PHYTOLOGIA BALCANICA 12 (2): 231–244, Sofia, 2006
Evolution of Veronica (Plantaginaceae)
on the Balkan Peninsula
Dirk C. Albach
Institut für Spezielle Botanik, Johannes Gutenberg-Universität Mainz, Bentzelweg 9b,
55099 Mainz, firstname.lastname@example.org
Received: October 10, 2005 ▷ Accepted: March 30, 2006
With more than 6500 species of native seed plants on the Balkans and almost a third of them endemic,
the Balkan Peninsula is known to be a place for diversification and formation of new species and an
important refugium during the Ice Ages. One plant group, which exemplifies this well, is the genus
Veronica (Plantaginaceae, formerly Scrophulariaceae). Four groups from this genus (V. subg. Stenocarpon;
V. subg. Chamaedrys; V. subg. Pseudolysimachium, V. alpina-complex) display putative tertiary relict
species, speciation within Pleistocene refugia and Pleistocene or Holocene speciation by hybridization and
polyploidization on the Balkan Peninsula. I here review earlier published results for these groups and present
new data. DNA sequence analyses from the nuclear ribosomal DNA (ITS) and plastid genome (trnLF region)
were examined so as to shed more light on the relationship of the species from the Balkans. In addition,
AFLP fingerprints were used to study V. subg. Pseudolysimachium, which exhibits limited DNA sequence
divergence. Results support the distinctiveness of taxa from the Balkans as a divergent group of plants on the
intra- and interspecific level. Limited resolution and support of the results further demonstrate the need for
another marker system to continue the study of evolution of these plants of the Balkan Peninsula.
Key words: AFLP, Balkan Peninsula, hybridization, phylogeography, polyploidy, Veronica
The plant-life on the Balkan Peninsula is richer than
any comparable area in Europe. There are more than
6500 species of native seed plants in the Balkans
(Turrill 1929; Horvat & al. 1974: 72–78; Polunin 1980:
22). Several factors are responsible for this diversity.
Apart from human impact, the combination of var-
ied topography and habitats at the crossroads of sev-
eral major floras (Central European, Mediterranean,
Anatolian, and Pontic), which had found favora-
ble refugia there to survive the Ice Age, is exception-
al. Almost one-third of the species is endemic to the
Balkans (Turrill 1929; Horvat & al. 1974: 72–78).
The Balkan Peninsula is known as an Ice Age ref-
ugium for many important plant species (Willis 1996;
Taberlet & al. 1998). The climate during the maxi-
mum glaciation did not allow continuous forest veg-
etation but was rather a steppe with scattered trees
and isolated pockets of forest (Frenzel 1968). Steppe
species included mainly species from Asia and led to
a major influx of Asian taxa into the European flora
(Frenzel 1968; Ozenda 1988). In addition to being a
source for the recolonisation of Europe (Taberlet & al.
1998; Hewitt 2000; Schmitt & Seitz 2002), the Balkan
Peninsula was a place for further diversification and
speciation, given its richness in endemic species.
The Pleistocene climate shifts have led to cycles of
isolation of the plants in refugia and to their subse-
quent spread in more favorable times for a more con-
tinuous distribution. Whereas the more temperate el-
ements were restricted to refugia in glacial times, the
Albach, D. & al. • Veronica on the Balkan Peninsula
cold-adapted plants had likely a more or less continu-
ous distribution during these times and were restrict-
ed to the refugia during the interglacials. Many alpine
plants might have originated in the Pleistocene in re-
stricted alpine regions and had diversified owing to
genetic drift (Zhang & al. 2001). Those taxa that were
restricted in the Pleistocene to small refugia and after-
wards expanded often did so into areas inhabited by
plants from other refugia, which provided ample op-
portunity for hybridization and formation of new spe-
cies (Hewitt 2000).
Veronica is a genus of the Plantaginaceae sensu
APG (1998) and Albach & al. (2005), to which have
been transferred most of the well-known genera of
Scrophulariaceae from the Northern Hemisphere flo-
ras. Veronica includes about 500 species (Albach & al.
2004a). It is distributed over much of the Northern
Hemisphere and beyond and is ecologically highly di-
verse, with species living in aquatic to dry steppe hab-
itats, from sea level to high alpine regions. This diver-
sity may explain the interest Veronica has evoked for a
long time. Recently, molecular techniques and phylo-
genetic analyses have been applied to Veronica and re-
lated genera (Wagstaff & Garnock-Jones 1998; Albach
& Chase 2001, Albach & al. 2004b, c). These studies
have helped revolutionise our ideas about the evolu-
tion of the genus and have led to a new phylogenet-
ic intrageneric classification of Veronica (Albach & al.
2004a). Combined with the vast amount of informa-
tion from other aspects of their biology, we have now
a much better understanding of how major groups in
Veronica are delimited, related, and evolved. In the
present article I summarise what we currently know
about the evolution of the species of Veronica na-
tive to the Balkan Peninsula. Apart from a review of
earlier publications, I present new sequence data on
the species from V. subg. Stenocarpon and V. subg.
Chamaedrys and AFLP-fingerprint data on V. subg.
Materials and Methods
DNA Extraction, Amplification, Sequencing
The total genomic DNA was extracted from herbarium
material and silica gel-dried samples according to the 2x
CTAB procedure of Doyle and Doyle (1987) and then
washed twice with 70 % ethanol or using NucleoSpin
plant DNA extraction kits (Macherey-Nagel; Düren,
Germany) following the manufacturer’s specifications.
Nine sequences (seven for the ITS-region, two for the
trnLF-region) are used here for the first time.
The trnLF region was amplified with primers c and
f of Taberlet & al. (1991) and includes the trnL intron,
3´ trnL-exon and the trnL-trnF spacer. ITS sequences
were amplified and sequenced using the primers 17SE
(Sun & al. 1994) and ITS4 (White & al. 1991) and in-
clude ITS1, 5.8S rDNA and ITS2. PCR products were
either separated on 1 % TBE-agarose gels; fragments
corresponding to the expected size were excised and
cleaned using the QIAquick™ PCR purification and gel
extraction kit (Qiagen GmbH, Hilden, Germany) fol-
lowing the manufacturer’s protocols or cleaned using
the Macherey-Nagel PCR purification kit (Macherey-
Nagel; Düren, Germany) following manufacturer’s
specifications. Sequencing reactions (10 μl) were car-
ried out using one μl of the BigDye Terminator Cycle
Sequencing mix (Applied Biosystems Inc.). Both
strands were sequenced on a Prism 377 automated se-
quencer (Applied Biosystems Inc.). Sequences were
assembled and edited using Sequence Navigator™
(Applied Biosystems Inc.) or Sequencher™4.1 (Gene
Codes Corp., Ann Arbor, MI, USA). Assembled se-
quences were manually aligned prior to analysis.
Origin, voucher information and GenBank accession
numbers for all sequences used in this study are giv-
en in Table 1.
AFLP profiles were generated following the estab-
lished procedures (Vos & al. 1995) and according to
the PE Applied Biosystems (1996) protocol with on-
ly minor modifications (Tremetsberger & al. 2003).
Genomic DNA (approx. 500 ng) was digested with
MseI (New England BioLabs) and EcoRI (Promega)
and ligated (T4 DNA-Ligase; Promega) to double-
stranded adapters in a thermal cycler for 2hr at 37°C.
Preselective amplification (5 μL reactions) was per-
formed using primer pairs with a single selective nu-
cleotide, MseI-C and EcoRI-A. Selective amplifica-
tions were performed with the primer combinations
MseI-CAT/EcoRI-ACT, MseI-CTC/EcoRI-AAG, and
MseI-CAT/EcoRI-ACC. Selective amplification prod-
ucts were run in a 5 % denaturing polyacrylamide gel
with an internal size standard (GeneScan®-500 [ROX],
PE Applied Biosystems) on an automated DNA se-
Phytol. Balcan. 12(2) • Sofia • 2006
quencer (ABI 377). Polyacrylamide gels run on au-
tomated sequencers increase the resolution and de-
crease the probability of scoring fragments of similar
size as homologues. Fragments from the polyacryla-
mide gel were analyzed using the ABI Prism Gene-
Scan® 2.1 Analysis Software (PE Applied Biosystems)
and Genographer 1.1 (Benham 1998). Peaks (i.e. frag-
ments) were scored manually as present (1) or absent
(0) in a readable region of bands from 50 to 500 base
pairs in length and used to construct a presence/ab-
sence data matrix with individual plants in rows and
bands in columns. Origin and vouchers for all speci-
men used in the AFLP analysis are given in Table 2.
Data Analysis – sequences
Data matrices were analysed separately and com-
bined using maximum parsimony and maximum
likelihood criteria in PAUP*4.0b10 (Swofford 2002).
Potentially parsimony informative gap characters
were coded as present/absent. Heuristic parsimony
searches were conducted using 10 replicates, start-
ing from random trees with tree bisection reconnec-
tion (TBR) branch swapping and MulTrees in effect
and no tree limit. Parsimony bootstrap analyses in-
volved 1000 replicates with the same search param-
eters as above but simple taxon addition. Maximum
Table 1. Species name, origin, voucher information and GenBank accession number of accessions (ITS/trnLF).
Veronica jacquinii Baumg. – cult. BG Bonn (Albach 70, WU – AF313000/AF513341); V. scutellata L. – Waldviertel, Austria (Dobes 7026,
WU – AF509805/AF486393); V. serpyllifolia L. – Bonn, Germany (Albach 64, WU – AF313017/AF486400); V. triloba Opiz – Aphrodisias,
Turkey (Albach 242, WU – AF509803/AF486366); V. triphyllos L. – Aphrodisias, Greece (Albach 244, WU – AF509795/AF486396);
Data set of V. subg. Stenocarpon (Boriss.) M.M. Mart. Ort., Albach, & M.A. Fisch
Paederotella pontica (Rupr.) Kem.-Nath. – Georgia (Sachokia, 01.9.1951, TBS – AF515214/AF486382); Veronica ciliata Fisch. – Qinghai,
China (Miehe & al., 98-33313, GOET – AF515215/AF486385); V. densiflora Ledeb. – Altai (Staudinger s.n., SALA – AY741521/AY776282); V.
erinoides Boiss. & Sprun. – Gioua Mts, Greece (Hagemann, Scholz & Schmitz 461, SALA – AY741523/n.a.); V. fruticans Jacq. – Scotland, UK
(Viv Halcro VH030, K – AY144462/n.a.); V. fruticulosa L. – cult. BG Bonn (Albach 71, WU – AF313004/AF486383); V. lanosa Royle ex Benth. –
Pakistan (Schickhoff 1377, GOET – AY540868/AY486442); V. lanuginosa Benth. – 13 km east of Mt Everest, Nepal (Dickore 6482, GOET –
AF509793/AF486386); V. macrostemon Bunge ex Ledeb. – Altai (Staudinger AL23-18, SALA – AY741522/AY486441); V. mampodrensis Losa &
P. Monts. – Velilla de Río Carrión , Spain (Martínez Ortega 713, SALA – DQ227331/ DQ227337); V. mexicana S. Watson – Mesa el Campanero,
Sonora, Mexico (Fishbein 2586, TEX – DQ227332/n.a.); V. monticola Trautv. – Georgia (Ivanisvili 26.07.1983, WU – DQ227333 & DQ227334/
n.a.); V. nummularia Gouan – Tosses, pr. La Molina, subida al pico pico Niu d‘Aliga, Spain (Martínez Ortega 718, SALA – DQ227335/n.a.);
V. saturejoides Vis. subsp. kellereri (Degen & Urum.) P. Monts. – Mt Pirin, Bulgaria (Albach 558, WU – AY144461/AY486450); V. thessalica
Benth. – Olymbos, Greece (Raus & Rogl 5072, SALA – AF509792/AF513343).
Data set of V. subgenus Chamaedrys (W.D.J. Koch) M.M. Mart. Ort., Albach & M.A. Fisch.
Veronica arvensis L. – Stromberg, Germany (Albach 147, WU – AF313002/AF486380); V. arvensis – Richmond Park, UK (M. Sheahan 9, K –
DQ227328/n.a.); V. chamaedryoides Bory & Chaub. – Peloponnes, Greece (Albach 393, WU – AF673611/AY673631); V. chamaedrys subsp.
chamaedrys L. – Homefield Wood, UK (M. Fay 149, K – DQ227329/n.a.); V. chamaedrys subsp. chamaedrys – cult. RBG Kew, ex Norway
(Albach 121, K – AF313003/ AF486377); V. chamaedrys subsp. micans M.A. Fisch. – Austria (Schönswetter 2567, WU – AY673616/AY673632);
V. krumovii (Peev) Peev – Eastern Rhodopes, Bulgaria (Albach 484, WU – AY673612/AY673633); V. laxa Benth. – Pakistan (Dickoree 13042,
GOET – AY673613/AF486378); V. magna M.A. Fisch. – Orbetia, Georgia (Albach 360, WU – AY6736152/AY673634); V. micrantha Hoffmanns. &
Link – Portugal (Martínez Ortega 1754, SALA – DQ227330/ DQ227336); V. verna L. – Bad Kreuznach, Germany (Albach 149, WU – AF509789/
AF486379); V. vindobonensis (M.A. Fisch.) M A. Fisch. – cult. BG Wien (M.A. Fischer s.n., WU – AY673614/AF510426).
Table 2. Samples (with vouchers) used in the AFLP analysis of V. subg. Pseudolysimachium.
Veronica bachofenii Heuff. – cult. Bot. Gard. Wien, ex Bot. Gart. Halle (Albach 978, MJG); V. barrelieri Schott ex Roem. & Schult. – Bulgaria
(Vladimirov s.n., WU); Croatia (Schneeweiss & Schönswetter, WU); Croatia (5 individuals, Martínez Ortega 908, SALA); V. crassifolia
Wierzb. ex Heuff. – Demogled-Herkulesbach, Romania (Köster s.n., WU); V. daurica Steven – cult. Bot. Gard. Jena (Albach s.n., WU); V.
incana L. – cult. Bot. Gard. Bonn (Albach 155, WU); V. longifolia L. – cult. Bot. Gard. Bonn (Albach 66, WU); Kühkopf, Germany (Fay et al.
s.n., K); Frey´s Is., UK (Sheahan 48, K); Altai (Tribsch 31.7.2002, WU); V. orchidea Crantz – Wien-Salmansdorf, Austria (Fischer 21.7.2000,
WU); Montana, Bulgaria (Albach 540, WU); White Carpathians, Czech Republic (Köster s.n., WU ); Klausenburg, Romania (Köster s.n.,
WU); V. porphyriana Pavlov – Altai (Tribsch 2002, WU); V. schmidtiana Regel – Kushiro-mashi, Hokkaido, Japan (Umezawa 20130, WU);
V. spicata L. – Stanner Rock, UK (5 samples, Jones s.n., K); Avon Gorge, UK (2 samples, M. Fay s.n., K); Uppsala, Sweden(Thulin 10035,
UPS); Neusiedler See, Austria (Krefft s.n., WU); Brey, Germany (Albach 224, WU); Thorenburg canyon (Koester s.n., WU); cult. RBG Kew,
ex Hainburg, Austria (Kew 1970-759 – Chase s.n., K); cult. Bot. Gard. Bonn (Albach 65, WU); ex Knauber, Bonn (Albach 214, WU); “Nana
Alba” cult. RBG Kew (Kew 1979-6053 – Chase s.n., K); „Sarabande“ cult. RBG Kew (Kew 1973-21579 – Chase s.n., K); Ukraine (2 samples,
WU); V. spuria L. – Burgenland, Austria (Fischer 04.06.2000, WU).
Albach, D. & al. • Veronica on the Balkan Peninsula
likelihood analyses were conducted using a mod-
el of sequence evolution as inferred by Modeltest
3.06 (Posada & Crandall 1998) using the Akaike
Information Criterion, three to six replicates of ran-
dom taxon addition starting from random trees with
TBR branch swapping, MulTrees in effect and no tree
limit. Likelihood bootstrap analyses for the com-
bined dataset involved 500 replicates with the same
search parameters as above but simple taxon addi-
tion. For analyses of V. subg. Stenocarpon, sequenc-
es of V. scutellata, V. serpyllifolia, V. triloba, V. triphyl-
los, V. jacquinii, and V. chamaedrys were designated
as outgroups. For analyses of V. subg. Chamaedrys,
the same taxa were designated as outgroups ex-
cept for replacing V. chamaedrys with V. thessalica.
To evaluate different phylogenetic hypotheses in V.
subg. Stenocarpon, the likelihood-based Shimodaira-
Hasegawa test (with the same model used in the max-
imum likelihood analysis and RELL optimization)
and parsimony-based Templeton test were used as
implemented in PAUP4.0b (Swofford 2002) with the
combined data set.
Data Analysis – AFLP
AFLP fingerprints were examined
using neighbour-joining and prin-
cipal coordinate analysis (PCoA).
Nei 1987) and neighbour-join-
ing bootstrap analyses using the
Nei-Li-distances have been con-
ducted in PAUP4.0b10 (Swofford
2002). Principal coordinate analysis
(PCoA) was conducted using Dice
genetic distances between all pairs of
accessions as implemented in R4.0
(Casgrain & Legendre 2001) and the
first two axes projected in two two-
Fig. 1. Results from the analysis of the ITS-
dataset. Left, optimal tree from the maximum
likelihood analysis. Right, strict consensus of
14 most parsimonious trees. Numbers above
the branches indicate bootstrap support.
Veronica subg. Stenocarpon
The ITS-dataset included 21 taxa with 745 aligned charac-
ters, 91 of them potentially parsimony-informative and no
indel character, whereas the trnLF-dataset with 17 taxa in-
cluded 998 aligned characters, 43 of them potentially parsi-
mony informative with one indel character. The combined
dataset included 17 taxa and 1743 characters with 135 of
them potentially parsimony-informative. Modeltest (Posada
& Crandall 1998) chose the GTR + Γ-model (Γ = 0.18) for the
ITS data set, the TVM + Γ-model model (Γ = 0.74) for the
trnLF dataset and the GTR + I + Γ-model (Γ = 0.72) for the
combined analysis as the optimal model. The 14 most parsi-
monious trees of the ITS-analysis required 365 steps (Fig. 1;
CI = 0.72; RI = 0.59), with the 1755 most parsimonious tree
from the trnLF-analysis requiring 239 steps (not shown;
CI = 0.90; RI = 0.69). The most likely tree of the maximum
likelihood analysis of the ITS-dataset (Fig. 1) required six
more steps under parsimony, whereas that from the analy-
Phytol. Balcan. 12(2) • Sofia • 2006
sis of the trnLF-dataset being one of the most par-
simonious trees (ITS ML vs. MP: -ln = 2858.12 vs.
2860.88-2862.06; trnLF: -ln = 2792.57). The com-
bined dataset had a single optimal parsimony tree
requiring 593 steps (Fig. 2; CI = 0.79; RI = 0.72;
-ln = 5583.40) and the best tree chosen by the like-
lihood analysis required 594 steps under parsimo-
ny (-ln = 5579.49).
Veronica subg. Stenocarpon is a well sup-
ported group by both the ITS-analysis and
the combined analysis (Figs 1, 2; >98 % BP).
Support for internal relationships within V.
subg. Stenocarpon is low, with internal branch-
es being short (Fig. 1). The combined analysis
(Fig. 2) reveals a clade of Asian species with
two Eastern Himalayan and the two Siberian
species, forming each well supported (>75 %
bootstrap percentage (BP)) sister-groups.
Support for other internal groups is lacking.
For example, V. densiflora groups in the results
of the ITS-analysis (Fig. 1) with the Western
Himalayan V. lanosa and the Mexican species,
V. mexicana, for which a trnLF-sequences is still miss-
ing and which therefore was not included in the trn-
LF- and combined analysis, but the relationship in the
ITS-analysis is not supported by the bootstrap analy-
sis. Results from the parsimony analysis of the trn-
LF dataset are largely unresolved and even some trees
in which the subgenus is not monophyletic are equal-
ly parsimonious (result not shown). The only clade re-
trieved in this analysis is the sister-group relationship
of V. ciliata and V. lanuginosa. The tree resulting from
the maximum likelihood analysis of the trnLF-data-
set is identical with the maximum likelihood analysis
of the combined dataset (Fig. 2) and is therefore also
not shown. Surprisingly, the rooting inferred by par-
simony and maximum likelihood analyses differs dra-
matically. Whereas parsimony analyses from both ITS
and the combined dataset (Figs 1, 2) reveal a grade of
Balkan-Iberian species (V. thessalica, V. saturejoides,
V. mampodrensis) as sisters to the rest of the subgenus,
all maximum likelihood analyses (Figs 1, 2) support a
sister-group relationship of V. fruticulosa (plus V. fru-
ticans in the ITS-analysis) to the rest of the subgenus.
All Central Asian-Himalayan species form a clade with
the exception of V. lanosa. Shimodaira-Hasegawa tests
and Templeton tests were insignificant (p = 0.29 and
p = 0.80, respectively). Increasing outgroup sampling
did not lead to different results (Albach, unpubl.).
Fig. 2. Unrooted phylogram from the maximum likelihood analysis of the
Numbers along the branches indicate parsimony branch lengths/ parsi-
mony bootstrap percentage/ maximum likelihood bootstrap percentage/
parsimony bootstrap percentage from the analysis of the trnL-F-dataset
only. Asterisk indicates branch not present in the most parsimonious tree.
ML – Maximum likelihood; BP – bootstrap percentage.
Veronica subg. Chamaedrys
The ITS-dataset included 18 taxa with 743 aligned
characters, 111 of them potentially parsimony-in-
formative with four indel characters, whereas the
trnLF-dataset included 16 taxa with 988 aligned
characters, 48 of them potentially parsimony in-
formative with two indel characters and the com-
bined dataset with 16 taxa had 1731 characters
and 151 of them potentially parsimony-informa-
tive. Modeltest (Posada & Crandall 1998) chose the
GTR + I + Γ-model (Γ = 0.38) for the ITS data set, the
GTR + Γ-model (Γ = 0.45) for the trnLF dataset and
the TIM + I + Γ-model (Γ = 0.63) for the combined
data set as the optimal model. The ten most parsi-
monious trees of the ITS-analysis required 367 steps
(Fig. 3A; CI = 0.73; RI = 0.73), with the four most par-
simonious tree from the trnLF-analysis requiring 223
steps (Fig. 3B; CI = 0.89; RI = 0.81). The most likely
trees (Fig. 3A, B) were just one step longer under par-
simony in both cases (ITS ML vs. MP: -ln = 2834.53
vs. 2835.07-2837.93; trnLF: ML vs. MP: -ln = 2618.47
vs. 2619.47). The combined dataset had a single op-
timal parsimony tree requiring 593 steps (CI = 0.79;
RI = 0.72; -ln = 5583.40) and the best tree chosen by
the likelihood analysis required 594 steps under par-
simony (Fig. 3C; -ln = 5579.49).
Albach, D. & al. • Veronica on the Balkan Peninsula
All analyses support four
groups in V. subg. Chamaedrys: V.
arvensis, V. verna, the Asian per-
ennial species, and the European
perennials. Support for V. arven-
sis as sister to the rest of the sub-
genus is high (99 BP in combined
analysis). The branching order of
the other three groups varies with
all three possibilities realized in at
least one analysis. Monophyly of
the perennials is, however, only
shown in the parsimony analysis
of the ITS dataset (Fig. 3A), with
most analyses supporting the sis-
ter-group relationship of V. ver-
na and the perennial Europeans
(Fig. 3B, C; 77 BP in maximum
likelihood analysis of combined
dataset). Within the European
perennials, the Balkan endemics,
V. chamaedryoides and V. krumov-
ii, are consistently occupying first
branches but support for relation-
ships within this group is low.
The AFLP dataset consists of 42
taxa and 732 bands scored with 73
of them constant. The number of
AFLP fragments per taxon ranged
between 97 and 159 bands. Ploidy
level is not known for the individ-
uals used in this study but there
was a tendency for taxa known to
be diploid (2n = 34) to have fewer
bands (e.g. V. porphyriana, which
had the lowest number of bands) than tetraploid
(2n = 68) taxa (V. spicata, which is mostly tetraploid
had the highest number of bands). Intraspecific varia-
tion was large in V. spicata (n = 104–159). Intraspecific
variation in V. barrelieri was significant with respect
to geographical origin with the Bulgarian sample hav-
ing markedly fewer bands than the Croatian samples
(109 vs. on average 140). No species-specific frag-
ments were found for species for which individuals
from more than one population was analyzed.
Fig. 3. Phylograms derived from the maximum likelihood analyses of V. subg. Chamaedrys.
Numbers above the branches indicate bootstrap support.
D, German accession; N, Norwegian accession; UK, English accessions; X, branch not pres-
ent in most parsimonious tree; A, optimal tree for the ITS-dataset. Dashed line indicates
branch present in most parsimonious tree; B, optimal tree for the trnL-F-dataset; C, optimal
tree for the combined dataset.
Coordinate Analysis (PCoA) did not cluster samples
into species specific groups but show similar results.
Hybridization and polyploidy are common in V. subg.
Pseudolysimachium (see below). Relationships based
on hybridization are difficult to show on a phylogenet-
ic tree. Therefore, only the results from the PCoA are
shown (Fig. 4). The x-axis explains 17.0 %; the y-axis
explains 9.0 % of the variation. Both are low values and
further emphasize the complexity of the results.
analysis and Principal
Phytol. Balcan. 12(2) • Sofia • 2006
Veronica in Europe
Based on detailed morphological analyses (e.g.,
Fischer 1972, 1973b, 1975a, b, 1978, 1991; Trávnícek
1998, 2000; Martínez-Ortega & Rico 2001; Albach &
Fischer 2003), we currently have a well-founded es-
timate of species diversity of Veronica, especially in
Europe. Molecular analyses have been contributing
to this (Martínez-Ortega & al. 2004; Albach & al. in
press; Albach, submitted). On the base of these anal-
yses our estimate that about 80 species of Veronica,
representing 10 of 13 subgenera recognized by
Albach & al. (2004a), are found in Europe. About 40
species found in nine different subgenera are endem-
ic to Europe. On the Balkan Peninsula, we can find
about 60 species of Veronica, with 14 of them endem-
ic (V. chamaedryoides, V. contandriopouli, V. crinita,
V. erinoides, V. glauca, V. krumovii, V. oetaea, V. or-
belica, V. orbiculata, V. rhodopea, V. sartoriana, V. sa-
turejoides, V. thessalica, V. turrilliana). Some of these
species have been investigated in the present study
using DNA sequence and AFLP data. Furthermore,
I discuss results from earlier analyses shedding more
light on the evolution of Veronica on the Balkan
Veronica subg. Stenocarpon –
relicts of alpine regions
Veronica subg. Stenocarpon is probably the most sur-
prising group revealed by molecular phylogenetic
analysis. The morphological diversity and especially
its biogeographical distribution pattern make it a fas-
cinating group to study. Veronica subg. Stenocarpon
includes approximately 25 species, most from
Central Asia but also seven species from Europe plus
the Mexican endemic representative of the genus,
V. mexicana. This Mexican-Himalayan disjunction
appearing in the analysis of ITS-sequences (Fig. 1),
however without bootstrap support, merits further
study. Among the European species most are endem-
ic in the Balkans, often known only from one or very
few localities, such as V. saturejoides subsp. kellere-
ri (Mt Pirin), V. saturejoides subsp. munellensis (Mt
Mnela), and V. thessalica (Mt Gjaliqa e Lumes; Mt
Olymbos; Mt Jakupica; Sar Planina; Mt Koritnik).
Veronica erinoides is known from only five locali-
ties (Giona, Vardousia, Lidorikiou Ori, Parnassos,
Fig. 4. First two axes of the principal coordinate analysis of V. subg. Pseudolysimachium.
13 • Phytol. Balcan. 12(2) • 2006
Albach, D. & al. • Veronica on the Balkan Peninsula
Killini). Veronica thessalica and V. erinoides were
thought to be synonyms (e.g., Stroh 1942) but even
before the results presented here (Fig. 1) a detailed
morphological analysis had shown that one is hav-
ing terminal and the other only lateral inflorescenc-
es (Fischer 1969), a character formerly used for di-
viding sections in Veronica. Molecular analyses on
the genus level (Albach & Chase 2001; Albach & al.
2004c) has shown that the character is labile and not
useful for delimiting natural groups in Veronica.
So far, all species for which chromosome num-
bers have been counted are diploid, although apart
from the seven European species only one Asian spe-
cies has been investigated (Albach & al. submitted).
Sequence analysis using parsimony (Figs 1, 2) revealed
that the Balkan endemics V. saturejoides, V. thessali-
ca and V. erinoides, together with V. mampodrensis
from the Iberian Peninsula constitute a paraphyletic
grade of consecutive sisters to the Asian species of the
subgenus, plus V. mexicana and three more European
species (V. fruticans, V. fruticulosa, V. nummularia).
Veronica mampodrensis has been included here for the
first time in a phylogenetic analysis. Its position in V.
subg. Stenocarpon is in general agreement with mor-
phology, although Fischer & Fischer (1981) suggest-
ed a closer relationship with V. fruticans and V. fru-
ticulosa. Based on their restricted distribution areas
(in comparison with other species of the subgenus),
diploid ploidy level and their position in parsimony
analyses, the Balkan endemic species could be consid-
ered to be relictual species, in accord with the idea of
the Balkan Peninsula as a refugium for plant groups
since the Tertiary (Horvat & al. 1974; Willis 1996;
Taberlet & al. 1998). However, support for the rela-
tionships is low and, more important, maximum like-
lihood analyses differ with regard to the topology of
V. subg. Stenocarpon in V. fruticulosa being a sister to
the rest of the subgenus (Figs 1, 2). The reason for this
difference is solely a difference in rooting the tree, be-
cause unrooted phylogenies from both kinds of anal-
yses are identical (Fig. 2). I evaluated several reasons
for the difference in rooting. For example, addition of
other outgroup taxa to the dataset does not change the
results (results not shown). Rather the lack of suffi-
cient informative characters leads to spurious differ-
ences in the results. Thus, the most parsimonious tree
is not rejected as significantly worse by the likelihood-
based Shimodaira-Hasegawa test, and the most like-
ly tree is not rejected as significantly worse by the par-
simony-based Templeton test. This lack of sufficient
informative characters currently renders it impossi-
ble to resolve the mysteries of the biogeographical and
morphological evolution in this clade. Faster evolving
DNA regions will be necessary for future analyses of
the evolution of the group.
Furthermore, no calibration point for molecular dat-
ing is available for Veronica, and extensive rate hetero-
geneity in the genus (Albach & Müller in prep.) renders
the common substitution rate unusable. Therefore, I
am currently unable to substantiate the claim that the
first branching event in V. subg. Stenocarpon repre-
sents a Tertiary event. Biogeographically, the situation
resembles Ramonda (Gesneriaceae), which is likewise
considered a Tertiary relict and found in the Pyrenees
and the Balkan Peninsula (Meyer 1970), although, in
contrast to Veronica, Gesneriaceae are of tropical ori-
gin. Work is in progress studying phylogeographical
patterns within the Balkan endemics V. saturejoides, V.
thessalica, V. erinoides, the Iberian endemics V. num-
mularia and V. mampodrensis and the more wide-
spread V. fruticans, so as to investigate the hypothesis
that the endemic species are relictual species that had
failed to disperse out of their glacial refugia, where-
as V. fruticans had managed to spread from a glacial
distribution area into and close to the Alps, north to-
wards to Scandinavia. Additional sequence data will
be helpful to estimate the genetic variation within the
Balkan endemic species in order to differentiate be-
tween the hypotheses of recent bottlenecks or long
persistence in refugial areas, so as to explain the limit-
ed distribution of these taxa and to help protect these
rare alpine species.
Veronica subg. Chamaedrys – Diploid taxa
marking hypothesized forest refugia
Veronica subg. Chamaedrys consists of four different
groups: V. arvensis (incl. V. sartoriana = V. subsect.
Microsperma (Römpp) Stroh), V. verna (most likely
together with V. dillenii and V. brevistyla; = V. subsect.
Microspermoides Albach1), V. laxa and V. magna ( = V.
1 Veronica subsectio Microspermoides Albach, subsect. nov. –
Typus: V. verna L. Spec. Pl. 1: 14, 1753; lectotypus (ab E. Fischer
1997 designatus): LINN 26.63. Annua aut raro biennis, −30 cm,
erecta, simplex. Folium sessile, pinnati- vel palmatipartitum
vel –lobatum. Inflorescentia terminalis, multiflora, glandulosa.
Pedicellus brevior quam calyx. Capsulae obcordatae, numquam
glandulosae. Semina plana. Numerus basalium chromosomatium
x = 9. Planta inutilis in Europa et Asia diffusa.
Phytol. Balcan. 12(2) • Sofia • 2006
subsect. Asiachamaedrys Albach2), and the European
perennial species (V. Chamaedrys, V. vindobonensis,
V. chamaedryoides, V. krumovii, and V. orbelica; = V.
subsect. Multiflorae Benth.). Surprisingly, the peren-
nial species do not form a monophyletic group, ex-
cept for the parsimony analysis of the ITS dataset
(Fig. 3A). Thus, a likely scenario is that first a for-
merly widespread Eurasian perennial taxon split in-
to a Mediterranean and an Asian taxon, and V. verna
(and relatives) evolved from the Mediterranean group.
Veronica verna is nowadays widespread across Eurasia
and nothing is known about its place of origin to con-
firm this hypothesis.
Veronica subsect. Multiflorae includes prominent
species of forest vegetation, mainly along the forest
edges, in open places and clearings. Originally con-
sidered one species, the group consists of five species
and three additional intraspecific taxa. Whereas V.
chamaedrys subsp. chamaedrys is a tetraploid plant, all
others are diploid. They differ not only in their ploi-
dy level but also in the distribution range, with the
tetraploid widespread throughout Europe and the on-
ly one that has reached North Europe. The diploids
occur in South, especially South and Central Europe,
and are mostly confined to small areas that have been
assumed as Pleistocene refugia. Veronica chamaedry-
oides is endemic to Greece, where it is fairly wide-
spread, and found in open forests, meadows and rocky
slopes up to 1900 m. Greece is commonly suggested
as a Pleistocene refugium for many submediterranean
taxa (e.g. Huntley & Birks 1983; Tzedakis & al. 2002).
Veronica krumovii is endemic to Bulgaria. It is mor-
phologically separated but its ecology and exact dis-
tribution has not been studied adequately (Mirek &
Fischer 1986). Although currently not restricted to the
western Black Sea Coast, a suggested Pleistocene refu-
gium by Huntley & Birks (1983), it could have expand-
ed from there to its current distribution area. Veronica
orbelica is another Bulgarian endemic. It is found in
Southwest Bulgaria and is morphologically close to V.
vindobonensis (Mirek & Fischer 1986). It grows in dri-
er habitats and may have originated in the drier inte-
rior part of Bulgaria. Unfortunately, no authenticated
2 Veronica subsectio Asiachamaedrys Albach, subsect. nov. –
Typus: V. laxa Benth. Scroph. Ind.: 35, 1835; typus: K n.v. Perennis.
35−90 cm, erecta. Caulis ubique pubescens. Folia brevipetiolata,
ovata, grosse dentate, plus quam 25 cm. Inflorescentia lateralis,
plus quam 10 cm longa, multiflora. Calyx longior quam pedicel-
lus, plus quam 4 mm. Corolla caerulea. Capsula obcordata, bre-
vior quam calyx, base cuneata, glabra.
material of V. orbelica was available for this study be-
cause much of the material seen in nature in Bulgaria
was intermediate between V. orbelica and V. vindobo-
nensis. Clear delimitation of these two species will be
an important task for the future.
Veronica vindobonensis is the most widespread
diploid taxon occurring from Northwestern Anatolia
to South Germany. It is probably adapted to the dri-
est types of forests within the group. Its wide distri-
bution argues either for rapid spread from a refugium
(possibly furthest to the north in Hungary, accord-
ing to Willis & al. 2000) after the last glaciation, or a
more widespread distribution even in the Pleistocene.
It sometimes reaches subalpine habitats in Bulgaria
(Albach pers. obs.), which might contend for the latter.
Veronica chamaedrys subsp. micans is mostly known
from the Austrian and German northern calcare-
ous Alps and is the only taxon that can be considered
truly subalpine. Its distribution is especially intrigu-
ing, because it is exactly confined to the hypothesized
Pleistocene refugia in the northeastern Alps (Tribsch
& Schönswetter 2003). Another diploid taxon is found
in South Austria and is informally called V. chamae-
drys „carintho-stiriaca“ (Fischer 1973a, pers. comm.).
It is similar to V. chamaedrys subsp. micans occur-
ring on drier habitats but is morphologically distinct
(Fischer 1973b) and may have survived in a forest re-
fugium in the southeastern Alps (Bennett & al. 1991).
Another diploid population can be found further
south in Dalmatia but has been not studied intensive-
ly (Fischer unpubl.). The tetraploid V. chamaedrys has
the widest distribution of all taxa. It occurs through-
out Europe, with the exception of the Mediterranean
islands and the Arctic. In Northern Greece it barely
overlaps with V. chamaedryoides (Fischer 1991).
Results from sequence analysis suggest that the
southern diploid taxa of the group, V. krumovii and
V. chamaedryoides, were not involved in the origin of
tetraploid V. chamaedrys subsp. chamaedrys but were
isolated in their Pleistocene refugia. Unfortunately,
molecular dating is not possible in Veronica and there-
fore the age of divergence is not inferable. Several sce-
narios regarding the origin of tetraploid V. chamaedrys
are possible on the base of sequence data. Veronica
chamaedrys subsp. micans is very likely a parent of
it, based on the high similarity of their ITS-sequenc-
es (Fig. 3A.). The other parent could be an unsampled
diploid taxon, possibly the “stiriaca-carinthiaca”-type
or the “dalmatica”-type, based on the fact that none of
Albach, D. & al. • Veronica on the Balkan Peninsula
the sampled diploid taxa had the tetraploids plastid-
sequence (Fig. 3B), although an autopolyploid event
cannot be excluded. Another possibility to be consid-
ered is the polytopic origin, supported by morpho-
logical variation in the tetraploid cytotype (Fischer
1973a). More intensive sampling will be needed to re-
veal the parents of tetraploid V. chamaedrys. So far,
we can be only reasonably certain that the tetraploid
V. chamaedrys sampled in our analysis had ancestors
that lived close to the ice-sheet during the Pleistocene.
Such a scenario would support Stebbins’ (1984) sec-
ondary-contact hypothesis, in which polyploids arise
at the margin of distribution. Thus, diploid taxa sur-
vived in their Pleistocene refugia and, after the ice re-
treated, spread out into the formerly glaciated area
and encountered other diploid survivors, with which
they formed polyploid hybrids. Such kinds of neo-
polyploids are known to surpass their ancestors in in-
vasion potential, especially in habitats that have be-
come available after the retreat of the Pleistocene ice
sheets (Ehrendorfer 1965, 1980). The involvement of
the subalpine V. chamaedrys subsp. micans in the or-
igin of V. chamaedrys rather than a true forest spe-
cies, such as V. krumovii, further supports such a sce-
nario. Consequently, the Balkan endemic species of
V. subsect. Multiflorae, in the sense applied here, rep-
resent relicts from Pleistocene refugia, which did not
contribute to the recolonization of Central and North
Europe – a situation resembling that of many animals,
such as the brown bears (Taberlet & Bouvet 1994),
grasshoppers (Cooper & al. 1995), shrews and bank
voles (Bilton & al. 1998).
Veronica subg. Pseudolysimachium –
speciation by hybridization
Veronica subg. Pseudolysimachium is one of the most
beautiful but also systematically complex groups of
Veronica. It is well differentiated from other species
of Veronica by its dense inflorescence, different flower
morphology, pollen, and embryology, as well as a dif-
ferent chromosome base number. The group contains
about forty species but is notorious for its hybridiza-
tion, intraspecific ploidy level changes, and phenotyp-
ic variability leading to vast amounts of synonyms and
subspecies or varieties. Crossing experiments by Härle
(1932) have shown the wide cross-compatibility with-
in a ploidy-level in the group. In recent times, studies
by Fischer (1974), Fischer & Bedalov (1988), Fischer
& Peev (1995), Albach & Fischer (2003) and Trávnícek
(1997, 1998, 2000) have helped greatly delimit mor-
phologically distinct units in Europe. However, AFLP
fingerprints failed to give support to those taxa stud-
Within the European members of V. subg.
Pseudolysimachium, we find about 11 diploid and
6–7 tetraploid taxa. Only two of the tetraploid taxa
are exclusively tetraploid, V. crassifolia and V. spicata
subsp. fischeri. Among the other tetraploids we find
the most common European species, V. spicata and
V. longifolia ( = V. maritima sensu Trávnícek 2000).
On the base of crossing experiments Graze (1933)
assumed widespread hybridization in nature, which
Fischer (1974) doubted on the basis of extensive her-
barium revision. DNA sequences from the ITS- and
trnLF-regions were unable to distinguish species be-
cause of the low variability (Albach & al. 2005), so we
depend on a faster evolving molecular marker. AFLP
were chosen here, because they have been success-
ful in detecting variation among closely related spe-
cies and within species in other groups of Veronica,
such as V. tenuifolia (Martínez-Ortega & al. 2004),
V. alpina (Albach & al. in press) and V. cymbalaria
(Albach submitted). AFLP have been useful in de-
tecting patterns of hybridization in an earlier study
of Veronica (V. hederifolia and V. cymbalaria; Albach
submitted) and other plant taxa (e.g., Hedrén & al.
2001; Gobert & al. 2002) and thus were thought to
be a reliable and cost-effective way to give some ini-
tial support for hypotheses of hybridization and its
importance on the Balkan Peninsula. Hybrids have
commonly been analyzed together with their par-
ents (e.g., Hedrén & al. 2001; Hodkinson & al. 2002;
Guo & al. 2005) but their effect on the analysis has
never been investigated in great detail. Results from
the analysis here show large-scale incongruence with
currently accepted species delimitation (Fig. 4) and
no fixed species-specific fragment was found. Several
reasons that do not mutually exclude each other
may explain this incongruence. Hybridization may
be more prevalent than discernible by morpholog-
ical inspection (Fischer 1974). Polyploidy increases
homoplasy in cases in which either different species
share one parent or one species is of polytopic origin
with genetically different parentage (Albach submit-
ted). Furthermore, V. subg. Pseudolysimachium is al-
ready of polyploid origin itself, which may increase
the chances for homoplasy.
Phytol. Balcan. 12(2) • Sofia • 2006
Hybridization and polyploidy may be very com-
mon in V. subg. Pseudolysimachium for historical rea-
sons. Pleistocene glaciations had considerable im-
pact on the vegetation of Europe. During glaciations,
the climate in Europe was more continental, favoring
steppe vegetation in larger parts of Europe (Mai 1995).
This may have led to invasions of steppe species, such
as V. subg. Pseudolysimachium, from West Siberia in-
to Eastern Europe. Trávnícek (1998) hypothesized
such an invasion and a subsequent retreat in more hu-
mid Holocene times for V. incana. Unfortunately, such
steppe species have seldom been investigated phyloge-
ographically in much detail (see Franzke & al. 2004).
Climatological oscillations during the Ice Ages may
have led to repeated advances of steppe species and re-
peated chances for hybridization. Another remarkable
fact that may indicate a high chance for hybridization
in this group is the prevalence of diploid cytotypes in
marginal areas and relictual areas (Trávnícek 1998,
2000, pers. comm.), whereas tetraploids predominate
in the central range of the species. Similar situations in
other plant groups led to the hypothesis that tetraploid
forms replaced diploid forms gradually during expan-
sion after the Pleistocene (Ehrendorfer 1962, 1965).
This situation seems on first sight in contrast to the
general notion that polyploids evolve at the margins of
distribution but supports Stebbins’ (1984) secondary-
contact hypothesis. Following this hypothesis, diploid
taxa survived in their Pleistocene refugia and after the
Ice Age spread and encountered other diploid survi-
vors, with which they formed polyploid hybrids. Such
kinds of neopolyploids are known to surpass their an-
cestors in invasion potential, e.g. in the habitats that
have become available after the retreat of Pleistocene
ice sheets (Ehrendorfer 1965, 1980). Consequently, a
phylogeographic study of V. subg. Pseudolysimachium
should retrieve a patchy distribution pattern of the
Pleistocene refugia in Europe for the diploids and a
pattern mirroring the post-Pleistocene migration for
the tetraploid cytotypes, which blurs the genealogical
history of the group. Unfortunately, it was not possi-
ble in this study to ascertain the ploidy level of the an-
alyzed individuals. Subsequent studies will be need to
confirm this factor.
Despite these fallacies, some results raise doubts
about the earlier taxonomic conclusions. Veronica bar-
relieri is known to have diploid and tetraploid popula-
tions and is found here in two widely divergent groups
(Fig. 4). It is also morphologically difficult to differen-
tiate and sometimes considered part of a broad V. spi-
cata (Elenevsky 1971; Walters & Webb 1972) but dif-
ferentiated by us in five subspecies (Albach & Fischer
2003). Therefore the species needs closer inspection.
Veronica porphyriana is also occasionally included in
a broader V. spicata (Elenevsky 1971) but is clearly dif-
ferentiated here from the European samples (Fig. 4).
Its status as a distinct species therefore seems to be val-
id. Veronica longifolia is also found in two widely sepa-
rated groups: the European samples in one group and
the Asian sample from the Altai in another (Fig. 4).
This would support Trávnícek´s (2000) hypothe-
sis of two species being hidden in V. longifolia, which
he termed V. maritima for the European populations
and V. longifolia for the Asian samples. It should be
borne in mind that the situation may be more com-
plex, because only one population from Asia was ana-
lyzed and no geographically intermediate sample was
included. Furthermore, it is not clear whether V. longi-
folia in Asia, where diploids are frequent (Albach &
al. submitted), represent diploid progenitors of their
mostly tetraploid European relatives, or just relative-
ly unrelated diploid relatives that were not involved in
the origin of the tetraploids.
Future investigations in V. subg. Pseudolysimachium
will be need to take the lessons from this study into
consideration. 1.) DNA sequence analyses are supe-
rior to anonymous markers, because hybridization
renders homology assessment of markers more dif-
ficult. However, even DNA sequence studies will be
problematic in the presence of hybridization and mul-
tiple markers will be necessary to infer a species tree.
2.) Ploidy level should be controlled in all samples.
3.) Morphological characters may be homoplastic due
to character shuffling by hybridization. 4.) Ancestral
genotypes are likely to be found in Asian members of
the group, which is in line with the inferred Asian or-
igin of the subgenus (Albach & al. 2005). The analysis
of additional diploid taxa from East Europe and West
Siberia will be crucial to understand the origin of the
Veronica subsect. Alpina –
relict populations on the Balkan Peninsula
Veronica alpina and V. bellidioides are closely related
species of V. subg. Veronica (Albach & al. in press).
They are both species of subalpine meadows but dif-
fer ecologically. This ecological difference is responsi-
Albach, D. & al. • Veronica on the Balkan Peninsula
ble for differences related to climate change as shown
by simulation studies (Guisan & Theurillat 2000)
and different phylogeographic histories (Albach & al.
in press). Phylogeographic analyses of AFLP finger-
print data revealed that V. alpina shows much high-
er intraspecific genetic diversity than V. bellidioides
(Albach & al. in press). Populations of V. bellidioides
from the Balkan Peninsula (Bulgaria) form a mono-
phyletic sister group to the populations from the Alps
and the Pyrenees (Albach & al. in press) representing
a long historic divergence. In the absence of molecu-
lar dating, we are unable to estimate the divergence
time, but given the absence of plastid DNA variation
(Albach & al. in press), a Pleistocene origin is likely.
In contrast, populations of V. alpina from the Balkan
Peninsula (Bulgaria) are not well separated from oth-
er populations of this species in Europe. Interestingly,
however, there does not seem to be a close connection
with populations from the Carpathians (Albach & al.
in press). Unfortunately, populations from other lo-
calities of both species on the Balkan Peninsula could
not be sampled. Populations of both V. alpina and V.
bellidioides may nevertheless be Pleistocene relicts of
similar age that were unable to migrate north after the
Ice Ages, similar to the case in V. subg. Chamaedrys
(see above). However, the cases of V. alpina and V. bel-
lidioides differ in that extinctions and genetic bottle-
necks in surviving populations of V. bellidioides led to
a clear bifurcation in the phylogenetic tree.
The chosen examples highlight the different respons-
es of plants to the Pleistocene climate changes. Plant
species probably had very different histories on the
Balkan Peninsula. They may have become relict taxa of
probable Tertiary origin such as species from V. subg.
Stenocarpon that differentiated into morphologically
clearly distinguishable species. Other groups may have
become separate taxa in the Pleistocene on the verge
of speciation by genetic drift, such as taxa in V. subg.
Chamaedrys. Or they may actively speciate by hybrid-
ization, such as taxa in V. subg. Pseudolysimachium.
Despite these differences, the Balkan Peninsula rep-
resents an important center of genetic diversity for
all investigated groups of Veronica. Other species
and species complexes of Veronica on the Balkan
Peninsula, such as V. sect. Pentasepalae, are currently
under investigation (Martínez Ortega, pers. comm.).
Undoubtedly, they will further highlight the impor-
tance of the Balkan Peninsula for the survival and di-
versification of plant taxa in the Pleistocene.
Acknowledgements. The author wishes to thank the Austrian
Science Fund (FWF, project P-15336-Bio) for the financial support
of this research. Antonio Abad (AFLP) and Marion Kever (DNA se-
quencing) provided invaluable help in the lab. Finally, I am grateful
to Montserrat Martínez Ortega (Univ. Salamanca), M. Fay (RBG
Kew, UK), Gerald Schneeweiss (Univ. Wien), and Andreas Tribsch
(Univ. Zalzburg) for providing plant material for this publication.
Albach, D.C. & Chase, M.W. 2001. Paraphyly of Veronica
(Veroniceae; Scrophularaceae): Evidence from the internal tran-
scribed spacer (ITS) sequences of nuclear ribosomal DNA. – J.
Pl. Res., 114: 9-18.
Albach, D.C. & Fischer, M.A. 2003. AFLP- and genome size
analyses: contribution to the taxonomy of Veronica subg.
Pseudolysimachium sect. Pseudolysimachion (Plantaginaceae),
with a key to the European taxa. – Phytol. Balcan., 9: 401-424.
Albach, D.C., Jensen, S.R., Özgökce, F. & Grayer, R.J. 2005.
Chemical characters for the support of phylogenetic relationships
based on nuclear ribosomal and plastid DNA sequence data in
Veronica. – Biochem. Syst. Ecol., 33: 1087-1106.
Albach, D.C., Martínez-Ortega, M.M., Fischer, M.A. & Chase,
M.W. 2004a. A new classification of the tribe Veroniceae − prob-
lems and a possible solution. – Taxon, 53: 429-452.
Albach, D.C., Martínez-Ortega, M.M. & Chase, M.W. 2004b.
Veronica: Parallel morphological evolution and phylogeography
in the Mediterranean. – Pl. Syst. Evol., 246: 177-194.
Albach, D.C., Martínez-Ortega, M.M., Fischer, M.A. & Chase,
M.W. 2004c. Evolution of Veroniceae: A phylogenetic perspec-
tive. – Ann. Missouri Bot. Gard., 91: 275-302.
Albach, D.C., Meudt, H.M. & Oxelman, B. 2005. Piecing together
the “new” Plantaginaceae. – Amer. J. Bot., 92: 297-315.
Albach, D.C., Schönswetter, P. & Tribsch, A. In press. Comparative
phylogeography of the Veronica alpina-complex in Europe and
North America. – Mol. Ecol.
Angiosperm Phylogeny Group. 1998. An ordinal classification
for the families of flowering plants. – Ann. Missouri Bot. Gard.,
Benham, J.J. 1998. Genographer, 1.1. Montana State Univ.,
Bennett, K.D., Tzedakis, P.C. & Willis, K.J. 1991. Quaternary
refugia of north European trees. – J. Biogeogr., 18: 103-115.
Bilton, D.T., Mirol, P.M., Mascheretti, S., Fredga, K., Zima, J.
& Searle, J.B. 1998. Mediterranean Europe as an area of ende-
mism for small mammals rather than a source for northwards
Phytol. Balcan. 12(2) • Sofia • 2006
postglacial colonization. – Proc. Roy. Soc. London, Ser. B, Bio.
Sci., 265: 1219-1226.
Casgrain, P. & Legendre, P. 2001 The R Package for multivariate
and spatial analysis. Version 4.0 d5. User’s manual. Dept. Sci.
Biol., Univ. Montreal, Montreal.
Cooper, S.J.B., Ibrahim, K.M. & Hewitt, G.M. 1995. Postglacial
expansion and genome subdivision in the European grasshopper
Chorthippus parallelus. – Mol. Ecol., 4: 49-60.
Doyle, J.J. & Doyle, J.L. 1987. A rapid DNA isolation procedure
for small quantities of fresh leaf tissue. – Phytochem. Bull. Bot.
Soc. Amer., 19: 11-15.
Ehrendorfer, F. 1962. Cytotaxonomische Beiträge zur Genese der
mitteleuropäischen Flora und Vegetation. – Ber. Deutsch. Bot.
Ges., 75: 137-152.
Ehrendorfer, F. 1965. Über stammesgeschichtliche Differen-
zierung smuster bei den Dipsacaceae. – Ber. Deutsch. Bot. Ges.,
Ehrendorfer, F. 1980. Polyploidy and distribution. – In: Lewis,
W.H. (ed.), Polyploidy: Biological relevance. Pp. 45-60. Plenum
Publishing Corp., New York.
Elenevsky, A. 1971. De Veronica spicata L. s. l. notulae systemati-
cae. – Novosti Sist. Vyssh. Rast., 8: 215-227.
Fischer, M.A. 1969. Veronica caespitosa Boiss., V. bombycina Boiss.
& Kotschy und V. thessalica Benth. gehören nicht in die Sektion
Veronicastrum Benth. – Österr. Bot. Z., 117: 54−63.
Fischer, M.A. 1972. Neue Taxa, Chromosomenzahlen und
Systematik von Veronica subsect. Acinifolia (Römpp) Stroh. –
Österr. Bot. Z., 121: 413-437.
Fischer, M.A. 1973a. Notizen zur Systematik, Chromosomenzahl
und Verbreitung einiger Veronica-Sippen in Kärnten. – Carinthia,
Fischer, M.A.1973b. Zur Cytotaxonomie von Veronica chamaedrys
L. agg., II.: subsp. micans M. Fischer, subsp. nova, eine weitere
diploide Sippe. – Österr. Bot. Z., 121: 73-79.
Fischer, M.A. 1974. Beitrag zu einer systematischen Neubearbeitung
der Gruppe um Pseudolysimachion spicatum (L.) Opiz ( = Veronica
spicata L.). – Phyton, 16: 29-47.
Fischer, M.A. 1975a. Untersuchungen über den Polyploidkomplex
Veronica cymbalaria agg. (Scrophulariaceae). – Pl. Syst. Evol.,
Fischer, M.A. 1975b. The Veronica hederifolia group: Taxonomy,
ecology, and phylogeny. – In: Walter, S.M. (ed.), European floris-
tic and taxonomic studies. Conf. Rep. Pp. 48-60. Bot. Soc. British
Isles, E.W. Classey Ltd, Farington.
Fischer, M.A. 1978. Veronica L. – In: Davis, P.H. (ed.), Flora
of Turkey and the East Aegean Islands. Vol. 6, pp. 689-753.
Edinburgh Univ. Press, Edinburgh.
Fischer, M.A. 1991. Veronica. – In: Strid, A. & Tan, K. (eds),
Mountain Flora of Greece. Vol. 2, pp. 209-234. Edinburgh Univ.
Fischer, M.A. & Bedalov, M. 1988. The genera Paederota and
Pseudolysimachion (Scrophulariaceae) in Croatia. – Acta Bot.
Croat., 47: 149-156.
Fischer, M.A. & Fischer, G. 1981. Distribution patterns of Veronica
species in the Balkan Peninsula. – In: Velčev, V.I. & Kožuharov,
S.I. (eds), Mapping the flora of the Balkan Peninsula. Pp. 175-232.
Publishing House Bulg. Acad. Sci., Sofia.
Fischer, M.A. & Peev, D. 1995. Pseudolysimachion Opiz. – In:
Kozhuharov, S. (ed.), Fl. Reipublicae Bulgaricae. Vol. 10,
pp. 190-202. Editio Acad. “Prof. Marin Drinov”, Serdicae (in
Franzke, A., Hurka, H., Janssen, D., Neuffer, B., Friesen, N.,
Markov, M. & Mummenhoff, K. 2004. Molecular signals for Late
Tertiary/Early Quaternary range splits of an Eurasian steppe plant:
Clausia aprica (Brassicaceae). – Mol. Ecol., 13: 2789-2795.
Frenzel, B. 1968. Grundzüge der pleistozänen Vegetationsgeschichte
Nord-Eurasiens. Franz Steiner Verlag, Wiesbaden.
Gobert, V., Moja, S., Colson, M. & Taberlet, P. 2002. Hybridization
in the section Mentha (Lamiaceae) inferred from AFLP mark-
ers. – Amer. J. Bot., 89: 2017-2023.
Graze, H. 1933. Die chromosomalen Verhältnisse in der Sektion
Pseudolysimachia Koch der Gattung Veronica. – Jahrb. Wiss.
Bot., 77: 507-559.
Guisan, A., Theurillat, J.-P. 2000. Assessing alpine plant vulner-
ability to climate change: a modelling perspective. – Integr.
Assessm., 1: 307-320.
Guo, Y.-P., Saukel, J., Mittermayr, R. & Ehrendorfer, F. 2005. AFLP
analyses demonstrate genetic divergence, hybridization, and
multiple polyploidization in the evolution of Achillea (Asteraceae-
Anthemideae). – New Phytol., 166: 273-290.
Härle, A. 1932. Die Arten und Formen der Veronica-Sektion
Pseudolysimachia Koch auf Grund systematischer und experi-
menteller Untersuchungen. – Bibl. Bot., 26: 1-86.
Hedrén, M., Fay, M.F. & Chase, M.W. 2001. Amplified frag-
ment length polymorphisms (AFLP) reveal details of polyploid
evolution in Dactylorhiza (Orchidaceae). – Amer. J. Bot., 88:
Hewitt, G. 2000. The genetic legacy of the quaternary ice ages. –
Nature, 405: 907-913.
Hodkinson, T.R., Chase, M.W., Takahashi, C., Leitch, I.J.,
Bennett, M.D. & Renvoize, S.A. 2002. The use of DNA sequenc-
ing (ITS and trnL-F), AFLP, and fluorescent in situ hybridization
to study allopolyploid Miscanthus (Poaceae). – Amer. J. Bot., 89:
Horvat, I., Glavac, V. & Ellenberg, H. 1974. Vegetation
Südosteuropas. G. Fischer, Stuttgart.
Huntley, B. & Birks, H.J.B. 1983. An atlas of past and present pollen
maps for Europe. Cambridge Univ. Press, Cambridge.
Mai, D. H. 1995. Tertiäre Vegetationsgeschichte Europas. Gustav
Fischer Verlag, Jena, Stuttgart, New York.
Martínez-Ortega, M.M. & Rico, E. 2001. Taxonomy of Veronica
subsect. Serpyllifolia (Scrophulariaceae). – Bot. J. Linn. Soc.,
Martínez-Ortega, M.M., Delgado, L., Albach, D.C., Elena-
Rosselló, J.A. & Rico, E. 2004. Species boundaries and phyloge-
ography of Veronica subsect. Pentasepalae (Scrophularaceae) in
the Western Mediterranean inferred from AFLP markers. – Syst.
Bot., 29: 965-986.
Meyer, F.K. 1970. Gesneriaceae als Glieder der Flora des Tertiär
in Europa. – Wiss. Z. Friedrich-Schiller-Univ. Jena, Math.-
Naturwiss. Reihe, 19: 401-411.
Albach, D. & al. • Veronica on the Balkan Peninsula
Mirek, Z., & Fischer, M.A. 1986. Additions to the ecogeography of
Veronica vindobonensis with special reference to Poland. – Phyton,
Ozenda, P. 1988. Die Vegetation der Alpen im europäischen
Gebirgsraum. Gustav Fischer Verlag, Stuttgart, New York.
PE Applied Biosystems. 1996. AFLP™ Plant Mapping Protocol.
Foster City, California.
Polunin, O. 1980. Flowers of Greece and the Balkans. Oxford Univ.
Posada, D. & Crandall, K. A. 1998. MODELTEST: testing the model
of DNA substitution. – Bio-informmation, 14: 817-818.
Saitou, N., & Nei, M. 1987. The neighbor-joining method: a new
method for reconstructing phylogenetic trees. – Mol. Biol. Evol.,
Schmitt, T., & Seitz, A. 2002. Postglacial distribution area expan-
sion of Polyommatus coridon (Lepidoptera: Lycaenidae) from its
Ponto-Mediterranean glacial refugium. – Heredity, 89: 20-26.
Stebbins, G.L. 1984. Polyploidy and the distribution of the arctic-
alpine flora: new evidence and a new approach. – Bot. Helv.,
Stroh, G. 1942. Die Gattung Veronica L. Versuch einer systema-
tischen Kodifizierung der Arten (mit Ausnahme der endemischen
Arten von Neuseeland). – Beih. Bot. Centralbl., 61: 384-451.
Sun, Y., Skinner, D.Z., Liang, G.H. & Hulbert, S.H. 1994.
Phylogenetic analysis of Sorghum and related taxa using internal
transcribed spacers of nuclear ribosomal DNA. – Theor. Appl.
Genet., 89: 26-32.
Swofford, D.L. 2002. PAUP* Phylogenetic Analysis Using
Parsimony (*and Other Methods), Version 4.10. Sinauer
Associates, Sunderland, Massachusetts.
Taberlet, P. & Bouvet, J. 1994. Mitochondrial DNA polymorphism,
phylogeography and conservation genetics of the brown bear
Ursus arctos in Europe. – Proc. Roy. Soc. London, Ser. B, Biol.
Sci., 255: 195-200.
Taberlet, P., Gielly, L., Pautou, G. & Bouvet, J. 1991. Universal
primers for amplification of three non-coding regions of chlo-
roplast DNA. – Pl. Mol. Biol., 17: 1105-1109.
Taberlet, P., Fumagalli, L., Wust-Saucy, A.-G. & Cosson, J.-F.
1998. Comparative phylogeography and postglacial colonization
routes in Europe. – Mol. Ecol., 7: 453-464.
Trávnícek, B. 1997. Pseudolysimachion (W.D.J. Koch) Opiz. – In:
Goliašová, K. (ed.), Flora of Slovakia. Vol. 5(2), pp. 264-298.
Veda Press, Bratislava (in Slovak).
Trávnícek, B. 1998. Notes on the taxonomy of Pseudolysimachion
sect. Pseudolysimachion (Scrophulariaceae) in Europe. I. P. inca-
num and P. spicatum. – Preslia, 70: 193-223.
Trávnícek, B. 2000. Taxa of the genus Pseudolysimachion
(Scrophulariaceae) in the Czech Republic. I. Identification key
to the species. Geographical distribution of species of sect.
Longifolia. – Preslia, 72: 411-439.
Tremetsberger, K., Stuessy, T.F., Samuel, R.M., Baeza, C.M. &
Fay, M.F. 2003. Genetics of colonization in Hypochaeris tenuifolia
(Asteraceae, Lactuceae) on Volcán Longiquimay, Chile. – Mol.
Ecol., 12: 2649-2659.
Tribsch, A. & Schönswetter, P. 2003. Patterns of endemism and
comparative phylogeography confirm palaeo-environmental
evidence for Pleistocene refugia in the Eastern Alps. – Taxon,
Turrill, W. 1929. The plant life of the Balkan Peninsula. Oxford
Univ. Press, Oxford.
Tzedakis, P.C., Lawson, I.T., Frogley, M.R., Hewitt, G.M. &
Preece, R.C. 2002. Buffered tree population changes in a qua-
ternary refugium: evolutionary implications. – Science, 297:
Vos, P., Hogers, R., Bleeker, M., Reijans, M., van de Lee, T.,
Hornes, M., Frijters, A., Pot, J., Peleman, J., Kuiper, M. &
Zabeau, M. 1995. AFLP: a new technique for DNA fingerprint-
ing. – Nucl. Acids Res., 23:4407-4414.
Wagstaff, S.J. & Garnock-Jones, P.J. 1998. Evolution and biogeog-
raphy of the Hebe complex (Scrophulariaceae) inferred from ITS
sequences. – New Zealand J. Bot., 36: 425−437.
Walters, S.M. & Webb, D.A. 1972. Veronica L. – In: Tutin, T.G. &
al. (eds), Flora Europaea. Vol. 3, pp. 242−251. Cambridge Univ.
White, T. J., Bruns, T., Lee, S. & Taylor, J. 1991. Amplification
and direct sequencing of fungal ribosomal RNA genes for phy-
logenetics. – In: Gelfand, D., Innis, M., Sninsky, J. & White, T.
(eds), PCR Protocols: A guide to Methods and Applications. Pp.
315-322. Acad. Press, San Diego.
Willis, K.J. 1996. Where did all the flowers go? The fate of tem-
perate European flora during glacial periods. – Endeavour, 20:
Willis, K.J., Rudner, E. & Sümegi, P. 2000. The full-glacial forests of
central and southeastern Europe. – Quatern. Res., 53: 203-213.
Zhang, L., Comes, H.P., & Kadereit, J.W. 2001. Phylogeny and
quaternary history of the European montane/alpin endemic
Soldanella (Primulaceae) based on ITS and AFLP variation. –
Amer. J. Bot., 88: 2331-2345.