Detection of abnormal fermentations in wine process by multivariate statistics and pattern recognition techniques

Three multivariate statistical techniques (Multiway Principal Component Analysis, Multiway Partial Least Squares, and Stepwise Linear Discriminant Analysis) and one artificial intelligence method (Artificial Neural Networks) were evaluated to detect and predict early abnormal behaviors of wine fermentations. The techniques were tested with data of thirty-two variables at different stages of fermentation from industrial wine fermentations of Cabernet Sauvignon. All the techniques studied considered a pre-treatment to obtain a homogeneous space and reduce the overfitting. The results were encouraging; it was possible to classify at 72h 100% of the fermentation correctly with three variables using Multiway Partial Least Squares and Artificial Neural Networks. Additional and complementary results were obtained with Stepwise Linear Discriminant Analysis, which found that ethanol, sugars and density measurements are able to discriminate abnormal behavior.     

Detection of abnormal processes of wine fermentation by support vector machines

The early detection of problematic fermentations is one of the main problems that appear in winemaking processes, due to the significant impacts in wine quality and utility. This situation is specially important in Chile because is one of the top ten wine production countries. In last years, different methods coming from Multivariate Statistics and Computational Intelligence have been applied to solve this problem. In this work we detect normal and problematic (sluggish and stuck) wine fermentations applying the support vector machine method with three different kernels: linear, polynomial and radial basis function. For the training algorithm, we use the same database of 22 wine fermentation studied in [1, 2] that contains approximately 22,000 points, considering the main chemical variables measured in this kind of processes: total sugar, alcoholic degree and density. Our main result establishes that the SVM method with third degree polynomial and radial basis kernels predict correctly 88 and 85 % respectively. The fermentation behavior results have been obtained for a 80–20 % training/testing percentage configuration and a time cutoff of 48 h.     

Temperature-Dependent Kinetic Model for Nitrogen-Limited Wine Fermentations

A physical and mathematical model for wine fermentation kinetics was adapted to include the influence of temperature, perhaps the most critical factor influencing fermentation kinetics. The model was based on flask-scale white wine fermentations at different temperatures (11 to 35°C) and different initial concentrations of sugar (265 to 300 g/liter) and nitrogen (70 to 350 mg N/liter). The results show that fermentation temperature and inadequate levels of nitrogen will cause stuck or sluggish fermentations. Model parameters representing cell growth rate, sugar utilization rate, and the inactivation rate of cells in the presence of ethanol are highly temperature dependent. All other variables (yield coefficient of cell mass to utilized nitrogen, yield coefficient of ethanol to utilized sugar, Monod constant for nitrogen-limited growth, and Michaelis-Menten-type constant for sugar transport) were determined to vary insignificantly with temperature. The resulting mathematical model accurately predicts the observed wine fermentation kinetics with respect to different temperatures and different initial conditions, including data from fermentations not used for model development. This is the first wine fermentation model that accurately predicts a transition from sluggish to normal to stuck fermentations as temperature increases from 11 to 35°C. Furthermore, this comprehensive model provides insight into combined effects of time, temperature, and ethanol concentration on yeast (Saccharomyces cerevisiae) activity and physiology.     

Ethanol Production and Maximum Cell Growth Are Highly Correlated with Membrane Lipid Composition during Fermentation as Determined by Lipidomic Analysis of 22 Saccharomyces cerevisiae Strains

Optimizing ethanol yield during fermentation is important for efficient production of fuel alcohol, as well as wine and other alcoholic beverages. However, increasing ethanol concentrations during fermentation can create problems that result in arrested or sluggish sugar-to-ethanol conversion. The fundamental cellular basis for these problem fermentations, however, is not well understood. Small-scale fermentations were performed in a synthetic grape must using 22 industrial Saccharomyces cerevisiae strains (primarily wine strains) with various degrees of ethanol tolerance to assess the correlation between lipid composition and fermentation kinetic parameters. Lipids were extracted at several fermentation time points representing different growth phases of the yeast to quantitatively analyze phospholipids and ergosterol utilizing atmospheric pressure ionization-mass spectrometry methods. Lipid profiling of individual fermentations indicated that yeast lipid class profiles do not shift dramatically in composition over the course of fermentation. Multivariate statistical analysis of the data was performed using partial least-squares linear regression modeling to correlate lipid composition data with fermentation kinetic data. The results indicate a strong correlation (R² = 0.91) between the overall lipid composition and the final ethanol concentration (wt/wt), an indicator of strain ethanol tolerance. One potential component of ethanol tolerance, the maximum yeast cell concentration, was also found to be a strong function of lipid composition (R² = 0.97). Specifically, strains unable to complete fermentation were associated with high phosphatidylinositol levels early in fermentation. Yeast strains that achieved the highest cell densities and ethanol concentrations were positively correlated with phosphatidylcholine species similar to those known to decrease the perturbing effects of ethanol in model membrane systems.     

Crabtree effect in aerobic fermentations using grape juice for the production of alcohol reduced wine

Summary Aerobic fermentations of grape juice to alcohol reduced wine were carried out by technical strains of wine yeast (S. cerevisiae var. ellipsoideus) at a temperature of 25 °C and an aeration rate of 1 vvm using a two-stage batch and fed-batch process. In the fed-batch phase of each fermentation Crabtree Effect [CE] limits between 0.2 and 0.5 g glucose/L have been detected.     

Kinetic model for nitrogen-limited wine fermentations.

A physical and mathematical model for wine fermentation kinetics has been developed to predict sugar utilization curves based on experimental data from wine fermentations with various initial nitrogen and sugar concentrations in the juice. The model is based on: (1) yeast cell growth limited by nitrogen; (2) sugar utilization rates and ethanol production rates proportional solely to the number of viable cells; and (3) a death rate for cells proportional to alcohol content. All but one parameter in the model can be estimated from existing data. However, experiments to find this final parameter, a constant describing cell death, indicate that cell death may not be the critical factor in determining fermentation kinetics as cell viability remains significant until sugar utilization has ceased. The model, nevertheless, predicts a transition from normal to sluggish to stuck fermentations as initial nitrogen levels decrease. It also predicts that fermentations with high initial Brix levels may go to completion when supplemented with nitrogen in the form of ammonia. Therefore, we hypothesize that the model is valid but that ethanol causes the yeast cells to become inactive while remaining viable. Experimental verification of the model has been performed using flask-scale experiments. The model has also been used to evaluate the possibility of using nitrogen or viable cell additions to avoid or correct problem (i.e., sluggish or stuck) fermentations.     

Creation and validation of a reactor engineering model for multiphase red wine fermentations

Red wine fermentations are performed in the presence of grape skins and seeds to ensure the extraction of color and other phenolics. The presence of these solids results in two distinct phases in the fermentor, as the solids float to the top to form a “cap.” Modeling of red wine fermentation is, therefore, complex and must consider spatial heterogeneity to predict fermentation kinetics. We have developed a reactor-engineering model for red wine fermentations that includes the fundamentals of fermentation kinetics, heat transfer, diffusion, and compressible fluid flow. To develop the heat transfer component of the model, the heat transfer properties of grapes were experimentally determined as a function of fermentation progression. COMSOL was used to solve all components of the model simultaneously utilizing a finite element analysis approach. Predictions from this model were validated using prior experimental work. Model prediction and experimental data showed excellent agreement. The model was then used to predict spatial profiles of active yeast cell concentration and ethanol productivity, as well as liquid velocity profiles. Finally, the model was used to predict how these gradients would change with differences in initial bioavailable nitrogen concentration, a key parameter in predicting fermentation outcome in nitrogen-limited wine fermentations.     

Coupling kinetic expressions and metabolic networks for predicting wine fermentations

Problematic fermentations are commonplace and cause wine industry producers substantial economic losses through wasted tank capacity and low value final products. Being able to predict such fermentations would enable enologists to take preventive actions. In this study we modeled sugar uptake kinetics and coupled them to a previously developed stoichiometric model, which describes the anaerobic metabolism of Saccharomyces cerevisiae. The resulting model was used to predict normal and slow fermentations under winemaking conditions. The effects of fermentation temperature and initial nitrogen concentration were modeled through an efficiency factor incorporated into the sugar uptake expressions. The model required few initial parameters to successfully reproduce glucose, fructose, and ethanol profiles of laboratory and industrial fermentations. Glycerol and biomass profiles were successfully predicted in nitrogen rich cultures. The time normal or slow wine fermentations needed to complete the process was predicted accurately, at different temperatures. Simulations with a model representing a genetically modified yeast fermentation, reproduced qualitatively well literature results regarding the formation of minor compounds involved in wine complexity and aroma. Therefore, the model also proves useful to explore the effects of genetic modifications on fermentation profiles.     

<Emphasis Type="Italic">Starmerella bombicola</Emphasis> influences the metabolism of <Emphasis Type="Italic">Saccharomyces cerevisiae</Emphasis> at pyruvate decarboxylase and alcohol dehydrogenase level during mixed wine fermentation

Background  

The use of a multistarter fermentation process with Saccharomyces cerevisiae and non-Saccharomyces wine yeasts has been proposed to simulate natural must fermentation and to confer greater complexity and specificity to wine. In this context, the combined use of S. cerevisiae and immobilized Starmerella bombicola cells (formerly Candida stellata) was assayed to enhance glycerol concentration, reduce ethanol content and to improve the analytical composition of wine. In order to investigate yeast metabolic interaction during controlled mixed fermentation and to evaluate the influence of S. bombicola on S. cerevisiae, the gene expression and enzymatic activity of two key enzymes of the alcoholic fermentation pathway such as pyruvate decarboxylase (Pdc1) and alcohol dehydrogenase (Adh1) were studied.     

Use of Artificial Neural Networks and a Gamma-Concept-Based Approach To Model Growth of and Bacteriocin Production by Streptococcus macedonicus ACA-DC 198 under Simulated Conditions of Kasseri Cheese Production

Growth of and bacteriocin production by Streptococcus macedonicus ACA-DC 198 were assessed and modeled under conditions simulating Kasseri cheese production. Controlled fermentations were performed in milk supplemented with yeast extract at different combinations of temperature (25, 40, and 55°C), constant pH (pHs 5 and 6), and added NaCl (at concentrations of 0, 2, and 4%, wt/vol). The data obtained were used to construct two types of predictive models, namely, a modeling approach based on the gamma concept, as well as a model based on artificial neural networks (ANNs). The latter computational methods were used on 36 control fermentations to quantify the complex relationships between the conditions applied (temperature, pH, and NaCl) and population behavior and to calculate the associated biokinetic parameters, i.e., maximum specific growth and cell count decrease rates and specific bacteriocin production. The functions obtained were able to estimate these biokinetic parameters for four validation fermentation experiments and obtained good agreement between modeled and experimental values. Overall, these experiments show that both methods can be successfully used to unravel complex kinetic patterns within biological data of this kind and to predict population kinetics. Whereas ANNs yield a better correlation between experimental and predicted results, the gamma-concept-based model is more suitable for biological interpretation. Also, while the gamma-concept-based model has not been designed for modeling of other biokinetic parameters than the specific growth rate, ANNs are able to deal with any parameter of relevance, including specific bacteriocin production.     

Temperature-dependent kinetic model for nitrogen-limited wine fermentations

A physical and mathematical model for wine fermentation kinetics was adapted to include the influence of temperature, perhaps the most critical factor influencing fermentation kinetics. The model was based on flask-scale white wine fermentations at different temperatures (11 to 35 degrees C) and different initial concentrations of sugar (265 to 300 g/liter) and nitrogen (70 to 350 mg N/liter). The results show that fermentation temperature and inadequate levels of nitrogen will cause stuck or sluggish fermentations. Model parameters representing cell growth rate, sugar utilization rate, and the inactivation rate of cells in the presence of ethanol are highly temperature dependent. All other variables (yield coefficient of cell mass to utilized nitrogen, yield coefficient of ethanol to utilized sugar, Monod constant for nitrogen-limited growth, and Michaelis-Menten-type constant for sugar transport) were determined to vary insignificantly with temperature. The resulting mathematical model accurately predicts the observed wine fermentation kinetics with respect to different temperatures and different initial conditions, including data from fermentations not used for model development. This is the first wine fermentation model that accurately predicts a transition from sluggish to normal to stuck fermentations as temperature increases from 11 to 35 degrees C. Furthermore, this comprehensive model provides insight into combined effects of time, temperature, and ethanol concentration on yeast (Saccharomyces cerevisiae) activity and physiology.     

A novel methodology independent of fermentation rate for assessment of the fructophilic character of wine yeast strains

The yeast Saccharomyces cerevisiae has a fundamental role in fermenting grape juice to wine. During alcoholic fermentation its catabolic activity converts sugars (which in grape juice are a near equal ratio of glucose and fructose) and other grape compounds into ethanol, carbon dioxide and sensorily important metabolites. However, S. cerevisiae typically utilises glucose and fructose with different efficiency: glucose is preferred and is consumed at a higher rate than fructose. This results in an increasing difference between the concentrations of glucose and fructose during fermentation. In this study 20 commercially available strains were investigated to determine their relative abilities to utilise glucose and fructose. Parameters measured included fermentation duration and the kinetics of utilisation of fructose when supplied as sole carbon source or in an equimolar mix with glucose. The data were then analysed using mathematical calculations in an effort to identify fermentation attributes which were indicative of overall fructose utilisation and fermentation performance. Fermentation durations ranged from 74.6 to over 150 h, with clear differences in the degree to which glucose utilisation was preferential. Given this variability we sought to gain a more holistic indication of strain performance that was independent of fermentation rate and therefore utilized the area under the curve (AUC) of fermentation of individual or combined sugars. In this way it was possible to rank the 20 strains for their ability to consume fructose relative to glucose. Moreover, it was shown that fermentations performed in media containing fructose as sole carbon source did not predict the fructophilicity of strains in wine-like conditions (equimolar mixture of glucose and fructose). This work provides important information for programs which seek to generate strains that are faster or more reliable fermenters.     

Field-flow fractionation as analytical technique for the characterization of dry yeast: correlation with wine fermentation activity

Important oenological properties of wine depend on the winemaking yeast used in the fermentation process. There is considerable controversy about the quality of yeast, and a simple and cheap analytical methodology for quality control of yeast is needed. Gravitational field flow fractionation (GFFF) was used to characterize several commercial active dry wine yeasts from Saccharomyces cerevisiae and Saccharomyces bayanus and to assess the quality of the raw material before use. Laboratory-scale fermentations were performed using two different S. cerevisiae strains as inocula, and GFFF was used to follow the behavior of yeast cells during alcoholic fermentation. The viable/nonviable cell ratio was obtained by flow cytometry (FC) using propidium iodide as fluorescent dye. In each experiment, the amount of dry wine yeast to be used was calculated in order to provide the same quantity of viable cells. Kinetic studies of the fermentation process were performed controlling the density of the must, from 1.071 to 0.989 (20/20 density), and the total residual sugars, from 170 to 3 g/L. During the wine fermentation process, differences in the peak profiles obtained by GFFF between the two types of commercial yeasts that can be related with the unlike cell growth were observed. Moreover, the strains showed different fermentation kinetic profiles that could be correlated with the corresponding fractograms monitored by GFFF. These results allow optimism that sedimentation FFF techniques could be successfully used for quality assessment of the raw material and to predict yeast behavior during yeast-based bioprocesses such as wine production.     

Influence of Williopsis saturnus yeasts in combination with Saccharomyces cerevisiae on wine fermentation

Aim: To examine the growth and survival of Williopsis saturnus strains along with wine yeast Saccharomyces cerevisiae in grape must. Methods and Results: For this study, fermentations were performed in sterilized grape must at 18°C. Inoculum level was 5 × 10⁶ cells per ml for each yeast. The results showed that W. saturnus yeasts exhibited slight growth and survival depending on the strain, but they died off by day 5. Saccharomyces cerevisiae, however, dominated the fermentation, reaching the population of about 8 log CFU ml^?1. It was observed that ethanol formation was not affected. The concentrations of acetic acid, ethyl acetate and isoamyl acetate were found higher in mixed culture experiments compared to control fermentation. The results also revealed that higher alcohols production was unaffected in general. Conclusion: Fermentations did not form undesirable concentrations of flavour compounds, but production of higher levels of acetic acid in mixed culture fermentations may unfavour the usage of W. saturnus in wine making. Significance and Impact of the Study: This study provides information on the behaviour of W. saturnus together with S. cerevisiae during the alcoholic fermentation.     

In situ recovery of 2,3-butanediol from fermentation by liquid–liquid extraction

End-product conversion, low product concentration and large volumes of fermentation broth, the requirements for large bioreactors, in addition to the high cost involved in generating the steam required to distil fermentation products from the broth largely contributed to the decline in fermentative products. These considerations have motivated the study of organic extractants as a means to remove the product during fermentation and minimize downstream recovery. The aim of this study is to assess the practical applicability of liquid–liquid extraction in 2,3-butanediol fermentations. Eighteen organic solvents were screened to determine their biocompatibility, and bioavailability for their effects on Klebsiella pneumoniae growth. Candidate solvents at first were screened in shake flasks for toxicity to K. pneumoniae. Cell density and substrate consumption were used as measures of cell toxicity. The possibility of employing oleyl alcohol as an extraction solvent to enhance end product in 2,3-butanediol fermentation was evaluated. Fermentation was carried out at an initial glucose concentration of 80 g/l. Oleyl alcohol did not inhibit the growth of the fermentative organism. 2,3-Butanediol production increased from 17.9 g/l (in conventional fermentation) to 23.01 g/l (in extractive fermentation). Applying oleyl alcohol as the extraction solvent, about 68% of the total 2,3-butanediol produced was extracted. An erratum to this article can be found at     

Occurrence of hydrogen sulfide in wine and in fermentation: influence of yeast strain and supplementation of yeast available nitrogen

Hydrogen sulfide (H₂S) is a powerful aroma compound largely produced by yeast during fermentation. Its occurrence in wines and other fermented beverages has been associated with off-odors described as rotten egg and/or sewage. While the formation of hydrogen sulfide (H₂S) during fermentation has been extensively studied, it is the final H₂S content of wine that is actually linked to potential off-odors. Nevertheless, factors determining final H₂S content of wine have received little attention, and it is commonly assumed that high H₂S-forming fermentations will result in high final concentrations of H₂S. However, a clear relationship has never been established. In this report, we investigated the contribution of yeast strain and nitrogen addition to H₂S formation during fermentation and its consequent occurrence the resulting wines. Five commercial Saccharomyces cerevisiae wine yeast strains were used to ferment a Chardonnay juice containing 110 mg/l of YAN (yeast assimilable nitrogen), supplemented with di-ammonium phosphate (DAP) to increase YAN concentration to moderate (260 mg/l) and high (410 mg/l) levels. In contrast to the widely reported decrease in H₂S production in response to DAP addition, a non-linear relationship was found such that moderate DAP supplementation resulted in a remarkable increase in H₂S formation by each of the five wine yeasts. H₂S content of the finished wine was affected by yeast strain, YAN, and fermentation vigor. However, we did not observe a correlation between concentration of H₂S in the finished wines and H₂S produced during fermentation, with low-forming fermentations often having relatively high final H₂S and vice versa. Management of H₂S in wine through nitrogen supplementation requires knowledge of initial YAN and yeast H₂S characteristics.     

Wine yeasts for the future

International competition within the wine market, consumer demands for newer styles of wines and increasing concerns about the environmental sustainability of wine production are providing new challenges for innovation in wine fermentation. Within the total production chain, the alcoholic fermentation of grape juice by yeasts is a key process where winemakers can creatively engineer wine character and value through better yeast management and, thereby, strategically tailor wines to a changing market. This review considers the importance of yeast ecology and yeast metabolic reactions in determining wine quality, and then discusses new directions for exploiting yeasts in wine fermentation. It covers criteria for selecting and developing new commercial strains, the possibilities of using yeasts other than those in the genus of Saccharomyces, the prospects for mixed culture fermentations and explores the possibilities for high cell density, continuous fermentations.     

The use of tamarind waste to improve ethanol production from cane molasses

Tamarind wastes such as tamarind husk, pulp, seeds, fruit and the effluent generated during tartaric acid extraction were used as supplements to evaluate their effects on alcohol production from cane molasses using yeast cultures. Small amounts of these additives enhanced the rate of ethanol production in batch fermentations. Tamarind fruit increased ethanol production (9.7%, w/v) from 22.5% reducing sugars of molasses as compared to 6.5% (w/v) in control experiments lacking supplements after 72 h of fermentation. In general, the addition of tamarind supplements to the fermentation medium showed more than 40% improvement in ethanol production using higher cane molasses sugar concentrations. The direct fermentation of aqueous tamarind effluent also yielded 3.25% (w/v) ethanol, suggesting its possible use as a diluent in molasses fermentations. This is the first report, to our knowledge, in which tamarind-based waste products were used in ethanol production. Received 2 April 1998/ Accepted in revised form 13 November 1998     

The timing of diammonium phosphate supplementation of wine must affects subsequent H2S release during fermentation

Aims:  The aim of this study was to evaluate the impact of supplementation by diammonium phosphate (DAP) on hydrogen sulfide (H₂S) production, when DAP given either prior to fermentation or during the early stationary growth phase of yeast. Methods and Results:  Three contrasting Saccharomyces cerevisiae wine strains were used to ferment synthetic grape juice (GJ) containing 67 mg l^?1 of initial yeast assimilable nitrogen (YAN), supplied either as DAP or as mixture of amino acids. Sufficient DAP was added either prior to or 72 h after the initiation of fermentation to achieve a final YAN concentration of 267 mg l^?1. Supplementation prior to fermentation stimulated H₂S production. The results obtained in model solutions were validated using natural GJ. Conclusion:  The timing of DAP supplementation is critical for ensuring that fermentation proceeds without excessive release of H₂S. Significance and Impact of the Study:  This result has important implications for the wine‐making industry, because it highlights the value of determining the initial nitrogen level of a GJ. It raises awareness of the dependence of wine quality on the correct timing of DAP supplementation.     

Yeast Diversity and Persistence in Botrytis-Affected Wine Fermentations

Culture-dependent and -independent methods were used to examine the yeast diversity present in botrytis-affected (“botrytized”) wine fermentations carried out at high (~30°C) and ambient (~20°C) temperatures. Fermentations at both temperatures possessed similar populations of Saccharomyces, Hanseniaspora, Pichia, Metschnikowia, Kluyveromyces, and Candida species. However, higher populations of non-Saccharomyces yeasts persisted in ambient-temperature fermentations, with Candida and, to a lesser extent, Kluyveromyces species remaining long after the fermentation was dominated by Saccharomyces. In general, denaturing gradient gel electrophoresis profiles of yeast ribosomal DNA or rRNA amplified from the fermentation samples correlated well with the plating data. The direct molecular methods also revealed a Hanseniaspora osmophila population not identified in the plating analysis. rRNA analysis also indicated a large population (>10⁶ cells per ml) of a nonculturable Candida strain in the high-temperature fermentation. Monoculture analysis of the Candida isolate indicated an extreme fructophilic phenotype and correlated with an increased glucose/fructose ratio in fermentations containing higher populations of Candida. Analysis of wine fermentation microbial ecology by using both culture-dependent and -independent methods reveals the complexity of yeast interactions enriched during spontaneous fermentations.