BY Kay Bogart, Linda Bisson,
Department of Enology & Viticulture
University of California, Davis
Researchers have gained more insight, in the last several years, into
one of the main categories of compounds responsible for vegetal or
herbaceous characters in some wnes, chiefly the class of pyrazines
that create the familiar green pepper/green bean aromas associated
with many Bordeaux-type varieties.
Although these pyrazinaes, chiefly methoxypyrazines (MP), are not
the only compounds responsible for many facets that define a grape
or wine as vegetal, they are the most common and the most
well understood today. They may also act synergistically with other
compounds to modify aroma and flavor in many different ways.
Below, we will summarize and present recent findings that will allow
readers to gain a better understanding of what these pyrazines are
and how they might be manipulated to meet specific quality standards
or style considerations.
We have reviewed pertinent literature on the methoxypyrazines
of grapes and the influence of viticultural and enological practices
on their formation and stability as a starting point to address
the persistence of vegetal characters in winegrapes and wine. Some
results of a preliminary nature reported from France, New Zealand,
and Oakville in Napa Valley are included. Such findings need to
be evaluated under the readers local growing conditions to
determine what role, if any, they might play in the characteristics
that define a wine as vegetal.
It is important to clarify if we are, indeed, actually
seeing the persistence of vegetal characters rather than a change
in the definition of enological ripeness. Perhaps those characters
that defined winegrapes as mature in the past may not be included
in the current definition of maturity, but instead, may be indicators
of a lack of ripeness. The characters or their evolution did not
change, but the definition of what they represent may have changed.
This is likely the case with pyrazines.
MPs are one of a group of pyrazines that are widely
distributed in nature. In winegrapes, they are present in such small
quantities that, until recently, their chemical detection and enumeration
has been difficult. Their threshold in red wine, especially Cabernet
Sauvignon, Cabernet Franc, and Merlot, has been reported to be as
low as 10 to 15 parts per trillion.1
But within the range of their occurrence, their aroma can be overpowering
MP concentrations in wines have always shown a consistent
relationship between winegrape variety and grapegrowing conditions.
Since MP concentrations are known to decrease during normal winegrape
ripening, a high MP concentration in grapes at harvest has become
commonly associated with a lack of ripeness and usually has a negative
impact on wine aroma and flavor qualities.2
Because the fruit with vegetal characteristics is
most often physiologically immature, there is often a corresponding
lack of acceptable color development and an undesirable, harsh tannic
quality in the wines produced from these grapes.
Winemaking decisions / methoxypyrazine content
It is important to understand how winemaking procedures may or may
not impact the pyrazine content in a finished wine. In a recent
published study, researchers in France were able to show that MP
in grapes is highly extractable in the traditional winemaking process.2
MP concentration in Sauvignon Blanc wine was shown
to be independent of pressing conditions of the juice. The final
MP concentration in the Sauvignon Blanc wine, in this study, differed
insignificantly from the first free run juice at pressing. Settling
of the juice did, however, decrease MP content by almost half in
the clarified portion. So, with white wines, settling may reduce
With Cabernet Sauvignon in the same study, all pyrazine,
found in the wine after racking, had already been extracted from
the grapes within 24 hours of crushing, before alcoholic fermentation
began. This final pyrazine concentration in the wine was not affected
by different cap management regimes or amount of time the wine spent
on its skins.2 This shows MP to
be highly extractable in winegrape must.
Press wines, however, have been shown to contain higher
levels of MP. This suggests that a fraction of MP remains in the
skins and is extracted during rigorous pressing. In the same study,
after ageing Cabernet Sauvignon for three years in a dark cellar,
no significant change in MP content was recorded. Therefore, dependence
on ageing or time in bottle to decrease vegetal character in wine
is probably not an effective strategy.
An interesting observation from some researchers is
that thermovinification of red grapes to promote extraction of phenolics
and destroy oxidases also leads to a decrease in MP concentration
in heat-treated wines.1 The wines
studied were heated to 60†C to 80†C for a short period of time.
The resultant MP content was decreased to a concentration below
the sensory threshold, so the original vegetal character was no
In concurrent laboratory trials, MP was found to volatilize
and dissipate into the headspace after heating to above 50†C. This
could be useful in dealing with grapes that are harvested before
optimal ripeness due to weather or growing conditions. Using thermovinification
conservatively could, in certain cases, lead to more highly colored,
fruity, less vegetal wines.
How vineyard decisions
affect wine MP content
Since it can be shown that winemaking procedures probably have a
minimal impact on pyrazine concentrations, it is most appropriate
to focus on recent viticultural research for clarification on pyrazine
management. This research seems to confirm that grape ripening conditions
in the vineyard are primarily responsible for the methoxypyrazine
content in wines.1 This means
that if your target wine style does not include a vegetal component,
you will have to start your winemaking process very early by understanding
what happens in the vineyard and how these compounds can be manipulated
A very significant study, recently completed at the
University of California, Davis, investigated the effects of yield
manipulation through vine pruning and cluster thinning on the resultant
sensory attributes and actual MP concentrations of Cabernet Sauvignon
wines.3 Researchers subjected vines to six different winter-pruning
treatments and eight different cluster-thinning treatments; thinning
being imposed at veraison.
This research is particularly significant because
the large number of treatments resulted in an equally broad range
of vine yields. In 2000 and 2001, vines were pruned to 12, 18, 24,
30, 36, or 48 buds per vine, with 24 buds per vine as the control.
Separately, in 2001, vines were pruned to 24 or 48 buds per vine
and clusters were removed at veraison to leave 12, 24, 36, or 48
clusters for the 24-bud vines and 48, 64, 72, and 96 clusters per
vine for the 48-bud vines.
Yields ranged from 1.75 tons per acre to 9 tons per
acre. In both experiments, treatments were imposed in a randomized
complete block design with six replications. All wines were made
in triplicate. Also, lending to importance of this research is the
fact that researchers used trained judges employing descriptive
analysis techniques to determine the sensory attributes of resultant
Analysis of variance and principal component analysis
showed that pruning vines to a very low bud number, produced wines
that were perceived by the panel of judges to be higher in vegetal
aroma and flavor, bell pepper aroma, bitterness, and astringency
than wines from high yield vines.
Wines from vines that carried more crop (higher bud
numbers) were consistently rated as higher in red/black berry aroma,
jam aroma, fresh fruit aroma, and fruity flavor. Regression analysis
showed that, in general, vegetal attributes decreased in intensity
and fruity attributes increased in intensity as bud number and yield
In contrast, there were few differences detected in
wines made from various cluster-thinning treatments, though the
yield range was greater. They concluded that Cabernet Sauvignon
aromas and flavors respond to yield manipulation, but do so significantly
only when yield is manipulated early in fruit development (winter
pruning, in this experiment).
This sensory evaluation was substantiated by a newly
developed chemical analysis using a rapid and automated solid phase
micro-extraction (SPME) stable isotope dilution gas chromatography/mass
spectrometry (GC-MS) method for quantifying 2-methoxy-3-isobutylpyrazine
(MP) concentrations in red wine. This assay showed that MP concentrations
in wines studied were significantly negatively-correlated with number
of buds per vine.5 As bud numbers increased, MP concentrations decreased
and, alternately, as bud numbers decreased, MP concentrations increased.
In addition, MP concentration determined by GC-MS was directly related
to sensory vegetal intensity ratings obtained by descriptive analysis.
The same study showed that crop thinning had a very
limited effect on wine quality, though it obviously impacted yield
significantly. These data are significant because they seem to challenge
the paradigm that as winegrape yield increases, resultant Cabernet
Sauvignon wine quality decreases. They further suggest that how
the crop yield is attained, rather than the exact yield itself,
is significant for resultant wine quality. This underscores the
importance of determining proper balance for a given set of growing
conditions as a function of the varietal, rather than having a pre-determined
crop level that ignores vine performance at the site and overall
Although the actual MP precursors have not been definitively
determined, some important features of their synthesis and degradation
are more clearly understood today. MP synthesis in grapes seems
to occur between fruit set and two to three weeks prior to veraison.2
This phenomenon seems to be related to grapevine water status before
veraison, as initial pyrazine concentrations were less in drier
years in France. At this pre-veraison stage, its concentration is
highest in the stems. It is also found in the berry pre-veraison,
but it is not known if it is actually synthesized in the berry or
if it is synthesized elsewhere and transported to the berry.
MP has also been identified in Cabernet Sauvignon
leaves, with basal leaves having the highest concentration, much
higher than in clusters. About three weeks before veraison, MP content
in berries decreases through to harvest.
Breakdown of malic acid and pyrazines seems to occur
simultaneously in winegrapes, irrespective of soil type, grape variety,
or vintage.7 This very strong correlation in their respective concentrations
may allow MP content to be considered a marker of grape immaturity.
MPs may serve as a biological deterrent to consumption of the fruit
before the seeds have attained sufficient ripeness to survive, and
may indeed be directly linked to ripening. However, other explanations
for the function of MPs have also been proposed.
Because the shape of the MP content-curve during ripening
does not change as a function of how the units are expressed, it
can be inferred that the decrease in MP content seen from about
three weeks before veraison through to harvest is independent of
the dilution that occurs during berry enlargement.1 Therefore, unlike
tartaric acid, MP is actually broken down as the berry swells and
Several studies have revealed a strong relationship
between exposure to light and decrease in MP content to explain
this decrease after mid-veraison. It is known that MP is broken
down in the laboratory by light.6 More work needs to be done in
the vineyard on intact berries and clusters to determine if light
is actually the chief agent of MP degradation. Regardless of phenological
stage, the pulp contains very little MP while stems and older leaves
contain a lot.2
At harvest in the cluster, MP is found mainly in the
stems or rachis. It has been shown that, under identical environmental
conditions, MP accumulates in leaves simultaneously with its degradation
in grapes during ripening. All data, so far, suggests that metabolism
of this pyrazine is completely different in grapes, stems, and leaves.1
Although actual quantification of MP content in grapes
is not practical for most wineries today, it may be possible to
track the curve of its degradation by monitoring the concurrent
decrease in malic acid during ripening.
Recent research from France has shown that there is
a significant correlation between degradation of MP content during
grape ripening and loss of malic acid during the same period.17
For the two years studied, Cabernet Sauvignon and Merlot grapes
grown under different environmental and cultural conditions, were
tested for both malic acid content and MP content from veraison
to harvest. These data show that the breakdown of malic acid and
MP occurred simultaneously, irrespective of soil type, grape variety,
or vintage. The determination coefficient (r2) between the breakdown
of these two compounds was 0.90 and 0.96 in 1996 and 0.91 and 0.99
in 1997. They are, therefore, highly correlated.
Knowing that many products that are synthesized in
the leaves are transported to the berries, in a preliminary study
(not yet published), researchers treated grape leaves with a solution
of a marked analogue of MP by putting the solution on leaf blades,
then analyzing its movement after several days.1 This was done between
the small pea stage and beginning of veraison.
Researchers observed a redistribution (movement) of
the marked compound to petioles and other growing parts of the vine,
including clusters. Amounts of MP confirmed the fact that, at this
stage of ripeness, metabolites are preferentially routed to clusters.
They found the marked MP in the rachis and berries, demonstrating
that this compound is transported by the phloem from leaves to berries.
Since it can be shown that MP is moved from leaves to the berries,
this fact underscores the importance of managing vegetative growth
to decrease the production of pyrazines. It is still unknown whether
the berry actually synthesizes pyrazines itself or whether they
are transported to the berry from the vines vegetative structures.
These results confirm the advantages of leaf removal
to decrease the perception of vegetal/green pepper characteristics.
Another aspect of this research measured the effects of various
canopy manipulation practices and their timing on the pyrazine content
of Cabernet Sauvignon and Merlot grapes. Summer thinning of laterals,
leaves, and clusters all had a direct effect on decreasing MP concentration
in these two varieties.1 The light-sensitive nature of MP may partially
explain this result.
Wines made from cluster-thinned or leaf-thinned groups
had higher alcohol, more color, and a lower MP content than those
of the control. But timing of the vineyard work was also shown to
be an important determinant of final MP content in the resultant
Leaf thinning between fruit set and before veraison
leads to grapes with a higher sugar content, smaller size, and less
green pepper character at harvest. It is interesting to note that
one year of the study in France, there was significant rainfall
during veraison that greatly slowed the usual MP breakdown that
occurs in this phase of berry ripening. This also resulted in an
unusual heterogeneity in berry ripeness. The vineyard had been leaf-thinned
after fruit set, prior to the rains, so MP synthesis had already
been inhibited somewhat.
Due to these lower levels of MP synthesis after leaf
thinning and decreased breakdown due to rainfall during veraison,
MP levels at harvest were close to zero. Comparing those results
to a more typical year with less dramatic rainfall, the typical
season showed much higher levels of MP at mid-veraison. These levels
held through to harvest. This highlights the assertion that it is
weather conditions just prior to and during veraison that most impact
the final MP concentration in grapes, rather than the period between
veraison and harvest.
Early lateral removal and leaf thinning also facilitate
later season vineyard work by eliminating some vegetation in the
fruiting zone. These practices improve vine balance, allow better
light penetration , lead to better aeration around clusters, and
remove vegetative structures requiring additional nutrients, which,
in turn, promotes better distribution of photosynthates for improved
Another aspect of the role of methoxypyrazines in wine production
concerns the relatively recent practice of extended hang time of
winegrapes to achieve more ripe flavors. There has been a great
deal of speculation about the reasons and ramifications of this
It is true that, from a winemaking perspective, the
fruit of some varietals in some regions simply does continue to
improve with extended ripening. In a study of Sauvignon Blanc in
New Zealand, fruit grown in a warmer region (2100 heat summation
units) had pyrazine levels below the threshold of detection while
fruit from a much cooler region (1430 heat summation units) had
a pyrazine level four-fold greater than the detection threshold
at harvest.5 This study is consistent with much earlier work stating
that the loss of such unripe characters was the basis
of recommendations of how varietals performed, or how they reached
enological ripeness, as a function of degree day summation units.
As mentioned earlier, pyrazine characters have been
shown to decrease with ripening, allowing a winemaker to time harvest
to the appropriate level of vegetal character for the wine style
Some wine writers and consumers do not appreciate
vegetal characters in wines, preferring instead wines of concentrated
fruitiness. This preference has resulted in a trend towards extended
ripening in an effort to further reduce vegetal character in winegrapes
and to enhance wine fruit aroma profile. The efficacy of a particular
grower or managers program of extended ripening is predicated
on the assumption that grapes are grown in an area that allows proper
Today, such errors in judgment are rare, yet there
is a belief that vegetal characters are persisting in the fruit
much longer than they should based on other indices of ripeness
or based on historical data obtained when most vines were planted
to AxR1 rootstock. Many new vineyard practices were introduced simultaneously
with extensive replanting of vines to rootstock other than AxR1.
Some of these practices resulted from extensive wine
quality evaluation, although not necessarily broadly applicable
beyond the region in which they were initially studied. But most
studies simply evaluated vine growth parameters without benefit
of accurate sensory evaluation of wines produced. Indeed, few were
evaluated in combination, and changes in ripening of the scion varieties
could simply reflect changes in vineyard practices.
However, it is also apparent that the definition of
enological ripeness has changed and planting recommendations based
on a historical target fruit composition may no longer be valid.
Should degree day summation recommendations be revisited
in terms of changes in consumer tastes, viticultural practices,
and definitions of enological ripeness?
Are some varietals simply not able to ripen to new
standards demanded in the marketplace in regions in which they have
been planted? Or are other forces in play? These are important questions
that need to be addressed.
Winemakers and grapegrowers know a lot more about methoxypyrazines
today than several years ago. A lot of that information can be put
to work almost immediately to help grow better grapes and make better
wine. The data clearly shows that MP content of a wine depends mainly
on the content in the grapes it was made from, and that it is only
very marginally affected by winemaking techniques. The one exception
appears to be thermovinification, or heating the must or wine to
approximately 60†C to 80†C for a short period, which has been shown
to greatly decrease MP and could be a solution in some cases.
Winegrowers need to focus on how to manipulate methoxypyrazines
to the preferred level in the vineyard. Vine balance is key to this.
Vines that are undercropped produce wines with high levels of green
pepper attributes, as do vines that are overly vigorous and cropped
Because the maximum MP content is achieved prior to
veraison, the period between set and veraison is when a grower can
employ practices that have the most impact on the vegetal concentration
of winegrapes There are significant data to show that early shoot
and leaf removal are beneficial in decreasing these vegetal characters
in grapes. Early season deficit irrigation, prior to veraison, can
also be used to decrease vegetal characters.
It appears that MP synthesis is related to vegetative
growth of grapevines, so any early intervention that decreases canopy
size/leaf area also decreases pyrazine content in berries. There
may be growing areas where some varietals cant achieve enological
ripeness, if the definition of ripeness includes absence
of vegetal characters. But in most areas, manipulation of pyrazines
in the vineyard is key to their expression in a resultant wine.
| 1. Roujou de Boubee, D., 2003 Amorim Academy Grand Prix.
2. Roujou de Boubee, D., A.M. Cumsille, M. Pons, D. Dubourdieu.
2002. Location of 2-methoxy-3-isobutylpyrazine in Cabernet
Sauvignon grape bunches and its extractability during vinification.
Amer. Jour. Enology & Viticulture. 53 (1): 15.
3. Chapman, D.M., M.A. Matthews, J.X. Guinard. 2004. Sensory
attributes of Cabernet Sauvignon wines made from vines with
different crop yields. Amer. Jour. Enology & Viticulture.
55 (4): 325334.
4. Chapman, D.M., J.H. Thorngate, M.A. Matthews, J.X. Guinard,
S.E. Ebeler. 2004. Yield effects on 2-methoxy-3-isobutylpyrazine
concentration in Cabernet Sauvignon using a solid phase microextraction
gas chromatography/mass spectrometry method. Jour. Agric.
Food Chem. 52 (17): 54315435.
5. Lacey, M.J., M.S. Allen, R.L.N. Harris, W.V. Brown.
1991 Methoxypyrazines in Sauvignon blanc grapes ands wine.
Amer. Jour. Enology & Viticulture. 42 (2): 103108.
6. Heymann, H., 1986. Studies of methoxypyrazines
and vegetative flavors of Cabernet Sauvignon wines. PhD
Thesis, University of California, Davis.
7. Roujou de Boubee, D., C. Van Leeuwen, D. Dubourdieu.
2000. Organoleptic impact of 2-methoxypyrazine on red
Bordeaux and Loire wines. Effect of environmental conditions
on concentrations in grapes during ripening. Jour. Agric.
Food Chem. 48 (10): 48304834.