to the elevated volatility of the acetaldehyde. The concentrations of acetaldehyde in yoghurts made with different commercial cultures are presented in Figure 5. In general, the content of acetaldehyde decreased until 14 days of storage, and then had a significant increase up to 21 days, when it reached its maximum concentrations, and then decreased again. Similar results were obtained by Laye et al. (1993), who found that the concentrations of acetaldehyde of Vol 62, No 2 May 2009 Table 1 Corresponding to Figure 1 (A and B) A Incubation (hours) 0.5 2 6 7 8 Product Mean SD Mean SD Mean SD Mean SD Mean SD 1 2 3 6.26 6.30 6.30 0.3512 0.4583 0.3055 6.19 6.20 6.12 0.4509 0.4041 0.3511 5.45 6.15 5.16 0.2466 0.3329 0.2507 5.32 6.02 4.93 0.5001 0.3464 0.4976 5.18 5.87 4.83 0.2275 0.4005 0.2007 B Storage (days) 1 Product Mean SD 1 2 3 4.80 4.79 4.25 7 14 21 28 35 42 Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD 0.3011 4.37 0.2456 4.20 0.3499 3.85 0.5437 4.69 0.3605 4.37 0.2291 4.00 0.3536 4.70 0.3743 4.39 0.5012 4.03 0.6245 4.70 0.3651 4.39 0.2517 4.06 0.2901 4.70 0.3483 4.39 0.3055 4.08 0.2950 4.70 0.2951 4.39 0.3253 4.08 0.2476 0.3005 0.5245 Table 2 Corresponding to Figure 4 (A and B) A Incubation (hours) 0.5 2 6 7 8 Product Mean SD Mean SD Mean SD Mean SD Mean SD 1 2 3 0.45 0.70 0.60 0.1501 0.1743 0.0513 1.20 0.70 1.80 0.2646 0.1212 0.2784 14.60 6.50 8.00 0.2011 0.6245 0.6083 15.01 7.97 14.93 0.8888 0.3464 0.9539 13.11 9.02 12.98 0.8718 0.2001 0.9001 B Storage (days) 1 Product Mean SD 1 2 3 7 14 21 28 35 42 Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD 11.99 0.7549 16.05 0.7001 12 30.00 0.9178 14.12 0.6557 14 16.02 0.7211 20.22 0.6083 14 0.6245 14.34 0.5196 9.03 0.5196 10.32 0.6083 10.29 0.7937 0.2646 29.76 0.8021 20.87 0.2646 17.89 0.2646 27.79 0.6083 0.0998 16.32 0.4001 7.89 0.5568 10.10 0.1732 11.88 0.1732 SD, standard deviation. Figure 5 Acetaldehyde concentrations during the storage period in yoghurts prepared with commercial starter cultures. © 2009 Society of Dairy Technology Figure 6 Acetaldehyde concentrations in yoghurts acquired from a market. 219 Vol 62, No 2 May 2009 four brands of yoghurt decreased from the onset up to 12 days of storage. Brandão (1980) reported that during the storage process the concentration of acetaldehyde decreased, with the highest peak appearing after 14 days and decreasing afterwards in all the yoghurts evaluated. The concentrations of acetaldehyde in yoghurts acquired in the market are presented in Figure 6. A remarkable variation was observed in the acetaldehyde concentrations in the yoghurts acquired from the markets (Figure 6), ranging from 11 to 35 mg/L at 21 days of storage. This behaviour may be due to different strains and technologies used by the industry in addition to variations in storage temperature at the market place. By comparing the results presented in Figures 5 and 6, a lower variation is observed in the acetaldehyde concentration during the storage period in the yoghurts prepared with commercial cultures compared to those acquired in the market. The difference is certainly due to the more accurate control of the temperature and the standardization in the technological process of production in the pilot plant. One of the problems in the commercial production of yoghurt in many places worldwide is the lack of the necessary low temperatures during the transport and storage of the product. In many cases, temperature fluctuations bring about the process of postacidification, which invariably shortens the shelf life of the yoghurt. To offset this problem, many industries utilize starter cultures that possess low post-acidification properties. Such characteristics are attained with high cocci/bacilli ratios, which in turn lead to low acetaldehyde production. Diacetyl Data in Figure 7 refer to the diacetyl production by a pure culture of S. thermophilus, by a pure culture of L. bulgaricus and by a mixed culture of L. bulgaricus and S. thermophilus, starting from incubation to the 42nd day after manufacturing. Runge et al. (2003) considered S. thermophilus as the main diacetyl producer, and the importance of L. bulgaricus was judged as not being significant for the production of diacetyl. Indeed, the pure culture of L. bulgaricus did not produce diacetyl throughout the experimental period, whereas the S. thermophilus pure culture produced a substantial amount of diacetyl when inoculated individually. This amount was followed by the mixed culture, which presented an intermediary production (Figure 7). During the incubation period (Figure 7) with pure S. thermophilus, the culture began diacetyl production 2 h after incubation, which is much earlier than the mixed culture (L. bulgaricus and S. thermophilus) did (7 h after incubation). During the storage period, pure S. thermophilus cultures and mixed cultures initiated a massive increase 220 © 2009 Society of Dairy Technology Figure 7 Diacetyl concentration, starting at incubation (a) to the 42nd storage day (b), in products prepared with a pure culture of Streptococcus thermophilus, a pure culture of Lactobacillus delbrueckii ssp. bulgaricus, and a mixed culture of L. bulgaricus and S. thermophilus. of diacetyl production after the 21st day. Single L. bulgaricus cultures started a moderate increase after the 28th storage day. Although diacetyl production was relatively low (varying from 0.09 to 0.44 mg/L) in yoghurts prepared with mixed cultures (Figure 7), it must be noted that the detection limit for this compound is even lower than the concentrations encountered, reinforcing the findings that diacetyl is among the main aroma-conferring compounds and the second most potent aromatic compound in yoghurt, after acetaldehyde. The evolution of diacetyl concentration throughout the storage period in yoghurts prepared with commercial cultures is depicted in Figure 8. The production peak of this compound was observed in all yoghurts after 21 days. The acetaldehyde production peak also presented after 21 storage days (Figure 5). The two yoghurts with the highest diacetyl production (yoghurts 2 and 3) also had the highest production of acetaldehyde (Figures 5 and 8). Similarly, those with the lowest diacetyl production (yoghurts 1 and 4) were also less aromatic when considering acetaldehyde (Figures 5 and 8). It is important to note that when the mixed cultures of cocci and bacilli were inoculated, there was little diacetyl production (0.09 to 0.44 mg/L) (Figure 7). On the other hand, in yoghurts prepared with mixed commercial cultures, the diacetyl Vol 62, No 2 May 2009 Figure 8 Diacetyl concentrations throughout the storage period in yoghurts prepared with commercial cultures. Figure 10 Ethanol concentrations from incubation to the 42nd day of storage in yoghurts prepared with pure Streptococcus thermophilus cultures, pure Lactobacillus delbrueckii ssp. bulgaricus cultures, and mixed cultures of L. bulgaricus and S. thermophilus. Figure 9 Diacetyl concentrations in yoghurts acquired in a local market. production range was greater (0.10 to 0.78 mg/L). This may be explained by the high ratio cocci/ bacilli in commercial starter cultures. The quantity of bacilli is insignificant in those cultures, justifying the high production of diacetyl, because the presence of L. bulgaricus could potentially restrain the production of that compound. A very substantial variation in diacetyl production was observed in yoghurts marketed locally (Figure 9). Yoghurt B, with the lowest diacetyl production, also had the lowest quantity of acetaldehyde (0.10 to 1.91 mg/L) throughout the experimental period (Figure 6). On average, yoghurts obtained at the local market were more aromatic than the ones prepared with a commercial culture starter when considering either diacetyl or acetaldehyde production. These differences can be attributed to particular characteristics of the strains or to variations in manufacturing technologies adopted by industry. Ethanol When a certain amount of acetaldehyde is accumulated in yoghurt, an enzyme called alcohol dehydrogenase starts reducing part of acetaldehyde to © 2009 Society of Dairy Technology ethanol. This reaction is unsuitable for the production of a high quality yoghurt since the flavour may be prominently affected, which is one of the most important causes of shelf-life curtailment in fermented milks. Throughout the incubation period, all microorganisms produced ethanol, although there was a relatively higher production in yoghurts prepared with mixed culture. When acetaldehyde production is compared to ethanol production in this phase (Figures 4 and 10), the lower acetaldehyde concentration is probably due to an increased alcohol dehydrogenase enzyme activity, which reduces acetaldehyde to ethanol, subsequently diminishing the acetaldehyde concentration and increasing the ethanol concentration. In yoghurts prepared with commercial culture starters and obtained at the local market, the acetaldehyde concentration reached its peak after 21 days (Figures 5 and 6), followed by a decline and a simultaneous increase in ethanol concentration (Figures 11 and 12). Among the yoghurts from the local market, the one that presented the lowest production of acetaldehyde (yoghurt B) also produced the most significant quantity of ethanol, which confirms the activity of alcohol dehydrogenase. After 21 storage days, those yoghurts (Figure 12) had a smaller range of ethanol production (3.3 to 40 mg/L) when compared to yoghurts prepared with commercial culture starters (Figure 11). 221 Vol 62, No 2 May 2009 obtained similar values for ethanol production when analysing the aromatic compounds in distinct cultures for yoghurt preparation. Figure 11 Ethanol concentrations throughout the storage period in yoghurts prepared with commercial culture starters. Methodology utilized in extraction, identification and quantification of volatile compounds through gaseous chromatography After some attempts to extract the sample using different kinds of flasks and septi, internal standards, chromatographer conditions, microsyringes, etc., it was concluded that the headspace method for sample extraction and chromatographer conditions would best suit the requirements for the identification and quantification of the evaluated peaks. Figure 13 shows the chromatographic profiles of seven samples of distinct yoghurts. Although the peaks are relatively small, they are of high resolution and provide satisfying results that are in agreement with the literature. CONCLUSIONS Figure 12 Average values for ethanol concentration in yoghurts obtained in a market. Both plain yoghurts obtained in a local market and those prepared with commercial starters presented high acetaldehyde concentrations and low diacetyl concentrations within the same storage period. There was no diacetyl production in yoghurts prepared with pure cultures of L. bulgaricus, but there was a larger production of diacetyl in milk fermented by pure cultures of S. thermophilus than in milk inoculated with both micro-organisms. Ethanol was produced during the incubation and storage period in all samples of yoghurts analysed, and this production increased considerably during storage from the beginning to the 21st day. A C K N OW L E D G E M E N T S The authors wish to thank CNPq—Brazil for financial support. REFERENCES Figure 13 Chromatographic profiles of seven samples of distinct yoghurts. This fact may explain the higher acetaldehyde concentrations in these yoghurts (Figure 6) than in the ones prepared with commercial starter cultures (Figure 5). There was a smaller reduction of acetaldehyde to ethanol in the locally marketed yoghurts. Brandão (1980) and Al-Kadamany et al. (2002) 222 © 2009 Society of Dairy Technology Abreu L R (1993) Factors affecting the biosynthesis of branched-chain fatty acids in milk fat. PhD Thesis, pp 163. Madison, WI: University of Wisconsin. Al-Kadamany E, Toufeili I, Khattar M, Abou-Jawdeh Y, Harakeh S and Haddad T (2002) Determination of shelf life of concentrated yoghurt (labneh) produced by in-bag straining of set yoghurt using hazard analysis. Journal of Dairy Science 85 1023–1030. Brandão S C C (1980) Determination of volatile flavour constituents and residual carbohydrates during the fermentation of yoghurt. PhD Thesis, pp 126. East Lansing, MI: Michigan State University. Chaves A C S D, Fernandez M, Lerayer A L S, Mierau I, Kleerebezem M and and. Hugenholtz J (2002) Metabolic engineering of acetaldehyde production by Streptococcus thermophilus. Applied Environmental Microbiology 68 5656–5662. Vol 62, No 2 May 2009 Ferreira C L L F (1993) Produtos Lácteos Fermentados: Aspectos Bioquímicos e Tecnológicos, pp 96. Viçosa, Minas Gerias, Brazil: Universidade Federal de Viçosa (technical booklet). Ferreira D B (1990) Sisvar—Sistema de Análise de Variância. Lavras, Minas Gerias, Brazil: UFLA. Gallardo-Escamilla F J, Kelly A L and Delahunty C M (2005) Influence of starter culture on flavour and headspace volatile profiles of fermented whey and whey produced from fermented milk. Journal of Dairy Science 88 3745– 3753. Gardini F, Lanciotti R, Guerzoni M E and Torriani S (1999) Evaluation of aroma production and survival of Streptococcus thermophilus, Lactobacillus delbrueckii subsp. bulgaricus and Lactobacillus acidophilus in fermented milks. International Dairy Journal 9 125–134. Gün Ö and Issikli N D (2006) The effects of fat and non fat dry matter concentration and storage time on the physical properties and acidity of yoghurts made with probiotic cultures. Food Science and Technology International 12 467–476. Laye I, Karleskind D and Morr C V (1993) Chemical, microbiological and sensory properties of plain nonfat yoghurt. Journal of Food Science 58 991–995. Ott A, Fay L B and Chaintreau A (1997) Determination and origin of the aroma impact compounds of yoghurt © 2009 Society of Dairy Technology flavour. Journal of Agricultural and Food Chemistry 45 850–857. Ott A, Hugri A, Baumgartner M and Chaintreau A (2000) Sensory investigation of yoghurt flavour perception: mutual influence of volatiles and acidity. Journal of Agricultural and Food Chemistry 48 441–450. Pinto S M (2001) Estudo do acetaldeído, diacetil e etanol em leites fermentados Doctorate Thesis. pp 87. Lavras, Minas Gerias, Brazil: Federal University of Lavras. Runge M O, Folkenberg D M, Guldager H S, Stenby E and Skriver A (2003) Sensory evaluation of yoghurt flavour. Proceedings of the IDF Seminar on Aroma and Texture of Fermented Milk Held in Kolding, pp 393–396. Kolding, Denmark: International Dairy Federation. Salvador A and Fiszman S M (2004) Textural and sensory characteristics of whole and skimmed flavoured set-type yoghurt during long storage. Journal of Dairy Science 87 4033–4041. Sodini I and Corrieu G and Latrille E (2000) Identification of interacting mixed cultures of lactic acid bacteria by their exclusion from a model predicting the acidifying activity of non-interacting mixed cultures. Applied Microbiology and Biotechnology 54 715–718. Tamime A Y and Robinson R K (2007) Yoghurt: Science and Technology, eds. pp 808. Cambridge, UK: Woodhead. 223