Malic acid Fermentation

Malolactic Fermentation
©Copyright Ben Rotter 2002-2009
www.brsquared.org/wine

1 Introduction
Malolactic fermentation (MLF, or "malo") is an important
winemaking process conducted on most red grape wines and some
white grape wines. It is also used with some fruit wines. The
following article gives information concerning the conditions
necessary for MLF, its affects, prevention, progress, suitable wine
type candidates for it and yeast compatibility.

2 Malolactic fermentation bacteria

MLF is conducted by lactic acid bacteria (LAB), of the genera
Lactobacillus, Oenococcus, Pediococcus, and Leuconostoc.
Not all LAB are desirable for MLF.
Oenococcus oeni (formerly Leuconostoc
oenos) is the most beneficial, and probably
the most frequently occurring species of
LAB in wine. Species associated with wine
spoilage are generally members of
Lactobacillus and Pediococcus genera. The
Lactobacillus genus, for example, can
cause acescence (excessive acetic acid) by
metabolising sugar or tartaric acid [Radler
and Yannissis, 1972]. Many LAB
metabolise pentoses, tartaric acid and
glycerol. The term "malolactic
fermentation bacteria" (MLB) is commonly
used to refer to those LAB strains which
are more desirable for MLF. They are more
resistant to low pHs such as those in wine
(and in which other LAB find it more
difficult to live) and they prefer to
metabolise malic acid over sugars and
citric acid (and they do not metabolise
tartaric acid or glycerol). Oenococcus oeni,
a desirable strain, may also metabolise
glucose to produce carbon dioxide, lactic
acid, acetic acid and ethanol (it follows the
heterolactic pathway (6-PG/PK pathway))
[Garvie, 1986; Axelsson, 1993; Henick-
Kling, 1993] but will degrade malic acid
before degrading any glucose present
under non-growing conditions.

The differentiation between LAB types is

based the basis of sugar metabolism, cell
shape and physiological features.
Homofermentary strains are defined as
those which transform sugar exclusively to
lactic acid, whereas heterofermentary
strains transform sugars into lactic acid,
ethanol, acetic acid, glycerol, mannitol and
other polyalcohols, and carbon dioxide.
Heterofermentative MLB can metabolise
citric acid to predominantly acetic acid,
lactic acid and carbon dioxide [Dittrich,
1977; Subramanian and SivaRaman,
1984; Martineau and Henick-Kling, 1995].
The co-fermentation of citric acid and
glucose by O. oeni has been seen to
increase its growth rate and biomass
production [Salou et al., 1994; Ramos and
Santos, 1996; Liu, 2002].

Lactobacillus contains both homo- and heterofermentative strains,

whilst Pediococcus species are all homofermentative, and
Leuconostoc species are all heterofermentary. Pediococcus and
Leuconostoc usually cease growing below pH 3.5.
The following table describes various species:

Table of some LAB Species

Pediococcus
Homofementant
cerevisiae
mainly present in musts and
Pediococcus
at high pH after malolactic
damnosus
fermentation
 Pediococcis sometimes present lower
pentosaceus populations
Pediococcis parvulus
sometimes present lower
Lactobacillus casei
populations
Lactobacillus
mainly present in musts
plantarum
Lactobacillus casei
Streptobacterium
Heterofermentant Leuconostoc gracile
Oenococcus oeni most resistant to low pH
Leuconostoc sometimes present lower
mesenteroides populations
Lactobacillus
hilgardii
Lactobacillus
fructivorans
Lactobacillus
desidiosus
Lactobacillus brevis
Lactobacillus sometimes present lower
hilgardi populations
sometimes present lower
Lactobacillus brevis
populations

(Cell shape and grouping may depend on the medium in which the
bacteria grow.)

3 Effects of MLF

3.1 General

The reaction of interest during MLF is the conversion of L(-) malic

acid to monocarboxilic (L-) or D(-) lactic acid and carbon dioxide.
(Whether L- or D- lactic acid is produced depends on the LAB strain
and substrate attacked.) All LAB follow this pathway for MLF. It is
worth noting that some commercial malic acid is racemate, a
mixture of L(-) and D(+) malic acid. Only L(-) malic acid will be
converted by MLB.
 Malic acid ---> Lactic acid + Carbon dioxide
COOH-H2OC-H2C-COOH ---> CH3-CHOH-COOH + CO2
For every 1 gram of malic acid metabolised, 0.67 grams of lactic
acid and 0.33 grams (or 165 ml) of CO2 are produced.

Bacteria also use only 0.3-2 g/l of sugars yielding about 100 mg/l of
D-lactic acid [Krieger et al., 2000].

3.2 Acidity

3.2.1 TA, pH, acid species

MLF affects TA, pH and acid species through the following factors:

 a chemical deacidification usually reducing titratable acidity by

about 1-4.6 g/l (as tartaric)
 a pH increase of between 0.1 and 0.45 units (more typically 0.1-
0.25)
 the fact that the newly formed lactic acid (like that in milk)
appears softer on the palate than the converted malic acid (like that
in apples).
These factors can give a wine a rounder, softer, mellower mouth-
feel.

MLB also produce substances which give these sensations more

directly, and/or integrate with bitter and astringent substances in
wine. This is predominantly seen in red wines and appears to be
dependant on MLB strain.

3.2.2 Volatile acidity (acetic acid)

Most LAB cause a slight increase in volatile acidity. Acetic acid

increases in dry wine due to citric acid metabolism and chetonic
compounds. This increase is by 0.05-0.2 g/l according to Krieger et
al. [2000], though Riesen [1999] quotes up to 0.3 g/l. This increase
is not usually perceptible, though the formation of acetic acid esters
does have an impact. [Riesen, 1999]. The increase is partly due to
LAB attacking trace sugar (usually 0.3-2 g/l of sugar [Krieger et al.,
2000]) and citric, pyruvic, and chetoglutaric acids, in addition to
acetaldehyde. Acetic acid production is higher during MLF at higher
pHs [see Wibowo et al., 1985].
Acetic acid is produced during bacterial growth in must and wine
(from sugar and organic acid metabolisation), not during malic acid
degradation. If the MLB grow fast, more acetic acid is produced.
Thus, when a high population MLB culture is inoculated into a wine
there is less growth, and therefore, a considerably lower production
of acetic acid.

3.3 Microbial stability

Under certain conditions, an MLF can increase the microbial stability

of a wine.
Some spoilage bacteria attack malic acid. Therefore, by reducing
the malic acid content of a wine using the favourable method of
controlled MLF, a more microbially stable wine can potentially
result.
MLB also consume nutrients (amino acids, nitrogen bases, vitamins)
and this reduction in nutrient availability has been thought to
increase microbial stability by limiting the potential growth of
spoilage orgaisms. However, it has been shown that wines which
have completed MLF can still support Oenococcus oeni, lactobacilli
or pediococci [Costello et al., 1983]. Indeed bacterial growth may
continue after MLF, particularly if SO2 levels remain low [Davis et
al., 1986].
Additionally, it should be kept in mind that MLFs which raise the pH
above 3.5 can cause a potential increase in microbial instability due
to providing a more favourable environment for spoilage organisms.

3.4 Aromatic/Taste Modification

Many people claim the alteration MLF makes in flavour alone

(without regard for acidity) is not perceptible. Some studies have
confirmed this [see Davis et al., 1985] and many might be used to
support this argument [Asmundson and Kelly, 1990; Axelsson,
1993; Buckenh 黶 kes, 1993; Henick-Kling, 1995; Henick-Kling et
al., 1989; Sandine and Heatherbell, 1985; Sharpe, 1981; Wibowo
et al., 1985; van Vuuren and Dicks, 1993]. However, at least one
other study showed that similar wines which have undergone MLF
can be distinguished from those which have not [McDaniel et al.,
1987]. Certainly, MLB can produce a whole host of products during
MLF (succinate, acetate, acetoin, lactate, diacetyl, mannitol, 2-
butanol, 1,3-propanedoil, and many more) depending on the
available substrates (acids, sugars, polyols, etc). These compounds
are claimed by many to potentially modify vegetal characters, and
potentially impart nutty, lactic, and/or earthy aromas.
Undesirable odours brought about by MLF are usually associated
with pediococci or lactobacilli, or MLF occurring above pH 3.5;
whereas MLF by Oenococcus oeni below pH 3.5 is less likely to
produce off-odours [Jackson, 1994].

3.5 Loss/Gain of Fruitiness and Vegetative/Grassy

Notes

Non-specific MLB strains can reduce or mask wine esters, resulting

in a reduction of varietal characteristics and fruit aromas.
However, some MLB strains (Leuconostoc oeni in particular) have
been shown to increase these compounds (blueberry, raspberry,
cherry, pineapple). Current theory is that a direct bacterial action
releases varietal compounds from odourless precursors in wine.
[Krieger et al., 2000; Riesen, 1999].
A reduction in excessive vegetative and grassy notes caused by MLB
has been seen in wines made from underripe grapes, though the
mechanism for this is unknown. [Krieger et al., 2000; Riesen,
1999].

3.6 Loss of Colour

A reduction in colour intensity can be caused by MLB directly (not

because of pH increase). This is due to the metabolic activity of
bacteria on phenolic compounds [Lonvaud-Funel, 1995]. MLF has
been shown to change colour hue (lower 520 nm) [Riesen, 1999].

3.7 Astringency

MLF can decrease astringency. Under the influence of MLF alone,

total phenols may decrease (gelatin index, measured at 280 nm).
[Riesen, 1999]

3.8 Tannins and Anthocyanins

MLF can increase the polymerisation of tannins and anthocyanins.

[Riesen, 1999]
3.9 Ethyl-lactate

Often exceeds the taste threshold of 60-110 mg/l giving a wider

and fuller palate.

3.10 Diacetyl

3.10.1 What is diacetyl?

Diacetyl is a substance produced by many LAB which smells like

warm/hot butter. It is commonly used to flavour "movie" popcorn
and gives the characteristic buttery aroma to MLFed, lees-stirred
Chardonnays that are so popular today.

3.10.2 Levels of diacetyl

Threshold levels vary for different wines. Martineau et al [1995]

claimed diacetyl thresholds of 0.2 mg/l in Chardonnay, 0.9 in Pinot
Noir, and 2.8 in Cabernet Sauvignon. Krieger et al. [2000] claim a
level of 2-7 mg/l can give a "butter or cheese" note that makes the
aroma heavy and unpleasant and Jackson [1994] writes that 1-4
mg/l adds "a desirable complexity to the fragrance" but at
concentrations over 5-7 mg/l the buttery aromas becomes overt
and undesirable. However, Dharmadhikari [2002] reported that 2-3
mg/l and 4-5 mg/l of diacetyl enhanced wine aroma in whites and
reds respectively. It is clear from this information that acceptable
levels vary depending on wine style and, most likely, personal
preference for diacetyl aswell.

On testing wines from 20 different regions, 28 different producers

and 8 vintages, Bartowsky et al. [2002] found that Chardonnay (of
24 wines) showed levels of diacetyl from 0.3 to 0.6 mg/l (mean
value 0.4 mg/l) whilst reds (Cabernet Sauvignon, Merlot and
Shiraz) showed 0.3 to 2.5 mg/L (mean value 1.1 mg/l) for 43
wines.

3.10.3 The production of diacetyl

Diacetyl appears to be a metabolic by-product of citric acid

metabolisation by Oenococcus oeni [Nielsen and Prahl, (1995) and
Shimazu et al., (1985)]. The amount produced depends on:
 the LAB strain
 cell multiplication (the lower the cell multiplication, the lower the
production of acetate and diacetyl; under stressed slow cell growth
conditions more diacetyl is produced ad less acetic acid produced)
 the environmental conditions affecting the LAB
 the amount of citric acid available
 sulphur dioxide (SO2) content
 the redox potential (oxygen content) of the wine.

Research suggests that

 citric acid metabolisation begins at the same time that malic acid
degradation occurs, but that the degradation of the citric acid is
slower
 the higher the initial concentration of citric acid, the more
diacetyl will be produced from MLF
 diacetyl concentration increases as citric acid is metabolised by
MLB and decreases again when most of the citric acid has been
consumed
 maximum diacetyl concentration tends to coincide with the
exhaustion of malic acid during MLF [Nielsen, 1995]
 semi-aerobic (2-4 mg/l oxygen) MLF yielded higher
concentrations of diacetyl, whereas anaerobic (<0.2 mg/l oxygen)
conditions yielded much lower concentrations [Nielsen, 1999]
 the higher the pH, the lower the diacetyl production [see
Wibowo et al., 1985]

The diagram shows the main metabolic pathway for citric acid
metabolisation by Leuconostoc oeni.
3.10.4 The breakdown or binding of diacetyl

Saccharomyces cerevisiae irreversibly reduces diacetyl, and sulphur

dioxide (SO2) binds reversibly with diacetyl. The addition of SO2
will initially decrease the concentration of diacetyl, and increase the
quantity of bound SO2. When the diacetyl-SO2 complex dissociates,
however, the diacetyl and free SO2 levels will increase.

Diacetyl is also reduced by LAB irreversibly to acetoin and 2,3-

butanediol (a sweet-tasting polyol (a polyalcohol is an alcohol with
more than one hydroxyl group per molecule)). These have no
influence on wine aroma in the normal concentrations they are
found in. [De Revel et al., 1989]

3.10.5 Minimising diacetyl

For a low diacetyl concentration, the best approach is to inoculate

with MLB at end of fermentation (when the yeast population is high)
and keep the wine on its lees (to allow the yeast and LAB to convert
the diacetyl to acetoin and 2,3-butanediol).

3.10.6 Maximising diacetyl

For a high diacetyl concentration, rack/clarify the wine in order to

reduce yeast population before inoculating MLB and stabilise the
wine as soon as the malic acid is catabolised. Malolactic nutrients
are often required in this case, since the necessary nutrients for
MLF have been removed due to the racking/clarifying procedure.
(0.13-0.15 g/l (0.5 g/U.S. Gal) is recommended.)

3.10.7 Diacetyl and red wines

Since diacetyl is not considered desirable in red wines, MLF is

usually conducted on yeast lees (or the wine is left in contact with
MLB lees). This results in the diacetyl being catabolised by the yeast
(or MLB).

4 Living conditions
The limiting factors on MLF include total and free sulphur dioxide
(SO2) concentration, alcohol content, pH, and temperature. Each
LAB has it's own limits (and culture manufacturers usually provide
these). The following limitations provide general guidelines (these
are based on Oenococcus oeni unless where otherwise stated).

4.1 Total and free SO2 concentration

Most LAB are sensitive to free SO2 concentrations above of 10-20

mg/l or above. Total SO2 is usually kept below a maximum value of
70 mg/l (for reds) and 40 mg/l (for whites). Molecular levels of 0.6
mg/l are inhibitory, whereas levels of 1.2-1.8 mg/l are strongly
inhibitory. Keeping molecular levels below 0.2 mg/l is desirable. It
should be noted that O. oeni can degrade acetaldehyde. Thus, if
SO2 were previously added to the wine/must and subsequently
became bound with acetaldehyde then the action of MLB upon this
bound complex will results in an increase in free SO2.

4.2 Alcohol content

MLB are sensitive to ethanol and usually struggle above an abv of

13.5% exhibiting very slow (or non-existent) growth. Leuconostoc
oeni appears to adapt to high alcohol environments over time but
loses this adaptation when returned to lower abv environments.
Lactobacillus is the most resistant to ethanol and Oenococcus oeni
(formerly named Leuconostoc oeni) appears to be the most
sensitive [Jackson, 1994].
4.3 pH

MLB favour higher pH's. The optimum pH will depend on the strain
and the culturing environment. For most strains, minimal growth
occurs at pH 3.0. Generally, pH's above 3.2 are advised. Optimal
growth for O. oeni occurs at pH 4.2-4.8. Optimal activity occurs
between pH 3.0 and 4.0 and decreases as pH rises - inhibition is
experienced at pH 4.5.

4.4 Temperature

MLF will proceed faster at higher temperatures. When no SO2 is

present in the wine the optimum temperature range for MLF is 23-
25°C (73-77°F). (Maximum malic acid degradation will occur at 20-
25°C (68-77°F) [Rib 閞 eau-Gayon et al., 1975].) However, this
decreases with increasing concentrations of SO2 causing 20°C
(68°F) to be more suitable. The following table indicates MLF speed
at given temperatures.

Delay Temperature
slowed by months / essentially no malic acid
10°C (50°F)
decarboxylation
12-13°C (54-
slowed by weeks
55°F)
begins to slow 15°C (59°F)
20-25°C (68-
optimum
77°F)
MLB death >30°C (> 86°F)

Most strains of Oenococcus oeni either cease to grow or grow very

slowly below 15°C (59°F). However, cells remain viable at low
temperatures. [Jackson, 1994].

4.4 Population

MLB require a certain population level to be reached before they can

begin MLF. A population of around 105 cells/ml is a good starting
level. Due to the stress inoculated MLB undergo in adapting to the
wine environment, inoculations of MLB are often increased to
population loads of at least 108 cells/ml.
All these factors act in synergy and the tolerance limit for one factor
can change if the other factors have an inhibiting affect on MLF.
Inhibition of MLB does not necessarily mean the bacteria have died.
A bacterial population may simply be dormant and become active
when conditions are more favourable. This explains the often seen
delay of MLF in bottled wines, as well as the increase in MLF activity
in cellared wines over the spring as cellar temperatures rise.

5 LAB development
MLB require a certain population level to be reached before they can
begin MLF. This means that natural/indigenous MLFs (where LAB
have not been inoculated into the must or wine) can have a lag
phase of weeks to several months before they begin MLF.
This lag phase is prolonged the lower the pH. Additionally, the pH
affects which species of LAB will be dominant in the must or wine
[Bousbouras and Kunkee, 1971]. At low pHs Oenococcus oeni is the
primary MLB genus, different strains of which will dominate
throughout MLF. At higher pHs, however, Lactobacillus and
Pediococcus dominate over Leuconostoc [Costello et al., 1983].
During MLF, the population often reaches 1 million cells/ml.
(Leuconostoc oeni in particular requires a population of more than
106-107 CFU/ml to begin MLF.) Their population is usually 10 3-104
CFU/ml following fermentation, and may rise to 10 6-108 CFU/ml
once growth initiates substantially.
After exponential growth, cell populations deline rapidly depending
on the conditions (e.g. elevated tempertures or SO2 will increase
the rate). It is possible that the population of strains of
Lactobacillus and Pediococcus may increase at this point.

6 Why inoculate
The MLF growth (lag) phase associated with spontaneous MLF
(wild/uncultured strains) presents a time of increased risk from
spoilage organisms due to the non-SO2 environment and the
potential production of volatile acidity. Inoculating with a prepared
MLB culture avoids the problems associated with the MLB growth
lag phase by immediately providing the population necessary to
conduct MLF. It is important that the MLB have enough nutrients to
develop. Yeast (Saccharomyces cerevisiae var. "bayanus" in
particular) can reduce the nutrients available to MLB considerably.
Winemakers often add a MLB nutrient when inoculating with MLB to
assist their development.

Additionally, spontaneous MLB can cause unpleasant aromas and

tastes, which can include sweat, mould, sulphate, phenols,
sauerkraut, or a bitter and oily aftertaste. The source and chemical
composition of these compounds is not known, but it has been
noted that they tend to occur in high pH and low SO2 environments,
indicating Lactobacillus and Pediococcus or wild strains of
Leuconostoc. Inoculation of cultured MLB avoids these unpleasant
flavours.

Cultured MLB strains are consistent. When relying on wild MLB,

there can be no certainty of consistent results in consecutive wines.
Some winemakers feel, however, that wild MLB tend to lend more
complexity to their wines. Additionally, some areas in the world
which have been making wine in the same cellar (and region) for
centuries may have resident MLB which are relatively stable and
favourable for MLF, thus providing reliable wild MLB strains for MLF.

7 Mixed cultures
Some believe there are advantages in conducting MLF with mixed
MLB cultures.

8 Timing of inoculation of cultured MLB

8.1 Sugars

Inoculation of MLB is usually conducted after all sugars (<2 g/l of

reducing sugars is a suitable level) have been fermented by yeast.
This avoids the possibility of MLB metabolising the sugar and
producing unwanted products such as acetic acid. Some research
suggests that pre-fermentation inoculation results in higher VA
concentrations [Semon et al., 2001]. When considering the timing
of inoculation, the compatibility of yeast and LAB compatibility
should also be considered. (Competition for nutrients can occur with
Saccharomyces cerevisiae var. "bayanus" strains, for example.)
Some winemakers, however, inoculate MLB a few days into alcoholic
fermentation in an attempt to minimise contamination through the
subsequent ability to add SO2 to the wine sooner.
8.2 Nutrients, alcohol and SO2

Interactions between co-existing yeast (Saccharomyces cerevisiae)

and Leuconostoc oeni can cause problems with MLF. Saccharomyces
cerevisiae releases ethanol, sulphur dioxide (SO2) and medium
chain fatty acids which inhibit O. oeni [Lonvaud-Funel et al., 1988].
Additionally, the yeast's utilisation of complex nutrients such as
amino acids during the early stages of fermentation can complicate
ensuing MLB growth.

Later inoculations result in more efficient MLB growth when

Saccharomyces cerevisiae var. "bayanus" strains are used. If a
bayanus strain is used, it is recommended that MLB inoculation be
conducted after complete alcoholic fermentation (avoiding
competition nutrients). Additionally, MLF nutrients should be added
to the wine since bayanus yeasts tend to monopolise and consume
most nutrients.

Bacterial inhibition decreases towards the end of fermentation. This

is probably connected with the death phase of yeast as SO2
produced during fermentation reduces and binds with other
compounds and dying yeast cells release nutrients useful to the
MLB.

However, some winemakers inoculate before the end of alcoholic

fermentation, taking advantage of the low alcohol, higher nutrient
environment. Assuming MLF completes before or soon after
alcoholic fermentation, this allows SO2 (which provides security
against bacterial and oxidative attack) to be added to the wine
sooner rather than later.

However, other work does not show higher VA concentrations, stuck

fermentations, or yeast-MLB incompatability [Beelman and Kunkee,
1985].

9 Measuring MLF progress

Most home winemakers do not have access to advanced lab
equipment to measure microbial populations or the concentration of
substrates and products. Therefore, the progress of MLF is usually
measured by paper chromatography (PC). In this test, drops of the
tested wines are placed on the chromatographic filter paper and
dried, with reference standards of tartaric, citric and malic acids.
The spotted paper is then (usually) rolled up like a cylinder and left
vertically to absorb chromatographic solvent via capillary action.
The solvent travels vertically up the paper, separating out the
organic compounds present. When the solvent has almost
penetrated to the top of the paper (usually after 8 hours), the paper
is removed and left to dry.

The chart can then be viewed and the presence of each individual
acid assessed. The background appears blue, with yellowish spots
appearing up the paper. Each acid can only travel a certain distance
up the paper. Tartaric, citric, malic, lactic, and succinic acids are
represented in order at heights progressively up the paper.

Below is a theoretical diagram of a chromatographic paper. Tartaric,

citric, malic and lactic acids are indicated by the letters T, C, M, and
L, respectively. Whilst each acid travels the full certain distance up
the paper, it can be seen that the interpretation of acids occupying
regions can be helpful.
Images 1 to 3 show wines which have progressively undergone MLF
to a greater extent. In wine 1, there is a small presence of lactic
acid and a high level of malic acid present. In wine 2, less malic is
shown to exist and more lactic has been produced as a result of
MLF. Wine 3 shows very little malic acid and a high level of lactic,
suggesting that this wine has nearly completed MLF. Paper
chromatography indicates malic acid concentrations above about
100 mg/l (wines may continue to undergo MLF over 30 mg/l without
showing any malic spot on the paper). Therefore, wine 3 may have
a little way to go until MLF is entirely complete, but it is almost
there.

The tartaric spot on 2 is weak. In wine's where the predominant

acids are tartaric and malic (such as from grapes), there is a
possibility that a high proportion of the acid in the wine is malic. If
this is the case, a complete MLF might leave the wine with an
unacceptably low TA and an excessively high pH. Therefore, these
parameters should be monitored.

If, alternatively, a wine has a weak malic acid concentration, there

is little point in conducting a MLF at all as it would have little affect
on the wine.

The diagram of wine 4 is included to indicate what the paper looks

like when citric acid is present.

10 MLF and Oak

LAB enzymes appear to react with soluble substances in oak
barrels, creating a wider range of flavours in a wine than would be
produced in an inert vessel.

LAB become resident in the wood or tartrate layer of cooperage. A

wine that has not been through, nor is intended to go through, MLF
(MLF- wine) cannot, therefore, safely be made in a (MLF+) barrel
which has previously stored MLFed wine.

11 MLF and Sorbate

Sorbate should not be added to a wine which is to (or may) undergo
MLF. MLB transform sorbate into ethoxyesadiene which has a strong
smell of grass or geranium. When this aroma is perceptible in wine,
the wine is generally considered ruined.

12 MLF and Lees

Lees provide nutrients for MLB. By stirring the yeast lees, nutrients
are re-exposed to the bacteria assisting their development. Lees in
the presence of MLF also tends to lead to lower diacetyl
concentrations, since the yeast metabolise the diacetyl to acetoin
and 2,3-butanediol.
Lees stirring is usually conducted on fine lees (mainly yeasts) and
not gross lees (that including fruit debris). However, some
winemakers conduct sur lie ageing on gross lees in barrels.
Substances which are toxic to MLB also exist in lees, however.
These include such substances as fatty acids C10 and C12, which
can be absorbed into lees (fine lees and yeast hulls) during alcoholic
fermentation.

13 Preventing MLF
Factors to help in preventing MLF occurring include the following:

 Use of lysozyme (a protein that destroys the cell walls of the

bacteria by catalyzing the hydrolysis of specific glucosidic links in
the cell walls)
 Keeping SO2 concentrations at MLF preventative levels - 0.8
mg/l molecular SO2 (or less accurately, a free SO2 concentration of
50 mg/l)
 Keeping pH low - below 3.2
 Sterile filtration
 Keeping wine lees contact to a minimum and racking
immediately upon completion of alcoholic fermentation
 Use of a bayanus yeast strain for alcoholic fermentation (limits
available nutrients for bacteria)
 Keeping wine below 10°C (50°F) can provide limited assistance
 Ensuring alcohol content is greater than 13%
 Keeping maceration on skins to a minimum
 Pasteurisation

14 MLF schedules: Partial MLF

Partial MLF is used when some crisp acid/fruity character is desired,
yet some softening and complexity from MLF is desired. There are
two techniques used to achieve this:

(1) The batch is split, one portion goes through full MLF and the
other no MLF. The two portions are then blended back together. The
problem to then overcome is a renewed MLF in the wine, since there
is still malic acid present. The best method in providing stability is
to sterile filter the wine. Alternatively, a thorough clearing and
filtration of the wine coupled with a 0.8 mg/l molecular SO2 level
can be successful in preventing renewed MLF.

(2) The wine undergoes MLF and is closely monitored. When the
winemaker is satisfied with the level of MLF (based on taste and
analytical methods), the MLF is arrested. MLF is terminated by
sterile filtration (<0.45 micron tight filter to remove yeast and lactic
bacteria), a 1.5-2 mg/l molecular SO2 level, keeping wine lees
contact to a minimum and racking when MLF termination is desired
to reduce the MLB population, and possibly the use of lysosyme and
cold temperatures/stabilisation.

15 Wine styles suitable for MLF

MLF is suitable for wines where a reduction in acidity is desired and
the wine style is not so much focused towards varietal aromas and
flavours. Where fruitiness and crispness/freshness (an impression
given by acidity) is desired, MLF is generally avoided. Certain wines
tend to lose their aroma through MLF because lactic aromas tend to
dominate the fruity aromas.

Some winemakers even use MLF in wines from aromatic grape

varieties such as Gew 黵 ztraminer and Muscat (Riesling seems an
exception where MLF is practically never practised) to improve body
and mouth feel, without a significant reduction in fruitiness.
Generally, however, it is avoided in these wine styles. That being
said, there are exceptions. For example, Sauvignon Blanc, anothe
aromatic variety, is sometimes put through (partial) MLF.

Complex, full bodied whites often benefit from MLF. Chardonnay is a

classic example. MLF gives a softer, more luscious wine with added
complexity.

Most reds undergo MLF. One of the main reasons for this is to
ensure that the wine will not undergo MLF later in bottle; leading to
lowered acidity, off aromas, and carbon dioxide.

MLF has traditionally been avoided with regard to fruit, flower and
vegetable wines. This is largely because an emphasis is place on
primary ingredient ("fruit" or varietal) character. However, many
non-grape wines may be suitable for MLF. Some winemakers
routinely MLF apple wine, for instance, and reds such as blackberry
are suitable candidates.

Overall, MLF is a stylistic option that will suit some wines and not
others. It is up to the winemaker to assess whether the individual
wine in question will benefit or not from MLF. The above information
hopefully provides clues as to how to discern this.

16 References
Asmundson, R.V. and Kelly, W.J. (1990). Temperature and
ethanol effects on growth of Leuconostoc oenos. In: Yu, P.-L. (ed.)
Fermentation technologies: Industrial applications, Elsevier, London,
pp 128-131.
Axelsson, L.T. (1993). Lactic acid bacteria: classification and
physiology. In: Salminen S. and von Wright A. (eds.), Lactic acid
bacteria, Marcel Dekker Inc., New York, pp 1-63.
Bartowsky, E.J., Francis, I.L., Bellon, J.R., and Henschke, P.A.
(2002). Is buttery aroma perception in wines predictable from the
diacetyl concentration?, AGJWR, Volume 8, Number 3.
Beelman, R.B., and Kunkee, R.E. (1985). Inducing simultaneous
malolactic-alcoholic fermentation in red table wines. In "Malolactic
Fermentation" (T.H. Lee, ed.), pp. 97-111. Aust. Wine Res. Inst.,
Urrbrae, South Australia.
Bousbouras, G.E., and Kunkee, R.E. (1971). Effect of pH on
malolactic fermentation in wine. Am. J. Enol. Vitic. 22, 121-126.
Buckenh 黶 kes, H.J. (1993). Selection criteria for lactic acid
bacteria to be used as start cultures for various food comodities.
FEMS Microbiol. Rev. 12, 253-271.
Costello, P.J., Morrison, R.H., Lee, R.H., and Fleet, G.H.
(1983). Numbers and species of lactic acid bacteria in wines during
vinification. Food Technol. Aust. 35, 14-18.
Davis, C.R., Wibowo, D., Eschenbruch, R., Lee, T.H., and
Fleet, G.H. (1985). Practical implications of malolactic
fermentation in wine. J. Appl. Bacteriol. 63, 513-521.
Davis, C.R., Wibowo, D.J., Lee, T.H. and Fleet, G.H. (1986).
Growth and metabolism of lactic acid bacteria during and after
malolactic fermentation of wines at different pH. Appl. Environ.
Microbiol. 51, 539-545.
De Revel, G., Bertrand, A., and Lonvaud-Funel, A. (1989).
Conn. Vigne Vin, 1, 39-45.
Dharmadhikari, Murli. (2002). Some Issues in Malolactic
Fermentation Acid Reduction and Flavor Modification, Vineyard and
Vintage View, Volume 17(4), p 4-6.
Dittrich, H.H. (1977). Handbuch der Getr 鋘 ketechnologie:
Mikrobiologie des Weines, Verlag Eugen Ulmer, Stuttgart, pp 233-
256. [In German]. Garvie, E.I. (1986). Genus Leuconostoc. In:
Sneath, P.H.A., Mair, N.S., Sharpe, E.J., Holt, J.G. (eds.) Bergey's
manual of systematic bacteriology vol 2, Williams and Wilkins,
Baltimore, pp 1071-1075.
Henick-Kling, T. (1993). Malolactic fermentation. In: Fleet, G.H.
(eds.) Wine microbiology and biotechnology, Harwood Academic
Publishers, Chur, pp 289-326.
Henick-Kling, T. (1995). Control of malo-lactic fermentation in
wine: Energetics, flavour modification and methods of starter
culture preparation. J. Appl. Bacteriol. Symp. Suppl. 79, 29S-37S.
Henick-Kling, T., Sandine, W.E. and Heatherbell, D.A. (1989).
Evaluation of malolactic bacteria isolated from Oregon wines. Appl.
Environ. Microbiol. 55, 2010-2016.
Jackson, R.S. (1994). "Wine science: principles and applications,"
San Diego Academic Press.
Krieger, S., Triolo, G., and Dulau, L. (2000). "Bacteria and Wine
Quality - State of the Art," www.lallemandwine.com
Liu, S.-Q. (2002). Malolactic fermentation in wine - beyond
deacidification: A review. J. Apll/ Microbiol. 92, 859-601.
Lonvaud-Funel, A. (1995). Microbiology of the malolactic
fermentation: molecular aspects. FEMS Microbiol. Lett. 126, 209-
214.
Lonvaud-Funel, A., Joyeux, A., Desens, C. (1988). Inhibition of
malolactic fermentation of wines by products of yeast metabolism.
Journal of the Science of Food and Agriculture, 44, 183-191.
Martineau, B. and Henick-Kling, T. (1995). Performance and
diacetyl production of commercial strains of malolactic bacteria in
wine. J. Appl. Bacteriol. 78, 526-536.
Martineau, B., Acree, T.E. and Henick-Kling, T. (1995). Effect
of wine type on threshold for diacetyl. Food Research International
28(2).
McDaniel, M., Henderson, L.A., Watson, B.T., Jr., and
Heatherbell, D. (1987). Sensory panel training and screening for
descriptive analysis of the aroma of Pinot noir wine fermented by
several strains of malolactic bacteria. J. Sens. Stud. 2, 149-167.
Nielsen, J.C. and Prahl, C. (1995). Metabolism of citric acid by
Leuconostoc oenos in direct inoculation. Effect on wine flavour.
Results presented at the 5th International Symposium of Enology,
Bordeuax, 15-17th June 1995.
Nielsen, J. C., and Marianne Richelieu. (1999). Control of flavor
development in wine during and after malolactic fermentation by
Oenococcus oeni. Applied and Environmental Microbiology 65(2):
740-745.
Radler, F., and Yannissis, C. (1972). Weins 鋟 reabbau bei Milchs
鋟 rebakterien. Arch. Mikrobiol. 82, 219-239.
Ramos, A. and Santos, H. (1996). Citrate and sugar
cofermentation in Leuconostoc oenos, a 13C nuclear magnetic
resonance study. Appl. Environm. M
Riesen, R. (1999). Modification of Wine Characteristics by
Malolactic Fermentation. http://www.oardc.ohio-
state.edu/grapeweb/vinevan/van0699.htm.
Rib 閞 eau-Gayon, J., Peynaud, E., Rib 閞 eau-Gayon, P., and
Sudraud, P. (eds.) (1975). "Trait?d'Oenologie: Sciences et
Techniques du Vin," Vol. 2. Dunod, Paris.
Sandine, W.E. and Heatherbell, D.A. (1985). Wine preparation
with new strains of Leuconostoc oenos. US Pat. 4,547,373,
application number 521,988.
Salou, P., Loubiere, P. and Pareilleux, A. (1994). Growth and
energetics of Leuconostoc oenos during cometabolism of glucose
with citrate and fructose. Appl. Environ. Microbiol. 60, 1459-1466.
Semon, J.M., Edwards, C.G., Forsyth, D., and Dinn, C. (2001).
Inducing malolactic fermentation in Chardonnay musts and wines
using different strains of Oenococcus oeni. AGJWR, Vol. 7, No. 1, p
52-59.
Sharpe, M.E. (1981). The genus Lactobacillus. In: Starr, M.P.,
Stolp, H., Tr 黳 er, H.G., Balows, A. and Schegel, H.G. (eds.).
Prokaryotes, vol. 2, Springer-Verlag. Berlin, pp 1653-1679.
Shimazu, Y., Uehara, M., and Watanbe, M. (1985).
Transformation of citric acid to acetic acid, acetoin and diacetyl by
wine making lactic acid bacteria. Agric. Biol. Chem. 49, 2147-2157.
Subramanian, S. and SivaRaman, C. (1984). Deacidification of
musts from the western United States by the calcium double-salt
precipitation process. Am. J. Enol. Vitic. 29, 153-160.
van Vuuren, H.J.J. and Dicks, L.M.T. (1993). Leuconostoc
oenos: A review. Am. J. Enol. Vitic. 44, 99-112.
Wibowo, D., Eschenbruch, R., Davis, C.R., Fleet, G.H., and
Lee, T.H. (1985). Occurrence and growth of lactic acid bacteria in
wine: A review. Am. J. Enol. Vitic. 36, 149-153 / 302-312.
 www.brsquared.org/wine

Enology Notes

Enology Notes #104, August 3, 2005

To: Regional Wine Producers

From: Bruce Zoecklein, Head, Wine/Enology-Grape Chemistry Group,

Virginia Tech

Subjects: Volatile Sulfur Compounds Workshop slide show posted;

Bleeding and Fermentation; Current Topics in Fermentation Workshop;
Berry Weight and Maturity Evaluation.

1. Volatile Sulfur Compounds Workshop slide show posted. The slides

from my presentation in New Zealand last November and from our recently-
concluded workshop on volatile sulfur compounds are posted at
www.vtwines.info and can be found in our On-line Publications and Current
Topics section.

2. Bleeding and Fermentation. Many conduct tank bleeding in order to help

increase the body and structural depth of red wines. Several have expressed
concern that some juices produced by bleeding do not ferment to dryness,
have a tendency toward reductive notes, and/or are difficult to get to complete
MLF.

As discussed in the recently-concluded workshop on volatile sulfur

compounds, amino acids are impacted by fruit maturity and are not equally
distributed in the grape berry.

For example, with mature Cabernet Sauvignon, about 8.5% of the total are in
the seeds, 15% in the skins and 77% in the pulp. It would seem that the
separation of the pulp juice from the skins as occurs with bleeding would not
have a large quantitative effect on juice that is removed. However, there is a
significant qualitative influence.

The two amino acids present in the greatest concentration in the fruit are
proline and arginine. Proline cannot be used by the yeast, while arginine can.
Indeed, because it has four atoms of N per molecule (three of which are
believed used by yeasts), arginine is a very good source of fermentable N.
The ratio of these two amino acids is impacted by maturity. Arginine
accumulation in the fruit begins before veraison, continues to increase,
plateaus, and declines. Proline, on the other hand, increases late in the
season and continues to increase with increased hang-time.

High levels of proline are, therefore, associated with increased maturity. High
proline is also associated with plant stress. The reduction in the arginine to
proline ratio can be significant and can negatively impact the amount of
fermentable nitrogen available to the yeast.

The ratio of arginine to proline is much greater in the skins than the pulp. In
other words, pulp juice taken off in bleeding has a relatively high
concentration of proline (approximately 55%) which cannot be used by the
yeast, and a small concentration of the more potent amino acid arginine and
others needed to carry out a healthy fermentation. The lower incidence of
incomplete fermentation in red compared with white musts supports the
concept that the slow release of nitrogen from grape skins during fermentation
is important.

Wines produced by bleeding should be tested separately for fermentable N

and likely given a higher concentration of supplemental fermentable N - higher
than is required to ferment the juice remaining in contact with the skins. This
will help assure ‘clean’ fermentations which go to completion.

If you desire to have your blush or rosé wines undergo malolactic

fermentation, I would also suggest that you add a commercial MLF
supplement at dryness and before bacterial inoculation.

As discussed at the workshop, all musts should be evaluated for fermentable

N. A simple method involves the formol titration (see the Wine/Enology –
Grape Chemistry Group website: www.vtwines.info, under On-line
Publications and Current Topics section).

Also as discussed at the volatile sulfur compound workshop, all wines should
undergo sulfite screening prior to bottling. See Enology Notes and Zoecklein
et al. (1999).

3. Current Topics in Fermentation Workshop. We will conduct a pre-

harvest workshop on current fermentation topics, and a sensory evaluation of
our training systems research wines on Monday, August 22nd beginning at
1:00 p.m. at Horton Vineyards.

Invited speakers to the meeting include:

  Lisa Van de Water, Director, Pacific Rim Oenology
 Patricia Roca, Technical Manager, Vinotec, Chile

Lisa and Patricia will discuss the latest practical information on conducting
healthy fermentations.

I will present our Traminette, Viognier and Cabernet Franc wines produced
from our training system trials: VSP, GDC and Smart-Dyson (Up vs. Down)
from 2003 and 2004.

Registration: To reserve a spot, send an email message to Terry Rakestraw,

at rakestra@vt.edu with the word "Pre-harvest" in the subject line.

Cost: $25 per person. The complete registration fee is due NO LATER THAN
August 18, 2005. Checks are to be written payable to Virginia Tech
Foundation and mailed to Terry Rakestraw, Department of Food Science and
Technology (0418), Virginia Tech, Blacksburg, VA 24061. Course fee is non-
refundable.

4. Berry Weight and Maturity Evaluation. Any measure of grape maturation

should correlate to wine aroma/flavor potential. In a number of issues of the
Vintner's Corner available on the Web at www.vtwines.info, I have suggested
the importance of using berry weight to help determine the progress of fruit
maturation, and aid in establishing consistent and optimum wine styles.

Many of the secondary metabolites (aroma/flavor and phenolic compounds)

are located in the skins. Therefore, a change in berry size (measured easily
by weight) can and should influence winemaking decisions. Berry weight
should also influence fruit maturity decisions.

The earlier the estimation of average berry weights, the more time the
winemaker has to evaluate the crop load, make adjustments, and plan for the
season. There is a relationship between berry weight at veraison and berry
weight at maturity.

For Syrah, McCarthy (1997) determined that relationship to be the following: y

= 1.35x + 0.53, where y = the berry weight at 23 Brix and x = the berry weight
at about 5 Brix. This relationship will differ by cultivar and site, but can easily
be determined by collecting veraison and harvest samples for several
seasons.

Changes in berry weight confound the measurement of degrees Brix, and are
yet another reason why Brix must not be the sole monitor of fruit maturation.
Research indicates that the maximum rate of production of aroma/flavor
 compounds in the fruit occurs at about the time when the berry stops
importing water from the phloem or shortly thereafter. Therefore, maximum
aroma/flavor occurs sometime after the berry reaches maximum weight in
most instances. This is the reason why Syrah producers, for example, monitor
the extent of berry shriveling.

Work conducted in Australia (Mike McCarthy, 2001) has determined the

optimum weight for this variety at harvest for maximum concentration of
secondary metabolites to be about 1.2g per berry. The decline in berry weight
is more closely related to the time from flowering than to degrees Brix. The
evaluation of berry weight is an important tool in stylistic winemaking - use it.

Subscription to Enology Notes. All past Enology Notes newsjournals are

posted on the Enology-Grape Chemistry Group's web site at:
http://www.vtwines.info/.

To be added to (or removed from) the Enology Notes listserve send an email
message to rakestra@vt.edu with the word "ADD" or "REMOVE" in the
subject line.

Dr. Bruce Zoecklein

Professor and Enology Specialist Head Enology-Grape Chemistry Group
Department of Food Science and Technology, Virginia Tech
Blacksburg VA 24061
Enology-Grape Chemistry Group Web address: http://www.vtwines.info/
Phone: (540) 231-5325
Fax: (540) 231-9293

Oxygen in winemaking - Part II Wynb

oer is
W. J. du Toit, Department of Viticulture and incorporated in
Oenology, Stellenbosch University WineLand,
magazine of the
O2 plays an important role in the winemaking SA wine
process. The winemaker must thus be informed Wessel du Toit producers.
on how much O2 occurs in his wine at which
stadium of the production process. This article will focus on the Subscribe to
amounts of O2 occurring at different stages in the winemaking WineLand
process.
The measurement of O2 is not easy in wine. One of the reasons for Visit our sister
this is that wine is quite a complex medium. Different O2 sites:
measurement models are available on the market. The less expensive
one has a probe that is inserted into the wine, which is connected to a
box that gives the reading in either mg per L (parts per million) or %.
More expensive models are also available which measures up to part South African wine
per billion. These also have a steel rod that can be inserted into a farmers'
cork, which makes it ideal for measuring the O2 concentration in representative
bottled wine. organisation

O2 comes into contact with must as just after crushing of the grapes,
which catalyses the activation of oxidation enzymes. These enzymes
oxidize phenolic molecules to their corresponding quinone, as Facts, figures,
discussed in the previous article. When juice or wine is saturated with contact details and
oxygen it contains about 7-8 mg/L of O2, depending on the much more in the
temperature. There are several ways available to the winemaker to 2009/10 Directory
avoid unwanted oxidation. The first entails removing the substrate for
oxidation, the phenolic molecules. Hyper oxidation has thus been
investigated as a means of removing phenolic molecules prior to
fermentation by oxidation, where large amounts of O2 are added to
the must. This can be done by sparging the juice in the tank for about
30 min with O2, racking the juice from one tank to another for a few
times or using O2 in stead of N2 when using flotation to clarify the
juice. According to literate white juice that did not receive skin contact
can be saturated once with O2, which should be sufficient to remove
most phenolic compounds. When skin contact has been given the
juice can be left for 30 min after saturating it and then be sparged
again for 30 min. It is obvious that no SO2 should be added prior to
these operations. If SO2 is added before fermentation it should also be
added after the juice has been racked from the precipitate after
saturation. After fermentation SO2 can also be added.

Phenolic molecules can also be removed from the must and wine by
fining with gelatin, PVPP or activated charcoal. The latter should
however be handled with care, as it can strip a wine from its flavour
compounds if added at too high dosages. Higher phenolics being
extracted into the juice and wine can also be avoided by not applying
too long skin contact periods, handling the grapes with care (thus not
pressing at too high pressures and using small amounts of press
must).
Fig.1: O2 pickup in sampling bottles due to different sampling methods.

Plus= CO2 added in the sampling bottles prior to sampling, minus= no

CO2 added in the sampling bottles prior to sampling. Chain: samples

taken with chain with container middle of tank. Sapl tap: samples taken

from sampling tap from the tank. Port: samples taken from the port and

transferred into sampling bottles. All data were repeatable within 10%.

The other way of avoiding oxidation is by avoiding O2 pickup in white

wine production, the so-called reductive manner of winemaking. This
is done by adding dry ice to the crusher, press and in fermentation
tanks and Vitamin C to the must. Inert gasses, such as N2 and CO2,
can also be used to fill tanks prior to pumping the wine or to regularly
replace the air space in a tank that has not been filled completely. It's
important to remember that CO2 is heavier than air and should thus
protect wine in a half filled tank better than N2, which is lighter than
air. Argon has also been investigated, but is not really financially
viable to due to it being expensive. O2 can also be removed from wine
in a tank. This can be done by bubbling an inert gas from the bottom
of the tank through an inlet for about 30 min, depending on the size
of the tank. Here N2 is the probably the preferred choice of gas, as
too high CO2 levels being dissolved into the wine is unwanted. The
addition of high enough concentrations of SO2 to must (30-50 mg/L in
healthy grapes) and in the wine (maintaining 35 mg/L free during the
production process and 35-40 mg/L free just prior to bottling) must
also be done. Certain white wines begin to start losing fruitiness after
only a few saturations with O2. Procedures like centrifugation (7
mg/L), transport (2-6 mg/L) and filtration (3-7 mg/L) has all been
shown to contribute to O2 pickup in wine, especially if the cellar
equipment, like pump, filters etc. are faulty. When wine is also
splashed into a tank or runs down the side of the tank in a wide film,
O2 pickup is also higher.

The question remains whether reductive winemaking is always

necessary in white wine production. It has been proven that in SA
Sauvignon blanc the addition of H2O2 (which is a stronger oxidation
agent than O2) to must did not decrease the methoxypirasine levels.
It has also been proven however, that the precursor of the granadilla
flavour in French Sauvignon blancs is sensitive to O2. It thus seems
that the practice of reductive winemaking is more justified in certain
white wines than others. During bottling white wine can also pickup
O2. This has been true in older bottling lines, but with modern bottling
lines O2 pickup should not exceed 1 mg/L during bottling. Bottling
lines should also have a QC practice in place where O2 pickup is
monitored during the bottling procedure. The practice of adding
Vitamin C just prior to bottling should also be re-considered, as recent
research has shown that Vitamin C actually enhanced ageing of white
wine in the bottle.

Red wine's quality can however improve with O2 additions to certain

extend. This is due to O2 catalyzing polymerization between
anthocyanin and tannin molecules. This can already be done by
adding O2 during fermentation with sparging or pumping overs. How
much O2 actually reacts with the phenolic compounds during
fermentation is however unclear, as the yeast also take up some O2.
A splash pump over should in theory lead to a few mg/L O2 being
dissolved in the wine, but red wine fermentations are normally
conducted at around 30 鳦 (at higher temperature lower amounts of
O2 dissolves in wine) and the CO2 produced by the yeast probably also
sparge some of the O2 out. O2 addition during fermentation can,
however, be used to treat stuck or sluggish fermentations.

After fermentation and before malolactic fermentation O2 can be

added in a process called macro-oxygenation (at a dosage of 1-4
mg/L/day for up to 4 days). This could be especially helpful for high
phenolic wine, especially press wine. We even added 6 times during
extended skin contact about 4 mg/L O2 to a Cabernet Sauvignon
without the wine becoming oxidized. This practice is not followed
normally, due to fears of VA formation, but has been done safely with
the wine having less tannin and being less reactive towards proteins.
Macro-oxygenation can be done with a machine, which injects tiny
bubbles of O2 through a sparger into the bottom of a tank in a
controlled manner. Here it's, however, important to check whether the
O2 actually dissolved in the wine and did not simply escape to the top
of the tank. Macro-oxygenation can also be applied in oak barrels
after malolactic fermentation, but care should be taken that
Brettanomyces or acetic acid bacteria growth is not stimulated.

O2 can also be added to red wine after malolactic fermentation with a

process called micro-oxygenation (1-4 mg/L/month for about 1-6
months). The dosage and time of these additions dependent on the
type of wine. To heavier style red wines with high anthocyanin and
tannin concentrations O2 can be added 3-4 mg/L for a few months.
These dosages actually imitate those, which the wine would have
received in an oak barrel and micro-oxygenation can thus be used in
combination with oak staves. During barrel ageing red wine normally
receive about 30-40 mg O2/L/year. This is introduces into the wine by
topping up the barrel, racking, fining etc. It is believed that O2 also
permeates through the oak staves.

O2 can however also enhance the unwanted growth of acetic acid

bacteria and Brettanomyces, both that are spoilage microorganisms in
wine. Acetic acid bacteria go into a viable, but non-cultural state at
low O2 concentrations in wine. They thus do not grow on selective
media, which is normally used to monitor their growth in wine, but
can only be enumerated with epifluoressent microscopy, with which
live cells can be distinguished from dead cells. The addition of O2 to
the wine however negates this state and can lead to their rapid
growth. When wine tanks or barrels are filled completely the changes
of AAB spoilage is however smaller, as they normally prefer to grow
on the surface of the wine. It thus seems that O2 inside the wine is
less of a risk than on top. Brettanomyces can also use O2 at high
concentrations to grow rapidly.

When samples are taken from a tank for analysis of SO2, O2 etc. it is
important to realize that wine can quickly take up O2 in the sampling
procedure, which may give a wrong impression of the oxidative state
of the wine. In Fig 1. one can clearly see the effect of this. A 2005
Sauvignon blanc wine at the US Welgevalen cellar was used to test
this. The O2 concentration was measured in the tank and was 0.48
mg/L. Wine was then taken from the top by means of a metal
container attached to a chain and added to 250 mL sampling bottles
in triplicate. In three of the bottles CO2 gas was blown prior to adding
the wine and in another three not. The same procedure was also done
by taking samples from the sampling tap, as well as from the bottom
port of the tank and then adding it in the bottles. The lids of the
bottles were then sealed immediately and the dissolved O2 (DO)
measured after a few seconds. It is clear from Fig 1 that where CO2
was added in the bottle little O2 came into contact with the wine,
especially in the wine that was taken from the sampling tap. The wine
picked up almost 1 mg/L O2 from the sampling tap where no CO2 was
added prior to the bottles. Wine that was first collected from the port
into an open container and then transferred into the bottles pickup up
very large amounts of O2. This should be kept in mind when collecting
wine samples for analyses, or when a winemaker wants to keep a
control sample in a bottle when micro oxygenation will be used in a
tank. It is thus possible to add more O2 to the control sample than the
treated wine if done incorrectly!

It is thus clear that O2 plays a pivotal role in the winemaking process,

be it negative or positive.

Literature cited

Du Toit, W.J. and Groenewald, D.P. 2003.The effect of micro-

oxygenation on a South African red wine composition and quality.
Proceedings of the 7th International Oenologie Symposium, Bordeaux.

Du Toit, W.J. an M.G. Lambrechts. 2002. The enumeration and

identification of acetic acid bacteria from South African red wine
fermentations. International Journal of Food Microbiology 74: 57-64.

Du Toit, W.J.and I.S. Pretorius. 2002. The occurrence, control and

esoteric effect of acetic acid bacteria in winemaking. Annals of
Microbiology, 52, 155-179.

Du Toit, W.J., Pretorius, I.S. and Lonvaud-Funel, A. (2005) The effect

of sulphur dioxide and oxygen on the viability and culturability of a
strain of Acetobacter pasteurianus and a strain of Brettanomyces
bruxellensis isolated from wine. Journal of Applied Microbiology 98,
862-871.

Ribereau-Gyon, P. et al. Handbook of Enology: Vol 2: The Chemistry

of Winemaking. 2000

Saucier, C., Little and D. Glories, Y. 1997. First evidence of

acetaldehyde condensation products in red wine. . American Journal of
Enology and viticulture. 48. 370-373.
Schneider,V. 1998 Must Hyperoxidation: A Review American Journal
of Enology and Viticulture; 49: 65 - 73.

Singleton, V. L. 1987. Oxygen with phenols and related reactions in

musts, wines and model solutions: observations and practical
implications. American Journal of Enology and Viticulture. 38. 69-77.

Picking grapes for optimum

quality
The Challenge

Let's assume:

 It is mid-September in the Okanagan and you have a well-tended field of grapes getting
close to harvest.  You have done all you can to ensure moderate cropping, good canopy
management, and enough, but not too much, water—all factors that affect early and
even ripening.
 There are no practical problems to get in the way of selecting the best picking date. For
example:
o The weather is good (days are getting cooler but no rain is coming).
o Pickers are available.
o Shipping or storage is ready.
 You know the style of wine you want to make.

Now all you have to do is to decide when the grapes are at their
optimum. This may not be an ideal optimum, just the best you can
expect in this year in this field for this variety of grape and for
the style of wine you want to make.

In making your decision you will probably consider three things:

1. Appearance.
2. Numbers (sugar and acid levels).
3. Taste.

Unless you are lucky (and in the Okanagan you are not likely to be
lucky very often), you will discover that appearance, numbers, and
taste do not point to the same picking time. You have to make some
compromises. In the recent past, the compromises were usually biased
in favor of sugar and acid numbers. Today, top winemakers are more
likely to prefer taste and almost ignore numbers. But any bias or
compromise will affect how you make your wine and the eventual
quality of this wine.

Picking on Appearance

The development stages of grapes

Grapes go through several stages of development and changes in

appearance during a growing season.  To better understand the final
stages, a brief overview is helpful.

1. Flowering
o Early flowering is preferred as long as the threat of cold is past—best vintages
are often associated with an early start and an early harvest.
o Weather can interfere with complete and even pollination.
2. Green berry growth
o Berries use their own chlorophyll to grow and accumulate acid and sugar.
o Canes and roots also grow rapidly.
3. Arrest of green berry growth
o Berries have reached more than half their final size, but are still green and
hard.
o Berries have all the basic acid they will ever have, but only a small proportion
of the sugar.
o Roots and shoots also stop growing. This is usually the recommended time for
pruning of vines for canopy management and for thinning of clusters.
4. Beginning of véraison
o Berries soften and the color changes (véraison means turning, ripening).  The
color change happens dramatically in individual berries (in 24 hours), but
unevenly among and within clusters.  Even under ideal conditions it will take
two weeks for a vineyard to turn completely. The date of véraison is based on
an estimate of when half the grapes have turned.
o Véraison comes earlier with a low leaf-to-fruit ratio and with warm, sunny, dry
weather—best vineyards tend to start véraison early
5. Sugar/water accumulation
o Berries gain size and weight, adding somewhere between 25% (a dry season)
and 75% (wet season) of their green berry weight.  A 40-50% increase is
normal for good quality grapes.
 o Sucrose is manufactured by the leaves and transported to the berries where it
is hydrolyzed and stored as glucose and fructose.
o Acidity declines because of grape growth and the dilution of the existing store
of acid. Malic acid also declines because it is used in grape respiration.
o Flavors begin to accumulate, although most varietal flavor development comes
later in the season.
o Grapes typically do best at about 25°C—temperatures below 15°C and above
35°C will stop the plant from working.  Warm (not hot), sunny, dry ripening
seasons tend to produce the best vintages.
o Berries ripen unevenly among and within clusters.  Within clusters, berries at
the top (next to the stem) are riper (more sugar, less acid) than those at the
tip.  Larger berries have more seeds, and these ripen more slowly than small
berries of the same variety.
6. Arrest of phloem transport
o Plants "shut down" and berries stop accumulating water and sugar—this
happens naturally at the "end" of the season but can happen earlier from
stress.
o Berries soften noticeably and are easily "deformed" with a squeeze. It is easier
to separate seeds and pulp and to detach grapes from their stems
o Stems go brown (in many varieties) and seeds go from green to brown.
o Flavors continue to develop, because most flavors are synthesized in the
grape.
7. Dehydration
o Berries can lose 10% or more of their weight through water loss over a period
of a couple of weeks.
o Sugar levels rise because of water loss.  The rise in sugar can suggest that the
plants are still working when they aren't.
o Acidity is reduced.  Tartaric acid level can rise a bit because of dehydration, but
malic acid continues to be lost (grapes can lose almost all their remaining
malic acid during this period). Potassium activity in the grapes can continue to
reduce pH even though TA may not decline much.
o Flavor synthesis comes to an end and and flavor deterioration begins (these
periods overlap in ways not well understood).
8. Raisining
o Berry flavors continue to deteriorate.

The pros and cons of picking on appearance

Pro: Grapes can be picked successfully on the basis of appearance.

Biologically, a grape is finished ripening when the phloem (water,
sugars) stops moving from the leaves to the grapes. People can tell
when a grape is ripe just by looking at and touching it. There is
full and distinctive color, the grapes are soft, and the the stem is
brown.  In fact, the best picking practices depend on this ability
to recognize ripeness. If labor costs are no object, a field of
grapes is best picked two or more times.  (The best Riesling
vineyards in Germany may be picked four or five times.)  Even when a
vineyard is only picked once, pickers will often be instructed to
leave behind the unripe or overripe clusters.  Pickers make the
distinction solely on appearance. If pickers can make these
distinctions, an experienced viticulturist and winemaker, who can
also examine seeds, should be able to determine optimum ripeness
quite well just by appearance.

Pro: There are several problems when grapes go beyond biologically

maturity: (a) Sugar rises from dehydration (not always bad). (b)
Malic acid continues to be lost (usually bad) and potassium activity
continues to increase pH (always bad).  And (c) the grape continues
to soften and become more susceptible to bruising, breakage, and
bacterial activity (always bad), especially if the grape has to be
shipped any distance. Picking on a ripe appearance can help avoid
these problems.

Con: For economic reasons, fields are picked all at once. To

understand what this will mean for the must, you need to measure
average sugar and acid levels.  You may want a slight concentration
of sugar through dehydration.

Con: Flavors can continue to develop after the physical maturity of

the grapes. It can be particularly important to let some of the least
ripe grapes catch up a bit to the riper ones.

Picking by the Numbers (sugar and acid)

Testing for numbers

In Napa and Sonoma you can drop off a sample of 100 berries at the
ETS lab in St. Helena before 2:00 pm, pay $110 (US), and by early
evening have the results for the following tests:

 °Brix (% by weight of soluble solids, mainly fermentable sugar). During the time they are
testing, ETS says this commonly varies between 19 and 30, which means people start
testing only well along in the growing season. For another $60 you can get a precise
measure of fermentable sugar.
 Various acid measurements:
 o pH—commonly varies between 2.90 and 4.20
o Titratable Acid or TA (g/100ml)—commonly varies between 0.35 and 1.20
o Tartaric Acid (g/L)—commonly varies between 1 and 11
o Malic Acid (g/L)—commonly varies between 0.5 and 11
o Potassium (mg/L)—commonly varies between 500 and 4000
 Two yeast available nitrogen (YAN) measures:
o Ammonia (mg/L)—commonly varies between 50 and 400
o Alpha Amino Nitrogen (mg/L)—commonly varies between 20 and 400

The YAN numbers are chiefly useful to indicate how much nutrient you
are going to have to add to the fermentation. However, the sugar and
acid numbers can serve as a guide to picking the grapes.

Recommended numbers for wines

As grapes ripen, sugar levels rise and acid levels fall. Both sugar
and acid are important. Higher sugar levels are associated with grape
maturity and the higher levels of alcohol expected of superior
vintages. Higher acid levels (especially a low pH) provide better
wine-making circumstances. Thus, the trick in picking by the numbers
is to get to the optimum cross-over point, where sugars are high
enough and acid numbers still allow for good wine-making. The current
wisdom suggests you should be picking your grapes when the following
specs are realized.

Recommended sugar and acid levels for wine grapes

Wine Type °Brix TA pH
Source: Boulton et al. (1977), Principles and practices of winemaking,
as cited by van Schalkwyk and Archer (2000).
Sparking 18.0-20.0 0.70-0.90 2.8-3.2
White Table 19.5-23.0 0.70-0.80 3.0-3.3
Red Table 20.5-23.5 0.65-0.75 3.2-3.4
Sweet 22.0-25.0 0.65-0.80 3.2-3.4
Dessert 23.0-26.0 0.50-0.75 3.3-3.7

Sugar acid combination measures

Because sugar rises and acid falls as grapes mature, there can be a
trade-off in desirable specifications. Do you wait for more sugar at
the risk of losing acid or increasing pH? This dilemma has led people
to propose various measures of optimal maturity in which both acid
and sugar figure.

One such index relies on the square of pH to assess critical acidity,

and consists of °Brix * pH2. According to researchers who use it,
the desirable range for this index is between 200 and 270. For
example, if you took the mid-points of the recommended values for red
wine in the previous table (Brix 22, pH 3.3), this sugar acid index
value would be (22.5 * 3.3 * 3.3) = 245, or nicely in the middle of
the desirable range. For whites, the index value based on mid-point
numbers would be (21.25 * 3.15 *3.15) = 211.

Another approach to combining sugar and acid is to divide the °Brix

by the TA in g/100ml. This is called the sugar acid ratio. Again it
is possible to generate ideal scores by using the mid-points of the
recommended ranges in the previous table. (NOTE: If you are reading
European sources, don't get confused by their higher values for the
Sugar-Acid Ratio. Europeans often measure acidity in the equivalent
of stronger sulfuric acid.)

The two combination measures are illustrated in the following table.

Measures combining sugar and acid

Wine Type B-pH Index B-TA Ratio
Source: Van Schalkwyk and Archer (2000) discuss the Index; Bisson
(2001) discusses both the India and Ratio.  Both these sources are
drawing on the work of thers.
Formula Brix * pH2 Brix ⁄ TA
Ideal White Table 211 28
Ideal  Red Table 245 33
Desirable Range 200-270 27-34

The numbers for our recent grapes

The problem with the numbers in the above tables is that we seldom
see them in the grapes we get. Our 2003 Black Sage grapes illustrate
the issue. The following table is based on typical (median) readings
by Club members.

Black Sage numbers for selected 2003 grapes

Grape Variety °Brix TA pH Brix-pHIndex Brix-TARatio
Source: Typical (median) values measured by at least three Club members, usually on very small
lots. Initials after Grape Variety are for the growers.
Chardonnay (JB) 22.5 0.74 3.39 259 30
Gewürztraminer (JB) 21.9 0.69 3.45 261 32
Pinot Blanc (JB) 21.0 0.47 3.74 294 45
Pinot Gris (RF) 19.8 0.55 3.71 273 36
Riesling (AMi) 22.4 0.69 3.27 240 32
Sauvignon Blanc (AMa) 22.9 0.64 3.66 307 36
Semillon (MW) 22.7 0.57 3.42 266 40
Cabernet Franc (LM) 22.4 0.54 3.64 297 42
Cab Sauv 15-169 (DC) 25.5 0.37 4.03 414 69
Merlot (RF) 23.1 0.43 3.76 327 54
Pinot Noir 115 (RF) 22.5 0.47 3.77 320 48
Pinot Noir 667 (RF) 20.0 0.49 3.71 275 41
Syrah (BF) 24.8 0.51 3.95 387 49

These parameters for 2003 Black Sage grapes suggest at least three
observations:

 Only Chardonnay, Gewürztraminer, and Riesling meet most of the suggested specs, but
even Chardonnay and Gewürztraminer have relatively high pH levels and this puts then
at the high end of the acceptable index range, where reds are more to be expected. 
Semillon also comes close to meeting the suggested specs.  Most people would consider
pH more important than TA, and the Index value for Semillon seems to capture the
quality of the Semillon better than the simple Brix/TA Ratio does.
 The general "numbers" problem with these Black Sage grapes is low acidity—a low TA
and a high pH. Only the Cab Sauv and perhaps the Syrah are clearly overripe in terms of
sugar. The phenomenon of low acidity without excessive sugar is a well-known
consequence of overripe grapes or a warm climate because of the respiration of malic
and other organic acids. This past summer in the Okanagan, high heat caused vines to
"shut down," stopping sugar production but increasing acid loss.
  The Index and Ratio numbers are not that bad for Pinot Gris and Sauvignon Blanc, but
these were the white grapes that arrived in poor condition with obvious VA problems. 
Appearance does matter! 

The pros and cons of picking by the numbers

Pro: Great, long-lived wines almost always come from grapes with
"good numbers" to start with. pH is particularly important for good
winemaking, and it is something that is very difficult to correct
even by adding acid later. No one wants to add water to an overly
sweet must and thereby dilute the intensity and flavor of the wine.

Con: The right numbers may occur fairly regularly in classic wine
areas like Bordeaux, but you will not get them in hotter, drier wine
areas like the Okanagan (or South Africa or Australia). In such
conditions, sugar development usually outpaces flavor development.

Con: Ultimate wine quality (as determined later by tasting) is not

well correlated with the numbers at harvest.  This is particularly
true of sugar levels. You can always adjust the numbers prior to
fermentation, by adding water or sugar or acid or other means. In
some commercial areas there will be restrictions on additions, but
there are no restrictions on the home winemaker. You cannot correct
for flavor that is not there.

Picking on Taste

Basic information about taste

Taste and its development in grapes and wines is not well understood
scientifically, but there are some basic things we do know.

 There are hundreds of different molecules that contribute to the flavor and aroma of
grapes and wine.  Some of these have a a sensory impact at concentrations of only a few
parts per billion.
 Flavors develop over the course of grape ripening, but much of the typical varietal flavor
and intensity comes very late.  In one Oregon test (Watson), Pinot Noir was made from
the same grapes, but one picking was one to two weeks later than the other.  Later,
when the finished wine was tested, more than twice as many flavor compounds were
 isolated from the grapes picked later—a rather astounding differences. Of course, more
flavor elements do not necessarily mean a better flavor, but in this case the later picking
was definitely judged to be better in overall flavor and aroma.
 Most flavor compounds are synthesized in the grape itself and this synthesis can
continue after grapes are biologically ripe and phloem flow has stopped.
 Identifiable tastes change as grapes develop. The exact tastes differ according to variety,
but the usual sequence is more vegetative tastes early and more fruit tastes later.  The
sequence for the appearance of flavors in Cabernet Sauvignon goes something like this:
1. Vegetation (plant matter)
2. Herbaceousness (straw, herb, vegetal, tobacco)
3. Unripe fruit (green apple, citrus rind)
4. Red fruit (cherry, strawberry, raspberry, cranberry)
5. Black fruit (plum, blackberry, black cherry)
6. Jam (prune, date, raisin)

This is not necessarily a sequence of tastes. Tastes deteriorate at different rates than new
ones appear. It is possible to have herbaceous and jammy tastes at the same time. This
leads Bisson (2001) to suggest that it may be better to taste for the absence of the
"unripe" tastes you want to avoid rather than to taste for mature complexity.

 Many of the flavors that are most desired in a mature wine are not taste-able in the
grape. They are locked-up with glucose in complex molecules that are technically known
as glycosides and more popularly known as flavor precursors. Some of these flavor
precursors can be unlocked with pectic enzyme, but some emerge only with
fermentation or aging. What can be tasted most easily in the grape are the fruit flavors
associated with a young wine.

o NOTE: There is a test called the Glycosyl-Glucose Assay (or G-G Assay) that is
used to establish the total level of glycosides in juice or wine. At one point in the
1990s there was hope that this test would provide a more objective basis for
establishing the flavor potential of grapes. It has not so far proved to be a magic
bullet.

The pros and cons of picking on taste

Pro: There is no point in making wine with a green or stemmy taste.

Even if you have to compromise your winemaking and settle for a wine
that is not for the ages, it is better to have good taste and aroma
than not. And while there is no science of taste, experience with
winemaking, particular varieties, and specific terroir and vineyards
can provide valuable information about when to pick grapes at their
optimum.
Con: If taste comes at the cost of "overripe" grapes, which spoil
easily during shipping and have a high pH and potassium level, you
could end up with poor and short-lived wine, even though the taste
potential was there. And there is always the risk that one is waiting
for flavors that are not going to appear, so that more time just
means more deterioration.

Con: Tastes are notoriously elusive and subjective at the best of

times. And many of the tastes we want in a wine are not discernable
in the grape—they only emerge later in the fermentation and aging.

Sources

Bisson, Linda (2001). "Optimal grape maturity," Practical Winery,

July/August. Bisson is at UC-Davis. This in a good overview.  It is
available at: http://www.practicalwinery.com/julaug01p32.htm

ETS Laboratories (2001). "Tools for Grape and Must Analysis." Useful
explanation of grape testing terminology. the 2003 version of this
paper is available at: http://www.etslabs.com/pagetemplate/blank.asp?
pageid=195. There are other useful technical papers available through
the ETS website.

Jackson, Ron S. (1994). Wine Science: Principles and Applications

(Academic Press).

Peynaud, Emile (1984). Knowing and Making Wine, second French

edition, trans. by Alan Spenser. John Wiley and Sons.

van Schalkwyk, Hanno, and Eben Archer (2000). "Determining Optimum

Ripeness in Wine Grapes," Wynboer.
http://www.wynboer.co.za/recentarticles/0500optimum.php3

Watson, Barney (2003) "Evaluation of Winegrape Maturity," in Edward

H. Hellman (ed.), Oregon Viticulture, Oregon State University Press,
pp.235-245. This excellent chapter is available as a download at
http://winegrapes.tamu.edu/grow/maturity.pdf

Zoecklein, Bruce W. (2000). "Yield and Quality," Vintner's Corner,

15:5, September-October. Available at
http://www.fst.vt.edu/extension/enology/septoct00.html
Prepared by Rod Church as a "backgrounder" for a meeting of the Nanaimo Winemakers
on March 5, 2004.

Master of Faults
Sam Harrop, a co-chair at the
International Wine Challenge in
London, spends half of his time
at the faults table. A New
Zealander by birth, a London-
based international winemaking
consultant by trade, Harrop
loves complex sulfides as much as minerality...
which, for him, are often one in the same.
   For the ten years before Harrop started with
the Challenge, the organizers had been tracking
corked wines, flagged by a range of different
judges. In those years, Harrop says, wines
presumed to have cork taint ranged between six
and seven percent of the total. Four years ago,
when Harrop joined the competition, he saw a
range of faults beyond cork taint coming through
from the panels. "On the hoof, I created a little
program on the computer and started logging
them," he recalls. Now he tracks faults through
14,000 to 15,000 bottles over the course of the
annual two weeks of judging.
   "I found in my first year that the figure for cork
taint was closer to three percent. Total faults
were around seven percent–a similar figure to
previous years; so the balance must be other
types of faults. The cork industry has been going
on for decades that seven percent is a load of
rubbish. And though I'm no apologist for the cork
producers, sure enough, tasters were flagging
cork taint when it was actually Brettanomyces."
   Harrop's interest in faults began when he was
winemaker at Marks & Spencer, the up-market
British retail chain. "I didn't actually make wine,"
he says. "But when you're a retailer selling five
million cases a year of own-label wine, you own
the product and have to make sure the faults are
minimized.
   "I started studying for the MW in 2001 and
realized that faults can actually be the making of
a wine," he recalls. Soon, for his MW dissertation,
he began to research Brettanomyces and its role
in high-quality syrah. "I'd been told that Brett was
the pox–its nickname in Australasia, where it's
treated as some kind of plague. Here I was in the
Rhône hearing winemakers wax lyrical about
wines I was taught to hate. I collected syrahs from
some of the most well-known producers around
the world, tasted them with MWs and asked them
to rate and rank them, and whether they
perceived Brett. Then I analyzed the wines for
volatile phenols. It was clear that just a little bit–
just over threshold–is fantastic. In the highest-
scoring wines, it brings out the varietal character
and complexity. Too little and the wine is simple.
A little Brett is good; too little is less good. On
the other end, the worst wines in the ranking
were full of volatile phenols."
   Harrop believes the New World winemakers who
insist that any Brett is bad Brett are just as
misguided as those who let it run rampant. "It's
easily controlled if you monitor it," he says, "and
easily stabilized through filtration. I don't know
who spread the word that unfiltered wines were
better. I've done trials and the filtered wines
come out best, both in texture and expression."
   Harrop surveys the damage from the Apple iMac
surrounded by white-sheathed bottles at his
faults table. "A lot of people think I'm some kind
of masochist. But I have the last laugh." For
Harrop, faults are the way not only to
understanding great wine, but to making great
wine. At below, he describes the marker for each
of the major faults, and some of the trends he has
found over time.

   Pavlova (a.k.a. Reduction) Maybe instead of

reduction, it's better to say complex sulfides.
These compounds express themselves under
reductive conditions. They vary from h2s (rotten
eggs) to DMS (sweet eggs–I often describe it as
Pavlova, the Kiwi dessert). And there are various
mercaptans, some flinty, some cabbagy. Some of
the great wines of the world are not sulfidic, but
on the cusp. To be honest, I struggle with it as a
good thing, with some Chablis or Côte d'Or
producers, who shall remain nameless. Cork taint
in years past has been three percent of total
faults, but as of the end of the first week of this
year's judging, reduction was the biggest category
in terms of total faults, at 26 percent of the
bottles; cork taint was at 24 percent.
(Preliminary results from the entire tasting show
cork taint and reduction contribute equally to
total faults, at 25 percent.) There's definitely
some correlation between more screwcap wines
and more reduction.
   If you taste sauvignon blanc in tank, it may be
full of guava. Then it goes into bottle, without
the winemaker managing the sulfur, the dissolved
oxygen, and you can end up with this cabbage or
green onion character. We've seen about 2
percent of all screwcap entries having this green
scallion character. It's not the closure, it's how
the wine was processed. Cork taint is terrible,
but the good thing is that it is random. Reduction
is not. If you have green onion in one bottle of
sauvignon blanc, it's in every bottle. The effect
on total wine in bottle is much greater if it's not
controlled.
   Cloves & Bullshit (Brettanomyces) Brett is about
15 percent of the total faults this year, up from
the last three years' average of 12.4 percent.
People describe the smell of Brett, but it's the
volatile phenols that you're smelling. Brett is only
a microorganism that can produce these
compounds–4–ethylphenol (4-EP) and 4–
ethylguaiacol (4–EG). 4–EP is the barnyard, fecal
scent. 4–EG is the clove, or country smoke,
burning, wet green wood. I think the reason we're
up from previous years in Brett is that we're
seeing more expensive wines in heavy bottles.
[The heavy bottles are often markers for wines
made from superripe fruit, which translates to
high alcohol and some residual sugar. Also, they
are usually fermented or aged in new oak; all that
unfermented sweetness is an invitation to Brett.]
It's a better reason not to buy wine in heavy
bottles than the carbon footprint: You might be
getting a dud.
   Balsamic Vinaigrette (Volatile Acidity) This can
be confused with oxidation because VA comes
from an oxidative reaction. But it's distinct–the
wine is not bruised but the volatile acidity is
excessive. It can work in elevated levels in some
wines, like Burgundy and zinfandel, by
heightening the intensity of the wine.
   Excessive volatility can come about through a
stressed fermentation, primarily if fruit sugars
are high. At 14 degrees or above, it's a risk.
Alcohol is toxic for the yeast; if you don't have
sufficient flexibility in the membrane of the yeast
cell, it fractures and more toxic activity develops.
The yeasts struggle and produce elevated levels
of VA and sulfides. Then the yeasts may die,
leaving residual sugar, which may allow Brett to
feed. Brett also feeds on the sugars in new oak.
   Hessian Cloth (Cork Taint) TCA (cork taint) is
dank and musty, like Hessian cloth–when you take
your socks off at the end of a run and leave them
in the bathroom for a few days and then go back.
   One thing that's helped is the Diam closure,
made from agglomerated cork. It's the same
process by which coffee is decaffeinated–
supercritical co2 extraction–applied to cork
granules, which are then glued together. We're
not only seeing more screwcap entries, we're also
seeing the number of Diam entries increasing.
   Stale Bread (Oxidation) We've seen the number
of oxidized wines increase this year. It ranges
from wines with stale, dank flavors to something
almost like stale bread, a stale, leesy character to
just completely screwed wine with no fresh fruit.
The wine is bruised.
   Death Warmed Over (Biogenic Amines) Other
faults include microbial or biogenic amines, like
putricine or cadaverine–which smells like a dead
body. That's a fault that has no place in wine.
   There's SO2, which is more of a sensation than
a scent–you can feel it in your nose and temples.
It has an eggy scent in high doses. And grey rot:
Grandma's dusty curtains, that mixed with a
green, unripe character.

Lactic Acid Bacteria and Wine Spoilage*

By Dr. Murli Dharmadhikari

Lactic acid bacteria (LAB) are responsible for many fermented foods such as sauerkraut, pickles and
yogurt. They have also been isolated from wines at various states of vinification. In wines they are
responsible for malolactic fermentation (MLF) which can be beneficial in some cases and undesirable in
others. Besides conducting MLF, these bacteria under certain conditions can also cause undesirable
changes in wine flavor which renders the wine undrinkable. Many species of LAB do not conduct MLF
and their growth in wine can cause some serious wine spoilage.

Nature of Lactic Acid Bacteria

Lactic Acid bacteria found in wine belong to three genera, namely: 
    1. Leuconostoc - Heterofermentative cocci, oval or spherical, occur in pairs or chains. 
    2. Pediococcus - Homofermentative cocci, often found in tetrads. 
    3. Lactobacillus - Homofermentative or heterofermentative rods, found singly or in chains.

These organisms are gram positive, catalase negative, nonsporing cocci, coccobacilli or rods. They are
microaerophilic that means they grow well under conditions of low oxygen content. Since they can grow
under low oxygen conditions, they can grow throughout the wine (as opposed to on the surface of the
wine) even though the container is kept full. The bacteria can metabolize sugars, acids and other
constituents in wine and produce several compounds. Some of these are undesirable and constitute
spoilage.

Source of Lactic Acid Bacteria

The bacteria can be found on the surface of grapes and grape leaves. During the harvest, the bacteria
gain entry into the winery with the grapes. Their population on the surface of the fruit is generally low
and it depends on the level of maturity and the condition of the fruit. Another source of these organisms
in a winery is the contamination equipment. These may include pumps, valves and storage containers.
Wooden barrels which are often difficult to clean and sanitize can be a source of these bacteria if the
barrels contain MLF wine and were not properly cleaned.

Occurrence of Lactic Acid Bacteria at Various Stages of Vinification

At crush the bacterial population is small, about 103 to 104 cells/mL. Species belonging to genera
Lactobacillus, Pediococcus and Leuconostoc have been found to occur at this stage. In the next stage,
i.e. alcoholic fermentation, the population of these bacteria declines. This may be due to the competition
by yeast and formation of ethanol and sulfur dioxide by the yeast during the alcoholic fermentation.
Following the alcoholic fermentation, the surviving bacteria grow vigorously and conduct MLF. The cell
population can reach as high as 106 to 108 cells/mL. Usually species of Leuconostoc grow and conduct
MLF but in the case of high pH wines (pH 3.5 and above), species of Pediococcus and Lactobacillus
could be involved in MLF.

After MLF the fate of lactic acid bacteria depends on wine composition and how the wine is handled. If
the wine pH is high (>3.5), and the SO 2 level is inadequate, then the spoilage causing species of LAB
can grow and spoil the wine. For this reason special attention should paid to the wines during storage
after MLF.

Nature of Spoilage by Lactic Acid Bacteria

The nature and the extent of wine spoilage by LAB depends on several factors such as the type of
bacteria, composition of the wine and vinification practices. Based on the substrate used lactic spoilage
has been classified as follows: 

1. Fermentation of Sugars
LAB, including those involved in MLF, metabolize sugars such as glucose and fructose, and produce
lactic acid and acetic acid. The resulting wine acquires a sour vinegar-like aroma due to high VA levels.
This is a serious spoilage and occurs in must with stuck fermentation or wines with higher residual
sugars (sweet wines).

A less serious form of lactic spoilage can occur in dry wines. In these wines the LAB utilizes pentose
sugars, trace amounts of glucose and fructose, and produces lactic and acetic acid as a by-product.
When sugars are attacked by LAB, lactic and acetic acids are produced. Formation of these acids
increases the titratable acidity and lowers the pH. The decrease in pH restricts the growth of those
organisms.
 
2. Degradation of Glycerol
Breakdown of glycerol by LAB results in the formation of lactic acid, acetic acid and acrolein. The wine
smells acetic, butyric and acquires a bitter taste due to acrolein. 

3. Fermentation of Tartaric Acid.

In this kind of spoilage, the LAB ferments tartaric acid and forms lactic acid, acetic acid and carbon
dioxide. Degradation of tartaric acid occurs especially in wines with low acidity and high pH (pH above
3.5). The titratable acidity is further reduced and the wine acquires an acetic aroma and disagreeable
taste. In advanced cases the wine is sometimes referred to as mousy.

4. Fermentation of Citric Acid

Citric acid content of a wine can decrease during MLF. Depending on the species of bacteria and the
wine pH. Citric acid degradation has been positively correlated with the formation of diacetyl and
acetone as well as acetic acid.
 
5. Ropiness
Certain species of Leuconostoc have been found to produce dextran slime or mucilaginous substances
in wine. The wine appears oily and may not necessarily have high volatile acidity.
 
In general, the lactic acid bacteria (as a group) are involved in the fermentation of malic acid and other
wine constituents. Their activity results in the formation of several components that impart off flavors to
wine. Some of the terms used in describing the lactic spoilage in wine include: acetic or sour, buttery,
cheesy, sauerkraut-like, bitter, pickle aroma, mousy and geranium.

6. Other Off Aromas

Very unpleasant odors associated with lactic spoilage include mousy and geranium-like aromas.

The mousy aroma has been attributed to the formation of a compound called acetyltetrahydropyridine.
Two species of lactobacillus have been shown to produce these mousy odor compounds.

Sometimes a wine can develop a geranium-like odor which makes it undrinkable. This odor is caused by
a compound known as 2-ethoxyhexa-3, 5-diene. This compound is produced from the decomposition of
sorbic acid by the LAB. In sweet wines sorbic acid is often added to prevent the growth of unwanted
yeast (yeast growth can cause refermentation). When the sorbic acid is attacked by LAB, 2-ethoxyhexa-
3,5-diene is formed which imparts the geranium-like odor to the wine. To prevent this odor the growth of
LAB in sweet wines containing sorbic acid should be controlled.

Factors Influencing the Growth of Lactic Acid Bacteria (LAB) in Wine

We have discussed the nature and extent of wine spoilage by LAB. In order to prevent wine spoilage by
these micro-organisms, it is important to know the various factors that influence the growth of these
microbes in wine so that one can manipulate these factors to reduce the risk of spoilage. The three
important factors that deserve some consideration include: 

    1. Must/wine composition 

    2. Vinification practices 
    3. Interrelationships with other organisms

Must/Wine Composition
Wine pH - Wine pH is one of the most important factors influencing the growth of LAB. It affects the
initiation and duration of malolactic fermentation MLF, it influences the type of species of bacteria that
may develop in wine and it also affects the metabolic behavior of the organism and thereby determines
the kind of by-products formed as a result of bacterial activity.
In the wine pH range of 3.0 to 4.0, the time needed for the completion of MLF decreases with an
increase in pH. Bousbouras and Kunkee (1971) reported that at pH 3.15 it took 23.4 weeks to complete
MLF; whereas at pH 3.83, it was completed in just two weeks.

Many researchers have noted the effect of pH on the species of bacteria that can grow in wine.
Generally at pH below 3.5, the MLF is often dominated by Leuconostoc, whereas; above pH 3.5,
species of Pediococcus and Lactobacillus seem to flourish. It should be noted here that many strains of
Lactobacillus are involved in wine spoilage.
Another important pH effect not commonly realized is the effect of pH on the metabolic behavior of the
organisms. For example at pH 3.5 and above, LAB are more likely to decompose sugars, tartaric acid
and citric acid. As mentioned earlier, fermentation of sugar leads to higher volatile acidity (VA) levels in
wine. From the foregoing discussion it should be obvious that controlling wine pH is one of the keys to
controlling wine spoilage by LAB.
Sulfur dioxide (S02) Sulfur dioxide is an effective germicide commonly used by the winemakers to
control the growth of harmful bacteria. The SO2 in wine exists in free and bound forms. All these forms
remain in an equilibrium which is influenced by pH. Concentration of the molecular SO2 form of free SO2
which is also the most toxic form increases with a decrease in wine pH. Therefore, maintaining low pH is
helpful in making SO2 the most effective tool to control LAB. The bound form of SO 2 has also been
reported to have a detrimental effect on LAB. In wine, SO 2 is bound to certain carbonyl compounds such
as acetaldehyde. When LAB attacks the carbonyl compound, the bound SO 2 is released. It is this
liberated free SO2 that prevents further growth of the bacteria.

SO2 is an effective germicide and concentrations of 0.8 ppm molecular SO 2 will be adequate to control
the growth of LAB in wine.
 
Alcohol - Generally LAB can survive and grow in table wines. There is some variation between various
species regarding alcohol tolerance. For example; Lactobacillus trichods has been found in wine
containing 20% alcohol. The alcohol tolerance is influenced by pH and storage temperature.

Oxygen and Carbon Dioxide - Although microaerophilic conditions are desirable for the growth of LAB,
the evidence suggests that a small amount of 02 may be necessary. It is however, widely recognized
that the presence of CO2 stimulates the growth of LAB. This may be a factor stimulating MLF in wines
left on the lees which would contain a fair amount of dissolved CO 2. Kelly, Asmudson, and Hopcroft
(1989) concluded that this was likely due to low levels of O2 as they observed the same effect using N2.

Nutrients - LAB require a source of energy such as carbohydrates and inorganic salts. In addition they
also need other growth factors such as vitamins and amino acids. Yeast autolysis (which occur during
prolonged lees contact) resulting in increased nutrient content can render a young wine prone to attack
by LAB.

Vinification Practices
Many vinification practices can influence growth of LAB in a winery. Some of the important practices
include: fruit condition, must treatment (adjustment), clarification, fermentation conditions, skin contact
time (in case of red wine), lees contact, wine clarification, storage and winery sanitation.

Sound fruit has a low population of LAB on the surface, therefore; using clean and healthy fruit is
important in reducing the number of microbes that would enter the winery at harvest. Sulfur dioxide is
often added at the crush. It is one of the most effective measures in controlling the growth of LAB.
Winemakers not sulfiting the must at crush in order to reduce sulfite's in wine are taking a bigger risk in
exposing their wines to bacterial spoilage. High pH musts usually contain low acid levels. The acidity
and pH of such a must should be adjusted with tartaric acid additions before fermentation. This will
enable fermentation to occur at low pH, and thus reduce the chances of spoilage by LAB. Clarifying
white must by settling or other means reduces the suspended solids in the must. This practice is
suggested for discouraging MLF in white wine.
 
Fermentation conditions affect the growth of LAB. For example, in case of a stuck fermentation, LAB
can attack sugar and increase V A levels in wine. Controlling the fermentation so that it proceeds rapidly,
evenly, and reaches dryness, is a sound enological practice to prevent any damage from LAB. A young
wine left on the lees for a long time will be prone to MLF. This is due to the availability of nutrients
released by yeast autolysis and a reduced CO2 environment. For controlling LAB, early racking is
recommended. Wine clarification, especially using tight filter pads or a .45 micron membrane filter will
reduce the bacterial population and consequently the chance of spoilage.

Of all the winery practices, cleaning and sanitization of equipment and containers is one of the most
important practices that a winemaker must employ to control the wine spoilage.

Interrelationships with Other Organisms

LAB does not seem to grow well in must during alcoholic fermentation. It seems that yeast has an
inhibitory effect on the growth of LAB. This could be due to several reasons such as competition and
depletion of nutrients by yeast, competition by natural yeast flora (eg. Pichia), formation of ethanol, SO 2
and other inhibitory compounds by the yeast.

Contrary to the antagonistic effects of the yeast, there are however, some reports that suggest that
yeast may have stimulatory influences on the growth of LAB. For example, prolonged contact with the
lees can result in enrichment of young wine by yeast autolysis. This in turn can stimulate
the growth of LAB.
 
Other microorganisms such as Botrytis cineria and acetic acid bacteria have been reported to have a
stimulating effect on LAB. LAB are often found in association with acetic acid bacteria and there is some
evidence indicating a symbiotic relationship between these organisms. 

Bacteriophages are known to destroy the LAB. These phages have been isolated from wine. Not much
is known about the inhibitory impact of these phages on LAB in wine and its influence on wine quality.

Recommendations to Winemakers
Since LAB are involved in MLF as well as wine spoilage, a winemaker needs to decide up front whether
to encourage MLF.

If the choice is to encourage MLF (and avoid spoilage), then the following recommendations should be
followed and MLF must be conducted under controlled conditions. 

    1. Use clean, healthy and high acid fruit. 

    2. Add a small dose of SO2 at crush. (About 25-30 ppm based on must pH.) 
    3. Adjust the must pH if necessary. A pH range of 3.3 to 3.5 is desirable for MLF. Since MLF causes
an increase in pH, it is advisable to conduct MLF at the lowest must pH as practically possible. 
    4. Inoculate the must with a pure starter culture of ML bacteria. The preferred time of inoculation is
the 2nd or 3rd day after the alcoholic fermentation has begun. Low ethanol, low SO 2 and warm
fermentation
conditions favor MLF. 
    5. Take precautions to avoid a stuck fermentation. This would include not using overripe or moldy
grapes, using a good dose of vigorously growing, pure culture of yeast, adding yeast nutrient and
maintaining controlled temperature conditions. Do not allow fermentation temperature to exceed about
30° C or 86° F. 
    6. Monitor MLF and as soon as it is completed, treat the wine to prevent further growth of any LAB.

If the winemaker's choice is not to encourage MLF then the following recommendations should be
followed as a guide to prevent MLF as well as spoilage due to LAB:
    1. Use sound fruit for making wine. 
    2. Add SO2 at crush, about 50 to 75 ppm based on must pH. 
    3. In a low acid and high pH must, add tartaric acid to bring the pH to 3.3 or lower. 
    4. In the case of white wine, clarify the must (reduce suspended solids) before fermentation. 
    5. Control fermentation temperature. Use well prepared, pure culture yeast starter. Use yeast nutrient
if needed. 
    6. In the case of red wine, prevent must temperature from exceeding 85° F. Punching the cap and
keeping the cap moist is important. 
    7. After the must is fermented dry, promptly rack the wine off the lees and add enough SO 2 to attain
0.8 ppm molecular SO2 level. 
    8. Clarify and stabilize the wine and store in clean containers. 
    9. Clean and sanitize equipment and containers before processing the wine. 
    10. Sterile filter and store wine at cool cellar temperatures. 

SUMMARY
Lactic acid bacteria are present on grapes, contaminated winery equipment and storage vessels. Some
of the LAB primarily decompose malic acid and under certain conditions, attack sugar and malic acid.
These are often involved in MLF and rarely in wine spoilage. Certain other LAB grows in low acid
conditions; metabolize sugars (pentose), tartaric acid and glycerol. These are more dangerous
organisms and cause serious spoilage. Conditions such as moldy fruit, low alcohol, low SO 2, high pH
(3.5.and above), low acidity, presence of fermentable sugars and warm temperatures such as 25° C
(78° F), favor the growth of LAB and can cause wine spoilage. Maintaining an adequate SO 2 level, low
pH, and sanitary conditions during processing can prevent the spoilage.

Literature cited.
Bousbouras, George E. and Ralph E. Kunkee.
1971. Effect of pH in malolactic fermentation in wine. Am. J. Enol. Vitic. 22:121-6.
Kelly, W.J., R.V. Asmundson, and D.H. Hopcraft. 1989. Growth of Leuconostoc olnos under anaerobic
conditions. Am. J. Enol. Vitic. 40:277-282.

*Previously published in Vineyard and Vintage View, Mountain Grove, MO