Influence of Temperatures and Fermentation Behaviour of Mixed Cultures of

Food Sci. Technol. Res., 19 (5), 781 793, 2013 Influence of Temperatures and Fermentation Behaviour of Mixed Cultures of Williopsis saturnus var. saturnus and Saccharomyces cerevisiae Associated with Winemaking Hasan Tanguler * Nigde University, Department of Food Engineering, 51245 Nigde Turkey Received September 12, 2012; Accepted June 25, 2013 Up to now, this has been the first study in which the influences of fermentation temperature on the yeast growth and the production of yeast-derived volatile compounds during the fermentation of Emir grape must at various temperatures were examined. The results demonstrate that the fermentation temperature plays an important role compared to all tested variables. Fermentations were completed in 6, 8 and 14 days at 12, 18 and 24, respectively. Increase in temperature resulted in an increase in glycerol, total acidity, acetic and tartaric acid, 2-methyl butanol, propan-1-ol, isobutanol, acetaldehyde and acetone, but in a decrease in ethanol, malic and citric acid, isoamyl acetate, ethyl acetate, isobutyl acetate, ethyl butyrate and ethyl hexanoate. Moreover, these mixed culture fermentations formed higher amounts of isoamyl acetate in comparison with pure culture of S. cerevisiae. According to chemical composition and volatile compounds the differences between obtained wines were found generally significant. Keywords: white wine, temperature, Williopsis saturnus var. saturnus, Saccharomyces cerevisiae, flavour compounds, isoamyl acetate Introduction Wine production is a combination of complex interactions involving grape variety, microbiota and winemaking technology (Torija et al., 2003). Among the most important factors affecting the quality of the wine are the clarification and composition of the grape juice, the sulfur dioxide level added, the interaction with other indigenous microorganisms, the supplementation with nutrients, strain and amount of inoculated yeasts, and the fermentation temperature. The fermentation temperature is one of the most important parameters for the production of wine since it can affect the biochemical reactions and metabolism of the yeasts especially non-saccharomyces and, as a result, the formation of secondary metabolites such as glycerol, acetic acid and succinic acid, thus determining the chemical and organoleptic qualities of the wine (Fleet and Heard, 1993; Torija et al., 2003). White wines are often fermented in the range of 10 20. Nevertheless, some European wineries still prefer *To whom correspondence should be addressed. E-mail: htanguler@nigde.edu.tr; tangulehasanr@gmail.com fermentation temperatures between 20 25 (Sener et al., 2007). Recently, low temperature (10 15 ) fermentations are becoming more frequent due to the winemaker s tendency to enhance the production of some volatile compounds and improve the wine aromatic profile (Beltran et al., 2008; Andorà et al., 2010). Another important parameter for the production of wine is yeast. It can strongly affect the quality and flavour of the final product. Among several yeasts, Saccharomyces (S.) cerevisiae is the most important specie existing during the fermentation process (Blanco et al., 2008). It has been used in wine fermentation due to its ability to induce reliable and rapid fermentation, convenience and ease of control and consistency of fermentations (Heard and Fleet, 1985; Fleet, 2003). Inoculation of grape juice with S. cerevisiae is also common practice in winemaking. The recommended dosage at the beginning of fermentation is approximately 10 6 10 7 cells/ml (Degre, 1993; Boulton et al., 1996; Erten, 2002). Lately, a trend of using non-saccharomyces yeasts is emerging in winemaking to take advantage of their positive role in imparting the organoleptic characteristics back to

782 wine (Fleet, 2003; Viana et al., 2011). But, previous studies have shown that small-scale fermentations carried out with single strains of Kloeckera (K.) apiculata, Candida (C.) stellata, C. pulcherrima and C. colliculosa were not able to complete the fermentation. High residual sugar levels remained at the end of fermentation, and these wines significantly differed from those produced by an industrial wine yeast strain (Jolly et al., 2003). This has led to the current trend toward using so-called mixed starter cultures, containing one or more non-saccharomyces yeasts as well as a selected industrial S. cerevisiae wine yeast (Styger et al., 2011; Viana et al., 2011). This combined use of different species has been evaluated to be able to impart better organoleptic characteristics of wine than single fermentation by pure Saccharomyces or non-saccharomyces yeasts in several studies (Soden et al., 2000; Toro and Vazquez, 2002; Ciani et al., 2006; Moreira et al., 2008; Viana et al., 2011; Clemente-Jimenez et al., 2005; Trinh et al., 2011). Some authors have reported the effect of temperature on yeast growth and the production of volatile compounds by using S. cerevisiae (Torija et al., 2003; Molina et al., 2007; Beltran et al., 2008; Reddy and Reddy, 2011), but there are limited studies on the effect of temperature on wine production by using non-saccharomyces yeasts such as K. apiculata (Erten, 2002), K. corcitis, Hanseniaspora osmophila (Granchi et al., 2002), Hanseniaspora uvarum, Torulaspora delbrueckii, Kluyveromyces thermotolerans (Ciani et al., 2006). However, more information is still necessary to understand the behaviour of non-saccharomyces yeasts during wine fermentation. Williopsis (W.) saturnus formerly known as Hansenula saturnus synthesizes important levels of volatile esters, especially isoamyl acetate and ethyl acetate (Yilmaztekin et al., 2009; Erten and Tanguler, 2010; Trinh et al., 2011). Generally, it is not found in the natural environment grape surfaces and winery equipments, but with the production of desirable flavour, W. saturnus can potentially enhance the fruity flavour in wines obtained from neutral cultivar characteristics (Erten and Tanguler, 2010). Although there are various studies on the effects of temperature of S. cerevisiae on wine and beer fermentation, there is no study on the effect of temperature by using W. saturnus on wine fermentation. The present work describes the influence of temperatures from 12 to 24 at 6 intervals on yeast growth, wine composition and volatile compounds in the mixed cultures of W. saturnus var. saturnus and S. cerevisiae at two different inoculation ratios (W. saturnus var. saturnus : S. cerevisiae = 1:1 and 10:1 cfu ml 1 ). Meanwhile, in this research, Emir which is a native grape variety of Vitis vinifera L. was used due to its potential to produce the best white wines of Turkey (Erten et al., 2006) and W. saturnus H. Tanguler var. saturnus was chosen because of its ability in producing high quantity of esters such as isoamyl acetate (Yilmaztekin et al., 2009; Lee et al., 2010). Materials and Method Yeast cultures Commercial wine yeast S. cerevisiae (Actiflore PM) was obtained from Laffort Company (Bordeaux, France). W. saturnus var. saturnus HUT 7087 was obtained from HUT Culture Collection (Higashi-Hiroshima, Japan). Yeasts were maintained on Malt Extract Agar (Merck 105398.0500), MEA, slants and re-cultured monthly. Fermentation conditions White wine grapes of cv. Emir were obtained from a vineyard in the Nevsehir-Urgup province of Cappadocia region. They were transported to the pilot winery of the Department of Food Engineering, University of Cukurova, Adana. They were de-stemmed and crushed and the must was mixed with 50 mg/kg of sulphur dioxide, and kept at 15 for 12 h. Then the must was pressed in a horizontal press and grape juice was sterilised by autoclaving at 115 for 10 min. On the other hand, all equipments in contact with the grape juices were autoclaved at 121 for 15 min. Emir must had a 5.26 g/l of titratable acidity (TA) (as tartaric acid), 3.32 of ph, and 18,80 of initial brix. All fermentations were conducted in duplicate in 1 L sterile erlenmeyer flasks containing 800 ml of sterile grape juice. The flasks fitted with foam bungs were incubated at three different temperatures, 12, 18 and 24 (± 0.5 ). Fermentations were monitored daily by measuring the density. Upon completion of alcoholic fermentation, samples were stored at 4 for two days to sedimentation before general wine analysis. For GC and HPLC analysis, the samples were cleared by centrifugation at 8000 rpm, 0, 5 min, and then stored at 18 until analysed. Yeast propagation and enumeration of yeasts S. cerevisiae was suspended in sterile warm water at 35 for 30 min according to the producer s instructions. The yeast cells were centrifuged at 4000 rev/min for 10 min at 4, and washed with cold sterile water. On the other side, a loopful of stock cultures of W. saturnus was plated onto MEA and incubated for 48 h at 25. Yeast cultures (Pellet of S. cerevisiae and a single colony of W. saturnus) were propagated aerobically in 100 ml of sterile grape juice in a 250 ml sterile conical flask fitted with foam bung and incubated for 48 h at 25 with orbital shaking at 160 rev/min. The yeast cells were centrifuged at 4000 rev/min for 10 min at 4, and washed with cold sterile water. Pellets were re-suspended in 5 ml of sterile grape juice. After counting by haemacytometer, W. saturnus var. saturnus and S. cerevisiae yeasts were added into all fermentation medium as mono- and mixed culture (Erten and Tanguler, 2010). Mixed culture fermentations

Effect of Temperature and W. saturnus on Wine were conducted with the addition of 5 10 6 cells/ml of S. cerevisiae + 5 106 cells/ml of W. saturnus var. saturnus (1:1) and 5 10 6 cells/ml of S. cerevisiae + 5 10 7 cells/ml of W. saturnus var. saturnus (1:10) for each temperature (12, 18 and 24 ). Fermentation trials from 12 to 24 were designated from W1 to W6, respectively. On the other hand, the addition of pure culture of 5 10 6 cells/ml S. cerevisiae (Sc) was used as a control. Samples were taken aseptically to count the yeasts during alcoholic fermentations every day. They were diluted in 0.25% saline as necessary, and 0.1 ml of diluted sample was spread onto plates of MEA and Lysine agar (Sigma-Aldrich L5910). Lysine agar was used to count non-saccharomyces yeasts because it is a synthetic medium with glucose, vitamins, inorganic salts and L-lysine as the sole nitrogen source, and Saccharomyces spp. are unable to grow on it. Total yeast and W. saturnus var. saturnus yeast were enumerated on MEA and Lysine agar, respectively. The plates were incubated at 25 for 3 to 5 days and then examined for the yeast counts. The total Saccharomyces spp. count was calculated from the total yeast and W. saturnus var. saturnus yeast counts (Fleet and Heard, 1993; Erten et al., 2006). Analytic determinations Density and ph were determined using a density meter (Densito 30PX Mettler Toledo Portable Lab TM ), and a ph meter (Inolab WTW, Germany), respectively. TA was measured by titrating sample and expressed as grams of tartaric acid/l (OIV, 1990). Determination of organic acid, sugar, ethanol content and volatile compounds Ethanol, glycerol, glucose, fructose and acetic, succinic, tartaric, malic and citric acids were analysed by a HPLC (Shimadzu LC-20AD, Kyoto, Japan) using an Aminex HPX-87H column (Bio-Rad, Richmont, CA, 300 7.8 mm, USA) at 50. The eluent was 5 mmol/l H 2 SO 4 in high-purity water at a flow rate of 0,6 ml/min. Concentrations of ethanol, glycerol, glucose and fructose were calculated from RI detector and amounts of organic acids from UV detector (Shimadzu) (Erten and Tanguler, 2010). Standards (Merck, Darmstadt, Germany) were used to determine the concentration of organic acids, sugars, glycerol and ethanol. Higher alcohols, esters and carbonyl compounds were measured by a Gas Chromatograph (HP 5890; Hewlett- Packard, Stockport, UK). Cell-free samples were diluted to 4% (v v) ethanol and then 5 ml of diluted sample, 2 g of NaCl and 50 µl of internal standard (200 mg/l of 3-heptanone) were sealed and a 1 ml sample was injected into a 60 m 0,25 mm i.d. 0,4-µm-thick Chrompac CP-Wax-57-CB column (Middleburgh, the Netherlands), temperature-programmed from 40 to 80. The stream from the column was split 1:1 to flame ionization detector (Erten et al., 2006). All 783 analytical determinations were carried out in duplicate. The samples were stored at 18 until GC and HPLC analyses. Statistical analysis Data of chemical composition and volatile compunds were analysed for statistical significance by one-way analysis of variance (ANOVA). Means were compared by Duncan test statistical analysis using the software SPSS 10.0 for Windows (SPSS Inc. Chicago, IL, USA). These programmes were run on a PC. Results and Discussion This is the first study in which fermentation of grape must by adding W. saturnus var. saturnus together with S. cerevisiae at various temperatures was examined. Fermentations were carried out at 12, 18 and 24 using cv. Emir grape must, in order to estimate how temperature and inoculation ratios affect fermentation, yeast growth, wine composition and the production of volatile compounds. Effect of temperature and inoculation ratios on fermentation and growth of yeast In spite of the different temperatures and inoculation ratios used in experiments, all of the mixed and pure fermentation trials were finished (final sugar concentration < 2 g/l). Fermentation was followed by measuring the density. The results of density determinations during alcoholic fermentation are presented in Figure 1. Fermentation of Emir grape juice was completed on 8 days at 18 by using pure S. cerevisiae. The fermentation characteristic of the mixed culture (W3) at 18 was similar to that of the pure S. cerevisiae in term of density. As expected, fermentations performed at 12 (14 days) were longer than those performed at 18 (8 days) and 24 (6 days), respectively. It is well known that fermentation rate is mainly dependent on temperature. The rate increases with an increase in fermentation temperature (Fleet and Heard, 1993). Similar observations were also obtained with sugar consumption (data not shown). Several authors (Torija et al., 2003; Masneuf-Pomarède et al., 2006; Sener et al., 2007; Beltran et al., 2008; Cortes et al., 2009) have also stated that increasing temperature results in a faster fermentation rate. Furthermore, increasing inoculation ratio accelerated the fermentation but, fermentation time did not changed. Similarly, other studies have showed that increasing the inoculum ratio leads to faster fermentations (Fleet and Heard, 1993; Mateo et al., 2001; Erten et al., 2006). Temperature affected not only the fermentation time, but also the yeast growth. Changes in S. cerevisiae and W. saturnus var. saturnus viability in mixed cultures at 12, 18 and 24 during alcoholic fermentation is given in Figures 2, 3 and 4, respectively. Figure 3 also shows the changes in pure S. cerevisiae viability at 18. As can be seen from figures, the growth

784 H. Tanguler Density 1,08 1,07 1,06 1,05 1,04 1,03 1,02 1,01 1 0,99 Sc W1 W2 W3 W4 W5 W6 0 2 4 6 8 10 12 14 Fermentation time (days) Fig. 1. The decrease in density during fermentations. Sc (5 10 6 cells/ml pure S. cerevisiae, +), W1 (5 10 6 cells/ml S. cerevisiae + 5 10 6 cells/ml of W. saturnus var. saturnus, ), W2 (5 10 6 cells/ml S. cerevisiae + 5 10 7 cells/ml W. saturnus var. saturnus, ), W3 (5 10 6 cells/ml S. cerevisiae + 5 10 6 cells/ml of W. saturnus var. saturnus, ), W4 (5 10 6 cells/ml S. cerevisiae + 5 10 7 cells/ml W. saturnus var. saturnus, ), W5 (5 10 6 cells/ml S. cerevisiae + 5 10 6 cells/ml of W. saturnus var. saturnus, ), W6 (5 10 6 cells/ml S. cerevisiae + 5 10 7 cells/ml W. saturnus var. saturnus, ), The values are averages of duplicate determinations. 8 Numbers of viable yeast cells (Log cfu/ml) 7 6 5 4 3 2 1 0 0 2 4 6 8 10 12 14 Fermentation time (days) Fig. 2. Growth of S. cerevisiae and W. saturnus var. saturnus in mixed cultures at 12 during alcoholic fermentation. (S. cerevisiae in W1), (W. saturnus var. saturnus in W1), (S. cerevisiae in W2), (W. saturnus var. saturnus in W2). Vertical bars indicate standard deviations. in yeast varied according to temperature. A low fermentation temperature affects the growth of the yeast and leads to a slow growth rate. For this reason, fermentations at 12 began more slowly, as we can see by their longer lag phase (Figure 2). This caused a delay in reaching the maximal population. The lag phase is an important technological aspect in wine-making because it determines the adaptation of the yeast cells after their inoculation in the grape must. One common feature among low temperature fermentations is very long lag phases (Torija et al., 2003). In contrast, in present study, the lag phase was not observed at 18 and 24 (Figures 3 and 4). These results correlated with the previous reports that the lag phase was not seen and fermentation was completed more rapidly at higher temperature than 18 (Erten, 2002; Sener et al., 2007; Cortes et al., 2009). Figure 3 shows the growth of pure and mixed cultures at 18. The main wine yeast, S. cerevisiae, in pure culture used as control increased rapidly after fermentation started, and the highest amounts achieved 7.90 log cfu/ml within the first 4 days of the fermentation. After maximum growth, it exhibited a very long stationary phase until the end of fermentation and its population was found as 7.35 log cfu/

Effect of Temperature and W. saturnus on Wine 785 Numbers of viable yeast cells (Log cfu/ml) 8 7 6 5 4 3 2 1 0 0 1 2 3 4 5 6 7 8 9 Fermentation time (days) Fig. 3. Growth of S. cerevisiae and W. saturnus var. saturnus in pure and mixed cultures at 18 during alcoholic fermentation. + (pure culture of S. cerevisiae), (S. cerevisiae in W3), (W. saturnus var. saturnus in W3), (S. cerevisiae in W4), (W. saturnus var. saturnus in W4). Vertical bars indicate standard deviations. Numbers of viable yeast cells (Log cfu/ml) 9 8 7 6 5 4 3 2 1 0 0 1 2 3 4 5 6 7 Fermentation time (days) Fig. 4. Growth of S. cerevisiae and W. saturnus var. saturnus in mixed cultures at 24 during alcoholic fermentation. (S. cerevisiae in W5), (W. saturnus var. saturnus in W5), (S. cerevisiae in W6), (W. saturnus var. saturnus in W6). Vertical bars indicate standard deviations. ml. Furthermore, S. cerevisiae in the grape juice fermented with mixed cultures also increased rapidly after fermentation started at all temperatures studied, and dominated the alcoholic fermentation at two inoculation ratios (S. cerevisiae : W. saturnus var. saturnus, Sc:Ws= 1:1 and 1:10) by the 5th and 6th days at 12, by the 1st and 4th days at 18 and by the 1st and 3rd days at 24, respectively. It can obviously be seen that S. cerevisiae dominated the fermentation earlier with the increasing temperature. After proliferation and domination the fermentation of S. cerevisiae, the counts of W. saturnus var. saturnus yeasts declined or disappeared (Figures 2, 3 and 4). On the other hand, at different temperatures, the maxi- mum S. cerevisiae cell population was obtained in different days. It was obtained within 3 (inoculation ratio of 1:1) and 3 days at 24 (inoculation ratio of 1:10) compared to 4 and 3 days at 18, and 7 and 7 days at 12, respectively. After maximum growth at different temperatures and inoculation ratios, S. cerevisiae exhibited a very long stationary phase until the end of alcoholic fermentation. At the end, S. cerevisiae values were obtained between 6.92 and 7.53 log cfu/ ml (Figures 2, 3 and 4). In our previous study on different species of W. saturnus yeast, we determined that S. cerevisiae was the dominant yeast in mixed culture fermentations, exhibited a stationary phase and had the numbers of 7 7.64 log cfu/ml at 18 (Erten and Tanguler, 2010). It is observed

786 that the starting of fermentation, dominating to the alcoholic fermentation, reaching of maximal S. cerevisiae population and decline of population at 12 are slow compared to fermentations at 18 and 24. The results obtained by this study confirmed the previous investigations on the S. cerevisiae growth in mixed cultures during alcoholic fermentation (Erten, 2002; Zohre and Erten, 2002; Reddy and Reddy, 2011). At 12, after fermentation started at both inoculation raios of 1:1 and 1:10, W. saturnus var. saturnus dominated over S. cerevisiae during the first 4 and 5 days, respectively. The maximum W. saturnus var. saturnus amounts were 7.20 and 8.06 log cfu/ml by day 4. However, the maximum amounts at 18 and 24 were obtained as 7.0 and 8.06 log cfu/ml at two inoculation ratios by day 2 and 3, and 6.74 and 7.84 log cfu/ml by day 1 and 2, respectively. After maximum growth, W. saturnus var. saturnus did not exhibited a stationary phase in all trials and, the decline phase occured during fermentation. In our previous study (Erten and Tanguler, 2010) on different species of W. saturnus yeast, we observed similar results. At the same time, Zohre and Erten (2002) and Erten et al. (2006) stated that after maximum growth, non-saccharomyces yeasts didn t show the stationary phase and a decline phase was observed. In contrast, (Trinh, 2011) described that W. saturnus var. mrakii at inoculation ratio of 1:1000 with S. cerevisiae var. bayanus showed stationary phase in mixed culture fermentation. Additionally, it was observed that as the inoculation ratios increased, the reduction in the number of W. saturnus var. saturnus slowed. Moreover, while W. saturnus var. saturnus disappeared from the 11th day in trial W1 (inoculation ratio, 1:1), it survived 14 day in trial W2 (inoculation ratio, 1:10). Similarly, at 24, it disappeared from the 5th day in trial W5, but it survived 6 day in trial W6. At 18, they died off on 6th and 8th day of fermentation at two inoculation ratios, respectively. In previous studies, the researchers reported that W. saturnus died off by 4 5 days at inoculation ratio of 1:1 (Erten and Tanguler, 2010), by 10 days at inoculation ratio of 1:1000 (Lee et al., 2012b) and by 14 days at inoculation rate of 1:100 (Trinh, 2011). In contrast, Lee et al. (2010) and Trinh (2011) stated that W. saturnus survived until the end of fermentation that was terminated by 21 days at inoculation ratio of 1:1000. In present study, W. saturnus var. saturnus multiplied at higher populations and survived longer at lower temperatures than higher temperatures. Similar conclusions were also reported by Erten (2002) and Sener et al. (2007) for non- Saccharomyces yeasts. It is generally accepted that non-saccharomyces yeasts grow and then begin to die off during the early stages of fermentations due to their inability to tolerate the increasing ethanol concentrations present in the must me- H. Tanguler dium. Contrary to general assumptions, recent studies have demonstrated that some non-saccharomyces yeasts such as Hanseniaspora guilliermondii, K. apiculata and C. pulcherrima are able to stand much higher ethanol concentrations than previously thought (Bilbao et al., 1997; Zohre and Erten, 2002; Perez-Nevado et al., 2006). Explanations for the early death of non-saccharomyces yeasts in mixed cultures with Saccharomyces spp., could rely upon other factors such as competition for sugar uptake, oxygen availability, nutrient limitation, presence of toxic compounds, cell-cell interaction and quorum sensing (Fleet, 2003; Perez-Nevado et al., 2006; Lee et al., 2012b). Therefore, this study indicates that the fermentation temperature and different inoculation ratios are important in order to determine the inhibition of the non- Saccharomyces yeast. They exist not only in the early wine fermentation stage but also even for longer periods. Effect of temperature and inoculation ratios on general wine composition Temperature and inoculation ratio effected not only the fermentation rate and length but also the yeast metabolism, which determined the general composition of the wine (Torija et al., 2003; Sener et al., 2007; Reddy and Reddy, 2011). The general composition of wines in present study is illustrated in Table 1. As can be seen, all parameter values were dependent on the fermentation temperature. Increasing temperature from 12 to 24 resulted in a decrease in ethanol, malic and citric acid (except W6) concentrations, but in an increase in glycerol, total acidity, acetic and tartaric acid concentrations. The highest ethanol values were found in W1 (9.56 %v/v) produced at 12 and the lowest values were in W6 (8.01 %v/v) produced at 24. However, it determined as 9.25%v/v in control produced at 18. Temperature and inoculation ratio significantly affected ethanol concentration (P < 0.001). Glycerol is quantitatively a very important wine constituent. During yeast fermentation, it is the major end product other than ethanol and carbon dioxide. Its concentrations in wine varying between 1.0 27.6 g/liter (Ciani and Ferraro, 1996; Cortes et al., 2009; Styger et al., 2011) and the maximum level acceptable in wines is 25.8 g/l (Toro and Vazquez, 2002; Mendoza and Farías, 2010). The amount of glycerol formed is influenced by several factors, such as grape variety, degree of ripeness, fermentation temperature and yeast strain (Ciani and Ferraro, 1996; Suarez-Lepe and Morata, 2012). In present study, glycerol were changed between 4.42 5.99 g/l (P < 0.001) and it was determined as 5.42 g/l in control. Wines inoculated with equal ratios of S. cerevisiae and W. saturnus var. saturnus have higher amounts of glycerol when compared with pure S. cerevisiae fermented wine used as a control. These results for glycerol were found in acceptable levels and correlated with the previous reports

Effect of Temperature and W. saturnus on Wine 787 Table 1. The general composition of wines. 18 12 18 24 Temperature Sc W1 W2 W3 W4 W5 W6 P Density (20 ) 0.9944 ab 0.99445 ab 0.9937 ab 0.9938 ab 0.9924 c 0.9948 a 0.9934 bc * Ethanol (v v) 9.25 ab ± 0.08 9.56 a ± 0.18 8.63 cd ± 0.17 9.42 a ± 0.01 8.2 de ± 0.16 8.94 bc ± 0.12 8.01 e ± 0.12 *** Glycerol (g/l) 5.42 b ± 0.07 5.45 b ± 0.07 4.42 d ± 0.1 5.56 b ± 0.06 4.46 cd ± 0.06 5.99 a ± 0.13 4.76 c ± 0.16 *** ph 3.165 bc 3.17 abc 3.21 a 3.175 abc 3.195 ab 3.145 c 3.2 ab * Total acidity as tartaric acid (g/l) 6.045 ab ± 0.04 5.93 abc ± 0.2 5.54 c ± 0.12 5.99 ab ± 0.03 5.66 bc ± 0.15 6.16 a ± 0.06 5.66 bc ± 0.1 * Organic acids (g/l) Acetic acid 0.40 c ± 0.02 0.53 bc ± 0.04 0.70 ab ± 0.05 0.70 ab ± 0.02 0.8 a ± 0.01 0.79 a ± 0.04 0.86 a ± 0.12 ** Tartaric acid 2.07 ± 0.02 1.785 ± 0.1 1.655 ± 0.07 2.04 ± 0.03 2.105 ± 0.2 2.09 ± 0.07 2.18 ± 0.12 ns Malic acid 1.72 a ± 0.03 1.73 a ± 0.03 1.48 bc ± 0.04 1.585 ab ± 0.06 1.47 bc ± 0.08 1.275 cd ± 0.09 1.16 d ± 0.1 ** Succinic acid 0.67 c ± 0.02 0.73 bc ± 0.06 1.02 a ± 0.06 0.7 bc ± 0.05 0.89 ab ± 0.07 0.7 bc ± 0.08 1.06 a ± 0.07 ** Citric acid (mg/l) 35 abc ± 4 43.18 a ± 1.9 28.53 bc ± 1.4 38 ab ± 4 25.5 c ± 3.5 30.62 bc ± 0.7 31.06 bc ± 1.1 * Sugars (g/l) Glucose 1.1 abc ± 0.02 1 abc ± 0.18 0.675 d ± 0.07 1.12 ab ± 0.02 0.79 cd ± 0.07 1.14 a ± 0.06 0.81 bcd ± 0.1 * Fructose 0.78 a ± 0.03 0.57 abc ± 0.03 0.385 c ± 0.08 0.705 ab ± 0.02 0.49 bc ± 0.04 0.67 ab ± 0.09 0.48 bc ± 0.11 * Total sugar 1.88 a ± 0.05 1.57 ab ± 0.21 1.06 b ± 0.14 1.825 a ± 0.04 1.28 b ± 0.11 1.81 a ± 0.15 1.29 b ± 0.22 * Sc (5 10 6 cells/ml pure S. cerevisiae), W1 (5 10 6 cells/ml S. cerevisiae + 5 10 6 cells/ml of W. saturnus var. saturnus), W2 (5 10 6 cells/ml S. cerevisiae + 5 10 7 cells/ml W. saturnus var. saturnus), W3 (5 10 6 cells/ml S. cerevisiae + 5 10 6 cells/ml of W. saturnus var. saturnus), W4 (5 10 6 cells/ml S. cerevisiae + 5 10 7 cells/ml W. saturnus var. saturnus), W5 (5 10 6 cells/ml S. cerevisiae + 5 10 6 cells/ml of W. saturnus var. saturnus), W6 (5 10 6 cells/ml S. cerevisiae + 5 10 7 cells/ml W. saturnus var. saturnus), P : Significance. ***, ** and * display the significance at 0.1%, 1% and 5% by LSD, respectively. a-f Values not sharing the same superscript letter within the horizontal line are different according to Duncan test. ns: not significant. (Soden et al., 2000; Toro and Vazquez, 2002; Comitini et al., 2011). On the other hand, Balli et al. (2003) stated that glycerol and ethanol concentrations were inversely related. Moreover, in present study it is observed that as ethanol concentration decreased, glycerol increased. Similarly, Torija et al. (2003), Balli et al. (2003), Molina et al. (2007), Cortes et al. (2009) and Reddy and Reddy (2011) also reported that ethanol concentration decreases, but glycerol increases as the temperature increases. In contrast, some researchers also stated that ethanol value increases (Sener et al., 2007; Erten, 2002), and/or glycerol decreases by increasing temperature (Bilbao et al., 1997). In addition, the inoculum ratio also influences the production of glycerol (Suarez-Lepe and Morata, 2012). In present study, it is determined that glycerol and ethanol decreased by increasing inoculum ratios of W. saturnus var. saturnus studied at all temperatures. On the contrary, Viana et al. (2011) reported that ethanol and glycerol levels were not affected by the type of inoculation. Fermentation temperature is also known to affect yeast metabolism, and as a result, the formation of other secondary metabolites (acetic acid, succinic acid, etc.) besides glycerol (Torija et al., 2003). Acetic acid, a by-product of yeast metabolism, leads to wine objectionable near its flavour threshold of 0.7 1.1 g/l of volatile acidiy as acetic acid (Henschke and Jiranek, 1993). Succinic acid is one of the major acids found in wine. It contributes a pleasant acidic taste (Patel and Shibamoto, 2003). Acetic acid and succinic acid values in wine produced by pure S. cerevisiae were 0.40 g/l and 0.67 g/l, respectively. Their value in mixed cultures were higher than pure S. cerevisiae and determined between 0.53 0.86 g/l and 0.70 1.06 g/l, respectively (P < 0.01). This high values are probably a result of the W. saturnus var. saturnus, which is produced high amounts acetic and succinic acid (Lee et al., 2010; Lee et al., 2012a). As acetic acid increased by temperature and inoculation, succinic acid decreased by temperature (except for W6), but increased by inoculation. Similarly, Torija et al. (2003) and Cortes et al. (2009) also stated that acetic acid concentration increased with increasing temperature. On the other hand, in present study, acetic acid concentrations obtained in some wines are in disagreement with the results found by Soden et al. (2000), Cortes et al. (2009) and Mendoza et al. (2011) who found the lower amounts of acetic acid (0.3 0.70 g/l) in mixed cultures. In this study, it was produced higher than the lower threshold value throughout the fermentations which leads to negative effect on wine quality (except for W1). Glucose, fructose, total sugar, total acidity and tartaric acid values in wines were increased with increasing temperature, except for fructose in W5 and W6, but decreased by increasing inoculation ratio (except for tartaric acid in W6). These results are in good accord with the data given by Cortes et al. (2009), by contrast

788 with Masneuf-Pomarède et al. (2006) and Ciani et al. (2006). The results also showed that fermentation temperature and inoculation ratio had a significant effect on glucose, fructose, sucrose, total sugar and total acidity values (P < 0.05), but did not affect tartaric acid concentration (P > 0.05). Effect of temperature and inoculation ratios on volatile compounds of wines There are many factors that may have an important impact on the formation of volatile compounds in wine, and thus on its quality, including yeast type, the population of individual yeasts in must, the composition of the must, the fermentation conditions and temperature, the winemaking technique followed, and the yeast strain used and inoculum ratio (Molina et al., 2007; Trinh et al., 2010; Suarez-Lepe and Morata, 2012). In this study, effect of temperature and also inoculation ratio on the volatile compositions of wine fermented by mixed cultures of yeasts were studied. Yeast is the main producer of flavour compounds such as esters, higher alcohols and carbonyl compounds during alcoholic fermentation (Sarris et al., 2009). For this reason, in this study, yeasts belonging to the Williopsis genus H. Tanguler was chosen. It is reported that this yeast is able to consume sugar oxidatively for cell growth as well as produce desirable fruity flavours. Particularly, W. saturnus strains are known to convert higher alcohols into the corresponding acetate esters. An example is the conversion of isoamyl alcohol into isoamyl acetate (Yilmaztekin et al., 2009). Isoamyl acetate (3-methylbutyl acetate) is important contributors to the pleasant fruity note of wine (Gil et al., 1996). In present study, mixed cultures of S. cerevisiae and W. saturnus var. saturnus formed relatively higher amounts of isoamyl acetate, between 3.18 5.12 mg/l, in comparison with control (2.49 mg/l) (Table 2). It could be said that W. saturnus var. saturnus was significantly able to enhance the production isoamyl acetate (banana and fruity like aroma). Zohre and Erten (2002) produced white wine with mixed cultures of Kloeckera apiculata, Candida pulcherrima and S. cerevisiae, and they found the concentrations of isoamyl acetate in the range of 0.995 1.905 mg/l. Findings in present work for isoamyl acetate are slightly higher than study reported by Zohre and Erten (2002), however slightly lower than previous study of Table 2. The effect of temperature on flavour compounds of wines. 18 12 18 24 Temperature Sc W1 W2 W3 W4 W5 W6 P ESTERS (mg/l) Ethyl acetate 40.97 b ± 2.86 56.35 b ± 5.09 103.76 a ± 5.88 49.08 b ± 1.17 93.09 a ± 5.03 48.18 b ± 3.86 99.76 a ± 7.6 *** Isobutyl acetate 0.115 ± 0.025 0.137 ± 0.004 0.143 ± 0.005 0.129 ± 0.013 0.112 ± 0.012 0.125 ± 0.006 0.112 ± 0.004 ns Ethyl butyrate 0.167 ± 0.015 0.203 ± 0.005 0.125 ± 0.004 0.193 ± 0.003 0.107 ± 0.007 0.114 ± 0.006 0.083 ± 0.004 ns Isoamyl acetate 2.49 c ± 0.253 3.98 ab ± 0.038 4.13 ab ± 0.026 3.90 ab ± 0.319 5.12 a ± 0.83 3.18 bc ± 0.041 3.31 bc ± 0.036 * Ethyl hexanoate 0.475 ± 0.103 0.488 ± 0.014 0.334 ± 0.006 0.427 ± 0.115 0.324 ± 0.102 0.29 ± 0.014 0.226 ± 0.01 ns Ethyl octanoate 0.26 ± 0.034 0.256 ± 0.008 0.21 ± 0.016 0.272 ± 0.069 0.206 ± 0.022 0.26 ± 0.008 0.178 ± 0.004 ns Total 44.477 b ± 2.43 61.41 b ± 5.03 108.70 a ± 5.86 54.001 b ± 1.02 98.959 a ± 4.3 52.15 b ± 3.8 104.0 a ± 8.01 *** HIGHER ALCOHOLS (mg/l) 2-Methyl butanol 23.79 ± 0.95 22.07 ± 1.03 20.91 ± 0.98 24.22 ± 2.14 25.79 ± 2.93 26.83 ± 0.95 28.73 ± 0.41 ns 3-Methyl butanol 116.16 a ± 2.56 95.79 b ± 0.9 68.09 d ± 0.98 123.22 a ± 3.1 84.01 c ± 3.08 115.32 a ± 1.62 76.71 c ± 2.61 *** Propan-1-ol 24.593 e ± 0.82 50.093 c ± 0.8 45.51 d ± 1.05 70.47 b ± 0.92 52.496 c ± 1.6 72.16 ab ± 0.94 74.4 a ± 1.16 *** Isobutanol (isobutyl alcohol) 29.55 e ± 0.91 22.30 f ± 0.07 32.79 d ± 0.57 31.11 de ± 0.15 39.74 b ± 0.89 35.58 c ± 0.96 49.55 a ± 0.91 *** Total 194.09 d ± 0.12 190.25 d ± 1 167.30 e ± 0.33 249.02 a ± 0.19 202.03 c ± 0.86 249.89 a ± 2.57 229.39 b ± 2.45 *** CARBONYL COMPOUNDS (mg/l) Acetaldehyde 20.78 c ± 1.73 21.47 c ± 1 30.24 b ± 1.45 21.58 c ± 2.06 31.28 b ± 1.12 30.42 b ± 1.56 56.47 a ± 0.87 *** Acetone (dimethyl ketone) 0.774 d ± 0.07 0.975 c ± 0.01 1.077 bc ± 0.02 1.113 bc ± 0.09 1.103 bc ± 0.1 1.573 a ± 0.02 1.264 b ± 0.01 *** Butanedione (diacetyl) 0.073 c ± 0.01 0.142 b ± 0.01 0.2 a ± 0.01 0.054 c ± 0.01 0.077 c ± 0.02 0.046 c ± 0.003 0.08 c ± 0.003 *** Pentanedione 0.026 bc ± 0.003 0.064 a ± 0.001 0.014 d ± 0.001 0.028 bc ± 0.008 0.035 b 0.015 d ± 0.001 0.021 cd ± 0.002 *** Total 21.653 c ± 1.67 22.65 c ± 0.99 31.531 b ± 1.41 22.775 c ± 1.97 32.495 b ± 1.23 32.054 b ± 1.54 57.835 a ± 0.86 *** MAIN TOTAL 260.22 d ± 4.21 274.31 d ± 5.02 307.531 c ± 7.6 325.796 bc ± 3.18 333.49 b ± 6.38 334.093 b ± 7.9 391.249 a ± 9.6 *** Sc (5 10 6 cells/ml pure S. cerevisiae), W1 (5 10 6 cells/ml S. cerevisiae + 5 10 6 cells/ml of W. saturnus var. saturnus), W2 (5 10 6 cells/ml S. cerevisiae + 5 10 7 cells/ml W. saturnus var. saturnus), W3 (5 10 6 cells/ml S. cerevisiae + 5 10 6 cells/ml of W. saturnus var. saturnus), W4 (5 10 6 cells/ml S. cerevisiae + 5 10 7 cells/ml W. saturnus var. saturnus), W5 (5 10 6 cells/ml S. cerevisiae + 5 10 6 cells/ml of W. saturnus var. saturnus), W6 (5 10 6 cells/ml S. cerevisiae + 5 10 7 cells/ml W. saturnus var. saturnus), P : Significance. *** and * display the significance at 0.1% and 5% by LSD, respectively. a-f Values not sharing the same superscript letter within the horizontal line are different according to Duncan test. ns: not significant.

Effect of Temperature and W. saturnus on Wine Viana et al. (2011) who found between 4.09 and 5.83 mg/l in wine fermented with Hanseniaspora vineae and S. cerevisiae. But, they used red grape must with an initial sugar content of 257 g/l and supplemented with 1 g/l of complex yeast nutrient in their study. Additionally, the production of isoamyl acetate dependent of fermentation temperature and inoculation rate (P < 0.05). Increasing temperature led to a marked decrease in the concentration of isoamyl acetate (except for W4), but its concentration increased when the inoculation ratio was raised. Simpson (1979) and Etievant (1991) stated that isoamyl acetate has 1 mg/l flavour threshold and it is the most abundant contributor to wine aroma. The results obtained by this study confirmed the previous investigations on the effects of temperature (Beltran et al., 2008), inoculation ratios (Erten et al., 2006; Trinh et al., 2011) and W. saturnus species (Erten and Tanguler, 2010). On the contrary, Erten (2002) stated that the production of isoamyl acetate was independent of fermentation temperature. Moreover, Lee et al. (2010) reported that the production of isoamyl acetate with pure S. cerevisiae is higher than produced with mixed cultures (the ratio of 1:100). Ethyl acetate is the most abundant ester existing in wine and its existence gives the positive effect on the fruity flavour of wines, because the levels are higher than flavour threshold (Reddy et al., 2008). Non-Saccharomyces yeasts such as Candida, Hansenula, Pichia and Williopsis species have a greater capacity to produce ethyl acetate than wine strains of S. cerevisiae (Lee et al., 2010; Manzanares et al., 2011). Similar to isoamyl acetate, it was observed that the lowest ethyl acetate (40.97 mg/l) and total ester (44.48 mg/l) were also determined in wine used as control (Table 2). Moreover, their concentrations in wines inoculated with equal ratios of S. cerevisiae and W. saturnus var. saturnus decreased when fermentation temperature increased to 24 with the exception for W6, but increased with increasing inoculation ratio. This is probably a result of longer growth and survival of W. saturnus var. saturnus at lower temperatures and and higher amount of W. saturnus var. saturnus than equal ratio. The concentrations of ethyl acetate and total esters in mixed cultures were ranging from 48.18 to 103.76 mg/l and from 52.15 to 108.70 mg/l, respectively. In addition, fermentation temperature and inoculation ratio were found significantly important with regard to the amount of ethyl acetate and total esters (P < 0.001). Etiévant (1991) and Lee et al. (2010) state that concentrations of ethyl acetate below 50 mg/l do not contribute to wine flavour, while amounts higher than 150 200 mg/l results in defects with solvent-like flavour in wine quality. In this study, the amounts of ethyl acetate were generally found between 50 150 mg/l and contribute positively to the wine flavour. The results obtained for ethyl 789 acetate in this study are in agreement with the findings of Kourkoutas et al. (2001), Erten (2002), Reddy et al. (2008) and Reddy and Reddy (2011) that there was an increase in ethyl acetate formation in wines at lower temperatures. Moreover, Trinh et al. (2011) have reported similar result that the concentration of ethyl acetate rose at different inoculation rates of W. saturnus. Erten and Tanguler (2010) reported that ethyl acetate concentrations were below 50 mg/ L in wines produced with equal ratios of S. cerevisiae and W. saturnus. However, some authors reported that ethyl acetate level decreased with decreasing temperature (Ikonomopoulou et al., 2003; Mallouchos et al., 2003). Isobutyl acetate (2-methylpropyl acetate), ethyl butyrate (ethyl butanoate) and ethyl hexanoate (ethyl caproate) are synthesized greater at low fermentation temperatures (Table 2). Their values were found as 0.11 0.14 mg/l, 0.08 0.20 mg/l and 0.23 0.49 mg/l, respectively. These results are consistent with previous reports (Erten, 2002; Molina et al., 2007; Beltran et al., 2008), but are not in agreement with data reported by Mallouchos et al. (2003). On the other hand, fermentation temperature and inoculation ratio did not affect isobutyl acetate, ethyl butyrate, ethyl hexanoate and ethyl octanoate (P > 0.05). Higher alcohols are quantitatively the largest group of volatile compounds. They are mainly formed from Ehrlich pathway in the presence of amino acids, and/or from sugars via biosynthesis by yeasts when amino acids are absent in the medium. Higher alcohols, characterized by their strong and pungent smell and taste, can significantly influence wine taste and character (Erten, 2002; Manzanares et al., 2011; Suarez-Lepe and Morata, 2012). The most important ones are 2-methyl butanol (active amyl alcohol), 3-methyl butanol (iso amyl alcohol) and propan-1-ol (n-propyl alcohol) (Kandylis et al., 2010; Suarez-Lepe and Morata, 2012). Wines produced at 18 fermented with W. saturnus var. saturnus at inoculation ratios of 1:1 and 1:10 have higher amounts of relative and total higher alcohol when compared with pure S. cerevisiae fermented wine used as a control (except for 3-methyl butanol in W4). However, it was observed that 2-methyl butanol (except W2) and isobutanol amounts increased, propan-1-ol (except W6), 3-methyl butanol and total higher alcohol decreased with the inoculation. These results correlated with the previous report given by Erten and Tanguler (2010) who studied with W. saturnus at equal inoculation ratio (1:1). In contrast, Trinh et al. (2011) studied also W. saturnus and they showed that 3-methyl butanol increased with two inoculation ratios of 1:100 and 1:1000. On the other hand, Lee et al. (2012b) reported that all relative higher alcohols increased at higher W. saturnus inoculation ratio (1:1000). However, Lee et al. (2010) stated that relative

790 higher alcohols decreased at higher W. saturnus inoculation ratio. On the other hand, Increasing at temperature led to an increase in concentration of relative (except for 3-methyl butanol) and total higher alcohol. 3-methyl butanol amount was increased by increasing temperature from 12 to 18 and the highest concentration was determined as 123.22 mg/l in W3, but later on decreased by increasing temperature from 18 to 24. In addition, relative and total higher alcohol values formed were found statistically important (P < 0.001), but 2-methyl butanol concentrations were trivial. Several authors (Bardi et al., 1997; Erten, 2002; Reddy and Reddy, 2011) have also stated that higher alcohols increased by increasing temperaure. The concentrations of 2-methyl butanol (20.91 28.73 mg/l), propan-1-ol (24.59 74.4 mg/l) and isobutanol (22.3 49.55 mg/l) were much below than their threshold values (300 750 mg/l, 75 500 mg/l and 300 330 mg/l, respectively) given by Etiévant (1991) but within the range of previous studies reported previously. Moreover, 3-methyl butanol concentrations in wines produced by mixed cultures were determined between 68.09 123.22 mg/l and their values were lower than control sample except for W3. Their concentrations in all wines were higher than its threshold value (14.5 mg/l) given in the literature (Etiévant, 1991). Higher alcohols generally exist in wine at the concentrations < 300 mg/l and they contribute to the aromatic complexity of the product. When their concentrations exceed 400 mg/ L, they are considered to have a negative effect on flavour (Manzanares et al., 2011). In present study, higher alcohol concentrations were determined below the 300 mg/l. For this reason, it could be said that higher alcohol had positive effect on flavour of wines produced at different temperatures with W. saturnus var. saturnus. Due to their low perception threshold and the characteristics that they confer on the wine, volatile aldehydes are among the most interesting carbonyl compounds. Acetaldehyde is the principal carbonyl compound in wine derived from pyruvate during alcoholic fermentation (Berry, 1995) and it constitutes more than 90% of the total aldehyde content of wines (Manzanares et al., 2011). In addition, it plays an important role in the flavour and bouquet of wine but not always desirable (Lee et al., 2010). Because, at higher concentrations this turns into a pungent irritating odor reminiscent of green grass or apples (Styger et al., 2011). In present study, the lowest concentration of acetaldehyde were found in control sample as 20.78 mg/l. However, in mixed cultures, its concentration increased with increasing temperature and inoculation ratio. Smogrovicova and Domeny (1999) reported that acetaldehyde concentration increased as the temperature increased in beers. Controversial results H. Tanguler were reported with the effect of temperature on acetaldehyde production in wines by Erten (2002), Torija et al. (2003) and Reddy and Reddy (2011). On the other hand, in this study, its value ranged from 21.47 to 56.47 mg/l in wines produced with mixed cultures and it was found that fermentation temperature and inoculation ratio significantly affected to acetaldehyde concentration (P < 0.001). The acetaldehyde concentration in wines usually ranges from 13 to 40 mg/l, but may reach 75 mg/l (Reddy et al., 2008; Kourkoutas et al., 2001). The results obtained in this study are in agreement with the results given by Kourkoutas et al. (2001) and Reddy et al. (2008). The lowest acetone value was determined in control sample as 0.77 mg/l. Acetone concentration were increased by increasing temperature and inoculum ratio in mixed cultures and its value ranged from 0.98 to 1.57 mg/l, whereas, butanedione and pentanedione amounts were decreased (except W6 for butanedione and except W1 for pentanedione). The effect of temperature and inoculation ratio on the amounts of these compounds were found important (P < 0.001). Conclusions This is the first study that effect of fermentation temperature at 12, 18 and 24 on wine fermentation by adding W. saturnus var. saturnus together with S. cerevisiae was evaluated based on fermentation rate, duration, yeast growth, wine composition and volatile compounds. Fermentations performed at lower temperature were found longer; and lag phase which is a desired trait of wine yeast was seen only at 12. The increase in temperature from 12 to 24 resulted in a decrease in ethanol, malic and citric acid concentrations, but an increase in glycerol, total acidity, acetic and tartaric acid concentrations at both inoculation ratios. Fermentation temperature affected also to volatile compounds. Higher amounts of esters, especially isoamyl acetate, ethyl acetate, isobutyl acetate, ethyl butyrate and ethyl hexanoate, and lower amounts of 2-methyl butanol, propan-1-ol, isobutanol, acetaldehyde and acetone were formed at low temperature. The results given in this study show the importance of temperature and inoculation ratio. Fermentations at 12 at two inoculation ratios appear to be more suitable options than fermentation conducted 18 and 24. On the other hand, it could be said that W. saturnus var. saturnus can be used in mixed starter cultures with S. cerevisiae. Regarding on wine production with W. saturnus var. saturnus, research is lacking with respect to: (1) application of multistarter fermentations with the other non-saccharomyces yeasts; (2) sequential inoculation of W. saturnus; (3) determination of bioactive compounds; (4) comparison of the glucophylic or/and

Effect of Temperature and W. saturnus on Wine fructophylic nature of W. saturnus; (5) determination of the effect of oxygen, lipid, and complex yeast nutrient supplementation on the volatile composition and (6) evaluation of organoleptic quality of wines. Therefore, more research is needed for concerning these subjects. Acknowledgements The author is grateful to Prof. Huseyin ERTEN and Niyazi CETINKAYA for critical reading and Dr. Adnan BOZDOGAN for his assistance in statistical analysis. References Andorrà, I., Landi, S., Mas, A., Esteve-Zarzoso, B., and Guillamón, J.M. (2010). Effect of fermentation temperature on microbial population evolution using culture-i ndependent and dependent techniques. Food. Res. Int., 43, 773-779. Balli, D., Flari, V., Sakellaraki, E., Schoina, V., Iconomopoulou, M., Bekatorou, A. and Kanellaki, M. (2003). Effect of yeast cell i mmobilization and temperature on glycerol content in alcoholic w respect to wine making. Proc. Biochem., 39, 499-506. Bardi, E., Koutinas, A.A., Psarianos, C. and Kanellaki, M. (1997). Volatile by-products formed in low-temperature wine-making using immobilized yeast cells. Proc. Biochem., 32, 579-584. Beltran, G., Novo, M., Guillamón, J.M., Mas, A. and Rozès, N. (2008). Effect of fermentation temperature and culture media on the yeast lipid composition and wine volatile compounds. Int. J. Food Microbiol., 121, 169-177. Berry, D.R. (1995). Alcoholic beverage fermentation, In Fermented Beverage Production eds. by Lea, A.G.H., Piggott, J.R. London: Blackie Academic, pp. 32-61. Bilbao, A., Irastorza, A., Duenas, M. and Fernandez, K. (1997). The effect of temperature on the growth of strains of Kloeckera apiculata and Saccharomyces cerevisiae in apple juice fermentation. Lett. Appl. Microbiol., 24, 37-39. Blanco, P., Vazquez-Alen, M. and Losada, A. (2008). Influence of yeast population on characteristics of the wine obtained in spontaneous and inoculated fermentations of must from Vitis vinifera Lado. J. Ind. Microbiol. Biotechnol., 35, 183-188. Boulton, R.B., Singleton, V.L., Bisson, L.F. and Kunkee, R.E. (1996). Principles and Practices of Winemaking. New York: Chapman & Hall. p. 604. Ciani, M., Beco, L. and Comitini, F. (2006). Fermentation behaviour and metabolic interactions of multistarter wine yeast fermentations. Int. J. Food Microbiol., 108, 239-245. Ciani, M. and Ferraro, L. (1996). Enhanced glycerol content in wines made with immobilized Candida stellata cells. Appl. Environ. Microbiol., 62, 128-132. Clemente-Jimenez, J.M., Mingorance-Cazorla, L., Martinez-Rodriguez, S., Heras-Vazquez, F.J.L. and Rodriguez-Vico, F. (2005). Influence of sequential yeast mixtures on wine fermentation. Int. J. Food Microbiol., 98, 301-308. 791 Comitini, F., Gobbi, M., Domizio, P., Romani, C., Lencioni, L., Mannazzu, I. and Ciani, M. (2011). Selected non-saccharomyces wine yeasts in controlled multistarter fermentations with Saccharomyces cerevisiae. Food Microbiol., 28, 873-882. Cortes, S., Salgado, J.M., Rivas, B., Torrado, A.M. and Dominguez, J.M. (2010). Fermentation kinetics and chemical characterisation of Vino tostado, A traditional sweet wine from galicia (NW Spain). J. Sci. Food Agric., 90, 121-131. Degre, R. (1993). Selection and commercial cultivation of wine yeast and bacteria, In Wine Microbiology and Biotechnology, Fleet, G.H. ed. by Chur, Switzerland: Harwood Academic Publishers, pp. 421-448. Erten, H. (2002). Relations between elevated temperatures and fermentation behaviour of Kloeckera apiculata and Saccharomyces cerevisiae associated with winemaking in mixed cultures. World J. Microbiol. Biotechnol., 18, 373-378. Erten, H., Tanguler, H., Cabaroglu, T. and Canbas, A. (2006). The influence of inoculum level on fermentation and flavour compounds of white wines made from cv. Emir. J. Inst. Brew., 112, 232-236. Erten, H. and Tanguler, H. (2010). Influence of Williopsis saturnus yeasts in combination with Saccharomyces cerevisiae on wine fermentation. Lett. Appl. Microbiol., 50, 474-479. Etiévant, P.X. (1991). Wine, In Volatile compounds in Food and Beverages, ed. by Maarse, H., New York: Marcel Dekker, pp. 483-546. Fleet, G.H. (2003). Yeast interactions and wine flavour. Int. J. Food Microbiol., 86, 11-22. Fleet, G.M. and Heard, G.M. (1993). Yeasts-growth during fermentation, In Wine Microbiology & Biotechnology, ed. by Fleet G.M. Chur: Harword Academic Publishers, pp. 27-54. Gil, J.V., Mateo, J., Jimenez, M. and Pastor Huerta, T. (1996). Aroma compounds in wine as influenced by apiculate yeasts. J. Food Sci., 61, 1247-1266. Granchi, L., Ganucci, D., Messini, A., Rosellini, D. and Vincenzini, M. (2002). Oenological properties of Hanseniaspora osmophila and Kloeckera corticis from wines produced by spontaneous fermentations of normal and dried grapes. FEMS Yeast Res., 2, 403 407. Heard, G.M. and Fleet, G.H. (1985). Growth of natural yeast flora during the fermentation of i noculated wines. Appl. Environ. Microbiol., 50, 727-728. Henschke, P.A. and Jiranek, V. (1993). Yeasts-metabolism of nitrogen compounds, In Wine Microbiology and Biotechnology ed. by Fleet, G.M., Chur: Harword Academic Publishers, pp. 77-164. Ikonomopoulou, M., Kanellaki, M., Soupioni, M. and Koutinas, A.A. (2003). Effect of freeze-dried immobilized cells on delignified cellulosic material in low-temperature and ambient-temperature wine making. Appl. Biochem. Biotechnol., 104, 23-36. Jolly, N.P., Augustyn, O.P.H. and Pretorius, I.S. (2003). The effect