G Model
ARTICLE IN PRESS
HORTI-6088; No. of Pages 11
Scientia Horticulturae xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Scientia Horticulturae
journal homepage: www.elsevier.com/locate/scihorti
Diversity of morpho-physiological traits in worldwide sweet cherry
cultivars of GeneBank collection using multivariate analysis
Ioannis Ganopoulos a,b,∗,1 , Theodoros Moysiadis a,1 , Aliki Xanthopoulou a,c ,
Maria Ganopoulou d , Evangelia Avramidou b , Filippos A. Aravanopoulos b , Eleni Tani e ,
Panagiotis Madesis a , Athanasios Tsaftaris a,c , Konstantinos Kazantzis f,∗∗
a
Institute of Applied Biosciences, CERTH, Thermi, Thessaloniki 57001, Greece
Forest Genetics & Tree Breeding, Faculty of Forest & Environmental Science, Aristotle University of Thessaloniki, Greece
c
Department of Genetics and Plant Breeding, School of Agriculture, Aristotle University of Thessaloniki, Thessaloniki 54124, Greece
d
Mathematics Department, Aristotle University of Thessaloniki, Thessaloniki 54006, Greece
e
Department of Plant Breeding and Biometry, Agricultural University of Athens, Iera Odos 75, 11855 Athens, Greece
f
Pomology Institute (Hellenic Agricultural Organisation—‘DEMETER’), P.O. Box 122, Naoussa, 59200, Greece
b
a r t i c l e
i n f o
Article history:
Received 23 June 2015
Received in revised form 25 August 2015
Accepted 30 September 2015
Available online xxx
Keywords:
Phenotypic diversity
Prunus avium
Germplasm collection
Hierarchical cluster analysis
a b s t r a c t
This study presents an assessment of morpho-physiological diversity for one hundred and forty-six
sweet cherry (Prunus avium) cultivars, originating from different countries and maintained in an ex
situ GeneBank collection. Data of thirty-five traits, describing phenology, plant morphology, yield and
fruit quality were recorded over three years and analyzed using principal component analysis (PCA). An
unsupervised hierarchical cluster analysis was performed the different cultivars using the Euclidean distance metric and the Ward’s agglomeration method. Significant positive and negative correlations were
detected among the different morpho-physiological traits. The sweet cherry cultivars were classified
into three main clusters, suggesting that the characterized sweet cherry collection has high potential for
specific breeding goals. Correlations among the traits, which will be useful for breeding in fruit size and
quality, are discussed. Sweet cherry cultivars, which were classified in diverse clusters, could be potential
parents for hybridization and new genotypes could be created with a combination of desirable traits that
complement one another. These new genotypes could have a high heterotic behavior and thus could
substantially contribute to existent sweet cherry breeding programs.
© 2015 Elsevier B.V. All rights reserved.
1. Introduction
Cherries (Prunus avium L.) include sweet cherry trees cultivated for human consumption and wild cherry trees, also called
mazzards, cultivated for their wood. Cherries are thought to have
originated in the Caucasus area, whereas at present they are
found across mainland Europe and western Asia (Webster, 1996).
Undoubtedly, cherries were an early food source for primitive
inhabitants of Europe, as pits have been recovered from cave
dwellings that date back to 4000–5000 B.C. Sweet cherries were
probably first cultivated in Greece (Hedrick, 1915; Marshall, 1954).
Today, cherry cultivation is one of the most popular fruit tree crop
∗ Corresponding author at: Institute of Applied Biosciences, CERTH, Thermi, Thessaloniki 570 01, Greece.
∗∗ Corresponding author.
E-mail addresses: giannis.ganopoulos@gmail.com (I. Ganopoulos),
nagrefpi@otenet.gr (K. Kazantzis).
1
These two authors contribute equally to this work.
in Greece (Koukourojiannis, 1996). Commercial demand for sweet
cherry resulted in the contemporary increased agricultural production.
Greek growers, in order to follow current market demands,
have changed their sweet cherry production from local cultivars,
mainly from ‘Van’ and ‘Burlat’ groups (Ganopoulos et al., 2011), to
the non-native sweet cherry cultivars. The replacement of local,
well adapted to the local climate conditions, by commercial sweet
cherry cultivars led to a dramatic loss of genetic diversity in terms
of adaptability, tolerance to diseases and fruit quality. Therefore, it
is of utmost importance the need for preservation of endangered
fruit germplasm and establishment of programs that target the conservation of genetic resources in Greece (Ganopoulos et al., 2011).
Keulemans (1993) underlined the contribution of local cultivars
to variability and better adaptability. The Experimental Station of
Institute of Pomology in Greece, which was established in 1970s,
created the first Greek germplasm collection with local and international sweet cherry cultivars (Kazantzis, 2013).
http://dx.doi.org/10.1016/j.scienta.2015.09.061
0304-4238/© 2015 Elsevier B.V. All rights reserved.
Please cite this article in press as: Ganopoulos, I., et al., Diversity of morpho-physiological traits in worldwide sweet cherry cultivars of
GeneBank collection using multivariate analysis. Sci. Hortic. (2015), http://dx.doi.org/10.1016/j.scienta.2015.09.061
G Model
HORTI-6088; No. of Pages 11
2
ARTICLE IN PRESS
I. Ganopoulos et al. / Scientia Horticulturae xxx (2015) xxx–xxx
One of the most important factors influencing plant breeding
is the existence of genetic diversity. Therefore, the identification
and estimation of the genetic diversity as well as its nature are of
paramount importance for a successful breeding program. Furthermore, it is crucial to have not only the information on the genetic
diversity but also the information on plant germplasm and the
physical presence of the germplasm in order to preserve it, exploit
it and sustainable use it in plant breeding and in agriculture or
even for other uses (Khadivi-Khub et al., 2014). To the best of our
knowledge, this is the first study describing the complete process
of morpho-physiological characterization of a worldwide sweet
cherry germplasm collection. Furthermore, similar research in fruit
crops has been conducted in a worldwide olive GeneBank collection by using morphological markers (Díez et al., 2012; Trujillo et al.,
2014).
In order to identify and analyze the genetic diversity of the
various sweet cherry cultivars used, one can rely solely on the phenotypic traits (IPGRI, 1985; UPOV, 1976). Morphological analysis is
a quick and commonly used method to identify and characterize
the germplasm through phenotyping.
To discover phenotypic traits that mostly contribute to the total
diversity in a germplasm collection and characterize the levels of
similarity/dissimilarity among the cultivars, a characterization of
phenotypic diversity and structure is needed (de Oliveira et al.,
2012; Furones-Pérez and Fernández-López, 2009; Mehmood et al.,
2014). Multivariate data analysis includes powerful statistical techniques for analyzing data with many variables simultaneously to
identify patterns and relationships. Since information obtained
from morphological characterization is derived from a large data
set consisting of qualitative and quantitative traits, the use of multivariate analysis is particularly preferred (de Oliveira et al., 2012;
Furones-Pérez and Fernández-López, 2009; Mehmood et al., 2014).
Furthermore, multivariate analysis has been used for genetic variability estimation. The most suitable multivariate techniques for
morphological characterization of genotypes are principal component analysis (PCA) and cluster analysis (Mohammadi and
Prasanna, 2003; Peeters and Martinelli, 1989). The combination of
these statistical methods could provide comprehensive information of characteristics that crucially contribute to genetic diversity
in plants (Khodadadi et al., 2011).
The aims of this study were (i) to evaluate the phenotypic diversity in 146 worldwide sweet cherry cultivars preserved in a Greek
GeneBank collection, (ii) to identify specific traits, and (iii) to detect
relationships among the studied cultivars.
2. Materials and methods
2.1. Plant material
Morpho-physiological genetic diversity was assessed in 146
worldwide sweet cherry (P. avium L.) cultivars. These cultivars are
part of the ex situ Greek Fruit GeneBank collection in Naoussa
(Table 1) and represent the total diversity of Greek sweet cherry
cultivars and a large part of international cultivars. Thirty-five variables were examined on the basis of the cherry descriptors at the
experimental collection for three consecutive years (2010–2012).
Different horticultural practices including fertilizer, application
spraying and irrigation and other cultural practices were made on
regular intervals each year. At the beginning of the study (2010), the
trees were mature (8 years old) and also healthy, and in cropping
condition.
2.2. Analysis of morpho-physiological traits
Mature leaves were collected, approximately at the end of July.
From each of the four trees studied per cultivar, seven leaves were
sampled per year, and the following parameters were measured
using a digital caliper with a sensitivity of ±0.01 mm. The flowers
were collected at full bloom; ten flowers were taken from each of
the four trees studied per cultivar and year. Fruits were collected at
maturity. Cherry fruits maturity was determined based on the color
characteristics of each cultivar, taking into account information
provided by growers and from personal experience and observation. A sample of a total of 106 cherry fruits was taken from each of
the four trees studied per cultivar and year. One hundred of them
were used to determine the mean fruit weight. The remaining six
cherries were used to study a series of quantitative and qualitative
descriptors.
The 21 quantitative traits evaluated, included Stone volume
[StV], Yield [YL], Ratio volume stone/volume fruit [VSt/VFr], Ratio
weight stone/weight fruit [WeSt/WeFr], Petiole length [PL], Petiole width [PeWi], Ratio petiole length/width [PL/Wi], Blade length
[BlLe], Blade width [BlWi], Ratio blade length/petiole length
[BlLe/PL], Stone length [StLe], Stone width [StWi], Stone thickness
[StTh], Soluble solids [SoSo], Titratable acidity [TiAc], ratio titratable acidity/soluble solids [TiAc/ SoSo], Pedicel length [PeLe], Fruit
polar diameter [PoDi], Fruit equatorial diameter [EqDi], Fruit width
[FrWi] and Fruit weight [FrWe].
The 14 qualitative traits investigated included tree vigour [TrVi],
tree habit [TrHa], tree branching [TrBr], lenticels size [LeSi], lamella
shape [LaSh], number of nectaries [NuNe], petal shape [PeSh], fruit
shape [FrSh], fruit size [FrSi], fruit skin colour [FrSCo], firmness of
flesh [FiFl], stone shape [StSh] and stone size [StSi].
These traits were selected from the International Union for
the Protection of New Cultivars of Plants descriptors proposed for
sweet cherry (UPOV, 1976; IPGRI, 1985). The scoring system is the
same as used by UPOV and IPGRI. When possible, all measurements
of a trait were made on the same date to avoid differences in the
environment or developmental stages of the tree.
2.3. Data scoring and analysis
Data for the 146 sweet cherry cultivars, involving 35 traits,
were analyzed via XLSTAT software (version 2014.1). The Principal
Components Analysis (PCA) was applied separately for the quantitative and qualitative traits. The missing data were estimated by
the mean in the first case and by the mode in the latter. In both
cases the correlation matrix was used, the main reason being that
standardization was necessary since the variables were measured
in different units. Within PCA, factor loadings >0.55 were regarded
as significant since the number of observations was 146 (see also
Mehmood et al., 2014). In each case, 3D plots were constructed,
with regard to the three most important principal components, to
facilitate the visualization of the results. In the correlation analysis, the Pearson coefficient (parametric) was used to measure the
correlation among quantitative characteristics and the Spearman
coefficient (non-parametric) was used to measure the correlation
among qualitative characteristics. The combined data from both
the quantitative and qualitative traits were used for dendrogram
construction. The chosen distance to estimate the genetic dissimilarity component was Euclidean and Ward’s method was used for
the agglomerative hierarchical clustering (AHC).
3. Results
3.1. Descriptive statistics and correlations for the quantitative
variables
The 21 quantitative traits were measured and the descriptive
statistics of minima, maxima, means, standard deviations and coefficients of variation (CV) are shown in Table 2. The results revealed
extensive morphological variability. Some traits displayed high CV.
These included YL (43.97%), StV (30.75%), TiAc/SoSo (30.49%), TiAc
(28.64%) and VSt/VFr (28.12%).
Please cite this article in press as: Ganopoulos, I., et al., Diversity of morpho-physiological traits in worldwide sweet cherry cultivars of
GeneBank collection using multivariate analysis. Sci. Hortic. (2015), http://dx.doi.org/10.1016/j.scienta.2015.09.061
G Model
HORTI-6088; No. of Pages 11
ARTICLE IN PRESS
3
I. Ganopoulos et al. / Scientia Horticulturae xxx (2015) xxx–xxx
Table 1
Origin and the main fruit characteristics sweet cherries cultivars analyzed.
Number (No)
Cultivar
Origin
Fruit skin colour
Flesh firmness
Fruit size
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
2e 48-28
Adriana
Angela
B. Producta Delbard
Bargioni I-137
Bargioni I-37
Bargioni I-38
Bargioni I-62
Bargioni I-63
Bargioni M-47
Belle Magnifique
Bianca Di Verona
Bigarreau Burlat
Bigarreau Burlat S-370
Bigarreau Geant D’ Hedelfingen
Bigarreau Goeur De Pigeon
Bigarreau Marmotte
Bigarreau Moreau
Bigarreau Napoleon
Bigarreau Reverchon
Bigarreau Stark Hardy Giant
Bigarreau Tigre
Bing
Black Russian
Black Tartarian
Blanka Kukleitska
Brooks
Burlat e1
Canada Giant
Chinook
Ciliegio If Roma BB2
Ciliegio If Roma T-57
Compact Stella
Corniola
Cristalina
Cuglyeva Acacra
Della Marka Modenese
Di Mauria
Droganova Zuta
Durona Di Cesena
Durone Di Vignola
Durone Di Vignola II
Early Rivers
Empereur Francis
Fercer (Arcina)
Ferrovia
Ferrovia spur
Germersdorfer
Giorgia
Glorius Stark Gold
Grossa Di Pistoia
Grossa Rossa
Guillaume
Hative De Bale
Hative De Berny
Hebros
Hedelfingen V 18775 × 20
Hudson
Jaboulay
Jubilee
Kordia
Kustendilska Hrustjalka
Lambert Òim
Lapins
Larian
Merton Bigarreau
Monnembegi
Napoleon S-787
Negre Di Bistrita
Nera Di Piemonde
Nero II clone 52 P3
Nero II clone 78 P2
Nero II e1 8
New Star 26-3-7.
Canada
Italy
Canada
France
Italy
Italy
Italy
Italy
Italy
Italy
France
Italy
France
France
Germany
France
France
France
Germany
Italy
USA
France
USA
USA
Europe
Bulgary
USA
Italy
Canada
USA
Italy
Italy
New Ziland
Italy
Canada
Bulgary
Italy
unknown
Bulgary
Italy
Italy
Italy
United Kingdom
Germany
France
Italy
Italy
Hungary
Italy
USA
Italy
Italy
France
Switcherland
France
Bulgary
Germany
USA
France
USA
Czech Republic
Bulgary
USA
Canada
USA
United Kingdom
Romania
France
Romania
Romania
Italy
Italy
Italy
Canada
Dark red
Dark red
Dark red
Orange red
Orange red
Dark red
Dark red
Dark red
Dark red
Dark red
Light red
Orange red
Dark red
Dark red
Darkish
Darkish
Dark red
Dark red
Orange red
Dark red
Dark red
Darkish
Darkish
Darkish
Darkish
Dark red
Dark red
Dark red
Dark red
Red
Brown red
Red
Brown red
Dark red
Dark red
Dark red
Orange red
Yellow
Yellow
Dark red
Darkish
Darkish
Dark red
Orange red
Dark red
Dark red
Dark red
Light red
Dark red
Yellow
Darkish
Orange red
Dark red
Darkish
Darkish
Dark red
Darkish
Light red
Dark red
Light red
Dark red
Dark red
Darkish
Red
Brown red
Light red
Dark red
Orange red
Brown red
Darkish
Dark red
Dark red
Dark red
Dark red
Strong
Strong
Weak
Medium
Medium
Strong
Strong
Strong
Strong
Strong
Strong
Strong
Medium
Medium
Medium
Strong
Strong
Strong
Strong
Strong
Strong
Strong
Strong
Medium
Weak
Strong
Medium
Strong
Strong
Strong
Medium
Strong
Strong
Strong
Strong
Strong
Weak
Weak
Strong
Strong
Medium
Strong
Weak
Strong
Strong
Strong
Strong
Strong
Strong
Weak
Strong
Strong
Strong
Weak
Weak
Strong
Medium
Strong
Strong
Medium
Strong
Medium
Medium
Strong
Strong
Strong
Strong
Strong
Strong
Weak
Strong
Strong
Strong
Strong
Very small
Large
Small
Very large
Very large
Small
Medium
Small
Very large
Medium
Very large
Large
Very large
Very large
Medium
Medium
Large
Medium
Small
Large
Very large
Large
Very large
Medium
Small
Medium
Large
Very large
Very large
Very large
Very large
Small
Small
Very small
Medium
Large
Small
Small
Very large
Large
Very large
Large
Very small
Very large
Very large
Very large
Very large
Very large
Very large
Small
Large
Large
Large
Very small
Very small
Very large
Large
Small
Small
Large
Large
Small
Very large
Very large
Very large
Very small
Very small
Small
Very small
Small
Very large
Medium
Large
Large
Please cite this article in press as: Ganopoulos, I., et al., Diversity of morpho-physiological traits in worldwide sweet cherry cultivars of
GeneBank collection using multivariate analysis. Sci. Hortic. (2015), http://dx.doi.org/10.1016/j.scienta.2015.09.061
G Model
HORTI-6088; No. of Pages 11
4
ARTICLE IN PRESS
I. Ganopoulos et al. / Scientia Horticulturae xxx (2015) xxx–xxx
Table 1 (Continued)
Number (No)
Cultivar
Origin
Fruit skin colour
Flesh firmness
Fruit size
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
New York 1143 (NY 1143)
New York T-27
Noble
Northstar
Pobeda
Precoce Bernard
Precoce Della Marca
Primavera
Prime Giant (Giant Red)
R3 Daritska Beluide
Rainier
Rana Cherne Edra
Regina
Rosii Di Bistrita
Sam
Schmidt’s Bigarreau
Seneca
Solymári Gömbölyü
Staccato (Splendid)
Stark Gold Bigarreau
Starkrimson
Stella
Sue
Summit
Sunburst
Ulster
V-1927
Valera
Valerij Tschkalov
Van
Vega
Verdel Ferbolus
Victor
Vittoria
Vogue
Windsor
Ziraat
Agiorgitika Lilantiou
Athinaika
Arkadias
Basiliadi
I.O.P. 1
I.O.P. 2
I.O.P. 3
Karamela Lilantiou
Kapsiotika
Kifisias
Kifisias Proimotero
Kokkina Anastasias
Koromilokeraso Vitalou
Lemonidi
Mavra Anastasias
Mavro Proimo Achaias
Mavro Proimo Vitalou
Mavro Tripoleos
Mesoproimo Tragano Evias
Moschato Tragano Opsimo Evias
Bakirtzeika
Napoleon Karamela
Opsimi Karamela Tripoleos
Opsimo Tragano Komotinis
Petrokeraso Tragano Achaias
Proimo Kolindrou
Proimo Tragano Komotinis
Samou
Tourkika
Tragana Edessis
Tragana Edessis-Naoussis
Tragana Edessis-Sarakinon
Tragano Komotinis
Fraoula Volou
Chalkidos Unknown
USA
USA
United Kingdom
Canada
Bulgary
France
Italy
Germany
USA
Bulgary
USA
Bulgary
Germany
Romania
Canada
Europe
USA
Hungary
Canada
USA
USA
Canada
Canada
Canada
Canada
USA
unknown
Canada
Hungary
Canada
Canada
France
Canada
Italy
Canada
Canada
Turkey
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Greece
Orange red
Orange red
Dark red
Dark red
Dark red
Brown red
Red
Dark red
Dark red
Orange red
Orange red
Orange red
Dark red
Dark red
Dark red
Darkish
Dark red
Dark red
Red
Yellow
Dark red
Light red
Orange red
Light red
Light red
Dark red
Dark red
Brown red
Dark red
Dark red
Orange red
Red
Orange red
Dark red
Dark red
Darkish
Dark red
Dark red
Dark red
Darkish
Dark red
Dark red
Dark red
Dark red
Orange red
Orange red
Dark red
Orange red
Dark red
Dark red
Dark red
Dark red
Dark red
Dark red
Dark red
Dark red
Orange red
Light red
Orange red
Orange red
Dark red
Orange red
Dark red
Dark red
Light red
Dark red
Dark red
Dark red
Dark red
Dark red
Yellow with blush
Dark red
Strong
Strong
Strong
Medium
Medium
Medium
Weak
Strong
Strong
Weak
Strong
Medium
Strong
Weak
Medium
Medium
Weak
Strong
Strong
Weak
Medium
Medium
Medium
Medium
Medium
Strong
Strong
Medium
Strong
Strong
Strong
Strong
Weak
Strong
Strong
Medium
Medium
Weak
Medium
Strong
Medium
Strong
Strong
Strong
Strong
Weak
Weak
Medium
Strong
Strong
Medium
Strong
Strong
Medium
Medium
Strong
Weak
Strong
Medium
Medium
Strong
Strong
Weak
Strong
Strong
Strong
Strong
Strong
Strong
Medium
Medium
Strong
Large
Medium
Very large
Very large
Small
Small
Small
Very small
Very large
Very small
Very large
Small
Large
Large
Large
Large
Very small
Very small
Medium
Small
Very large
Very large
Very large
Very large
Very large
Medium
Medium
Small
Very large
Medium
Very large
Large
Large
Medium
Large
Medium
Very large
Very small
Small
Medium
Very large
Very small
Very small
Very small
Small
Small
Very small
Large
Medium
Small
Very large
Large
Very large
Small
Medium
Very small
Very small
Very large
Small
Very small
Small
Large
Very small
Very small
Medium
Very large
Large
Very large
Very large
Very small
Large
Very large
Please cite this article in press as: Ganopoulos, I., et al., Diversity of morpho-physiological traits in worldwide sweet cherry cultivars of
GeneBank collection using multivariate analysis. Sci. Hortic. (2015), http://dx.doi.org/10.1016/j.scienta.2015.09.061
G Model
ARTICLE IN PRESS
HORTI-6088; No. of Pages 11
5
I. Ganopoulos et al. / Scientia Horticulturae xxx (2015) xxx–xxx
Table 2
Descriptive statistics for 21 quantitative traits in 146 sweet cherry cultivars.
Variable
Minimum
Maximum
Mean
Std. deviation
CV (%)
Stone volume
Yield
Ratio volume stone/volume fruit
Ratio weight stone/weight fruit
Petiole length
Petiole width
Ratio petiole length/width
Blade length
Blade width
Ratio blade length/petiole length
Stone length
Stone width
Stone thickness
Soluble solids
Titratable acidity
Ratio titratable acidity/soluble solids
Pedicel length
Fruit polar diameter
Fruit equatorial diameter
Fruit width
Fruit weight
0.190
25.000
7.700
1.700
9.900
4.540
0.400
2.720
0.110
2.060
8.400
7.000
5.400
9.600
3.000
0.900
2.700
11.500
18.100
15.500
3.100
0.750
190.000
35.300
26.600
16.000
8.080
0.600
6.220
0.220
5.250
18.000
15.700
11.700
22.300
17.400
7.100
7.300
26.600
28.500
24.000
12.100
0.439
71.096
16.931
14.714
12.437
6.058
0.487
4.195
0.158
3.102
11.120
8.936
7.102
16.644
8.539
2.112
4.515
21.387
23.586
19.838
6.896
0.135
31.264
4.762
3.320
1.266
0.583
0.040
0.562
0.019
0.593
1.099
0.896
0.720
2.470
2.446
0.644
0.719
1.947
2.016
1.486
1.502
30.75
43.97
28.12
22.56
10.17
9.62
8.21
13.39
12.02
19.11
9.88
10.02
10.13
14.84
28.64
30.49
15.92
9.11
8.54
7.49
21.78
Strong positive, linear correlations were observed among all
the 21 quantitative traits (Table 3). The highest significant, positive correlation was between FrWe and EqDi (0.858). Significant
positive correlations were also observed between EqDi and FrWi
(0.844), PL and PeWi (0.709), StLe and StWi (0.686), VSt/VFr and
WeSt/WeFr (0.646), PeWi and BlWi (0.565), EqDi and WeSt/WeFr
(0.541), FrWe and WeSt/WeFr (0.495), and FrWi and WeSt/WeFr
(0.458). On the other hand, there were also high, significant, negative correlations between some quantitative traits (Table 3). These
included for instance, the TiAc/ SoSo and TiAc (-0.714) and VSt/VFr
and StVo (−0.712) (Table 3). We reported significant correlations
between traits contributing to fruit yield and quality, which is
helpful for plant breeding. For example, to attain high yield and
superior quality cultivars, cross combinations could be performed
between cultivars with large fruit size (‘Lemonidi’, ‘Mpakirtzeika’,
‘Starkrismson’ and ‘Ziraat’), low stone volume (‘Brooks’, ‘Hative
De Berny’, ‘Hative De Bale’, ‘Seneca’, ‘Merton Bigarreau’ and ‘Monnembegi’) and high total soluble solids (‘2e 48-28′ , ‘Monnembegi’,
‘Napoleon S-787′ , ‘R3 Daritska Beluide’, ‘Rainier’, ‘Rosii Di Bistrita’,
‘Sue’, ‘Arkadias’, ‘Kapsiotika’ and ‘Tragana Edessis-Naoussis’).
3.2. Principal component analysis of quantitative variables
The distribution of cultivars, based on the PC-1, PC-2 and PC3, shows the phenotypic variation among the cultivars and how
widely dispersed they are along both axes (Fig. 1). Using Kaiser’s
criterion (“Eigenvalue” >1) (Kaiser, 1958), we reduce the dimension
implied by the 21 quantitative traits to six significant components
that explained 73.59% of the total variation (Table 4; Fig. 1). The
first component, which accounted for 22.90% of the total variation, included stone volume, petiole length, blade width, fruit
polar diameter, fruit equatorial diameter, fruit width and fruit
weight. The second component, which explained 16.75% of the total
variation, was mainly correlated to characters related to ratio of
volume stone/volume fruit, ratio weight stone/weight fruit, stone
length, stone width and stone thickness. The third component that
explained 10.15% of the total variation included petiole length, petiole width and stone length. The fourth component, accounted for
9.55% of the total variation, was determined by titratable acidity
and ratio of titratable acidity/soluble solids. The fifth component
explained 7.71% of the total variation, and included blade length
Fig. 1. ‘Scree plot’ of eigenvalues obtained from the PCA process for (A) quantitative and (B) qualitative traits.
Please cite this article in press as: Ganopoulos, I., et al., Diversity of morpho-physiological traits in worldwide sweet cherry cultivars of
GeneBank collection using multivariate analysis. Sci. Hortic. (2015), http://dx.doi.org/10.1016/j.scienta.2015.09.061
G Model
HORTI-6088; No. of Pages 11
6
StV
YL
VSt/VFr
StV
1
YL
0.199 1
−0.712 0.008 1
VSt/VFr
WeSt/WeFr −0.227 −0.027 0.646
PeLe
0.130 0.115 0.050
PeWi
0.158 0.095 −0.028
PL/Wi
0.059 −0.041 −0.118
BlLe
−0.225 −0.244 0.200
BlWi
0.262 0.277 −0.025
0.085 0.282 −0.055
BlLe/ PeLe
0.353 0.091 −0.280
StLe
0.427 0.142 −0.284
StWi
0.352 0.214 −0.243
StTh
SoSo
0.024 0.049 −0.013
0.282 −0.034 −0.186
TiAc
TiAc/ SoSo −0.271 0.150 0.190
PeLe
0.212 0.106 −0.216
0.206 0.081 0.302
PoDi
0.389 0.070 0.231
EqDi
FrWi
0.460 0.245 0.169
FrWe
0.334 0.142 0.330
WeSt/WeFr
1
0.100
0.012
−0.094
0.155
0.065
−0.131
−0.471
−0.434
−0.487
0.066
0.096
−0.048
−0.272
0.457
0.541
0.458
0.495
PeLe
PeWi
1
0.709 1
−0.389 0.316
0.167 0.077
0.671 0.565
0.431 0.333
0.068 0.087
0.126 0.219
0.181 0.223
0.007 0.058
0.134 0.150
−0.104 −0.142
−0.027 0.058
0.287 0.195
0.267 0.221
0.310 0.258
0.328 0.257
PL/Wi
BlLe
BlWi
1
−0.118
−0.171
−0.259
0.063
0.123
0.054
0.060
0.020
−0.052
0.096
-0.114
-0.025
-0.038
-0.070
1
0.141 1
−0.614 0.277
−0.156 −0.005
−0.176 0.182
−0.216 0.228
−0.003 0.000
−0.042 0.123
0.050 −0.123
0.150 0.096
−0.038 0.216
0.010 0.312
−0.050 0.360
0.041 0.330
BlLe/PeLe
StLe
StWi
StTh
SoSo
TiAc
TiAc/SoSo
PeLe
PoDi
EqDi
FrWi
FrWe
1
0.032
0.094
0.263
0.027
0.088
−0.067
−0.059
0.050
−0.030
0.085
0.020
1
0.686
0.542
−0.123
−0.080
0.049
0.060
0.335
0.118
0.129
0.161
1
0.790
−0.078
−0.002
−0.053
0.091
0.192
0.298
0.320
0.259
1
−0.108
0.016
−0.099
0.126
0.103
0.191
0.313
0.208
1
0.460
0.135
0.214
90.021
−0.005
0.090
0.021
1
−0.714
0.062
0.004
0.137
0.256
0.142
1
0.058
0.034
−0.133
−0.210
−0.109
1
−0.172
−0.104
−0.025
−0.046
1
0.595
0.660
0.701
1
0.844
0.858
1
0.843
1
Values in bold are different from 0 with a significance level alpha = 0.05. Trait abbreviation: Stone volume [StV], Yield [YL], Ratio volume stone/volume fruit [VSt/VFr], Ratio weight stone/weight fruit [WeSt/WeFr], Petiole length
[PL], Petiole width [PeWi], Ratio petiole length/width [PL/Wi], Blade length [BlLe], Blade width [BlWi], Ratio blade length/petiole length [BlLe/ PL], Stone length [StLe], Stone width [StWi], Stone thickness [StTh], Soluble solids
[SoSo], Titratable acidity [TiAc], ratio titratable acidity/soluble solids [TiAc/SoSo], Pedicel length [PeLe], Fruit polar diameter [PoDi], Fruit equatorial diameter [EqDi], Fruit width [FrWi] and Fruit weight [FrWe].
ARTICLE IN PRESS
Variables
I. Ganopoulos et al. / Scientia Horticulturae xxx (2015) xxx–xxx
Please cite this article in press as: Ganopoulos, I., et al., Diversity of morpho-physiological traits in worldwide sweet cherry cultivars of
GeneBank collection using multivariate analysis. Sci. Hortic. (2015), http://dx.doi.org/10.1016/j.scienta.2015.09.061
Table 3
Correlation coefficients (Pearson) among 21 quantitative traits in 146 sweet cherry cultivars.
G Model
ARTICLE IN PRESS
HORTI-6088; No. of Pages 11
7
I. Ganopoulos et al. / Scientia Horticulturae xxx (2015) xxx–xxx
Table 4
First 6 components from the PCA analysis of 21 quantitative traits in 146 sweet cherry cultivars.
Traits
F1
F2
F3
F4
F5
F6
StV
YL
VSt/Vfr
WeSt/WeFr
PL
PeWi
PL/Wi
BlLe
BlWi
BlLe/ PeLe
StLe
StWi
StTh
SoSo
TiAc
TiAc/ SoSo
PeLe
PoDi
EqDi
FrWi
FrWe
Cumulative%
Variability (%)
0257
0.132
0.005
0.102
0.266
0.240
−0.033
−0.050
0.270
0.133
0.165
0.242
0.228
0.013
0.125
−0.114
0.000
0.305
0.360
0.394
0.373
22.906
22.906
−0.256
−0.078
0.414
0.483
0.043
−0.041
−0.105
0.189
0.003
−0.125
−0.300
−0.320
−0.328
0.031
−0.015
0.043
−0.161
0.185
0.189
0.136
0.200
39.661
16.755
0.014
0.016
−0.128
−0.004
0.340
0.300
−0.102
0.038
0.298
0.275
−0.318
−0.243
−0.144
0.225
0.395
−0.296
0.100
−0.232
−0.167
−0.096
−0.158
49.815
10.155
0.266
−0.214
−0.202
0.086
−0.329
−0.167
0.266
0.059
−0.225
−0.357
−0.043
−0.014
−0.096
0.200
0.438
−0.377
0.101
−0.029
0.145
0.145
0.083
59.374
9.558
−0.037
−0.196
−0.019
−0.112
0.188
0.261
0.118
0.653
0.225
−0.440
0.079
0.098
0.029
−0.022
−0.140
0.125
0.314
−0.052
−0.007
−0.069
0.008
67.083
7.710
0.043
0.431
0.054
0.037
−0.146
−0.007
0.178
−0.093
0.007
0.021
−0.091
−0.058
−0.054
0.567
−0.029
0.445
0.447
−0.011
0.001
0.094
0.050
73.590
6.506
Trait abbreviation: Stone volume [StV], Yield [YL], Ratio volume stone/volume fruit [VSt/VFr], Ratio weight stone/weight fruit [WeSt/WeFr], Petiole length [PL], Petiole width
[PeWi], Ratio petiole length/width [PL/Wi], Blade length [BlLe], Blade width [BlWi], Ratio blade length/petiole length [BlLe/ PL], Stone length [StLe], Stone width [StWi], Stone
thickness [StTh], Soluble solids [SoSo], Titratable acidity [TiAc], ratio titratable acidity/soluble solids [TiAc/ SoSo], Pedicel length [PeLe], Fruit polar diameter [PoDi], Fruit
equatorial diameter [EqDi], Fruit width [FrWi] and Fruit weight [FrWe].
and ratio of blade length/petiole length. The sixth component
explained 6.50% of the total variation, included yield, soluble solids,
pedicel length and ratio of titratable acidity/soluble solids. Furthermore, a PCA scatter plot was constructed based on the first three
components (Fig. 2). The plot grouped the cultivars according to
their phenotypic similarity and morphological traits. For instance,
cultivars Bakirtzeika and Lemonidi with the highest fruit weight,
largest fruit width and longest fruit length were placed closely in
the upper right plane. These results demonstrate that fruit equatorial diameter, fruit weight and fruit width are highly positively
correlated and as a result, these morphological traits led to the
highest loading factors in this PCA analysis. Thus, PCA showed that
some traits had the highest loadings in the first two components.
These traits included stone volume, petiole length, blade width,
fruit polar diameter, fruit equatorial diameter, and fruit width and
fruit weight. These results indicate that such traits are suitable both
for the assessment of genetic diversity and for phenotypic characterization of sweet cherry germplasm (Fig. 3).
3.3. Correlations for qualitative traits
There were more significant negative correlations than positive
ones among the 14 qualitative traits (Table 5). Negative correlations were observed between the following traits: FiFl and TrBr
(−0.297), StSi and TrBr (−0.293), FrSh and FiFl (−0.269), TrVi and
FrSi (−0.262), FrSh and FrSi (−0.244), TrBr and TrVi (−0.175), and
LaSh and TrVi (−0.174). Positive correlations included TrBr and
NuNe (0.364), FrSi and StSi (0.342), FrSi and LaSi (0.283) LaSh and
NuNe (0.280) and FiFl and FrSi (0.229).
3.4. Principal component analysis (PCA) of qualitative variables
Using Kaiser’s criterion (“Eigenvalue” >1) (Kaiser, 1958), we
obtained 6 significant components, which explained 65.11% of the
total variation (Fig. 1). The first component, which accounted for
17.24% of the total variation, included tree branching, firmness
of flesh, fruit size, number of nectaries and lamella size. The sec-
Fig. 2. (A) Three-dimensional PCA plot of the 146 sweet cherry cultivars with regard to the first three principal components. (B) Three-dimensional PCA plot of the 21
quantitative traits with regard to the first three principal components. Variability explained: F1 (22.91%), F2 (16.76%), F3 (10.16%).
Please cite this article in press as: Ganopoulos, I., et al., Diversity of morpho-physiological traits in worldwide sweet cherry cultivars of
GeneBank collection using multivariate analysis. Sci. Hortic. (2015), http://dx.doi.org/10.1016/j.scienta.2015.09.061
G Model
ARTICLE IN PRESS
HORTI-6088; No. of Pages 11
8
I. Ganopoulos et al. / Scientia Horticulturae xxx (2015) xxx–xxx
Fig. 3. (A) Three-dimensional PCA plot of the 146 sweet cherry cultivars with regard to the first three principal components. (B) Three-dimensional PCA plot of the 14
qualitative traits with regard to the first three principal components. Variability explained: F1 (17.25%), F2 (10.95%), F3 (10.43%).
Table 5
Correlation coefficients (Spearman) among 14 qualitative traits in 146 sweet cherry cultivars.
Variables
TrVi
TrHa
TrVi
TrHa
TrBr
LeSi
LaSh
LaSi
NuNe
PeSh
FrSh
FrSi
FrSCo
FiFl
StSh
StSi
1
0.108
−0.175
−0.042
0.012
−0.174
−0.055
0.076
0.017
−0.262
0.007
−0.028
−0.114
0.037
1
−0.115
−0.172
−0.071
−0.141
−0.025
−0.151
−0.009
−0.051
0.099
−0.065
−0.073
0.035
TrBr
1
−0.262
0.005
−0.112
0.364
−0.021
0.251
−0.138
0.127
−0.297
−0.105
−0.293
LeSi
1
0.107
−0.015
−0.067
−0.086
−0.016
0.115
−0.046
0.135
0.160
0.060
LaSh
LaSi
NuNe
1
−0.160
0.280
0.128
0.043
−0.145
0.097
−0.075
−0.128
−0.168
1
−0.119
−0.029
−0.161
0.283
−0.031
0.181
0.069
0.279
1
0.008
0.217
−0.075
0.208
−0.259
−0.175
0.051
PeSh
FrSh
FrSi
FrSCo
FiFl
StSh
StSi
1
−0.087
−0.105
0.122
0.115
−0.059
−0.059
1
−0.244
−0.166
−0.269
−0.237
−0.084
1
−0.014
0.229
−0.067
0.342
1
0.069
0.158
−0.062
1
0.138
0.152
1
−0.143
1
Values in bold are different from 0 with a significance level alpha = 0.05. Trait abbreviation: Tree vigour [TrVi], tree habit [TrHa], tree branching [TrBr], lenticels size [LeSi],
lamella shape [LaSh], number of nectaries [NuNe], petal shape [PeSh], fruit shape [FrSh], fruit size [FrSi], fruit skin colour [FrSCo], firmness of flesh [FiFl], stone shape [StSh]
and stone size [StSi].
ond component, which explained 10.94% of the total variation, was
determined by fruit size, stone size and stone shape. The third component, explained 10.43% of the total variation, and was mainly
correlated to characters related to tree vigour, fruit skin colour and
tree habit. The fourth component, accounted for 9.31% of the total
variation included lamella shape and stone shape. The fifth component, explained 9.14% of the total variation, included tree habit,
lenticels size and fruit skin colour. The sixth component, explained
8.03% of the total variation, and included petal shape, tree habit and
lenticels size (Table 6).
Table 6
First 7 components from the PCA analysis of 14 qualitative traits in 146 sweet cherry cultivars.
F1
F2
F3
F4
F5
F6
F7
TrVi
TrHa
TrBr
LeSi
LaSh
LaSi
NuNe
PeSh
FrSh
FrSi
FrSCo
FiFl
StSh
StSi
0.077
0.043
0.394
−0.184
0.218
−0.329
0.341
0.032
0.345
−0.365
0.057
−0.389
−0.179
−0.301
−0.329
−0.057
0.252
−0.128
−0.164
0.302
0.272
−0.275
0.243
0.387
−0.204
−0.184
−0.352
0.363
−0.447
−0.385
0.306
0.094
0.224
0.165
0.267
0.207
-0.204
0.
0.406
0.121
0.285
−0.182
−0.290
−0.010
0.223
−0.045
−0.431
0.017
−0.327
−0.392
0.073
−0.117
−0.195
−0.
0.440
-0.380
−0.063
−0.484
−0.155
0.586
0.287
−0.076
−0.046
−0.045
0.287
−0.049
−0.460
−0.019
−0.033
−0.066
0.076
−0.411
0.147
−0.415
−0.202
0.224
−0.260
0.559
0.034
−0.122
−0.290
0.103
−0.197
−0.099
−0.458
0.316
−0.052
−0.151
0.349
−0.254
−0.244
0.056
−0.193
0.274
−0.246
0.119
−0.328
−0.348
Cumulative%
Variability (%)
17.249
17.249
28.196
10.947
38.627
10.431
47.937
9/310
57.086
9.149
65.117
8.031
71.329
6.212
Trait abbreviation: Tree vigour [TrVi], tree habit [TrHa], tree branching [TrBr], lenticels size [LeSi], lamella shape [LaSh], number of nectaries [NuNe], petal shape [PeSh], fruit
shape [FrSh], fruit size [FrSi], fruit skin colour [FrSCo], firmness of flesh [FiFl], stone shape [StSh] and stone size [StSi].
Please cite this article in press as: Ganopoulos, I., et al., Diversity of morpho-physiological traits in worldwide sweet cherry cultivars of
GeneBank collection using multivariate analysis. Sci. Hortic. (2015), http://dx.doi.org/10.1016/j.scienta.2015.09.061
G Model
HORTI-6088; No. of Pages 11
ARTICLE IN PRESS
I. Ganopoulos et al. / Scientia Horticulturae xxx (2015) xxx–xxx
9
Fig. 4. Dendrogram using agglomerative hierarchical clustering (AHC) for 146 sweet cherry cultivars based on 21 quantitative and 14 qualitative traits.
3.5. Dendrogram using agglomerative hierarchical clustering
(AHC)
Unsupervised agglomerative hierarchical cluster analysis was
used in order to divide the available data up into groups of increasing dissimilarity. The Euclidean distance was used as a metric to
measure the genetic dissimilarity of the 146 sweet cherry cultivars,
based on the combined quantitative and qualitative data, and the
Ward’s method was used for the agglomeration. The dendrogram
of Fig. 4 pointed out that, all sweet cherry cultivars were different
from each other, and several clusters are identified. The dendrogram revealed three distinct groups. C1 contained 75 cultivars, C2
had 5 cultivars, and finally C3 included 66 cultivars. Among all 146
cultivars from different world regions, there were no specific clusters based on locality. The highest genetic distance exists between
C2 and C3 (115.60), followed by C1 and C2 (71.14) and C1 and C3
(46.06).
4. Discussion
The main goal of germplasm management is to collect and to
characterize diverse species or forms of the same species at national
and regional level. To do so it is important to perform, as a first step,
the evaluation of morphological and agronomic traits of interest.
Plant breeders then routinely use this morphological characterization for the initial description and classification of the germplasm
under consideration.
We estimated the morpho-physiological traits variation in
sweet cherry cultivars established in the Greek GeneBank and
provided basic knowledge on the range of variation of several
morphological traits. The studied cultivars displayed significant
variation in fruit weight traits as well as moderate variation for
yield, stone volume, ratio of titratable acidity/soluble solids, titrat-
able acidity, and ratio of volume stone/volume fruit (Pommer,
2012).
Herein, the largest variation among the fruit traits corresponded
to fruit weight and yield as it has been previously reported
(Christensen, 1974; Ganopoulos et al., 2011; Hjalmarsson and Ortiz,
2000; Petruccelli et al., 2013). In addition, these authors have
also suggested that the fruit weight could represent the utmost
significant trait for distinguishing sweet cherry cultivars. Furthermore, the fruit weight (fruit size) is a very crucial trait due to its
economic importance: some local cultivars, e.g., ‘Lemonidi’ and
‘Mpakirtzeika’, have shown relative large fruits with considerably
elevated weights and volumes. The morphological traits that were
observed as crucial for sweet cherry cultivars characterization in
this study were fruit weight and fruit size which were also found to
be important in sweet cherry from Greece (Ganopoulos et al., 2011;
Petruccelli et al., 2013), which indicates similar diversity patterns
in the Mediterranean region.
The fruit weight showed positive correlation with stone weight
(r = 0.495, p < 0.001) and also fruit equatorial diameter with fruit
width (r = 0.844, p < 0.001) in agreement with (Khadivi-Khub,
2014). Our results also showed a very close correlation among
fruit weight and fruit dimensions (length, width and diameter;
r = 0.82, r = 0.90 and r = 0.86, respectively); therefore, these parameters could potentially be used to predict each other. A degree
of correlation between fruit weight and fruit dimensions was
also detected earlier in cherries (Khadivi-Khub, 2014; Rakonjac
et al., 2010). All variables related to fruit size (fruit weight and
dimensions) showed significant positive correlation with leaf traits
(petiole length and petiole width) at p = 0.05, indicating a role of
leaf in the increased fruit size (Demirsoy and Demirsoy, 2004). A
positive correlation between the sweet cherry fruit size and leaf
traits was also determined before by Rakonjac et al. (2010). Posi-
Please cite this article in press as: Ganopoulos, I., et al., Diversity of morpho-physiological traits in worldwide sweet cherry cultivars of
GeneBank collection using multivariate analysis. Sci. Hortic. (2015), http://dx.doi.org/10.1016/j.scienta.2015.09.061
G Model
HORTI-6088; No. of Pages 11
10
ARTICLE IN PRESS
I. Ganopoulos et al. / Scientia Horticulturae xxx (2015) xxx–xxx
tive correlations were also obtained between leaf traits with each
other.
Most of the qualitative traits evaluated are considered crucial
for species registration and discrimination in the test guidelines
proposed by UPOV (UPOV, 1976). Notably, the fruit characteristics, such as skin color, flesh colour and firmness, are those that
typically differentiate sweet cherry cultivars (Antonius et al., 2012;
Hjalmarsson and Ortiz, 2000; Petruccelli et al., 2013). The skin color
of the fruit is a very important quality characteristic, which also
helps the assessment of the stage of fruit maturity as it has been
shown when (Esti et al., 2002) evaluated the changes in color of
sweet cherry and used this trait for monitoring pigment evolution.
Moreover, fruit color has a significant impact on consumer perception of fruit quality, especially as regards the attractiveness of fruit
(Ruiz and Egea, 2008). Consumers generally seem to prefer dark red
cherries (Petruccelli et al., 2013). Another important characteristic
is fruit firmness which is relevant to an assessment of the quality
of fruit, fruit shelf life, and consumer acceptance. Fruit firmness is
a combination of skin and flesh strength, and in general cultivars
with the firmest fruit are preferred by consumers increasing as a
result their value in the market (Kappel et al., 1996; Petruccelli et al.,
2013).
Multivariate statistical methods such as PCA and cluster analysis
could be valuable tools for monitoring cultivars, and characterizing
and classifying plant germplasm in GenBank collections (Iezzoni
and Pritts, 1991). Further, PCA is useful in defining the number of
main factors, thus decreasing the number of efficient parameters to
differentiate genotypes. Additionally, associations between characteristics emphasized by this method may correspond to genetic
linkage between loci controlling traits with a pleiotropic effect
(Iezzoni and Pritts, 1991; Rakonjac et al., 2010). Many studies,
in agreement with our study, indicated that fruit and leaf traits
are crucial factors in phenotyping and morphologically characterizing the diversity in sweet cherry breeding materials (Antonius
et al., 2012; Ganopoulos et al., 2011; Hjalmarsson and Ortiz, 2000;
Lacis et al., 2009; Rakonjac et al., 2010). Furthermore, PCA analysis
has been widely used for evaluation of sweet cherry germplasm
(Beyer et al., 2002; Christensen, 1974; Ganopoulos et al., 2011;
Hjalmarsson and Ortiz, 2000; Khadivi-Khub, 2014; Lacis et al., 2009;
Petruccelli et al., 2013; Rodrigues et al., 2008; Sánchez et al., 2008).
We have used PCA for the identification of the most significant variables in the data set presented herein. For each factor, a principal
component loading of more than 0.55 was considered significant,
which indicated that seven components explained 73.59% of the
total variance. The first three components, consisted of 21 quantitative variables, explained 49.81% of the total variability observed
(Table 4), indicating that these attributes have the highest variation
between the cultivars and had the greatest impact on separation of
the cultivars (Iezzoni, 2008; Khadivi-Khub, 2014; Lacis et al., 2009).
PCA analysis of the qualitative variables has shown similar results
(Table 5). Further, PCA analysis achieved the characterization of
cultivars groups by identifying highly discriminating variables like
traits regarding the fruit. Thus, PCA analysis suggested that future
evaluations can rely on a reduced number of traits with a minimum
loss of information for the discrimination and characterization of
the different varieties/cultivars resulted in a reduction in labour,
time and cost.
Fig. 4 depicts the unsupervised hierarchical cluster analysis of
146 sweet cherry cultivars. The cultivars analyzed grouped into
three main clusters. (Khadivi-Khub, 2014) characterized 41 sweet
cherry genotypes with morphological descriptors and found that
all genotypes were divided into five clusters. The dendrogram
produced showed high diversity between sweet cherry cultivars
indicating that analyzed germplasm collection could be assumed
in breeding programs as a good gene pool for contrasting traits.
For instance, sweet cherry cultivars ‘Cuglyeva Acacra’ from Bulgaria
and ‘Merton Bigarreau’ from U.K. were highly distinguished from
other cultivars.
5. Conclusions
PCA combined with unsupervised cluster analysis for agronomic, morphological and fruit quality traits revealed a wide
diversity in sweet cherry maintained in Germplasm Bank at the
Institute of Pomology (Greece). Understanding the phenotypic
diversity among the sweet cherry cultivars is crucial for the
conservation of traditional genetic material that is endangered. Furthermore, this information is very useful for registration of new
sweet cherry cultivars carried out by EU-Community Plant Variety
Office.
The study presented herein provides useful information on
agronomic, morphological and fruit quality traits of sweet cherry
cultivars grown under the climatic conditions of Imathia area in
North Greece. Future breeding programs could take advantage of
those genetic materials and use them as parental genotypes in order
to achieve genetic recombination and improvement in agronomic,
morphological and fruit quality traits. Thus, new strategies could
be created for new breeding programs of sweet cherry cultivars
with better adaptation to the limiting agro-climatic conditions of
Greece.
When separate traits are analyzed, information of paramount
importance about the internal similarities of cultivars and genetic
structure is lost. Contrariwise, when combined application of PCA
and cluster analysis is used, important traits for sweet cherry cultivars characterization are revealed and provide comprehensive
information. Hence, in order to thoroughly analyze the diversity
of a GeneBank collection, a multivariate statistics approach taking
into account both the genetic and the phenotypic data is of utmost
importance.
Conflict of interest
The authors declare that they have no conflict of interest.
References
Antonius, K., Aaltonen, M., Uosukainen, M., Hurme, T., 2012. Genotypic and
phenotypic diversity in Finnish cultivated sour cherry (Prunus cerasus L.).
Genet. Resour. Crop Evol. 59, 375–388.
Beyer, M., Hahn, R., Peschel, S., Harz, M., Knoche, M., 2002. Analysing fruit shape in
sweet cherry (Prunus avium L.). Sci. Hortic. 96, 139–150.
Christensen, J.V., 1974. Numerical Studies of Qualitative and Morphological
Characteristics of 41 Sweet Cherry Cultivars 2. Tidsskrift for Planteavl.
de Oliveira, E.J., Dias, N.L.P., Dantas, J.L.L., 2012. Selection of morpho-agronomic
descriptors for characterization of papaya cultivars. Euphytica 185, 253–265.
Demirsoy, H., Demirsoy, L., 2004. A study on the relationships between some fruit
characteristics in cherries. Fruits 59, 219–223.
Díez, C.M., Imperato, A., Rallo, L., Barranco, D., Trujillo, I., 2012. Worldwide core
collection of olive cultivars based on simple sequence repeat and
morphological markers. Crop Sci. 52, 211–221.
Esti, M., Cinquanta, L., Sinesio, F., Moneta, E., Di Matteo, M., 2002. Physicochemical
and sensory fruit characteristics of two sweet cherry cultivars after cool
storage. Food Chem. 76, 399–405.
Furones-Pérez, P., Fernández-López, J., 2009. Morphological and phenological
description of 38 sweet chestnut cultivars (Castanea sativa Miller) in a
contemporary collection. Span. J. Agric. Res., 829–843.
Ganopoulos, I.V., Kazantzis, K., Chatzicharisis, I., Karayiannis, I., Tsaftaris, A.S., 2011.
Genetic diversity, structure and fruit trait associations in Greek sweet cherry
cultivars using microsatellite based (SSR/ISSR) and morpho-physiological
markers. Euphytica 181, 237–251.
Hedrick, U.P., 1915. The Cherries of New York. J. B. Lyon, Albany, N.Y.
Hjalmarsson, I., Ortiz, R., 2000. In situ and ex situ assessment of morphological and
fruit variation in Scandinavian sweet cherry. Sci. Hortic. 85, 37–49.
Iezzoni, A.F., 2008. Cherries Temperate Fruit Crop Breeding. Springer, pp. 151–176.
Iezzoni, A.F., Pritts, M.P., 1991. Applications of principal component analysis to
horticultural research. HortScience 26, 334–338.
IPGRI, 1985. Cherry Descriptors. International Plant Genetic Resources Institute,
Rome, Italy, pp. 33.
Kaiser, H.F., 1958. The varimax criterion for analytic rotation in factor analysis.
Psychometrika 23, 187–200.
Please cite this article in press as: Ganopoulos, I., et al., Diversity of morpho-physiological traits in worldwide sweet cherry cultivars of
GeneBank collection using multivariate analysis. Sci. Hortic. (2015), http://dx.doi.org/10.1016/j.scienta.2015.09.061
G Model
HORTI-6088; No. of Pages 11
ARTICLE IN PRESS
I. Ganopoulos et al. / Scientia Horticulturae xxx (2015) xxx–xxx
Kappel, F., Fisher-Fleming, B., Hogue, E., 1996. Fruit characteristics and sensory
attributes of an ideal sweet cherry. HortScience 31, 443–446.
Kazantzis, K., 2013. Monograph of Sweet Cherry Cultivars Evaluated from Institute
of Pomology in Greece. Institute of Pomology, Elgo ‘Demeter’, 218.
Khadivi-Khub, A., 2014. Assessment of cultivated cherry germplasm in Iran by
multivariate analysis. Trees 28, 669–685.
Khodadadi, M., Fotokian, M.H., Miransari, M., 2011. Genetic diversity of wheat
(Triticum aestivum L.) genotypes based on cluster and principal component
analyses for breeding strategies. Aust. J. Crop Sci. 5, 17–24.
Koukourojiannis, V., 1996. O ˛
´ ς ␣␣␥␥´ ␣ о´␣
´ Ŵ␥´␣-Kоо´␣ (Greek) 2, 24–31.
␣.
Lacis, G., Kaufmane, E., Trajkovski, V., Rashal, I., 2009. Morphological variability and
genetic diversity within Latvian and Swedish sweet cherry collections. Acta
Univ. Latv. 753, 19–32.
Marshall, R.E., 1954. Cherries and cherry products. Economic Crops, vol. 5.
Inter-science, New York.
Mehmood, A., Jaskani, M.J., Khan, I.A., Ahmad, S., Ahmad, R., Luo, S., Ahmad, N.M.,
2014. Genetic diversity of Pakistani guava (Psidium guajava L.) germplasm and
its implications for conservation and breeding. Sci. Hortic. 172, 221–232.
Mohammadi, S.A., Prasanna, B.M., 2003. Analysis of genetic diversity in crop
plants—salient statistical tools and considerations. Crop Sci. 43, 1235–1248.
Peeters, J.P., Martinelli, J.A., 1989. Hierarchical cluster analysis as a tool to manage
variation in germplasm collections. Theor. Appl. Genet. 78, 42–48.
Petruccelli, R., Ganino, T., Ciaccheri, L., Maselli, F., Mariotti, P., 2013. Phenotypic
diversity of traditional cherry accessions present in the Tuscan region. Sci.
Hortic. 150, 334–347.
11
Pommer, C.V., 2012. Guava world-wide breeding: major techniques and cultivars
and future challenges. In III International Symposium on Guava and other
Myrtaceae 959, 81–88.
Rakonjac, V., Akšić, M.F., Nikolić, D., Milatović, D., Čolić, S., 2010. Morphological
characterization of ‘Oblačinska’sour cherry by multivariate analysis. Sci. Hortic.
125, 679–684.
Rodrigues, L.C., Morales, M.R., Fernandes, A.J.B., Ortiz, J.M., 2008. Morphological
characterization of sweet and sour cherry cultivars in a germplasm bank at
Portugal. Genet. Resour. Crop Evol. 55, 593–601.
Ruiz, D., Egea, J., 2008. Phenotypic diversity and relationships of fruit quality traits
in apricot (Prunus armeniaca L.) germplasm. Euphytica 163, 143–158.
Sánchez, R.P., Sánchez, M.A.G., Corts, R.M., 2008. Agromorphological
characterization of traditional Spanish sweet cherry (Prunus avium L.), sour
cherry (Prunus cerasus L.) and duke cherry (Prunus x gondouinii Rehd.) cultivars.
Span. J. Agric. Res. 6, 42–55.
Trujillo, I., Ojeda, M.A., Urdiroz, N.M., Potter, D., Barranco, D., Rallo, L., Diez, C.M.,
2014. Identification of the worldwide olive germplasm bank of Córdoba (Spain)
using SSR and morphological markers. Tree Genet. Genomes 10, 141–155.
UPOV, 1976. Guidelines for the Conduct of Test for Dis- Tintness, Homogeinity and
Stability of the Cherry. International Union for the Protection of New Varieties
of Plants, Genova, Italy, pp. 15.
Webster, A.D., 1996. The Taxonomic classification of sweet and sour cherries and a
brief history of their cultivation. In: Webster, A.D., Looney, N.E. (Eds.), Cherries.
Cab International, Wallingford OX 10 8DE, UK, pp. 3–25.
Please cite this article in press as: Ganopoulos, I., et al., Diversity of morpho-physiological traits in worldwide sweet cherry cultivars of
GeneBank collection using multivariate analysis. Sci. Hortic. (2015), http://dx.doi.org/10.1016/j.scienta.2015.09.061