TAXON 68 (2) • April 2019: 340–369
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
SYSTEMATICS AND PHYLOGENY
Genetic and morphological data reveal new insights into the
taxonomy of Campanula versicolor s.l. (Campanulaceae)
Ivana Janković,1
Zlatko Satovic,2,4
Zlatko Liber,3,4
Nevena Kuzmanović,1
3,4
6
Ivan Radosavljević,
Zoran Nikolov & Dmitar Lakušić1
Romeo Di Pietro,5
1 Institute of Botany and Botanical Garden, Faculty of Biology, University of Belgrade, Takovska 43, 11 000 Belgrade, Serbia
2 Department of Seed Science and Technology, Faculty of Agriculture, University of Zagreb, Svetošimunska 25, 10 000 Zagreb, Croatia
3 Division of Botany, Department of Biology, Faculty of Science, University of Zagreb, Marulićev trg 9A, 10 000 Zagreb, Croatia
4 Centre of Excellence for Biodiversity and Molecular Plant Breeding (CroP-BioDiv), Svetošimunska 25, 10 000 Zagreb, Croatia
5 Department PDTA (Section Environment and Landscape), “Sapienza” University of Rome, Via Flaminia 72, 00196 Rome, Italy
6 Macedonian Museum of Natural History, Boulevard Ilinden 86, 91000 Skopje, Republic of North Macedonia
Address for correspondence: Ivana Janković, ijankovic@bio.bg.ac.rs
DOI https://doi.org/10.1002/tax.12050
Abstract Campanula versicolor is a constituent of the Campanula pyramidalis complex, distributed in the central and southern Balkan
Peninsula with a small disjunct range in SE Italy. The taxonomy of the broadly defined C. versicolor was controversial in the past due to
high morphological variability of the populations across the range. We used microsatellite DNA data along with morphometric analyses
on an extensive population sampling covering the entire range to reconstruct the relationships among populations of the intricate
C. versicolor s.l. Based on the results three distinct entities can be distinguished, circumscribed here on the subspecies level, with newly
established combinations: C. versicolor subsp. versicolor, C. versicolor subsp. korabensis and C. versicolor subsp. tenorei. We propose
a new taxonomic treatment of C. versicolor s.l. with description, distribution and habitat data for each of the subspecies and a key for
their identification.
Keywords Apennine Peninsula; Balkan Peninsula; infraspecific variation; microsatellites; morphometrics; taxonomy
Supporting Information may be found online in the Supporting Information section at the end of the article.
■ INTRODUCTION
The genus Campanula L. (Linnaeus, 1753: 163) includes
350–420 (Kovačić, 2004; Lammers, 2007) or even up to
580–600 species when derived genera currently recognized as
distinct based on selected morphological characters are also included in it (Mansion & al., 2012). It is one of the most important genera that contribute to the floristic diversity of the
Holarctic Kingdom (Jones & al., 2017). The center of species
diversity of Campanula is considered to be located in the Mediterranean region (Park & al., 2006) where 250 species are recorded (Geslot, 1984). The largest number of endemic taxa is
found in the eastern Mediterranean (Phitos, 1964, 1965),
circum-Adriatic and west Balkan regions (Kovačić, 2004),
and the Caucasus (Gagnidze, 2005).
The Campanula pyramidalis complex is a group of morphologically specific “isophylloid” bellflowers (Damboldt,
1965; Kovačić & Nikolić, 2006; Park & al., 2006; Lakušić &
al., 2013). Previous studies have shown that the C. pyramidalis
complex is most closely related to the C. waldsteiniana aggregate (Kovačić & Nikolić, 2006; Park & al., 2006; Liber & al.,
2008; Mansion & al., 2012; Bogdanović & al., 2014, 2015;
Crowl & al., 2016), together representing the “core” of the
isophylloid bellflowers (Park & al., 2006). This complex,
along with C. pyramidalis L. (Linnaeus, 1753: 164), comprises
C. austroadriatica Kovačić & D.Lakušić (Lakušić & al., 2013:
519), C. montenegrina I.Janković & D.Lakušić (Janković & al.,
2016: 77), C. secundiflora Vis. & Pančić (Visiani & Pančić,
1862: 20) and C. versicolor Andrews (1804: 396). The distribution of the complex stretches from the Gulf of Trieste (Italy) in
the northwest to the Peloponnese Peninsula (Greece) in the
south and Mt. Konjevska (Bulgaria) in the east, with a small
disjunction in southern Italy (Lakušić & al., 2013). Based on
phylogenetic analyses of combined plastid and nuclear sequence data (Lakušić & al., 2013), as well as microsatellite
markers (Janković & al., 2016), it has been shown that
C. versicolor s.l. represents a well-identifiable group characterized by the largest distribution range if compared to that of the
other species of the C. pyramidalis complex. It is distributed in
the central and southern parts of the Balkan Peninsula—southern Serbia, southwestern Bulgaria, Republic of North
Macedonia, Albania, Greece—and in the southeastern part of
the Apennine Peninsula—Apulia and Basilicata administrative
regions of Italy (Janković & al., 2017). The southernmost
known occurrence of C. versicolor is located on the island of
Kythira (Strid & Tan, 1996; Yannitsaros, 2004) which lies
Article history: Received: 28 May 2018 | returned for (first) revision: 2 Aug 2018 | (last) revision received: 15 Jan 2019 | accepted: 15 Jan 2019
Associate Editor: Alessia Guggisberg | © 2019 International Association for Plant Taxonomy
340
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
TAXON 68 (2) • April 2019: 340–369
between the Peloponnese and Crete in the South Aegean Sea
(Strid & Tan, 1996), while the northernmost is at Mt.
Konjevska in the Znepole region in Bulgaria (Ančev, 1984;
Asenov, 2009). Campanula versicolor mainly grows in limestone rock crevices, rarely in rocky places in meadows, from
lowland to 2100 m (Hartvig, 1991). It also occurs in rocky
places and walls in cities and villages (Blionis & Vokou, 2005).
Reflecting the morphological variability observed in this
group, 15 taxa at the specific and infraspecific levels were described over the years in the broadly defined C. versicolor
(Lammers, 2007) (Table 1). However, their taxonomic status
and morphological and geographical peculiarities were not
clear, to the point that the modern floristic and taxonomic literature does not recognize any of these taxa but consider them as
synonyms of C. versicolor (Pignatti 1982; Hartvig, 1991;
Lammers, 2007; Castroviejo & al., 2010; Govaerts &
Lammers, 2018).
The results of the phylogenetic study of Lakušić & al.
(2013) suggested a geographic structure in north–south direction within C. versicolor s.l. populations. The authors observed
a differentiation between northern (Bulgaria, southern Serbia,
North Macedonia) and southern populations (Greece, southern
Italy) based on analyses of chloroplast DNA (cpDNA) data.
On the other hand, based on the nuclear ribosomal internal
transcribed spacer (nrITS) data, they found a weakly supported
west–east subdivision: circum-Ionian and western Greece populations, and the remainder of the populations from eastern
Greece, North Macedonia, Serbia, and Bulgaria. However,
their population sampling was not extensive (14 populations),
calling for further analyses based on a broader sampling
(Lakušić & al., 2013). Therefore, for this study we increased
Table 1. Investigated Campanula taxa and informal names assigned to
the populations from the loci classici.
Taxon
Informal name
1.
C. korabensis subsp. bicajensis
bicajensis
2
C. korabensis
korabensis
3.
C. longipetiolata
longipetiolata
4.
C. versicolor subsp. thessala
subvar. lancifolia
lancifolia
5.
C. mrkvickana
≡ C. versicolor f. mrkvickana
mrkvickana
6.
C. plasonii
plasonii
7.
C. corymbosa
≡ C. tenorei
≡ C. rosanoi
≡ C. versicolor var. rosanii
tenorei
8.
C. versicolor var. thessala
thessala
9.
C. versicolor var. tomentella
≡ C. versicolor var. thessala f. tomentella
tomentella
10.
C. versicolor*
versicolor
*All other investigated populations are marked as versicolor. For
details about homotypic synonyms cited in the table see Janković
& al. (2017).
population sampling from 14 to 41 populations, in order to allow for further taxon segregation within C. versicolor s.l. Considering all this, the main aim of the present study was to get
insight into the genetic diversity and morphological variability
of C. versicolor s.l. across the entire range of the group based
on extensive population sampling.
■ MATERIALS AND METHODS
Plant material. — Material for molecular and morphometric study was sampled from the entire distribution area of
C. versicolor. Additionally, recently described taxa from
Albania—C. korabensis F.K.Mey. (Meyer, 2011: 149),
C. korabensis subsp. bicajensis F.K.Mey. (Meyer, 2011: 150)
and C. longipetiolata F.K.Mey. (Meyer, 2011: 151)—were
included as well, based on the protologues, type materials
and personal observations.
For molecular analysis, leaves were collected from 8 to 22
individuals per population yielding a total of 667 individuals
from 41 populations (Appendix 1; Fig. 1). For morphometric
study, flowering stems, flowers and leaves were collected at
full anthesis from 10 to 17 individuals per population yielding
a total of 456 individuals from 37 populations (Appendix 1).
Sampled material was preserved in glycerol–ethanol (1:1)
solution.
Voucher specimens for each investigated population were
deposited at the Herbarium of the Institute of Botany and Botanical Garden of the Faculty of Biology, University of
Belgrade (BEOU), at the Herbarium of the Natural History
Museum Rijeka (NHMR) and at the Herbarium Croaticum of
the University of Zagreb (ZA). Details regarding sampled populations are given in Appendix 1. Furthermore, additional material was examined to define the distribution ranges of the
investigated taxa (see the Appendix 2).
The Articles of the International Code of Nomenclature
for algae, fungi, and plants (ICN) cited in the text follow
Turland & al. (2018), while the quoted syntaxa follow Mucina
& al. (2016). For an easier interpretation of the results and discussion, informal names, without a taxonomic rank, were
assigned to the investigated taxa collected from the locus
classicus. Those informal names represent the specific or
infraspecific epithets of the investigated taxa (Table 1). To all
other investigated populations of C. versicolor s.l. we assigned
the informal name “versicolor”. The population of
C. longipetiolata was included only in the molecular analysis.
Campanula planiflora Willd. (Willdenow, 1809: 210) (≡
C. willdenowiana Schult.) was not included in this study since
the locus classicus is not specified in the protologue (Janković
& al., 2017).
Chorological data. — Chorological data (Appendix 2)
were collected during field investigations (personal observations) as well as from relevant herbarium collections.
Material at BEO, BEOU, BP, HMMNH, SO, SOA and SOM
herbaria has been revised and examined, as well as digital
images available via online databases or images of specimens
341
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
sent by the curators from AMD, B, BM, BP, BRNM, C, CAT,
CGE, E, FI, G (G-DC), GE, HAL, HFLA, HMMNH, JE, K,
L, LD, LE, MANCH, MNHN, MPU, NAP, NYC, OXF, P,
PAD, PAL, PH, PI, PRC, RO, TCD, TO, TOU, U, W and WAG.
Microsatellite analysis. — Genomic DNA samples were
extracted from silica gel-preserved leaf tissue using the
GenElute Plant Genomic DNA Miniprep Kit (Sigma-Aldrich,
St. Louis, Missouri, U.S.A.) or NucleoSpin Plant II
(Macherey-Nagel, Düren, Germany). The molecular analysis
was carried out on seven SSR loci (CpUZ001, CpUZ003,
CpUZ004, CpUZ005, CpUZ006, CpUZ009, CpUZ010) developed and characterized by Radosavljević & al. (2015), which
had shown cross-amplification in C. versicolor. A two-step polymerase chain reaction (PCR) protocol with an initial touchdown cycle (Radosavljević & al., 2015) was employed for
amplification of the microsatellite loci. The amplified fragments
were run on an ABI 3730XL analyser (Applied Biosystems,
Foster City, California, U.S.A.) and the results were analyzed
with GeneMapper v.4.0 (Applied Biosystems, 2005).
Genetic diversity and bottlenecks. — Polymorphism
information content (PIC) of each microsatellite marker was
calculated as described in Botstein & al. (1980). GENEPOP
v.4.0 (Raymond & Rousset, 1995) was used to estimate
population genetic parameters (average number of alleles per
locus Nav; observed heterozygosity HO; expected heterozygosity HE; inbreeding coefficient FIS) and to test for
deviation from Hardy-Weinberg equilibrium (HWE).
TAXON 68 (2) • April 2019: 340–369
Sequential Bonferroni adjustments (Holm, 1979; Rice,
1989) were applied for correction of the effect of multiple
tests with SAS release 9.1 (SAS Institute, 2004). MicroChecker v.2.2.3 (Van Oosterhout & al., 2004) was used to
check for the presence of null alleles and potential problems
related to allele dropout. Frequencies of null alleles were estimated using FreeNA (Chapuis & Estoup, 2007) based on the
expectation-maximization algorithm (Dempster & al., 1977).
FSTAT v.2.9.3.2 (Goudet, 1995, 2002) was used to estimate
allelic richness (Nar), while private allelic richness (Npar)
was calculated in HP-Rare v.1.0 (Kalinowski, 2004, 2005).
BOTTLENECK v.1.2.02 (Cornuet & Luikart, 1996; Piry
& al., 1999) was used to test for evidence of recent bottleneck
events on the basis of the theoretical expectation. The observed gene diversity (HO) was compared with the gene diversity expected at mutation–drift equilibrium (HEQ) and
calculated from the observed number of alleles under different
mutation models: infinite allele model (IAM), stepwise mutation model (SMM) and an intermediate two-phase model
(TPM). The TPM was applied assuming 30% multistep
changes and a variance of 30 (Pascual & al., 2001). Based
on the number of loci in our dataset, the Wilcoxon signedrank test (Luikart & al., 1998) was chosen for the statistical
analysis of heterozygote excess or deficiency as recommended by Piry & al. (1999).
Genetic differentiation and structure. — Genetic
differentiation between all pairs of populations was measured
Fig. 1. Location of sampled Campanula versicolor s.l. populations. Symbols represent informal names assigned to the
investigated taxa. Population numbering and informal names correspond to Appendix 1.
342
TAXON 68 (2) • April 2019: 340–369
with pairwise fixation indices (FST) estimates. Pairwise FST
and their respective P-values for significant differences from
zero were calculated in FSTAT. Pairwise Cavalli-Sforza and
Edwards’s chord distances (Cavalli-Sforza & Edwards,
1967) were calculated, and an unrooted tree was constructed
using the Fitch-Margoliash algorithm. To obtain bootstrap
support values for branches, cluster analysis was performed
with 1000 bootstraps (Felsenstein, 1985) over microsatellite
loci as implemented in SEQBOOT, GENDIST, FITCH and
CONSENSE programs in the PHYLIP v.3.6b software
package (Felsenstein, 1993). Analysis of molecular variance
(AMOVA; Excoffier & al., 1992) was performed by means
of Arlequin v.3.0 (Excoffier & al., 2005) to partition the
total microsatellite diversity among and within populations,
among groups identified with STRUCTURE analysis,
among populations within groups and within populations
within groups. The variance components were tested
statistically by non-parametric randomization tests with
10,000 permutations. A model-based clustering method was
applied to infer genetic structure and to define the number
of clusters using STRUCTURE v.2.3.3 (Pritchard & al.,
2000). Thirty runs per cluster (K; from K = 1 to K = 21) were
carried out on the Isabella computer cluster at the University
of Zagreb, University Computing Centre (SRCE). Each run
consisted of a burn-in period of 200,000 steps, followed by
1 million Monte Carlo Markov chain (MCMC) replicates,
assuming an admixture model and correlated allele
frequencies. The choice of the most likely number of
clusters was carried out by comparing the average estimates
of the likelihood of the data, ln (Pr(X |K)) for each value of
K (Pritchard & al., 2000) and calculating the statistic ΔK
based on the rate of change in the log probability of the data
between successive K values, as described by Evanno & al.
(2005) using STRUCTURE HARVESTER v.0.6.92 (Earl &
Holdt, 2012). Runs were clustered and averaged in
CLUMPAK (Kopelman & al., 2015). Isolation by distance
(IBD) among populations was tested by the method of
Rousset (1997). A Mantel test (106 permutations of
population locations among all locations) on the matrix of
pairwise FST / (1 – FST) ratios and on the natural logarithm of
geographical distances (in km) between pairs of populations
was performed using NTSYS-pc v.2.02 (Rohlf, 1997).
Morphometric data analysis. — Plant material for
morphometric analyses was prepared as described in
Janković & al. (2016). In total, 50 quantitative morphological
characters (23 floral characters; height of the plant,
inflorescence length and stem height; 8 characters each of
basal, middle and upper leaves) of 456 individuals from 37
populations were analyzed (Appendix 1, 3). All characters
were measured with Digimizer Image Analysis 4.0.0.0
(MedCalc Software, 2005). Statistical analyses were
performed using Statistica v.5.1 (StatSoft, 1996).
Descriptive statistics (mean, standard deviation, minimum,
maximum and coefficient of variation) were calculated for
each quantitative character. The number of characters included
in further multivariate analyses was reduced according to the
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
results of pairwise Spearman correlations, retaining only one
out of character pairs with absolute values of correlation coefficients exceeding 0.9.
To get insight into the relationships between the analyzed
populations, we performed an UPGMA (unweighted pair
group method with arithmetic mean) clustering analysis based
on Mahalanobis distances calculated with discriminant analysis
defining 37 populations as a priori groups. Thereafter, canonical discriminant analysis (CDA) was used to test the hypothesis of morphological segregation of the populations a priori
classified into three groups according to the results from the
STRUCTURE analysis for K = 3. Canonical scores for each
case (individual) were calculated with the aim to measure the
distances between the individuals, and a scatterplot of canonical scores was made to visualize the relationships between the
a priori defined groups. The percentage of correctly classified
cases in each group was calculated with the classification function. Finally, a discriminant function analysis (DFA) was performed to estimate the contribution of individual characters
to the overall discrimination.
■ RESULTS
Intrapopulation genetic diversity. — For this study,
seven microsatellite markers were selected (Table 2). One
hundred and twenty-one alleles were detected, a mean of 17
alleles per locus. The number of alleles per locus ranged
from 9 at CpUZ001 to 33 at CpUZ003. Six of seven
microsatellites showed a high PIC >0.7 (Table 2).
Generally, the observed heterozygosities (HO) showed
slightly lower values than the expected heterozygosities
(HE), while the expected heterozygosities were high for most
populations, except for the mrkvickana population 1 BGZemen (HE = 0.091). This population had the lowest values
among all populations for all the genetic variability estimates
(Table 3). Multilocus estimates of Wright’s inbreeding coefficient within populations (FIS) showed no significant deviation from zero (P > 0.05) in 32 of 41 populations, suggesting
that HWE was met for each population. A significant deficit
Table 2. Allelic diversity of seven microsatellite loci scored in 41
populations of Campanula versicolor s.l.
No.
Locus
Repeat motif
Size range (bp)
Na
PIC
1
CpUZ001
ACT
148–172
9
0.761
2
CpUZ003
GA
184–296
33
0.937
3
CpUZ004
AGA
133–232
20
0.752
4
CpUZ005
GT
127–163
13
0.385
5
CpUZ006
GT
212–246
15
0.778
6
CpUZ009
GA
175–217
16
0.814
7
CpUZ010
ATC
154–196
15
0.796
Average
17
Na = total number of alleles per locus; PIC = polymorphism information content.
343
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
TAXON 68 (2) • April 2019: 340–369
Table 3. Genetic diversity and sample size of 41 populations of Campanula versicolor s.l.
No.
Pop. code
n
Nav
Nar
Npr
Npar
HO
HE
FIS
1
BG-Zemen a
10
1.429
1.254
0
0.000
0.086
0.091
0.061 ns
3
RS-Preševo
20
3.000
2.125
1
0.154
0.279
0.387
0.278***
4
RS-Gotovuša
10
3.714
2.898
0
0.044
0.456
0.564
0.192 ns
5
RS-Dušanov grad
11
3.286
2.362
0
0.002
0.293
0.409
0.283*
7
AL-Mt. Gjallica
8
3.429
2.891
0
0.071
0.527
0.540
0.024 ns
8
AL-Shija Gorge b
20
4.429
2.694
0
0.064
0.346
0.439
0.212**
9
AL-Mt. Korab c
21
3.714
2.700
0
0.079
0.462
0.508
0.09 ns
10
AL-Kurbnesh
23
3.714
2.492
1
0.159
0.478
0.475
0.008 ns
11
MK-Ostruga
13
2.429
2.000
0
0.000
0.418
0.382
0.092 ns
12
MK-Matka Canyon d
20
6.429
3.631
2
0.146
0.614
0.675
0.09*
14
MK-Katlanovo
21
3.429
2.498
0
0.002
0.442
0.481
0.081 ns
15
MK-Veles
13
3.714
2.726
0
0.109
0.539
0.496
0.085 ns
16
MK-Kozjak Lake
9
4.429
3.461
0
0.168
0.619
0.642
0.036 ns
18
MK-Radika Gorge
17
4.143
2.840
0
0.005
0.475
0.536
0.115 ns
19
MK-Mt. Bistra
9
3.000
2.443
0
0.000
0.404
0.450
0.104 ns
21
MK-Mt. Ilinska pl.
12
2.000
1.754
0
0.001
0.262
0.316
0.171 ns
23
MK-Ohrid Lake
14
2.571
2.154
0
0.000
0.429
0.387
0.108 ns
25
MK-Mt. Galičica
19
4.714
2.890
1
0.132
0.368
0.510
0.278***
26
MK-Demir Kapija e
12
2.857
2.153
0
0.000
0.298
0.345
0.137 ns
28
GR-Oros Askion
10
3.286
2.575
1
0.165
0.357
0.484
0.262 ns
29
GR-Mt. Olympus f
9
4.000
3.319
0
0.095
0.603
0.692
0.128 ns
30
GR-Prionia g
20
4.143
2.838
0
0.004
0.561
0.548
0.023 ns
31
GR-Tempi Gorge
14
4.143
3.183
1
0.157
0.642
0.637
0.008 ns
33
GR-Mt. Giona
12
5.571
3.808
2
0.202
0.695
0.711
0.022 ns
34
GR-Mt. Parnassus
17
5.429
3.589
0
0.092
0.542
0.660
0.178**
35
GR-Alonistaina
20
5.571
3.606
0
0.090
0.555
0.675
0.179***
37
GR-Mt. Mavrovouni
18
6.429
3.976
2
0.259
0.594
0.719
0.174***
38
GR-Lefkada
15
5.714
3.831
0
0.040
0.673
0.730
0.078 ns
39
GR-Mt. Tzoumerka
22
6.286
3.386
0
0.079
0.571
0.611
0.065 ns
42
GR-Vikos Gorge
20
5.286
3.186
1
0.126
0.573
0.609
0.059 ns
43
GR-Mt. Tymfi
20
4.429
2.812
1
0.116
0.515
0.526
0.022 ns
44
GR-Corfu
16
6.143
3.951
2
0.298
0.607
0.734
0.173*
45
AL-Bistrica
17
5.000
3.205
1
0.193
0.597
0.634
0.059 ns
46
AL-Luzat h
20
6.143
3.642
0
0.040
0.586
0.660
0.112 ns
47
AL-Tepelenë
20
5.429
3.442
0
0.078
0.586
0.631
0.072 ns
48
AL-Vlorë
20
5.857
3.532
1
0.115
0.604
0.648
0.067 ns
49
IT-Matera i
19
2.429
1.999
0
0.001
0.252
0.306
0.175 ns
50
IT-Castellaneta i
20
2.857
2.335
0
0.047
0.403
0.403
0.001 ns
51
IT-Punta Palascia i
19
3.429
2.544
0
0.096
0.450
0.528
0.147 ns
52
IT-Alessano i
17
3.000
2.420
0
0.007
0.461
0.455
0.014 ns
55
IT-Gallipoli i
20
2.571
2.068
0
0.000
0.293
0.338
0.133 ns
No. = population numbering; Pop. code = population code based on the two-letter code defined by ISO 3166-1 and population locality; n = sample size; Nav = average number of alleles; Nar = allelic richness; Npr = number of private alleles; Npar = private allelic richness; HO = observed
heterozygosity; HE = expected heterozygosity; FIS = inbreeding coefficient. Significant values are printed in bold (ns = non-significant value;
*= significant at P < 0.05; **= significant at P < 0.01; ***= significant at P < 0.001). Bold-face letters after the population code indicate a population from the locus classicus: a, mrkvickana; b, bicajensis; c, korabensis; d, lancifolia; e, plasonii; f, tomentella; g, thessala; h, longipetiolata; i,
tenorei. Population numbering, codes and details correspond to Appendix 1.
344
TAXON 68 (2) • April 2019: 340–369
of heterozygotes was found in 9 populations, ranging from
FIS = 0.09 (12 MK-Matka Canyon) to FIS = 0.283 (5 RSDušanov grad).
A statistically significant departure (P < 0.05) from
mutation–drift equilibrium was detected in 15 populations (9,
11, 14, 29, 31, 33, 35, 38, 39, 44, 47, 48, 50, 51, 52) when
the IAM was applied to the analyses. In contrast, there was
no significant departure when the SMM was applied. Finally,
according to the TPM, considered as the most appropriate
model given the empirical evidence (Di Renzo & al., 1994),
significant departures from mutation–drift equilibrium (P <
0.05) were detected in four populations (29, 31, 50, 52).
One-tailed probability values of the Wilcoxon signed-rank test
for heterozygosity excess based on the three mutation models
(IAM, TPM, SMM) are presented in Table 4.
Interpopulation genetic diversity. — Genetic differentiation between pairs of populations was low, with FST
values ranging from 0.042 to 0.736 (Appendix 4). The
highest FST values were observed between the Apennine
populations (49–52, 55) and all the other populations,
especially central Balkan ones (1, 3–5, 7–12, 14–16, 18,
19, 21, 23, 25, 26). High FST values were also found
between the mrkvickana population (1 BG-Zemen) and all
the other populations.
The unrooted Fitch-Margoliash tree (Fig. 2) revealed two
main branches which illustrate a clear genetic differentiation
between the populations from the northern part of the range of
C. versicolor s.l. (Bulgaria 1; Serbia 3–5; northeastern
Albania 7–10; North Macedonia 11, 12, 16, 18, 19, 21, 23,
25) and those from the southern part (Greece 28–31, 33–35,
37–39, 42–44; southern Albania 45–48; Italy 49–52, 55). The
populations from the northern part of the range of C. versicolor
s.l. are distributed in the central part of the Balkan Peninsula,
while the populations from the southern part of the range are
distributed in the southern part of the Balkan Peninsula and
the southeastern part of the Apennine Peninsula. The branch
of the Fitch-Margoliash tree, where the central Balkan populations were positioned, was marked as “versicolor 1”. Within
this branch were also nested the populations informally named
bicajensis, korabensis, lancifolia and mrkvickana. The other
branch was marked as “versicolor 2” and included the southern
Balkan populations and the longipetiolata, thessala and
tomentella populations along with the Apennine tenorei populations (Fig. 2). The exception were three central Balkan populations from central (14 MK-Katlanovo, 15 MK-Veles) and
southeastern North Macedonia (plasonii, 26 MK-Demir
Kapija), which were positioned within the southern Balkan
populations (“versicolor 2”), but close to “versicolor 1” (Fig. 2).
The results of the STRUCTURE analysis showed that the
highest ΔK was obtained at K = 2 (ΔK = 1509.42) followed by
that at K = 3 (ΔK = 124.67). Being aware that the maximal ΔK
at K = 2 might appear to be an artefact resulting from markedly
low likelihoods for K = 1, as suggested by Vigouroux & al.
(2008) and Janes & al. (2017), we also provide the results for
K = 3. At Ks larger than 3, the variances increased substantially and ΔKs were lower than 5.
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
Table 4. Probability results of the Wilcoxon signed-rank test to assess
population bottlenecks in 41 populations of Campanula versicolor s.l.
Mutation model
No.
Pop. code
IAM
TPM
SMM
1
BG-Zemen
0.875*
0.875
0.875
3
RS-Preševo
0.344
0.422
0.922
0.531
4
RS-Gotovuša
0.406
0.469
5
RS-Dušanov grad
0.719
0.961
0.984
7
AL-Mt. Gjallica
0.289
0.469
0.531
8
AL-Shija Gorge
0.813
0.961
0.988
9
AL-Mt. Korab
0.039
0.078
0.500
10
AL-Kurbnesh
0.289
0.813
0.992
11
MK-Ostruga
0.016
0.078
0.406
12
MK-Matka Canyon
0.188
0.344
0.813
14
MK-Katlanovo
0.023
0.078
0.344
15
MK-Veles
0.219
0.422
0.578
16
MK-Kozjak Lake
0.055
0.344
0.656
18
MK-Radika Gorge
0.188
0.594
0.961
19
MK-Mt. Bistra
0.531
0.711
0.945
21
MK-Mt. Ilinska pl.
0.078
0.078
0.578
23
MK-Ohrid Lake
0.055
0.344
0.578
25
MK-Mt. Galičica
0.281
0.945
0.977
26
MK-Demir Kapija
0.500
0.891
0.953
28
GR-Oros Askion
0.406
0.711
0.766
29
GR-Mt. Olympus
0.012
0.027
0.234
30
GR-Prionia
0.148
0.531
0.813
31
GR-Tempi Gorge
0.008
0.012
0.148
33
GR--Mt. Giona
0.039
0.406
0.766
34
GR-Mt. Parnassus
0.148
0.234
0.656
35
GR-Alonistaina
0.008
0.289
0.766
37
GR-Mt Mavrovouni
0.148
0.234
0.766
38
GR-Lefkada
0.020
0.344
0.469
39
GR-Mt Tzoumerka
0.023
0.656
1.000
42
GR-Vikos Gorge
0.188
0.766
0.973
43
GR-Mt Tymfi
0.531
0.852
0.988
44
GR-Corfu
0.012
0.188
0.852
45
AL-Bistrica
0.188
0.406
0.813
46
AL-Luzat
0.289
0.469
0.766
47
AL-Tepelenë
0.020
0.406
0.945
48
AL-Vlorë
0.020
0.406
0.945
49
IT-Matera
0.063
0.156
0.844
50
IT-Castellaneta
0.016
0.016
0.078
51
IT-Punta Palascia
0.008
0.148
0.188
52
IT-Alessano
0.016
0.023
0.078
55
IT-Gallipoli
0.219
0.578
0.578
Wilcoxon test P-values represent one-tailed probabilities for heterozygosity excess based on three mutation models. Values of P < 0.05 are
indicated in bold. No. = population numbering; Pop. code = population code based on the two-letter code defined by ISO 3166-1 and population locality; IAM = infinite allele model; TPM = two-phase model;
SMM = stepwise mutation model. Population numbering, codes and
details correspond to Appendix 1.
345
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
At K = 2, the central Balkan populations were grouped in
a single cluster A – “versicolor 1”, where bicajensis,
korabensis, lancifolia and mrkvickana were also grouped.
The southern Balkan populations were grouped in cluster B –
“versicolor 2” along with longipetiolata, thessala and
tomentella, and the Apennine tenorei populations. These two
clusters are geographically well defined (Fig. 3A). The results
of the STRUCTURE analysis as a model-based clustering
method are congruent with the results of the distance-based
model (unrooted Fitch-Margoliash tree based on CavalliSforza and Edwards’s chord distance; Fig. 2).
Populations from central (14 MK-Katlanovo, 15 MKVeles) and southeastern (26 MK-Demir Kapija) North
TAXON 68 (2) • April 2019: 340–369
Macedonia were the most admixed. The proportions of
ancestry of each population in each of the two clusters
ranged from 60.1% (14 MK-Katlanovo) to 99.1% (21 MKMt. Ilinska pl. for cluster A and 50 IT-Castellaneta for
cluster B).
At K = 3, cluster B defined for K = 2 split up into clusters
B1 – “versicolor 2” and B2 – “tenorei”, the latter representing
the Apennine tenorei populations (Italy 49–52, 55) and some
southern Balkan versicolor populations (southern Albania
46–48; northwestern Greece 39, 43), as well as two versicolor
populations from central North Macedonia (14, 15) (Figs. 2,
3B). These three clusters were also geographically well differentiated (Fig. 3B).
Fig. 2. Unrooted Fitch-Margoliash tree based on Cavalli-Sforza and Edwards’s chord distances between 41 populations of
Campanula versicolor s.l. Bootstrap support values >50% based on 1000 replicates are shown. Population numbering, locality codes and informal names correspond to Appendix 1, while color coding and symbols correspond to Fig. 1.
346
TAXON 68 (2) • April 2019: 340–369
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
Fig. 3. Geographically presented proportions of assignment of each Campanula versicolor s.l. population (pie chart) to
each of the (A) two clusters and (B) three clusters as defined by the model-based clustering methods of Pritchard & al.
(2000). Population numbering corresponds to Appendix 1.
347
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
AMOVA showed that 31.43% of the total microsatellite
diversity was attributable to among-population differences
(Table 5A). A two-way nested AMOVA (Table 5B) revealed
that the differences between the two clusters identified by
STRUCTURE analysis amounted to 11.79% while the
23.93% of the total variation was attributable to amongpopulation differences.
The FST / (1 FST) ratio for pairs of populations increased linearly with the natural logarithm of the geographical distance (r = 0.229; PMantel < 0.01; R2 = 0.053) (suppl.
Fig. S1).
Morphological variability. — The analysis of morphological variation revealed that the largest number of the
characters showed a moderate degree of variability (CV =
20%–50%). Highly variable characters with a coefficient of
TAXON 68 (2) • April 2019: 340–369
variation higher than 50% were flower pedicel length (CaPe-L),
inflorescence length (Inf-L), petiole length (FbPe-L, FmPe-L,
FuPe-L) and leaf lamina area (Fb-Ar, Fm-Ar, Fu-Ar). Contrary
to them, stable characters with low variability were width of the
corolla lobe base (CoL-W), distance from maximal width of the
lobe to the corolla base (CoL-h) and stamen base perimeter
(StB-Per) (Appendix 3).
UPGMA clustering identified two clusters of populations:
one with central Balkan populations from Bulgaria, Serbia,
North Macedonia and northeastern Albania and a second one
with southern Balkan populations from Greece and southern
Albania and Apennine populations from southeastern Italy
(Fig. 4). Within the second cluster, tenorei populations were
grouped in subcluster B2 – “tenorei” (Fig. 4), while southern
Balkan populations formed subcluster B1 – “versicolor 2”
Table 5. Analysis of molecular variance (AMOVA) for the partitioning of microsatellite diversity (A) among and within 41 populations of
Campanula versicolor s.l. and (B) between clusters A and B obtained by STRUCTURE analysis, among populations within clusters and within
populations.
Source of variation
A
B
d.f.
Variance components
% Total variation
Among populations
40
0.831
31.43
Within populations
1293
1.814
68.57
ϕ ST
P (ϕ ST)
0.314
<0.0001
Between clusters
1
0.333
11.79
0.118
<0.0001
Among populations within clusters
39
0.675
23.93
0.271
<0.0001
Within populations
1239
1.814
64.29
0.357
<0.0001
P (ϕ ST), ϕ statistics probability level after 10,000 permutations.
Fig. 4. UPGMA dendrogram for populations of C. versicolor s.l. Population numbering and locality acronyms correspond to Appendix 1.
Population markers and lines color were assigned to populations based on the genetic clusters obtained by STRUCTURE analysis for K = 3.
348
TAXON 68 (2) • April 2019: 340–369
(Fig. 4). The populations informally named mrkvickana,
lancifolia, bicajensis and korabensis as well as plasonii,
thessala and tomentella were positioned within the central
Balkan cluster (A – “versicolor 1”, Fig. 4).
With CDA we tested the morphological segregation of
populations which were a priori classified into three groups according to the STRUCTURE analysis of the molecular data
(for K = 3). The classification function showed that, when
using this criterion, quite some individuals were not correctly
classified, resulting in a total percentage of correctly classified
individuals of 80%. In case of the “versicolor 2” and “tenorei”
groups the percentage of correctly classified individuals was
low (74.7% and 71.4%, respectively). Namely, according to
the STRUCTURE analysis for K = 3, the Apennine tenorei
populations were related with a few southern Balkan populations (“versicolor 2”) from southern Albania and northwestern
Greece as well as two populations from central North
Macedonia (Figs. 2, 3). However, according to the results of
the UPGMA clustering analysis (Fig. 4), the Apennine tenorei
populations differ morphologically from the Balkan populations. Therefore, we performed an additional CDA, with the
exception of a priori classification of tenorei populations according to the STRUCTURE analysis for K = 3. Namely,
tenorei populations were kept alone in the third separate group.
With this classification, the total percentage of correctly classified individuals was 97.4%, while in case of the “versicolor 2”
and “tenorei” groups it was 96.5% and 94.4%, respectively.
The result of the CDA showed that all three groups were
well differentiated and positioned in distinct parts of the
DA1–DA2 space (Fig. 5). Finally, DFA showed that morphological characters that contributed most to discrimination were
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
petiole width and length of middle leaf (FmPe-W, FmPe-L),
stamen anther length (StA-L), inflorescence length (Inf-L)
and calyx height (Ca-H). The results of descriptive statistics
for all measured morphometric characters included in DFA
are shown in Table 6.
■ DISCUSSION
The results of our study showed that the investigated populations have abundant allelic variation over seven investigated
loci and high overall genetic diversity. Substantially lower
values of genetic parameters of the mrkvickana population in
relation to all the other investigated populations (Table 3) are
probably a consequence of restricted gene flow due to geographic isolation of this population. Between the mrkvickana
population in southwestern Bulgaria (1 BG-Zemen) and the
closest populations in southern Serbia and the northern part of
North Macedonia, there is an area about 80 km wide, 100 km
long and with altitudes up to about 1000 m where no other populations of C. versicolor occur. In addition, the FST values between the mrkvickana population (1 BG-Zemen) and all the
other populations are high and mostly statistically significant.
The deviation from HWE detected in nine populations
(Table 3) can be related to geographic barriers which reduce
gene flow, since those populations are narrowly distributed
in canyons (3, 5, 8, 12, 35), on islands (44) or high mountains (25, 34, 37). Most of those populations are small and
restricted so the inbreeding rate is higher than in the more
numerous populations. The deviation from HWE can also
be related to the infections of ovaries with larvae of the weevil
Fig. 5. Canonical discriminant analysis (CDA)
for populations of C. versicolor s.l. a priori
classified into three groups: versicolor 1,
versicolor 2, and tenorei.
349
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
Miarus abnormis Solari (1947: 75) after which the number of
flowers and the production of pollen are lower (25). The cause
of recent bottleneck signs detected in four populations (29, 31,
50, 52) may be anthropogenic since these populations grow on
rocks along roads or on walls in cities.
The results of the AMOVA showed a clear differentiation
between clusters of populations and populations within clusters. This might be a result of the populations’ isolation
and/or restricted gene flow. However, relatively high levels
of intrapopulation variability are maintained by the presence
of allogamy (Table 5).
The statistically significant IBD pattern indicates that only
5.3% of the genetic differentiation between populations can be
explained by geographical distance, although FST values are
generally the highest (and statistically significant) between
populations that are geographically the most distant from
each other, i.e., between the Apennine populations (49–52,
55) and all other populations, as well as between the population from Bulgaria (1) and all other populations (Fig. 1;
Appendix 4).
The results of the unrooted Fitch-Margoliash tree (Fig. 2)
and Bayesian model-based clustering analysis using STRUCTURE at K = 2 (Fig. 3A) revealed two groups/clusters of
populations having a geographic structure in north–south direction. This differentiation between central Balkan (southwestern Bulgaria, southern Serbia, northeastern Albania,
North Macedonia) and southern Balkan and Apennine populations (Greece, southern Albania, Italy) of C. versicolor was
also suggested by the cpDNA tree obtained in the molecular
phylogenetic study of Lakušić & al. (2013). Lakušić & al.
(2013) revealed that all analyzed individuals here treated as
members of the “versicolor 2” group formed one strongly
supported clade (Bayesian posterior probability 1 and parsimony bootstrap value 86), while the main core of individuals
here treated as members of the “versicolor 1” group formed
the second strongly supported clade (Bayesian posterior
probability 1 and parsimony bootstrap value 84) (fig. 2 in
Lakušić & al., 2013). Interestingly, individuals from Demir
Kapija, which in our analysis showed admixed characteristics, formed an independent, well-supported clade on the
cpDNA tree (Bayesian posterior probability 0.99 and parsimony bootstrap value 64). At the same time, a clear genetic
connection between Apennine “tenorei” (IT-Alessano, ITTricase) and one western Balkan “versicolor 2” population
(GR-Mt. Tymfi) was revealed on the majority-rule consensus
tree of combined plastid and nuclear non-coding sequence
data (fig. 4 in Lakušić & al., 2013).
The results of STRUCTURE for K = 3 (Fig. 3B) indicate that the Apennine tenorei populations are genetically
distinct, which is also supported by higher FST values between the tenorei and Balkan populations, as well as with
the position of the tenorei populations at the tip of the second main branch of the Fitch-Margoliash tree (Fig. 2).
However, the tenorei populations were grouped into the
same genetic cluster with a few populations from the western part of the Balkan Peninsula (southern Albania 46–48
350
TAXON 68 (2) • April 2019: 340–369
and northwestern Greece 39, 43). This relatedness may
provide evidence that some time ago the Balkan and Apennine Peninsulas were geographically closer. As a result,
nowadays there are plant species distributed on both peninsulas. The phytogeographical connection between these two
peninsulas could have happened in the period of the
Messinian salinity crisis during the late Miocene when
the level of the Adriatic Sea was lower than today (Hsü
& al., 1973; De Giuli & al., 1987; Krijgsman & al.,
1999; Surina & al., 2014). This assumption is supported
by similar disjunct distributions of several other plant species, e.g., Bromus parvispiculatus (Karl & Scholz, 2009),
Erica manipuliflora (Valdés, 2009–), Euphorbia apios,
Helictotrichon convolutum and Satureja cuneifolia (Di
Pietro & Misano, 2010), Edraianthus graminifolius (Surina
& al., 2014) and Linum elegans (Wagensommer & al.,
2017).
According to the UPGMA clustering, bicajensis,
korabensis, lancifolia and mrkvickana are similar to the central
Balkan populations (Fig. 4), which is in accordance with the
obtained relationships in genetic cluster A – “versicolor 1”
(Fig. 3). The main differences between genetic clusters and
grouping on the morphological level are in the positions of
the plasonii, thessala and tomentella populations. The plasonii
population is more closely related to the populations from genetic cluster B – “versicolor 2”; Figs. 2, 3), while on the morphological level according to the UPGMA clustering it is more
similar to the populations (Fig. 4) from genetic cluster A –
“versicolor 1”. This might be explained by the geographic position of the plasonii population close to the valley and gorge
of the Vardar River through which gene flow among central
and southern Balkan populations can be hypothesized. This is
further supported by the fact that the other two populations that
are geographically close to the Vardar River (14 MKKatlanovo, 15 MK-Veles) share significant portions of genetic
clusters A and B. The presence of several C. versicolor populations has also been recorded in the gorge of the Vardar River
and in the gorges of its tributaries Babuna and Topolka. Although these populations were not included in this study, they
are for sure contributing to the gene flow and genetic diversity
of the surrounding populations in this region. The thessala and
tomentella populations are more related to the populations
from genetic cluster B – “versicolor 2” (Figs. 2, 3), but on
the morphological level according to the UPGMA clustering
they are more similar to “versicolor 1” (Fig. 4). These populations are influenced by the arid Aegean variety of the Mediterranean climate which comes from the East and which is
probably the main reason for the occurrence of small individuals in the western central part of the Balkan Peninsula (northern part of the C. versicolor range).
The CDA of the morphometric data for three groups
(Fig. 5) strongly suggested that the “versicolor 1” (central
Balkan) group is most distinct, as well as that the “tenorei”
(Apennine) group is more similar to the “versicolor 2” (southern Balkan) group. This pattern of morphological relationships
between the investigated groups coincides with their
Discriminant function
analysis
Character
code
Character
Co-H
Length
Co-De
Depth
CoL-H
Corolla lobe height
CoL-h
Distance from maximal width of
the lobe base to the corolla base
0.067
Descriptive statistics (mm)
“versicolor 1”–
C. versicolor subsp.
korabensis
“versicolor 2”–
C. versicolor subsp.
versicolor
“tenorei”–
C. versicolor subsp.
tenorei
0.344
(9.4–)11.1–14.8(–18.3)
(12–)14.5–19.2(–26.2)
(14.2–)16.2–20.1(–22)
0.006
0.994
(5.5–)7–9.9(–12.7)
(8.4–)10.1–13.4(–17.8)
(8.3–)9.6–12.3(–14.3)
8.618
0.000
(3.5–)4.3–5.9(–7.4)
(3.4–)5–7.3(–11)
(5.3–)7.2–9.5(–10.4)
0.010
(4.6–)5.5–7.1(–8.5)
(5.7–)6.9–9(–11.7)
(6.6–)7.5–9.4(–11.4)
Wilk’s
Lambda
F-remove
(2.417)
0.066
1.070
0.066
0.069
4.690
p
Corolla
CoL-Wm
Maximal width of the corolla lobe
0.068
8.516
0.000
(4.6–)7.5–10.3(–12.6)
(6.1–)9.8–13.2(–16.2)
(8.7–)10.3–13.2(–16.7)
Co-Per
Base perimeter
0.066
0.550
0.577
(3.5–)4.4–6.1(–7.6)
(3.5–)5–7.2(–9.8)
(5.6–)7–9.3(–10.1)
TAXON 68 (2) • April 2019: 340–369
Table 6. Summary of discriminant function analysis and results of descriptive statistics of morphometric characters used for the discriminant function analysis of Campanula versicolor s.l.
populations a priori classified into three groups: “versicolor 1”, “versicolor 2” and “tenorei”
Calyx
CaD-W
Width of the teeth base
0.068
8.212
0.000
(1.1–)1.4–1.9(–2.2)
(1.3–)1.7–2.3(–2.9)
(1.7–)2–2.6(–3)
CaD-L
Teeth length
0.067
5.398
0.005
(2.6–)3.7–7.3(–12.1)
(4.4–)6.3–9.5(–13.3)
(3.8–)5–7.6(–9.8)
Height
0.069
11.621
0.000
(1.5–)2–2.9(–4.3)
(1.7–)2.2–2.9(–3.4)
(2.1–)2.3–3.1(–4.2)
Pedicel length
0.066
1.925
0.147
(0.6–)1.2–4.8(–13.6)
(0.4–)1.2–4.2(–9.4)
(0.2–)1.1–4.6(–11.6)
Pi-L
Length
0.068
5.733
0.003
(8.7–)10.8–14.9(–17.7)
(13.1–)15.6–19.1(–21.9)
(14.8–)16.6–20.6(–22.8)
Style length
0.066
0.064
0.938
(3.6–)4.9–7.5(–9.9)
(5.7–)6.8–8.9(–10.8)
(5.9–)7.2–9.2(–10.8)
0.068
7.000
0.001
(1.8–)2.4–3.1(–4)
(2.3–)2.8–3.9(–4.9)
(3.1–)3.6–4.5(–5)
Pistil
Pi-Sy
Stamen
StB-H
Base height
StF-L
Filament length
0.067
2.673
0.070
(0.7–)1.2–2.1(–2.9)
(0.9–)1.5–2.3(–3.5)
(0.9–)1.2–2(–2.9)
StA-L
Anther length
0.074
25.465
0.000
(3.1–)4.3–5.6(–6.3)
(5–)6.4–8.2(–9.8)
(6–)6.9–8.7(–10.5)
Maximal width of the base
0.067
4.749
0.009
(1.6–)1.9–2.4(–2.8)
(1.9–)2.4–3.1(–3.7)
(2.4–)2.6–3.2(–3.8)
StB-Wm
Habitus
Pl-H
Height of the plant
0.066
2.006
0.136
(78–)257.3–667.4(–1182.5)
(155–)347.3–779.1(–1220)
(130–)230.1–489.8(–720)
Inf-L
Inflorescence length
0.070
11.835
0.000
(30–)63.4–340.1(–760.9)
(50–)122.7–351.7(–655)
(30–)33–159(–310)
Stem-H
Stem height
0.067
2.863
0.058
(23–)120.6–400.4(–680)
(65–)148.6–515.1(–1070)
(100–)173.8–354.1(–510)
351
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
Ca-H
CaPe-L
Discriminant function
analysis
Character
code
Character
Fb-Lm
Descriptive statistics (mm)
“versicolor 1”–
C. versicolor subsp.
korabensis
“versicolor 2”–
C. versicolor subsp.
versicolor
“tenorei”–
C. versicolor subsp.
tenorei
(20.4–)34.6–57.9(–74.2)
Wilk’s
Lambda
F-remove
(2.417)
Lamina length
0.069
8.640
0.000
(24.4–)36.4–71.4(–114.1)
(25.3–)47.8–91.7(–133.3)
Fb-Wm
Width of left side
0.066
0.597
0.551
(3.8–)7.8–15(–20.4)
(8.3–)11–21.3(–36)
(6.4–)10.3–17.7(–24.1)
Fb-L
Distance from maximal width
to the base of the lamina
0.067
2.418
0.090
(5.3–)10.4–24.8(–42.1)
(6.6–)11–27.8(–44)
(7.9–)10.4–18.9(–26.1)
FbPe-L
Petiole length
0.068
5.494
0.004
(7.3–)23.1–82.1(–149.1)
(7–)27.2–84.1(–149.7)
(5.6–)15.9–62(–133.3)
FbPe-Wb
Petiole base width
0.066
0.554
0.575
(0.5–)2.7–4.8(–7.5)
(1.5–)3.4–5.6(–7.9)
(2.3–)3.6–5.3(–6.5)
FbPe-W
Petiole width
0.066
1.346
0.261
(0.5–)1.5–2.8(–4.1)
(1–)1.3–2.6(–4.6)
(1.5–)1.8–2.4(–2.9)
p
Basal leaf (Fb)
Middle leaf (Fm)
Fm-Lm
Lamina length
0.068
8.494
0.000
(13.1–)26.9–56.8(–113.9)
(24.3–)36.9–65(–91.3)
(21.6–)25.6–45.4(–63.3)
Fm-Wm
Width of left side
0.069
9.636
0.000
(2.7–)5.3–12.5(–26)
(4.9–)8.9–16.1(–22.6)
(5.1–)7.9–14.3(–19.1)
Fm-L
Distance from maximal width
to the base of the lamina
0.067
4.195
0.016
(4.2–)8.7–21.2(–45.1)
(7.3–)11.4–21.8(–30)
(6.6–)9.2–16.2(–25.8)
FmPe-L
Petiole length
0.071
15.275
0.000
(0.8–)5.2–33.9(–100.6)
(2.5–)6.8–26.3(–61.1)
(1.5–)5.4–23.4(–39.8)
FmPe-Wb
Petiole base width
0.069
9.502
0.000
(0.7–)2.3–3.9(–5.5)
(0.4–)2.9–4.6(–5.9)
(2.3–)3.1–4.7(–6.3)
FmPe-W
Petiole width
0.076
33.500
0.000
(0.8–)1.8–3(–4.5)
(0.9–)1.4–2.9(–4.6)
(1.3–)1.7–2.6(–3.4)
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
352
TABLE 6 Continued.
Table 6. Continued.
Upper leaf (Fu)
Lamina length
0.066
1.060
0.347
(7.7–)17.3–39.2(–70.7)
(17.6–)23.7–43.8(–73.3)
(10.7–)15.9–29(–48.5)
Fu-Wm
Width of left side
0.066
2.189
0.113
(1.6–)3–8.1(–13.1)
(3.2–)5.1–10.4(–16.5)
(2.9–)4.3–7.9(–13)
Fu-L
Distance from maximal width
to the base of the lamina
0.066
0.006
0.994
(2–)6.2–15.8(–32.9)
(4.3–)7.9–17.1(–36.7)
(4.4–)6.2–12(–22)
FuPe-L
Petiole length
0.068
5.669
0.004
(0–)0.2–10.3(–40)
(0–)1.5–8.4(–17.4)
(0–)1.2–8.1(–19.2)
FuPe-Wb
Petiole base width
0.067
4.125
0.017
(0–)1.4–3.5(–4.8)
(0–)1.6–4.1(–5.7)
(0–)2–4(–5)
FuPe-W
Petiole width
0.067
3.358
0.036
(0–)1.2–3.2(–4.6)
(0–)1.1–3(–4.4)
(0–)1.4–2.6(–3.1)
Significant P-values are in bold (P < 0.05). For descriptive statistics, value ranges correspond to the mean ± standard deviation, with the minimal and maximal values in parentheses.
TAXON 68 (2) • April 2019: 340–369
Fu-Lm
TAXON 68 (2) • April 2019: 340–369
distribution (Figs. 1, 5). Furthermore, the position of
“versicolor 1” on the positive part of the first DA axis and
the positions of the “versicolor 2” and “tenorei” on the negative part of the first DA axis correspond with the obtained genetic clusters A, B1 and B2 (Fig. 3B).
According to the results presented in this study, it can be
seen that both genetic and morphological groups obtained are
mostly geographically well defined. Thus, it is surprising that
two populations from the central Balkan (14 MK-Katlanovo,
15 MK-Veles) are related to the Apennine populations from
the genetic cluster B2 – “tenorei” based on the results from
the STRUCTURE analysis for K = 3 (Fig. 3B). By inspecting
allele frequencies by locus for each population, it was found
that these two populations have in high frequency the allele
C008178 that became predominant (or even fixed) in the Apennine tenorei populations. However, on the FitchMargoliash tree, these two populations are not positioned
close to the tenorei populations, but close to the “versicolor
1” group (Fig. 2). Their position on the tree coincides with
their intermediate geographical position between the central
and southern Balkan populations. Furthermore, the population from Veles (15) and the Apennine populations were also
included in the molecular phylogenetic study of Lakušić &
al. (2013). According to the phylogenetic trees based on
ITS and on combined plastid and nuclear non-coding sequence data, the population from Veles is more related to
the central Balkan populations than to the Apennine
populations.
■ TAXONOMIC TREATMENT
Given that the population sampling of Campanula
versicolor s.l. analyzed by Lakušić & al. (2013) was poor (14
populations), the present study was focused on a more comprehensive population sampling (41 populations), spanning the
whole distribution area of C. versicolor. In our study, we
employed faster-evolving markers (microsatellites – SSR), capable of differentiating among closely related and recently diverged taxa. Additionally, we performed a morphometric
analysis on an extensive population sampling – 456 individuals
from 37 populations.
However, our proposed taxonomic treatment is based on
both the previous analyses of cpDNA, microsatellite, morphometric and chorological data for all members of the Campanula pyramidalis complex (Lakušić & al., 2013; Janković &
al., 2016), as well as the results of the current study. According
to the cpDNA tree of Lakušić & al. (2013: fig. 2), populations
of C. versicolor s.l. were grouped into two strongly supported
clades, which correspond to the two main branches/groups revealed on the Fitch-Margoliash tree (Fig. 2) based on the microsatellite data. Apennine tenorei populations are positioned
at the tip of the second main branch of the Fitch-Margoliash
tree (Fig. 2) and according to the results of STRUCTURE for
K = 3 (Fig. 3B), those populations are genetically distinct.
Along with the tenorei populations in genetic cluster B2 –
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
“tenorei” are also grouped some populations from the western
part of the Balkan Peninsula (southern Albania and northwestern Greece). This genetic connection between Apennine and
Balkan populations coincides with the results of the majorityrule consensus tree of combined plastid and nuclear noncoding sequence data of Lakušić & al. (2013: fig. 4).
Taking into account all the results, we propose a new taxonomic treatment of C. versicolor s.l. Three distinct entities
can be distinguished, circumscribed here on the subspecies
level.
A detailed morphological description of C. versicolor s.l.
is provided, while morphological descriptions of the three subspecies cover only diagnostic morphological characters. In the
morphological descriptions, value ranges correspond to mean
± standard deviation, with minimal and maximal values in parentheses. Distribution and habitat data are also provided for all
three subspecies.
Campanula versicolor Andrews, Bot. Repos.: t. 396. 1804 –
Lectotype (designated by Janković & al. in Phytotaxa
323(3): 265. 2017): [illustration] “Campanula versicolor”
in Andrews, Bot. Repos.: t. 396. 1804 [fig. 1A in Janković
& al., 2017]. Epitype (designated by Janković & al. in
Phytotaxa 323(3): 265. 2017): GREECE. Parnas [Mt. Parnassus], stene, krečnjak [limestone cliffs], 38.56761°N,
22.56891°E, 1564.2 m a.s.l., 21 Aug 2014, D. Lakušić,
N. Kuzmanović & I. Janković s.n. (BEOU No. 40045!)
[fig. 2 in Janković & al., 2017].
= Campanula versicolor Sibth. in Sibthorp & Smith, Fl.
Graec. Prodr. 1: 138. 1806, nom. illeg., non Andrews
1804 – Lectotype (designated by Janković & al. in
Phytotaxa 323(3): 267. 2017): [illustration] “Campanula
versicolor” in Sibthorp & Smith, Fl. Graec. 3: t. 207.
1819) [fig. 1B in Janković & al., 2017].
= Campanula planiflora Willd., Enum. Pl.: 210. 1809, nom.
illeg., non Lam. 1785 ≡ C. willdenowiana Schult. in
Roemer & Schultes, Syst. Veg. 5: 107. 1819 – Lectotype
(designated by Janković & al. in Phytotaxa 323(3): 267.
2017): “Hort. bot. Berol W.” [Horto Botanico
Berolinensis/Berlin Botanical Garden Willdenow] (B
barcode B -W 03805-01 0 [photo!]).
Description. – Perennial plant with 1 or few (usually 1 or
3, rarely 10–30) simple herbaceous flowering stems which are
generally semipendulous and slightly arched upward or erect
or just arched in the lower part; usually 25–70 cm long.
Specimens from high-mountain regions can have only 8 cm
long erect stems, while some specimens from low-altitude
and shady habitats can have thinner, gentle, pendulous or
semipendulous stems up to 1 m long. Plants typically
glabrous, rarely with short, bright patent hairs on stems,
leaves and calyx. Glabrous and hairy plants can often be
found in the same population. The rhizome is a brown,
brittle, woody stock with scarfs from old rosette leaves on
the elongated parts from where stems arise so that stems are
up to 10 cm woody in their lower part (caudex). This is
especially noticeable in older specimens. Leaves dark green,
353
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
symmetric or somewhat asymmetric, usually once-folded
along the central nerve and slightly bent down. Leaf margin
glandular, crenulate, crenate or serrate. Rosette usually dense,
having many leaves or made of few broadly cordate,
triangular-deltoid or broadly elliptic leaves. Lower cauline
leaves elliptic, lanceolate or ovate; apex obtuse, subobtuse,
mucronate or acute; base cuneate, attenuate, rounded, cordate
or truncate; (2–)4–8(–13) cm long and 2–7 cm wide. Petioles
narrowly winged. Length of petioles gradually decreases
towards the top of the plant while petiole “wings” are wider.
Uppermost leaves elliptic, lanceolate or cuneate, with short
petioles or sessile; (0.7–)2–4(–8) cm long and 1–2(–4) cm
wide. Inflorescence paniculate, (3–)7–30(–75) cm long, with
clusters of 3–5 flowers, rarely with single flowers. Pedicels
(0.2–)1.5–4.5(–10) cm long; rarely with bracteole. Calyx
green, conical, with few shallow furrows or trilocular with 3
prominent furrows, (2–)3–4(–5) mm in diameter and (1.5–)2–
3(–4) mm long. Calyx teeth green, narrowly triangular or
subulate with wider base; (2.5–)4.5–9(–13) mm long and
(1–)1.5–2.5(–3) mm wide at the base. Corolla widely
campanulate, rotate to flat, divided to one-half, two-thirds or
three-fifths; (9.5–)12–19(–26) mm long. Corolla color pale
violet to violet, rarely white, with or without dark violet eye
in the center of the flower around the base of stamens.
Corolla lobes deltoid or triangular, (5.5–)8–12(–18) mm long
and (4.5–)6–8(–11) mm wide. Style (9–)12–19(–23) mm
long, usually curved upwards, upper half encrusted with
pollen-collecting hairs; pollen grains pale yellow, rarely pale
violet. Ovary trilocular, with numerous ovules. Stamens 5,
(6.5–)9–14(–18) mm long; anthers (3–)5–8(–10) mm long;
filaments (1–)1.5–2(–3.5) mm long, basal part of filaments
triangular to deltoid, (2–)2.5–4(–5) mm long. Capsule
obovoid, shallowly sulcate, pale brownish, dehiscing by basal
pores or irregularly rupturing laterally and apically. Seeds
numerous, reticulate, elliptic-ovate, light brown.
Key to the subspecies of Campanula versicolor
The following key makes no allowance for extreme or
intermediate individuals (potential hybrids).
1. Calyx wide, conical with 3 prominent furrows; calyx teeth
(1.3–)1.7–2.4(–3) mm wide; corolla with dark violet eye in
the center; anther length (5.4–)6.7–8.4(–10.5) mm..........2
1. Calyx narrow, conical with few shallow furrows; calyx
teeth (1–)1.5–2 mm wide; corolla pale violet, without dark
violet eye in the center; anther length (3–)4.5–5.5(–6) mm;
southwestern Bulgaria, southern Serbia, northeastern
Albania and North Macedonia ...........................................
....................................(2) C. versicolor subsp. korabensis
2. Corolla widely campanulate, rotate and flat with clear dark
violet eye; petals bent backwards; inflorescence elongated,
(5–)12–35(–66) cm long; Greece and southern and western Albania....................(1) C. versicolor subsp. versicolor
2. Corolla widely campanulate and conical with a slightly
less clear dark violet eye; petals mostly upright;
354
TAXON 68 (2) • April 2019: 340–369
short internodes; inflorescence thickened, 3–16(–31) cm
long; entire stems distinctively thick; southeastern Italy
...........................................(3) C. versicolor subsp. tenorei
There is no clear boundary between the central and southern populations in the Balkan Peninsula, and this is shown by
the occurrence of populations that are transitional at both genetic and morphological levels. The contact zone is located between southern Republic of North Macedonia and northern
Greece, especially in the western part of this belt in the area
of Lake Prespa and the eastern part of the belt in the gorge of
the Vardar River and its tributaries (Fig. 6).
(1) Campanula versicolor subsp. versicolor
= Campanula versicolor var. thessala Boiss., Fl. Orient. 3:
915. 1875 – Lectotype (designated by Janković & al. in
Phytotaxa 323(3): 269. 2017): GREECE. In rupibus reg.
sylvat. [regione sylvatica] m. Olympi [mount Olympus]
Thessaliae, 21 Jul 1851, T. de Heldreich 400 (G barcode
G00330725 [second sheet G00330725_a] [photo!]).
= Campanula versicolor var. tomentella Halácsy in Österr.
Bot. Z. 42: 372. 1892 ≡ C. versicolor [var. thessala] f.
tomentella Halácsy, Consp. Fl. Graec. 2: 263. 1902 – Lectotype (designated by Janković & al. in Phytotaxa 323(3):
270. 2017): GREECE. Litachori [Lithochori]: Fels an den
Schluchten [rock at the canyons] Megarema (am Olymp),
7 Sep 1891, P.E.E. Sintenis & J.F.N. Bornmüller 1354 (B
barcode B 10 0365469 [photo!]).
= Campanula longipetiolata F.K.Mey. in Haussknechtia
Beih. 15: 151–152, fig. 25. 2011 – Holotype: ALBANIA:
Tepelena, Tal von Luzat, ca. 300 m, 9 Sep 1961, F.K.
Meyer 6035 (JE barcode JE00016710 [photo!]; isotype:
JE barcode JE00016711 [photo!]).
Note. – On the basis of the obtained results, as well as
based on the information from relevant protologues (Boissier,
1875; Halácsy, 1892; Meyer, 2011), study of the original herbarium specimens and personal observations, we concluded
that C. versicolor var. tomentella, C. versicolor var. thessala
and C. longipetiolata should be synonymised with the typical
subspecies. The population of C. longipetiolata was not collected for morphometric study. However, according to the
morphological features of the holotype, the protologue
(Meyer, 2011) and personal observations we made during
our field investigation, it emerged that C. longipetiolata is
not substantially different from C. versicolor. One of the characters usually considered as diagnostic was the long leaf petioles (Meyer, 2011), but this character is no longer reliable
since based on our extensive morphometric study we observed that it is highly variable. Longer leaf petioles are probably the consequence of wetter and shaded habitats.
Furthermore, the acuminate shape of the leaves with a truncate or rounded base has also been found in individuals of
other populations (i.e., populations from Vlorë and Tepelenë
in southern Albania).
Description. – Perennial plants usually with 1 up to 10
(rarely up to 20) simple herbaceous flowering stems which
TAXON 68 (2) • April 2019: 340–369
are generally semipendulous and slightly arched upward, rarely
erect. Leaves with a prominent central nerve and pale,
conspicuous glands at the tips of the teeth. Leaf margin
serrate, rarely crenate or deeply serrate with larger teeth
oriented to the apex of the leaf lamina. Lower cauline leaves
widely elliptic, with acuminate, cuspidate, mucronate or acute
apex and rounded, obtuse, cordate or truncate base; (3–)5–
9(–13) cm long and 2–4(–7) cm wide; widest in the lower
part, in the first third or quarter. Petioles typically almost of
the same length as leaf lamina or 3 or 4(–6) times shorter.
Uppermost leaves elliptic to cuneate with short petioles,
rarely sessile; 2–5(–7) cm long and (0.7–)1–2(–3) cm
wide. Inflorescence elongated; (5–)12–35(–66) cm long.
Calyx widely conical with 3 prominent furrows; (2–)3–4
(–5) mm in diameter and (1.7–)2–3 mm long. Calyx teeth
narrowly triangular or subulate with wider base; (5–)6–9
(–13) mm long and 1.5–2.5(–3) mm wide; bent backward
or clinging to the corolla; twice as long as the ovary;
longer than the connate part of the corolla. Corolla widely
campanulate, rotate or flat, (12–)14.5–19.5(–26) mm long;
violet to blue; with clear dark violet eye; divided to two-
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
thirds, rarely to one-half. Corolla lobes (8.5–)10–13.5(–18)
mm long and (5.5–)6.5–8.5(–11) mm wide. Pistil
(13.5–)16–19(–22) mm, stamens (9–)11–14(–15.5) mm,
anthers (5.5–)6.5–8(–10) mm long. Basal parts of the
filaments are violet, on the edges brighter violet and with
a short pale violet longitudinal line from the apex to the
middle of the filament base (Fig. 7).
Distribution and ecology. – Populations of this
subspecies can be found in the southern part of the Balkan
Peninsula, in Greece and southern Albania (Fig. 6; Appendix
2). The geographical distribution ranges from the south area
of Lake Prespa in northwestern Greece to the island of Kythira
(Strid & Tan, 1996; Yannitsaros, 2004) south of the Peloponnese and from Mt. Pangeon in northeastern Greece to the island
of Othonoi (Halácsy, 1902) in northwestern Greece. Populations occur in the vegetation of rocky crevices (Asplenietea
trichomanes–Potentilletalia speciosae, Onosmetalia frutescentis) and more rarely within limestone rocky slopes and screes,
at altitudes ranging between 20 and 2000 m a.s.l. This subspecies can also be found on the walls of fortresses, old monasteries and old bridges.
Fig. 6. Distribution map of three subspecies of C. versicolor and position of transitional populations.
355
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
TAXON 68 (2) • April 2019: 340–369
Fig. 7. Campanula versicolor subsp. versicolor. A & B, Habitus; C, Flower; D, Rhizome; E, Calyx; F, Part of
inflorescence. — Photographs: I. Janković.
(2) Campanula versicolor subsp. korabensis (F.K.Mey.)
I.Janković & D.Lakušić, comb. & stat. nov. ≡
C. korabensis F.K.Mey. in Haussknechtia Beih. 15: 149–
150, fig. 22. 2011 – Holotype: ALBANIA. Korab, Wiesen
oberhalb Radomir, ca. 1400 m, an Felsen, 5 Aug 1959,
F.K. Meyer 4759 (JE barcode JE00016705 [photo!];
isotype: JE barcode JE00016706 [photo!]).
= Campanula plasonii Formánek in Verh. Naturf. Vereins
Brünn 37: 155–156. 1899 – Lectotype (first-step designated by Janković & al. in Phytotaxa 323(3): 271. 2017):
Republic of NORTH MACEDONIA. Pržigrad Dudica,
1898, de Ed. Formánek s.n. (BRNM No. 13259/36
356
[photo!]), second-step (designated here): Republic of
NORTH MACEDONIA. (Pržigrad) Dudica, 1895, de
Ed. Formánek s.n. (BRNM No. 13259/36, the upper
right-hand plant turned face down [photo!]).
= Campanula mrkvickana Velen. in Allg. Bot. Z. Syst. 11:
44–45. 1905 ≡ C. versicolor f. mrkvickana (Velen.)
Hayek in Repert. Spec. Nov. Regni Veg. Beih. 30(2):
543. 1930 – Lectotype (designated by Ančev in
Kožuharov & Ančev, Fl. Reipubl. Bulgaricae 11: 108.
2013): BULGARIA. In siccis calcareis collinis Konjovo
Planina ad Kistendil, Aug 1904, I. Mrkvička s.n. (PRC
barcode PRC 451230!).
TAXON 68 (2) • April 2019: 340–369
= Campanula versicolor [subsp. thessala] subvar. lancifolia
Bornmüller in Bot. Jahrb. Syst. 59: 2–3. 1925 – Lectotype
(designated by Janković & al. in Phytotaxa 323(3): 271.
2017): Republic of NORTH MACEDONIA. In valle fl.
Treska, ad rupes, 300–500 m, 27 Apr 1918, J. Bornmüller
4398 (JE barcode JE00007087 [photo!]).
= Campanula korabensis subsp. bicajensis F.K.Mey. in
Haussknechtia Beih. 15: 150, fig. 23. 2011 – Holotype:
ALBANIA. Gjalica e Lumes, Bicaj, Eingang der Schlucht,
ca. 400 m, 31 Jul 1959, F.K. Meyer 4562 (JE barcode
JE00016708 [photo!]).
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
Note on the second-step lectotypification. – The first-step
lectotypification was published in Janković & al. (2017: 271).
The specimen selected as lectotype is part of the original
material, and the plants on the sheet fit the description
provided in the protologue. Therefore, we do not have
arguments to supersede it according to Art. 9.19 of the ICN.
However, two localities (two gatherings) were cited on the
label of this specimen, both mentioned in the protologue
(Formánek, 1899: 155) and visited earlier by Formánek in
1895 (Formánek, 1896: 255). The locality Prži Grad is
located in the Kavadarci municipality in the Republic of
Fig. 8. Campanula versicolor subsp. korabensis (F.K.Mey.) I.Janković & D.Lakušić, comb. & stat. nov. A &
B, Habitus; C, Flower; D, Rhizome; E, Calyx; F, Part of inflorescence. — Photographs: I. Janković.
357
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
North Macedonia, at the altitude of 1620 m a.s.l. The locality
Dudica (2132 m a.s.l.) is the peak of Mt. Kožuf, which is a border mountain between the Republic of North Macedonia and
Greece. The upper right-hand plant, selected here as the
second-step lectotype, was most probably collected on Dudica,
as this plant is much smaller than the other plants on the herbarium sheet. This pattern of a smaller habitus of plants
inhabiting higher altitudes observed during our field investigations was confirmed in this study.
Note. – The earliest legitimate name whose epithet could
be used for the new combination is Campanula plasonii
published by Formánek (1899: 155). However, most of the
localities Formánek cited in the protologue are situated in a
transitional zone between the central and southern Balkan
populations. We included in our analyses the population from
Demir Kapija, provided in the protologue for C. plasonii, and
this population is genetically admixed. Bearing in mind the
original circumscription of Formánek’s C. plasonii, as well as
the results of our study, we consider that using the epithet
plasonii would be taxonomically problematic and decided not
to use it for making the new combination. The next-earliest
legitimate name, Campanula mrkvickana of Velenovský
(1905: 44), is also taxonomically disruptive: the analyzed
population from the locus classicus slightly differs genetically
and morphologically from other central Balkan populations
because of its isolated position. Therefore, we used the name
Campanula korabensis of Meyer (2011) as basionym. The
designation “Campanula versicolor f. matkae” was published
in a master thesis (Nikolov, 2005: 67), being therefore not
validly published according to Art. 29.1 of the ICN.
Based on our results, as well as the information from relevant protologues (Formánek, 1899; Bornmüller, 1925;
Velenovský, 1905; Meyer, 2011), the original herbarium specimens and personal observations, it is concluded that the following taxa should be synonymised with C. versicolor subsp.
korabensis: C. versicolor subsp. thessala subvar. lancifolia,
C. mrkvickana, C. plasonii and C. korabensis subsp. bicajensis.
Description. – Perennial plants with 1–5 (rarely 10 or
more) simple herbaceous flowering stems which are typically
erect and arched only in the lower part; rarely whole stems
are arched upward. Leaf margin slightly crenulate or obtusely
serrate, sometimes slightly undulate like the whole lamina.
Glands are rare, small and not present on every leaf tooth.
Lower cauline leaves linear elliptic, lanceolate, oblong or
elongated cordate; apex subobtuse or acute; base rounded,
cuneate, truncate or attenuate, rarely slightly cordate;
(2.5–)4–7(–11) cm long and (0.8–)1.5–3(–4) cm wide; widest
in the first half or the first third. Petioles are half as long or
longer than the leaf lamina. Leaves in the middle part of the
stems sessile or with a very short petiole which gradually
passes into the attenuate leaf base. Uppermost leaves
narrowly elliptic, cuneate or lanceolate, leaf base attenuate to
sessile; (0.7)1.5–4(–7) cm long and (0.3–)0.7–1.5(–2.5) cm
wide. Inflorescence elongated; (3–)7–33(–74) cm long. Calyx
narrower than calyx of flowers of the typical subspecies,
conical, with few shallow furrows; (1.7–)2.5–3(–4) mm in
358
TAXON 68 (2) • April 2019: 340–369
diameter and (1.5–)2–3(–4.5) mm long. Calyx teeth
triangular; (2.5–)3.5–7(–12) mm long and (1–)1.5–2 mm
wide; slightly bent backward or in line with the upper surface
of the ovary (ovary disk); 2–3 times longer than the ovary;
shorter than the connate part of the corolla. Corolla widely
campanulate and conical, pale violet to almost white, without
dark violet eye; divided to one-half, rarely to two-thirds;
(10–)11–15(–18) mm long. Corolla lobes triangular to
deltoid, upright; (5.5–)7–10(–13) mm long and 5–7(–8) mm
wide. Pistil (9–)10.5–15(–18) mm, stamens (6.5–)8–11(–12)
mm, anthers (3–)4.5–5.5(–6) mm long. Basal parts of the
filaments are pale violet, the same color as the corolla (Fig. 8).
Distribution and ecology. – Populations of this
subspecies can be found in the central part of the Balkan
Peninsula, in southwestern Bulgaria, southern Serbia, Republic of North Macedonia and northeastern Albania (Fig. 6;
Appendix 2). Populations mostly occur in the vegetation of
rocky crevices (Asplenietea trichomanes–Potentilletalia
speciosae), rarely in the vegetation of screes (Drypidetea
spinosae), or on rocky ground; usually on limestone between
200 and 1900 m a.s.l., rarely on serpentine.
(3) Campanula versicolor subsp. tenorei (Moretti) I.Janković
& D.Lakušić, comb. & stat. nov. ≡ C. tenorei Moretti in
Schouw, Prosp. Descr. Geogr. Piante Ital.: 19 [= in Giorn.
Fis. Chim. Storia Nat. Med. Arti, dec. 2, 7: 41]. 1824,
nom. nov. pro C. corymbosa Ten. ≡ C. corymbosa Ten.,
Fl. Napol. 1: XV. 1811, nom. illeg., non Desf. 1808 ≡
C. rosanoi Ten., Fl. Napol. 3: 205–206. 1827, nom. illeg.
(Art. 52.1 and 52.2 of the ICN) ≡ C. versicolor var.
rosanii (Ten.) Nyman, Consp. Fl. Eur.: 480. 1879, comb.
illeg. – Lectotype (designated by Janković & al. in
Phytotaxa 323(3): 267. 2017): ITALY. Basilicata,
Ginosa, s.d., M. Tenore s.n. (NAP!) [fig. 3 in Janković
& al., 2017].
Note. – This is a new combination based on C. tenorei
which was published by Moretti (1824) as a replacement
name for the illegitimate name C. corymbosa Ten.
(Tenore, 1811; Art. 53.1 of the ICN). Morphological
differentiation between subsp. tenorei and subsp. versicolor is
less clear than between subsp. versicolor and subsp.
korabensis. However, C. versicolor subsp. tenorei is
characterized by some qualitative morphological features by
which it can be easily recognized, i.e., distinctively thick
stems and leaves, short inflorescence, stem with a long woody
basal part.
Description. – Perennial plants with up to 5 (rarely up to
20) simple herbaceous glabrous flowering stems which are
generally semipendulous and whole stems are arched
upward, rarely erect. Stems distinctively thick, in the upper
part up to 8 mm in diameter. Lower part of stems can be up
to 10 cm woody. Leaves also thick, glabrous. Leaf margin
irregularly serrate, rarely crenate or deeply serrate with
larger teeth oriented to the apex. Glands at the tips of the
leaf teeth are pale and small. Lower cauline leaves widely
ovate, apex obtuse or subobtuse, base blunt, slightly cordate
TAXON 68 (2) • April 2019: 340–369
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
Fig. 9. Campanula versicolor subsp. tenorei (Tenore) I.Janković & D.Lakušić, comb. & stat. nov. A
& B, Habitus; C, Flower; D, Rhizome; E, Calyx; F, Part of inflorescence. — Photographs: I. Janković.
or truncate; (2–)3.5–6(–7) cm long and (1.5–)2–4(–5) cm
wide; widest in the first third or quarter. Petioles typically
half as long as the leaf lamina. Uppermost leaves narrowly
elliptic, obovate or cuneate with short petioles, rarely
sessile; (1–)1.5–3(–5) cm long and (0.5–)1–1.5(–2.5) cm
wide. Inflorescence thickened and with short internodes
between flower clusters; 3–16(–31) cm long. Calyx green,
glabrous, conical with 3 prominent furrows; 3–4(–5) mm in
diameter and 2–3(–4) mm long. Calyx teeth green, narrowly
triangular; (4–)5–7.5(–10) mm long and (1.7–)2–2.5(–3) mm
wide in the base; typically clinging to the corolla; half as
long as the ovary. Corolla widely campanulate and conical,
divided up to two-thirds, rarely to one-half; (14–)16–20(–22)
mm long; violet to pale violet, with or without dark violet
eye in the center which is less clearly expressed than in
individuals of the typical subspecies. Corolla lobes
triangular to deltoid; upright; (8–)9.5–12(–14) mm long and
(6.5–)7–9(–11) mm wide. Pistil (15–)17–21(–23) mm,
stamens (11–)12–15(–18) mm, anthers (6–)7–9(–10.5) mm
long. Basal parts of stamen filaments of darker color than
the corolla if a dark violet eye is present. If the dark violet
eye is not present, then the basal part of the filaments are
the same color as the corolla (Fig. 9).
Distribution and ecology. – Populations of this
subspecies can be found in southeastern Italy, in the Apulia
and Basilicata administrative regions (Fig. 6; Appendix 2).
Populations mostly occur in the vegetation of rocky crevices
on limestone as well as on old walls and fortresses; at altitudes
ranging from 8 to 400 m a.s.l (Asplenietea trichomanes–Caro
multiflori–Aurinion megalocarpae).
359
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
■ AUTHOR CONTRIBUTIONS
DL designed the research. DL, NK, RDP, ZN and IJ collected almost all the analyzed population samples. IR performed the main part
of the molecular laboratory work. ZŠ performed the statistical data analysis of microsatellites. ZL and ZŠ contributed to the design of the research and the analysis of the microsatellite data as well as to the
interpretation of the results. IJ designed figures, performed the measurements of morphometric characters and with the help of NK conducted
the statistical analysis of the morphometric data. NK wrote the part
regarding nomenclatural issues. IJ wrote the manuscript with the support
of all authors. All authors contributed to the interpretation of the
results and to the final version of the manuscript. — IJ, https://orcid.
org/0000-0003-1501-0167; ZŠ, https://orcid.org/0000-0002-9848-1601;
ZL, https://orcid.org/0000-0002-2370-6650; NK, http://orcid.org/00000003-3463-5541; RDP, https://orcid.org/0000-0003-4983-8931; IR,
https://orcid.org/0000-0001-8900-5112; DL, http://orcid.org/00000001-6708-6652
■ ACKNOWLEDGEMENTS
We gratefully acknowledge the financial support provided by the
Serbian Ministry of Education, Science and Technological Development
(Project No. 173030 to D. Lakušić) as well as of Institutional Fundings
of Scientific and Artistic activities of the University of Zagreb (Project
No. 20282319 to Z. Liber) and Centre of Excellence for Biodiversity
and Molecular Plant Breeding (CroP-BioDiv), Zagreb, Croatia. Many
thanks to Dr. G. Tomović, Dr. M. Niketić, U. Buzurović, Dr. S. Đurović,
Š. Duraki, Dr. M. Lazarević, Dr. P. Lazarević (Belgrade, Serbia), Dr. B.
Zlatković (Niš, Serbia), Dr. L. Shuka (Tirana, Albania), Dr. S. Škondrić
(Banja Luka, Bosnia and Herzegovina), Dr. B. Surina (Rijeka, Croatia),
Dr. S. Bogdanović (Zagreb, Croatia) and Dr. D. Iamonico, Dr. G.
Misano, Dr. G. Gambetta and Dr. P. Medagli (Italy) for their assistance
in collecting the plant material. Thanks to Dr. J. Šinžar-Sekulić
(Belgrade, Serbia) for the help with the statistical analysis of the morphometric data. Many thanks go to the curators of the herbaria for sending scans of the herbarium specimens from the relevant collections.
■ LITERATURE CITED
Ančev, M. 1984. Campanula versicolor Andrews. P. 359 in: Velchev,
V. (ed.), Red Data Book of the Republic of Bulgaria, vol. 1, Plants.
Sofia: BAS & MoEW.
Andrews, H.C. 1804. The botanists repository: Comprising colour’d
engravings of new and rare plants, vol. 6. London: printed by
T. Bensley, and published by the author. https://doi.org/10.5962/
bhl.title.51972
Applied Biosystems 2005. GeneMapper, version 4.0. https://www.
thermofisher.com/order/catalog/product/4440915
Asenov, A. 2009. Reports 1–6. P. 274 in: Vladimirov, V., Dane, F. &
Tan, K. (eds.), New floristic records in the Balkans: 11. Phytol.
Balcan. 15(2): 273–289.
Blionis, J.G. & Vokou, D. 2005. Reproductive attributes of Campanula
populations from Mt Olympos, Greece, Pl. Ecol. 178: 77–88.
https://doi.org/10.1007/s11258-004-2495-6
Bogdanović, S., Brullo, S., Rešetnik, I., Lakušić, D., Satovic, Z. &
Liber, Z. 2014. Campanula skanderbegii: Molecular and morphological evidence of a new Campanula species (Campanulaceae) endemic to Albania. Syst. Bot. 39: 1250–1260. https://doi.org/
10.1600/036364414X682571
Bogdanović, S., Rešetnik, I., Brullo, S. & Shuka, L. 2015. Campanula
aureliana (Campanulaceae), a new species from Albania. Pl.
Syst. Evol. 301(6): 1555–1567. https://doi.org/10.1007/s00606014-1171-0
360
TAXON 68 (2) • April 2019: 340–369
Boissier, E. 1875. Flora orientalis, vol. 3. Genevae et Basileae [Geneva &
Basel]: apud H. Georg. https://doi.org/10.5962/bhl.title.20323
Bornmüller, J. 1925. Beiträge zur Flora Mazedoniens 1 – Sammlungen
in den Kriegsjahren 1916–1918. Bot. Jahrb. Syst. 59: 294–504.
Botstein, D., White, R.L., Sholnick, M. & Davis R.W. 1980.
Construction of a genetic linkage map in man using restriction
fragment length polymorphisms. Amer. J. Human Genet. 32:
314–331.
Castroviejo, S., Aldasoro, J.J. & Alarcón, M.; with contributions
from Hand, R. 2010. Campanulaceae. – In: Euro+Med Plantbase
– The information resource for Euro-Mediterranean plant diversity.
http://ww2.bgbm.org/EuroPlusMed/PTaxonDetail.asp?
NameCache=Campanulaceae&PTRefFk=7400000
Cavalli-Sforza, L.L. & Edwards, A.W.F. 1967. Phylogenetic analysis:
Models and estimation procedures. Amer. J. Human Genet. 19:
233–257.
Chapuis, M.P. & Estoup, A. 2007. Microsatellite null alleles and estimation of population differentiation. Molec. Biol. Evol. 24:
621–631. https://doi.org/10.1093/molbev/msl191
Cornuet, J.M. & Luikart, G. 1996. Description and power analysis of
two tests for detecting recent population bottlenecks from allele frequency data. Genetics 144: 2001–2014.
Crowl, A., Miles, N., Visger, C., Hansen, K., Ayers, T., Haberle, R. &
Cellinese, N. 2016. A global perspective on Campanulaceae: Biogeographic, genomic, and floral evolution. Amer. J. Bot. 103(2):
233–245. https://doi.org/10.3732/ajb.1500450
Damboldt, J. 1965. Zytotaxonomische Revision der isophyllen
Campanulae in Europa. Bot. Jahrb. Syst. 84: 302–358.
De Giuli, C., Masini, F. & Valleri, G. 1987. Paleogeographic evolution
of the Adriatic area since Oligocene to Pleistocene. Rivista Ital.
Paleontol. Stratigrafia 93(1): 109–126.
Dempster, A.P., Laird, N.M. & Rubin, D.B. 1977. Maximum likelihood from incomplete data via the EM algorithm. J. Roy. Statist.
Soc. Ser. B 39: 1–22. https://doi.org/10.1111/j.2517-6161.1977.
tb01600.x
Di Pietro, R. & Misano G. 2010. Shrubland and garrigue vegetation in
the « Gravine » gorges (Apulia region, south-eastern Italy). Acta
Bot. Gallica 157(1): 195–229. https://doi.org/10.1080/12538078.
2010.10516199
Di Renzo, F., Peterson, A., Garza, J.C., Valdes, A., Slatkin, M. &
Freimer, N.B. 1994. Mutational processes of simple-sequence repeat loci in human populations. Proc. Natl. Acad. Sci. U.S.A. 91:
3166–3170. https://doi.org/10.1073/pnas.91.8.3166
Earl, D.A. & von Holdt, B.M. 2012. STRUCTURE HARVESTER: A
website and program for visualizing STRUCTURE output and
implementing the Evanno method. Conservation Genet. Resources
4: 359–361. https://doi.org/10.1007/s12686-011-9548-7
Evanno, G., Regnaut, S. & Goudet, J. 2005. Detecting the number of
clusters of individuals using the software STRUCTURE: A simulation study. Molec. Ecol. 14: 2611–2620. https://doi.org/10.1111/
j.1365-294X.2005.02553.x
Excoffier, L., Smouse, P.E. & Quattro, J.M. 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: Application to human mitochondrial DNA restriction sites.
Genetics 131: 479–491.
Excoffier, L., Laval, G. & Schneider, S. 2005. Arlequin ver. 3.0: An
integrated software package for population genetics data analysis.
Evol. Bioinf. Online 1: 47–50. https://doi.org/10.1177/117693430
500100003
Felsenstein, J. 1985. Confidence limits on phylogenesis: An approach
using the bootstrap. Evolution 39: 783–791. https://doi.org/
10.1111/j.1558-5646.1985.tb00420.x
Felsenstein, J. 1993. PHYLIP: Phylogeny inference package,
version 3.5c. Department of Genetics, University of Washington,
Seattle.
Formánek, E. 1896. Zweiter Beitrag zur Flora von Serbien, Macedonien
und Thessalien. Verh. Naturf. Vereins Brünn 34: 255–368.
TAXON 68 (2) • April 2019: 340–369
Formánek, E. 1899. Fünfter Beitrag zur Flora von Macedonien. Verh.
Naturf. Vereins Brünn 37: 124–220.
Gagnidze, R. 2005. Vascular plants of Georgia: A nomenclatural
checklist. Tbilisi: Georgian Academy of Sciences.
Geslot, A. 1984. Campanula L. In: Greuter, W., Burdet, H.M. & Long,
G. (eds.), Med-Checklist: A critical inventory of vascular plants of
the circum-Mediterranean countries, vol. 1. Geneva: Conservatoire
et Jardin botaniques.
Goudet, J. 1995. FSTAT (Version 1.2): A computer program to calculate F-statistics. J. Heredity 86: 485–486. https://doi.org/10.1093/
oxfordjournals.jhered.a111627
Goudet, J. 2002. FSTAT, version 2.9.3. http://www2.unil.ch/popgen/
softwares/fstat.htm
Govaerts, R. & Lammers, T.G. 2018. World checklist of
Campanulaceae. Facilitated by the Royal Botanic Gardens, Kew.
Published on the Internet: http://wcsp.science.kew.org/ (accessed
15 May 2018).
Halácsy, E. 1892. Beiträge zur Flora der Balkanhalbinsel. Oesterr. Bot.
Z. 42: 368–375.
Halácsy, E. 1902. Conspectus florae Graecae, vol. 2. Lipsiae [Leipzig]:
sumptibus Guilelmi Engelmann.
Hartvig, P. 1991. Campanula L. Pp. 369–387 in: Strid, A. & Tan, K.
(eds.), The mountain flora of Greece, vol. 2. Edinburgh: Edinburgh
University Press.
Holm, S. 1979. A simple sequentially rejective multiple test procedure.
Scand. J. Statist. 6: 65–70. https://doi.org/10.5962/bhl.title.9919
Hsü, K.J., Ryan, W.B.E. & Cita, M.B. 1973. Late Miocene desiccation
of the Mediterranean. Nature 242: 240–244. https://doi.org/
10.1038/242240a0
Janes, J.K., Miller, J.M., Dupuis, J.R., Malenfant, R.M., Gorrell,
J.C., Cullingham, C.I. & Andrew, R.L. 2017. The K = 2 conundrum. Molec. Ecol. 26: 3594–3602. https://doi.org/10.1111/
mec.14187
Janković, I., Satovic, Z., Liber, Z., Kuzmanović, N., Radosavljević, I.
& Lakušić, D. 2016. Genetic diversity and morphological variability in the Balkan endemic Campanula secundiflora s.l.
(Campanulaceae). Bot. J. Linn. Soc. 180: 64–88. https://doi.org/
10.1111/boj.12359
Janković, I., Lakušić, D., Di Pietro, R. & Kuzmanović, N. 2017. Nomenclatural notes and typifications in Campanula versicolor
(Campanulaceae) and related names. Phytotaxa 323(3): 264–274.
https://doi.org/10.11646/phytotaxa.323.3.5
Jones, E.K., Korotkova, N., Petersen, J., Henning, T., Borsch, T. &
Kilian, N. 2017. Dynamic diversification history with rate upshifts
in Holarctic bell-flowers (Campanula and allies). Cladistics 33(6):
637–666. https://doi.org/10.1111/cla.12187
Kalinowski, S.T. 2004. Counting alleles with rarefaction: Private alleles
and hierarchical sampling designs. Conservation Genet. 5:
539–543. https://doi.org/10.1023/B:COGE.0000041021.91777.1a
Kalinowski, S.T. 2005. HP-Rare 1.0: A computer program for
performing rarefaction on measures of allelic diversity. Molec. Ecol.
Notes 5: 187–189. https://doi.org/10.1111/j.1471-8286.2004.00845.x
Karl, R. & Scholz, H. 2009. Bromus parvispiculatus H. Scholz. P. 340
in: Greuter, W. & Raus, T. (eds.), Med-Checklist Notulae 28.
Willdenowia 39: 335–345.
Kopelman, N.M., Mayzel, J., Jakobsson, M., Rosenberg, N.A. &
Mayrose, I. 2015. Clumpak: A program for identifying clustering
modes and packaging population structure inferences across K.
Molec. Ecol. Resources 15: 1179–1191. https://doi.org/10.1111/
1755-0998.12387
Kovačić, S. 2004. The genus Campanula L. (Campanulaceae) in
Croatia, circum-Adriatic and west Balkan region. Acta Bot. Croat.
63: 171–202.
Kovačić, S & Nikolić, T. 2006. Relations of the western Balkan endemic Campanula L. (Campanulaceae) lineages based on comparative floral morphometry. Pl. Biosyst. 140: 260–272. https://doi.org/
10.1080/11263500600947665
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
Krijgsman, W., Hilgen, F.J., Raffi, I., Sierro, F.J. & Wilson, D.S.
1999. Chronology, causes and progression of the Messinian salinity
crisis. Nature 400: 652–655. https://doi.org/10.1038/23231
Lakušić, D., Liber, Z., Nikolić, T., Surina, B., Kovačić, S., Bogdanović,
S. & Stefanović, S. 2013. Molecular phylogeny of the Campanula
pyramidalis species complex (Campanulaceae) inferred from chloroplast and nuclear non-coding sequences and its taxonomic implications. Taxon 62: 505–524. https://doi.org/10.12705/623.1
Lammers, T.G. 2007. World checklist and bibliography of
Campanulaceae. Richmond: Royal Botanic Gardens, Kew.
Liber, Z., Kovačić, S., Nikolić, T., Likić, S. & Rusak, G. 2008.
Relation between western Balkan endemic Campanula L.
(Campanulaceae) lineages: Evidence from chloroplast DNA.
Pl. Biosyst. 142(1): 40–50. https://doi.org/10.1080/112635007018
72283
Linnaeus, C. 1753. Species plantarum, vol. 1. Holmiae [Stockholm]:
impensis Laurentii Salvii. https://doi.org/10.5962/bhl.title.669
Luikart, G., Allendorf, F.W., Cornuet, J.M. & Sherwin, W.B. 1998.
Distortion of allele frequency distributions provides a test for recent
population bottlenecks. J. Heredity 89: 238–247. https://doi.org/
10.1093/jhered/89.3.238
Mansion, G., Parolly, G., Crowl, A.A., Mavrodiev, E., Cellinese, N.,
Oganesian, M., Fraunhofer, K., Kamari, G., Phitos, D.,
Habarle, R., Akaydin, G., Ikinci, N., Taus, T. & Borsch, T.
2012. How to handle speciose clades? Mass taxon-sampling as a
strategy towards illuminating the natural history of Campanula
(Campanuloideae). PLoS ONE 7(11): e50076. https://doi.org/
10.1371/journal.pone.0050076
MedCalc Software 2005–2011. Digimizer image analysis software,
version 4.0.0.0. Belgium. http://www.digimizer.com/
Meyer, F.K. 2011. Beiträge zur Flora von Albanien. Haussknechtia,
Beiheft 15. Jena: Thüringische Botanische Gesellschaft e.V.
Moretti, G. 1824. [Notes on Italian Campanulas.] In: Scho[u]w, J.F.,
Prospetto di una descrizione geografica delle piante d’Italia e
di Sicilia. Pavia. https://gallica.bnf.fr/ark:/12148/bpt6k98642b/f2.
image
Mucina, L., Bültmann, H., Dierßen, K., Theurillat, J.P., Raus, T.,
Čarni, A., Šumberová, K., Willner, W., Dengler, J., García,
R.G., Chytrý, M., Hájek, M., Di Pietro, R., Iakushenko, D.,
Pallas, J., Daniëls, F.J.A., Bergmeier, E., Guerra, A.S.,
Ermakov, N., Valachovič, M., Schaminée, J.H.J., Lysenko,
T., Didukh, Y.P., Pignatti, S., Rodwell, J.S., Capelo, J., Weber,
H.E., Solomeshch, A., Dimopoulos, P., Aguiar, C., Hennekens,
S.M. & Tichý, L. 2016. Vegetation of Europe: Hierarchical floristic classification system of vascular plant, bryophyte, lichen,
and algal communities. Appl. Veg. Sci. 19: 3–264. https://doi.
org/10.1111/avsc.12257
Nikolov, Z. 2005. Taksonomija i horologija na rodot Campanula L.
(Campanulaceae) vo Skopskata kotlina = Taxonomy and
chorology of the genus Campanula L. (Campanulaceae) in the
Skopje basin. M.Sc. thesis, Faculty of Natural Sciences and
Mathematics, Saint Cyrili and Methodius University of Skopje,
Skopje, Republic of North Macedonia. [In Macedonian — available from: Botanical Library, Faculty of Natural Science and
Mathematics, Skopje]
Park, J.M., Kovačić, S., Liber, Z., Eddie, E.M.M. & Schneeweiss,
G.M. 2006. Phylogeny and biogeography of isophyllous species
of Campanula (Campanulaceae) in the Mediterranean area. Syst.
Bot. 31: 862–880. https://doi.org/10.1600/036364406779695924
Pascual, M., Aquadro, C.F., Soto, V. & Serra, L. 2001. Microsatellite
variation in colonizing and palearctic populations of Drosophila
subobscura. Molec. Biol. Evol. 18: 731–740. https://doi.org/
10.1093/oxfordjournals.molbev.a003855
Phitos, D. 1964. Trilokuläre Campanula-Arten der Ägäis. Österr. Bot.
Z. 111: 208–230. https://doi.org/10.1007/BF01373765
Phitos, D. 1965. Die quinquelokulären Campanula-Arten. Österr. Bot.
Z. 112: 449–498. https://doi.org/10.1007/BF01373181
361
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
Pignatti, S. (ed.) 1982. Flora d’Italia, vol. 2. Bologna: Edagricole.
Piry, S., Luikart, G. & Cornuet, J.M. 1999. BOTTLENECK: A computer programme for detecting recent reductions in the effective
population size using allele frequency data. J. Heredity 90:
502–503. https://doi.org/10.1093/jhered/90.4.502
Pritchard, J.K., Stephens, M. & Donnelly, P. 2000. Inference of population structure using multilocus genotype data. Genetics 155:
945–959.
Radosavljević, I., Jakse, J., Satovic, Z., Javornik, B., Janković, I. &
Liber, Z. 2015. New microsatellite markers for Campanula
pyramidalis (Campanulaceae) and cross-amplification in closely related species. Applic. Pl. Sci. 3(3): 1400117. https://doi.org/10.3732/
apps.1400117
Raymond, M. & Rousset, F. 1995. GENEPOP (version 1.2): Population genetics software for exact tests and ecumenicism. J.
Heredity 86: 248–249. https://doi.org/10.1093/oxfordjournals.
jhered.a111573
Rice, W.R. 1989. Analyzing tables of statistical tests. Evolution 43:
223–225. https://doi.org/10.1111/j.1558-5646.1989.tb04220.x
Rohlf, F.J. 1997. NTSYS-pc: Numerical taxonomy system, version 2.0.2.
Setauket (NY): Exeter Software.
Rousset, F. 1997. Genetic differentiation and estimation of gene flow
from F-statistics under isolation by distance. Genetics 145:
1219–1228.
SAS Institute 2004. SAS/STAT® 9.1 user’s guide. Cary (NC): SAS
Institute. https://support.sas.com/documentation/onlinedoc/91pdf/
sasdoc_91/stat_ug_7313.pdf
Solari, F. 1947. Curculionidi nuovi o poco conosciuti della fauna
paleartica. XIII. Note preliminari sui Miarus ed un nuovo genere
di Mecinini (Col. Curc.). Mem. Soc. Entomol. Ital. 26: 72–79.
StatSoft 1996. STATISTICA: Data analysis software system, version
5.1. Tulsa (OK): StatSoft.
Strid, A. & Tan, K. 1996. Flora and vegetation of the Peloponnese and
Kithira. Report of a student excursion form the University of Copenhagen, May 14–28, 1995. Copenhagen, Botanical Inst.
Surina, B., Schneeweiss, G.M., Glasnović, P. & Schönswetter, P.
2014. Testing the efficiency of nested barriers to dispersal in the
Mediterranean high mountain plant Edraianthus graminifolius
(Campanulaceae). Molec. Ecol. 23: 2861–2875. https://doi.org/
10.1111/mec.12779
TAXON 68 (2) • April 2019: 340–369
Tenore, M. 1811–1815. Flora Napolitana Prodromus, vol. 1.
Napoli [Naples]: nella Stamperia Francese. http://www.ortobotanico
napoli.it/paginadimenu.htm
Turland, N.J., Wiersema, J.H., Barrie, F.R., Greuter, W.,
Hawksworth, D.L., Herendeen, P.S., Knapp, S., Kusber,
W.-H., Li, D.-Z., Marhold, K., May, T.W., McNeill, J., Monro,
A.M., Prado, J., Price, M.J. & Smith, G.F. (eds.) 2018. International Code of Nomenclature for algae, fungi, and plants (Shenzhen
Code): Adopted by the Nineteenth International Botanical Congress Shenzhen, China, July 2017. Regnum Vegetabile 159.
Glashütten: Koeltz Botanical Books. https://doi.org/10.12705/
Code.2018
Valdés, B. 2009–. Erica. – In: Euro+Med Plantbase – The information
resource for Euro-Mediterranean plant diversity. http://ww2.bgbm.
org/EuroPlusMed/PTaxonDetail.asp?NameCache=Erica%
20manipuliflora&PTRefFk=7100000
Van Oosterhout, C., Hutchinson, W.F., Wills, D.P.M. & Shipley, P.
2004. Micro-Checker: Software for identifying and correcting
genotyping errors in microsatellite data. Molec. Ecol. Notes 4:
535–538.
Velenovský, J. 1905. Beiträge zur Flora des Orients. Allg. Bot. Zeitschr.
Syst. 11: 43–45.
Vigouroux, Y., Glaubitz, J., Matsuoka, Y., Goodman, M.,
Sánchez, J. & Doebley, J. 2008. Population structure and genetic
diversity of New World maize races assessed by DNA
microsatellites. Amer. J. Bot. 95(10): 1240–1253. https://doi.org/
10.3732/ajb.0800097
Visiani, R. de & Pančić, J. 1862. Plantae serbicae rariores aut novae,
Decas, I. Mem. Reale Ist. Veneto Sci. 10: 425–450.
Wagensommer, R.P., Bartolucci, F., Forentino, M., Licht, W.,
Peccenini, S., Perrino, E.V. & Venanzoni, R. 2017. First record
for the flora of Italy and lectotypification of the name Linum
elegans (Linaceae). Phytotaxa 296(2): 161–170. https://doi.org/
10.11646/phytotaxa.296.2.5
Willdenow, C.L. 1809. Enumeratio plantarum, vol. 1. Berolini [Berlin]:
in taberna libraria scholae realis. http://bibdigital.rjb.csic.es/ing/
Libro.php? Libro=1684
Yannitsaros, A. 2004. Addition to the flora of Kithira (Greece) II.
Willdenowia 34(1): 117–128. https://doi.org/10.3372/wi.34.34110
Appendix 1. Information regarding populations of Campanula versicolor s.l. used in this study (No. = population numbering; Locality code = country twoletter code defined by ISO 3166-1 plus population locality; G = genetic clusters obtained with STRUCTURE analysis for K = 3; S = populations used for
molecular analysis; M = populations used for morphometric analysis; Alt. = altitude).
No. Locality code
Taxon
Informal name G
S M
Herbarium and
voucher no.
Country and locality
Alt. (m)
A
X
BEOU 30450
Bulgaria, Zemen
750
593
C. versicolor subsp. korabensis
1 BG-Zemen
C. mrkvickana
mrkvickana
2 BG-Zemen
C. mrkvickana
mrkvickana
X
BEOU 40039
Bulgaria, Zemen
3 RS-Preševo
C. versicolor
versicolor
A
X X
BEOU 40040
Serbia, Preševo
590
4 RS-Gotovuša
C. versicolor
versicolor
A
X
BEOU 33143
Serbia, Mts. Sharr, Gotovuša
1080
5 RS-Dušanov grad
C. versicolor
versicolor
A
X
BEOU 28087
Serbia, Prizren, Dušanov grad
500
6 RS-Prizren
C. versicolor
versicolor
BEOU 40190
Serbia, Prizren, Prizrenska
Bistrica Canyon
582
X
7 AL-Mt. Gjallica
C. versicolor
versicolor
A
X
NHMR 1094
Albania, Mt. Gjallica
500
8 AL-Shija Gorge
C. versicolor
subsp. bicajensis
bicajensis
A
X X
BEOU 46507
Albania, Mt. Gjallica,
Shija River Gorge
526
9 AL-Mt. Korab
C. korabensis
korabensis
A
X X
BEOU 46501
Albania, Mt. Korab, Radomirë
1467
(Continues)
362
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
TAXON 68 (2) • April 2019: 340–369
Appendix 1. Continued.
Taxon
Informal name G
S
10
AL-Kurbnesh
C. versicolor
versicolor
A
X
BEOU 42620
Albania, Lezhë, Kurbnesh
789
11
MK-Ostruga
C. versicolor
versicolor
A
X X
BEOU 37599
North Macedonia, Mt. Skopska
Crna Gora, Ostruga
1250
12
MK-Matka Canyon
C. versicolor subsp.
lancifolia
thessala subvar. lancifolia
A
X X
BEOU 37606
North Macedonia, Treska river,
Matka Canyon
355
13
MK-Skopje
C. versicolor
versicolor
X
BEOU 37609
North Macedonia, Skopje,
quarry Govrlevo
435
14
MK-Katlanovo
C. versicolor
versicolor
B2 X X
BEOU 37605
North Macedonia, Katlanovo,
Katlanovo spa
945
15
MK-Veles
C. versicolor
versicolor
B2 X X
BEOU 37596
North Macedonia, Veles, Mladost Lake
250
16
MK-Kozjak Lake
C. versicolor
versicolor
A
X X
BEOU 37601
North Macedonia, Kozjak Lake, Zdunje
530
17
MK-Radika Gorge
C. versicolor
versicolor
X
BEOU 40038
North Macedonia, Radika River Gorge
1309
18
MK-Radika Gorge
C. versicolor
versicolor
A
X
BEOU 33441
North Macedonia, Radika River Gorge
850
19
MK-Mt. Bistra
C. versicolor
versicolor
A
X
BEOU 30785
North Macedonia, Mt. Bistra, Galičnik
1450
A
X
A
X
X
20
MK-Mt. Bistra
C. versicolor
versicolor
21
MK-Mt. Ilinska pl.
C. versicolor
versicolor
22
MK-Mt. Ilinska pl.
C. versicolor
versicolor
23
MK-Ohrid Lake
C. versicolor
versicolor
24
MK-Ohrid Lake
C. versicolor
versicolor
25
MK-Mt. Galičica
C. versicolor
versicolor
A
26
MK-Demir Kapija
C. plasonii
plasonii
B1 X
27
MK-Demir Kapija
C. plasonii
plasonii
M
Herbarium and
voucher no.
Country and locality
No. Locality code
X
X
X
X
Alt. (m)
BEOU 37602
North Macedonia, Mt. Bistra, Galičnik
1500
BEOU 29700
North Macedonia, Mt. Ilinska planina
1869
BEOU 40035
North Macedonia, Mt. Ilinska planina
1830
BEOU 27924
North Macedonia, Ohrid Lake, Peštani
700
BEOU 37575
North Macedonia, Ohrid Lake, Peštani
708
BEOU 37614
North Macedonia, Mt. Galičica
1750
BEOU 37593
North Macedonia, Demir Kapija
190
BEOU 37593
North Macedonia, Demir Kapija
190
C. versicolor subsp. versicolor
28
GR-Oros Askion
C. versicolor
versicolor
B1 X
BEOU 39009
Greece, Oros Askion
580
29
MK-Mt. Olympus
C. versicolor
subsp. tomentella
tomentella
B1 X X
BEOU 40042
Greece, Mt. Olympus, Prionia
680
30
MK-Prionia
C. versicolor
subsp. thessala
thessala
B1 X X
BEOU 40043
Greece, Mt. Olympus, Prionia
1077
31
GR-Tempi Gorge
C. versicolor
versicolor
B1 X
BEOU 28150
Greece, Pineios river, Tempi Gorge
50
32
GR-Tempi Gorge
C. versicolor
versicolor
BEOU 40044
Greece, Pineios river, Tempi Gorge
40
33
GR-Mt. Giona
C. versicolor
versicolor
B1 X
BEOU 28090
Greece, Mt. Giona, Gravia – Amfissa
850
34
GR-Mt. Parnassus
C. versicolor
versicolor
B1 X X
BEOU 40045
Greece, Mt. Parnassus
1546
35
GR-Alonistaina
C. versicolor
versicolor
B1 X X
BEOU 40052
Greece, Alonistaina
1160
36
GR-Parorio
C. versicolor
versicolor
X
BEOU 40051
Greece, Mt. Taygetus, Parorio
350
37
GR-Mt. Mavrovouni C. versicolor
versicolor
B1 X
BEOU 40049
Greece, Mt. Mavrovouni
1700
38
GR-Lefkada
C. versicolor
versicolor
B1 X X
BEOU 44458
Greece, Lefkada, Exanthia
602
39
GR-Mt. Tzoumerka
C. versicolor
versicolor
B2 X X
BEOU 40055
Greece, Mt. Tzoumerka
1500
40
GR-Voreia
C. versicolor
versicolor
X
BEOU 40056
Greece, Mt. Tzoumerka, Voreia
840
41
GR-Ioannina
C. versicolor
versicolor
X
BEOU 40057
Greece, Ioannina
650
42
GR-Vikos Gorge
C. versicolor
versicolor
B1 X X
BEOU 40058
Greece, Vikos Gorge, Kokkoros bridge
750
43
GR-Mt. Tymfi
C. versicolor
versicolor
B2 X X
BEOU 40059
Greece, Mt. Tymfi, Mikro Papigo
970
44
GR-Corfu
C. versicolor
versicolor
B1 X
BEOU 44460
Greece, Corfu, Makrades
204
45
AL-Bistirca
C. versicolor
versicolor
B1 X X
BEOU 40060
Albania, Bistrica
90
46
AL-Luzat
C. longipetiolata
longipetiolata
B2 X
BEOU 40061
Albania, Tepelenë, Luzat
220
47
AL-Tepelenë
C. versicolor
versicolor
B2 X X
BEOU 40063
Albania, Tepelenë
200
48
AL-Vlorë
C. versicolor
versicolor
B2 X X
BEOU 40066
Albania, Vlorë
20
49
IT-Matera
tenorei
B2 X X
BEOU 41745
Italy, Matera
347
X
C. versicolor subsp. tenorei
C. tenorei
(Continues)
363
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
TAXON 68 (2) • April 2019: 340–369
Appendix 1. Continued.
Herbarium and
voucher no.
Country and locality
Alt. (m)
B2 X X
BEOU 41741
Italy, Castellaneta
223
B2 X X
BEOU 41746
Italy, Punta Palascia
55
B2 X
ZA 30841
Italy, Alessano
9
X
BEOU 41748
Italy, Giuliano del Capo, Il Ciolo
26
tenorei
X
BEOU 41749
Italy, Leuca, Santa Maria di Leuca
46
tenorei
B2 X X
BEOU 41750
Italy, Gallipoli, Conchiglie–Alto
10
No. Locality code
Taxon
Informal name G
50
IT-Castellaneta
C. tenorei
tenorei
51
IT-Punta Palascia
C. tenorei
tenorei
52
IT-Alessano
C. tenorei
tenorei
53
IT-Il Ciolo
C. tenorei
tenorei
54
IT-Leuca
C. tenorei
55
IT-Gallipoli
C. tenorei
S M
Appendix 2. Specimens examined in this study.
Campanula versicolor subsp. versicolor. ALBANIA: Gjirokastër — Dhëmbel (pers. obs., Lakušić, D. 2012); Gjirokastër (BEOU 31444); Mount Dhëmbel,
saddle Dhëmbel (BEOU 32737); Poliçan (BEOU 42510); Tepelenë, Bence gorge (BEOU 34817, BEOU 40065); Tepelenë, Luzat (BEOU 40061); Tepelenë,
rocks below fortress (BEOU 40064, ZAGR 39388); Tepelenë (BEOU 40063); Vjosa river, Këlcyrë gorge, between Këlcyrë and Dragot (pers. obs., Lakušić, D.
2012; BEOU 32837, BEOU 40062); Korçë — Mali i Thatë, Shëngjergj (BEOU 38307); Prespa Lake, Maligrad (BEOU 38751); Sarandë — Bistrica (BEOU
32621); Butrint (BM 001191859); Maja e Dhema, between Ksamil Pasqyra (W 2012-00847); Vlorë — Bistrice (pers. obs., Lakušić, D. 2012; BEOU 40060);
Dukat (pers. obs., Lakušić, D. 2012); Palasë, below fortress (ZAGR 39389); Vlorë, Dukat i Ri (BEOU 32618); Vlorë (BEOU 31443, BEOU 32617, BEOU
40066). GREECE: East Central — Osa (BEOU 27739); North Central — Edessa, Edessa–Agras (BP 380965); Kalabaka, Grand Meteora (MNHN
P04442298); Kalabaka, Meteora (W 11674); Kastoria, Kastoria Lake (SOM 135684); Kastoria (E 00616299); Kozani, Servia (W 9490); Mt. Olympus,
Litohoro, Enipeas (W 2013-04281, E 00616290); Mt. Olympus, Litohoro, Mega Rema (AMD 44175, B 10-0365469, BEOU 22436, BEOU 22439, BEOU
22440, BM 001191866, BM 001191873, BP 142008, BP 389977, L 2997008, LD 1323599, LD 1330319, LD 1331519, LD 1367798, LD 1371473, LE
1017320, LE 1017324, MANCH, MNHN P00177996, MNHN P00177997, MNHN P00177998, MNHN P00177999, MNHN P00185840, MNHN
P00185841, PH 0000426, PH 00025238, SOM 72821, SOM 72822, SOM 72824, SOM 75231, W 4240, W 9324, W 9325); Mt. Olympus, Litohoro (L
2997009, L 2997010, W 18988); Mt. Olympus, monastery of St. Dionysius, refugi Stavros (BEO, LD 1323539, LD 1380098, MNHN P00178000, SOM
72825, SOM 75230, W 01254, W 18987); Mt. Olympus, Prionia (BEOU 24729, BEOU 39010, BEOU 40041, BEOU 40042, BEOU 40043, BEOU
8045); Mt. Olympus (BP 389975, C, FI 050345, MNHN P00177995); Oros Askion, east of Namata (E 00616301); Oros Askion (BEOU 39009); Mts.
Pindhos, Kalabaka, Meteora (SOM 165164); Pineios river, Tempi gorge, Rapsani (W 01148); Pineios river, Tempi gorge (BEOU 28150, BEOU 40044);
Prespa Lake, Mikro Prespa (pers. obs., Lazarević, P. & al. 2017); Ionian islands — Kefalonia, Argostolion (BM 001191871, FI 050366); Kefalonia, Ainos
(BEO); Kefalonia, Poros (E 00616296); Corfu, Ag. Dimitrios (C); Corfu, Lakones (BEOU 44460); Corfu, Makrades (BEOU 44461); Corfu, Pantocratoras (W
2004-00954, W 2006-16339); Lefkada, Apolpena (BEOU 40054); Lefkada, Exanthia (BEOU 44458); Lefkada, Nydri, Nydri waterfall (BEOU 44459);
Lefkada, from village Kalamitsi toward beach Kavalikefta, W exp. (BEOU 31441); Lefkada (BEOU 40053); Southern Pindhos — Mts. Acarnanian, monastery Romvos, between Gdyves and Flamboura summits (W 2008-20961); Klissoura gorge, Chrisobergion, Byzantine church Agia Eleousa (C); Mts.
Pindhos, Agrafa, Karoplesi (BM 001191860, E 00616295); Sermeniko, Kokkino Lithari (E 00616291, LD 1374174, MANCH); Mt. Tomaros, Kopani (C);
Mt. Tomaros (BM 001191864, FI 050367); Mt. Tzoumerka, Katafigio, Kostelata (BEOU 37023); Mt. Tzoumerka, Strungula (BEOU 32341); Mt. Tzoumerka,
Voreia (BEOU 40056); Mt. Tzoumerka (BEOU 40055); Peloponnese — Alonistaina (BEOU 40052); Mt. Helmos, Kalavrita, from ski center to the summit
(pers. obs., Lakušič, D. & al. 2017); Mt. Helmos, Krioneri, Zarouchla (C); Mt. Helmos (BM 001191867, MNHN P00185837, PI-GUAD); Mt. Kyllini, east of
Ano Trikala; along the road towards the summit (pers. obs., Lakušić D. & al. 2017); Mt. Kyllini, Ano Trikala (C); Mt. Kyllini, Trikala (BEO, BM 001191861,
BM 001191862, BM 001191872, BP 389971, FI 050362, MANCH, MNHN P00185838, W 62601); Langadia (BEOU 28142); Mt. Mainalo, Levidi, ski center
(C); Mt. Mainalo, east of ski center (BEOU 46977); Mt. Mainalo, along the road towards the ski center (pers. obs., Lakušić D. & al. 2017); Mani, Cape Matapan, Porto Kagio (W 9494); Mt. Mavrovouni (BEOU 40049, BM 001191870, C, CGE 25027, FI 050360, LE 017321); Mt. Parnon, Polidroso, drinking fountain (BEOU 46983); Mt. Parnon (BP 142009, BP 389972, FI 050347, FI 050365, GE, LD 1320542, MANCH, MNHN P00185834, MNHN P00185835,
MNHN P00185836, MNHN P04444771, W 2186); Mt. Taygetus, Parorio (BEOU 40051); Mt. Taygetus (BEOU 36868, BEOU 40050, BM 001191863,
BM 001191869, COI 0054764, E 00616293, E 00616297, E 00616298, E 00616300, FI 050359, LE 1017322, MANCH, MNHN P00185825, MNHN
P00185826, MNHN P00185839, NCY 004736, W 8160); Tripoli, Neochori (BEOU 40048); Northern Pindhos — Gliki, Aheron river, tunnel (pers. obs.,
Janković, I., Stevanoski, N. 2016); Ioannina, Mitsikeli (BEOU 39012); Ioannina, Konitsa (C); Ioannina (BEOU 40057); Mitsikeli, Kriovrisi (W 200716737); Mt. Tymfi, Astraka (W 9498); Mt. Tymfi, Mikro Papigo (BEOU 39011, BEOU 40059, L 3744198); Mt. Tymfi, Vikos, Kokkoros bridge (BEOU
40058); Mt. Tymfi, Vikos (BEOU 27894); North East — Mt. Panagaion Hills, Lofkes (C); Sterea Ellas — Chelidona, Palio Mikro Chorio (BP 389370,
CGE 25028, GE, W 9852); Chelidona (MNHN P00185832, MNHN P00185833); Delphi, Castalian, Hyampeia (NCY 018637); Delphi, Castalian (NCY
018634, NCY 018635); Mt. Giona (BEOU 28090, BEOU 40047, BEOU 47093, NCY 018633, NCY 018636); Mt. Helicon, Palaeovouni (C); Mt. Helicon,
Paliovouna, Koukoura (C); Levadia, Trophonios (U 1181190); Mt. Parnassus, between Amfissa and Karoute (C); Mt. Parnassus, Eptalofos (BEOU
40046); Mt. Parnassus, Gourna (C, FI 050348, L 2996997, MNHN P00185827, MNHN P00185828, MNHN P00185829, PI, W, W 19896); Mt. Parnassus,
Langadha (C); Mt. Parnassus, at ski center (BEOU 27700); Mt. Parnassus (BEO, BEOU 31921, BEOU 40045, E 00616292, MNHN P00185824, MPU, U
1181191); Mts. Agrafa, south of the village Petralona, ESE of the peak Svoni, peak Pouli (C); Mts. Pindhos, Mt. Tymfristos, Sympetherikon (BP 389370,
FI 050346, MNHN P00185830, MNHN P00185831, W 17161); Prevezis, between Nikopoleos and Pargas, Kato Mirsini (C); Mt. Vardousia, Vustinitza
(BM 001191868, MANCH, MNHN P00185843, MNHN P00185844, MNHN P00185845, MNHN P04444770, BP 142010). Campanula versicolor subsp.
korabensis. ALBANIA: Dibër — Mt. Korab, Radomirë, Fuša i Korabit (BEOU 46501); Mt. Korab, Radomirë (pers. obs., Lakušič, D. & al. 2016); Kukës
— Mt. Gjallica, between Lusen and Kolesjan (BEOU 46503); Mt. Gjallica, Shija river gorge, close to Bicaj (BEOU 46507, NHMR 1094); Mt. Koritnik,
Vana-Tal (W 2008-02139); Mt. Paštrik, west side (W 14919); Lezhë — Kurbnesh (BEOU 42620). BULGARIA: Kyustendil — Mt. Konjevska (BP
380964); Zemen, Zemen gorge (BEOU 40039, SOM 157675, SOM 75227, SOM 75228); Zemen, Zemenski prolom (BEOU 30450, BEOU 37646); Mt.
Zemenska (SOM 102124, SOM 165495). MACEDONIA: Southeastern region — Rabrovo (BM 001191865); Southwestern region — Black Dri river
gorge, Debar Lake, HE Globočica (HMMNH 3825); Black Dri river gorge, Debar Lake, HE Špilje (HMMNH 2235 ⇒ HMMNH 2239); Black Dri river gorge,
Lukovo (BEO, BEOU 37573); Debar, between Debar and Izrovi (C); Mt. Galičica, Baba (BEOU 33464, BEOU 37579, BEOU 37614, BEOU 38237, BEOU
40037); Mt. Galičica, Gjura (HMMNH 6986, HMMNH 6987); Mt. Galičica, toward Tomoros (BEO); Mt. Galičica, Ljubaništa (BEOU 37577); Mt. Galičica,
(Continues)
364
TAXON 68 (2) • April 2019: 340–369
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
Appendix 2. Continued.
saddle (BEOU 21848, BEOU 40036); Mt. Galičica, old ski lift (BEOU 37620); Mt. Ilinska planina, village Golemo Ilino, summit Liska (BEOU 37582, BEOU
37611, HMMNH 7027, HMMNH 7043 ⇒ HMMNH 7045); Mt. Ilinska planina, village Golemo Ilino (BEOU 29700, BEOU 40035); Mt. Ilinska planina,
Železnec (BEOU 37610); Makedonski Brod, Pešnica river gorge, at the cave (BEOU 38464); Mt. Momina čuka (BRNM 13254/36, BRNM 13261/36); Ohrid
Lake, Peštani (BEOU 27924, BEOU 37574, BEOU 37575); Poreče, Breznica (BEO); Stogovo, Gari, above village (HMMNH 2234); Suhi Dol, Kičevo
(BEO); Treska river, Kozjak Lake, Zdunje (BEOU 30005, BEOU 37601); Treska river, Poreče, Kapina (BEO); Treska river, Poreče, Selišta (BEO); Pelagonia
— Kičevo (BEOU 22437); Prespa Lake, Oteševo (BEOU 37623, BEOU 46665); Prespa Lake, Oteševo–Carina (BEOU 46666); Prespansko jezero, Stenje
(BEOU 46667, BRNM 13255/36, HMMNH 6985); Polog Region — Mt. Bistra, Galičnik, between Janče and Galičnik (BEOU 37569, BEOU 37570); Mt.
Bistra, Galičnik (BEOU 30785, BEOU 37602); Mt. Bistra, summit Maskarovec (HMMNH 2240 ⇒ HMMNH 2251); Radika river gorge, Adžina reka (BEOU
40038); Radika river gorge, between Debar and Mavrovi anovi, crossroad for Nichpur (C); Radika river, Barič gorge (BEOU 33441, BEOU 33442, BEOU
37567); Radika river, Rec gorge (BEOU 33459, BEOU 37566); Radika river, Sence (BEOU 33439, BEOU 33440, BEOU 37568); Northeastern region
— Kumanovo, village Pčinja (BEOU 37598, BEOU 38516); Skopje — Katlanovo, Katlanovo spa (BEOU 37605, HMMNH 2843 ⇒ HMMNH 2865);
Katlanovo, Kožle, Prnar (BEO); Katlanovo, Kožle (BEOU 37604, HMMNH 2866 ⇒ HMMNH 2893); Mts. Sharr, Kobilica (BP 389973, BP 389974);
Mts. Sharr, Raduša (BEO); Skopje, quarry Govrlevo (BEOU 37609); Mt. Skopska Crna Gora, Banjani, monastery of St. Ilija (BEOU 37600, HMMNH
2106 ⇒ HMMNH 7026); Mt. Skopska Crna Gora, Ostruga (BEOU 37599); Treska, HE Matka (BEO, BEOU 27448, HMMNH 2225 ⇒ HMMNH 2229,
HMMNH 2947, HMMNH 2950 ⇒ HMMNH 2970, HMMNH 4413 ⇒ HMMNH 7024, HMMNH 7028 ⇒ HMMNH 7042); Treska river, monastery of
St. Nikola Šiševski (BEO); Treska river, Matka canyon, monastery of the Holy Mother of God (BEO, HMMNH 2898 ⇒ HMMNH 2921, HMMNH 2947
⇒ HMMNH 2949, HMMNH 2966); Treska river, Matka canyon (BEOU 37606); Treska river (JE 00007087); Vardar — Demir Kapija, Beli Kamen
(BEO); Demir Kapija, Kadovo gorge (BEOU 25040); Demir Kapija (BEO, BEOU 33715, BEOU 37593, E 00113261, W 1995-06892); Drenovo gorge, between Kvadarci and Prilep, Raec (BEOU 30306, BEOU 37591); Mt. Flora (BRNM 13256/36, BRNM 13258/36); Mt. Klepa (pers. obs., Zlatković, B. 2018);
Mt. Kožuf, Dudica (BRNM 13260/36); Mt. Kožuf, Pržigrad (BRNM 13259/36); Veles, Babuna river, Pešti gorge (HMMNH 2230 ⇒ HMMNH 2233); Veles,
Babuna river gorge (BEO, BEOU 37594, BEOU 38590); Veles, Mladost Lake (BEOU 37596, BEOU); Veles, Navičanska reka (BEO); Veles, Topolka river
(BEOU 22438, BEOU 37595); Veles, Vardar river gorge, close to railroad station “Pčinja” (HMMNH 2210, HMMNH 2211); Veles, Veles, Vardar river gorge
(BEOU 25035, SOA 18768); Veles (BEOU 27450, BEOU 27451). SERBIA: South — Preševo (BEO, BEOU 40040, BEOU 40192); Kosovo — White Drin
river canyon, Našec (BEOU 1855); Mt. Kodža Balkan (BEOU 2545); Prizren, Duvska Klisura gorge (BEOU 2581); Prizren, Prizrenska Bistrica river (BEOU
40190); Prizren, Prizrenska Bistrica river gorge, Dušanov Grad (BEOU 24520, BEOU 28087); Prizren, Prizrenska Bistrica river gorge, Kaljaja (BEO); Prizren,
Prizrenska Bistrica river gorge (BEOU 2893); Prizren, Sinan Pasha mosque (BEOU 31446); Mt. Sharr, Brod selo – Gradski kamen (BEOU 2316/91); Mt.
Sharr, Drven grad (BEO); Mt. Sharr, Globočki kamen (BEOU 22710); Mt. Sharr, village Gotovuša (BEOU 33143); Mt. Sharr, Gotovuška river gorge (BEOU
7792); Mt. Sharr, Kokošinje, Ljuboten (slopes toward village Gotovuša) (BEOU 7826); Mt. Sharr, Kule, towards Lešačka Bistrica river valley (BEOU 1308/
92); Mt. Sharr, Tumba peak (BEOU 7811). Campanula versicolor subsp. tenorei. ITALIA: Apulia — Brindisi (K 00814350); Castellaneta (BEOU 41741);
Gallipoli, Conchiglie-Alto, Montagna Spaccata (BEOU 41750, FI 050351); Gallipoli, Madonna Dell’Alto Mare (BM 001191882); Gallipoli, Rupi di San
Mauro (BM 001191882); Gallipoli (RO); Guliano del Capo, Il Ciolo (BEOU 41748); Laterza, gravina di Laterza (BEOU 41739, BEOU 41740); Leuca, Santa
Maria di Leuca (BEOU 41749, PignS982: 692); Marina Serra (BEOU 41747); Massafra, gravina della Madonna della Scala (BEOU 41751); Mottola, gravina
di Petruscio (BEOU 41743); Otranto (FI 050364); Palagianello, gravina San Biagio (BEOU 41742); Porto Miggiano, Santa Cesarea Terme (pers. obs.,
Vydrova, A., Grulich, V. 2014); Punta Palascia (BEOU 41746, BM 001191874, BM 001191875, BM 001191876, BM 001191877, BM 001191878, BM
001191879, BM 001191880, BM 001191881, BM 001191883, BM 001191885, BP 380969, E 00616303, FI 050352, FI 050353, FI 050354, GE, MNHN
P00185822, PI-GUAD, RO); Taranto, Murgie di Leucaspidi (K 00814351); Palagianello, Gravina di Palagianello (RO). Basilicata — Matera, gravina di
Matera (BEOU 41744), Matera (BEOU 41745, FI 050356, FI 050357, FI 050358, RO).
Appendix 3. Descriptive statistics of morphological characters of Campanula versicolor s.l. (Max = maximum value; Min = minimum value; Std. Dev. = standard
deviation; CV = coefficient of variation).
Character code
Character
Min (mm)
Mean (mm)
Max (mm)
Std. Dev.
CV (%)
Co-H
Length
9.4
15.2
26.2
3.0
19.8
Co-De
Depth
3.5
6.0
10.1
1.4
23.4
CoL-H
Corolla lobe length
5.5
10.1
17.8
2.2
21.5
CoL-h
Distance from maximal width of the lobe base to the corolla base
3.4
6.0
11.0
1.5
24.9
CoL-W
Width of the corolla lobe base
4.7
7.0
10.9
1.2
17.5
CoL-Wm
Maximal width of the corolla lobe
4.6
7.3
11.7
1.3
17.8
Co-Per
Base perimeter
4.6
10.3
16.7
2.0
19.6
Height of connate part
2.6
5.1
9.7
1.4
26.5
Corolla
Co-Di
Calyx
CaD-W
Width of the teeth base
1.1
1.9
3.0
0.4
19.9
CaD-L
Teeth length
2.6
6.5
13.3
2.0
30.4
Ca-H
Height
1.5
2.5
4.3
0.4
16.6
CaPe-L
Pedicel length
0.2
2.9
13.6
1.7
59.0
(Continues)
365
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
TAXON 68 (2) • April 2019: 340–369
Appendix 3. Continued.
Character code
Character
Min (mm)
Mean (mm)
Max (mm)
Std. Dev.
CV (%)
Pistil
Pi-L
Length
8.7
15.4
22.8
3.1
20.4
Pi-Sy
Style length
3.6
7.1
10.8
1.4
20.2
Pi-St
Length of hairy part of pistil
4.2
8.3
13.1
2.0
23.8
Stamen
StB-H
Base height
1.8
3.2
5.0
0.7
21.0
StF-L
Filament length
0.7
1.7
3.5
0.4
25.4
StA-L
Anther length
3.1
6.3
10.5
1.5
23.4
StB-W
Width of the base
1.3
2.1
3.4
0.4
18.2
StB-Wm
Maximal width of the base
1.6
2.5
3.8
0.5
18.5
StB-h
Distance from maximal width to the base
0.5
1.0
2.1
0.3
27.1
StB-Ar
Base area
2.2
5.8
14.7
2.2
38.9
StB-Pe
Base perimeter
5.7
9.1
14.5
1.7
19.1
Habitus
Pl-H
Height of the plant
78.0
478.9
1220.0
210.0
43.8
Inf-L
Inflorescence length
30.0
195.7
760.9
129.4
66.1
Stem height
23.0
287.8
1070.0
155.0
53.9
Stem-H
Basal leaf (Fb)
Fb-Lm
Lamina length
20.4
58.3
133.3
20.4
35.1
Fb-Wm
Width of left side
3.8
13.5
36.0
4.8
35.3
Fb-L
Distance from maximal width to the base of the lamina
5.3
17.8
44.0
7.5
41.9
FbPe-L
Petiole length
5.6
51.7
149.7
28.7
55.6
FbPe-Wb
Petiole base width
0.5
4.1
7.9
1.1
27.0
FbPe-W
Petiole width
0.5
2.1
4.6
0.6
29.4
Fb-Ar
Lamina area
134.3
1172.1
6119.6
817.9
69.8
Fb-Pe
Lamina perimeter
53.7
139.4
327.1
46.8
33.6
Lamina length
13.1
44.1
113.9
15.0
33.9
Fm-Wm
Width of left side
2.7
10.6
26.0
3.9
36.8
Fm-L
Distance from maximal width to the base of the lamina
4.2
15.2
45.1
5.7
37.2
FmPe-L
Petiole length
0.8
17.7
100.6
12.3
69.4
FmPe-Wb
Petiole base width
0.4
3.5
6.3
0.9
25.0
FmPe-W
Petiole width
0.8
2.3
4.6
0.6
28.6
Fm-Ar
Lamina area
52.5
684.5
2805.8
462.4
67.6
Fm-Pe
Lamina perimeter
30.5
104.7
245.3
34.7
33.2
Middle leaf (Fm)
Fm-Lm
Upper leaf (Fu)
Fu-Lm
Lamina length
7.7
29.2
73.3
10.7
36.8
Fu-Wm
Width of left side
1.6
6.4
16.5
2.7
41.6
Fu-L
Distance from maximal width to the base of the lamina
2.0
11.2
36.7
4.6
41.3
FuPe-L
Petiole length
0.0
5.1
40.0
4.3
85.1
FuPe-Wb
Petiole base width
0.0
2.7
5.7
1.1
42.6
FuPe-W
Petiole width
0.0
2.1
4.6
0.9
43.8
Fu-Ar
Lamina area
17.4
276.5
1678.2
216.4
78.3
Fu-Pe
Lamina perimeter
18.4
67.6
166.8
24.3
35.9
366
Table A.
No.
1
1
3
**
4
*
**
5
*
7
ns
8
**
9
**
10
**
11
**
12
**
14
**
15
**
16
*
18
**
19
ns
21
**
23
**
25
**
26
**
28
*
29
*
0.652
4
0.587
0.477
**
*
**
**
**
**
**
**
**
**
**
ns
**
**
**
**
**
**
ns
ns
**
**
**
**
**
**
**
ns
**
ns
**
**
**
**
*
ns
5
0.536
0.328
0.281
7
0.543
0.392
0.212
0.148
ns
**
**
**
*
**
**
*
ns
**
ns
*
*
**
*
ns
ns
ns
**
**
*
**
**
*
ns
*
ns
ns
*
ns
*
ns
ns
8
0.559
0.465
0.256
0.253
0.133
9
0.520
0.351
0.329
0.187
0.179
0.236
**
**
**
**
**
**
**
**
ns
**
**
**
**
**
**
**
**
**
**
**
**
**
ns
**
**
**
**
**
10
0.519
0.409
0.251
0.263
0.179
0.181
0.185
**
**
**
**
**
**
**
ns
**
**
**
**
**
11
0.568
0.505
0.431
0.394
0.359
0.419
0.364
0.376
**
**
**
**
**
**
ns
**
**
**
**
**
**
12
0.422
0.328
0.260
0.195
0.141
0.235
0.176
0.193
0.281
14
0.583
0.417
0.402
0.376
0.317
0.423
0.302
0.340
0.430
0.165
**
**
*
**
ns
**
**
**
**
**
**
**
**
**
ns
**
**
**
**
**
**
15
0.609
0.440
0.416
0.324
0.307
0.400
0.277
0.348
0.409
0.174
0.172
16
0.471
0.327
0.233
0.217
0.184
0.258
0.246
0.210
0.335
0.085
0.274
0.266
**
**
ns
**
**
**
**
**
**
**
ns
*
**
**
**
*
ns
18
0.491
0.354
0.209
0.207
0.141
0.164
0.141
0.136
0.276
0.165
0.319
0.298
0.208
19
0.598
0.421
0.212
0.184
0.043
0.113
0.113
0.084
0.391
0.154
0.353
0.355
0.210
0.074
ns
**
**
**
**
**
**
ns
ns
ns
ns
ns
21
0.669
0.472
0.469
0.399
0.367
0.419
0.354
0.411
0.475
0.213
0.391
0.407
0.280
0.357
0.427
ns
**
**
**
**
23
0.644
0.460
0.423
0.354
0.288
0.361
0.335
0.307
0.451
0.175
0.276
0.323
0.291
0.270
0.348
0.380
**
ns
**
**
**
25
0.534
0.379
0.335
0.229
0.150
0.251
0.205
0.201
0.314
0.102
0.231
0.245
0.206
0.178
0.192
0.283
0.042
26
0.594
0.533
0.494
0.408
0.398
0.487
0.405
0.441
0.444
0.329
0.452
0.441
0.353
0.398
0.456
0.541
0.487
0.382
**
**
**
**
28
0.604
0.444
0.402
0.359
0.364
0.451
0.397
0.387
0.438
0.240
0.374
0.385
0.199
0.354
0.399
0.474
0.452
0.367
0.328
**
29
0.491
0.413
0.333
0.279
0.244
0.334
0.291
0.279
0.332
0.151
0.289
0.252
0.177
0.251
0.269
0.416
0.365
0.267
0.287
0.240
30
0.500
0.429
0.405
0.352
0.314
0.405
0.350
0.310
0.345
0.229
0.315
0.276
0.244
0.334
0.353
0.453
0.390
0.305
0.291
0.252
0.089
31
0.360
0.361
0.289
0.289
0.252
0.323
0.288
0.241
0.280
0.180
0.306
0.305
0.148
0.225
0.252
0.376
0.356
0.279
0.350
0.234
0.127
33
0.517
0.338
0.330
0.224
0.264
0.355
0.301
0.292
0.315
0.183
0.251
0.258
0.177
0.253
0.288
0.399
0.360
0.271
0.350
0.244
0.133
34
0.458
0.388
0.323
0.281
0.267
0.339
0.314
0.298
0.333
0.190
0.276
0.252
0.182
0.268
0.284
0.395
0.358
0.287
0.304
0.270
0.110
35
0.493
0.392
0.361
0.292
0.300
0.373
0.320
0.327
0.375
0.205
0.284
0.279
0.248
0.295
0.304
0.398
0.370
0.306
0.322
0.277
0.146
37
0.488
0.383
0.325
0.263
0.280
0.342
0.293
0.298
0.349
0.190
0.284
0.250
0.194
0.269
0.298
0.399
0.374
0.306
0.344
0.256
0.125
38
0.419
0.388
0.263
0.245
0.211
0.286
0.279
0.286
0.294
0.164
0.308
0.278
0.156
0.233
0.229
0.366
0.335
0.265
0.312
0.270
0.113
39
0.507
0.402
0.344
0.313
0.322
0.389
0.322
0.310
0.394
0.240
0.299
0.276
0.254
0.295
0.337
0.443
0.408
0.351
0.406
0.322
0.130
42
0.469
0.315
0.366
0.277
0.298
0.373
0.275
0.281
0.326
0.225
0.297
0.250
0.234
0.251
0.301
0.426
0.382
0.305
0.387
0.313
0.177
43
0.602
0.407
0.443
0.372
0.391
0.445
0.341
0.408
0.473
0.225
0.313
0.251
0.307
0.350
0.407
0.465
0.435
0.373
0.493
0.423
0.232
44
0.400
0.330
0.241
0.206
0.177
0.273
0.213
0.216
0.262
0.129
0.253
0.201
0.118
0.168
0.196
0.341
0.306
0.220
0.282
0.191
0.081
45
0.534
0.429
0.368
0.333
0.309
0.344
0.298
0.298
0.404
0.224
0.360
0.321
0.243
0.292
0.293
0.450
0.433
0.345
0.388
0.320
0.139
*
367
(Continues)
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
3
TAXON 68 (2) • April 2019: 340–369
Appendix 4. FST values between all pairs of 41 × 21 (Table A) and 41 × 20 (Table B) populations of Campanula versicolor s.l. (lower diagonal) with pairwise significance after sequential Bonferroni corrections
(upper diagonal). No. = population numbering; ns = non-significant value; * = significant at P < 0.05; ** = significant at P < 0.01; *** = significant at P < 0.001. Population numbering and details are listed in
Appendix1.
Table A.
1
3
4
5
7
8
9
10
11
12
14
15
16
18
19
21
23
25
26
28
29
46
0.498
0.312
0.304
0.257
0.259
0.345
0.245
0.287
0.299
0.150
0.193
0.147
0.181
0.234
0.277
0.354
0.323
0.250
0.337
0.233
0.125
47
0.563
0.420
0.407
0.322
0.325
0.420
0.344
0.376
0.408
0.242
0.314
0.252
0.297
0.336
0.373
0.463
0.410
0.342
0.424
0.368
0.134
48
0.484
0.357
0.318
0.280
0.226
0.358
0.310
0.304
0.367
0.211
0.240
0.227
0.218
0.282
0.301
0.411
0.336
0.269
0.277
0.248
0.101
49
0.736
0.555
0.577
0.516
0.514
0.566
0.489
0.486
0.589
0.363
0.336
0.342
0.450
0.497
0.560
0.601
0.500
0.417
0.521
0.516
0.311
50
0.685
0.529
0.546
0.468
0.472
0.540
0.455
0.467
0.547
0.344
0.282
0.337
0.433
0.458
0.518
0.555
0.458
0.397
0.518
0.509
0.315
51
0.573
0.458
0.418
0.364
0.364
0.431
0.353
0.378
0.440
0.279
0.297
0.315
0.319
0.336
0.385
0.489
0.462
0.376
0.437
0.359
0.186
52
0.631
0.513
0.427
0.408
0.380
0.469
0.354
0.417
0.497
0.286
0.249
0.299
0.367
0.360
0.416
0.516
0.483
0.410
0.483
0.416
0.289
55
0.718
0.579
0.545
0.513
0.488
0.536
0.413
0.440
0.557
0.315
0.352
0.317
0.431
0.432
0.502
0.567
0.508
0.426
0.500
0.462
0.310
Table B.
No.
30
31
33
34
38
39
42
43
44
45
46
47
48
49
50
51
52
55
**
**
*
**
35
**
37
1
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
3
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
4
**
**
*
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
5
**
*
ns
**
**
*
**
**
**
**
**
**
**
**
**
**
**
**
**
**
7
**
ns
ns
**
**
**
*
**
**
**
**
*
**
**
**
*
**
**
*
**
8
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
9
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
10
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
11
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
12
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
14
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
15
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
16
**
**
*
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
18
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
19
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
21
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
23
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
25
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
26
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
28
**
**
*
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
29
**
**
*
**
**
**
**
**
**
**
ns
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
30
31
0.181
33
0.198
0.172
**
(Continues)
TAXON 68 (2) • April 2019: 340–369
No.
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
368
Appendix 4. Continued.
31
0.158
33
0.09
34
35
0.19
0.201
0.132
0.096
37
0.19
0.168
0.11
0.08
0.077
38
0.215
0.158
0.159
0.127
0.141
0.119
39
0.238
0.243
0.19
0.2
0.212
0.158
0.216
42
0.195
0.193
0.197
0.194
0.204
0.173
0.217
0.187
43
0.32
0.331
0.282
0.248
0.24
0.222
0.246
0.247
0.258
44
0.126
0.116
0.131
0.128
0.152
0.11
0.119
0.176
0.114
0.221
45
0.222
0.218
0.203
0.185
0.125
0.123
0.167
0.214
0.223
0.254
46
0.171
0.202
0.151
0.153
0.157
0.141
0.175
0.158
0.161
0.155
0.1
0.184
47
0.24
0.3
0.203
0.222
0.215
0.152
0.233
0.129
0.227
0.218
0.171
0.211
0.125
48
0.146
0.226
0.145
0.161
0.171
0.168
0.193
0.151
0.199
0.251
0.118
0.196
0.135
0.137
49
0.299
0.428
0.31
0.281
0.282
0.282
0.38
0.302
0.359
0.361
0.315
0.371
0.256
0.291
0.191
50
0.356
0.431
0.26
0.332
0.317
0.316
0.374
0.282
0.351
0.375
0.338
0.408
0.285
0.267
0.22
0.226
51
0.278
0.272
0.151
0.204
0.256
0.225
0.243
0.237
0.276
0.338
0.223
0.298
0.211
0.243
0.205
0.347
0.297
52
0.341
0.37
0.248
0.288
0.308
0.288
0.328
0.249
0.327
0.334
0.264
0.364
0.239
0.282
0.231
0.425
0.288
0.204
55
0.3
0.407
0.329
0.306
0.315
0.295
0.398
0.295
0.375
0.364
0.304
0.359
0.264
0.326
0.269
0.323
0.336
0.347
35
**
37
**
38
**
39
**
42
**
43
**
44
**
45
**
46
**
47
**
48
**
49
**
50
**
51
**
52
**
55
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
0.101
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
**
0.307
369
Janković & al. • New taxonomic concept in Campanula versicolor s.l.
Table B.
No.
30
34
0.171
TAXON 68 (2) • April 2019: 340–369
Appendix 4. Continued.
Keep reading this paper — and 50 million others — with a free Academia account
Used by leading Academics
Jeffrey Schwartz
University of Pittsburgh
Richard Cloutier
Université du Québec à Rimouski
Jorge Urbán
Universidad Autonoma de Baja California Sur (UABCS)
Dirk K F Meijer
University of Groningen
