See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/309412694

Cluster architecture of old, neglected Croatian grapevine varieties (Vitis

vinifera L.)

Article · January 2015

CITATIONS READS

2 121

6 authors, including:

Zdunić Goran Ana Domagoja Mucalo

Institut za jadranske kulture i melioraciju krša Institute for Adriatic Crops and Karst Reclamation; Institut za jadranske kulture i …
54 PUBLICATIONS   128 CITATIONS    10 PUBLICATIONS   22 CITATIONS   

SEE PROFILE SEE PROFILE

Irena Budić-Leto Iva Humar

Institut za jadranske kulture i melioraciju krša Institut za jadranske kulture i melioraciju krša
35 PUBLICATIONS   172 CITATIONS    13 PUBLICATIONS   25 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Unravelling the varietal typicity of wines and olive oils from Croatian native varieties View project

Biotechnological Parameters of Premium-quality Dalmatian Dessert Wine Prošek View project

All content following this page was uploaded by Zdunić Goran on 09 January 2017.

The user has requested enhancement of the downloaded file.

Vitis 54 (Special Issue), 177–180 (2015)

Cluster architecture of old, neglected Croatian grapevine varieties

(Vitis vinifera L.)

G. ZDUNIĆ1), A. MUCALO1), I. BUDIĆ-LETO1), I. HUMAR1), I. PEJIĆ2) and E. MALETIĆ2)

1)
Institute for Adriatic Crops and Karst Reclamation, Croatia
2)
University of Zagreb, Faculty of Agriculture, Croatia

Summary 2004). Although cluster architecture is a very important trait

for grape production, little is known about its genetic ba-
Grape cluster architecture is an important trait af- sis. Recently, 19 quantitative trait loci (QTL) and 50 genes
fecting fruit composition and yield. Damage caused by were identified that are involved in genetic control of clus-
gray mold (Botrytis cinerea) is directly related to clus- ter architecture (CORREA et al. 2014). This suggests a com-
ter density and some other environmental factors. The plex genetic base for grape cluster architecture.
aim of this study was to quantify characteristics of fully Damage caused by gray mold (Botrytis cinerea) is di-
mature clusters of eleven old Croatian grapevine varie- rectly related to cluster density and some other environ-
ties (Vitis vinifera L.) which so far have not been phe- mental factors. Cultivars with compact clusters are more
notypically described in detail, and to examine Botrytis sensitive to gray mold than those with loose clusters (VAIL
cinerea disease severity on their grapes under field con- and MAROIS 1991, MLIKOTA GABLER et al. 2003). There
ditions. Eight variables that describe the appearance are some viticultural techniques like cluster thinning or
of clusters were quantified in three consecutive years cluster division (MOLITOR et al. 2012), leaf removal (PONI
including cluster weight, length, width, volume, com- et al. 2006), or anti-transpirants application (PALLIOTTI
pactness, rachis weight, number of berries and single et al. 2010) that can induce cluster loosening, but resist-
berry weight. Two-factor analysis of variance provided ant cultivars are still intriguing for viticultural practis. The
estimates of varietal and annual differences. Significant cluster compactness is usually estimated visually accord-
differences among eleven old Croatian varieties for all ing to OIV system descriptors (OIV descriptor No. 204),
cluster traits were determined. Results suggest diverse and indirectly through several quantitative criteria that are
cluster morphology within the old Croatian varieties calculated from cluster measurement variables (TELLO and
revealing three variety groups by PCA analysis and IBÁŇEZ 2014).
their varying degrees of sensitivity to Botrytis. These The aim of this study was to quantify characteristics
phenotypic data on cluster structure represent a basis of fully mature clusters of eleven old Croatian grapevine
for future research, such as QTL analysis or assessment varieties (Vitis vinifera L.), and to examine Botrytis cinerea
of productivity of these varieties. disease severity on their grapes under field conditions. The
varieties have been selected to represent different cluster
K e y w o r d s : cluster morphology; Botrytis; ampelogra-
structure and degree of cluster density.
phy; PCA analysis.

Introduction Material and Methods

Grape cluster architecture is an important trait affect- P l a n t m a t e r i a l : Eleven grapevine varieties (Vitis
ing fruit composition and yield. At full maturity, there are vinifera L.) previously identified by microsatellite markers
clearly visible differences among cultivars in cluster shape, (ZDUNIĆ et al. 2013) were selected for this study: 'Babić',
from very small to large, compact to very loose, and long 'Babica', 'Crljenak Viški', 'Glavinuša', 'Grk', 'Lasina', 'Plava
to very short. Various cluster shapes and forms were likely Lovora', 'Plavac Mali Sivi', 'Pošip Bijeli', 'Svrdlovina', and
one of the earliest criteria in human selection of grape cul- 'Vrškajica'. They are very poorly represented in current
tivars, and later assisted in the morphological classification grape production, but could be very important for future
of grapevine cultivars into eco-geographical groups (NE- breeding programs and/or production revitalization. The
GRUL 1938). Yield, one of the most important characteris- clusters from each variety were collected from the grape
tics in viticulture, is influenced by numerous genetic loci germplasm repository of the Institute for Adriatic Crops
and environmental factors. The cluster weight, berry size, and Karst Reclamation, which is located in Split (latitude,
and number of berries per cluster are essential components 43°30.335N; longitude, 016º29.855E; 14 m asl) in Croatia.
that significantly determine yield. The varieties were grown under identical cultivation and
There are many cluster components, but cluster ar- climatic conditions in a vineyard established in 2005. They
chitecture is mainly determined by the total rachis length, are grafted onto 1103 Paulsen rootstocks and trained on the
weight, and length of lateral internodes (SHAVRUKOV et al. bilateral spur cordon system with spacing of 1.0 m among

Correspondence to: Dr. G. ZDUNIĆ, Institute for Adriatic Crops and Karst Reclamation, Put Duilova 11, 21000 Split, Croatia,
Fax: +385-21-316-584. E-mail: [email protected]
178 G. ZDUNIĆ et al.

vines in a row and 2.0 m between rows. The soil in the Results and Discussion
vineyard is clay-sandy. The average annual temperature
of the location was 16.1 °C; the average rainfall rate was Significant differences among eleven old Croatian va-
782.8 mm (average values for 1971 to 2000 recorded at the rieties for all cluster traits were determined during three
meteorological station in Split). consecutive vintages (Table). Clusters of 'Grk' were long-
Determination of cluster structure: est, with average length of 22.6 cm, while 'Babić' clusters
Ten representative clusters were collected from each vari- were the widest (average 14.3 cm). Cluster weight varied
ety at full maturity in three consecutive years (2008, 2009 from 176.0 to 453.0 g and the number of berries per clus-
and 2010) to determine cluster dimensions. Cluster weight ter ranged from 68 to 298, depending on the variety. 'Grk'
(g), length (cm) and width (cm) were determined. The had the most berries per cluster but also the smallest ber-
number of berries was counted for each cluster and berry ries; the average single berry weight was 1.3 g. 'Babić' had
weight determined. Rachis weight (g) was determined af- the largest single berry weight (3.1 g) and rachis weight
ter manual removal of berries. The morphological volume (15.9 g), resulting also in the largest cluster volume (1200
of each cluster was calculated using the standard formula cm3). The lowest cluster volume observed was in 'Svrd-
for a cone, Vcone = 1/3πr2l, where the radius r was taken lovina' (496 cm3). The relative differences among varieties
as equivalent to half of the cluster width, and l = cluster were not consistent among years, which resulted in a sig-
length. Cluster compactness was evaluated using the index nificant interaction between year and variety.
Cl-12 = cluster weight (g) / [cluster length (cm)]2 (TELLO At full maturity, there were clearly visible differences
and IBANEZ 2014). between the compact and loose clusters. We quantified
Assessment of Botrytis cinerea dis- cluster compactness based on the Cl-12 index (ratio of
e a s e s e v e r i t y : The severity of Botrytis cinerea at- cluster weight to length), which correlated significantly
tack was assessed at full maturity by evaluating all clusters with the visual cluster compactness categories of OIV de-
of ten vines from each variety using the EPPO guide for scriptor No. 204 proposed for rapid evaluation (TELLO and
classifying visually observed symptoms of infection in IBÁŇEZ 2014). Cluster compactness significantly varied
seven classes: (0 %, 1 to 5 %, 6 to 10 %, 11 to 25 %, 26 to among varieties and ranged from 0.41 for 'Plava Lovora'
50 %, 51 to 7 5%, 76 to 100 %). (very loose cluster) to 1.21 for 'Babica' (dense cluster).
D a t a a n a l y s i s : Two-factor analysis of variance Differences among other varieties were significant and
(ANOVA) for variety x year using the statistical software suggest a different degree of cluster compactness among
Statistica 8.0 (StatSoft, Inc., USA) provided estimates of them. Numerous studies have demonstrated that varie-
varietal and annual differences. Standard error was deter- ties with compact clusters result in high Botrytis infection
mined using Fisher LSD with a 0.05 significance level. while varieties with loose clusters are less susceptible (VAIL
Principal component analysis (PCA) was used as a multi- and MAROIS 1991, MLIKOTA GABLER et al. 2003). Among
variate analytical tool to differentiate among varieties, us- our varieties, visual symptoms of B. cinerea infection on
ing average values for eight grape cluster variables. clusters ranged from 0 % to 50 % depending on variety.

Ta b l e

Average values and standard errors of cluster dimensions at harvest of eleven old Croatian varieties in 2008, 2009 and 2010

Cluster
Cluster length Cluster width Cluster weight Rachis fresh Cluster Number of Single berry
Variety compactness
(cm) (cm) (g) weight (g) volume (cm3) berries weight (g)
index*
Babić 20.5 ± 0.6d 14.3 ± 0.6e 453.0 ± 34.6e 15.9 ± 1.4e 1200 ± 132.5f 1.07 ± 0.06de 153 ± 12.8ac 3.1 ± 0.2e
Babica 17.1 ± 0.5bc 12.7 ± 0.5abc 349.7 ± 25.1c 14.4 ± 1.1bde 769 ± 70.4abd 1.21 ± 0.08f 158 ± 16.3ad 2.4 ± 0.1g
Crljenak viški 18.1 ± 0.5c 12.0 ± 0.5a 271.8 ± 15.1a 12.6 ± 0.7bc 728 ± 70.6abd 0.88 ± 0.07a 209 ± 20.6e 1.6 ± 0.2a
Glavinuša 17.8 ± 0.5c 12.3 ± 0.6ab 209.8 ± 16.2bd 9.1 ± 0.7a 784 ± 99.4bd 0.68 ± 0.05b 122 ± 8.9b 1.8 ± 0.1b
Grk 22.6 ± 0.5e 13.5 ± 0.4ce 377.3 ± 17.3c 13.6 ± 0.5bd 1136 ± 81.9ef 0.74 ± 0.03bc 298 ± 21.6g 1.3 ± 0.1f
Lasina 16.8 ± 0.4abc 12.0 ± 0.5a 277.5 ± 16.1a 13.1 ± 1.0bc 700 ± 73.1abc 0.99 ± 0.05ad 147 ± 15.0abc 2.1 ± 0.1d
Plava Lovora 21.0 ± 0.8d 11.7 ± 0.8ad 176.0 ± 11.6b 9.7 ± 0.9a 927 ± 162.5d 0.41 ± 0.03g 133 ± 15.2abc 1.5 ± 0.1a
Plavac Mali Sivi 15.5 ± 0.3a 12.3 ± 0.3abc 280.5 ± 17.0a 11.1 ± 0.6 ac
640 ± 39.7abc 1.15 ± 0.05ef 156 ± 11.8a 1.9 ± 0.1c
Pošip Bijeli 21.5 ± 0.6de 13.3 ± 0.5bce 383.6 ± 18.0c 15.3 ± 1.2 de
1083 ± 119.1ef 0.86 ± 0.05ac 184 ± 9.7de 2.1 ± 0.1d
Svrdlovina 15.9 ± 0.3ab 10.7 ± 0.4d 231.6 ± 11.9ad 10.2 ± 0.5a 496 ± 35.5c 0.93 ± 0.04a 128 ± 8.1bc 1.8 ± 0.1bc
Vrškajica 16.4 ± 0.6ab 10.7 ± 0.5d 182.3 ± 11.9b 9.2 ± 0.7a 555 ± 66.9ac 0.70 ± 0.04b 68 ± 6.5f 3.0 ± 0.2e
Year
2008 19.9 ± 0.4b 14.0 ± 0.3b 290.4 ± 11.2a 14.2 ± 0.5b 1115 ± 65.9b 0.76 ± 0.02a 209 ± 8.0c 1.4 ± 0.0a
2009 18.0 ± 0.3a 11.3 ± 0.3a 292.9 ± 14.4a 11.3 ± 0.5a 672 ± 40.6a 0.89 ± 0.03b 163 ± 10.4b 2.0 ± 0.1b
2010 17.5 ± 0.3a 11.6 ± 0.2a 287.5 ± 12.4a 11.1 ± 0.5a 672 ± 35.7a 0.97 ± 0.03c 106 ± 4.6a 2.7 ± 0.1c
Interaction Variety x Year < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001

Note: Values in columns followed by different letters are significantly different (p < 0.05).
*calculated according to cluster compactness index Cl-12 (TELLO and IBANEZ 2014).
 Cluster architecture of old, neglected Croatian grapevine varieties 179

'Crljenak Viški' was the most sensitive variety and berry ability, and are defined based on the real average values
damage was drastic in 2010 (50 % infected berries). 'Babić' of eight cluster properties. PC1 accounts for 56.66 % of
and 'Lasina' showed very similar sensitivity to 'Crljenak the total variability, while PC2 accounts for 26.75 %. The
Viški' (11 to 25 % infected berries over = three consecutive dispersion of varieties can be observed from their distribu-
years), while 'Babica', 'Plavac Mali Sivi', and 'Svrdlovina' tion in the coordinate system, which indicates the existence
had medium sensitivity (6 to 10 % infected berries). A very of at least three groups among varieties. Group I ('Babić',
small proportion of Botrytis-infected berries (1-5 %) were 'Pošip Bijeli', and 'Grk') had long, heavy (> 370 g), large
found in 'Glavinuša', 'Grk', and 'Vrškajica'. No symptoms volume (>1000 cm3), and medium loose clusters. Group
of Botrytis infection were observed in 'Plava Lovora' in II ('Crljenak Viški', 'Glavinuša', and 'Plava Lovora') had
any year, which can be explained by its very open clusters. medium length, low weight (~ 200 g), medium volume
However, a more detailed understanding of the relationship (700 to 900 cm3), and loose clusters. Group III ('Babica',
between cluster compactness and Botrytis susceptibility 'Plavac Mali Sivi', 'Lasina', 'Vrškajica', and 'Svrdlovina')
would include several other factors related to the morphol- had short to medium length, medium cluster weight (200 to
ogy, anatomy and physiology of berries, such as thickness 350 g), small cluster volume (500 to 700 cm3), and com-
and number of cell layers in the epidermis, number of len- pact clusters. PCA of the cluster data revealed clear differ-
ticels and pores, the thickness of the cuticle and wax, and ences among varieties.
the chemical composition of berry skins (MLIKOTA GABLER
et al. 2003).
Principal Component Analysis (PCA) was used to Conclusion
group similar varieties based on eight cluster-relevant
properties. The Figure shows the distribution of the varie- The results of this study indicate different cluster struc-
ties in the coordinate system defined by two latent factors tures among the old Croatian varieties, with different sensi-
of the PCA. PC1 and PC2 comprise 83.41 % of total vari- tivity to Botrytis. The significant variety by year difference
in the magnitude of all characters expressed very sensitive
A variety reaction to annual variation in climatic condition
that are closely related to the Botrytis attacks. PCA analysis
revealed at least three different groups of varieties which
were favored and selected for traditional wine growing in
Croatia. These phenotypic data on cluster structure repre-
sent a basis for future research, such as QTL analysis or
assessment of productivity of these varieties.

Acknowledgements

The authors thank the Croatian Ministry of Science, Educa-

tion and Sports for financial supports (project ID: 091-0910468-
0279). Joint publication of the COST Action FA1003 "East-West
Collaboration for Grapevine Diversity Exploration and Mobiliza-
tion of Adaptive Traits for Breeding".
B

References

CORREA, J.; MAMANI, M.; MUNOZ-ESPINOZA, C.; LABORIE, D.; MUNOZ, C.;
PINTO, M.; HINRICHSEN, P.; 2014: Heritability and identification of
QTLs and underlying candidate genes associated with the architec-
ture of the grapevine cluster (Vitis vinifera L.). Theor. Appl. Genet.
127, 1143-1162.
MLIKOTA GABLER, F.; SMILANICK, J. L.; MONSOUR, M.; RAMMING, D. W.;
MACKEY, B. E.; 2003: Correlation of Morphological, anatomical,
and chemical features of grape berries with resistance to Botrytis
cinerea. Phytopathology 93, 1263-1273.
MOLITOR, D.; BEHR, M.; HOFFMANN, L.; EVERS, D.; 2012: Impact of grape
cluster division on cluster morphology and bunch rot epidemic. Am.
J. Enol. Vitic. 63, 508-514.
NEGRUL, A. M.; 1938: Evolution of cultivated forms of grapes. Comptes
Figure: Projection of eight cluster characteristics on which the Rendus Acad. Sci. l'URSS 18, 585-588.
definition of latent PC factors 1 and 2 are based. PC factors 1 PALLIOTTI, A.; PONI, S.; BERRIOS, J.G.; BERNIZZONI, F.; 2010: Vine per-
and 2 together comprise 83.41 % of total variability. B: Principal formance and grape composition as affected by early-season source
component analysis (PCA) of eleven varieties on the plane de- limitation induced with anti-transpirants in two red Vitis vinifera L.
fined by the first two principal components. cultivars. Aust. J. Grape Wine Res. 16, 426-433.
 180 G. ZDUNIĆ et al.

PONI, S.; CASALINI, L.; BERNIZZONI, F.; CIVARDI, S.; INTRIERI, C.; 2006: Ef- TELLO, J.; IBÁŇEZ, J.; 2014: Evaluation of indexes for the quantitative
fects of Early Defoliation on Shoot Photosynthesis, Yield Compo- and objective estimation of grapevine bunch compactness. Vitis 53,
nents, and Grape Composition. Am. J. Enol. Vitic. 57, 397-407. 9-16.
SHAVRUKOV, Y. N.; DRY, I. B.; THOMAS, M. R.; 2004: Inflorescence and VAIL, M. E.; MAROIS, J. J.; 1991: Grape cluster architecture and the suscep-
bunch architecture development in Vitis vinifera L. Aust. J. Grape tibility of berries to Botrytis cinerea. Phytopathology 81, 188-191.
Wine Res. 10, 116-124. ZDUNIĆ, G.; PREECE, J. E.; DANGL, G. S.; KOEHMSTEDT, A.; MUCALO, A.;
MALETIĆ, E.; PEJIĆ, I.; 2013: Genetic characterization of grapevine
cultivars collected throughout the Dalmatian region. Am. J. Enol.
Vitic. 64, 285-290.

View publication stats