Analytica Chimica Acta 660 (2010) 183–189

Contents lists available at ScienceDirect

Analytica Chimica Acta

journal homepage: www.elsevier.com/locate/aca

Mechanical properties, phenolic composition and extractability indices of

Barbera grapes of different soluble solids contents from several growing areas
Fabrizio Torchio, Enzo Cagnasso, Vincenzo Gerbi, Luca Rolle ∗
DIVAPRA, Settore Tecnologie Alimentari, Università degli Studi di Torino, Via L. da Vinci 44, 10095 Grugliasco, Torino, Italy

a r t i c l e i n f o a b s t r a c t

Article history: Phenolic compounds, extractable from grape skins and seeds, have a notable influence on the quality of
Received 24 July 2009 red wines. Many studies have clearly demonstrated the relationship between the phenolic composition
Received in revised form 9 October 2009 of the grape at harvest time and its influence on the phenolic composition of the red wine produced.
Accepted 10 October 2009
In many previous works the evolution of phenolic composition and relative extractability was normally
Available online 17 October 2009
studied on grapes sampled at different times during ripening, but at the same date the physiological
characteristics of grape berries in a vineyard are often very heterogeneous. Therefore, the main goal
Keywords:
of the study is to investigate the differences among mechanical properties, phenolic composition and
Skin hardness
Cellular maturity index
relative extractability of Vitis vinifera L. cv Barbera grape berries, harvested at the same date from several
Texture analysis vineyards, and calibrated according to their density at three levels of soluble solids (A = 235 ± 8, B = 252 ± 8
Anthocyanin and C = 269 ± 8 g L−1 sugar) with the aim of studying the influence of ripeness stages and growing locations
Proanthocyanidin on these parameters.
Results on mechanical properties showed that the thickness of the berry skin (Spsk ) was the parameter
most affected by the different level of sugars in the pulp, while different skin hardnesses, evaluated by
the break skin force (Fsk ), were related to the cultivation sites. The latter were also observed to influence
the mechanical characteristics of seeds.
Generally, the anthocyanin content increased with the level of soluble solids, while the increase in the
tannin content of the berry skin and seeds was less marked. However, significant changes in flavanols
reactive to vanillin in the seeds were found.
The cellular maturity index (EA%) was little influenced by the soluble solids content of grapes.
© 2009 Elsevier B.V. All rights reserved.

1. Introduction ripeness and harvesting date [6,7]. Skin proanthocyanidins, instead,

The elaboration of high quality red wines requires the
assessment of grape phenolic maturity by the determination of con-
centration of phenolic compounds and their extractability during
the winemaking process. Knowledge of these parameters allows
for the management of red wine maceration and prediction of the
color and phenol potential of wines. Many studies have demon-
strated the relationship between grape phenolic composition and
its influence on the phenolic composition of red wine [1,2].
Biosynthesis and concentration of phenols in red grapes
depends on the cultivar, management vineyard practices, climatic
conditions, soil features and crop load [3]. Quantitative and qual-
itative modifications of tannins and anthocyanins were observed
during the ripening of the grapes [4,5]. Accumulation of antho-
cyanins in skins starts at veraison and it is at its maximum around
harvest, but the final total content and profile depend on fruit
∗ Corresponding author. Tel.: +39 011 6708558; fax: +39 011 6708549.
E-mail address: luca.rolle@unito.it (L. Rolle).
are mainly accumulated before veraison [8], but the literature
reports contrasting results on whether their amount and degree
of polymerization (mDP) increase or decrease during maturation
[9–12]. The highest flavanol concentrations in the seed are at verai-
son, after which they decline slowly until near maturity, at which
time they remain relatively constant [13,14]. The decrease is more
rapid for flavan-3-ol monomers than for oligomers, so the average
degree of polymerization increases with maturity [15,16].
The extractability of phenol compounds from the skins,
increases throughout grape ripeness as a consequence of degra-
dation of the cellular wall by pectolytic enzymes [17]. Differences
in polysaccharides based on galactose and arabinose, together with
the cellulose content and degree of methylation of the pectins could
be responsible for the different extractability [18]. In the seeds,
because of the histological and histochemical modifications that
occur during fruit development, the capacity for phenol release
also changes [13,19,20]. Many studies have been conducted aim-
ing at defining the index which best reflects the ease with which
the phenols are released. The cellular maturity index (EA%) is cur-
rently one of the most used indices to assess the extractability

0003-2670/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.aca.2009.10.017
184 F. Torchio et al. / Analytica Chimica Acta 660 (2010) 183–189
Table 1
The operative conditions applied for execution of texture analysis tests on grapes and mechanical parameters measured.
Test Probe–platform
Berry skin hardness SMS P/2N needle; HDP/90 platform
Berry skin thickness SMS P/2 Ø 2 mm; HDP/90 platform
Seed hardness SMS P/35 Ø 35 mm; HDP/90 platform
of anthocyanins [1,17,18,21–23]. Regarding the physical proper-
ties of skin, the break skin force has proved to be the mechanical
parameter best able to estimate the extractability of anthocyanins
with adequate reliability, at least in Brachetto and Nebbiolo grapes
[24,25]. Furthermore, a relationship between skin hardness and
thickness with EA% index was observed in Galician grapes [21],
focusing the importance of the study of mechanical parameters as
phenols extractability markers. Respect to Glories’ protocol [26,27],
the mechanical methods are rapid and inexpensive, and show a
clear potential as routine monitoring tool of the grapes quality [28].
Furthermore, because of high correlation of mechanical parameters
with sensory descriptors, these instrumental indices can be favor-
ably used in match with the judgments of sensory analysis panel
for the assessment of winegrapes ripeness [28,29].
Much of this knowledge about the evolution of mechanical char-
acteristics, phenolic composition and relative extractability was
acquired on grapes sampled at different times during ripening, but
at any given date the physiological characteristics of grape berries
in a vineyard are very heterogeneous [30]. According to Fournand
et al. [31], homogeneous samples of berries at different advanced
physiological stages (berries with the same density), show changes
in phenol extractability, proanthocyanidin content and antho-
cyanin composition, content and profile, with the increase of sugar
levels in the pulp. However, an influence of the succession of har-
vest date was also observed [31].
Therefore, the object of this work is to investigate the differences
of mechanical properties, phenolic composition and extractability
indices of grape berries of cv Barbera, harvested on the same date
from several vineyards and calibrated, according to their density, at
three levels of soluble solids contents, in order to determine how
these parameters are affected by the ripeness stage and growing
location. Working with homogeneous grape samples, the evolution
of chemical–physical characteristics of the grapes at different sugar
contents [27] can be better understood, permitting to separate the
influence on these parameters of the several stages of ripeness from
that of cultivation sites.
2. Materials and methods
2.1. Grapes and sampling
In 2008, Vitis vinifera L. cv Barbera grape samples, were har-
vested at the same date (6 October) from four vineyards when the
technological maturity was optimal for the production of Barbera
wines. The vineyards are located in the Alessandria (I), Asti (II) and
Cuneo (III, IV) provinces of Piedmont (North West Italy). The study
was conducted in commercial vineyards employing the same clone
(AT 84) and rootstock (S.O.4). The vines have the same age (14–16
years) and are planted at 2.5 m × 1 m. The vines were Guyot pruned
and shoots were vertically trained. The soil and climatic data were
reported in zoning study of Barbera productions areas [32,33].
3000 berries for each vineyard, were randomly picked with
attached pedicels (three berries for cluster). The berries were cali-
Test speed Force Mechanical properties
−1
1 mm s Compression, 3 mm Fsk = berry skin break force (N);
Wsk = berry skin break energy (mJ);
Esk = skin Young’s modulus (N mm−1 )
0.2 mm s−1 – Spsk = berry skin thickness (␮m)
1 mm s−1 Compression, 50% deformation Fs = seed break force (N); Ws = seed
break energy (mJ); Es = seed Young’s
modulus (N mm−1 ); DIs = Seed
Deformation Index (%) (distance of
seed break point/seed height × 100)
brated according to density (i.e., total soluble solids). Density was
estimated by flotation of berries in different salt solutions (from
130 to 190 g L−1 NaCl) so that the difference in total soluble solids
of two consecutive batches of berries was ∼17 g L−1 sugar (i.e., 1%,
v/v in potential alcohol)[31]. Three ripening stages were studied:
A (235 ± 8 g L−1 sugar), B (252 ± 8 g L−1 sugar) and C (269 ± 8 g L−1
sugar). Two representative sub-samples of 30 sorted berries were
used for the determination of the mechanical proprieties. Three
sub-samples of 10 sorted berries were used for the determination
of skins and seeds phenolic composition. Other three sub-samples
of 100 berries were used for the phenol extractability indices.
2.2. Mechanical properties
For Texture Analysis tests, a Universal Testing Machine (UTM)
TAxT2i Texture Analyzer (Stable Micro Systems—SMS, Surrey, UK)
equipped with a HDP/90 platform and a 25 kg load cell was used.
All the acquisitions were made at 400 Hz; data were evaluated
using the Texture Expert Exceed software package (version 2.54
in Windows 2000). The operating conditions applied, probe and
mechanical parameters measured, are reported in Table 1. Skin
hardness was evaluated using the puncture test [34]. Thirty berries
were placed on the metal plate of the UTM with the pedicel in a
horizontal plane in order to be consistently punctured in the lat-
eral face. For the measurement of berry skin thickness, a piece of
skin (ca. 0.25 cm2 ) was removed from the lateral side of all the other
30 berries of each sample with a razor blade. Care was taken, when
removing the pulp from the skin and when positioning the skin
sample on the UTM platform, to prevent folds in the skin. After
calibrating the probe position, skin thickness was calculated as
the distance between the point corresponding to the probe con-
tact with the berry skin (trigger) and the platform base during a
compression test [35]. It was convenient to insert an instrumental
trigger threshold equal to 0.05 N to enable the plane surface of the
probe to adhere completely to the skin sample before the acqui-
sition started, thus reducing or eliminating the ‘tail’ effect [35].
For assessing seed hardness a compression test was used [35]. For
each of the 30 analyzed berries, the same used for skin thickness,
one seed was carefully removed from the pulp and cleaned with
adsorbent paper before analysis [35].
2.3. Phenolic composition of grapes
2.3.1. Extraction
Three replicates of 10 berries for each soluble solids class and
growing area were weighed before phenolic extraction [36]. The
berry skins, removed manually from the pulp and dried with paper,
were quickly immersed in 25 mL of a buffer solution containing
12% (v/v) ethanol, 600 mg L−1 sodium metabisulfite, 50 mg L−1
NaN3 , 5 g L−1 tartaric acid and titrated to pH 3.20 by the addition
of NaOH 1 M [37]. After homogenization with an Ultraturrax T25
(IKA Labortechnik, Staufen, Germany), the extract was centrifuged
for 10 min at 3000 × g at 20 ◦ C. The supernatant was then used
 F. Torchio et al. / Analytica Chimica Acta 660 (2010) 183–189 185

Table 2
Berry skin mechanical properties of Barbera grapes: Fsk = berry skin break force, Wsk = berry skin break energy, Esk = skin Young’s modulus, Spsk = berry skin thickness.

Areas of production Soluble solid class Fsk (N) Wsk (mJ) Esk (N mm−1 ) Spsk (␮m)

A 0.710 ± 0.082 a,␣

0.507 ± 0.178 a,␣
0.492 ± 0.109 a,␤
163 ± 26a,␣
I B 0.687 ± 0.075a,␣ 0.461 ± 0.133a,␣ 0.501 ± 0.104a,␤ 171 ± 30ab,␣␤
C 0.696 ± 0.089a,␣ 0.477 ± 0.132a,␣ 0.489 ± 0.101a,␤ 184 ± 24b,␣␤

A 0.774 ± 0.108a,␣ 0.594 ± 0.146a,␣ 0.474 ± 0.072a,␣␤ 165 ± 28a,␣

II B 0.769 ± 0.097a,␤ 0.562 ± 0.163a,␣ 0.509 ± 0.094ab,␤ 170 ± 32a,␣
C 0.781 ± 0.122a,␤ 0.522 ± 0.158a,␣ 0.565 ± 0.105b,␥ 180 ± 32a,␣␤

A 0.739 ± 0.083a,␣ 0.509 ± 0.130a,␣ 0.520 ± 0.110a,␤ 160 ± 41a,␣

III B 0.727 ± 0.123a,␣␤ 0.512 ± 0.163a,␣ 0.498 ± 0.091a,␤ 167 ± 37a,␣
C 0.741 ± 0.102a,␣␤ 0.527 ± 0.195a,␣ 0.520 ± 0.095a,␤␥ 169 ± 22a,␣

A 0.915 ± 0.119a,␤ 0.918 ± 0.185a,␤ 0.412 ± 0.093a,␣ 177 ± 30a,␣

IV B 0.913 ± 0.114a,␥ 0.957 ± 0.251a,␤ 0.401 ± 0.105a,␣ 193 ± 30ab,␤
C 0.937 ± 0.114a,␥ 1.002 ± 0.232a,␤ 0.393 ± 0.100a,␣ 200 ± 35b,␤

Average value ± standard deviation (n = 30). Different Latin letters indicate significant difference (p < 0.05) among the three level of sugar content in the same area of production.
Different Greek letters indicate significant difference (p < 0.05) among the grape of the different four areas of production at the same levels of sugar content. A = 235 ± 8,
B = 252 ± 8 and C = 269 ± 8 g L−1 sugars content.

for analysis. The seeds removed from the mesocarp were placed (A1 and A3.2), total flavonoids (TF1 and TF3.2) and flavonoids non-
in 50 mL of the same buffer solution used for skin extraction at anthocyanin (FNA1 and FNA3.2) [23,37]. The phenols extractability
pH 3.20 and then introduced in a controlled temperature room at indices determined were: cellular maturity index (EA%) and seed
25 ◦ C for a week [37]. The extract was then used for analysis. maturity index (Mp%) [1,22,23]. The latter index was determined by
taking into consideration the average ratio (TAR) between the total
2.3.2. Spectrophotometric methods phenols (A280 ) and the total anthocyanins of the skin (expressed as
Phenolic compounds of berry skin (sk) and seed (s) were g L−1 ), equal to the value 40 [23].
determined by spectrophotometric methods [37,38] using a UV- The EA% and Mp% indices were calculated as follows:
1601PC spectrophotometer (Shimazdu Scientific Instruments Inc.,
A1 − A3.2
Columbia, MD, USA). The total anthocyanins index (TAI) was EA% = × 100
A1
expressed as malvidin-3-glucoside chloride while flavanols reac-
tive to vanillin (flavanols vanillin assay, FVA) and total flavonoids A280 − ((A3.2/1000) × TAR)
Mp% = × 100
index (TFI) were expressed as (+)-catechin. The proanthocyanidin A280
content (PRO) was determined after acid hydrolysis with warming
(Bate–Smith reaction) using a ferrous salt (FeSO4 ) as catalyst. It was
expressed as cyanidin chloride. The relative standard deviations 2.5. Statistical analysis
of phenolic compound determination, based on repeated analyses
(n = 20) of ten sample extracts, were 1.14, 2.80, 0.93 and 1.74% for Statistical analyses were performed using the statistical soft-
TAI, FVA, TFI and PRO respectively. ware package SPSS (version 17.0; SPSS Inc., Chicago, IL, USA).

2.3.3. HPLC method 3. Results

Analysis of individual anthocyanins was performed after con-
centration of berry skins extract using a SEP-PAK C18 cartridge
(Waters Corporation, Milford, MA, USA) and elution with methanol.
The chromatograph system employed was a P100 pump equipped
with an AS3000 autosampler (Spectra Physics Analytical, Inc., San
Jose, CA, USA), a 20-mL Reodyne sample loop, a LiChroCART column
(25 cm × 0.4 cm i.d.) from Merck (Darmstadt, Germany) packed
with LiChrosphere 100 RP-18 5-␮m particles by Alltech (Deerfield,
IL, USA) and a Spectra Focus Diode Array Detector (Spectra Physics
Analytical, Inc., San Jose, CA, USA) operating at 520 nm. Formic acid
(10%, v/v) in water and formic acid (10%, v/v) with methyl alcohol
(50%, v/v) in water were used as mobile phases. Compounds were
identified according to chromatography conditions and peak iden-
tifications reported in literature [39,40]. Individual anthocyanin
concentrations were determined by comparing the area of the
appropriate peak with the total peak area and data were expressed
in percentages.
2.4. Phenol extractability indices
Phenol extractability indices were assessed in accordance with
the procedure described by Glories and Saint-Criq [26,27] modified
for Barbera grapes studies [23] using homogenized grapes by three
replicates of 100 berries for each soluble solids class and growing
area. The following parameters were determined in pH 1 and pH
3.2 solutions: total phenolic content (A280 ) [41], total anthocyanins
3.1. Mechanical properties
The mechanical characteristics of the berry skins of Barbera
grapes containing different levels of sugars sampled at each site of
cultivation are reported in Table 2. With the increase of the soluble
solids content in the pulp, no significant changes in the parameters
that characterize the skin hardness (Fsk , berry skin break force, and
Wsk berry skin break energy,) were observed. Similarly, for Esk (skin
Young’s modulus), the parameter used to measure the rigidity or
stiffness of tissues [35], no modifications were noted except for the
production area II in which a lower elasticity of the skin was shown
at higher levels of sugars in the berries. In contrast, the average skin
thickness increases in the order of increasing soluble solids in all
areas of production studied, in particular, in I and IV. In the same
soluble solids class, the berry skin mechanical properties show sig-
nificant differences in the different areas of production considered.
Environmental factors seem to play a dominant role in the char-
acterization of mechanical properties of Barbera berry skin with
respect to the ripening stage. However, specific studies allowing the
influence of single environmental factors (soil, temperature, water
stress) to be assigned have not yet been carried out, even if recent
studies showed an influence of rainfalls on berries firmness and in
particular on cohesiveness parameter [28,29]. Similar remarks can
be made about the mechanical properties of seeds which do not
present differences with the increase of soluble solids content for
186 F. Torchio et al. / Analytica Chimica Acta 660 (2010) 183–189

Table 3
Seed mechanical properties of Barbera grapes: Fs = seed break force, Ws = seed break energy, Es = seed Young’s modulus, DIs = seed deformation index.

Areas of production Soluble solid class Fs (N) Ws (mJ) Es (N mm−1 ) DIs (%)

A 47.742 ± 7.909 a,␤␥

11.315 ± 3.658 a,␤␥
I B 44.732 ± 9.110a,␤ 10.314 ± 3.156a,␣␤
C 48.128 ± 7.315a,␤ 11.214 ± 3.081a,␤
88.085 ± 15.752 a,␤
16.155 ± 2.723a,␤
84.538 ± 19.161a,␤ 15.761 ± 2.604a,␣
93.753 ± 16.720a,␤ 15.785 ± 2.375a,␤

A 44.353 ± 11.040a,␣␤ 9.266 ± 3.554a,␣␤ 101.872 ± 24.285a,␥ 13.701 ± 2.342a,␣

II B 44.283 ± 7.389a,␣␤ 9.234 ± 2.586a,␣ 97.555 ± 14.362a,␥ 13.952 ± 1.708a,␣
C 40.316 ± 5.823a,␣ 7.895 ± 1.813a,␣ 95.419 ± 12.561a,␤ 12.840 ± 1.566a,␣

A 39.249 ± 6.241a,␣ 8.476 ± 1.988a,␣ 74.354 ± 19.024a,␣ 16.273 ± 2.606a,␤

III B 37.303 ± 6.088a,␣ 8.226 ± 2.065a,␣ 68.534 ± 16.430a,␣ 16.181 ± 3.404a,␣
C 40.545 ± 5.457a,␣ 9.167 ± 2.137a,␣ 74.900 ± 16.464a,␣ 16.238 ± 3.840a,␤

A 51.190 ± 9.100a,␥ 11.552 ± 2.929a, ␥ 94.520 ± 19.197a,␤␥ 16.521 ± 2.477a,␤

IV B 51.278 ± 7.836a,␥ 11.652 ± 3.080a,␤ 95.073 ± 16.402a ,␤␥ 15.781 ± 2.643a,␣
C 50.644 ± 7.432a,␤ 11.802 ± 2.911a,␤ 89.303 ± 15.016a,␤ 16.459 ± 2.338a,␤

Average value ± standard deviation (n = 30). Different Latin letters indicate significant difference (p < 0.05) among the three level of sugar content in the same area of production.
Different Greek letters indicate significant difference (p < 0.05) among the grape of the different four areas of production at the same levels of sugar content. A = 235 ± 8,
B = 252 ± 8 and C = 269 ± 8 g L−1 sugars content.

Table 4
Skin phenolic composition of Barbera grapes: TAIsk = total anthocyanin index, TFIsk = total flavonoid index, PROsk = proanthocyanidins, FVAsk = flavanol vanillin assay.

Areas of Soluble TAIsk (mg kg−1 TFIsk (mg kg−1 (+)-catechin) PROsk (mg kg−1 FVAsk (mg kg−1
production solid class malvidine-3-glucoside cyanidin chloride) (+)-catechin)
chloride)

A 2144 ± 125a,␤ 6425 ± 228a,␤ 1030 ± 19a,␤␥ 43 ± 8a,␤

I B 2477 ± 146a,␤ 7015 ± 363a,␤ 1196 ± 55a,␣ 103 ± 4c,␥
C 2553 ± 249a,␤ 7088 ± 481a,␤␥ 832 ± 340a,␣ 80 ± 13b,␣

A 1822 ± 155a,␣␤ 5628 ± 328a,␣ 1163 ± 57a,␥ 100 ± 13b,␥

II B 1996 ± 119ab,␣ 5764 ± 267a,␣ 977 ± 167a,␣ 45 ± 12a,␣␤
C 2258 ± 162b,␣␤ 6238 ± 268a,␣␤ 664 ± 399a,␣ 83 ± 19ab,␣

A 2150 ± 124a,␤ 5856 ± 271a,␣␤ 595 ± 279a,␣ 41 ± 19a,␤

III B 2705 ± 150b,␤ 7160 ± 354b,␤ 854 ± 327a,␣ 57 ± 6a,␤
C 3107 ± 129c,␥ 7820 ± 416b,␥ 1017 ± 560a,␣ 86 ± 37a,␣

A 1684 ± 150a,␣ 5301 ± 293a,␣ 746 ± 14a,␣␤ 7 ± 3a,␣

IV B 1809 ± 231b,␣ 5620 ± 203b,␣ 965 ± 30b,␣ 29 ± 10ab,␣
C 2101 ± 141c,␣ 6066 ± 130b,␣ 965 ± 94b,␣ 58 ± 19b,␣

Average value ± standard deviation (n = 3). Different Latin letters indicate significant difference (p < 0.05) among the three level of sugar content in the same area of production.
Different Greek letters indicate significant difference (p < 0.05) among the grape of the different four areas of production at the same levels of sugar content. A = 235 ± 8,
B = 252 ± 8 and C = 269 ± 8 g L−1 sugars content.

each cultivation site (Table 3), but do allow the differentiation of FTs and PROs of several production areas were significant only in
seeds belonging to different growing areas. the soluble solids class A, i.e., that with the lowest level of sugar
(A = 235 ± 8 g L−1 ), while the differences for FVAs were significant
in all three categories of sugar level.
3.2. Phenolic composition of grapes

The phenolic composition of berry skins is shown in Table 4. The

values of TAIsk and TFIsk in Barbera grapes increase with the rise of
the sugar class in all cases. The differences were significant, partic-
ularly for the production areas in the province of Cuneo (sites III and
IV). Furthermore, the TAIs index shows significant differences also
in cultivation site II. However, no relationship was found between
PROsk and FVAsk values in dependence of different classes of soluble
solids.
Results detailing the variation of anthocyanin patterns are
reported in Table 5. With the increase of the soluble solids content,
some differences in the anthocyanin profile were found. In particu-
lar, significant changes were observed in the acylated anthocyanin
forms (sum of acetyl and cinnamoyl anthocyanins). At the same
stage of ripeness, differences in all skin phenolic parameters were
observed as a function of the cultivation site.
The phenolic composition of the berry seeds of Barbera grapes at
different levels of sugars and for each cultivation site is reported in
Table 6. PROs and FVAs do not show changes with respect to soluble
solids class. However, the differences of TFIS among the three sugar
content classes were significant in the production areas III and IV,
i.e., the same ones already shown for TFIsk . The differences between
3.3. Phenols extractability indices
Results concerning changes in the phenolic maturity indices
are shown in Table 7. With the rise of content of soluble solids
in the berries, a significant increase in the values of all measured
parameters was found in each Barbera production area. In fact, the
anthocyanin, flavonoid and flavonoid non-anthocyanin contents
increase in all cases, both in the extract at pH 1 and that at pH 3.2
with significant differences among different soluble solids classes.
Consequently, the calculated cellular maturity index (EA%) shows
no significant differences in almost all the zones studied. Only in
the production area III different values of EA% among soluble solids
classes were observed. Hence, at the same date, the EA% index
appears to be a rather homogeneous parameter for a given vine-
yard. On the contrary, the Mp% index shows a significant decrease
with the rise of sugars content in all Barbera vineyards studied.
On the other hand, for all phenolic extractability indices and in
particular for the EA% index, the area of production is deemed to
be the most dominant factor for the variability of these values.
 F. Torchio et al. / Analytica Chimica Acta 660 (2010) 183–189 187

Table 5
Anthocyanin profile of Barbera grapes.

Areas of Soluble solid Simple Acetyl- Cinnamoyl- Delphinidin Cyanidin Petunidin Peonidin Malvidin
production class glucosides (%) glucoside glucosidea derivatives derivatives derivatives derivatives derivatives
(%) (%) (%) (%) (%) (%) (%)

A 82.9 ± 0.3a,␤ 10.9 ± 0.1a,␣ 6.2 ± 0.3a,␣ 20.1 ± 0.2a,␥ 12.4 ± 1.4a,␤ 17.5 ± 0.2a,␤ 10.9 ± 0.6a,␤ 39.1 ± 1.8a,␣
I B 82.8 ± 0.7a,␥ 10.8 ± 0.3ab,␣ 6.4 ± 0.4a,␣ 20.0 ± 0.9a,␤ 12.4 ± 0.8a,␤ 17.7 ± 0.3a,␣␤ 10.5 ± 0.7a,␤ 39.4 ± 2.0a,␣
C 83.5 ± 0.1a,␥ 10.1 ± 0.4b,␣ 6.3 ± 0.4a,␣ 20.1 ± 0.8a,␤ 12.2 ± 0.3a,␥ 17.6 ± 0.2a,␣ 10.2 ± 0.3a, ␥ 39.8 ± 0.4a,␣

A 78.8 ± 1.6a,␥ 11.0 ± 1.5a,␣␤ 9.9 ± 0.1a,␥ 16.1 ± 1.2a,␤ 5.9 ± 0.3a,␣ 18.6 ± 0.4a,␥ 5.6 ± 0.1a,␣ 53.4 ± 1.2a,␤␥
II B 79.2 ± 0.7a,␤ 11.7 ± 0.4a,␣ 9.1 ± 0.4b,␤ 17.7 ± 1.4a,␣␤ 6.3 ± 0.7a,␣ 18.8 ± 0.9a,␤ 6.0 ± 0.3ab,␣ 51.2 ± 2.3a,␤␥
C 80.5 ± 0.9a,␤ 11.0 ± 0.9a,␣ 8.4 ± 0.1c,␤ 17.6 ± 1.0a ,␣ 7.9 ± 1.2a,␣␤ 18.4 ± 0.5a,␣ 6.9 ± 0.6b,␣␤ 49.0 ± 3.0a,␤

A 80.7 ± 0.1a,␤␥ 11.0 ± 0.4a,␣␤ 8.3 ± 0.4a,␤ 18.1 ± 1.3a,␤␥ 7.2 ± 0.8ab,␣ 18.5 ± 0.5a,␤␥ 8.4 ± 0.6a,␤ 47.7 ± 2.0a,␤
III B 80.8 ± 0.6a,␤␥ 10.9 ± 0.5a,␣ 8.2 ± 0.2a,␤ 19.2 ± 0.5ab,␤ 6.7 ± 0.4a,␣ 19.3 ± 0.6a,␤ 7.1 ± 0.5a,␣ 47.6 ± 0.2a,␤
C 82.4 ± 0.5b,␤␥ 10.4 ± 0.2a,␣ 7.1 ± 0.4b,␣ 20.8 ± 0.5b,␤ 9.9 ± 1.7b,␤␥ 18.9 ± 0.8a,␣ 8.3 ± 1.0a,␤␥ 42.1 ± 2.3b,␣

A 75.7 ± 0.7a,␣ 13.2 ± 0.2a,␤ 11.1 ± 0.6a,␦ 13.5 ± 0.2a,␣ 6.1 ± 1.1a,␣ 15.6 ± 0.3a,␣ 9.4 ± 1.7a,␤ 55.3 ± 2.8a,␥
IV B 73.5 ± 1.5a,␣ 15.0 ± 0.5b,␤ 11.4 ± 1.1a,␥ 14.8 ± 1.7a,␣ 5.3 ± 0.7a,␣ 17.0 ± 0.5a,␣ 6.6 ± 0.6b,␣ 56.3 ± 3.4a,␥
C 75.0 ± 1.2a,␣ 14.5 ± 0.6ab,␤ 10.6 ± 0.6a,␥ 15.5 ± 0.9a,␣ 5.3 ± 0.3a,␣ 17.4 ± 0.5a,␣ 6.4 ± 0.6b,␣ 55.4 ± 1.2a,␥

Average value ± standard deviation (n = 3). Different Latin letters indicate significant difference (p < 0.05) among the three level of sugar content in the same area of production.
Different Greek letters indicate significant difference (p < 0.05) among the grape of the different four areas of production at the same levels of sugar content. A = 235 ± 8,
B = 252 ± 8 and C = 269 ± 8 g L−1 sugars content.
a
Cinnamoyl-glucosides included both p-coumaroyl and caffeoyl anthocyanin forms.

Table 6
Seeds phenolic composition of Barbera grapes: TFIs = total flavonoid index, PROs = proanthocyanidins, FVAs = flavanol vanillin assay.

Areas of production Soluble solid class TFIs (mg kg−1 (+)-catechin) PROs (mg kg−1 cyanidin chloride) FVAs (mg kg−1 (+)-catechin) FVAs/PROs

A 4567 ± 214 a,␥

1032 ± 146 a,␤
519 ± 57 a,␥
0,50 ± 0,03a,␤
I B 4267 ± 203a,␣ 958 ± 36a,␣ 474 ± 41a,␤ 0.50 ± 0.05a,␣
C 4494 ± 279a,␣ 927 ± 148a,␣ 565 ± 22a,␤ 0.62 ± 0.12a,␣

A 3939 ± 78a,␤ 912 ± 18a,␤ 374 ± 16a␤ 0.41 ± 0.01a,␣

II B 4026 ± 132a,␣ 966 ± 68a,␣ 428 ± 5b,␤ 0.44 ± 0.03ab,␣
C 4245 ± 319a,␣ 1092 ± 120a,␣ 539 ± 20c,␤ 0.50 ± 0.04b,␣

A 3433 ± 171a,␣ 630 ± 48a,␣ 309 ± 27a,␣␤ 0.49 ± 0.01a,␣␤

III B 4003 ± 67b,␣ 785 ± 62a,␣ 380 ± 47a,␣ 0.48 ± 0.02a,␣
C 4030 ± 161b,␣ 821 ± 109a,␣ 393 ± 84a,␣ 0.48 ± 0.04a,␣

A 3703 ± 107a,␣␤ 516 ± 64a,␣ 232 ± 17a,␣ 0.45 ± 0.06a,␣␤

IV B 4334 ± 183b,␣ 826 ± 49b,␣ 385 ± 47b,␣␤ 0.46 ± 0.03a,␣
C 4210 ± 79b,␣ 837 ± 47b,␣ 458 ± 50b,␣␤ 0.55 ± 0.03b,␣

Average value ± standard deviation (n = 3). Different Latin letters indicate significant difference (p < 0.05) among the three level of sugar content in the same area of production.
Different Greek letters indicate significant difference (p < 0.05) among the grape of the different four areas of production at the same levels of sugar content. A = 235 ± 8,
B = 252 ± 8 and C = 269 ± 8 g L−1 sugars content.

4. Discussion demonstrated on Cabernet franc growing in different terroirs of the

4.1. Mechanical properties
Previous studies have indicated the potential of grape texture
measurement to estimate grape ripeness [42]. In particular, dur-
ing ripening, the grapes become more and more soft; this is the
result of significant changes in parietal constituent composition
notably in pulp cells [17]. Therefore, the texture test conduced on
whole berry (Double Compression Test), which assess the parame-
ters such as firmness, cohesiveness, gumminess, is actually the best
test to monitoring the ripeness, although the values of parameters
measured can be affected by rainfalls [28,43]. In this type of test,
pulp and skin data are aggregate. On the contrary, by puncture test
conduced with needle probe [34] or thin rounded probe [28] only
skin characteristics can be defined. Actually, the break skin force
(Fsk ) could be considered an important parameter to be monitored
for the assessment of the anthocyanins extractability. During mac-
eration in a model hydroalcoholic solution, the grapes with higher
values of Fsk produced extracts with a higher content of total antho-
cyanin [24]. Significant interactions between stage of ripening and
skin hardness were found in the individual anthocyanin composi-
tion of extracts [25]. From veraison to ripeness, above all in the first
phases, an increase of Fsk is shown, with a steady or slight decrease
in the proximity of the harvest and a new increase in the phases of
over ripeness and on-vine drying process [44]. However, as already
Loire Valley region [28] and confirmed in this work on Barbera
grape, the behaviour of the Fsk values close to harvest could place
a limit on the choice of this parameter as an indicator of maturity
of grapes. Nevertheless, the skin hardness at each level of soluble
solids content is an effective tool to discriminate among different
vineyards of cv Barbera, although Fsk values are strongly affected by
climate trends of the vintages [35]. Knowledge about the mechan-
ical properties of seeds is scarce in the literature and thus, as yet,
it is not possible to understand the importance of the significant
differences noted in Fs , Ws and Es parameters among the different
areas of production of Barbera grapes. However, from our current
data, probably because of their high variability, trends in mechan-
ical parameters, possibly correlated with the increase of soluble
solids content, could not be discerned, although such trends might
be anticipated by the numerous changes in tissues present in the
seeds during the ripening [13]. Instead, during withering on-vine
this trend is clearly demonstrated [44]. Further research examining
mechanical properties and extractability of phenols from seeds is
necessary.
4.2. Phenolic composition of grapes
In the literature there are many studies describing the change of
the content in phenolic compounds during ripening [3,17]. In the
same vineyards [26] and also among the grapes produced by the
188 F. Torchio et al. / Analytica Chimica Acta 660 (2010) 183–189

Average value ± standard deviation (n = 3). Different Latin letters indicate significant difference (p < 0.05) among the three level of sugar content in the same area of production. Different Greek letters indicate significant difference
Phenols extractability indices of Barbera grapes: A3.2 = total anthocyanins extracted at pH 3.2; TF3.2 = total flavonoids extracted at pH 3.2, FNA3.2 = flavonoids non-anthocyanin extracted at pH 3.2, A280 = total phenolic content, same vine in different clusters, differences in the sugar content are

75 ± 1a,␤␥
69 ± 1b,␣
69 ± 1b,␤

69 ± 1b,␤

65 ± 1b,␣
73 ± 1a,␤

73 ± 2a,␤

70 ± 1a,␣
69 ± 1a,␣

75 ± 1a,␤
72 ± 1b,␥
76 ± 1a,␥
present [45]. Therefore, sampling methods can be a determinant
Mp (%) factor for the study of phenolic compounds [45]. In this work on
Barbera grapes, many differences were found in several phenolic

56 ± 2a,␣␤
parameters of skins and seeds among the different classes of sol-

64 ± 2b,␤
60 ± 2a,␤
60 ± 2a,␤
61 ± 4a,␣

58 ± 1a,␤
58 ± 1a,␤
56 ± 2a,␣

54 ± 2a,␣
53 ± 1a,␣
57 ± 3a,␣
65 ± 3b,␥
EA (%)

uble solids in the same vineyard at the same date of harvesting.

Nevertheless, these parameters and, in particular TAIsk , were also
strongly influenced by the production area. Anthocyanins synthe-
sis in the same growing location is influenced by many factors such

1615 ± 109b,␣
2035 ± 156c,␤

1476 ± 29a,␣␤

1613 ± 24a,␤␥
(+)-catechin)

as temperature, sunlight, vigor and leaf surface area [3,15]. Differ-

1370 ± 127a,

1782 ± 27b,␣

1726 ± 44b,␣
1474 ± 21a,␣

1563 ± 62a,␤
1677 ± 19b,␥
1720 ± 69a,␥
1947 ± 28b,␦
A1 = total anthocyanins extracted at pH 1; TF1 = total flavonoids extracted at pH 1, FNA1 = flavonoids non-anthocyanin extracted at pH 1, EA% = cellular maturity index, Mp% = seed maturity index.

ences observed in the anthocyanins profile of Barbera grapes do not

(mg kg−1

seem to be correlated with a different level of accumulation of sug-

FNA1

ars. In fact, the anthocyanin di-substitued forms were not prevalent

in less ripe grapes (level A) and the tri-substituted ones do not pre-
vail in those grapes with a high level of sugar content (C), as already
TF1 (mg kg−1 (+

3233 ± 393a,␣␤

reported frequently in literature [3,14,46].

3300 ± 24a ,␣␤
4280 ± 243b,␤
3594 ± 170b,␣

4738 ± 170b,␤

3216 ± 121a,␣
4572 ± 313c,␤
4143 ± 75b,␤

3769 ± 80b,␣
3552 ± 73a,␤

2976 ± 58a,␣
3178 ± 0a,␣

For the FTIsk , FVAsk and PROsk parameters the differences noted
catechin)

were significant only in a few cases. Also, in other studies no signif-

icant increase in skin tannin content was observed during sugars
(p < 0.05) among the grape of the different four areas of production at the same levels of sugar content. A = 235 ± 8, B = 252 ± 8 and C = 269 ± 8 g L−1 sugars content. accumulation in the pulp [45]. But, for these parameters, differ-
ences among the growing locations were very evident (Table 4).
The seed tannins play an important role during red winemak-
malvidin-3-glucoside

ing. During ripening a general decline in seed tannins has been

observed using several analytical methods. This may reflect their
1280 ± 183a,␤
1787 ± 179b,␥
1742 ± 108c,␤

covalent attachment to the insoluble matrix of the seed, render-

1508 ± 32b,␤

1171 ± 15b,␣

1403 ± 85b,␣
1135 ± 41a,␣
1031 ± 20a,␣

1359 ± 42c,␣

2030 ± 98c,␥

1159 ± 0a,␣␤
A1 (mg kg−1

1259 ± 3a,␤

ing them unavailable for extraction during winemaking [3,5,15]. In

chloride)

accordance with this, the seed maturity index decreases (Table 7),
but the phenolic concentration (TFIs , PROs and FVAs ) showed no
significant differences (Table 6).
83 ± 1b,␣␤
78 ± 3a,␣␤

82 ± 3ab,␤

87 ± 4ab,␤
85 ± 2b, ␤
73 ± 4a,␣
86 ± 4b,␤

79 ± 0b,␣
70 ± 2a,␣

75 ± 6a,␣

91 ± 1a,␤
75 ± 0a,␣

4.3. Phenols extractability indices

A280

To produce a wine with a high color intensity, the accumulation

of anthocyanins in the skin represents a necessary, but not always
sufficient condition. In addition, it is also necessary that cells are
FNA3.2 (mg kg−1

886 ± 118b,␣␤

able to completely release the anthocyanins during the maceration

914 ± 181a,␤

611 ± 165a,␣

985 ± 70b,␣␤
692 ± 48a,␣␤

912 ± 42ab,␤
(+)-catechin)

1068 ± 14b,␤
894 ± 10b,␣
1089 ± 82a,␤

742 ± 52a,␣
1102 ± 31a,␥

833 ± 6a,␣␤

[3,5,10,11,17,30]. For Barbera and other Italian grape varieties [23],

a relationship between the grape A1 and A3.2 values and relative
anthocyanin content in the wines was found, while EA% for these
varieties turns out to be an efficient index to assess the kinetics of
release of anthocyanins. In these studies it was observed that, at the
2075 ± 119b,␤

1795 ± 102b,␤
1652 ± 218a,␣

1429 ± 252a,␣
1463 ± 58b,␣

1869 ± 85b,␣
1330 ± 42a,␣

1614 ± 29a,␣
1692 ± 51a,␤
1982 ± 12b,␥

2136 ± 19c,␤

same date, the A1 and A3.2 values depend significantly on the con-
1766 ± 0c,␣
catechin)
(mg kg−1

tent of sugars present. In accordance with previous studies [47], the

TF3.2

changes of EA% values seem to depend mainly on the date of harvest

(+)-

rather than on the sugar contents in the berries. Moreover, in accor-

dance with Rio Segade [21], the Fsk and Spsk parameters correlated
with EA%, show the same behaviour as this index. In fact, they do
507 ± 25a,␣␤
3-glucoside

537 ± 24a ,␤
605 ± 13b,␤

607 ± 10b,␣
625 ± 11a,␤
562 ± 60a,␤
677 ± 25c,␤

495 ± 4b,␣
438 ± 4a,␣

733 ± 4b,␥
599 ± 7c,␣

536 ± 6a,␥
malvidin-

not change with the increase of the content of soluble solids, but,
chloride)
(mg kg−1

instead, they are more influenced by the production area (Table 2).
A3.2

The ripening process determines the reduction of extractability

of tannins from seeds during the winemaking because of the for-
mation of covalent bond among the polymer flavan-3-ols sub-unit
Soluble solid class

[11,15]. These results are in agreement with the diminution of the

Mp% index, observed in this work with the increase of sugar content
(Table 7).
A

5. Conclusions
B

B
C

The separate effects of concentration of soluble solids and the

Areas of production

location of growth on the physical–chemical characteristics of

Barbera grapes were defined. The areas of production and stage
of ripening do not influence the several parameters analyzed to
the same extent. Higher influences on all phenolic composition,
Table 7

extractability indices and mechanical parameters measured are

IV
III
II
I

attributable to terroirs.
 F. Torchio et al. / Analytica Chimica Acta 660 (2010) 183–189
At the same date, in grapes ripened in a vineyard, the mechanical
proprieties of the skin and the cellular maturity index were quite
uniform even if the berries did present different levels of sugars.
These indices were, however, greatly influenced by the produc-
tion areas. In contrast, the content of phenols generally increases
parallel to the sugar content of the grape although the differences
were more marked among the growing areas. The increase was
particularly significant for the anthocyanin indices TAI, A1 and A3.2.
However, further studies are necessary to increase our knowl-
edge about grape mechanical properties. In particular, defined the
relevance of the terroirs, new researches will be able to find, for
each production area, the relationships among soil characteristics,
bioclimatic indices, such as Huglin index (heliothermal index), STA
index (sum of the active temperatures) or SET index (sum of daily
thermal excursions), and their expression on physical berries fea-
tures. Moreover, histological studies will be necessary to explain
the relation between tissues characteristics and seeds and skins
hardness.
References
[1] G. González-Neves, D. Charamelo, J. Balado, L. Barreiro, R. Bochicchio, G. Gato,
G. Gil, A. Tessore, A. Carboneau, M. Moutounet, Anal. Chim. Acta 513 (2004)
191.
[2] B.S. Sun, T. Pinto, M.C. Leandro, J.M. Ricardo Da Silva, M.I. Spranger, Am. J. Enol.
Vitic. 50 (1999) 179.
[3] M. Downey, N.K. Dokoozlian, M.P. Krstic, Am. J. Enol. Vitic. 57 (2006) 257.
[4] J.M. Ryan, E. Revilla, J. Agric. Food Chem. 51 (2003) 3372.
[5] A. Ortega-Regules, I. Romero-Cascales, J.M. Ros-García, A.B. Bautista-Ortín, J.M.
López-Roca, J.I. Fernández-Fernández, E. Gómez-Plaza, J. Int. Sci. Vigne Vin. 42
(2008) 147.
[6] A.M. Jordão, J.M. Ricardo Da Silva, O. Laureano, Vitis 37 (1998) 93.
[7] N. Mateus, J.M. Machado, V. De Freitas, J. Sci. Food Agric. 82 (2002) 1689.
[8] Y. Cadot, M.T. Minana Castello, M. Chevalier, Anal. Chim. Acta 563 (2006) 65.
[9] V.A.P. de Freitas, Y. Glories, A. Monique, Am. J. Enol. Vitic. 51 (2000) 397.
[10] J.A. Kennedy, Y. Hayasaka, S. Vidal, E.J. Waters, G.P. Jones, J. Agric. Food Chem.
49 (2001) 5348.
[11] J.F. Habertson, J.A. Kennedy, D.O. Adams, Am. J. Enol. Vitic. 53 (2002) 54.
[12] J.A. Kennedy, M.A. Matthews, A. Waterhouse, Am. J. Enol. Vitic. 53 (2002) 268.
[13] Y. Cadot, M.T. Minana Castello, M. Chevalier, J. Agric. Food Chem. 54 (2006)
9206.
[14] J.A. Kennedy, M.A. Matthews, A. Waterhouse, Phytochemistry 55 (2000) 77.
[15] M.O. Downey, J.S. Harvey, S.P. Robinson, Aust. J. Grape Wine Res. 9 (2003) 15.
189
[16] J.A. Kennedy, G.J. Troup, J.R. Pilbrown, D.R. Hutton, D. Hewitt, C.R. Hunter, R.
Ristic, P.G. Iland, G.P. Jones, Aust. J. Grape Wine Res. 6 (2000) 244.
[17] P. Ribéreau-Gayon, Y. Glories, A. Maujean, D. Dubourdieu, Handbook of Enology,
vol. 2. The Chemistry of Wine Stabilization and Treatments, John Wiley and Sons
Ltd. Baffins Lane, Chichester, 2000.
[18] A. Ortega-Regules, I. Romero-Cascales, J.M. Ros-García, J.M. López-Roca, E.
Gómez-Plaza, Anal. Chim. Acta 563 (2006) 26.
[19] F. Mattivi, U. Vrhovsek, D. Masuero, D. Trainotti, Aust. J. Grape Wine Res. 15
(2009) 27.
[20] S. Vidal, D. Cartalade, J.-M. Souquet, H. Fulcrand, V. Cheynier, J. Agric. Food
Chem. 50 (2002) 2261.
[21] S. Rio Segade, L. Rolle, V. Gerbi, I. Orriols, J. Food Comp. Anal. 21 (2008) 644.
[22] I. Romero-Cascales, A. Ortega-Regules, J.M. López-Roca, J.I. Fernández-
Fernández, E. Gómez-Plaza, Am. J. Enol. Vitic. 56 (2005) 212.
[23] E. Cagnasso, L. Rolle, A. Caudana, V. Gerbi, Ital. J. Food Sci. 20 (2008) 365.
[24] L. Rolle, F. Torchio, G. Zeppa, V. Gerbi, J. Int. Sci. Vigne Vin. 42 (2008) 157.
[25] L. Rolle, F. Torchio, G. Zeppa, V. Gerbi, Am. J. Enol. Vitic. 60 (2009) 93.
[26] Y. Glories, M. Augustin, Acta Colloque J. Tech. (1993) 56.
[27] N. Saint-Criq, N. Vivas, Y. Glories, Rev. Fr. Oenol. 173 (1998) 22.
[28] C. Maury, E. Madieta, M. Le Moigne, E. Mehinagic, R. Siret, F. Jourjon, J. Texture
Stud. 40 (2009) 511.
[29] M. Le Moigne, C. Maury, D. Bertrand, F. Jourjon, Food Qual. Prefer. 19 (2008)
220.
[30] J.M. Cortell, M. Halbleib, A.V. Gallagher, T.L. Righetti, J.A. Kennedy, J. Agric. Food
Chem. 53 (2005) 5798.
[31] D. Fournand, A. Vicens, L. Sidhoum, J.-M. Souquet, M. Moutounet, V. Cheynier,
J. Agric. Food Chem. 54 (2006) 7331.
[32] C. Lovisolo, A. Schubert, A. Cellino, Quad. Vitic. Enol. Univ. Torino 28 (2006) 249,
http://www.arboree.unito.it/QSVE28/21Lovisolo.pdf.
[33] http://www.regione.piemonte.it/agri/suoli terreni/dwd/atla carto250/piem
legsuoli 250.pdf.
[34] H. Letaief, L. Rolle, G. Zeppa, V. Gerbi, J. Sci. Food Agric. 88 (2008) 1567.
[35] H. Letaief, L. Rolle, V. Gerbi, Am. J. Enol. Vitic. 59 (2008) 323.
[36] S. Guidoni, F. Mannini, A. Ferrandino, N. Argamante, R. Di Stefano, Am. J. Enol.
Vitic. 48 (1997) 438.
[37] R. Di Stefano, M.C. Cravero, Riv. Vitic. Enol. 44 (1991) 37.
[38] A. Rigo, F. Vianello, G. Clementi, M. Rossetto, M. Scarpa, U. Vrhovšek, F. Mattivi,
J. Agric. Food Chem. 48 (2000) 1996.
[39] F. Pomar, M. Novo, A. Masa, J. Chromatogr. A 1094 (2005) 34.
[40] L. Rolle, S. Guidoni, J. Int. Sci. Vigne Vin. 41 (2007) 193.
[41] P. Ribereau-Gayon, Chim. Anal. 52 (1970) 627.
[42] J.P. Robin, P. Abbal, J.M. Salmon, J. Int. Sci. Vigne Vin. 31 (1997) 127.
[43] M. Grotte, Y. Cadot, A. Poussier, D. Loonis, E. Pietri, F. Duprat, G. Barbeau, J. Int.
Sci. Vigne Vin. 35 (2001) 87.
[44] L. Rolle, F. Torchio, S. Giacosa, V. Gerbi, J. Sci. Food Agric. 89 (2009) 1973.
[45] A. Carbonneau, A. Moueix, N. Leclair, J.L. Renoux, Bull. OIV 64 (1991) 679.
[46] P.K. Boss, C. Davies, S.P. Robinson, Plant Mol. Biol. 32 (1996) 565.
[47] E. Cagnasso, A. Caudana, L. Rolle, V. Gerbi, Quad. Vitic. Enol. Univ. Torino 28
(2006) 61, http://www.arboree.unito.it/QSVE28/06Cagnasso.pdf.