Content uploaded by Sophie Nadot
Author content
All content in this area was uploaded by Sophie Nadot
Content may be subject to copyright.
Relationships among pansies (Viola section Melanium)
investigated using ITS and ISSR markers
R. Yockteng
1
, H. E. Ballard
2
Jr, G. Mansion
3
, I. Dajoz
4
, and S. Nadot
1
1
Laboratoire Ecologie, Syste
´
matique et Evolution, Universite
´
Paris-Sud, Orsay, France
2
Department of Environmental and Plant Biology, Por ter Hall, Ohio University, Athens, OH, USA
3
Institute of Botany, University of Neuchaˆ tel, Neuchaˆ tel
4
Laboratoire Ecologie, Ecole Normale Supe
´
rieure, Paris, France
Received October 28, 2002; accepted June 11, 2003
Published online: November 4, 2003
Springer-Verlag 2003
Abstract. Sequences of the nuclear region ITS and
the variable molecular markers ISSR were used to
estimate the phylogeny of the section Melanium of
the genus Viola. We confirm that the so-called
pansies form a derived and monophyletic group.
Two floral characters, the upturned side petals and
the large size of pollen grains appear to be synapo-
morphies in Melanium. The Melanium species are
very closely related, as shown by the reduced genetic
variation compared to the other section s of Viola.
Our analyses suggest x ¼5orx¼7 as the possible
base chromosome number of the section Melanium.
Polyploidy and hybridization would have played an
important role in the evolutionary history of this
clade resulting in a wide range of chromosome
number. The low genetic differentiation and the
complex cytological evolution suggest that
diversification in Melanium is the result of a reticu-
late evolution and rapid radiation in Europe and
Northern Africa.
Key words: Viola, phylogeny, ISSR, ITS, chromo-
some numbers, rapid radiation.
Introduction
Viola L. is the largest genus of the Violaceae
family, with 525–600 species (Clausen 1964,
Ballard 1996), distributed throughout most
frost-free regions of the world. The genus
probably arose in South America but most
centers of morphological and taxonomic diver-
sity occur in the Northern Hemisphere.
Numerous taxonomic studies on particular
species complexes have been published over
the 20th century based largely on morphology
(Becker 1925) and chromosome numbers
(Clausen 1927, 1929). A recent molecular
phylogenetic analysis (Ballard 1996, Ballard
et al. 1999) has clarified the composition and
relationships of the main groups and reevalu-
ated the placement of controversial
assemblages.
Section Melanium Ging., which includes
the so-called pansies, is derived and probably
monophyletic. It is a morphologically well-
defined group of about 80–100 species (Bal-
lard 1996). Most species are herbaceous,
caulescent, with frontally flattened flowers, a
yellow throat, big divided leaf-like stipules as
well as entire stipules and a well-developed
spur of variable length among species. The
side petals are upturned (downturned in the
other sections) and the bottom petal is
Plant Syst. Evol. 241: 153–170 (2003)
DOI 10.1007/s00606-003-0045-7
enlarged, serving probably as a landing plat-
form for insect pollinators. The style has a
characteristic capitate or globose shape with a
stigmatic orifice on a ventral rostellum. Cleis-
togamy has not been reported for members of
this section (Knuth 1908, Herrera 1993)
although it occurs in most species in the rest
of the genus. The geographical distribution
spans Europe and westernmost Asia. A few
species are found in Northern Africa and one
species is disjunct and probably native in
North America (Clausen et al. 1964). The
composition of section Melanium varies
according to authors. Ballard (1996) hypoth-
esizes that the woody sections Xylinosium and
Delphiniopsis may have arisen from within
Melanium. The woody stems, linear leaves
and long spurred flowers, adapted to hawk
moth pollination in the three Delphiniopsis
species, have been traditionally interpreted as
primitive characters by some taxonomists
(Beattie 1974) However, these are readily
argued to represent instead evolutionary spe-
cializations, as are the woody stems of the
four Xylinosium species in the Fynbos and
Mediterranean regions. Both sections have
the frontally flattened flowers with a yellow
corolla throat, which characterize the section
Melanium, and share cytological features with
different groups of this section (Ballard 1996).
If this most recent viewpoint is maintained,
section Melanium is composed of three
groups: Elongatae, Delphiniopsis and Xylino-
sium.
Cytological diversity is a striking feature of
this section. Whereas the base chromosome
number is relatively stable in the rest of the
genus (x ¼12, x ¼10 or x ¼6), it is extremely
diverse in the section Melanium where it ranges
from x ¼5tox¼17, with polyploid series
based on most of them (Ballard 1996, Erben
1996). Erben (1996) proposes x ¼11 as the
original base number, while Ballard has sug-
gested either x ¼5, x ¼6orx¼10 based on the
molecular phylogenetic relationships of the
whole clade containing Melanium and its
nearest sister groups (Ballard et al. 1999)
Determination of which hypothesis is correct
would depend on a robust phylogeny for
Melanium species themselves and placement
of certain ploidy levels within that phylogeny.
Intra-specific variation of chromosome num-
ber has been reported (Ku
¨
pfer 1971).
Hybridization is common throughout the
genus and often leads to fertile derivatives
(Stebbins et al. 1963, Ballard 1996, Erben
1996, Ko et al. 1998, Neuffer et al. 1999) in
spite of ploidy differences. Many species from
the section Melanium are believed to originate
from hybridization events (Clausen 1927,
Ku
¨
pfer 1971, Erben 1996), which suggests
potentially close relationships among species.
One can wonder to what extent hybridization
has driven the expansion of the section. Could
hybridization combined with allopolyploidy
and aneuploidy account for the high diversity
in chromosome numbers? Reconstructing the
cytological history of such a group, with a
putatively reticulate evolution, is a challenge.
A few attempts, based on chromosome num-
bers, have been made in the past (Clausen
1927, Ku
¨
pfer 1971), but there is no compre-
hensive study of the section. The increasing
sophistication of molecular markers offers the
possibility to investigate the phylogenetic rela-
tionships of closely related species (Small et al.
1998, Wolfe and Randle 2001), to detect the
hybrid origin of species (Mes et al. 1997,
Morrell and Rieseberg 1998, Sang and Zhang
1999) and reticulate evolution (Rieseberg 1991,
Sang et al. 1995) and to ultimately place
cytogenetic evolution within a broader, inde-
pendently derived context.
In this paper, we present a molecular
phylogeny of the section Melanium. Two kinds
of markers have been used for this purpose:
Internal Transcribed Spacer (ITS) nuclear
DNA sequences, which are routinely used for
resolving phylogenetic relationships within
genera (Yuan 1996, Compton et al. 1998,
Torrell et al. 1999, Blattner et al. 2001), and
Inter Simple Sequence Repeat (ISSR) markers,
which have been used for assessing intraspe-
cific as well as interspecific genetic diversity
and potential hybridization (Salimath et al.
1995, Fang et al. 1998, Ge and Sun 1999, Joshi
154 R. Yockteng et al.: Phylogeny of pansies using molecular markers
et al. 2000, Martin and Sa
´
nchez-Ye
´
lamo 2000,
Wolfe and Randle 2001). The phylogenetic
position of species belonging to the groups
Xylinosium and Delphiniopsis is examined, and
we propose scenarios for cytological evolution
in the section Melanium.
Materials and methods
Plant material and DNA isolation. Table 1 lists the
46 species studied and their taxonomic position
according to Ballard et al. (1999). Section Melani-
um Ging. is represented by 25 species of 80–100 and
sections Delphiniopsis and Xylinosium by 4 species;
the rest of the species are used as outgroups to root
the trees. The 25 Melanium species used in the
study were selected to represent the geographic
range of pansies. However, the number species
included was limited by the accessibility to the
plant material. A handful of species could not be
amplified successfully for both approaches, but
most taxa yielded data for both ITS sequences and
ISSR markers. Specimens were obtained either in
the field, from botanical gardens or from the
herbarium of the University of Neuchaˆ tel. Voucher
specimens are deposited at University of Neuchaˆ tel.
Total DNA was extracted from either fresh or dried
leaves, using the Dneasy Plant Mini Kit (Qiagen)
according to the manufacturer’s instructions.
ITS sequenc ing. The ITS1 and ITS 2 regions
were amplified simultaneously using universal
primers (5¢ GGA AGT AGA AGT CGT AAC
AAG 3¢ and 5¢ TCC TCC GCT TAT TGA TAT
GC 3 ¢, respectively). PCR reactions were per-
formed in a total volume of 50 ll containing 2 ll
DNA, 0.32 lM of each primer, 140 lM dNTPs,
2.5 mM MgCl
2
and 1U Taq Polymerase (Promega,
Madison, WI), in 1X buffer A (Promega, Madison,
WI). We used a PTC-100 Thermal cycler (MJ-
Research, Inc.) or, for some ITS sequences, a
Perkin Elmer 2400 model, programmed for an
initial step of 3 min at 94 C, followed by 35 cycles
of 40 s at 94 C, 50 s at 50–55 C, 1.5 min at 72 C
and a final step of 5 min at 72 C. PCR products
were visualized on a 2% agarose gel, purified using
the QIAquick PCR purification kit (Qiagen) or the
Promega Wizard PCR Preps kit and sequenced on
both strands with the DNA sequencing Big Dye
terminator kit (ABI Prism). Cycle sequencing
products were run on an ABI Prism 3700 DNA
Analyzer automated sequencer or a 310 capillary
sequencer (PE Biosystems, Foster City, California,
USA). Sequences were deposited in GenBank.
ISSR. The seven primers employed wer e
based on SSR motifs ((CA)
6
-GC, (CT)
8
-AC,
(CT)
8
-TG, (CA)
6
-AC, (CT)
8
-GC, (CA)
6
-AG and
(CA)
6
-GT) reported for flowering plants (Wolfe
et al. 1998, Esselman et al. 1999). Reactions con-
sisted of 1X Taq DNA polymerase buffer, 0.2 mM
of each dNTP, 3 mM MgCl
2
, 1.3 lM primer, 2%
Formamide, 1 ll DNA and 1U Taq DNA poly-
merase (Promega Corp.) in a total volume of 15 ll.
Amplifications were performed in a PTC-100
thermal cycler (MJ Research, Inc.) using the
following temperature profile: 90 s at 94 C, 35
cycles [40 s at 94 C, 45 s at 44 C and 90 s at
72 C] followed by 45 s at 94 C, 45 s at 44 C and
5 min at 72 C. Amplification products were
visualized on 1.5% agarose gels stained with
ethidium bromide and photographed under UV
light. Band sizes were estimated with the 200–
10,000 bp Ladder and the 100–1000 bp Ladder
(Smart Ladder, Eurogentec, Belgium).
Data analys es. ITS Sequences were aligned
using Clustal X (Thompson et al. 1997) with final
corrections made manually. The 5.8S coding region
was excluded from the alignment. Phylogenetic
analyses were performed with PAUP 4.0b8a*
(Swofford 2001). Indel regions were scored as
additional characters using Gapcoder (Young and
Healy 2002). Maximum Parsimony analysis (MP)
was performed using the heuristic search algorithm,
with the MULTREES option on, TBR branch
swapping, and 1000 replicates with random addi-
tion saving 100 trees per replicate. Bootstrapping
(100 replicates) was done using the previous
parameters. Jukes-Cantor pairwise distances were
used for distance analysis with the Neighbor
Joining (NJ) method. Maxim um Likelihood anal-
ysis was performed using the heuristic search
algorithm with TBR branch swapping, 10 replicates
with random addition option and the following
model of evolution: TIM+G, alpha shape param-
eter equal to 1.3044, estimated using Modeltest
version 3.06 (Posada 2001).
The DNA fragments obtained using the ISSR
primers were scored as present (1) or absent (0).
Only fragments with a strong and clear signal were
scored. Bands of identical size were assumed
homologous across species sampled in this study.
The 0/1 matrix was used to calculate a similarity
R. Yockteng et al.: Phylogeny of pansies using molecular markers 155
Table 1. List of species included in this study. Taxonomic classification and haploid chromosome number are given. Voucher data include
primary collector or herbarium voucher are deposited. GenBank numbers are included. Extractions from herbarium material are indicated by an
asterisk
Sections Species Species Distribution n Ref. Voucher GenBank
Number
Melanium
Viola aetolica Boiss.
& Heldr.
Greece, Albany,
Yougoslavia
8 1 HR1 AY148225,
Viola alpina Jacq. NE Alps, Carpaths
(Carinthia)
11 2 SN-0007 AY148245
Viola arvensis Murr. Europe 17 2 SN-0003 AY148226,
AY148246
Viola bertolonii De Salis Corsica 20 2 HR3 AY148227,
AY148247
Viola calcarata L. Alps, Balkans (Europe) 20 2, 4 SN-0005 AY148229,
AY148249
Viola cheiranthifolia
Humb. & Bonpl.
Canary Isl. Tenerife circa 32
c
3 RBGE:
19841491
AY148231,
AY148251
Viola comollia Messara* S Alps (Italy) 11 3 HIB 25831 AY148232,
AY148252
Viola cornuta L.* Pyre
´
ne
´
es (S Europe) 11 2, 4 HIB 03064
Viola corsica Nyman Corsica, Sardinia 60, 26 2, 3 HR5
Viola dyris Maire* Morocco 10, 11 2, 6 HIB 03336
Viola eugeniae Parl. Appenines (Italy) 17
b
2 HR7 AY148233,
AY148253
Viola gracilis Sibth. & Sm. Greece 10
q
17 SN-0004 AY148234,
AY148254
Viola kitaibeliana
Roem. & Schultes
Europe, Asia, USA 7, 8, 12, 18, 24 2 SN-0008 AY148235,
AY148255
Viola lutea Hudson* CE Europe, S-Pyrenees 24 2, 4 HIB 03032 AY148236,
AY148256
Viola magellensis Porta
& Rigo ex Strobl.
Appenines, Greece,
Albany
11 2 HIB? AY148237,
AY148257
Viola minuta Bieb.* Caucase ? HIB 03456
Viola montcaunica Pau* Spain 11 7 HIB 03024
Viola argenteria B. Moraldo
& G.Forneris*
Alps, Corsica (Europe) 7
c
3 HIB 03038
156 R. Yockteng et al.: Phylogeny of pansies using molecular markers
Table 1 (continued)
Viola oreades Bieb.* Anatolia, Caucas, Crimea ? HIB 03026
Viola palmensis Webb
& Berth.
Canary Isl. La Palma ? HR8 AY148239,
AY148259
Viola parvula Tineo* S Europe, Syria 5 2.8 HIB 03025 AY148240,
AY148260
AY148241,
AY148261
Viola perinensis W. Becker Greece, Bulgaria
(Macedonia)
10, 11 2, 1 HR9
Viola stojanowii W. Becker Greece, Bulgaria
(Macedonia)
13 1 SN-0001
Viola tricolor L. Europe, North America 13 2, 4 SN-0010 AY148243,
AY148263
Viola valderia All. Alps 10 2, 4 HIB 03056 AY148244,
AY148264
Xylinosium
Viola arborescens L. Mediterranean region 26 3 KEW 316-
74.02529 -
10220
Viola saxifraga Maire Marocco 26 4 HR10
Viola decumbens L.f. South Africa
Delphiniopsis
Viola cazorlensis Gand. Spain 10 4, 5 SN-0006 AY148230,
AY148250
Viola I –
Outgroup
Viola rotundifolia Michx. N America 6 Ballard et al.
1999
AF0972241,
AF097287
Viola II –
Outgroup
Viola blanda Brainerd N America 22 14.15 Ballard et al.
1999
AF097238,
AF097284
Viola jalapaensis W. Becker Mexico Ballard et al.
1999
AF097235,
AF097281
Viola pinnata L. Eurasia, N America 24 3 Ballard et al.
1999
AF097240,
AF097286
R. Yockteng et al.: Phylogeny of pansies using molecular markers 157
Table 1 (continued)
Sections Species Species Distribution n Ref. Voucher GenBank
Number
Viola III –
Outgroup
Viola odorata L. C Europe,
S and N America,
Asia
10 2 SN-0002 AY148238,
AY148258
Viola reichenbachiana
Jordan ex Boreau
Europe W-Asia,
Middle East,
North Africa
10 AF097248,
AF097294
Viola riviniana Reichb. Europe, Mediterranean
region, Middle East
20 SN-0009 AY148242,
AY148262
Dischidium –
Outgroup
Viola biflora L.* Boreal and Temperate
Regions
6 2 HIB 03068 AY148228,
AY148248
Chamaemelanium I –
Outgroup
Viola barroetana Hemsley Mexico Ballard et al.
1999
AF097224,
AF097270
Viola pubescens Aiton N America 6 13 Ballard et al.
1999
AF097225,
AF097271
Viola purpurea Kellog N America,
Baja California
6 11 Ballard et al.
1999
AF097229,
F097275
Viola sheltonii Torr. N America 6 11 Ballard et al.
1999
AF097226,
AF097272
Viola vallicola A. Nelson N America,
Baja California
6 16 Ballard et al.
1999
AF097230,
AF097276
Chamaemelanium II –
Outgroup
Viola cuneata S. Watson N America,
N Mexico
Ballard et al.
1999
AF097234,
AF097280
Viola flagelliformis Hemsley Mexico 6 13 Ballard et al.
1999
AF097233,
AF097279
158 R. Yockteng et al.: Phylogeny of pansies using molecular markers
matrix using the Dice coefficient (Salimath et al.
1995). A phenogram was constructed with UP-
GMA using NTSYS 1.8 (Rohlf 1999). An analysis
of maxi mum parsimony was also done with a
heuristic search using PAUP 4.0b8a (Swofford
2001). Robustness of the UPGMA and Maximum
Parsimony trees was evaluated by a bootstrap
analysis. Principal coordinates analysis (PCoA)
was used to find relationships between ISSR
markers and specimens without a priori division
of the samples into discrete groups (Wiley 1981).
PCoA was conducted under MatLab (version 5,
The Math Works Inc.) using a Dice dist ance matrix
(1 – Dice similarity coefficient).
We traced the distribution of haploid chromo-
some number on the Neighbor Joining ITS tree and
the consensus ISSR cladogram (MP) using Mac
Clade (Maddison and Maddison 1992).
Results
ITS. The alignment of ITS1 and ITS2 se-
quences consisted of 510 positions (ITS1: 293
positions and ITS2: 217 positions). 127 indels
were coded as additional characters. This
alignment included the sequences from 20
individuals from the section Melanium (repre-
senting 17 species), from 19 species represent-
ing the other sections of the genus and from
Hybanthus concolor (Violaceae), used as an
outgroup (Table 1). 227 sites were parsimony
informative (35.5%). The parsimony informa-
tive positions were reduced to 14 when only
species from section Melanium were included,
revealing highly reduced genetic divergence
within the pansies.
The heuristic search generated 8300 most
parsimonious trees of 875 steps with a consis-
tency index of 0.644 and a retention index of
0.752, indicating a moderate level of homo-
plasy. The strict consensus tree resulted in a
largely unresolved polytomy for the pansies,
but their monophyletic origin is supported
with a 100% bootstrap value (Fig. 1). The
species representing the sections Delphiniopsis
and Xylinosium are located apart from the
Melanium species. Viola cazorlensis (section
Delphiniopsis) appears as sister group to Mel-
anium in all analyses, although this relation-
Table 1 (continued)
Leptidium –
Outgroup
Viola scandens Willd.
Ex Roem. & Schult.
Mesoamerica, West Indies,
S America
27 12 Ballard et al.
1999
AF097221,
AF097267
Chilenium –
Outgroup
Greece, Albany,
Yougoslavia
Viola reichei Skottsberg Chili, Argentina Ballard et al.
1999
AF097222,
AF097268
Outgroup
Hybanthus concolor
(T. F. Forst.) Spreng.
N America Ballard et al.
1999
AF097218,
AF097264
1: (Erben 1985); 2: (Bolkhovskikh et al. 1969); 3: (Merxmu
¨
ller and Lippert 1977); 4: (Clausen 1927); 5: (Ku
¨
pfer 1972); 6: (Favarger et al. 1979);
7: (Merxmu
¨
ller 1974); 8: (Schmidt 1962); 9: (Galland 1985); 10: (Strid 1996); 11: (Clausen 1964); 12: (Damboldt and Phitos 1971); 13: (Clausen
1929); 14: (Gershoy 1932); 15: (Canne 1987); 16: (Davidse 1976); 17: (Tutin et al. 1968)
HR: Collection Heinz Rehfeld, D.-Erxleben Str.9 06484 Quedlinburg, Germany; RBGE: Royal Botanic Garden, Edinburgh, UK ; KEW: Royal
Botanic Gardens, Kew, UK; HIB: Herbarium of Institute of Botany, University of Neuchaˆ tel, Switzerland; SN: New samples deposited in the
HIB; Ballard et al. 1999: Data used from Ballard et al. (1999)
R. Yockteng et al.: Phylogeny of pansies using molecular markers 159
Fig. 1. Strict consensus tree of 8300 most parsimonious trees resulting from a heuristic search with TBR,
random addition sequence (nreps ¼1000) and MulTrees ON, based on a data set of ITS sequen ces. Bootstrap
values higher than 50% are indicated above branches. Sections are according to (Ballard et al. 1999)
160 R. Yockteng et al.: Phylogeny of pansies using molecular markers
ship is only moderately supported in the
Maximum Parsimony tree (bootstrap value of
62%). In both distance (NJ) and Maximum
Likelihood analyses (not shown), Xylinosium is
sister group to a clade comprising Melanium,
Delphiniopsis and Viola.
The level of ITS sequence divergence with-
in section Melanium was estimated using
uncorrected pairwise distance and compared
to other sections of Viola. Significantly lower
values were found for the pansy group (mean
value = 0.023) compared to related sections
(0.032 to 0.092) and subsections (0.051 to
0.068) (Kolmogorov-Smirnov test, p < 0.01 in
SAS System).
ISSR. ISSR data from 7 primers were
obtained for 32 species (3 from section Viola,1
from sections Dischidium and Delphiniopsis,2
from section Xylinosium and 25 from section
Melanium). Species from sections other than
Melanium were used as outgroups. We selected
primers that displayed interspecific variation
but no or little intraspecific variation.
A total of 177 polymorphic fragments
ranging from 230 to 1556 bp were scored. We
scored 27 bands for the primer (CA)
6
-GC, 15
bands for (CT)
8
-AC, 20 bands for (CT)
8
-TG,
33 bands for (CA)
6
-AC, 24 bands for (CT)
8
-
GC, 33 bands for (CA)
6
-AG and 25 bands for
(CA)
6
-GT. 84% of the 177 characters con-
sisted of bands shared by at least two species,
and were therefore parsimony informative.
sixty-six bands were found exclusively in spe-
cies belonging to Melanium and were therefore
autapomorphies of this section, although none
of the bands were shared by all 25 pansies. The
smallest pairwise similarities were found
among outgroups, ranging between 0 and
0.355. Values ranged between 0.127 and
0.759 among pansies, the highest similarity
being between V. cheiranthifolia and V. palm-
ensis, from the Canary Islands. Whatever
method was used for phylogenetic inference,
MP (Fig. 2) or distance (not shown), all species
belonging to Melanium were clustered together
and all the species chosen as outgroups were
basal to Melanium. Viola arborescens and
V. saxifraga were grouped together, according
to their placement in section Xylinosium (Gal-
land 1998) Within Melanium, the tree topology
shows a basal group of 5 species including V.
parvula, confirming the result obtained with
the ITS sequence data for this species (Fig. 1).
The rest of the species are distributed in three
main cluster. One regroups 4 species widely
distributed in Europe (V. arvensis, V. lutea, V.
cornuta and V. tricolor). Another cluster
regroups 7 more narrowly endemic species
(V. stojanowii, V. bertolonii, V. corsica, V.
calcarata, V. perinensis, V. cheiranthifolia and
V. palmensis) found in Southern Europe or the
Canary Islands. The position of V. aetolica, V.
eugeniae, V. gracilis and V. alpina varies
between distance and MP analyses. The prin-
cipal coordinates analysis (Fig. 3) supports the
inclusion of these 4 species in the two cluster
above mentioned, as found in the MP analysis.
Viola kitaibeliana has an intermediate position
between this group and the more basal species.
Chromosome numbers. Chromosome
number evolution was reconstructed with
MacClade using the MP tree based on ISSR
data (Fig. 4). It gives x ¼7 as a hypothetical
base chromosome numbers for the section
Melanium. Reconstruction using the NJ tree
based on ITS data gives x ¼5. The number
x ¼11 appears to be the ancestral state for the
most derived groups of Melanium, which
display the highest diversity of chromosome
numbers.
Discussion
For sorting out phylogenetic relationships in a
group of closely related plants, the most widely
used markers are ITS and non-coding chloro-
plast regions such as trnL-F (Gielly and
Taberlet 1994, Gielly et al. 1996, Johansson
1998, Fukuda et al. 2001). As a database of
ITS sequences of Viola species was already
available, we chose to start our investigation of
phylogenetic relationships in section Melanium
using ITS sequences, in the hope of finding
phylogenetic structure in the section and
identifying clusters of species. Based on our
results, section Melanium, which consists of all
R. Yockteng et al.: Phylogeny of pansies using molecular markers 161
the species considered as ‘‘pansies’’, is clearly
monophyletic, as shown by the strong boot-
strap support. It confirms that the orientation
of lateral petals, used as a criterion for
distinguishing between pansies and violets, is
a synapomorphy of pansies. This particular
shape of flower is encountered in only one
other complex species of violets, the circum-
boreal V. biflora species complex. ITS
sequence data confirm morphological classifi-
Fig. 2. Strict consensus of 4 trees resulting from a heuristic search based on ISSR data from 25 Melanium
species and 7 outgroup species. Bootstrap values higher than 50% are indicated above branches. Sections in
bold are according to Ballard et al. (1999), divisions following species names are according to Melchior (1960)
162 R. Yockteng et al.: Phylogeny of pansies using molecular markers
cations, which consider this feature as a
convergence: this species is placed in section
Dischidium, not directly related to section
Melanium (Fig. 1). Another morphological
synapomorphy of pansies is the size of pollen
grains, which was measured for a sample of the
species included in this study. Pollen grains are
significantly larger in section Melanium than in
outgroups (mean length 56 lm within Melani-
um,33lm for outgroups; T test p < 0.01). It
can be noted that the size of pollen in sections
Xylinosium and Delphiniopsis is not signifi-
cantly different from the size of pollen in the
other outgroups which confirms the distinction
between section Melanium and sections Xyli-
nosium and Delphiniopsis.
However, the alignment of ITS1+ITS2
sequences revealed very little variation among
pansy species. The number of nucleotide
differences between species ranges from 0 to
16 (0 to 4.3%). In fact, intraspecific variation
evaluated by the sequences of two individuals
for V. lutea (3 differences), V. arvensis (2
differences) and V. calcarata (6 differences),
was comparable to interspecific variation. The
large polytomy observed in the tree presented
in Fig. 2 shows a complete lack of resolution.
Consequently, whereas ITS sequence data
proved useful to clarify relationships among
sections (Compton et al. 1998, Esselman et al.
1999, Torrell et al. 1999, Bell and Patterson
2000), it clearly offers no help within section
Melanium. Chloroplast regions are now used
routinely for infrageneric phylogenetic studies
(Gielly and Taberlet 1994, Gielly et al. 1996,
Johansson 1998, Wang et al. 1999). The
Fig. 3. Principal Coordinates Analysis (PCoA) plot basedonISSRdata.Thetwoprincipalcoordinates
account for an accumulate variation of 21.2%. Chromosome numbers (n) are indicated
R. Yockteng et al.: Phylogeny of pansies using molecular markers 163
Fig. 4. Reconstruction of the evolution of chromosomal number using Equivocal Cycling option of MacClade
3.08 (Maddison and Maddison 1992), using the ISSR MP tree. First of 684 most parsimonious reconstructions.
Branch numbers are haploid (meiotic) chromosome numbers, whereas those in parentheses following species
names are diploid (sporophytic) numbers
164 R. Yockteng et al.: Phylogeny of pansies using molecular markers
potential variability of the chloroplast ge-
nome within Melanium was evaluated by
using the PCR-RFLP technique. Two regions,
rpoC1 and trnH-trnK, were amplified and
digested with an array of 4-base and 6-base
cutting enzymes (data not shown). No varia-
tion at all was detected among pansies,
indicating a high degree of sequence identity.
Instead of sequencing very long regions, with
limited chances to find variation, our choice
was to use ISSR markers. These markers have
a level of variability similar to or greater than
that of RAPDs, but do not generally suffer as
much from the problem of reproducibility.
They have been used mostly for assessing
genetic diversity among populations (Essel-
man et al. 1999, Gilbert et al. 1999, Camacho
and Liston 2001) but they have also been
used successfully for studying phylogenetic
relationships and potential hybridization and
introgression among closely related species
(Fang et al. 1998, Joshi et al. 2000, Wolfe and
Randle 2001).
Although most nodes on the distance or
the MP trees (Fig. 2) are not supported by
high bootstrap values, it can be noted that the
general clustering pattern of species is similar
in both trees and in the PCoA (Fig. 3). None
of the classifications of the section proposed
previously are in agreement with our results
(Drabble 1909, Shishkin 1949, Melchior 1960),
suggesting a high level of homoplasy in mor-
phological distinctions of subsections. Melani-
um was divided in two groups by Melchior
(1960): Scaposae (absence of above-ground
stem) and Elongatae (presence of above-
ground stem), the latter being subdivided again
in Integrifoliae (entire leaves) and Crenatifoliae
(crenate leaves). Our tree (Fig. 2) does not
support this classification. Most pansies have
crenate leaves and the few species with entire
leaves have evolved independently. The only
species representing the group Scaposae,
V. alpina, appears derived from Elongatae.
The occurrence of hybridization events could
affect the morphological characters as the leaf
form leaving the phylogenetic value of these
type of character very questionable. Addition-
ally, the ISSR results do not follow a geo-
graphic pattern: no clear correlation appears in
the trees between species relationship and
geographic distribution. Although the ISSR
markers bring valuable information about the
relationships among pansies, the interpretation
could be complex if Melanium species have
undergone a reticulated evolution. In this case,
introgression events could affect the pattern of
molecular markers in species originated by
hybridization. To confirm the occurrence of
reticulation processes during the evolutionary
history of Melanium, it will be necessary to
conduct a thorough molecular analysis in
order to find specific markers to identify the
hypothetical parental species and verify the co-
occurrence of these markers in species formed
by hybridization.
Some remarks can be made about chro-
mosomal evolutionary pattern. In a paper
discussing the role of hybridization in forming
species in Melanium, Erben (1996) suggested
that this section would have evolved from an
ancestor with x ¼11, with decreasing and
increasing dysploidy taking place through
structural changes. Our results do not contra-
dict this hypothesis, since a paraphyletic group
of species with 2n ¼22 is located near the base
of the section. The number 2n ¼20 would be
derived from 2n ¼22 by the fusion of 2
chromosomes, forming a large metacentric
chromosome (Erben 1996). Under such a
scenario of evolution, the low chromosome
numbers of V. parvula (2n ¼10) and V. argen-
teria (2n ¼14) would be interpreted as derived,
resulting from decreasing dysploidy. However,
our reconstruction of the chromosome number
evolution suggest that the base number of the
section is either x ¼5 using the ITS tree or
x ¼7 using the ISSR tree. The derived chro-
mosome number x ¼11 would be then the
result of hybridization process between species
with x ¼7 and x ¼5 and further fusion of
chromosomes. Under this scenario, V. parvula
(and possibly also V. argenteria) is wrongly
positioned in some of the ISSR trees, and is in
fact sister group to a clade including all other
Melanium species, as confirmed by the ITS
R. Yockteng et al.: Phylogeny of pansies using molecular markers 165
data set. Pollen morphology also supports this
scenario. Both V. parvula and V. argenteria
have smaller pollen grains than the remaining
species of the section, and pollen grains are
predominantly 3-aperturate, like violets,
whereas all other pansies examined produce
mostly 4-aperturate pollen grains. These two
species could then be representative of a sister
group of section Melanium in which decreasing
dysploidy would have happened. Further
addition of species and use of other types of
DNA data, would be interesting in order to
examine more closely the phylogenetic position
of Melanium species with low base numbers.
In the ISSR tree, the most derived group of
pansies shows extremely variable chromosome
numbers, ranging from 2n ¼16 to 120, and no
clear pattern of evolution appears from the
tree (Fig. 4). It has been suggested that pansies
have undergone a reticulate evolution (Ku
¨
pfer
1971, Erben 1996): indeed, crosses between
species are easy in pansies, even with different
chromosome numbers, and often give fertile
hybrids. This phenomenon would explain the
weakly supported relationships among pan-
sies. It would also explain why most species of
pansies display pollen heteromorphism: this
phenomenon has been shown to arise as a
consequence of polyploidization (Bronckers
1963, Mignot et al. 1994, Nadot et al. 2000)
which is often associated with hybridization.
Polyploidization and/or hybridization could
also account for the larger size of pollen grains
in pansies, as it has been shown that hybrid-
ization can affect pollen morphology (Cha-
turvedi et al. 1999) The low level of genetic
differentiation revealed by the ITS analysis
could be interpreted as resulting from explo-
sive and quite recent radiation (Hodges and
Arnold 1994, Givnish and Sytsma 1997, Yuan
and Ku
¨
pfer 1997, Ainouche and Bayer 1999,
Hahn and Sytsma 1999, Blattner et al. 2001).
When examining different sections and sub-
sections of Viola, pansies appear to have the
lowest level of genetic differentiation of ITS
(0.023%), even compared to the most derived
group in the genus, namely the Hawaiian
violets of section Nosphinium (0.032%)
(Ballard et al. 1999, Ballard and Sytsma
2000). Their sequence identity is therefore
likely to reflect an explosive radiation, and
not simply a recent origin. The apparition of a
key innovation, such as the different orienta-
tion of lateral petals perhaps coupled with
lateral petal trichomes, corolla color differen-
tiation and modification of the shape, could
have triggered the diversification of section
Melanium, the changes in flower morphology
allowing them to access a broader array of
pollinators compared to the ‘‘ancestral’’ types
(Hodges and Arnold 1994).
The molecular study presented here is only
preliminary and further work is needed in
order to understand the evolution of pansies.
To clarify conclusively the relationships
among pansies, it would be helpful to increase
the ISSR data set by adding extra species or
using markers such as Amplified Fragment
Length Polymorphisms (AFLP’s). In addition,
such a complex reticulate evolution will be
better understood by using cytogenetic meth-
ods such as Fluorescent or Genomic In Situ
Hybridization, by conducting artificial hybrid-
izations and observing meiotic behavior of
both synthetic and putative hybrids, and by
comparison of data from biparentally and
uniparentally inherited genomes (e.g. nuclear
and chloroplast simple sequence repeats).
We are very grateful to Philippe Ku
¨
pfer (Uni-
versite
´
de Neuchaˆ tel) and Sonia Yakovlev for
helpful discussions and to Michel Baylac (MNHN
Paris) for his help with MatLab. We are also
indebted to Franz Tod and W alter Till of Institute
of Botany and Botanical Garden of Vienna Uni-
versity, Philip Ashby of Royal Botanical Garden in
Edinburgh, Olga Baeta of Madeira Botanical
Garden, Boris Turk of Alpine Botanical Garden
Juliana in Slovenia, Simonetta Peccenini, Ire
`
ne Till-
Botraud and Carlos Herrera for generously pro-
viding leaf and seed material for DNA extraction.
References
Ainouche A. K., Bayer R. J. (1999) Phylogenetic
relationships in Lupinus (Fabaceae: Papilionoi-
deae) based on internal transcribed spacer (ITS)
166 R. Yockteng et al.: Phylogeny of pansies using molecular markers
of nuclear ribosomal DNA. Amer. J. Bot. 86:
590–607.
Ballard H. E. (1996) Phylogenetic relationships and
infrageneric groups in Viola (Violaceae) based
on morphology, chromosome numbers, na tural
hybridization and Internal Transcribed Spacer
(ITS) sequences. PhD Thesis, University of
Wisconsin, Maddison.
Ballard H. E., Sytsma K. J. (2000) Evolution and
biogeography of the woody hawaiian violets
(Viola, Violaceae): arctic origins, herbaceous
ancestry and bird dispersal. Evolution 54(5):
1521–1532.
Ballard H. E., Sytsma, K. J., Kowal, R. R. (1999)
Shrinking the violets: Phylogenetic relationships
of infrageneric groups in Viola (Violaceae) based
on internal transcribed DNA sequences. Syst.
Bot. 23(4): 439–458.
Beattie A. J. (1974) Floral evolution in Viola. Ann.
Missouri Bot . Gard. 61: 781–793.
Becker W. (1925) Viola. In: Engler A., Prantl K.
(eds.) Die natu
¨
rlichen Pflanzenfamilien. Verlag
von Wilhelm Engelmann, Leipzig, Germany,
pp. 363–376.
Bell C. D., Patterson, R. W. (2000) Molecular
phylogeny and biogeography of Linanthus (Pol-
emoniaceae). Amer. J. Bot. 87(12): 1857–1870.
Blattner F. R., Weising K., Ba
¨
nfer G., Maschwitz
U., Fiala B. (2001) Molecular analysis of phy-
logenetic relation ships among myrmecophytic
Macaranga species (Euphorbiaceae). Mol.
Phylogent. Evol. 19(3): 331–344.
Bolkhovskikh Z., Grif V., Matvejeva, T., Zakh-
aryena O. (1969) Chromosome numbers of
flowering plants. Leningrad.
Bronckers F. (1963) Variations polliniques dans
une se
´
rie d’autopolyploı
¨
des artificiels d’Arabid-
opsis thaliana (L.) Heyhn. Pollen Spores 5: 233–
238.
Camacho F. J., Liston A. (2001) Popul ation
structure and genetic diversity of Botrychium
pumicola (Ophioglossaceae) based on inter-sim-
ple sequences repeats (ISSR). Amer. J. Bot.
88(6): 1065–1070.
Canne J. M. (1987) Determinations of chromosome
numbers in Viola (Violaceae). Canad. J. Bot. 65:
653–655.
Chaturvedi M., Datta K., Pal M. (1999) Pollen
anomaly – a clue to natural hybridity in Arge-
mone (Papaveraceae). Grana 38(6): 339–342.
Clausen J. (1927) Chromosome number and the
relationship of species in the genus Viola. Ann.
Bot. 41(156): 677–714.
Clausen J. (1929) Chromosome number and the
relationship of some North Ame rican species of
Viola. Ann. Bot. 63: 741–764.
Clausen J. (1964) Cytotaxonomy and distributional
ecology of western North American violets.
Madron
˜
o 17: 173–204.
Clausen J., Channell R. B., Nur U. (1964) Viola
rafinesquii, the only Melanium violet native to
North America. Rhodora 66: 32–46.
Compton J. A., Culham A., Gibbings J. G., Jury S.
L. (1998) Phylogeny of Actaea including Cimi-
cifuga (Ranunculaceae) inferred from nrDNA
ITS sequence variation. Biochem. Syst. Ecol 26:
185–197.
Damboldt J., Phitos D. (1971) IOPB Chromosome
number reports 34. Taxon 20: 785–797.
Davidse G. (1976) A study of some intermountain
violets (Viola, sect. Chamaemelanium). Madron
˜
o
23: 274–283.
Drabble E. (1909) The British pansies. J. Bot. 47:
1–32.
Erben M. (1985) Cytotaxonomische Untersuchun-
gen an su
¨
dosteuropa
¨
ischen Viola – Arten der
Sektion Melanium. Mitt. Bot. Staatssamm.
Mu
¨
nchen 21: 339–740.
Erben M. (1996) The significance of hybridization
on the forming of species in the genus Viola.
Bocconea 5: 113–118.
Esselman E. J., Jianqiang L., Crawford D. J.,
Winduss J. L., Wolfe A. D. (1999) Clonal
diversity in the rare Calamagostris porteri ssp.
insperata (Poaceae): comparative results for
allozymes and random amplified polymorphic
DNA (RAPD) and intersimple sequence repeat
(ISSR) markers. Molec. Ecol. 8: 443–451.
Fang D., Krueger R. R., Roose M. L. (1998)
Phylogenetic relationships among selected Citrus
germplasm accessions revealed by Inter-Simple
Sequence Repeat (ISSR) markers. J. Amer. Soc.
Hort. Sci. 123(4): 612–617.
Favarger C., Galland N., Ku
¨
pfer P. (1979) Recher-
ches cytotaxonomiques sur la flore orophile du
Maroc. Nat. Monspel., Ser. Botanique 29: 1–64.
Fukuda T., Yokoyama J., Ohashi H. (2001) Phy-
logeny and biogeography of the genus Lycium
(Solanaceae): inferences from chloroplast DNA
sequences. Mol. Phylogenet. Evol. 19(2): 246–258.
R. Yockteng et al.: Phylogeny of pansies using molecular markers 167
Galland N. (1985) IOPB Chromosome Number
Reports 87. Taxon 34: 346–351.
Galland N. (1998) Recherche sur l’origine de la
flore orophile du Maroc: e
´
tude caryologique et
cytoge
´
ographique. Travaux de l’Institut Sci.,
Univ. Mohammed V.; Se
´
ries Botanique 35: 1–
168.
Ge X. J., Sun M. (1999) Reprodu ctive biology and
genetic diversity of a cryptoviviparous mangrove
Aegiceras corniculatum (Myrsinaceae) using
allozyme an d intersimple sequence repeat (ISSR)
analysis. Mol. Ecol. 8: 2061–2069.
Gershoy A. (1932) Descriptive notes for Viola
exhibits. I. General considerations, Vermont
Agricultural Experiment Station Publication.
Gielly L., Taberlet P. (1994) Chloroplast DNA
polymorphism at the intrageneric level and plant
phylogenies. Com pt. Rend. Acad. Paris 317:
685–692.
Gielly L., Yuan Y., Ku
¨
pfer P., Taberlet P. (1996)
Phylogenetic use of noncoding regions in the
genus Gentiana L.: chloroplast trnL (UAA)
intron versus nuclear ribosomal internal tran-
scribed spacer sequences. Mol. Phylogent. Evol.
5(3): 460–466.
Gilbert J. E., Lewis R. V., Wilkinson M. J., Cal igari
P. D. S. (1999) Developing an appropriate
strategy to asses s genetic variability in plant
germplasm collections. Theor. Appl. Genet. 98:
1123–1131.
Givnish T. J., Sytsma K. J. (1997) Molecular
evolution and adaptive radiation. New York,
NY, Cambridge University Press.
Hahn W. J., Sytsma K. J. (1999) Molecular
systematics and biogeography of the Southeast
Asian genus Caryota (Palmae). Syst. Bot. 24(4):
558–580.
Herrera C. M. (1993) Selection on floral morphol-
ogy and environmental determinants of fecun-
dity in a hawkmoth-pollinated violet. Ecol.
Monogr. 63: 251–275.
Hodges S. A., Arnold M. L. (1994) Columbines: A
geographically widespread species flock. Proc.
Natl. Acad. Sci. USA 91: 5129–5132.
Johansson J. T. (1998) Chloroplas t DNA restric-
tion site mapping and the phylogeny of Ranun-
culus. Plant Syst. Evol. 213: 1–19.
Joshi S. P., Gupta V. S., Aggarwal R. K., Ranjekar
P. K., Brar D. S. (2000) Genetic diversity and
phylogenetic relationshi ps as revealed by inter
simple sequence repeat (ISSR) polymorphism in
the genus Oryza. Theor. Appl. Genet. 100: 1311–
1320.
Knuth P. (1908) Handbook of flower pollination
Vol II. London, Clarendon Press.
Ko M. K., Yang J., Jin Y. H., Lee C. H., Oh B. J.
(1998) Genetic relationships of Viola species
evaluated by ran dom amplified polymorphic
DNA analysis. J. Horticult. Sci. Bi otech 73(5):
601–605.
Ku
¨
pfer P. (1971) Contribution a
`
l’e
´
tude cytologi-
que et phyloge
´
ne
´
tique de la section Melanium
Ging. du genre Viola L. Compt. Rend. Acad.
Sci.; Se
´
rie D 272: 1085–1088.
Ku
¨
pfer P. (1972) Cytotaxonomie et cytoge
´
ographie
de quelques groupes d’orophytes du bassin
occidental de la Me
´
diterrane
´
e et des Alpes.
Compt. Rend. Hebdo. Se
´
ances Acad. Sci.; Se
´
ries
D 275: 1753–1756.
Maddison W. P., Maddison D. R. (1992) Mac
Clade: analysis of phylogeny and character
evolution. Sinauer, Sunderland, Mass.
Martin J. P., Sa
´
nchez-Ye
´
lamo M. D. (2000)
Genetic relationships among species of the genus
Diplotaxis (Brassicaceae) using inter-simple
sequence repeat markers. Theor. Appl. Genet.
101: 1234–1241.
Melchior H. (1960) Violaceae. In: Engler A., Prantl
K. (eds.) Die natu
¨
rlichen Pflanzenfamilien.
Duncker and Humblot, Berlin.
Merxmu
¨
ller H. (1974) Veilchenstudien I-I Viola.
Phyton 16: 137–158.
Merxmu
¨
ller H., Lippert W. (1977) Veilchenstudien
V-VII. Mitt. Bot. Staatssamml. Mu
¨
nchen 13:
503–534.
Mes T. H. M., Wijers G. J., Hart H. (1997)
Phylogenetic relationships in Monanthes (Cras-
sulaceae) based on morphological, chloroplast
and nuclear DNA variation. J. Evol. Biol. 10(2):
193–216.
Mignot A., Hoss C., Dajoz I., Leuret C., Henry J.
P., Dreuillaux J. M., Heberle-Bors E., Till-
Bottraud I. (1994) Pollen aperture polymor-
phism in the angiosperms: importance, possible
causes and consequences. Acta Bot. Gallica 141:
109–122.
Morrell P. L., Rieseberg L. H. (1998) Molecular
tests of the proposed diploid hybrid origin of
Gilia achilleifolia (Polemoniaceae). Amer. J. Bot.
85(10): 1439–1453.
Nadot S., Ballard H. E., Creach J. B., Dajoz I.
(2000) The evolution of pollen heteromorphism
168 R. Yockteng et al.: Phylogeny of pansies using molecular markers
in Viola: a phylogenetic approach. Plant Syst.
Evol. 223: 155–171.
Neuffer B., Auge H., Mesch H., Amarell U., Brandl
R. (1999) Spread violets in pollut ed pine forest:
morphological and molecular evidence for the
ecological importance of interspecific hybridiza-
tion. Molec. Ecol. 8: 365–377.
Posada D. (2001) ModelTest. Department of
Zoology Brigham Young University, Provo
USA.
Rieseberg L. H. (1991) Homoploid reticulate evo-
lution in Helianthus (Asteraceae): Evidence from
ribosomal genes. Amer. J. Bot. 78(9): 1218–1237.
Rohlf J. (1999) NTSYS-pc version 1.80. Applied
Biostatistics.
Salimath S. S., deOliveira A. C., Godwin I. D.,
Bennetzen J. L. (1995) Assessment of genome
origins and genetic diversity in the genus Eleusine
with DNA markers. Genome 38(4): 757–763.
Sang T., Crawford D. J., Stuessy T. F. (1995)
Documentation of reticulate evolution in paeon-
ies (Paeonia) using internal transcribed spacer
sequences of nuclear ribosomal DNA – Impli-
cations for biogeography and concerted evolu-
tion. Proc. Natl. Acad. Sci. USA 92(15): 6813–
6817.
Sang T., Zhang D. (1999) Reconstructing hybrid
speciation using sequences of low copy nuclear
genes: hybrid origins of five Paeonia species
based on Adh gene phylogenies. Syst. Bot. 24(2):
148–163.
Schmidt A. (1962) Eine neue Grundzahl in der
Gattung Viola (Zytotaxonomische Unter such-
ungen an Viola parvula Tin. und Viola occulta
Lehm.). Ber . Deutsch. Bot. Ges. 75: 78–84.
Shishkin B. K. (1949) Flora of the USSR. Vol. XV.
Malvales, Parietales, Myrtiflorae, Academy of
Science of the USSR.
Small R. L., Ryburn J. A., Cronn R. C., Seelanan
T., Wendel J. F. (1998) The tortoise and the
hare: choosing between noncoding plastome and
nuclear Adh sequences for phylogeny reconstruc-
tion in a recently diverged plant group. Amer. J.
Bot. 85(9): 1301–1315.
Stebbins G. L., Harvey B. L., Cox E. L., Rutger J.
N., Jelencovic G., Yagil E. (1963) Identification
of the ancestry of an amphidiploid Viola with the
aid of paper chromatography. Amer. J. Bot.
50(8): 830–839.
Strid A. (1996) Mountain flora of Greece vol 1.
Cambridge, Cambridge UP.
Swofford D. L. (2001) PAUP*. Phylogenetic Anal-
ysis Using Parsimony and other methods. Sinau-
er Assoc., Sunderland, MA.
Thompson J. D., Gibson T. J., Plewnoak F.,
Jeanmougin F., Higgins D. G. (1997) The
ClustalX windows interface: flexible strategies
for multiple sequence alignment aided by quality
analysis tools. Nucleic Acids Res. 24: 4876–4882.
Torrell M., Garcia-Jacas N., Alfonso S., Valle
`
sJ.
(1999) Phylogeny of Artemisia (Asteraceae, Anthe-
mideae) inferred from nuclear ribosomal DNA
(ITS) sequences. Taxon 48: 721–736.
Valentine D. H., Merxmu
¨
ller H., Schmidt A. (1968)
Viola L. In: Tutin T. G., Heywood V. H., Burges
N. A., Moore D. M., Valentine D. H., W alters S.
M., Webb D. A. (eds.) Flora Europaea, Vol 2.
(Rosaceae to Umbelliferae). Cambridge Univer-
sity Press, Cambridge.
Wang X., Tsumura Y., Yoshimaru H., Nagasaka
K., Szmidt A. E. (1999) Phylogenetic relation-
ships of Eurasian pines (Pinus, Pinaceae) based
on chloroplast rbcL, matK, rpl20-rps18 spacer,
and trnV intron sequences. Amer. J. Bot. 86(12):
1742–1753.
Wiley E. O. (1981) Phylogenetics: The theory and
practice of phylogenetic systematics. New York,
Wiley.
Wolfe A., Randle C. P. (2001) Relationships within
and among species of the holoparasitic genus
Hyobanche (Orobanchaceae) inferred from ISSR
banding patterns and nucleotide sequences. Syst.
Bot. 26(1): 120–130.
Wolfe A. D., Xiang Q., Kephart S. R. (1998)
Assessing hybridization in natural populations
of Penstemon (Scrophulariaceae) using hyper-
variable intersimple sequence repeat (ISSR)
bands. Mol. Ecol. 7: 1107–1125.
Young N. D., Healy J. (2002) Gap Coder: A
computer program for including indels in phy-
logenetic analysis. http://www.trinity.edu/
nyoung/GapCoder, Trinity University Biology
Dept.
Yuan Y. (1996) Infrageneric phylogeny of the
genus Gentiana (Gentianaceae) inferred from
nucleotide sequences of the internal transcribed
spacers (ITS) of nuclear ribosomal DNA. Amer.
J. Bot. 83(5): 641–652.
Yuan Y., Ku
¨
pfer P. (1997) The monophyly and
rapid evolution of Gentiana sect. Chondrophyllae
Bunge s.l. (Gentianaceae) Evidence from the
nucleotide sequences of the internal transcribed
R. Yockteng et al.: Phylogeny of pansies using molecular markers 169
spacers of nuclear ribosomal DNA. Bot. J. Linn.
Soc. 123: 25–43.
Addresses of the authors: Roxana Yockteng,
(e-mail: Roxana.Yockteng@ese.u-psud.fr) Sophie
Nadot, Laboratoire Ecologie, Syste
´
matique et
Evolution, Universite
´
Paris-Sud, Baˆ timent 360,
F-91405 Orsay Cedex, France. Harvey E. Ballard
Jr, Department of Environmental and Plant Biol-
ogy, Porter Hall, Ohio University, Athens, OH
45701, USA. Guilhem Mansion, Institute of Bot-
any, University of Neuchaˆ tel, Chantemerle 22,
Neuchaˆ tel 2007. Isabelle Dajoz, Laboratoire Ecol-
ogie, Ecole Normale Supe
´
rieure, Paris, France.
170 R. Yockteng et al.: Phylogeny of pansies using molecular markers