Exploring the biodiversity of a wine region: Saccharomyces yeasts associated with wineries and vineyards

Exploring the biodiversity of a wine region: Saccharomyces yeasts associated with wineries and vineyards L. Mercado 1 and M. Combina 1,2 1 EEAMendoza, Instituto Nacional de Tecnología Agropecuaria (INTA), 2 Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET). San Martín 3853 (5507), Luján de Cuyo, Mendoza, Argentina Saccharomyces yeasts have been used in the fermentation of food and drink products for thousands of years. Wine production is a complex microbiological process involving different S. cerevisiae populations during fermentation. Argentina is the fifth largest wine producer in the world and most of its grape and wine production is located in the west side of the country on the eastern side of the Andes Mountains (between 22º and 42º S). In characterizing a wine region, it is necessary to not only define its microclimate and soil characteristics, but also the yeast populations that are present in the vineyards, grapes and wineries. The objective of the present study was to determine the native strains of Saccharomyces cerevisiae present in the grapes, wineries and wines of the Zona Alta del Río Mendoza (ZARM) region. This region is known worldwide for producing high quality Malbec wines. Two different, but related, aspects were evaluated using a molecular approach for Saccharomyces strain differentiation. First, we tried to elucidate the origin of S. cerevisiae yeasts involved in the spontaneous fermentation of grape musts by evaluating the distribution of these yeasts on winery equipment and determining their contribution to the fermentation process. Their genetic relationships with commercial yeasts were also evaluated. Secondly, we explored S. cerevisiae strain diversity in the ZARM vineyards. The contribution of vineyard Saccharomyces strains to the population responsible for industrial spontaneous fermentations and the level of genetic relationships between these populations were also evaluated. The results showed S. cerevisiae biota resident in both wineries and vineyards. A wide diversity and dynamic behaviour were also found within and between seasons, as was a variable contribution to the fermentation process as well as complex interactions with the commercial yeasts used in the wineries. There was no evidence of a representative strain distributed throughout the viticultural region evaluated. Complex genetic relationships at the molecular level between isolated yeasts which shared the same ecological environment were found. Although a wide diversity was observed, these yeasts shared many characteristics, as evidenced by molecular markers, which suggests that the strong selection pressure exerted by the fermentation process could have generated variability at different levels. Knowledge about the biodiversity of native Saccharomyces strains is essential for the preservation and exploitation of the oenological potential of wine grape growing regions. Although microorganism biodiversity has hardly been considered before, it could be used alongside other tools to help face the effects of climate change on viticulture and the winemaking process. Key words: yeasts, Saccharomyces, wine, winery, vineyard 1. Introduction The quality of wine is a direct consequence of the evolution of the microbiota of must during fermentation. Yeasts play a central role in the fermentation process during winemaking. Saccharomyces cerevisiae, the wine yeast, is the most important species involved in alcoholic fermentation [1]. In the past, wine fermentation was spontaneously carried out by indigenous yeasts. This method is still applied by several wineries in Argentina to preserve the tipicity (or regional character) of their wines. Yeasts naturally present in musts transform sugars into alcohol, carbon dioxide and other important metabolites [2]. It is generally assumed that these yeasts are present on grapes and winery equipment, although some controversy about this still exists [3]. Due to the extremely low occurrence and the difficulty in isolating Saccharomyces from healthy undamaged grapes by direct plating, some authors have excluded a natural origin of these yeasts, postulating instead only a winery origin for them [4, 5, 6]. However, it has been shown that damaged grapes are very rich depositories of microorganisms including Saccharomyces [7]. The contribution to the fermentation process by flora present on winery equipment surfaces has been widely postulated [8, 9, 10, 11], but only confirmed recently by direct sampling and the isolation of yeasts from these surfaces [12, 13 14]. Yeasts are part of the natural microbial communities of grapes [15]. It is generally thought that unique strains of yeasts are associated with particular grape varieties in specific geographical locations and that significant diversity and regional character, or terroir, are introduced into the winemaking process via this association [16, 17, 18, 19, 20].Thus, the grapes of a region represent an important source of yeasts for starter culture development when trying to preserve both yeast biodiversity and the regional influence on the characteristics of a wine [3]. Argentina represents an important wine producer in South America. Although it has an extensive history in oenology and viticulture, very little is known about the ecology of the microorganisms involved in local fermentation. The development of knowledge on the microbial ecology of local ecosystems is essential for understanding the winemaking process and for generating products with local characteristics, allowing the development of modern winemaking practices and the diversification of wine products. In the present study, two different, but related, aspects were evaluated using a molecular approach for S.cerevisiae strain differentiation. First, we tried to elucidate the origin of S. cerevisiae yeast involved in the spontaneous

fermentation of grape musts by evaluating the distribution of these yeasts on winery equipment and determining their contribution to the fermentation process. Their genetic relationships with commercial yeasts were also evaluated. Secondly, we explored S.cerevisiae strain diversity in ZARM vineyards. The contribution of vineyard S.cerevisiae strains to the population responsible for industrial spontaneous fermentations and the level of genetic relationships between these populations were also evaluated. 2. Saccharomyces winery populations and the origin of fermentative yeasts It is generally assumed that spontaneous grape must fermentation occurs via the development of yeast naturally present in the must, although the origin of these yeasts is a matter of discussion [4, 5, 6, 12, 13]. In order to elucidate the participation of winery and grape Saccharomyces populations in spontaneous fermentations of Malbec musts from the Zona Alta del Río Mendoza region (Argentina), Saccharomyces yeast associated with grapes and winery equipment were analysed over two consecutive years [14]. The total winery yeast population was evaluated on all winery surfaces in contact with grapes or must during processing, from harvest to the end of fermentation. Two sampling points were considered: before vintage (BV) and during vintage (DV). At before vintage time, the processing equipment had been cleaned, disinfected and not used since the previous year. The aim of this first sampling (BV) was to evaluate the ability of the yeasts to survive between two consecutive seasons. At the second sampling point (DV) the winery surfaces were sampled before the must was processed. This sampling (DV) was carried out to evaluate the winery micobiota present during the must processing and its contribution to fermentation. The winery equipment was used to process other grapes between both sampling points. Grape samples were directly obtained upon the arrival of grapes at the winery by random sampling. The grapes were crushed and the tanks were filled. Samples of fresh must (M) were taken once the fresh must was inside the fermentation tanks. During fermentation, two sampling points were considered: the beginning of fermentation (BF), when the initial density of the must diminished in 0.01 g/ml and the end of fermentation (EF when the density remained constant. The fermentations were conducted following standard winery practices without the addition of commercial yeasts. Several commercial S.cerevisiae strains normally employed in ZARM viticulture region wineries were also included in this study. To find out the total yeast population, samples were spread onto two culture media (MEA+BR and WL) carried out in triplicate using a serial dilution method. At the same time, an enrichment procedure (SelMed selective enrichment medium according to Mortimer and Polsinelli [7]) was employed to allow multiplication of Saccharomyces from grapes, where these yeasts could be present at numbers below the detection limit for direct isolation by plating. Saccharomyces and non-saccharomyces yeasts were rapidly discriminated according to their ability to grow in L-lysine medium (Oxoid, Basingtoke, UK). This assignation was confirmed by a conventional yeast identification method following some of the taxonomic criteria described by Kurtzman and Fell (1998) [21]. S.cerevisiae isolates were subsequently differentiated at the strain level by using two molecular methods. Many techniques have been developed using the tools offered by molecular biology and many of them are useful for identifying and characterizing yeasts at the molecular level. The main molecular techniques proposed for the studies of S. cerevisiae strain diversity include: pulsed field electrophoresis [22], mitochondrial DNA restriction analysis [23], interdelta element PCR amplification [24, 25] and the amplification of polymorphic microsatellite loci [26-30]. The two molecular markers selected in this study still represent the simplest and most widely used techniques for studying Saccharomyces biodiversity. We discussed about their differences, usefulness and advantages in a previous work [31]. Firstly, we used interdelta PCR analysis, which allows the amplification of DNA fragments between two delta elements. Delta elements are direct repeat elements which flank the Ty1 retrotransposons that are dispersed on the S.cerevisiae nuclear genome at an amplifiable distance; their number and position in the genome is strain specific and stable in about 50 generations [24]. Mitochondrial DNA restriction fragment length polymorphism was also applied. This produces an unambiguous mitochondrial pattern supported on specific restriction sites. The endonuclease Hinf I recognizes a high number of restriction sites in the nuclear yeast DNA but only a few sites in the mitochondrial genome. Therefore, mitochondrial restriction fragments can be easily separated by agarose gel electrophoresis [23]. Patterns of bands generated with each method were combined to define the pattern, i.e., the strain. 2.1. Total yeast counts in different winery surfaces and grapes. The total yeast counts on different winery surfaces showed an increasing number of total yeasts according to the advancing vintage. The isolation frequency of Saccharomyces also increased. This confirms that the continuous passage of must on winery equipment throughout the vintage season supports the development of yeasts present and may introduce new ones. Furthermore, grape must exerts a positive selective pressure on Saccharomyces because of its high sugar content, low ph and the presence of SO 2 [15]. The yeast counts obtained from grapes and fermentation phases were in agreement with counts previously reported [15, 32, 33]. As we expected, it was not possible to isolate Saccharomyces by direct sampling from the grapes. Recovery by the enrichment method confirmed their low population on this substrate. This fact has been reported before by several authors who questioned the presence of this species in the vineyard ecosystem and postulated a winery origin for it [5, 6, 12, 34]. Several studies demonstrated that

Saccharomyces on grapes occurred at percentages below 0.1% of naturally occurring yeast biota, and that they were not systematically widespread, since neither all plants nor all grape clusters harboured wine yeast [7, 35, 36]. 2.2. Distribution of S.cerevisiae strains in winery Between 9 and 20 different Saccharomyces patterns were found on winery equipment at BV during 2001 and 2002 respectively, (Table 1 shows the results for 2002 as an example). Some surfaces exhibited a unique pattern, i.e. a single yeast strain, while others harboured up to 10 different strains. The equipment surfaces evaluated previously for Malbec grape processing (DV) exhibited an increasing number of different patterns, 22 patterns in 2001 and 35 in 2002. Despite the diversity found in the winery, some strain patterns were present simultaneously and at both sampling times on more than one equipment surface sampled (Table 1). All of the S.cerevisiae isolates found in the winery at BV were considered as resident or perennial strains. However, the S.cerevisiae patterns recovered from the winery for both years, independent of the isolation sampling point, were also included as resident biota under the assumption that they were present in low numbers at BV and were undetectable by the sampling method used. It is important to point out the great diversity of S.cerevisiae strains found on the winery equipment. There are no previous reports of such extensively recovered S.cerevisiae strains. The amount and diversity of wine yeasts present on equipment depends on the standards of cleanliness of the winery and the nature of the surface. Irregular, unpolished surfaces which are difficult to clean, for example pipes and crushers, may support dense populations of winery yeasts [2]. Our results confirmed the presence of a particular Saccharomyces strain population resident in the winery. This population was found to have a dynamic behaviour since it fluctuated from year to year and throughout the vintage season, even though the weather conditions and cleaning protocols of the winery in both years were similar. Despite this dynamism, it is important to note the existence of stable strains throughout the season and in consecutive years. Several authors have observed the persistence of some yeast strains and their subsequent contribution to spontaneous fermentations in wineries [4, 9, 13]. 2.3. S.cerevisiae winery strains in grapes and fermentation Despite the fact that the grapes were sampled from the same vineyard they exhibited different yeast patterns each year. Although S.cerevisiae was found in low number on grapes, one of the predominant strains on the grapes was also isolated from winery surfaces in 2001. This situation could be explained by contamination of the grapes with these yeasts being present on the equipment used in harvest and transport. The presence of these yeasts on different equipment and tools used in harvest operations has been previously demonstrated [15]. Our results showed a change in yeast population composition on the grapes from year to year. The great diversity and heterogeneity in distribution of the wine yeasts in vineyards have been previously demonstrated [35, 36, 37]; these differences can be attributed to the fact that different bunches of grapes were sampled which may have had a distinct flora. The S.cerevisiae population on fresh must consisted of three to five strains which showed different molecular patterns from those previously described on grapes, although some of the yeast strains previously found in winery also appeared in the fresh must samples. On the other hand, the whole strain population on fresh must could not be detected later during fermentation. Interestingly, a strain population change was also observed between the grapes and fresh must once it had been in contact with the winery equipment (Table 1). During spontaneous fermentation, a succession of different S.cerevisiae strains was observed in both years. An increased strain diversity was found during fermentation (Table 1) and a few strains were present at more than one fermentation stage. Similar situations have been described in spontaneous fermentations where the dynamics of different subpopulations throughout the fermentation stages were observed [11, 13, 37, 38, 39, 40]. In the present study, a change in the yeast population during fermentation was observed from year to year. While in 2001 8 different strains participated in must fermentation, in 2002 a total of 22 patterns were found during this process (Table 1). These results agreed with those of other studies, where the micobiota of each year was characterized by the appearance of new strains and by different isolation frequencies than the previously detected strains [10, 11, 37, 41]. Two different situations a large number of strains at low percentages and a smaller number of strains with one dominant strain were found in this study and have been previously reported [13, 36, 40, 41]. Curiously, and in agreement with Santamaría et al. (2005) [13], in the fermentation where one strain was dominant, in the 2001 fermentation one strain represented 67 % and 29 % of strain population at beginning and end of fermentation respectively, it was coincident with a commercial yeast which had not been previously inoculated in the must. Some S.cerevisiae yeast strains found during fermentation were previously isolated from the winery equipment. Around 30% and 60% of the yeast population at the end of fermentation in 2001 and 2002 corresponded to the winery strains, respectively, most being perennial yeast strains. This fact provides evidence of the contribution of winery yeast to the industrial process through simple contact. It has been postulated that numerous different Saccharomyces strains are present in the winery ecosystem and that the different conditions at each harvest (the chemical composition of must, the winemaking process, the level of sulphitation and temperature, for example) could determine which specific strains will develop during fermentation [13]. Similar results were previously observed by Ciani et al. (2004) [12], who demonstrated that autochthonous S.cerevisiae strains in a winery predominate in natural fermentations and

concluded that the contribution to the process by Saccharomyces resident on grapes was not significant or was absent. Strains which had not previously been found on equipment surfaces, nor on the grapes or fresh must, were also found to be involved in the fermentation process. Under our experimental sampling procedure for evaluating the winery surfaces, some strains might not have been detected, resulting in a limited picture of the kinds of strains that actually occur in winery. It could also be possible that these strains came from other equipment at the winery or other fermentation tanks. Table 1 Distribution, as percentage, of Saccharomyces strains, defined by combination of amplification and restriction patterns, isolated during 2002 season at ZARM region. Pattern Code Winery before vintage Winery during vintage Fresh Fermentation Grapes must R CR PTR P F a P F b R CR TI TC BF EF Impact of the pattern I 02 20 Unique in winery II 02 43 6.5 5 Winery and fermentation III 02 8.3 Unique in winery IV 02 8.3 6 Winery two surfacees V 02 20 20.5 5 Winery and fermentation VI 02 12.5 14 Winery repeated VII 02 20 Unique in winery VIII 02 5 Fermentation IX 02 6 20 Commercial XLI 6 Unique in winery XLII 12.5 6.5 6 10 Winery and fermentation XLIIII 5 Fermentation XLIV 6 20 Winery and fresh must XLV 6 Unique in winery XLVI 100 Unique in winery XLVII 100 29 Winery repeated XLVIII 20 Unique in winery XLIX 6 Unique in winery L 8.3 Unique in winery LI 13.5 Unique in winery LII 5 Fermentation LIII 20 Unique in winery LIV 14 Unique in winery LV 33 Grape LVI 6 7 Winery and fermentation LVII 14 8.3 Repeated in winery LVIII 6 Unique in winery LIX 43 8.3 Repeated in winery LX 6 Unique in winery LXI 29 Unique in winery LXII 14 Unique in winery LXIII 8.3 Unique in winery LXIV 13.5 12 15 5 Winery and fermentation LXV 6 Unique in winery LXVI 20 Fresh must LXVII 8.3 Unique in winery LXVIII 8.3 5 Winery and fermentation LXIX 8.3 Unique in winery LXX 6.5 7 Winery and fermentation LXXI 6 Unique in winery LXXII 6.5 Unique in winery LXXIII 12.5 8.3 Unique in winery LXXIV 8.3 Unique in winery LXXV 6.5 21 5 Winery and fermentation LXXVI 6 5 Winery and fermentation LXXVII 7 Fermentation LXXVIII 7 Fermentation LXXIX 13.5 16 Repeated in winery LXXX 67 Grape LXXXI 20 Fresh must LXXXII 20 Fresh must LXXXIII Unique in winery LXXXIV 12 7 10 Winery and fermentation LXXXV 15 5 Commercial LXXXVI 7 Fermentation LXXXVII 7 Fermentation LXXXVIII 5 Fermentation LXXXIX 6.5 Unique in winery XC 12.5 Unique in winery XCI 6 Unique in winery XCII 5 Fermentation XCIII 8.3 6 10 Winery and fermentation XCIV 10 Fermentation XCV 20 Unique in winery R: reception equipment; CR: crusher; PTR: pipe for must transport; PF: pipe for filling; TI: surface interior of tank; TC: exterior connections of the tank; BF: beginning of fermentation; FF: final of fermentation. Our results also indicated commercial strain participation in fermentations conducted without yeast inoculation. Similar findings were reported by other authors [9, 13]. Although the Malbec must under study was spontaneously fermented in this winery, white wine fermentations are usually conducted by the inoculation of commercial yeasts. These yeasts could remain on the equipment and may become predominant when spontaneous fermentations are delayed, although their participation would decrease as the fermentation progressed, as was shown by our results. Nevertheless, in this winery, no commercial strains were isolated on the equipment at the first sampling point,

indicating that this yeast could not remain in the winery from year to year. Taking the contribution to the fermentative population of the winery and commercial strains as a whole, they represented in average a 70 and 68% of strains found in fermentation every season respectively. In summary, the following findings can be highlighted: a) the low occurrence of Saccharomyces on grapes and its limited participation in fermentation was confirmed; b) the population of S.cerevisiaes on fresh must showed a different strain composition to the populations previously described on grapes, with only one exception. Moreover, 40% of the strains in these samples were similar to those previously found in the winery equipment; c) a sequential substitution of S.cerevisiae strains was observed during fermentation. Around 30% and 60% of the yeast population at the end of fermentation had originated at the winery in the 2001 and 2002 vintages, respectively; d) the presence of Saccharomyces on winery surfaces at the two different sampling points in both seasons was observed. A stable and resident S.cerevisiae microbiota in the winery was confirmed, consisting of very diverse strains with most of them showing a dynamic behaviour. Additionally, an important participation by the winery yeasts in fermentation was demonstrated and their contribution was found to be dependent on the native yeast populations on must and the oenological practices employed; e) commercial yeast strains were found during fermentation at different percentages even though they were hardly found on the winery surfaces. 3. Vineyards and the noteworthy yeast biodiversity One outcome of numerous reports is that the quality of a wine is a direct consequence of the yeast biota which developed during fermentation [42]. In different wine regions around the world, there is currently a growing demand for autochthonous strains of fermentative yeast with typical oenological characteristics representative of those particular regions. These strains could be better adapted to the conditions of a particular region, including the soil and climate conditions, the grape variety and the viticulture management and oenological techniques which are used there. Since the importance of S. cerevisiae in wine production was already established, the use of foreign commercial yeast cultures has been transformed into one of the most common practices for reducing the risk of spoilage of the wine. However, this practice could prevent the production of some desirable or typical organoleptic characteristics of the wines and reduce the diversity of the natural microbiota [43]. On the other hand, some studies have demonstrated a certain correlation between S. cerevisiae strains isolated from grapes and wines and their geographical origin. The results of some studies on yeast biodiversity in vineyards supported the theory that in spite of the great polymorphism observed, a population of yeasts considered as particular to a viticulture region or terroir might exist [44, 45, 46, 47]. Each viticulture region constitutes an ecosystem where yeasts are included in the microbial diversity. The characterization of yeasts from a viticulture region allows the determination of their population structure, their distribution in the vineyard and the genetic relationships between them in relation to their geographical origins, and it also allows the evaluation of their participation in fermentations. All of this information could be used in the selection of native yeasts and the conservation of genetic resources. In order to microbiologically characterize the Zona Alta del Río Mendoza viticulture region, eight different vineyards spread all over this region were selected. The grape sampling design included ten different sampling points distributed throughout the sub-area of the vineyard evaluated. The grapes obtained from each sample point were individually processed to determine the distribution of Saccharomyces within the vineyard. Afterwards, the vineyards were harvested and the grapes were transported to six different wineries located in the same viticulture region, avoiding mixing with other grapes, for spontaneous fermentation. 3.1. Saccharomyces populations from vineyards We previously reported the low populations of Saccharomyces on mature healthy grapes and the difficulty in isolating these yeasts from grapes using direct isolation methods [14, 48]. Therefore, an enrichment procedure was applied in order to recover the S.cerevisiae strains from the grapes. The samples were aseptically crushed and the musts obtained were allowed to ferment. The yeasts were later isolated. Using this procedure means that the results obtained reflect the S.cerevisiae strains able to develop in the conditions imposed. It must be taken into account that conditions in a vineyard are quite different from those in fermentation, and that the populations which are isolated and characterized only reflect those yeasts with particular competitive traits allowing them to survive the fermentation process, i.e. yeasts with some oenological interest. This methodology could give a distorted picture of the Saccharomyces populations present in vineyards, but it did not contradict the aims of this work and it is a commonly used strategy for the study of vineyard yeast populations [36, 49, 50, 51, 52]. A combination of molecular patterns obtained by interdelta PCR and RFLP mtdna allowed 1020 S.cervisiae patterns to be differentiated. The selection of these molecular markers for the characterization of Saccharomyces populations was found to be very useful in a study of closely related S.cerevisiae strains, as was as their simplicity and low cost, as previously discussed [31]. The vineyards of ZARM exhibited a variable number of S.cerevisiae strains: 9 to 36 different molecular patterns in total were observed per vineyard (Table 2). A non-homogeneous distribution of these yeasts was verified; the different sampling sites in the vineyards harboured between 1 to 12 different patterns. In some cases, one or two sites showed great diversity whereas in others S.cerevisiae strains could not be isolated (Figure 1).

Current Research, Technology and Education Topics in Applied Microbiology and Microbial Biotechnology A. Méndez-Vilas (Ed.) FORMATEX 2010 1047

Additionally, pattern 1CB, found in two sites of vineyard CB, also coincided with another commercial strain. In addition to the important presence and spread of commercial yeasts, the low diversity vineyards showed a drastic decrease in S.cerevisiae diversity, as evidenced by the small number of strains isolated here. Moreover, high biodiversity values (ratio of number isolates/number of strains) of 4.9 and 6.2 were observed for vineyards S and CB, respectively. Coincidentally, both vineyards are located in a very close proximity to the winery, indicating that the currently used oenological practices, such as the intensive use of commercial inoculates, could affect the biodiversity of the vineyard ecosystem, decreasing polymorphism and replacing the native biota with foreign organisms [59]. High polymorphism Low polymorphism Table 2 Diversity of S.cerevisiae strains found in vineyards from ZARM wine region of Mendoza, Argentina Total yeast Number of Number of Number of Commercial patterns in Vineyard populations Number patterns patterns present vineyard vineyard Common patterns code (Log cfu/ml of distinct present in in more than patterns found between vineyards Incidence ±sd) patterns unique one vineyard in Pattern Site (%) vineyard site site fermentations T 4.92 ± 0.31 36 30 6-2 - - - 2 = 12 CZ NL 3.31 ± 1.12 25 21 4 3 = 5 CZ 2 - - - 6 = 4 CZ CR 3.21 ± 0.76 31 27 4-3 - - - CZ 4.97 ±0.61 35 32 3 4 = 6 NL 5 = 3 NL 12 = 2 NL 2 - - - ID 4.05 ±0.71 26 22 4-2 1 1 3 5 45.5 25 75 25 3 25 NP 3.58 ±1.13 23 18 5-1 6 4 12. 5 1 100 1 5 80 2 55.6 CB 2.81 ±1.17 9 7 2-2 4 100 2 7 100 9 100 S 2.31 ±0.71 10 5 5-3 3 9 1 16.7 3 100 4 61.5 6 85.7 8 60 7 33 8 20 3.2. Saccharomyces populations in industrial spontaneous fermentations With the aim of examining the dynamics of Saccharomyces during spontaneous fermentations at an industrial scale in different wineries from the ZARM region, the previously sampled vineyards were harvested and the grapes were processed in six different wineries also located in the ZARM region. The spontaneous fermentations were conducted according to the standard protocols of the different companies. Samples were taken at specific stages of the fermentation process, defined in terms of must density, for comparison. The fermentations evaluated under industrial conditions showed different levels of S.cerevisiae present at the beginning of the processes (BF), even though mature healthy Malbec grapes were always used and the wineries have a similar level of technology, knowledge and prestige in the production of quality wines. On the other hand, the intermediate stage (MF) of fermentation showed yeast counts more homogeneous than those initially recorded, with a predominance of S.cerevisiae in all cases, which became the exclusive species by the end of fermentation. The fermentations evaluated showed a highly polymorphic nature and a variable behaviour of Saccharomyces populations in each case. Moreover, the Saccharomyces populations were different in different fermentations, with 5-21 patterns observed in total. In the different fermentation stages, different numbers of patterns were observed, with a maximum of 13. The dynamics of S.cerevisiae patterns during the fermentations were different and two main tendencies were observed: fermentations with a high polymorphism at the start with the number of patterns involved decreasing during the process; and the other fermentations which showed a low polymorphism in the beginning with increasing populations co-existing until the end of the process. The low polymorphic vineyards (S, CB) showed only a few strains at the beginning of fermentation and the presence of commercial strains, previously found in the vineyards, participated in both processes. In addition to the variability in the number of patterns found in different stages of the process, a different permanence of these patterns during fermentation was also observed. In general, a substitution of patterns was verified, with multiple strains conducting the fermentation with no predominance. This result suggests that strains found in the different stages could exhibit different physiological characteristics, with some of them more adapted to tolerate the high osmotic pressure at the initial stage, with others having faster growth or more tolerance to ethanol [60]. This

alteration in the strains is characteristic of spontaneous fermentation and constitutes a strength whereby the risk of sluggish and stuck fermentations is reduced, and the strains could contribute metabolites which provide a greater complexity to the wine [59]. The participation of vineyard strains in spontaneous fermentation was low or even nonexistent. Although in most fermentations the patterns originated from vineyard (one to three in each case), these patterns did not display a significant prevalence in the different fermentations (Table 3). Curiously, in just one case (the low polymorphic vineyard S), the fermentative population mainly came from the vineyard. The population involved in this process had a very limited diversity and was homogenous throughout the process with three patterns present at all stages; moreover, two of these persistent patterns corresponded to a commercial strain. The different yeast patterns found in the fermentations did not coincide with the vineyard patterns but they could have corresponded to isolates originally present in the vineyard samples but which were not recovered under the conditions used for isolation, or they could have represented winery strains incorporated by contamination during the processing of musts in the winery. Table 3 Diversity of S.cerevisiae strains found in spontaneous fermentations from ZARM wine region of Mendoza, Argentina Vineyard name Fermentation stage Total number of different patterns Number of patterns present in unique stage Number of patterns presents in more than one stage (pattern code) Number of patterns from vineyard (pattern code) Number of commercial pattern (pattern code) B 10 7 3 (36-37-40) 1 (36) 1 (40) T M 7 4 3 (36-37-40) 1(36) 1(40) F 7 4 2 (36-40) 2 (16-36) 1 (40) B 3 3-1(20) - NL M 6 5 2 (29-33) - - F 13 11 2 (29-33) 1(13) - B 1 1 0 0 1(33) CR M 10 8 4(3-32-36-44) 2(3-31 - F 9 6 4(3-32-36-44) 1(3) - B 6 5 6(4-28-36-37-38-39) 2(4-28) - CZ M 7 5 7(4-28-36-37-38-39-40) 1(4) 1(40) F 2 1 1(40) - 2(40-46) B - - - - - ID M 6 3 3(1-25-28) 2(1-25) 2(1-25) F 7 4 3(1-25-28) 2(1-25) 2(1-25) B 9 7 2(23-28) 1 (23) - NP M 7 3 4(23-28-33-36) 1 (23) - F 8 5 3(23-33-36) 1 (23) - B 2 1 1(10) - - CB M 10 8 3(2-10-18) 2(1-2) 2(1-2) F 8 6 2(2-18) 1(2) 1(2) B 3 0 3(3-9-10) 2(3-9) 3(3-9-10) S M 3 0 3(3-9-10) 2(3-9) 3(3-9-10) F 5 2 3(3-9-10) 3(3-9-11) 3(3-9-10) -: none; ns: not sampled; B:beginning of fermentation; M: middle of fermentation; F: final of fermentation 3.3. The impact of commercial yeasts in vineyards and fermentations The molecular patterns of S. cerevisiae strains isolated from the vineyards and fermentations were compared with a set of 30 commercial strains which included those widely used in the wineries of the ZARM region. S. cerevisiae isolates from four vineyards coincided with commercial strains (Table 1). Moreover, within this group, the low polymorphic vineyards CB and S showed a wide distribution and high incidence of such strains. As commented above, both vineyards are located next to the respective cellar, suggesting a transfer of yeasts from the winery to the vineyard. Two other vineyards which belonged to the highly polymorphic group of vineyards (NP and ID) also presented some patterns coincident with commercial strains. The NP vineyard is located approximately 200 m from the winery where the grapes were processed; this distance could explain why the commercial strains were not spread throughout this vineyard and why the global diversity of S.cerevisiae populations found there was not affected, as it was the case of vineyards S and CB. Moreover, another difference found was that the commercial strains present in vineyard NP were not found later in the corresponding fermentation. On the other hand, vineyard ID is also located near a winery (300 m) and this could be the cause of the presence of commercial strains on the grapes from this vineyard. These results agreed with those of previous works, which suggested that the dispersal of commercial strains is mainly mediated by water runoff, macerated grape skin at dumping sites and different vectors, such as insects or others, which would be responsible for their presence at distances greater than 1000 m from the cellar [51]. The commercial strains were also found in some spontaneous fermentation evaluated in the present study, even though commercial strains were not found in the corresponding grapes (Table 1 and 2). In these cases, it could be

inferred that they were incorporated during winemaking in the cellar. Previous studies demonstrated the horizontal transfer of yeasts in wineries by cross contamination, mediated by winery equipment employed daily in the processing of musts, which transferred microorganisms from tank to tank [9, 61]. The present work emphasizes the real situation in Argentinean oenology, where all the wineries evaluated corresponded to commercial industrial companies where commercial yeast cultures are commonly used for fermentation. This fact could explain the presence of commercial yeasts in the spontaneous fermentations evaluated. It is noteworthy that only four different commercial strains were detected in the grapes and fermentations. Moreover, one of them was repetitively found in four vineyards and five fermentations. This commercial strain is intensively and widely used in the viticulture region evaluated. This fact could explain its generalized presence, in addition to its characteristics which allow it to survive in the vineyard and in the winery, permitting it to compete with the other strains present in these ecosystems. The Saccharomcyes biodiversity results arising from this ecological area, the ZARM wine region, provide evidence of the importance of evaluating different aspects of microbial diversity in order to assess the complete picture of yeasts involved in the winemaking environment. Moreover, knowledge of the biological patrimony of yeasts is essential to their maintenance, and is the source of the genetic background needed to obtain starter strains which are able to fully develop the typical flavours and aromas of wines originating from different grapevine cultivars [2] and to ensure the conservation of gene pools of primary importance for the preservation of productive activities based on yeast mediated processes [62]. 4. Molecular relationships among S.cerevisiae strains These results show a picture of Saccharomyces populations present in the ZARM wine region and illustrate the populations of S. cerevisiae present in grapes, winery equipment and spontaneous fermentations. The analysis of coincidences in the molecular patterns clearly visualized a huge polymorphism at this level. The identity among two isolates was assigned based on total coincidence in the corresponding patterns of bands. But this criterion had some limitations, for example: were the two patterns which were considered different by only one band totally different? Or did some relationship exist among them that could be detected? In order to answer these questions a cluster approach was proposed. Pattern of bands were compared after the estimation of size band and similarities based on the Dice coefficient and dendrograms were constructed using the UPGMA method. A subset of S. cerevisiae strains representing winery-fermentation isolates and commercial strains was selected and the corresponding patterns were compared. A third molecular marker was used in order to gain information about the molecular relationships between the groups of strains. The microsatellite analysis using six different loci were applied according to Jubani et al. (2007) [30].. As shown in the dendrograms, different similitude coefficients were observed according the molecular marker used (Figure 2). These results provided evidence of polymorphism at different genetic levels, with the nuclear and mitochondrial markers allowing different groupings of strains. Three clusters of strains were repetitively conserved independently of the molecular pattern utilized in the construction of the dendrogram; they mostly included isolates from winery equipment. Moreover, the repetitive cluster of these isolates, inferring a monophyletic origin, and the slight relationship with the commercial strains suggested an American origin accompanied by microevolutionary events in recent times. On the other hand, the rest of the isolates from the winery and the fermentations showed a random clustering according to the molecular marker applied; this result suggests some kind of change at the nuclear level, which could occur at a different rate in the nucleus with respect to the mitochondria, during the lifecycle of the yeast. Recent studies demonstrated a low stability of the genome in wine yeast [2], which may be due to the high reorganization capacity of its genome by Ty-promoted translocation, mitotic recombination and gene conversion [19, 63, 64]. Alternatively, it has also been proposed that ethanol and acetaldehyde introduce breaks in the DNA, with a much higher mutation rate on the mitochondrial genome. This may be due to a higher efficiency of the yeast nuclear DNA repair system compared with the mitochondrial system that lacks a proofreading activity [19, 65]. The set of commercial strains included in this analysis showed different relationships with those isolated from the winery equipment and fermentation strains. Some of them clustered separately in the different analysis but others did not. Only one was always clustered with an isolate from spontaneous fermentation. According to these results, the European strains were not restricted to a sole cluster, as was previously found [19]. The existence of some genetic relationships between a few winemaking-related strains with some of the commercial strains would support the hypothesis that some native American strains proceeded from European strains, as was recently suggested [30]. Conversely, it was also suggested that the existence of genomic resemblance among the native Saccharomyces and commercial strains, caused by centuries of positive selection during wine production could lead to the same characteristics being explored during the process of selection for commercial cultures [66]. It was suggested that resemblance in phenotype is reflected in genotypic characteristics [66]. The results of the studies presented here showed the complex relationships found at the molecular level among the yeast isolates that share the same ecological environment. Moreover, this study revealed that, despite an abundant diversity, these yeasts share many genetic characteristics.

I A II III II B I III I C III II Figure 2 Dendrograms showing molecular relationships based on PCR interdelta (A), RFLP mtdna (B) y SSR (C) for 28 S.cerevisiae isolates and 7 commercial selected strains. Cophenetic correlation: (A),r = 0,88019; (B), r = 0,85498; (C), r = 0,88078. Clusters repetitively grouped were indicated with roman numbers I, II and III.

References [1] Fleet GH. The commercial and community significance of yeasts in food and beverage production. In: Querol A, Fleet GH, eds. The Yeast Handbook - Yeasts in Food and Beverages. Berlin Heidelberg, Springer-Verlag, 2006. [2] Pretorius IS. Tailoring wine yeast for the new millennium: novel approaches to the ancient art of winemaking. Yeast. 2000, 16: 1-55. [3] Fleet GH. Wine yeasts for the future. FEMS Yeast Res. 2008, 8: 979 995 [4] Rosini G. Assessment of dominance of added yeast with fermentation and origin of Saccharomyces cerevisiae in wine making. J Gen Appl Microbil. 1984, 30: 249-256. [5] Martini, A. Origin and domestication of the wine yeast Saccharomyces cerevisiae. J.Wine Res. 1993, 4(3): 165-176 [6] Vaughan-Martini A., Martini A. Facts, myths and legends on the prime industrial microorganism. J Ind Microbiol. 1995, 14: 514-522 [7] Mortimer R, Polsinelli M. On the origins of wine yeast. Res Microbiol. 1999, 150: 199-204. [8] Freizer V, Dubourdieu D. Ecology of yeast strains of Saccharomyces cerevisiae during spontaneous fermentation in Bordeaux winery. Am J Enol Vitic. 1992, 43: 375-380. [9] Constantí M, Poblet M, Arolo L, Mas A, Guillamón JM. Analysis of yeast population during alcoholic fermentation in a newly established winery. Am J Enol Vitic. 1997, 48: 339-344. [10] Zilio F, Lombardi A, Galeotto A, Comi G. Mitochondrial DNA restriction patterns of Saccharomyces strains isolated in the area of production of Soave DOC. Riv Vitic Enol. 1998, 3: 33-41 [11] Torija MJ, Rozes N, Poblet M, Guillamon JM, Mas A. Yeast population dynamics in spontaneous fermentations: comparison between two different wine-producing areas over a period of three years. Antonie van Leeuwenhoek. 2001,79: 345-352. [12] Ciani M, Mannazzu I, Marinangeli P, Clementi F, Martini A. Contribution of winery-resident Saccharomyces cerevisiae strains to spontaneous grape must fermentation. Antonie van Leeuwenhoek. 2004, 85: 159-164. [13] Santamaría P, Garijo P, López R, Tenorio C, Gutierrez R. Analysis of yeast population during spontaneous alcoholic fermentation: Effect of the age of the cellar and the practice of inoculation. Int J Food Microbiol. 2005, 103: 49-56 [14] Mercado L, Dalcero A, Masuelli R, Combina M. Diversity of Saccharomyces strains on grapes and winery surfaces: Analysis of their contribution to fermentative flora of Malbec wine from Mendoza (Argentina) during two consecutive years. Food Microbiol. 2007, 24: 403 412. [15] Ribéreau-Gayon P, Dubordieu D, Donèche B, Lombaud A. (Eds.) Citology, taxonomy and ecology of grape and wine yeasts. In: Handbook of Enology. The microbiology of wine and vinifications 2 nd edition. Chichester, John Wiley& Soons, Ltd. (2006). [16] Vezinhet F, Hallet J, Valade M, Poulard A Ecological survey of wine yeast strains by molecular methods of identification. Am J Enol Vitic. 1992, 43: 83 86. [17] Pretorius IS, van der Westhuizen TJ, Augustyn OPH. Yeast biodiversity in vineyards and wineries and its importance to the South African wine industry. A review. S Afr J Enol Vitic. 1999, 20: 61 74. [18] Raspor P, Milek DM, Polanc J, Mozina SS, Nadez N. Yeasts isolated from three varieties of grapes cultivated in different locations of the Dolenjska vine-growing region, Slovenia. Int J Food Microbiol. 2006, 109: 97 102. [19] Martínez, C., Cosgaya, P., Vásquez, C., Gac, S. and Ganga, A. High degree of polymorphism and geographic origin of wine yeast strains J Appl Microbiol. 2007, 103: 2185-95 [20] Valero E, Cambon B, Schuller D, Casal M, Dequin S Biodiversity of Saccharomyces yeast strains from grape berries from wineproducing areas using starter commercial yeasts. FEMS Yeast Res. 2007, 7: 317 329. [21] Kurtzman CP, Fell JW, eds. The Yeast: A taxonomic study. Fourth Revised and Enlarged Edition. Amsterdam: Elsevier Science Publisher. 1998. [22] Blondin B, Vezinhet F. Identification de souches de levures oenologiques par leurs caryotypes obtenus en électrophorèse en champ pulsé. Rev Fr Oenol. 1988, 115:7 11 [23] Querol A, Barrio E, Ramon D. A comparative study of different methods of yeast strains characterization. Syst Appl Microbiol. 1992, 15: 439-446. [24] Ness C, Lavalle F, Dubourdieu D, Aigle M, Dulau L. Identification of yeast strains using PCR. J Sci Food Agric. 1993, 32: 89-94 [25] Legras JL, Karst F. Optimisation of interdelta for Saccharomyces cerevisiae strain characterization. FEMS Microbiol Lett. 2003, 221:249 255 [26] Gallego FJ, Perez G, Martinez I, Hidalgo P. Microsatellites obtained from database sequences are useful to characterize Saccharomyces cerevisiae. Am J Enol Vitic. 1998, 49: 350 351. [27] Howell, K.S., Bartowsky, E.J., Fleet, G.H. and Henschke, P.A. Microsatellite PCR profiling of Saccharomyces cerevisiae strains during wine fermentation. Lett Appl Microbiol. 2004, 38:315 320 [28] Pérez MA, Gallego F, Hidalgo P. Evaluation of molecular techniques for the genetic characterization of Saccharomyces cerevisiae strains. FEMS Microbiol Lett. 2001, 205: 375-378. [29] Legras JL, Ruh O, Merdinoglu D, Karst F. Selection of hypervariable microsatellite loci for the characterization of Saccharomyces cerevisiae strains. Int J Food Microbiol. 2005, 102: 73 83. [30] Jubany S, Tomasco I, Ponce de León I, Medina K, Carrau F, Arrambide N, Naya H, Gaggero C. Toward a global database for the molecular typing of Saccharomyces cerevisiae strains. FEMS Yeast Res. 2008, 8: 472-484. [31] Mercado L, Jubany S, Gaggero C, Masuelli RW, Combina M. Molecular relationships between Saccharomyces cerevisiae strains involved in winemaking from Mendoza, Argentina. Curr Microbiol. 2010, On line first, DOI 10.1007/s00284-010-9645- y [32] Fleet, G.H., Heard, G.M. Yeasts-Growth during fermentation. In: Fleet GH, ed. Wine-Microbiology and Biotechnology. Singapore: Harwood Academic Publishers; 1993. [33] Fungelsang K, Edwards C, eds. Wine microbiology, practical applications and procedures 2 nd ed. New York: Springer Science + Business Media, LLC; 2007.

[34] Sabate J, Cano J, Esteve Zardozo B, Guillamon JM. Isolation and identification of yeast associated with vineyard and winery by RFLP analysis of ribosomal genes and mitochondrial DNA. Microbiol Res. 2002, 157: 1-8. [35] Török T, Mortimer RK, Suzzi G, Polsinelli M. Quest for wine yeast-and old story revisited. J Ind Microbiol. 1996, 17: 303-313. [36] van der Westhuizen TJ, Augustyn OHP, Pretorius IS. Geographical distribution of indigenous Saccharomyces cerevisiae strains isolated from vineyards in the coastal regions of the western Cape in South Africa. S Afr J Enol Vitic. 2000, 21: 3-9. [37] Schuller D, Alves H, Dequin S, Casal M. Ecological survey of Saccharomyces cerevisiae strains from vineyards in the Vinho Verde Region of Portugal. FEMS Microbiol Ecol. 2005, 51: 167-177. [38] Querol A, Barrio E, Ramon D. Population dynamics of natural Saccharomyces strains during wine fermentation. Int J Food Microbiol. 1994, 21: 315-323. [39] Sabate J, Cano J, Querol A, Guillamon JM. Diversity of Saccharomyces strains in wine fermentations: analysis for two consecutive years. Lett Appl Microbiol. 1998, 26: 452-455. [40] Pramateftaki PV, Lanaridis P, Typas MA Molecular identification of wine yeast at species or strain level: a case study with strains from two vine-growing areas of Greece. J Appl Microbiol. 2000, 89: 236-248. [41] Schütz M, Gafner J. Analysis of yeast diversity during spontaneous and induced alcoholic fermentation. J Appl Bacteriol. 1999, 75: 551 558. [42] Fleet GH. Yeast interactions and wine flavour. Int J Food Microbiol. 2003, 86: 11 22. [43] Romano P, Capece A, Jespersen L. Taxonomic and ecological diversity of food and beverage yeasts. In: Querol A, Fleet GH, eds. The Yeast Handbook - Yeasts in Food and Beverages. Berlin Heidelberg: Springer-Verlag. 2007. [44] Versavaud A, Courcoux P, Roulland C, Dulau L, Hallet JN. Genetic diversity and geographical distribution of wild Saccharomyces cerevisiae strains from the wine producing area of Charentes, France. Appl Environ Microbiol. 1995, 61:3521-3529. [45] Guillamón JM, Barrio E, Querol A. Characterization of wine yeast strains of the Saccharomyces genus on the basis of molecular markers: relationships between genetic distance and geographic or ecological origin. Syst Appl Microbiol. 1996, 19:122 132. [46] Khan W, Augustyn OPH, van der Westhuizen TJ, Lambrechts MG, Pretorius IS. Geographic distribution and evaluation of Saccharomyces cerevisiae strains isolated from vineyards in the warmer inland regions of the Western Cape in South Africa. S Afr J Enol Vitic. 2000, 21: 17-29 [47] González SS, Barrio E, Querol A. Molecular identification and characterization of wine yeasts isolated from Tenerife (Canary Island, Spain). J Appl Microbiol. 2007, 102: 1018 1025. [48] Combina M, Elía A, Mercado L, Catania C, Ganga A, Martínez C. Dynamics of indigenous yeast populations during spontaneous fermentation of wines from Mendoza, Argentina. Int J Food Microbiol. 2005, 99: 237-243. [49] Polsinelli M, Romano P, Suzzi G, Mortimer R. Multiple strains of Saccharomyces cerevisiae on a single grape vine. Lett Appl Microbiol. 1996, 23: 110-114. [50] Martínez C, Gac S, Lavin A, Ganga M. Genomic characterization of Saccharomyces cerevisiae strains isolated from wineproducing areas of South America. J Appl Microbiol. 2004, 96:1161 1168. [51] Valero E, Schuller D, Cambon B, Casal M, Dequin S. Dissemination and survival of commercial wine yeast in the vineyard: a utilarge-scale, three years study. FEMS Yeast Res. 2005, 10: 959-969. [52] Schuller D, Casal M. The genetic structure of fermentative vineyard-associated Saccharomyces cerevisiae populations revealed by microsatellite analysis. Antonie van Leeuwenhoek. 2007, 91:137 150. [53] Gutierrez A.R., Lopez R., Santamaria P. Sevilla M. (1997) Ecology of inoculated and spontaneous fermentations in Rioja Spain musts, examined by mitochondrial DNA restriction analysis. International Journal of Food Microbiology. 20: 241-245. [54] Cavallieri D, Barbeiro C, Casalone E, Pinzauti F, Sebastián F, Mortimer R, Polsinnelli M. Genetic and Molecular diversity in Saccharomyces cerevisiae natural populations. Food Technol Biotechnol. 1998, 36: 45-50. [55] Comi G, Maifreni M, Manzano M, Lagazio C, Cocolin L. Mitochondrial DNA restriction enzyme analysis and evaluation of the enological characteristics of Saccharomyces cerevisiae strains isolated from grapes of the wine-producing area of Collio Italy. Int J Food Microbiol. 2000, 58:117 121. [56] Schütz M, Gafner J. Analysis of yeast diversity during spontaneous and induced alcoholic fermentation. J Appl Bacteriol. 1993, 75:551-558. [57] Lopes CA, van Brook M, Querol A, Caballero A. Saccharomyces cerevisiae wine yeast populations in a cold region in argentinean Patagonia. A study at different fermentation scales. J Appl Microbiol. 2001, 93: 608-615. [58] Goode J. Terroir: how do soils and climate shape wine? In: Lumsden H, Williams-Leedham Y, Hayes G, eds. The science of wine. Berkeley and Los Angeles: University of California Press. 2005: 25-31 [59] Mercado L. Biodiversidad de Saccharomyces en viñedos y bodegas de la región vitícola Zona Alta del Río Mendoza. PhD thesis. Universidad Nacional de Cuyo, Argentina. 2009. [60] Querol A, Belloch C, Fernández-Espinar MT, Barrio E. Molecular evolution in yeast of biotechnological interest Int Microbiol. 2003, 6: 201 205. [61] Beltran G, Torija M, Novo M, Ferrer N, Poblet M, Guillamon JM, Rozes N, Mas A. Analysis of yeast populations during alcoholic fermentation: A six years follow-up study. Syst Appl Microbiol. 2002, 25: 287-293. [62] Di Maro E, Ercolini D, Coppola S. Yeast dynamics during spontaneous wine fermentation of the Catalanesca grape Int J Food Microbiol. 2007, 117:201 210. [63] Puig S, Querol A, Barrio E, Pérez-Ortín JE. Mitotic recombination and genetic changes in Saccharomyces cerevisiae during wine fermentation. Appl Environ Microbiol. 2000, 66:2057 2061 [64] Rachidi N, Barre P, Blondin B. Multiple Ty-mediated chromosomal translocation lead to karyotype changes in a wine strain of Saccharomyces cerevisiae. Mol Gen Genet. 1999, 261:841 850 [65] Castrejón F, Codón A, Cubero B, Benítez T. Acetaldehyde and ethanol are responsible for mitochondrial DNA (mtdna) restriction fragment length polymorphism (RFLP) Syst Appl Microbiol. 2002, 25: 462-467. [66] Carreto L, Eiriz M, Gomes A, Pereira PM, Schuller D, Santos M. Comparative genomics of wild type yeast strains unveil important genome diversity. BMC Genomics (2008) 9:524.