360 Influence of agitation on yeast metabolism Australian Journal of Grape and Wine Research 24, 360–367, 2018

Agitation impacts fermentation performance as well as carbon and

nitrogen metabolism in Saccharomyces cerevisiae under winemaking
conditions
S. ROLLERO, S. ROBERTS, F.F. BAUER and B. DIVOL
Institute of Wine Biotechnology, Department of Viticulture and Oenology, Stellenbosch University, Matieland 7602,
South Africa
Corresponding author: Professor Benoit Divol, email [email protected]

Abstract
Background and Aims: Under industrial conditions, fermenting juice is normally not mechanically agitated; thermody-
namic forces will lead to some rotating motion of the medium. At laboratory scale, research aims to mimic industrial condi-
tions using a standardised synthetic grape juice-like medium. The level of agitation applied, however, to fermentations
differs between the studies. In this study, we investigated the influence of agitation speed on fermentation parameters,
including fermentation kinetics, yeast cell production and volatile aroma production.
Methods and Results: These parameters were evaluated during fermentations under different agitation conditions. Agita-
tion speed was found to significantly influence all the parameters monitored. In the evaluated conditions, a speed of 80 rpm
was optimal for the production of fermentative aromas (lead to the highest production) while guaranteeing the complete
and rapid fermentation and limiting the production of off-flavours. Fermentation of a Chardonnay grape juice in 50 L tanks
confirmed the results, and the panellists described the wine produced by the stirred fermentation as sweeter but it is likely
that the panellists meant softer because the wines were dry.
Conclusions: Agitation is an important parameter to control during fermentation under laboratory conditions. The data also
suggest that the type and strength of agitation in commercial fermentations may have a more significant impact than previ-
ously thought.
Significance of the Study: This study demonstrates that researchers must take into account the importance of agitation
during their studies. Similarly, winemakers might consider constant agitation of their tanks.

Keywords: agitation speed, aroma compounds, fermentation kinetics, Saccharomyces cerevisiae, winemaking

Introduction NAD(P)H (oxidized Nicotinamide Adenine Dinucleotide

Microbiological research frequently involves cell culture (Phosphate) oxidized / reduced Nicotinamide Adenine
experiments using chemically defined synthetic media in Dinucleotide (Phosphate)) (Fariña et al. 2012, Bloem
small flasks. In biotechnological research, the applicability of et al. 2015). Some studies highlighted the influence of phys-
such laboratory generated data to industrial conditions is of ical parameters such as the temperature of fermentation
primary importance (Farid et al. 2002, Ercan et al. 2013, (Sablayrolles and Barre 1993, Molina et al. 2007, Mouret
Jayakar and Singhal 2013, Saad et al. 2014, Ibrahim et al. 2014, Rollero et al. 2015) or the effect of the scale of
et al. 2015, Leite et al. 2016). In wine microbiology, it has the fermenter (Casalta et al. 2016, Liccioli et al. 2011).
already been reported that several parameters influence the Unlike other research fields, however, such as enzyme syn-
successful completion of the fermentation and the produc- thesis, bioethanol production or brewing (Farid et al. 2002,
tion of desirable flavour compounds. In particular, several Ercan et al. 2013, Ibrahim et al. 2015), investigation of the
studies focused on the impact of the grape juice constitu- effect of agitation in grape juice fermentation has been lim-
ents, such as sugars (Erasmus et al. 2004, Ferreira ited. Indeed, Casalta et al. (2010) are the only authors to
et al. 2006), nitrogenous compounds (Bely et al. 1990, have compared the outcome of a static fermentation with
Blateyron and Sablayrolles 2001, Jiménez-Martí et al. 2007, that of a stirred fermentation, but their study was limited to
Mouret et al. 2014), lipids (Lafon-Lafourcade et al. 1979, one agitation speed, however, and a few yeast-derived com-
Luparia et al. 2004, Rollero et al. 2015, Casalta et al. 2016), pounds. Agitation speed should clearly be an important
oxygen (Sablayrolles et al. 1996, Julien et al. 2000, Coetzee parameter when evaluating such impacts. It provides
et al. 2013), vitamins (Bataillon et al. 1996, Thomas and adequate mixing of the medium components within the
Surdin-Kerjan 1997, Wang et al. 2003, Bohlscheid fermenter, which may facilitate access to nutrients, while
et al. 2007) and minerals (Pereira 1988, Birch et al. 2003, improving mass and heat transfer as well as the dissolution
Ibanez et al. 2008). The redox metabolism of the yeast can and dispersion of oxygen through the culture medium.
also play a substantial role in the synthesis of aroma com- Casalta et al. (2010) observed that only half of the nitrogen
pounds, because their synthesis involves a wide range of available is consumed in static fermentation, while this
dehydrogenases, thus impacted by the balance of NAD(P)+/ nutrient is fully depleted in stirred fermentation. Recently,

doi: 10.1111/ajgw.12338
© 2018 Australian Society of Viticulture and Oenology Inc.
 17550238, 2018, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ajgw.12338 by Nat Prov Indonesia, Wiley Online Library on [03/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Rollero et al. Influence of agitation on yeast metabolism 361

de Koker (2015) also highlighted that in static fermentation magnesium sulfate heptahydrate; 1.23 g/L of calcium chlo-
neither amino acids nor ammonium were completely con- ride dehydrate; vitamins (mg/L): myo-inositol (100), cal-
sumed. Indeed, regardless of the initial amount of assimila- cium pantothenate (1), thiamine hydrochloride (0.5),
ble nitrogen, the yeasts consumed about half of the nicotinic acid (2), pyridoxine hydrochloride (2), biotin
theoretically assimilable amount. These data were in contra- (0.125), PABA.K (para-aminobenzoate acid K) (0.2), ribofla-
diction with several studies using the same fermentation vin (0.2), folic acid (0.2); trace elements (μg/L): manganese
medium (Crepin et al. 2014, Lage et al. 2014, Taillandier (II) chloride tetrahydrate (200), zinc chloride (135), iron
et al. 2014, Rollero et al. 2015). The only difference chloride (30), copper chloride (15), boric acid (5), cobalt
between these studies was the stirring in the latter studies nitrate hexahydrate (1), sodium molybdate dehydrate
versus the absence thereof (de Koker 2015). Nevertheless, (25) and potassium iodate (10).
the holistic impact of agitation on fermentation kinetics, The nitrogen sources comprised ammonium chloride
yeast cell production and metabolite production (especially and amino acids. The composition of the stock solution
the volatile compounds playing a role in the sensory proper- of amino acids and ammonium was (in g/L): tyrosine
ties of wine) has never been investigated systematically. Yet, (1.8), tryptophan (17.9), isoleucine (3.2), aspartate (4.4),
the implementation of this physical parameter could have a glutamate (12.0), arginine (37.4), leucine (4.8), threonine
major impact on the possible extrapolation of the results to (7.5), glycine (1.8), asparagine (5.3), glutamine (50.5),
the industrial scale and potentially render the results alanine (14.5), valine (4.4), methionine (3.1), phenylala-
obtained in laboratory inapplicable in industry. nine (3.7), serine (7.8), histidine (3.2), lysine (1.7), cyste-
The overall aim of the study was to evaluate the impact ine (1.3), proline (61.2) and ammonium chloride (46).
of different agitation conditions (125, 80 and 40 rpm or no To obtain 200 mg/L of assimilable nitrogen in the
agitation) and means on fermentation parameters (kinetics SM, 6.67 mL of this solution was added to 1 L of
and yeast growth) and on the synthesis of fermentative medium.
aromas. Orbital agitation will be used in this work because it Instead of adding ergosterol (yeast sterol) as Bely
is a typical agitation system in the incubators and allows a et al. (1990), the SM medium was initially supplemented
large number of fermentations to be carried out simulta- with anaerobic factors composed of phytosterols (85 451,
neously. It is also easier to measure the agitation speed in Sigma-Aldrich, St Louis, MO, USA), sterols naturally present
this way. The data provide insights into the impact of agita- in the grape juice (Le Fur et al. 1994) and Tween 80 for a
tion on fermentative metabolism, and suggest conditions final concentration of 10 mg/L. The stock solution was com-
that are most representative of large-scale industrial wine- posed of 5 g/L of phytosterols in Tween 80 and ethanol
making conditions. (1:1, v/v).
The pH of the SM was adjusted to 3.3 with potassium
Materials and methods hydroxide (Saarchem, Krugersdorp, South Africa). The trace
elements, vitamins, nitrogen sources and anaerobic factors
Yeast strains and preculture conditions were filtered through a 0.22 μm syringe filter (Starlab
The commercial wine yeast strains Saccharomyces cerevisiae
Scientific, Cape Town, South Africa) and added into the
Lalvin EC1118 (Lallemand, Montréal, QC, Canada) and
autoclaved SM.
VIN13 (Anchor Yeast, Cape Town, South Africa) were used
Each laboratory-scale fermentation was performed in
for the laboratory-scale experiments. The cryopreserved
triplicate. The fermentations were carried out in cylindrical
yeast cultures were thawed at room temperature and
streaked out on yeast peptone dextrose (YPD) agar (Biolab- fermentors of 3.5 cm diameter and 10 cm height. The fer-
Merck, Modderfontein, South Africa). Starter cultures of all mentors contained 70 mL medium, so that the headspace
yeast strains were prepared by inoculating a single colony occupied 30% of the volume of the fermentors. In order to
into 5 mL YPD broth for each strain. The cultures were incu- maintain anaerobiosis, the fermentors were equipped with
bated at 30 C with shaking on a test tube rotating wheel for fermentation locks filled with water, at 25 C, with stirring
24 h. These starter cultures were used to inoculate YPD pre- bar agitation (130 rpm), orbital agitation (125, 80 or
cultures at an initial cell density of 1 × 106 cells/mL, which 40 rpm) or without agitation. A polystyrene plate was
were incubated at 30 C with shaking (125 rpm) for 9 h. In applied between the fermentors and the stirring platform in
order to deplete the nitrogen present in the cells, the yeasts order to avoid excessive heat transfer.
were incubated for 8 h in yeast nitrogen base (YNB) con- The fermentation progress was monitored by determina-
taining neither amino acids nor ammonium (Difco Labora- tion of CO2 release extrapolated from the measurement of
tories, Cape Town, South Africa) supplemented with 20 g/L the mass loss once a day throughout the process. In Control
of glucose at 30 C with shaking (125 rpm). The nitrogen experiments (cell-free systems), resazurin was used to con-
depletion was reached when the growth stopped. For the firm anaerobiosis (not shown).
two yeasts, this appeared after 6–7 h. The pilot-scale experiments were carried out in 50 L
The commercial wine strain S. cerevisiae VIN13 was used stainless steel cylindrical tanks of 40 cm diameter and 60 cm
for the pilot-scale experiments, in which fermentation tanks height with stirring bar agitation (100 rpm) or without agi-
were inoculated with active dry yeasts previously rehydrated tation. These fermentations were undertaken in a Chardon-
as recommended by the supplier (0.2 g/L were added). nay grape juice with the following composition: total sugars
(150 g/L), TA (7.33 g/L), pH (3.17) and yeast assimilable
Fermentation conditions and sampling nitrogen (YAN) (330 mg/L) at 20 C. The fermentation pro-
Fermentations at laboratory scale were carried out in syn- gress was monitored by estimating the sugar concentration
thetic medium (SM) that simulates standard grape juice daily with a hydrometer.
(Bely et al. 1990). The SM contained 230 g/L of sugar At the end of each fermentation, different samples were
(115 g/L of glucose and 115 g/L of fructose); 2.5 g/L of centrifuged at 4000 g for 5 min, after which the superna-
potassium L-tartrate; 3 g/L of malic acid; 0.2 g/L of citric tants were filtered through a 0.22 μm syringe filter (Starlab
acid; 1.14 g/L of potassium hydrogen phosphate; 0.44 g/L of Scientific) and stored at −20 C for further chemical analysis.

© 2018 Australian Society of Viticulture and Oenology Inc.

 17550238, 2018, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ajgw.12338 by Nat Prov Indonesia, Wiley Online Library on [03/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
362 Influence of agitation on yeast metabolism Australian Journal of Grape and Wine Research 24, 360–367, 2018

The static fermentations were manually stirred just before Sensory evaluation of wines
sampling. The wines from the Chardonnay juice fermentations were
subjected to sensory analysis 5 months after bottling. A
Monitoring of yeast population panel of 32 judges (25 women and 7 men, between the age
For the laboratory-scale fermentations, the yeast cell popu- of 19 and 65) carried out triangle tests. The panel consisted
lations were monitored by plating the appropriate dilutions of staff members at the Department of Viticulture and
onto YPD nutrient agar every day during the first 3 days Oenology, Stellenbosch University. All had experience in
and then every second day. Plates were incubated at 30 C, winemaking and participated in previous studies as sensory
generally for 2 days, until colonies were formed. judges. Three wine samples were presented to the judges
identified by random three-digit codes. The order of presen-
tation was randomly assigned for each judge, verifying that
Quantification of residual sugars and ammonium
for the whole panel, presentation order of the samples was
For the concentration of residual glucose, fructose, and
balanced. Wine (25 mL) was served in black tulip-shaped
ammonium, 400 μL of filtered sample was enzymatically
ISO tasting glasses at a constant temperature of 20 C, and
analysed using the Arena 20XT (Thermo Fisher Scientific,
covered with plastic Petri dishes to allow the volatiles to
Waltham, MA, USA), which makes use of automated spec-
equilibrate in the headspace. Tests were performed in a sen-
trophotometric readings to determine the concentration of
sory laboratory complying with normal requirements, such
the various compounds. The different enzymatic assay kits
as proper light and temperature control and isolation from
are: Enzytec Fluid D-Glucose (Id-No: 5140, R-Biopharm,
noises and odours. No information about the aim of the
Darmstadt, Germany) for glucose, Enzytec Fluid D-Fructose
study or about wine samples was given to the judges prior
(Id-No: 5120, R-Biopharm) for fructose and Enzytec Fluid
to the tests. Judges were asked to evaluate samples from left
Ammonia (Id-No: 5390, R-Biopharm) for ammonium.
to right, looking for differences in aroma and taste. Judges
were informed that two samples were identical and one
Quantification of individual amino acids sample was different. They had to select the odd sample.
Amino acids were separated and quantified by HPLC (Agilent When a significant difference was detected, the judges were
1100; Agilent Technologies, Waldbronn, Germany) by pre- asked to freely note the descriptors that made the odd sam-
column derivatisation and fluorescence detection based upon ple unique.
a method previously described (Henderson and Brooks 2010)
with some modifications to the derivatisation and injection.
A Poroshell HPH-C18 column (4.6 × 150 mm, 2.7 μm particle Statistical analysis
size) (Agilent Technologies) was used following derivatisation The R software, version 3.2.3 (http://cran.r-project.org/)
of the amino acids. The amino acids were derivatised with was used for statistical analyses. Each variable was then
three different reagents (Sigma-Aldrich): iodoacetic acid for tested using a one-way ANOVA with the production of
cysteine, o-phthaldialdehyde (OPA) for primary amino acids aroma compounds as a factor to describe the diversity
and fluorenylmethyloxycarbonyl chloride for secondary between the different agitation speeds to detect a global
amino acids. Internal standards, norvaline (Sigma-Aldrich) effect at a P-value threshold of 0.05. For each parameter,
and sarcosine (Sigma-Aldrich) were spiked to each sample normality of residual distributions and homogeneity of vari-
prior to derivatisation. One millilitre of each filtered sample ance were studied using standard diagnostic graphics; no
was analysed. violation of the assumptions was detected. As the effect was
significant at a P-value threshold of 0.05, all pairwise com-
Analysis of by-products of central carbon metabolism parisons for agitation speed were tested using Tukey’s hon-
Ethanol, glycerol and succinic acid were quantified by HPLC estly significant difference (HSD) test.
with the use of an Aminex HPX-87H (300 × 7.8 mm) col-
umn (Agilent, Wilmington, DE, USA) at 55 C with 5 mmol
H2SO4 as mobile phase at a flow rate of 0.5 mL/min Results
(Eyéghé-Bickong et al. 2012). Peaks were detected and Effect of agitation speed on fermentation and population
quantified using Agilent RID and UV detectors in tandem. kinetics and nitrogen consumption
HPChemstation software (Agilent, Wilmington, DE, USA) The two S. cerevisiae strains, Lalvin EC1118 and VIN13, dis-
was used for data analysis. played a similar behaviour in response to the different agita-
tions provided (Figures 1, S1). Fermentation replicates were
Analysis of major volatile compounds highly reproducible under each condition. In all treatments,
The major volatiles (i.e. a selection of higher alcohols, ace- all the sugars were fermented, but the time necessary to
tate esters, fatty acids, fatty acid ethyl esters and acetic acid) reach dryness as well as the overall fermentation kinetics
were analysed by GC equipped with a flame ionisation were treatment dependent. The overall fermentation kinet-
detector (GC/FID) using the Agilent GC System HP 6890 ics was virtually identical for the agitation speeds of 125 and
Series (Agilent) as described by Louw et al. (2009) with 80 rpm (Figures 1, S1); the duration of fermentation was
minor modifications. Five millilitres of each of the filtered 224 and 240 h for Lalvin EC1118 and VIN13, respectively
samples were used with 100 μL of 4-methyl-2-pentanol (Tables 1,S1). The maximum cell population reached under
(internal standard, Sigma-Aldrich). Diethyl ether (1 mL) these two conditions was also identical (Tables 1,S1). The
was added to the mixture, which was then placed in an fermentations performed with an agitation speed of 40 rpm
ultrasonic bath for 5 min to extract the volatile compounds. or without agitation ended at the same time (312 h for both
Thereafter, the samples were centrifuged at 4000 g for strains, Tables 1,S1), but the kinetic profiles were slightly
3 min. Sodium sulfate was added to remove any water from different (Figures 1, S1). Indeed, the fermentation without
the non-polar layer. HPChemstation software (Agilent) was agitation appeared to be slower during the major part of the
used for data analysis. process but ultimately ended the fermentation at the same

© 2018 Australian Society of Viticulture and Oenology Inc.

 17550238, 2018, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ajgw.12338 by Nat Prov Indonesia, Wiley Online Library on [03/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Rollero et al. Influence of agitation on yeast metabolism 363

central carbon metabolism (acetic acid, acetoin and ethyl

acetate), those arising from the catabolism of amino acids
(via the Ehrlich pathway) and of sugars (higher alcohols,
fusel acids and acetate esters) and those produced through
lipid metabolism [short- and medium-chain fatty acids
(MCFA) and their corresponding esters]. As observed for
the fermentation kinetics, both strains displayed a similar
response to the changes of agitation speed (Figures 2, S2).
The production of volatile compounds, however, synthe-
sised through the same metabolic pathway was similarly
impacted by the agitation speed. Although no significant
difference in the production of short and MCFAs
(e.g. 1.08 mg/L of hexanoic acid) and their corresponding
esters (e.g. 0.60 mg/L of ethyl hexanoate) was observed
Figure 1. Fermentation kinetics for VIN13 with an orbital agitation of under all conditions tested (data not shown), fermentative
125 ( ), 80 ( ), 40 rpm ( ), no agitation ( ) or with a stirring bar ( ). aromas produced through the Ehrlich pathway or via cen-
tral carbon metabolism were greatly affected by the agita-
tion speed (Figures 2, S2). The concentration of higher
time as the 40 rpm fermentation. No difference in maxi- alcohols (Figures 2a,S2a), of acetate esters (Figures 2b,d,
mum cell population was observed (Tables 1,S1). S2b,d) and of fusel acids (including acetic acid)
As reported in the literature (Casalta et al. 2016, Rollero (Figures 2c,e,S2c,e) displayed the same pattern (from
et al. 2015), agitation is often provided by the use of a stir- higher to lower concentration): 80 rpm (e.g. 132.95 mg/L
ring bar in laboratory-scale experiments. In order to investi- of isoamyl alcohol), 125 rpm (e.g. 82.05 mg/L), 40 rpm
gate the impact of the type of agitation on fermentation (e.g. 55.39 mg/L) and no agitation (e.g. 27.56 mg/L).
performance, fermentations were also carried out using a When agitation was driven by a stirring bar, the highest
bar stirring at a speed of 130 rpm. The use of a stirring bar production of higher alcohols (e.g. 200.52 mg/L of isoamyl
significantly decreased the duration of fermentation com- alcohol; Figures 2a,S2a) and acetic acid (1.73 g/L of acetic
pared to that of the orbital agitation of 125 rpm (from acid; Figures 2e,S2e) of all treatments was obtained. Acetoin,
224 or 240 h to 144 h, for Lalvin EC1118 and VIN13, however, reached one of its lowest concentration
respectively) (Figures 1, Tables 1,S1). No difference in the (Figures 2d,S2d) with this type of agitation (3.57 mg/L with a
maximum cell population, however, was observed between stirring bar compared to 13.35 mg/L at 125 rpm). The pro-
the two types of agitation (Tables 1,S1). duction of the acetate esters (Figures 2b,d,S2b,d) and the
Nitrogen consumption was also impacted by the agitation fusel acids (Figures 2c,S2c) was similar between the agitation
speed used (Tables 2,S2,S3). Indeed, the YAN – determined with a stirring bar and the orbital agitation of 80 rpm with
by measuring the concentration of ammonium and amino the stirring bar (e.g. 0.53 and 0.51 mg/L of phenylethyl ace-
acids remaining in the medium by enzymatic kit and HPLC, tate, 1.84 and 1.81 mg/L of propionic acid for 80 rpm and
respectively – was totally consumed in the fermentations per- stirring bar, respectively).
formed with the stirring bar and with an agitation of 125 and
80 rpm. Neither amino acids nor ammonium, however, were
Validation in grape juice
depleted at the end of fermentation for an agitation of In order to assess whether the observations made in syn-
40 rpm or without any agitation (Tables 2,S2,S3). thetic grape juice-like medium were comparable with those
in a real grape juice fermentation, a Chardonnay must was
Effect of agitation speed on the production of fermentative fermented in 50 L tanks with a 100 rpm agitation (stirring
aromas bar) or without any agitation. As observed in synthetic
First, the effect of agitation on the production of the non- grape juice-like medium, the stirred fermentation completed
volatile compounds arising from central carbon metabolism 5 days earlier than that of the non-stirred (Table 3). Simi-
(i.e. ethanol, glycerol and succinic acid) was assessed. No larly, only half of the YAN was consumed in the non-
significant difference was observed between the different agitated fermentation, whereas it was fully depleted in the
agitation speeds (data not shown). agitated tank (Tables 3,S3). The production of the major vol-
The impact of agitation speed on the synthesis of vola- atile compounds also described the same trend as that
tile compounds was also investigated. Molecules with observed in SM (Table 4) with an increase in the production
diverse origins were quantified: those produced from of higher alcohol (e.g. 8.44 and 31.86 mg/L of isobutanol

Table 1. Duration of fermentation and maximal population with VIN13.

Table 2. Nitrogen assimilation by VIN13.
Agitation Duration of Maximal
fermentation (h) population Agitation YAN in must Nitrogen
(106 cells/mL) (mg/L) assimilated by
yeasts (%)
125 rpm 240a 115  8.8a
80 rpm 240a 112  9.7a 125 rpm 200 100
40 rpm 312b 89  5.2b 80 rpm 100
No agitation 312b 87  3.3b 40 rpm 60
Stirring bar 146c 114  8.6a No agitation 61
Stirring bar 0
Mean values  SD. Agitation speeds sharing the same letter are not signifi-
cantly different at a P < 0.05 threshold. YAN, yeast assimilable nitrogen.

© 2018 Australian Society of Viticulture and Oenology Inc.

 17550238, 2018, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ajgw.12338 by Nat Prov Indonesia, Wiley Online Library on [03/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
364 Influence of agitation on yeast metabolism Australian Journal of Grape and Wine Research 24, 360–367, 2018

In order to verify that the observed differences in the

chemical composition of the wines had an impact on their
sensory properties, a sensory analysis by a triangle test was
carried out. The triangle test was divided into different parts.
First, the panellists had to differentiate the two wines by
their aroma and then by the taste. The glasses were changed
between the two identifications. If they wished, the panel-
lists could attribute one or more descriptors to the wines.
According to the Roessler statistical tables (Roessler
et al. 1978), the results of the triangle test indicated that the
two wines were perceived differently (P < 0.001) both for
aroma and taste (23 judges out of the 32 found the odd
samples). Most of the judges described the taste of the stir-
red wine as being sweeter and softer than that of the non-
stirred wine.

Discussion
Most research studies focusing on yeast fermentations are
carried out at laboratory scale with continuous agitation of
the fermentation medium (Mendes-Ferreira et al. 2004,
Garde-Cerdán and Ancín-Azpilicueta 2008, Barbosa
et al. 2012, Rollero et al. 2015). In industrial winemaking
fermentation, however, the grape juice is not normally
mechanically agitated. Nevertheless, the release of CO2
combined with the presence of solid particles can provide
nucleation site for the CO2 bubbles and create a natural
motion within the tank (Siebert et al. 1986, Casalta
et al. 2016). In this context, the objective of this work was
to determine which agitation speed was realistic to mimic
the industrial process not only in terms of fermentation
kinetics but also of aroma production. The study also aimed
to assess the impact of agitation speed on the consumption
of nutrients and their metabolism in S. cerevisiae.
A significant difference in fermentation kinetics, yeast
population and nitrogen consumption was observed
between agitation speeds. Slower fermentations associated
with lower maximum cell population and partial nitrogen
consumption. Although the observations are consistent with
previous studies, this was the first time that a low agitation
speed was included in the evaluation. Indeed, de Koker
(2015) showed that regardless of the amount of YAN ini-
tially present in the medium only half was consumed in the
non-agitated fermentation compared to that in a fermenta-
Figure 2. Production of volatile compounds by yeast VIN13 with an orbital tion with an agitation of 150 rpm. Casalta et al. (2016) also
agitation of 125 ( ), 80 ( ), 40 rpm ( ), no agitation ( ) or with a stirring found a lower fermentation rate for the non-agitated fer-
bar ( ). Agitation speeds sharing the same letter are not significantly mentation. Indeed, agitation induces an adequate mixing of
different at a 0.05 threshold. the nutrients present in the medium as well as a better dis-
persion of the cells thus facilitating the assimilation of nutri-
Table 3. Duration of fermentation and yeast assimilable nitrogen con- ents by the yeasts (Ibrahim et al. 2015). Differences in
sumed in Chardonnay fermentations.
fermentation rate and cell formation were probably due to
Duration Proportion of differences in nitrogen consumption. When no agitation or
(days) YAN consumed (%) a too low agitation speed (40 rpm) was provided, the yeasts
were not completely dispersed in the medium and tended to
Stirred fermentation 4 100
settle at the bottom of the flasks (not shown). Subsequently,
Non-stirred 9 51
fermentation they most probably did not access all the nitrogen available
in the medium (Figures 1, Tables 1,S1). At stronger agitation
YAN, yeast assimilable nitrogen. (i.e. 80 or 125 rpm), the YAN was fully depleted resulting in
higher fermentation rate and maximal population formed,
produced during static or stirred fermentation, respectively), consistently with results reported in the literature (Crepin
acetate esters (e.g. 13.60 and 42.66 mg/L of ethyl acetate), et al. 2014, Lage et al. 2014, Mouret et al. 2014, Rollero
acetoin (0.93 and 18.06 mg/L) and acetic acid (431.48 and et al. 2015). Interestingly, while the speed of agitation was
879.95 mg/L) during the stirred fermentation, while the similar to the most vigorous orbital agitation, the fermenta-
MCFA and MCFA ethyl esters were produced at the same tion rate was even higher when a stirring bar was used
concentration in the two conditions (e.g. 1.65 mg/L of octa- instead of an orbital agitation (Figures 1, S1, Tables 1,S1).
noic acid and 1.35 mg/L of ethyl octanoate). Thus, the type of agitation appeared to impact the

© 2018 Australian Society of Viticulture and Oenology Inc.

 17550238, 2018, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ajgw.12338 by Nat Prov Indonesia, Wiley Online Library on [03/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Rollero et al. Influence of agitation on yeast metabolism 365

Table 4. Major volatile compound produced in Chardonnay fermentations. were used by a greater number of cells, while during static
and slowest agitated fermentations a lower number of cells
Concentration (mg/L)
was present and needed less lipids. Thus, no impact on the
Non-stirred Stirred lipid metabolism was observed.
fermentation fermentation The synthesis of higher alcohols, fusel acids and acetate
esters is directly related to nitrogen assimilation. Certain
Higher alcohols
amino acids are the direct precursors of these volatile com-
Propanol 23.80 59.40
Isobutanol 8.44 31.86 pounds via the Ehrlich pathway. It was therefore not sur-
Isoamyl alcohol 44.53 105.53 prising that their production was directly affected by
Phenylethanol 7.60 33.98 agitation, with the latter impacting the availability of assimi-
Acetate esters lable nitrogen. Overall, the more agitated fermentations pro-
Ethyl acetate 13.60 42.66
Isoamyl acetate 0.28 0.83 duced more higher alcohols as well as acetate esters and
Phenylethyl acetate 0.34 1.11 fusel acids than that of the slowest agitation or with no agi-
Fusel acids tation at all. These observations were consistent with the
Propionic acid 0.62 1.46 results found by de Koker (2015) and Casalta et al. (2016),
Isobutyric acid 0.32 1.37
but this study we deepened this aspect by comparing differ-
Isovaleric acid 0.42 1.06
MCFA ethyl esters ent agitation speeds. Higher production of aromatic higher
Ethyl hexanoate 0.47 0.49 alcohols was also previously observed in the context of quo-
Ethyl octanoate 1.37 1.34 rum sensing (Zupan et al. 2013).
Ethyl decanoate 0.40 0.41 Surprisingly, differences in the concentration of these
MCFA
Hexanoic acid 2.06 2.08 molecules have also been observed between the three fer-
Octanoic acid 1.64 1.66 mentations where all the nitrogen was consumed (80 and
Decanoic acid 0.95 0.95 125 rpm, stirring bar). An agitation of 80 rpm led to the
Carbon compounds maximal production of Ehrlich pathway-derived com-
Acetoin 0.93 18.06
pounds compared to agitation at 125 rpm or provided by a
Acetic acid 431.48 879.95
stirring bar. As the quantity of nitrogen consumed and the
maximal population produced were the same under these
MCFA, medium-chain fatty acids.
conditions, it can be assumed that this difference was not
due to a metabolic effect but perhaps to a physical one.
fermentation metabolism of yeasts. One possible hypothesis Several authors have previously demonstrated that the
to explain this phenomenon is that the stirring bar agitation increase of the agitation speed creates shear forces, which
could improve the dissolution of oxygen present in the have a strong impact on the cells (Darah et al. 2011, Ibra-
headspace at the beginning of fermentation in the culture him et al. 2015). In the latter study, the shear forces
medium and therefore promote the overall yeast metabo- caused morphological changes and eventually damage to
lism (Julien et al. 2000, Blateyron and Sablayrolles 2001). the intra- and extracellular structures of the cells. This
Moreover, even though the fermentors were insulated from stress may lead to a lower production of enzymes by the
the stirring plate heat, the temperature in the flasks was not cells (Ibrahim et al. 2015). It can therefore be assumed
monitored; an increase of the temperature, even transient, that, beyond a certain speed threshold, the production of
could also be responsible for the increase of the the enzymes responsible for the synthesis of aroma com-
fermentation rate. pounds derived from the Ehrlich pathway was lower,
Depending on the metabolic origins of the major volatile thereby leading to decreased concentration. As Darah
compounds, the agitation did not yield the same outcome et al. (2011) showed, the agitation speed threshold is
with regard to their production. The synthesis of MCFAs as strain- and enzyme-dependent, which could explain why
well as their corresponding ethyl esters was affected neither only certain compounds were affected.
by the different speeds nor by the type of agitation. Casalta Interestingly, the production of acetic acid and acetoin
et al. (2016) reported previously the same observation but was also affected by stirring. The production of these com-
only for the ethyl octanoate. Our results made it possible to pounds is known to be closely linked with redox balance.
generalise this observation to all MCFAs and their corre- During the fermentations stirred at 125 and 80 rpm, the
sponding esters, and also to a wide range of agitation speed. yeast population obtained was higher, which can impact the
de Koker (2015) also found that there was no impact of agi- demand in the different redox cofactors. This can play a role
tation on the ethyl ester production, but a significant differ- in the redox balance and maybe explain, at least partially,
ence in the synthesis of MCFAs was observed. The different the difference of production of these compounds. Fariña
sterols provided in the medium between this study (phytos- et al. (2012) have previously shown that oxygen plays an
terols) and the latter study (ergosterol) may explain a mod- important role in redox balancing and impacts aroma pro-
erately different behaviour for the compounds deriving duction, including MCFAs and their esters. These authors,
from the lipid metabolism. As observed for the nitrogen however, compared fermentations conducted in strict reduc-
sources, it is likely that not all the lipids provided were con- tive condition versus microaerobiosis, while in our study
sumed during static and slowest agitated fermentations. The there was no difference in terms of oxygen exposure
maximal population reached under these conditions, how- between the different conditions of agitation (i.e. self-
ever, was lower than that reached with higher agitation. induced anaerobiosis), which explains why we did not
The amount of lipids needed to fulfil the requirement of observe any difference in the concentration of the MCFAs
the yeasts is therefore less important. Thus, the availability and their esters. More recently, Bloem et al. (2015) have
of lipids per cell was most likely similar under all the condi- also demonstrated the crucial impact of redox balance on
tions of agitation speed tested. Indeed, in conditions with the production of fermentative aromas by increasing the
high agitation speed, more lipids were available but they demand in reduced cofactors via the addition of acetoin.

© 2018 Australian Society of Viticulture and Oenology Inc.

 17550238, 2018, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ajgw.12338 by Nat Prov Indonesia, Wiley Online Library on [03/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
366 Influence of agitation on yeast metabolism Australian Journal of Grape and Wine Research 24, 360–367, 2018

All the observations highlighted at laboratory scale were Bataillon, M., Rico, A., Sablayrolles, J.-M., Salmon, J.-M. and
also found during the pilot-scale Chardonnay fermentation. Barre, P. (1996) Early thiamin assimilation by yeasts under eno-
logical conditions: impact on alcoholic fermentation kinetics. Jour-
The greatest production of acetoin in the stirred fermenta- nal of Fermentation and Bioengineering 82, 145–150.
tion may be the explanation for the perception of sweetness Bely, M., Sablayrolles, J. and Barre, P. (1990) Automatic detection
and roundness. Indeed, acetoin was associated with these of assimilable nitrogen deficiencies during alcoholic fermentation
descriptors according to the Handbook of flavour ingredients in enological conditions. Journal of Fermentation and Bioengi-
(Burdock 2004). The perceived sweetness cannot come from neering 70, 246–252.
Birch, R.M., Ciani, M. and Walker, G.M. (2003) Magnesium, cal-
residual sugars as both wines were dry (<1 g/L of residual cium and fermentative metabolism in wine yeasts. Journal of
sugars). Moreover, the overall difference in chemical com- Wine Research 14, 3–15.
pounds appears to translate into sensory differences, thus Blateyron, L. and Sablayrolles, J.M. (2001) Stuck and slow fermen-
reinforcing the idea that agitation should be considered tations in enology: statistical study of causes and effectiveness of
combined additions of oxygen and diammonium phosphate. Jour-
carefully when doing small-scale fermentation in the nal of Bioscience and Bioengineering 91, 184–189.
laboratory. Bloem, A., Sanchez, I., Dequin, S. and Camarasa, C. (2015) Meta-
bolic impact of redox cofactor perturbations on the formation of
aroma compounds in Saccharomyces cerevisiae. Applied and Envi-
Conclusion ronmental Microbiology 82, 174–183.
This study illustrates the effect of agitation on alcoholic fer- Bohlscheid, J.C., Fellman, J.K., Wang, X.D., Ansen, D. and
mentation under winemaking conditions. Significant differ- Edwards, C.G. (2007) The influence of nitrogen and biotin interac-
ences were found, both in kinetics and flavour production, tions on the performance of Saccharomyces in alcoholic fermenta-
tions. Journal of Applied Microbiology 102, 390–400.
between the different speeds of agitation. Even if no agita- Burdock, G.A. (2004) Fenaroli’s handbook of flavor ingredients, 5th
tion is enforced during industrial alcoholic fermentation, the edn (CRC Press: Boca Raton, FL, USA).
solid particles present in grape juice combined with the bub- Casalta, E., Vernhet, A., Sablayrolles, J.-M., Tesnière, C. and
bles of CO2 induce a gentle agitation within the tank. Agita- Salmon, J.-M. (2016) Review: characterization and role of grape
tion speed is therefore an important factor that needs to be solids during alcoholic fermentation under enological conditions.
American Journal of Enology and Viticulture 67, 133–138.
controlled during research experiments under winemaking Casalta, E., Aguera, E., Picou, C., Rodriguez-Bencomo, J.-J.,
conditions in order to mimic accurately the industrial pro- Salmon, J.-M. and Sablayrolles, J.-M. (2010) A comparison of lab-
cess. A gentle orbital agitation (between 40 and 80 rpm) oratory and pilot-scale fermentations in winemaking conditions.
might approach the natural motion observed within indus- Applied Microbiology and Biotechnology 87, 1665–1673.
Coetzee, C., Lisjak, K., Nicolau, L., Kilmartin, P. and du, Toit, W.J.
trial tanks but this agitation will also depend on the level of (2013) Oxygen and sulfur dioxide additions to Sauvignon blanc
clarification of the grape juice. In future work, it would be must: effect on must and wine composition: oxygen and sulfur
interesting to measure the natural motion created by the dioxide: effect on Sauvignon blanc. Flavour and Fragrance Journal
CO2 released during fermentation to compare with labora- 28, 155–167.
tory conditions and to asses if this motion is more similar to Crepin, L., Sanchez, I., Nidelet, T., Dequin, S. and Camarasa, C.
(2014) Efficient ammonium uptake and mobilization of vacuolar
orbital shaking or stirring with a bar. arginine by Saccharomyces cerevisiae wine strains during wine fer-
Moreover, this study suggests that agitated fermentations mentation. Microbial Cell Factories 13, 109.
could produce wines with more complexity and fruity notes Darah, I., Sumathi, G., Jain, K. and Lim, S.H. (2011) Influence of
but also a softer palate combined with faster fermentation. It agitation speed on tannase production and morphology of Aspergil-
lus niger FETL FT3 in submerged fermentation. Applied Biochemis-
might therefore be interesting for winemakers to consider try and Biotechnology 165, 1682–1690.
introducing a constant agitation in their tanks. The roto-fer- Erasmus, D.J., Cliff, M. and van, Vuuren, H.J. (2004) Impact of
mentors, fully automated horizontal spinning tanks, are yeast strain on the production of acetic acid, glycerol, and the sen-
already used during red winemaking to favour the extraction sory attributes of icewine. American Journal of Enology and Viti-
of the compounds responsible for the wine colour. They could culture 55, 371–378.
Ercan, Y., Irfan, T. and Mustafa, K. (2013) Optimization of ethanol
therefore be used more widely to diversify wine styles. The production from carob pod extract using immobilized Saccharomy-
data might also provide an explanation and support for anec- ces cerevisiae cells in a stirred tank bioreactor. Bioresource Technol-
dotal evidence that the shape of fermentation tanks impacts ogy 135, 365–371.
aroma production. This impact might indeed be linked to the Eyéghé-Bickong, H.A., Alexandersson, E.O., Gouws, L.M.,
Young, P.R. and Vivier, M.A. (2012) Optimisation of an HPLC
type and strength of the motion that can be generated within method for the simultaneous quantification of the major sugars
a tank depending on its shape. Nevertheless, further studies and organic acids in grapevine berries. Journal of Chromatography
are necessary in order to find the balance between a sufficient B 885–886, 43–49.
agitation speed to impact aroma production but not too great Farid, M.A., El-Enshasy, H.A. and Noor El-Deen, A.M. (2002) Alco-
to avoid the deleterious effects of shear forces. hol production from starch by mixed cultures of Aspergillus awa-
mori and immobilized Saccharomyces cerevisiae at different agitation
speeds. Journal of Basic Microbiology 42, 162–171.
Fariña, L., Medina, K., Urruty, M., Boido, E., Dellacassa, E. and
Acknowledgements Carrau, F. (2012) Redox effect on volatile compound formation in
The authors thank Stellenbosch University and Winetech wine during fermentation by Saccharomyces cerevisiae. Food Chem-
for funding this project. They also thank Ms Jeanne Brand istry 134, 933–939.
for her technical assistance with the sensory analysis. The Ferreira, J., Toit, M.D. and Toit, W.j.D. (2006) The effects of copper
pilot-scale tanks used in this study were specially engineered and high sugar concentrations on growth, fermentation efficiency
and volatile acidity production of different commercial wine yeast
and sponsored by the Institute for Grape and Wine Sciences, strains. Australian Journal of Grape and Wine Research 12, 50–56.
Stellenbosch University. Garde-Cerdán, T. and Ancín-Azpilicueta, C. (2008) Effect of the addi-
tion of different quantities of amino acids to nitrogen-deficient must
on the formation of esters, alcohols, and acids during wine alcoholic
References fermentation. LWT: Food Science and Technology 41, 501–510.
Barbosa, C., Mendes-Faia, A. and Mendes-Ferreira, A. (2012) The Henderson, J.W.J. and Brooks, A. (2010) Improved amino acid
nitrogen source impacts major volatile compounds released by Sac- methods using Agilent ZORBAX Eclipse Plus C18 columns for a vari-
charomyces cerevisiae during alcoholic fermentation. International ety of Agilent LC instrumentation and separation goals. Application
Journal of Food Microbiology 160, 87–93. note, 5990-4547EN (Agilent Technologies: Santa Clara, CA, USA).

© 2018 Australian Society of Viticulture and Oenology Inc.

 17550238, 2018, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ajgw.12338 by Nat Prov Indonesia, Wiley Online Library on [03/05/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Rollero et al. Influence of agitation on yeast metabolism 367

Ibanez, J.G., Carreon-Alvarez, A., Barcena-Soto, M. and Casillas, N. Rollero, S., Bloem, A., Camarasa, C., Sanchez, I., Ortiz-Julien, A.,
(2008) Metals in alcoholic beverages: a review of sources, effects, Sablayrolles, J.-M., Dequin, S. and Mouret, J.-R. (2015) Combined
concentrations, removal, speciation, and analysis. Journal of Food effects of nutrients and temperature on the production of fermenta-
Composition and Analysis 21, 672–683. tive aromas by Saccharomyces cerevisiae during wine fermentation.
Ibrahim, D., Weloosamy, H. and Lim, S.-H. (2015) Effect of agita- Applied Microbiology and Biotechnology 99, 2291–2304.
tion speed on the morphology of Aspergillus niger HFD5A-1 hyphae Saad, N., Abdeshahian, P., Kalil, M.S., Yusoff, W.M.W. and
and its pectinase production in submerged fermentation. World Hamid, A.A. (2014) Optimization of aeration and agitation rate for
Journal of Biological Chemistry 6, 265–271. lipid and gamma linolenic acid production by Cunninghamella bai-
Jayakar, S.S. and Singhal, R.S. (2013) Kinetic modeling and scale nieri 2A1 in submerged fermentation using response surface meth-
up of lipoic acid (LA) production from Saccharomyces cerevisiae in a odology. Scientific World Journal 2014, 280146.
stirred tank bioreactor. Bioprocess and Biosystems Engineering 36, Sablayrolles, J. and Barre, P. (1993) Kinetics of alcoholic fermenta-
1063–1070. tion under anisothermal conditions. 2. Prediction from the kinetics
Jiménez-Martí, E., Aranda, A., Mendes-Ferreira, A., Mendes- under isothermal conditions. American Journal of Enology and
Faia, A. and li del Olmo, M. (2007) The nature of the nitrogen Viticulture 44, 134–138.
source added to nitrogen depleted vinifications conducted by a Sablayrolles, J.-M., Dubois, C., Manginot, C., Roustan, J.-L. and
Saccharomyces cerevisiae strain in synthetic must affects gene expres- Barre, P. (1996) Effectiveness of combined ammoniacal nitrogen and
sion and the levels of several volatile compounds. Antonie Van oxygen additions for completion of sluggish and stuck wine fermen-
Leeuwenhoek 92, 61–75.
tations. Journal of Fermentation and Bioengineering 82, 377–381.
Julien, A., Roustan, J.-L., Dulau, L. and Sablayrolles, J.-M. (2000)
Siebert, K.J., Blum, P.H., Wisk, T.J., Stenroos, L.E. and Anklam, W.
Comparison of nitrogen and oxygen demands of enological yeasts: J. (1986) The effect of trub on fermentation. MBAA Technical
technological consequences. American Journal of Enology and Quarterly 23, 37–43.
Viticulture 51, 215–222.
Taillandier, P., Lai, Q.P., Julien-Ortiz, A. and Brandam, C. (2014) Inter-
de Koker, S. (2015) Nitrogen utilisation of selected non-Saccharomyces
actions between Torulaspora delbrueckii and Saccharomyces cerevisiae in
yeasts and the impact on volatile compound production. MSc Thesis, wine fermentation: influence of inoculation and nitrogen content.
Stellenbosch University, Matieland, South Africa. World Journal of Microbiology and Biotechnology 30, 1959–1967.
Lafon-Lafourcade, S., Larue, F. and Ribéreau-Gayon, P. (1979) Evi-
Thomas, D. and Surdin-Kerjan, Y. (1997) Metabolism of sulfur
dence for the existence of “survival factors” as an explanation for
amino acids in Saccharomyces cerevisiae. Microbiology and Molecular
some peculiarities of yeast growth, especially in grape must of high Biology Reviews 61, 503–532.
sugar concentration. Applied and Environmental Microbiology 38, Wang, X.D., Bohlscheid, J.C. and Edwards, C.G. (2003) Fermenta-
1069–1073.
tive activity and production of volatile compounds by Saccharomy-
Lage, P., Barbosa, C., Mateus, B., Vasconcelos, I., Mendes-Faia, A.
ces grown in synthetic grape juice media deficient in assimilable
and Mendes-Ferreira, A. (2014) H. guilliermondii impacts growth nitrogen and/or pantothenic acid. Journal of Applied Microbiology
kinetics and metabolic activity of S. cerevisiae: the role of initial 94, 349–359.
nitrogen concentration. International Journal of Food Microbiol-
Zupan, J., Avbelj, M., Butinar, B., Kosel, J., Sergan, M. and
ogy 172, 62–69.
Raspor, P. (2013) Monitoring of quorum-sensing molecules during
Le Fur, Y., Hory, C., Bard, M.-H. and Olsson, A. (1994) Evolution of minifermentation studies in wine yeast. Journal of Agricultural
phytosterols in Chardonnay grape berry skins during last stages of and Food Chemistry 61, 2496–2505.
ripening. Vitis 33, 127–131.
Leite, L.G., Shapiro-Ilan, D.I., Hazir, S. and Jackson, M.A. (2016)
The effects of nutrient concentration, addition of thickeners, and Manuscript received: 5 June 2017
agitation speed on liquid fermentation of Steinernema feltiae. Jour-
nal of Nematology 48, 126–133. Revised manuscript received: 17 November 2017
Liccioli, T., Tran, T.M.T., Cozzolino, D., Jiranek, V., Chambers, P.J. Accepted: 22 November 2017
and Schmidt, S.A. (2011) Microvinification—how small can we
go? Applied Microbiology and Biotechnology 89, 1621–1628.
Louw, L., Roux, K., Tredoux, A., Tomic, O., Naes, T., Nieuwoudt, H.
H. and van, Rensburg, P. (2009) Characterization of selected Supporting information
South African young cultivar wines using FTMIR spectroscopy, gas Additional supporting information may be found in the
chromatography, and multivariate data analysis. Journal of Agri- online version of this article at the publisher’s website:http://
cultural and Food Chemistry 57, 2623–2632.
onlinelibrary.wiley.com/doi/10.1111/ajgw.12338/abstract
Luparia, V., Soubeyrand, V., Berges, T., Julien, A. and Salmon, J.-
M. (2004) Assimilation of grape phytosterols by Saccharomyces cere-
visiae and their impact on enological fermentations. Applied Table S1. Duration of fermentation and maximal popula-
Microbiology and Biotechnology 65, 25–32. tion with Saccharomyces cerevisiae strain Lalvin EC1118.
Mendes-Ferreira, A., Mendes-Faia, A. and Leao, C. (2004) Growth
and fermentation patterns of Saccharomyces cerevisiae under differ- Table S2. Nitrogen assimilation by Saccharomyces cerevisiae
ent ammonium concentrations and its implications in winemaking strain Lalvin EC1118.
industry. Journal of Applied Microbiology 97, 540–545.
Molina, A.M., Swiegers, J.H., Varela, C., Pretorius, I.S. and Table S3. Residual nitrogen in the 40 rpm-stirred and static
Agosin, E. (2007) Influence of wine fermentation temperature on fermentations of a synthetic medium and Chardonnay juice
the synthesis of yeast-derived volatile aroma compounds. Applied
Microbiology and Biotechnology 77, 675–687. by two Saccharomyces cerevisiae strains.
Mouret, J.R., Camarasa, C., Angenieux, M., Aguera, E., Perez, M.,
Figure S1. Fermentation kinetics for Saccharomyces cerevisiae
Farines, V. and Sablayrolles, J.M. (2014) Kinetic analysis and gas–
liquid balances of the production of fermentative aromas during strain Lalvin EC1118 with an orbital agitation of 125 ( ),
winemaking fermentations: effect of assimilable nitrogen and tem- 80 ( ), 40 rpm ( ), no agitation ( ) or with a stirring
perature. Food Research International 62, 1–10. bar ( ).
Pereira, C.F. (1988) The importance of metallic elements in wine. A
literature survey. Zeitschrift für Lebensmittel-Untersuchung und Figure S2. Production of volatile compounds by Saccharomy-
Forschung 186, 295–300. ces cerevisiae strain Lalvin EC1118 with an orbital agitation of
Roessler, E.B., Pangborn, R.M., Sidel, J.L. and Stone, H. (1978)
125 ( ), 80 ( ), 40 rpm ( ), no agitation ( ) or with a stirring
Expanded statistical tables for estimating significance in paired—
preference, paired–difference, duo–trio and triangle tests. Journal bar ( ). Agitation speeds sharing the same letter are not sig-
of Food Science 43, 940–943. nificantly different at a P < 0.05 threshold.

© 2018 Australian Society of Viticulture and Oenology Inc.