Recent findings on rosé wine aromas. Part II: optimising winemaking

Winemaking
Recent findings on rosé wine
aromas. Part II: optimising
winemaking techniques
Marie-Laure Murat
Florent Dumeau
SARCO BP
40, 33015 Bordeaux cedex, France
Société d’Application de Recherche et de Conseil
oenologique, research subsidiary of Laffort oenologie.
[email protected]
SARCO BP
[email protected]
This article follows the publication of Recent findings on Rose
wine aromas. Part 1: identifying aromas studying the aromatic
potential of grapes and juice, first published in The Australian &
New Zealand Grapegrower & Winemaker Annual Technical Issue,
2005.
juice extraction methods, then study the factors that affect
alcoholic fermentation. We then describe the impact of yeast
strains and fermentation temperatures on rosé wine aromas.
Finally, we discuss the ageing methods that best preserve rosé
wine aromas.
Introduction
Juice extraction methods
The must of many “non-muscat” grape varieties, such as
Merlot, Cabernet Sauvignon, Cabernet Franc, Syrah, Grenache,
etc. is almost odourless, whether the grapes are pressed
immediately, given pre-fermentation skin contact, or the juice is
bled off from a vat used to make red wine. Their aromas develop
during alcoholic fermentation. Our previous article (Part I)
described how three of the main compounds involved in the fruity
aromas of these wines were identified. 3-mercaptohexan-1-ol
(3MH) and its acetate (3MHA) contribute to grapefruit, passion
fruit, and boxwood nuances. Phenyl-ethyl acetate (PEA) is
responsible for floral nuances. Finally, the role of isoamyl acetate
(IA) in the amyl character of wines has been known for some
time24.
The research reported in our previous article demonstrated the
existence of the cysteinylated precursor of 3MH (P-3MH) in must
used to make rosé wines, as well as the correlation between the P3MH content of the must and the quantity of 3MH released into
the wine (Part I). Temperature and pre-fermentation skin-contact
time also have an impact on the P-3MH concentration of the must.
Extensive studies using Sauvignon Blanc have shown that the
conversion of P-3MH into aroma is due to the metabolism of
winemaking yeasts during alcoholic fermentation25, 26, 27 and is
variable depending on the strain17. A3MH is also formed by yeast
via acetylation of 3MH25. The impact of the yeast strain on the
concentrations of fermentation esters (eg. PEA and IA) in wine
has been known for many years3, 22, 24, 29.
Thus, the intensity and character of rosé wine aromas is
very closely linked to the juice extraction and alcoholic
fermentation methods used. In this article, we present various
The methods used to obtain juice for making rosé wines varies
considerably from one winemaking area to another.
This is the most common method of making rosé wines. It
consists of pressing red grapes immediately after picking, as is the
case in making dry white wines. The pressing cycle must be
adapted with long, slow pressing and a minimum of mechanical
handling and crushing (number of times the press cake is broken
up) to reduce the release of tannins. Direct pressing usually
produces very pale-coloured juice. When grapes are hot (above
25°C) on arrival in the winery run the must through a heatexchanger after crushing, or keep hand-picked grapes in a
refrigerated room for half a day to reduce the temperature to around
15°C. If this is not possible, it is essential to press the grapes
immediately. The juice must be cooled after pressing so that the
must can settle.
Presses equipped with closed cages, cooling systems, or even
nitrogen blankets may be used for pre-fermentation skin contact,
to enhance the extraction of anthocyanins and aroma precursors.
Contact time may be reduced by adding specific enzymes to the
grapes. These preparations accelerate anthocyanin extraction and
increase the yield of free-run juice by over 10%, compared with
an untreated control15 in a press operated at reasonable pressures
(see Figure 1). As is the case in white winemaking, the greatest
attention must be paid to selecting the juice. When the grapes are
simply left in the press for a few hours, up to 50-70% of the total
volume of free-run juice may be obtained by gravity. Low
pressures, 0.2-0.8bar, are sufficient to obtain the required amount
The Australian & New Zealand Grapegrower & Winemaker 49
▲
August 2005
Direct pressing
winemaking
Little is known about the impact of grape ripeness in different
grape varieties on the organoleptic characteristics of wines made
by bleeding the vats. For example, when some grape varieties are
picked too early, the wine made from bled-off juice has
unflattering, herbaceous nuances. Paradoxically, in other cases,
juice from incompletely ripe grapes is the most aromatic.
Irrespective of the method used to obtain the juice, the greatest
care must be taken to protect the must from oxidation (see Part I).
Fermentability of musts intended for rosé wines
Fig. 1. Impact of adding enzymes addition on the free-run juice yield during
pressing (grape variety: Cinsault, 2003, skin contact: 20 min in the press at
18°C). Source: Centre de Recherche et d’expérimentation sur le vin Rosé (rosé
wine research center)15.
of juice. The juice from the last pressing cycles (p>0.8 bar) may
be kept apart and fermented separately.
Pre-fermentation skin contact
When the grapes are healthy, pre-fermentation skin
contact is quite common in making both dry white and rosé
wines. This may be done directly in the press, as described
above, or in a maceration tank, after destemming and
crushing. Some vats are equipped with a membrane,
drains, and cooling systems for optimum juice extraction.
Between 500 and 600 litres of juice may be obtained per
metric ton of grapes, depending on the variety. This juice
is generally clearer (<300NTU) as it is filtered naturally
by the pomace. When the pomace has been pressed, a
variable amount of juice from the first pressings may be added to
the free-run juice, depending on the overall yield and quality
objectives.
Frequent pumping over under a nitrogen blanket is sometimes
used to accelerate extraction of the colouring matter. Specific
enzymes also improve juice yields and extract colour more
rapidly. The rate of colour extraction is the main factor in
determining contact time.
During this phase, P-3MH is extracted at the same time as
anthocyanins. However, the vats with the deepest colour are not
necessarily the most aromatic (see Part I).
Bleeding
In Bordeaux and other areas with a strong tradition of red
winemaking, rosé wines are traditionally made by bleeding juice
off from vats. The winemaker bleeds 10-30% of the free-run juice
off 12-24 hours after the must has been put into the vat. Juice bled
off from several vats is often gathered in a single vat and
maintained at low temperature until the vat is full. This technique
also increases the pomace/juice ratio in the vats used to make red
wine.
50 The Australian & New Zealand Grapegrower & Winemaker
Winemakers generally agree that rosé must is often difficult to
ferment, independently of temperature, turbidity, or sugar content.
Stuck fermentation is a frequent problem and the precise reasons
for it have not yet been determined.
Among the factors that affect alcoholic fermentation kinetics,
the role of available nitrogen used by yeast (ammonium cations +
amino acids) has been clearly demonstrated1,4,19. The available
nitrogen content of ripe grapes varies from one grape variety to
another, as well as according to the vine’s nitrogen and water
supply, the vintage, etc.12,30. The threshold for available nitrogen
deficiency in must has been assessed at approximately 140mg/L2.
The available nitrogen content must be systematically adjusted
when it is below 160mgN/L. For reference, 20g/hL ammonium
salts represent 42mg/L available nitrogen. We determined the
available nitrogen content (Sörensen method or formol titration)
of a large number of must samples from Bordeaux vineyards used
to make rosé wines. As shown in Table 1, 68% of the rosé musts
analysed had a nitrogen deficiency.
Table 1. Available nitrogen content (mgN/L) of 150 musts used to make rosé
wines (Bordeaux, 1998, 1999, 2001-2004 vintages).
Minimum value
Maximum
value
Mean
Standard
deviation
% nitrogen
deficient must
(N < 140 mg/L)
40
296
125
50
68
The role of lipids (unsaturated and saturated long-chain fatty
acids) in the very structure of membrane phospholipids was also
clearly demonstrated. Associated with sterols, another survival
factor, their special conformation regulates membrane fluidity, an
essential parameter in sugar and amino acid transport protein
function and, therefore, their assimilation by yeast. Yeast is only
capable of synthesising these membrane compounds in the
presence of molecular oxygen. It is, therefore, vital to aerate the
must during the yeast multiplication phase, or during the first
third of fermentation, before depletion of the first 50 grams of
reducing sugars. Unless oxygen is added, the lipid content of the
must is determined by the level of clarification (turbidity)18. The
majority of lipid compounds come from the lees. The turbidity
August 2005
winemaking
levels of rosé musts before fermentation have a major impact on
their fermentability. Excessively clarified must has a very low lipid
content and the end of fermentation is often very sluggish. It has
recently been demonstrated that nitrogen-deficient musts may also
have a lipid deficiency. In cases of nitrogen deficiency, excessive
clarification, or anaerobic fermentation, it is, thus, essential to
adjust the lipid content to ensure that the must ferments properly.
Adding gross lees from juice with sufficient nitrogen (see Table 2)
or using yeast activators with a high survival factor content (eg.
Dynastart®) provide effective solutions to this problem (see
Figure 2).
Table 2. Improving the fermentation of a rose must with an available nitrogen
deficiency by adding solids from a nitrogen-rich must (Cabernet Sauvignon,
Bordeaux, 1999).
Composition
1
Length of alcoholic Reducing sugars
fermentation
(g/L)
Volatile acidity
(g/L acetic acid)
DS wo N
Stop
105
0.22
2
Stop
5.4
0.06
3
14 days
1.3
0.08
DS + N
RS + N
1: Turbidity adjusted with gross lees from nitrogen-deficient juice, without adding
ammonia nitrogen.
2: Turbidity adjusted with gross lees from nitrogen-deficient juice, adding ammonia
nitrogen to obtain 200 mg/L available nitrogen.
3: Turbidity adjusted with gross lees issues from nitrogen-rich juice, adding
ammonia nitrogen to obtain 200 mg/L available nitrogen.
Impact of yeast strain on rosé wine aromas
experimental strains that are not yet commercially available,
resulting from interspecific cross (S. cerevisiae/S. bayanus var.
uvarum), strain A8,13,16 or intraspecific cross (S. cerevisiae x S.
cerevisiae), strains X5 and B (A and B are experimental strains,
not yet commercialised). At mid-fermentation, the analysis of the
karyotypes of the total biomass for each yeast tested confirmed
the implantation of all strains in the two experiments (data not
shown). In those experiments, we used at mid fermentation for
each tested yeast, the analysis of the karyotypes of the total
biomass to confirm the implantation of strains.
The results in Table 3 clearly demonstrate the role of the yeast
strain in releasing 3MH and producing APE: strain A generated
the largest quantity of both compounds. The impact of the yeast
strain on A3MH formation was not statistically significant.
However, strains A, VL3c, and VL1 apparently formed more of
this compound than strain 522d. These results indicate that the
Made for each other.
At T.W. Boswell, bringing out
the true character of wines is a
passion we share with you. Our
Médoc, Côte d'Or and Côtes
du Rhône barrels are carefully
handcrafted of the finest oak
and then toasted to enhance
and complement particular
grape varieties. This attention
to detail produces harmonies
in texture, balance and
flavor, while imparting an
Old-World complexity to
your finest wines.
Great wine’s silent partner
www.twboswell.com
August 2005
117 Murray Street
Shop 3
Tanunda, SA 5352
(08) 8353 8177
The Australian & New Zealand Grapegrower & Winemaker 51
▲
As described in Part I, the yeast strain has a major impact on
the fruity aromas of rosé wines. Since 1998, we have tested a
number of yeast strains in the Saccharomyces cerevisiae species,
commonly used to make white and rosé wines as well as some
Fig. 2. Effect of adding a yeast activator (Dynastart) when rehydrating the
yeasts – Temperature :16°C – Turbidity: 100NTU – Nitrogen: 130mg/L (adjusted
to 180mg/L)
Table 3. Impact of the yeast strain on the 3MH, A3MH and APE content of rose
wines (Bordeaux, 1998, mean of 5 trials).
1
Compounds
Strain A
VL3c
VL1
522d
3MH (ng/L)
(grapefruit passion fruit)
848 a
603 ab
583 ab
177 b
3MHA (ng/L)
(passion fruit boxwood)
7.6
8.5
9,2
1.6
PEA (mg/L)
(floral)
1.93 a
0.6 b
0.62 b
0.46 b
2
1
A = interspecific hybrid Saccharomyces cerevisiae / Saccharomyces bayanus var.
uvarum.
2
: Values followed by the letters a,b are statistically different (Two-factor variance
analysis without repetition, p < 0.01).
choice of a suitable yeast strain is a significant factor in making
fruity rosé wine.
Breeding is a selection method used to cross yeast strains
chosen for their winemaking qualities: capacity to ferment under
difficult conditions, capacity to release the aromas of certain
grape varieties, etc. This technique represents a major
technological progress in yeast selection. As an example, we show
the results obtained with two strains produced by an intraspecific
cross (see Table 4). Strain X5 showed a remarkable aptitude to
release 3MH from its cysteinylated precursor and produce its
acetate (3MHA). It also generated very small quantities of
fermentation esters. On the contrary, strain B released less of the
volatile thiols and produced more fermentation esters, particularly
isoamyl acetate. It is, therefore, possible to adapt the choice of
yeast strain to suit the desired aromatic profile.
Table 4. Impact of the yeast strain on the aromatic profile of rose wines (Merlot,
Bordeaux, 2004).
Compounds
Strain X 5
Strain B
3MH (ng/L)
(grapefruit - passion fruit)
988
598
3MHA (ng/L)
(passion fruit - boxwood)
185
78
PEA (mg/L) (floral)
0.75
0.91
IA (mg/L) (banana)
1.3
6
Impact of fermentation temperature on rosé wine aromas
Fermentations at low temperature enhanced the wine content
of some volatile compounds produced by the yeast during the
alcoholic fermentation (esters, acetates, medium-chain fatty
acids)5,6,7,9,23,28. More recently, it has been demonstrated that a high
fermentation temperature (18-20°C) is preferable to the revelation
of volatile thiols from their odourless precursors in Sauvignon
Blanc wines14. As some of those compounds are present in rosé
wines, we studied the impact of fermentation temperature on the
release of 3MH, 3MHA. Our research considered two
fermentation temperatures: 13º and 20°C, and four yeast strains.
As shown in Figures 3a and 3b, a higher fermentation temperature
produced wines with significantly higher concentrations of 3MH
and 3MHA, as in the case of Sauvignon Blanc14. It is, therefore,
possible to adapt fermentation temperature depending on the
aromatic profile desired, to enhance varietal or fermentation
aromas.
Impact of ageing on fine lees on the volatile thiol content of
rosé wines.
52 The Australian & New Zealand Grapegrower & Winemaker
August 2005
▲
Once we had determined the positive impact of certain factors
during alcoholic fermentation on the quality of rosé wines, we
TM
Nothing
to declare
Unlike some of it’s competitors,
GrapEXTM natural tannins are wine
grape derived and do not need
to be declared on the wine label.
So when winemakers use GrapEXTM
they can do so with confidence.
Ask about all the benefits of
GrapEXTM today.
TARAC TECHNOLOGIES
Australia
Samuel Road [PO Box 78] Nuriootpa
South Australia 5355
T. + 61 8 8562 1522 F. +61 8 8562 2031
Freecall 1800 827 228 www.tarac.com.au
New Zealand Distributor
Kauri New Zealand Ltd
T. + 64 4 476 0105 F. + 64 4 476 0161
GrapEX is a registered Trade Mark of Tarac Technologies.
winemaking
Fig. 3a. Impact of fermentation temperature on the 3MH content of rosé wines
(mean obtained using four yeast strains, Bordeaux, 2003).
Fig. 3b. Impact of fermentation temperature on the 3MHA content of rosé wines
(mean obtained using four yeast strains, Bordeaux, 2003).
Table 5. Impact of ageing on lees on the volatile thiol content of rose wines
(Merlot, Bordeaux, 2004).
It is thus probable that this compound contributes to preserving
rosé wine aromas as well.
Compounds
Control
Wine + lees
3MH (ng/L)
(grapefruit - passion fruit)
250
490
3MHA (ng/L)
(passion fruit - boxwood)
10
25
studied the impact of ageing on fine lees on their volatile thiol
content. We measured the 3MH and 3MHA content of several rosé
wines after two months’ ageing with or without lees. The wines
always had a higher 3MH and A3MH content when they were
aged on the lees (see Table 5), which are known to have reducing
properties8,10,20,21. This effect had already been demonstrated for
dry white wines11. Recent research by Lavigne and Dubourdieu
showed that ageing wines on the lees minimised the decrease in
glutathion and varietal thiol content of Sauvignon Blanc wines11.
Conclusion
Recent enthusiasm for rosé wines in international markets
proves that it is by no means a secondary product. Winemakers
should, therefore, make every effort to ensure that their rosé wines
suit consumer tastes.
The research presented in both Part I and Part II highlight the
need to follow certain rules in making this type of wine. Indeed,
rosé winemaking requires certain specific techniques. More
detailed knowledge of the key compounds in the fruity aroma of
rosé wines and the quantities present have made it possible to
optimise winemaking methods. We are continuing our research,
particularly focusing on yeast strain selection and optimising the
use of enzyme preparations, and our findings will be published at
a later date.
References
1 Agenbach W. A. I. (1977) A study of must nitrogen content in relation to
incomplete fermentations, yeast production and fermentation activity. In
processing of South African Society for Enology and Viticulture. Cape Town,
South Africa, Stellenbosh.
2 Bely M. (1990) Détection automatique et correction des carences en azote
assimilable des fermentations alcooliques en conditions oenologiques : étude
cinétique et approche physiologique. Thèse de biochimie, biologie cellulaire et
moléculaire. Université de Montpellier.
3 Bertrand A., Marly-Brugerolle C. et Sarre C. (1978) Influence du débourbage
des moûts et du sulfitage sur les teneurs en substances volatiles des vins et
eaux-de-vie. Conn. Vigne Vin. 12 :35-48.
4 Bezenger M. C. et Navarro J.M. (1987) Influence de l'azote sur la fermentation
alcoolique en milieu modèle simulant les conditions de l'oenologie. Sci. Aliments.
7:41-60.
5 Bisson J., Daulny B. et Bertrand A. (1980) Influence de la température de
fermentation sur la composition d’un vin blanc sec. Conn. Vigne Vin, 14. N°3
:195-202.
6 Cotrell, T.H.E. MC Lellan, M. R. (1986) The effect of fermentation temperature
on chemical and sensory characteristics of wines from seven white grape
cultivars grown in New York State. Am. J. Enol. Vitic. 37,3:190-194.
7
Daudt C.E. and Ough C.S. (1973) Variations in some volatile
acetate esters formed during grape juice fermentation, effects of fermentation
temperature, SO2, yeast strain, and grape variety. Am. J. Enol. Vitic. 24,3:130195.
8 Dubourdieu D., Tominaga T., Masneuf I., Peyrot des Gachons C., Murat ML.
(2000) The role of yeast in grape flavor development during fermentation : the
example of Sauvignon blanc. Proceedings of the ASEV 50th Anniversary
Meeting, Seattle, Washington, Am J. Enol. Vitic. 51,5:196-203.
9 Killiam, E. and Ough, C.S. (1979) Fermentation esters- Formation and retention
as affected by fermentation temperature. Am. J. Enol. Vitic. 30,4:301-305.
10 Lavigne-Cruege V. (1996) Recherches sur les composés soufrés volatils
formés par la levure au cours de la vinification et de l'élevage des vins blanc
secs. Thèse de Doctorat ès Science. Université Victor Ségalen Bordeaux II.
54 The Australian & New Zealand Grapegrower & Winemaker
August 2005
winemaking
11 Lavigne-Cruège V, Pons A., Choné X. et Dubourdieu D. (2003) Rôle du
glutathion sur l’évolution aromatique des vins blancs secs. In : ?nologie VII
Symposium International d’oenologie. Editions Tec & Doc. 2003 :385-388.
21 Salmon J.M. (2004) Réactivité entre la levure de vinification et l'oxygène :
conséquences pratiques en oenologie. Premières rencontres Internationales du
Rosé. 7-8.
12 Masneuf I. et Dubourdieu D. (1999) L'azote assimilable: intérêt de son dosage
par formoltitration ; Etude de quelques paramètres à l'origine des variations de
sa teneur dans les moûts. Rev. des Oenologues. 93 :31-32.
22 Soufleros E. et Bertrand A. (1979) Rôle de la “souche de levure” dans la
production de substances volatiles au cours de la fermentation du jus de raisin.
Conn. Vigne Vin. 13:181-198.
13 Masneuf I., Murat M.L, Naumov G.I., Tominaga T. and Dubourdieu D. (2002)
Hybrids Saccharomyces cerevisiae x S. bayanus var uvarum having a high
liberating ability of some sulfur varietal aromas of Vitis vinifera Sauvignon Blanc
wines. J. Int. Sci. Vigne Vin. 36,4, 205-212.
23 Soufleros E. et Bertrand A. (1980) Incidence de l’action conjuguée de la
14 Masneuf I., Mansour C., Murat M.L, Tominaga T. and Dubourdieu
D. (2005) Influence of fermentation temperatue in volatile thiols
concentrations in Sauvignon Blanc wines. Submitted to publication in Food
Microbiol.
température de fermentation et de l’acidité du milieu sur les teneurs en
substances volatiles formées par les levures. Conn. Vigne Vin. 14,2:97-109.
24 Suomalainen H. (1971) Yeasts and its effect on the flavour of alcoholic
beverages. J. Inst. Brew. 77:164-177.
25 Tominaga T., Masneuf I., Dubourdieu D. (1995) Mise en évidence d’un S-
conjugué de la cystéine, précurseur d'arôme du Sauvignon. J. Int. Sci. Vigne Vin.
29,4:227-232.
15 Masson G. Cayla L., Guérin L. (2003) Compte rendu technique : Incidence de
l’utilisation de préparations enzymatiques pour la vinification en rosé.
16 Murat M.L. (2001) Recherches sur la vinification des vins rosés et clairets de
26 Tominaga T. (1998) Recherches sur l’arôme variétal des vins de Vitis vinifera L.
cv. Sauvignon blanc et sa genèse à partir de précurseurs inodores du raisin.
Thèse Doctorat ès Sciences. Université Victor Ségalen Bordeaux II.
Bordeaux. Diplôme d’Etudes et de Recherches de l’Université de Bordeaux II.
17
Murat M.L., Masneuf I., Darriet P., Lavigne V., Tominaga T.,
Dubourdieu D. (2001a) Effect of the Saccharomyces cerevisiae yeast strains on
the liberation of volatile thiols in Sauvignon blanc wine. Am. J. Enol. Vitic.
52,2:136-139.
18 Ollivier, C., Stonestreet, T., Larue, F., and Dubourdieu, D. (1987) Incidence de
la composition colloidale des moûts blancs sur leur fermentescibilité. Conn.
Vigne vin. 21:59-70.
27 Tominaga T., Peyrot des Gachons C., Dubourdieu D. (1998b) A new-type of
flavors precursors in Vitis Vinifera L. Cv. Sauvignon S-cystéine conjugates. J.
Agric. Food Chem. 46:5215-5219.
28 Torija, M. J., Rozès, N, G., Poblet, M., Guillamon, J. M., Mas, A. (2003) Effects
of fermentation temperature on the strain population of Saccharomyces. Int. J.
Food Microbiol. 80:47-53.
29 Usseglio Tomasset L. (1967) L’alcol b-feniletilico nei vini. Riv. Viticol. Enol.
20:10-35.
19 Salmon J. M. (1989) Effect of sugar transport inactivation in Saccharomyces
cerevisiae on the appearance of sluggish and stuck enological fermentations.
Applied And. Env. Microbiol. 55,4:953-958.
20 Salmon J.M., Mauzauric J.P., Fornairon C., Moutounet M. (2000) Les lies de
levures et la consommation d'oxygène. V colloque des Sci. Tech. Tonnellerie,
Vigne Vin publication internationale. 39-43.
August 2005
30 Van Leeuwen C., Choné X., Chery P., Molot C. et Soyer J-P. (1999) Etude de
quatre sols viticoles A.O.C. Saint-Julien (Haut-Médoc, Bordeaux) ; incidences du
régime hydrique et de l'alimentation en azote de la vigne sur la maturation du
raisin et la qualité du vin (Vitis vinifera var. Cabernet Sauvignon, 1997). In
Oenologie 99, Coordinateur A. Lonvaud-Funel, Ed. Lavoisier Tec & Doc, Paris.
■
92-97.
The Australian & New Zealand Grapegrower & Winemaker 55