ArticlePDF Available

Phylogeny, karyotype evolution and taxonomy of Cerinthe L. (Boraginaceae)

Authors:

Abstract and Figures

A phylogenetic and karyological analysis of the small and poorly known genus Cerinthe L. (Boraginaceae-Lithospermeae) was performed using ITS sequences and standard chromosomal techniques. All taxa are diploid with 2 n = 16 or 2 n = 18 and show a variable degree of infraspecific variation, in particular in the polymorphic C. major and C. minor. Change in base number is associated with an early split between the two well-supported clades of C. major, corresponding to Cerinthe sect. Cerinthe, and that of all other taxa belonging to C. sect. Ceranthe, with the base x = 8 found only in the strictly annual C. major group, and x = 9 in the other five species of the genus: C. minor, C. glabra, C. tenuiflora, C. retorta and C. palaestina . The latter section is subdivided into the E Mediterranean, annual lineage of C. palaestina - C. retorta and the mainly continental, perennial group of C. minor-C. glabra, the sister of which is the Corsican endemic C. tenuiflora. The hypothesis that x = 9 represents the primary haploid number and x = 8 is derived through descending aneuploidy, is discussed. A taxonomic revision of the genus is provided and the following formal taxonomic changes are proposed: C. major L. subsp. oranensis (Batt.) Selvi & Cecchi, stat. nov.; C. major L. subsp. purpurascens (Boiss.) Selvi & Cecchi, stat. nov.; C. minor L. subsp. cleiostoma (Boiss. & Sprun.) Selvi & Cecchi, stat. nov. Cerinthe tenuiflora, C. retorta and the poorly known C. palaestina are well-defined species with little internal variation.
Content may be subject to copyright.
1
Selvi & al. • Phylogeny and taxonomy of Cerinthe
TAXO N • 07 October 2009: 19 pp.
INTRODUCTION
Cerinthe L. is a small genus of Boraginaceae tribe
Lithospermeae DC. distributed in the circum-Mediterra-
nean area from the Atlantic region of Morocco to the west-
ern part of the Irano-Turanian region. Morphologically it
is strongly characterised by the glabrous to subglabrous
vegetative parts, the nearly dialisepalous calyx with un-
equal sepals inserted at different heights, the tubulose-
cylindrical corolla without faucal scales and the sagittate
anthers, appendaged and connate at the base (Fig. 1A,
B). The most relevant autapomorphic trait, however, is
represented by the fruit, a schizocarp splitting at maturity
in only two, rather than four as usual in Boraginaceae,
mericarpids formed by two connate, monospermic nut-
lets. Each of the two carpels of the ovary consists of two
attached locules with a single ovule, separated by a par-
enchymatous tissue becoming thick and hard at maturity
(Fig. 1B). This feature was interpreted by Johnston (1924)
as a derived condition resulting from the conf luence of
two distinct mericarpids, while Popov (1953) viewed it as
an “atavism repressing to the ancestral structure typical
of the ancient arboreal subfamily Ehretieae, by reduction
through which Cerinthe originated”. Mainly based on this
feature, Popov (1953) kept this genus in the “remarkable
anomalous” tribe Cerintheae DC. instituted by Candolle
(1846).
The affinities of Cerinthe are still unclear. Pre-
liminary data from atpB DNA sequences (Långström
& Chase, 2002) showed it to represent a clade of Old-
World Lithospermeae with uncertain relationships. More
Phylogeny, karyotype evolution and taxonomy of Cerinthe L.
(Boraginaceae)
Federico Selvi, Lorenzo Cecchi & Andrea Coppi
Dipartimento di Biologia Vegetale dell’Università, sezione di Botanica Sistematica, Via G. La Pira 4, 50121
Firenze, Italy. selvi@unifi.it (author for correspondence)
A phylogenetic and karyological analysis of the small and poorly known genus Cerinthe L. (Boraginaceae-
Lithospermeae) was performed using ITS sequences and standard chromosomal techniques. All taxa are diploid
with 2n = 16 or 2n = 18 and show a variable degree of infraspecific variation, in particular in the polymorphic
C. major and C. minor. Change in base number is associated with an early split between the two well-supported
clades of C. major, corresponding to Cerinthe sect. Cerinthe, and that of all other taxa belonging to C. sect.
Ceranthe, with the base x = 8 found only in the strictly annual C. major group, and x = 9 in the other five species
of the genus: C. minor, C. glabra, C. tenuiflora, C. retorta and C. palaestina. The latter section is subdivided
into the SE Mediterranean, annual lineage of C. palaestinaC. retorta and the mainly continental, perennial
group of C. minor–C. glabra, the sister of which is the Corsican endemic C. tenuiflora. The hypothesis that x =
9 represents the primary haploid number and x = 8 is derived through descending aneuploidy, is discussed. A
taxonomic revision of the genus is provided and the following formal taxonomic changes are proposed: C. major
L. subsp. oranensis (Batt.) Selvi & Cecchi, stat. nov.; C. major L. subsp. purpurascens (Boiss.) Selvi & Cecchi,
stat. nov.; C. minor L. subsp. cleiostoma (Boiss. & Sprun.) Selvi & Cecchi, stat. nov. Cerinthe tenuiflora, C. re-
torta and the poorly known C. palaestina are well-defined species with little internal variation.
KEYWORDS: Boraginaceae, Cerinthe, chromosome evolution, molecular phylogeny, systematics, taxonomy
Fig. 1. Calyx (A) and floral diagram (B) of Cerinthe. 1, carpel
with two connate locules; 2, stamens; 3, sympetalous co-
roll a; 4, dialisepalous calyx; 5, b rac t; 6, pedunc le. Drawing
by L. Cecchi.
2
Selvi & al. • Phylogeny and taxonomy of Cerinthe TA XON • 07 October 2009: 19 pp.
recently, chloroplast matK sequences (Cecchi & Selvi,
2009) suggested an affinity with Onosma L. as inferred
by past authors from f lower morphology (Gürke, 1894;
Popov, 1953), while nuclear ITS DNA sequences sup-
ported a relationship with a clade comprising the mono-
typic genera Neatostema I.M. Johnst., Moltkiopsis I.M.
Johnst., Mairetis I.M. Johnst., Paramoltkia Greuter and
Halacsya Boiss. However, the limited taxon sampling of
the aforementioned studies (one or two species only) did
not allow to assess either the affinities of this genus or
relationships at a lower taxonomic level.
From the karyological point of view, available litera-
ture data suggest that species of Cerinthe are diploid with
2n = 16 or 18, although chromosome reports are still scanty.
If confirmed, the existence of two different base numbers
poses the question of whether such variation is correlated
with evolutionary events of divergence between lineages,
and which one of the two may represent the plesiomorphic
condition. Although increase in chromosome number due
to processes such as fission, fragmentation, polyploidy or
amphidiploidy are driving evolutionary forces in many
angiosperm groups, there is evidence that descending an-
euploidy or disploidy has played a key role in the evolution
of some Borag inaceae groups like Nonea Medik. (Selvi &
Bigazzi, 2002; Bigazzi & Selvi, 2003), Cynoglottis (Guşul.)
Vural & Kit Tan (Bigazzi & Selvi, 2001) and Borago L.
(Selvi & al., 2006). Therefore, Cerinthe offers a good op-
port unit y to test whethe r decrease i n chromosome number
and associated rearrangements of the karyotype can be
interpreted as an evolutionary trend associated with phy-
logenetic divergence between lineages.
In this paper we first analyse the karyotype morphol-
ogy of most described taxa and attempt to bring light
on the possible processes behind the observed variation.
Then we correlate chromosome data with the results of a
phylogenetic analysis based on DNA sequences from the
internal transcribed spacer region (ITS) of the nuclear
genome. Phylogenetic and karyological data are finally
integrated with morphological evidence to propose an
updated taxonomic treatment aiming at resolving old
problems in especially the polymorphic groups of C. ma-
jor L. and C. minor L. Here, “doubtful” and little-known
taxa such as C. palaestina Eig & Sam. or C. tristis Tey be r
still cause contrasting treatments in standard floras. To
this purpose, nomenclatural types, synonymies, distribu-
tion maps, original iconographies and analytical keys for
identification are given.
MATERIALS AND METHODS
Plant material.
Populations of Cerinthe were
sampled in the field over most of the Mediterranean re-
gion from Morocco to Turkey. Herbarium vouchers, seeds,
fixed samples of reproductive structures and fragments of
leaf tissue in silica gel were collected for each accession.
The complete list of accessions used for the karyological
study, the phylogenetic analysis and the morphological
observations is reported in the Appendix.
Karyology.
At least ten root tips were collected in
early spring and/or autumn from germinating seeds (an-
nual taxa) or cultivated plants (peren nial taxa). They were
pretreated with 0.002 M 8-hydroxyquinoline or 0.05%
colchicine, 2.5 hrs at room temperature and then fixed
overnight in ethanol : glacial acetic acid (3 : 1). When neces-
sary, they were preserved in 70% ethanol at 3°C–4°C until
preparation. The meristematic tissue was then thoroughly
rinsed in disti lled water, hyd roly zed in 1N HCl at 60°C for
6–7 minutes, and stained in lacto-propionic orcein over-
night (Dyer, 1979). The meristems were finally dissected
and squashed on glass slides in a drop of 45% acetic acid.
Metaphase plates were examined with a Zeiss Axioscop
light microscope under oil immersion (×100), and photo-
graphed with a Canon digital system. Karyotype formulas
were determined on selected enlarged prints by carefully
measuring the length of chromosome arms and satellites.
The centromeric index was calculated as the long : short
arm ratio to classify chromosomes following Levan & al.
(1964): m = metacentric (r = 1.00–1.69), sm = submetacen-
tric (r = 1.70–2.99), st = subtelocentric (r = 3.00–6.99). SAT
chromosomes were recognised based on the occurrence
and position of secondary constrictions corresponding to
the nucleolar organising region(s) (NORs). The intrachro-
mosomal asymmetry index (A1) was calculated according
to Romero Zarco (1986), while the interchromosomal index
(A
2
) was measured as the ratio between standard deviation
of chromosome length and mean chromosome length. Fi-
nal karyotype formulas, chromosome size and asymmetry
indexes for each accession resulted from the mean values
of at least five individual karyotypes.
DNA
extraction and amplification.
Genomic
DNA was extracted from silica-gel dried samples of leaf
tissue following a modified 2×CTAB protocol (Doyle &
Doyle, 1990). The extracted DNA was quantified after aga-
rose gel electrophoresis (0.6% w/v) in TAE buffer (1 mM
EDTA, 40 mM Tris-acetate) containing 1 μg/ml of ethidium
bromide by comparison with a known mass standard.
Amplification of the ITS region of nuclear DNA, in-
cluding ITS1, 5.8S and ITS2, was done using the primers
ITS4 and ITS5 of White & al. (1990). Polymerase chain
reactions were performed in a total volume of 25 μl con-
taining 2.5 μl of reaction buffer (Dynazyme II; Finnzyme,
Espoo, Finland), 1.5 mM MgCl2, 10 pmol of each primer,
200 μM of each dNTP, 1 U of Taq DNA polymerase (Dy-
nazyme II; Finnzyme) and 10 ng of template DNA. Re-
actions were performed in a MJ PTC-100 thermocycler
(Peltier Thermal Cycler; MJ Research). Fourty amplifica-
tion cycles were run with annealing temperature 50°C,
3
Selvi & al. • Phylogeny and taxonomy of Cerinthe
TAXO N • 07 October 2009: 19 pp.
annealing time 30 s and final extension for 45 s at 72°C.
Subsequently, 5 μl of each amplification mixture were
analysed by agarose gel (1.5% w/v) electrophoresis in TAE
buffer containing 1 μg/ml ethidium bromide. The PCR
reactions were purified from excess salts and primer with
the PCR Purification Kit (Roche, Mannheim, Germany).
Automated DNA sequencing was performed directly
from the purified PCR products using BigDye Termina-
tor v.2 chemistry and an ABI310 sequencer (PE-Applied
Biosystems, Norwalk, Connecticut, U.S.A.).
Sequence alignment and phylogenetic infer-
ence.
Sequences were checked for orthology by
comparison with Onosma echioides (L.) L., Halacsya
sendtneri (Boiss.) Dörfl. and Echium arenarium Guss.,
all of tribe Lithospermeae. These three species were then
used as outgroups based on their position in our recent
ITS-matK phylogenetic study of circum-Mediterranean
Lithospermeae (Cecchi & Selvi, 2009). Multiple align-
ments were performed with Multalin (Corpet, 1988), and
then further examined and slightly modified manually
based on eye examination of the sequencer output chro-
matof iles. All charact ers we re weighted e qually, and char-
acter state transitions were treated as unordered. Gaps
were coded according to the “simple gap” method (Sim-
mons & Ochoterena, 2000) using the program FastGap
(Borchsenius, 2007), and appended at the end of the se-
quences. Maximum parsimony (MP), was used to analyse
the aligned sequences and to search for phylogenetic trees.
MP t rees were calculat ed with PAUP* ver. 4.0b (Swofford,
1998) through Branch-and-Bound search with Tree-Bisec-
tion-Reconnection branch swapping, MulTrees, AccTran
and Collapse options in effect, addition sequence = fur-
thest. Internal support to the branches was estimated by
means of 50% majority-rule bootstrap analysis (Felsen-
stein, 1985), with 1,500 random addition replicates. Sup-
port to the branches (BS) was divided in the following
levels: weak (BS 50%–74%), moderate (BS 75%–84%),
good (BS 85%–94%) and strong (BS 95%–100%).
Taxonomy.
For each taxon we adopt the follow-
ing format: name, main homotypic synonyms and type,
heterotypic synonyms with types (when traced), diagnos-
tic morphological characters based on field observations
and herbarium specimens in B, B-W, BM, BOLO, BP, FI,
FI-W, G, G-BOIS, K, LD, MPU, NAP-GUSS, NAP-TEN,
OXF, PAV, W and ZA. In the descriptions, measurements
of calyx sepals refer to the external ones, which are wider
than the internal ones. Distribution and habitat are based
on reliable literature data, herbarium material and per-
sonal field experience. Finally, explanatory comments are
added wherever appropriate.
RESULTS
Karyotype variation.
Chromosome numbers,
karyotype formulas, mean chromosome size and asym-
metry indices of the twelve examined accessions are re-
ported in Table 1.
Cerinthe glabra. Plants from the Italian Maritime
Alps showed 2n = 18, in line with previous reports from
Slovakia (Murin & Majovsky, 1979) and Bulgaria (Mar-
kova & Goranova, 1995). With seven pairs of metacentrics
Table 1. Chromosome number, ploidy level, karyotype formula, mean chromosome length, intra- (A1) and inter- (A2)
chromosomal asymmetr y for each investigated accession of Cerinthe.
T a xon 2nPloidy
level Formula
Mean
length [μm] A1A2
C. glabra 18 2x14m + 4sm 2.6 0.29 0.23
C. major subsp. major 16 2x6m + 8sm + 2st 1.6 0.38 0.13
C. major subsp. oranensis 16 2x12m + 4smSAT 1.9 0.24 0.17
C. major subsp. purpurascens 16 2x10m + 2sm + 2smSAT + 2st 1.5 0.34 0.10
C. minor subsp. minor 18 2x10m + 8sm 2.5 0.31 0.24
C. minor subsp. cleiostoma (Croatia) 18 2x8m + 10sm 1.7 0.34 0.16
C. minor subsp. cleiostoma (C Greece) 18 2x8m + 10sm 1.8 0.35 0.17
C. minor subsp. auriculata (S Italy) 18 2x14m + 2sm+ 2smSAT 1.9 0.23 0.14
C. minor subsp. auriculata (N Albania) 18 2x6m + 10sm + 2smSAT 2.1 0.42 0.27
C. palaestina 18 2x10m + 4sm + 4st 3.4 0.42 0.22
C. retorta 18 2x8m + 8sm + 2smSAT 1.9 0.32 0.26
C. tenuiflora 18 2x14m + 2mSAT + 2sm 2.0 0.20 0.20
4
Selvi & al. • Phylogeny and taxonomy of Cerinthe TA XON • 07 October 2009: 19 pp.
and only two of submetacentrics, the intrachromosomal
asymmetry (A1) of the alpine plants was lower than in the
eastern European accessions, and showed no satellited
pairs (Figs. 2B, 3B, 4).
Cerinthe major. Plants from Sardinia of C. major
subsp. major showed 2n = 16 and a complement of six
metacentrics, eight submetacentrics and a pair of sub-
telocentrics; no satellites or secondary constrictions were
observed (Figs. 2J, 3J). Intrachromosomal asymmetry
was relatively high, while differences in size between
chromosomes were negligible, accounting for a low A
2
asymmetry (Fig. 4). Our finding is in accordance with
previous analyses on plants from the Sevilla region, Spain
(Luque, 1990) and from peninsular Italy (Altamura & al.,
1984; Coppi & al., 2005).
Cerinthe major subsp. purpurascens from S Spain
was characterised by a microsatellited pair of submeta-
centrics and by a higher number of metacentrics, account-
ing for a slightly lower A1 asymmetry (Figs. 2K, 3K, 4).
Cerinthe major subsp. oranensis from sand dunes near
Tunis was also diploid with 2n = 16. The complement was
similar to that of C. major subsp. purpurascens, in espe-
cially the presence of satellited submetacentrics (Figs. 2I,
3I), but had lower A
1
values due to the presence of six pairs
of metacentrics. To our knowledge this a first report for
both C. major subsp. purpurascens and subsp. oranensis.
Cerinthe minor. All examined accessions were dip-
loid with 2n =18, matching previous countings from other
parts of Europe (Kliphuis, & Wieffering, 1979; Markova
& Goranova, 1995; Dobeš & al., 1997; Favarger, 1997).
However, minor differences in terms of intrachromosomal
asymmetry were detected in relation to the geographic
origin of the accessions. With ten metacentrics and eight
submetacentrics (Figs. 2C, 3C), the complement of C. mi-
nor subsp. minor from peninsular Italy was more symmet-
rical than Bulgarian accessions of the same taxon, which
showed a higher proportion of submetacentrics and a pair
of SAT-chromosomes (Markova & Goranova, 1995). A
similar difference occurs in C. minor subsp. auriculata.
Compared with the asymmetrical karyotypes of plants
from Albania (Figs. 2D, 3D, 4; see also Baltisberger &
Baltisberger, 1995) and Bulgaria (Markova & Goranova,
1995), the accession from S Italy showed a more symmetri-
cal karyotype due to the prevalence of metacentrics (Figs.
2F, 3F, 4). One pair of SAT-submetacentrics occurred in
both examined accessions. Likewise, plants of C. minor
Fig. 2. Micrographs of metaphase chromosome plates. A, Cerinthe palaestina, 2n = 18; B, C. glabra, 2n = 18; C, C. minor
subsp. minor, 2n = 18; D, C. minor subsp. auriculata (Albania), 2n = 18; E, C. tenuiflora, 2n = 18; F, C. minor subsp. auriculata
(S Italy), 2n = 18; G, C. retorta, 2n = 18; H, C. minor subsp. cleiostoma (Croatia), 2n = 18; I, C. major subsp. oranensis, 2n =
16; J, C. major subsp. major (Sardinia), 2n = 16; K, C. major subsp. purpurascens, 2n = 16. Scale bars: 10 μm.
5
Selvi & al. • Phylogeny and taxonomy of Cerinthe
TAXO N • 07 October 2009: 19 pp.
subsp. cleiostoma from Mt. Dhirfis in Evia, Greece and
from the Biokovo range (Croatia), previously referred to
C. tristis, also showed a similar prevalence of submetacen-
trics, accounting for a relatively high A1 asymmetry. No
satellites were observed in these accessions.
Cerinthe palaestina. All examined plants from two
localities in northern Israel were diploid with 2n = 18.
Chromosomes were larger than in other accessions, and
showed secondary constrictions and heterochromatic
segments on both short and long arms. The karyotype
Fig. 3. Karyotypes of A, Cerinthe palaestina; B, C. glabra; C, C. minor subsp. minor; D, C. minor subsp. auriculata (Albania);
E, C. tenuiflora; F, C. minor subsp. auriculata (S Italy); G, C. retorta; H, C. minor subsp. cleiostoma (Croatia); I, C. major
subsp. oranensis; J, C. major subsp. major (Sardinia); K, C. major subsp. purpurascens.
0.280A2
C. minor subsp. auriculata (N Albania)
0.240
0.260
C. minor subsp. minor
C. retorta
0.200
0.220
C. tenuiflora
C. palaestina
0 160
0.180 C. minor subsp. cleiostoma (C Greece)
Ci
cleiostoma (Croatia)
C. major subsp. oranensis
0.140
0
.
160
C
. m
i
nor subsp.
C. minor subsp. auriculata (S Italy)
Cj
major (Sardinia)
0.100
0.120
C
. ma
j
or
C. major subsp.purpurascens
0.150 0.200 0.250 0.300 0.350 0.400 0.450 0.500 0.550
0.080 A1
subsp.
C.glabra (N Italy)
Fig. 4. Bidimensional
ordination of the twelve
examined accessions
of Cerinthe as a func-
tion of the intrachro-
mosomal (A1) and
interchromosomal (A2)
karyotype asymmetry.
6
Selvi & al. • Phylogeny and taxonomy of Cerinthe TA XON • 07 October 2009: 19 pp.
Neatostema apulum
C. tenuiflora (Corsica)
C. minor subsp. auriculata 4 (N Albania)
C. minor subsp. auriculata 3 (C Greece)
C. glabra subsp. glabra 1 (N Italy)
C. glabra subsp. glabra 2 (E Turkey)
C. minor subsp. cleiostoma 1 (S Croatia)
C. minor subsp. cleiostoma 2 (C Greece)
C2
C. major subsp. purpurascens (S Spain)
C1
C1
Onosma visiani
A
10
A
C. minor subsp. auriculata 2 (N Italy)
C. minor subsp. auriculata 1 (S Italy)
C. minor subsp. minor (C Italy)
C. palaestina (N Israel)
C. retorta (Greece)
C. major subsp. oranensis (N Tunisia)
C. major subsp. major 2 (N Morocco)
Echium arenarium
C. major subsp. major 1 (Sardinia)
strictly
perennials
perennials/
biennials
(rar. annual)
strictly
annuals
Fig. 5. A, One of the most
parsimonious phylograms of
Cerinthe (L = 361; CI = 0.85,
RI = 0.87) with optimized
branch lenghts; B, consen-
sus tree from 50% majority-
rule bootstrap analysis.
Suppor t to the branches is
shown when > 50%; nodes
with BS < 50% are indicated
by dotted lines; main clades
(C1, C2) and corresponding
sections with respective
base chromosome number
(x), nutlet morphology and
life- cycle are shown.
Neatostema apulum
Echium arenarium
Onosma visianii
sect
Cerinthe
100
63 x=8
sect
.
Cerinthe
C1
96 96 C. palaestina
94
C. retorta
C. tenuiflora
C2
97
sect. Ceranthe
71
63
61
B
B
C. major subsp. purpurascens
C. major subsp. oranensis
C. major subsp. major 1 (Sardinia)
C. major subsp. major 2 (Morocco)
C. minor subsp. auriculata 4 (N Albania)
C. minor subsp. auriculata 2 (N Italy)
C. minor subsp. cleiostoma 1 (Croatia)
C. minor subsp. cleiostoma 2 (C Greece)
C. minor subsp. auriculata 1 (S Italy)
C. minor subsp. minor
C. glabra subsp. glabra 1 (N Italy)
C. glabra subsp. glabra 2 (E Turkey)
C. minor subsp. auriculata 3 (C Greece)
7
Selvi & al. • Phylogeny and taxonomy of Cerinthe
TAXO N • 07 October 2009: 19 pp.
consisted of medium-sized chromosomes, ten of which
metacentrics, four submetacentrics and four subtelocen-
trics (Figs. 2A, 3A). Intrachromosomal asymmetry was
higher than in most of the other taxa, while A2 was com-
parable (Fig. 4). This is the first report for this species.
Cerinthe retorta. The 18 chromosomes of this Ae-
gean endemic were small-sized and showed no second-
ary constrictions. The complement was formed by four
pairs of metacentrics and ten of submetacentrics, one of
which bearing microsatellites on the short arms (Figs. 2G,
3G); the present report is in line with a previous finding
by Constantinidis (1996). Intrachromosomal asymmetry
was on average values, while difference in size between
chromosomes (A2) was relatively high (Fig. 4).
Cerinthe tenuiflora. Plants from the SW foot of the
Renoso ma ssif i n Corsica were diploid with 18 small ch ro-
mosomes (Figs. 2E, 3E). Our finding does not match the
only previous counting available for this taxon, which is
probably erroneous (2n = 16; Contandrioupoulos, 1964
cited in Bolkhovskikh & al., 1969). The strong prevalence
of metacentrics, a pair of which bearing microsatellites,
accounts for the very low intrachromosomal (A1) asym-
metry (Fig. 4).
Phylogenetic analysis.
The aligned ITS se-
quences, available in TreeBASE (SN4028; P.I.N. 21684)
or directly from the authors in Nexus format, are 700 bp
long including coded gaps which were added at the end
of the matrix (47 bp); rate of variation within the ingroup
is relatively low, with 119 variable positions (17%) in at
least one accession. In the MP analysis, 436 positions were
constant, 126 were variable but uninformative and 138
were parsimony-informative. Branch and Bound search
yielded 75 most parsimonious trees with length (L) = 361,
consistency index (CI) = 0.85 and retention index (RI) =
0.87. One of the most parsimonious phylograms with op-
timized branch lenghts and the consensus tree from 50%
majority-rule bootstrap analysis are shown in Figs. 5A, B.
Cerinthe is retrieved as a monophyletic group (96%
BS) with an early split in two well-supported clades, in
Fig. 5 indicated as C1 and C2. The two lineages differ in
six indels, two in the ITS1 and four in the ITS2, and in
thirty single nucleotide polymorphisms (SNPs), of which
nine are in the ITS1 region, three in the 5.8S and 18 in the
ITS2 region. Clade C1 includes all accessions of the C. ma-
jor group (100% BS), while clade C2 consists of all other
taxa/accessions (94% BS). In clade C1, C. major subsp.
purpurascens and subsp. oranensis are sister to typical
C. major and C. majorgymnandra”, which have almost
identical ITS sequences and cluster together with 63% BS.
Clade C2 shows a basal split in two strongly supported
sister lineages, i.e., the strictly annual group of C. palaes-
tina and C. retorta (96% BS), both from the E Mediter-
ranean, and the primarily perennial group of C. glabra and
C. minor (97% BS). In the latter, the Corsican endemic
C. tenuiflora is consistently retrieved as sister to a weakly
supported clade (71% BS) comprising all accessions of
C. glabra and C. minor s.l., in respect of which it shows
several subst itutions. A polytomy forme d by four lineages
with unresolved relationships is found in the C. minor–
C. glabra clade. A first lineage, supported by 63% BS,
retrieves all but two accessions of C. minor s.l., i.e., subsp.
minor from the Italian peninsula, subsp. cleiostoma from
Greece and Croatia and subsp. auriculata from peninsular
Italy and the western Alps; the accessions from north Al-
bania and central Greece were not retrieved in this group
and remained unresolved. This result suggests the lack of
genetic separation between the short-lived (annual or bien-
nial), and the perennial forms corresponding to C. minor
subsp. auriculata. Finally, the two accessions of C. glabra
from the western Alps and the Pontic mountain range are
retrieved together with 61% BS support, but relationships
of this species with respect to C. minor s.l. are not resolved.
DISCUSSION
Karyotype evolution.
Cerinthe is a strictly dip-
loid group, but inter- and intraspecific variation occurs in
terms of chromosome number and karyotype morphology.
As already suggested (Luque, 1990) two base numbers are
found, x = 9 and x = 8. Mapping their distribution onto the
molecular phylogenet ic cladograms it is clear that they are
associated with a key evolutionary event, i.e., the early
split between the lineage of C. major s.l., invariably with
2n = 16, and that of all other taxa, always with 2n = 18.
Although the two numbers represent true synapomorphies
for their respective clades, the sister relationship between
these lineages does not allow to infer which one of the
two may represent the ancestral karyotypic condition.
Based on previous studies on karyological evolution of
other groups of Old-World Boraginaceae (Bigazzi & Selvi,
2001; Selvi & Bigazzi, 2002; Selvi & al., 2006), however,
we see descending aneuploidy as the most likely process
for the origin of x = 8 from progenitors with x = 9. The
tandem centric fusion of two homologous pairs does not
imply the loss of the requisite genetic material but simply a
decrease in the number of chromosomes as separate link-
age groups at either the diploid or polyploid level (Jackson
& Casey, 1980; Levin, 2002). Evidence for the role of this
process has been provided for several dicot groups (i.e.,
Vicia, Hollings & Stace, 1974; Genista, Sañudo, 1979;
Minuartia, Favarger, 1999; Houstonia, Church, 2003;
various genera of Asteraceae, Garnatje & al., 2004). As
supposed for Nonea (Bigazzi & Selvi, 2003), decrease in
the number and size of chromosomes in the C. major clade
is likely to be associated with the evolution of the strictly
therophytic habit from perennial ancestors during radia-
tion into dry habitats of the Mediterranean. On the other
8
Selvi & al. • Phylogeny and taxonomy of Cerinthe TA XON • 07 October 2009: 19 pp.
hand, our findings on C. palaestina and C. retorta, both
with 2n = 18, show that descending aneuploidy did not
occur in all of the annual xerophytic species of Cerinthe.
A different evolutionary trend emerges therefore in C. pa-
laestina, which has undergone an increase in chromo-
some size and intrachromosomal asymmetry in contrast
with its sister species C. retorta, which has maintained
the presumably ancestral karyotypic traits of the primar-
ily perennial taxa of clade C2, such as C. tenuiflora, C.
glabra and C. minor subsp. auriculata. Compared with
C. tenuiflora, a slightly higher asymmetry and degree of
karyotype variation is instead found in C. glabra and the
C. minor s.l. group, matching the remarkable phenotypic
polymorphism shown by most of the members of this
clade. Compared with SE European material of C. minor
subsp. minor, C. minor subsp. auriculata and C. glabra
(Markova & Goranova, 1995), the Italian accessions of the
C. minor group show a lower asymmetry, suggesting an
East-to-West geographic trend of increase in the number
of submetacentric and subtelocentric chromosomes.
Finally, this study shows that karyotype variation
within the C. major group is low and concerns mostly
satellited chromosomes. Their presence in C. major subsp.
purpurascens from S Spain and subsp. oranensis from
North Africa emerged as a significant diagnostic character
with respect to typical C. major and C. majorgymnan-
dra” from Morocco, in which they are lacking (Altamura
& al., 1984; Luque, 1990; Coppi & al., 2005). Such a dif-
ference is matched by ITS sequences, which are nearly
identical in the two latter accessions but slightly different
in C. major subsp. oranensis and subsp. purpurascens.
Phylogeny, taxonomy and life cycle.
Cerinthe
is traditionally subdivided in the two sections Cerinthe,
including the annual taxa with tubulose corolla and very
short teeths recurved outwards (C. major), and Ceranthe
Rchb. comprising all taxa of the C. minor group with
longer, often connivent teeth (Candolle, 1846; Popov,
1953). Molecular data confirm that the genus comprises
two early-diverging lineages differing in six indels and
thirty SNPs, in the ITS1, ITS2 and 5.8 S regions. However,
there is evidence that corolla structure is not the proper
morphological character on which to base separation of
the two lineages, since C. palaestina and, less, C. retorta
have basically the same corolla structure of C. major but
are phylogenetically closer to the C. minor group. The long
tubulose-cylind rical corolla w ith tiny teeths recurved out-
wards is therefore to be interpreted as either a homoplasy
or a conserved plesiomorphism. On the contrary, synapo-
morphies for the two sections are the chromosome number
and the shape of the cicatrix at base of the mericarpid.
Cerinthe sect. Cerinthe is characterized by the base x = 8
and a concave cicatrix (or attachment areola) forming the
whole base of the nutlet, while C. sect. Ceranthe should in-
clude all taxa with x = 9 and with a smaller, plane cicatrix
(Fig. 5B). Unlike in Candolle (1846), C. retorta should
therefore be placed in C. sect. Ceranthe, in spite of its
resemblance to C. major in terms of life-cycle and mor-
phology of the corolla. The same applies to C. palaestina,
which was not assigned to one of the two sections, but
always been supposed close to C. major. Consequently,
there is evidence that the annual habit has originated inde-
pendently in the two clades, and in particular in C. major
s.l. (C1), C. palaestina/C. retorta and, to a much lesser
extent, in the group of C. minor/C. glabra (C2).
At a lower taxonomic level, the present analysis cor-
roborates the recognition of North African C. major subsp.
oranensis and the Spanish endemic subsp. purpurascens
as independent taxa, based also on morphological and
ecological differences with respect to typical C. major,
while it shows the close proximity between the latter and
the variant with partially exserted anthers described from
S Italy as C. gymnandra (see below).
In clade C2, there is strong support for the indepen-
dent status of C. palaestina, previously considered as a
doubtful entity (Greuter & al., 1984) or even included in
C. major (Edmondson, 1979). This Syro-Palestinian en-
demic is closely related to C. retorta as ably supposed by
Samuelsson (1943) based on the narrow and plane cicatrix
of the nutlet (Fig. 9 on p. 12). A second important point
in clade C2 concerns C. tenuiflora, mostly considered
a subspecies of C. glabra (Domac, 1972; Jeanmonod &
Gamisans, 2007), with a few exceptions (e.g., Pignatti,
1982). ITS data show this Corsican endemic to be geneti-
cally different from C. glabra and possibly the earliest
diverging lineage in the C. glabra–C.minor clade, sug-
gesting it to be the species closer to the basal ancestor
of this perennial group. C. tenuiflora occurs with a few
populations on the siliceous massifs of Corsica which were
part of the “Protoligurian” mountain chain of Hercynian
origin that connected the Eastern Pyrenées to the Western
Alps until the split-off of the Corso-Sardinian microplate
in the late Oligocene (ca. 30 million years ago; Westphal
& al., 1976; Cherchi & Montadert, 1982; Speranza & al.,
2002). Cerinthe glabra is instead strictly continental and
discontinuously distributed in the Pyrenees, Alps, Car-
pathians, Pontic Alps and Caucasus. Given the lack of
long-distance seed dispersal in Cerinthe (mericarpids are
relatively large, heavy and smooth), an ancient vicariance
event triggered by the split-off of the Corso-Sardinian
plate with the consequent break of the continuous con-
tinental range of the ancestral lineage may provide an
explanatory model for the divergence between C. glabra
and C. tenuiflora.
Within the C. minor group, the relatively low karyo-
type and molecular variation show the genetic affinity
between the perennial forms of higher altitudes, often
referred to as C. minor L. subsp. auriculata (Ten.) Domac
and the annual/biennial forms of lowlands such as typical
9
Selvi & al. • Phylogeny and taxonomy of Cerinthe
TAXO N • 07 October 2009: 19 pp.
C. minor and C. minor subsp. cleiostoma from the Bal-
kans. These forms represent geographical races with weak
morphological differentiation and are here treated at the
infraspecific rank.
TAXONOMIC SYNOPSIS
Cerinthe L., Sp. Pl.: 136. 1753 – Type: C. major L.
Annual, biennial or perennial herbs. Leaves subgla-
brous and ± glaucous, often adaxially white-spotted or
scabrid-tuberculate. Cymes terminal, branched, nodding
at first and often strongly elongated in fruit, prominently
bracteate. Calyx dissected to base, with unequal sepals.
Corolla tubulose-cylindrical with lobes of variable length
and shape, usually yellow, often with reddish or violet
spots at especially the throat, rarely entirely dark vio-
let; faucal scales and folds absent. Stamens included or
equalling corolla, rarely exserted, filaments inserted near
middle of tube, often short, flat and dilated, anthers sag-
ittate to hastate, acuminate at apex and connate at base
through narrow coherent appendages. Style gynobasic,
often exserted. Mericarpids two, each bilocular, develop-
ing into globular, shortly beaked nutlets formed by two
connate locules; surface smooth, basal attachment scar
(cicatrix) of variable size, plane or concave.
Key to species
1. Corolla divided to at least 1/2 with erect, ± connivent,
lanceolate and acute lobes. . . . . . . . . . . .4. C. minor
1. Corolla divided to max. 1/4 with tiny, broadly trian-
gular lobes recurved outside, never connivent . . . . 2
2. Corolla l5–25(–30) mm long and 4–7 mm wide . . . 3
2. Corolla 8–12 mm long and 2.5–4 mm wide . . . . . . 4
3. Corolla sulphur-yellow with or without a purple belt;
stamen filaments dilated at base; mericarpids with a
concave cicatrix encircled by a thin rim and occupy-
ing the whole basal surface . . . . . . . . . . . 1. C. major
3. Corolla pale yellow with dark tiny spots between
lobes; filaments f lat but not dilated at base; basal
areola of the mericarpids plane, smaller, not occupy-
ing the total surface of the base . . . 2. C. palaestina
4. Annual; bracts deep purple-violet; corolla curved
above, narrow, somewhat gibbous and constricted at
throat, pale yellow and dark violet distally. . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. C. retorta
4. Perennial (rarely biennial); bracts green, corolla
clavate to cylindrical, straight and regular, not dark-
violet distally. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
5. Basal leaves intensely white-spotted; bracts lanceo-
late, up to 3.5 mm wide; sepals ca. 2 mm wide, corolla
ca. 14 m m long, slightly clavate, entirely orange-rust;
mericarpids 3.3 × 2.9 mm. . . . . . . . 5. C. tenuiflora
5. Basal leaves weakly white-spotted; bracts ovate-
lanceolate, up to 6 mm wide; sepals ca. 3 mm wide,
corolla ca. 12 mm long, tubulose-cylindrical, yel-
low with reddish belt at throat; mericarpids ca.
4.0 × 3.2 mm. . . . . . . . . . . . . . . . . . . . . 6. C. glabra
Cerinthe sect. Cerinthe
1. Cerinthe major L., Sp. Pl.: 136. 1753.
Annual. Leaves often with ± dense, small, rough tu-
bercles and shortly ciliate margins. Corolla 14–22 mm
long, 4–8 mm broad, cylindrical, sulphur-yellow, with or
without purple belt at throat or at the lower half, rarely
entirely purple-violet (subsp. purpurascens). Filaments
dilated at base; anthers included or partially exserted
(“C. gymnandra”). Mericarpids 3.2–6.5 mm long, 2.5–
5.0 mm wide, with cicatrix occupying the whole basal
surface, slightly concave, with a thin rim all around.
A widespread Mediterranean species with strong
variation in the density of the tubercles on leaves, the
colour of corolla, and the position of the anthers. Based
on morphology, karyology and ITS sequence variation,
three subspecies are recognised here.
1. Plant up to 30 cm; corolla ca. 14 mm long and 4–6
mm wide; mericarpids 3.0–3.5 mm long, shiny white-
ivory with dark-brown spots . . . b. subsp. oranensis
1. Plant up to 50(60) cm; corolla 18–22 mm long,
6–8 mm broad; mericarpids 4.0–6.5 mm long, grey-
brown with dark spots, not shiny . . . . . . . . . . . . . . 2
2. Corolla completely or partially yellow, with or without
dark red belt at throat; mericarpids ca. 5 mm long and
4 mm wide. . . . . . . . . . . . . . . . . . . . a. subsp. major
2. Corolla entirely purple-violet, usually without yellow
component; mericarpids ca. 6.5 mm long and 5 mm
wide. . . . . . . . . . . . . . . . . . . c. subsp. purpurascens
a. subsp. major Lectotype (Verdcourt, 1993: 32): Herb.
Clifford: 48, Cerinthe 2 (BM-000557935).
= C. aspera Roth, Catal. Bot. 1: 33. 1797 – Type not
traced.
=
C. gymnandra Gasp. in Rendiconti Accad. Sci. Soc.
Borbon. Napoli 1(3): 72. 1842 C. major var. gym-
nandra (Gasp.) Guadagno in Fiori & Béguinot, Sched.
Fl. Ital. Exsicc., ser. 3: 139. 1917 C. major subsp.
gymnandra (Gasp.) Rouy, Fl. France 10: 279. 1908
Lectotype (Valdés, 1981: 101): “prope Neapolim”,
Gasparrini (PAV !).
=
C. gymnandra
var.
macrosiphonia
Murb. in Acta
Univ. Lund 34 (2,7): 17. 1898
described from
Morocco and Algeria. Type not traced.
=
C. major
var.
elegans
Fiori, Nuov. Fl. Italia 2: 267.
1926
C. major subsp. elegans (Fiori) Giardina &
Raimondo in Bocconea 20: 10. 2007 Lectotype (des-
ignated here): “Cerinthe aspera form. semipurpurea
10
Selvi & al. • Phylogeny and taxonomy of Cerinthe TA XON • 07 October 2009: 19 pp.
Strobl, Prov. di Girgenti: prope Porto Empedocle, in
ar vis col lium, alt. 60 m. circa, solo argilloso 23 Febr.
1906, Fiori, Sched. Fl. Ital. Exs. No. 937 (FI!).
=
C. gymnandra var. iberica Caball. in Anales Jard.
Bot. Madrid 5: 509. 194 4 Holot ype:habitat a d vias
et in dumosis prope Baños de Montemayor (Cáceres)”,
20 Mai 1944, Caballero (MA).
=
C. gymnandra
var.
quichiotis
Car rasco in Bot. Com-
plutensis 18: 170. 1993
Holot
ype: “Ciudad Real,
Lagunas de Ruidera, base de los cerros frente a
la laguna La Colgada al lado de la carretera”, 09
Mai 1980,
Carrasco
&
Velayos
(MACB 47373, MA;
iso- MACB 47374, MA).
Diagnostic characters. – Annual up to 50(60) cm.
Corolla cylindrical, 18–22 mm long and 6–8 mm wide,
completely sulphur-yellow or with a dark purple belt at
middle or lower half. Anthers included or partially ex-
serted (“C. gymnandra”). Mericarpids 4–5 mm long, 3.5–
4.2 mm wide, grey-brown with blackish spots. Fig. 6.
Distribution. – Mediterranean region from the Ibe-
rian Peninsula and North Africa (eastwards to Libya),
to Greece, Aegean islands and western Turkey (no sure
records from Syria, Jordan, Lebanon, Israel and Egypt).
Fig. 7.
Habitat. – Dry fallow fields, cultivated land, waste
and ruderal places, meadows, road margins, often on
nutrient-rich soils.
Note. – Cerinthe gymnandra Gasp., described from
Naples, is a variant with prominent anthers (Guadagno,
1917; Fiori, 1926: 265–267; Pignatti, 1982) which is com-
mon especially in the W Mediterranean (Iberian peninsula
Fig. 6. Cerinthe major subsp. major. A, h ab it; B, fl ower with
bract; C, open corolla; D, mericarpid. Scale bar: A = 30 mm;
B & C = 6 mm; D = 3 mm. Drawing by A. Maury.
Fig. 7. Distribution of Cerinthe major subsp. major, C. major subsp. oranensis, and C. major subsp. purpurascens.
11
Selvi & al. • Phylogeny and taxonomy of Cerinthe
TAXO N • 07 October 2009: 19 pp.
and NW Africa). However, anther position has a weak
taxonomic value due to broad variation even within
populations and lack of correlation with other diagnostic
characters. In addition, dried plants of C. major which
had anthers included when fresh often show exserted
stamens due to the shortening of corolla during exsic-
cation. According to Valdés (1981), C. gymnandra has
yellow corollas without purple belt in the middle pa r t and
smaller nutlets. However, completely yellow f lowers are
found in several populations of typical C. major in Italy
and Greece, while mixed populations with either entirely
yellow or yellow/red corollas are found in the type area
around Naples (Guadag no, 1917). Finally, the mer icar pids
of “C. gymnandra” from southern Italy are as large as in
typical C. major, while they are considerably smaller in
C. major subsp. oranensis from the coa sts of Nor th Af rica.
b. subsp. oranensis (Batt.) Selvi & Cecchi, stat. nov.
C. oranensis Batt. in
Compt. Rend. Assoc. Franç.
Avancem. Sci., 16(2): 572. 1888
C. gymnandra var.
oranensis (Batt.) Murb. in Acta Univ. Lund 34 (2,7):
18. 1898
Lectotype (Valdés, 1981: 103): [Algeria]
“La Macta”, April 1878,
Battandier
(MPU!).
Diagnostic characters. An nual, up to 30 c m tall. Usu-
ally much branched from base. Corolla 14 mm long and
4.5–5.5 mm wide, often slightly gibbous and restricted
at the apex. Anthers more or less exserted. Mericarpids
small, ca. 3.2 mm long, shiny white-ivory spotted of dark-
brown; nutlets are often asymmetrical due to the abortion
of one of the two embryos in the two attached locules.
Fig. 8.
Distribution. – Mediterranean belt of North Africa
from Morocco to Lybia (Fig. 7). This taxon is largely
sympatric with typical C. major, but it is ecologically re-
stricted to coastal sandy places.
Habitat. – Sandy places and fields near to the sea,
maritime dunes.
Note. – Often considered within C. gymnandra (e.g.,
Murbeck, 1898; Valdés, 1981) due to the exserted anthers,
it is distinguished from C. major subsp. major by karyo-
logical, molecular and morphological characters, as well
as its mainly psammophytic ecology.
c. subsp. purpurascens (Boiss.) Selvi & Cecchi, stat. nov.
C. major var. purpurascens Boiss., Voy. Bot. Midi
Esp. 2: 421. 1841 Lectotype (Burdet & al., 1983:
406): “in sylvaticis circa San Roque”, Mai 1837, Bois-
sier (G !).
Diagnostic characters. – Leaves glaucous-pruinose
and nearly glabrous, bright green. All cymes, bracts, ca-
lyces and corollas tinged with dark purple-violet (at least
in the wild accessions). Mericarpids large, up to 6.5 mm
long and 5 mm wide.
Distribution. – Endemic to S Spain (Fig. 7).
Habitat. – Fields, pastures, road margins, ruderal sites.
Note. – This subspecies differs from typical C. major
in karyotype features, ITS sequences (see above), and
morphology.
Cerinthe sect. Ceranthe
Rchb., Iconogr. Bot. Pl. Crit.:
tab. 491. 1827 – Type: C. minor L.
2. Cerinthe palaestina Eig & Sam. in
Bot. Not. 1943,
353. 1943
Holotype: “Transjordan, Wadi Shaib,
in glareosis, ca. 300 m”, 26.3.1933,
Samuelsson
2558 (S).
Robust annual, up to 50 cm tall, glabrous. Leaves
green-glaucous, abundantly white-spotted, amplexicaul.
Bracts often tinged with bluish. Calyx with five strongly
unequal, free sepals inserted at different heights, mi-
nutely ciliate at margins, ca. 15 mm. Corolla 20–25 mm
long, with tube progressively narrowed towards base
and narrowly campanulate limb, pale yellow with tiny
dark-purplish dots between lobes, which are recurved
outwards. Filaments flattened but not dilated at base,
Fig. 8. Cerinthe major subsp. oranensis. A, habit; B, f lower
with bract; C, open corolla; D, mericarpid. Scale bar: A =
30 mm; B & C = 6 mm; D = 3 mm. Drawing by A . Maur y.
12
Selvi & al. • Phylogeny and taxonomy of Cerinthe TA XON • 07 October 2009: 19 pp.
with a dark spot at the attachment point on the corolla
tissue. Mericarpids ca. 5.0 × 3.2 mm, variegated brown
and blackish, not shiny; cicatrix plane, smaller than the
basal surface. Fig. 9.
Distribution. – Israel, southern Syria and western
Jordan (Fig. 10).
Habitat. – Fields, cultivated areas, ruderal places, on
nutrient-rich soils.
Note. – Placed by
Edmondson (1979) in synonymy
of
C. major
and considered as doubtful in Med-Check-
list (Greuter & al., 1984), this is a remarkable endemic
sister to
C. retorta
and well distinct in the morphology
of corolla and mericarpids (Feinbrun-Dothan, 1978:
74). It belongs to
C.
sect.
Ceranthe
(clade C2), in spite
of superficial resemblance to
C. major
(sect.
Cerinthe
,
clade C1).
3.
Cerinthe retorta
Sibth. & Sm., Fl.
Graec
. Prodr. 1:
120. 1806
Lectot
ype (designated here): “C. intorta
Sibth.”
Sibthorp
(Sib-0419a, OXF!).
Annual, up to 40 cm tall. Cymes, bracts and ca-
lyces intensely tinged with purple-violet. Corolla ca.
13.0 × 3.5 mm, irregular cylindrical, narrow and gibbous,
semiclosed at apex, yellowish in the middle part and dark
purplish-violet at tip; lobes tiny, reflexed. Mericarpids ca.
4.0 × 3.5 mm, dark brown with blackish spots, not shiny;
cicatrix plane, smaller than the whole basal surface.
Fig. 11 (also nicely illustrated
in Smiths
Fl. Graeca
2:
171, 1816)
.
Distribution. – Endemic to the Aegean region, fre-
quent in mainland Greece, most Aegean islands, Crete;
Fig. 9. Cerinthe palaestina. A, h abi t; B, fl ow er wi th br ac t; C,
open corolla; D & E, mericarpids. Scale bar: A = 30 mm; B
& C = 6 mm; D & E = 3 mm. Drawing by A. Maury.
Fig. 10. Distribution of Cerinthe glabra subsp. glabra, C. glabra subsp.smithiae, C. tenuiflora, C. retorta and C. palaestina.
13
Selvi & al. • Phylogeny and taxonomy of Cerinthe
TAXO N • 07 October 2009: 19 pp.
it extends eastwards to SW Anatolia and northwards to
S Macedonia (Fig. 10).
Habitat. Rocky slopes, fields, ruderal places, road
margins, olive groves.
4. Cerinthe minor L.,
Sp
. Pl.: 137. 1753.
Annu al t o perennia l. Cymes shor t or strongly elongate d
in fruit. Corolla divided to 1/2–2/3 into narrowly triangular-
lanceolate to subulate, connivent lobes, with or without
purple belt or spots at throat. Mericarpids 3.1–4.3 mm long
and 2.7–3.2 mm wide, blackish when fresh then turning
grey-brown, cicatrix small and plane, without rim.
A widespread species occurring in Central and SE Eu-
rope eastwards to Iran and Black Sea region with weakly
differentiated races. According to Rothmaler (1944) “the
whole group is too little studied and it appears that it must
be further subdivided; up to now it has been attempted
only with very li mited success”. Evidence f rom morphol-
ogy, distribution, ecology and DNA sequences suggests
that three main races can be recognised at the subspecific
rank mainly due to continuous variation.
1. Perennial with woody rootstock; basal and lower cau-
line leaves present at anthesis; cauline leaves broadly
ovate-amplexicaul up to 5.5–7.5 × 3.5–5.0 cm; inflo-
rescence usually short and poorly branched; corolla
11–13 mm; mericarpids ca. 4.3 × 3.2 mm. . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .a. subsp. auriculata
1. Annual to biennial with parenchymatous root; basal
and lower cauline leaves withering before anthesis,
upper cauline leaves ovate, semiamplexicaul, up to
3.5–5.0 × 2.0–3.5 cm; cymes much-branched and
elongated; corolla 7–10 mm long; mericarpids ca.
3.1 × 2.7 mm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Mainly annual; basal leaves oblong-lanceolate; corolla
7–8 mm, without red belt at throat, divided to 1/2 into
triangular-lanceolate lobes ca. 3.5 × 1.0 mm . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . b. subsp. minor
2. Mainly biennial; basal leaves broadly ovate; co-
rolla 9–10 mm, with or without red belt at throat,
divided to 1/2–2/3 into subulate, acuminate lobes ca.
4.3 × 0.7 mm . . . . . . . . . . . . . . .c. subsp. cleiostoma
a. subsp. auriculata (Ten.) Domac in B
ot. J. Linn. Soc.
65(2): 260. 1972
C. auriculata Ten., Ind. Se m. Hor t.
Neap.: 10. 1830 Lectotype (Selvi & Cecchi, 2009a:
622): “in pascuis montosis Lucaniae, Pollino”, Tenore
(FI!).
= C. maculata L., Sp. Pl.: 137. 1753 Neotype (Ed-
mondson, 2004: 800): “Italy, Potenza, Monte Pol-
lino, Coppola di Paola towards Col di Dragone, ca.
1500 m”, 21 Jun 1979, Davis & al. 65757 (BM!; iso-
neo- E).
Diagnostic characters. – Robust perennial with
thick rootstock and basal leaves persisting at anthesis,
somewhat leathery; cauline leaves largely ovate-amplex-
icaul. Inflorescence poorly branched, usually with short
cymes bearing flowers at the axil of foliaceous bracts.
Pedicels, bracts and sepals often ciliate at the margins.
Corolla 11–12 mm long, divided to ca. 1/2 into narrowly
triangular, connivent lobes, usually with a broad purple
belt at throat. Mericarpids ca. 4.1 × 3.2 mm, dark-brown,
blackish-spotted, cicatrix plane and relatively small, not
occupying the whole basal surface. Fig. 12.
Distribution. – Widespread in the mountains of south-
ern Europe, from SE France (Hautes Alpes, Maritime
Alps) to Iran, and especially in Italy (from central Appe-
nines to Calabria and Sicily), Balkan peninsula, Greece
southward to Mt. Taygetos in the Peloponnese, Anatolian
Turkey, Syria and Lebanon (Fig. 13).
Habitat. – Mountain pastures and slopes, clearings
and margins of mountain forests with beech and conifers,
on a variety of substrates.
Note.Cerinthe maculata L. has been recently neo-
typified with a specimen (BM!) from the same locality of
the type collection of C. auriculata Ten., i.e., Mt. Pollino
Fig. 11. Cerinthe retorta. A, habit; B, flower with bract; C,
open corolla; D & E, mericarpids. Scale bar: A = 30 mm; B
& C = 6 mm; D & E = 3 mm. Drawing by A. Maury.
14
Selvi & al. • Phylogeny and taxonomy of Cerinthe TA XON • 07 October 2009: 19 pp.
in southern Italy (Edmondson, 2004). The two specimens
are nearly identical and clearly belong to the same taxon.
The correct name at species rank would, therefore, be-
come C. maculata, but maintaining the taxon as a subspe-
cies the epithet “auriculata” in current use has priority
and is maintained.
Populations from Mt. Taygetos in the Peloponnese
were named C. macrophylla Boiss. based on Heldreich’s
collection n. 384 of July
1844, today conserved
in several
herbaria. However, this name does not seem to be actu-
ally published but only applied to herbarium specimens.
As observed by past Italian authors (Gola, 1904;
Lacaita, 1913), C. minor subsp. auriculata is distinct
from typical C. minor in the perennial habit with thick
rootstock, the basal leaves persisting at anthesis, the
shorter inflorescence with large foliaceous bracts, the
corolla up to 11–12 m long and the larger nutlets. Cer-
inthe minor subsp. auriculata is usually found at higher
altitudes and in more mesic habitats than subsp. minor and
subsp. cleiostoma. However, the occurrence of interme-
diate forms with the latter in the S Balkans and with the
former in SE Europe, the broad geographic sympatrism
and the lack of consistent variation in ITS sequences and
karyotype morphology, show the conspecificity of the
three races.
b. subsp. minor Lectotype (Edmondson, 2004: 801):
Herb. Burser XIV(2): 38 (UPS).
Diagnostic characters. – Annual or biennial up to
50 cm tall. Basal leaves oblong-lanceolate usually wither-
ing before anthesis, weakly white-spotted; cauline leaves
Fig. 12. Cerinthe minor subsp. auriculata; A, habit; B, flow-
er with bract; C, open corolla; D & E, mericarpids. Scale
bars: A = 30 mm; B & C = 6 mm; D & E = 3 mm. Drawing by
A. Maury.
Fig. 13. Distribution of Cerinthe minor subsp. minor, C. minor subsp. auriculata, and C. minor subsp. cleiostoma.
15
Selvi & al. • Phylogeny and taxonomy of Cerinthe
TAXO N • 07 October 2009: 19 pp.
semiamplexicaul, smaller. Cymes elongated in fruit, more
or less st raight and erect, bracts narrowly oblong, rou nded
at base. Sepals ca. 3.5 mm wide, sparsely ciliate at the
margins. Corolla 7–8 mm, divided to ca. 1/2 into narrowly
triangular lobes (0.9–1.0 mm wide at base), yellow, with-
out purple belt at throat but sometimes with tiny purplish
spots at the base of lobes. Mericarpids 3.1 × 2.7 mm, dark
grey-brown, not shiny. Fig. 14C, D.
Distribution. – Central and eastern Europe southward
to central Italy and the Black Sea region (Fig. 13).
Habitat. – Fields, cultivated areas, ruderal places,
road margins, usually in plains and lowlands.
Note. – In its typical form it is an annual species grow-
ing in ruderal and cultivated lowland sites of Central and
eastern Europe (“habitat in Austriae, Styriae agris”).
c. subsp. cleiostoma (Boiss. & Sprun.) Selvi & Cecchi,
comb. & stat. nov. C. cleiostoma Boiss. & Sprun.
in Boiss., Diagn. Pl. Orient. 4: 44. 1844 Lecto-
type (designated here): “in lapidosis mobilium fau-
cium Hymetti prope Athenas”, Apr 1842, Boissier &
Spruner (G-BOIS !; isolecto- FI!, NAP!).
= C. lamprocarpa Murb. in Acta Univ. Lund 27(5): 85.
1891 Lectotype (designated here): “Cerinthe lam-
procarpa nov. spec.// Hercegovina: ad latera montis
Bjelašica planina 12–1300 m 17/8/1889”, Murbeck
(LD 1213158!).
= C. tristis Teyb e r in
Oesterr. Bot. Z., 63: 491. 1913
Holotype: “in rupibus declivium australium montis
Biokovo prope oppidum Makarska, ca. 1000 m,
floret Julio Augusto”,
Teyber
(W!).
Diagnostic characters. – Biennial with parenchyma-
tous root, up to 75 cm tall. Basal leaves forming a rosette
the first year and withering before anthesis the second
year, oblong-spathulate, somewhat glaucous and white-
spotted, sometimes tinged of dark-purple at the margins.
Lower cauline leaves also withering before anthesis.
Upper cauline leaves and bracts oblong to suborbicular.
Cymes much branched, with many flowers clustered at
first, then elongating and becoming straight and erect.
Bracts ovate-cordate. Sepals ca. 2.3 mm wide, minutely
ciliate at margins. Corolla ca. 9 mm, slightly swollen in
the middle part, deeply divided (to 1/2–2/3) into narrow
(ca. 0.7 mm wide at base), acuminate lobes connivent
in flower, often, but not always, tinged with purplish at
throat. Mericarpids ca. 3.2 × 2.8 mm, dark brown and
blackish spotted. Fig. 14A, B
Distribution. – Croatia, Montenegro, Bosnia-Her-
cegovina and Greece, southwards to the Peloponnese
(Fig. 13).
Habitat. – Steep rocky slopes, ravines, cliffs, usually
facing to the west and on limestone, 200–1,300 m.
Note. This subspecies represents the southern Balkan
race of C. minor, in respect of which it shows a weak
morphological differentiation in the larger, suborbicular
bracts and the more deeply divided corolla with narrow,
subulate lobes. Unlike C. minor, it grows in primary habi-
tats such as steep rocky slopes and ravines, usually on
limestone massifs near to the coast. Based on examination
of the type material, C. cleiostoma is the first specific
name given to this race, which was later described from
also Bosnia-Hercegovina as C. lamprocarpa (Murbeck,
1891) and from the mountains of Biokovo as C. tristis
based on a sterile collection (Teyber, 1913). The identity
of the latter has remained uncertain (see Domac, 1972;
Greuter & al., 1984) until its rediscovery in the type area
(Domac, 1988). Based on our observations on the Biokovo
populations, however, this taxon is
not sufficiently dis-
tinct from the other Balkan accessions with a biennial
habit to be kept as a separate taxon. The purple ma rgins
of the basal leaves, upon which it was described, are
not constant, while other characters show its conspe-
cific nature with
C. minor
and its correspondence with
C. minor
subsp.
cleiostoma
.
Fig. 14. Cerinthe minor subsp. cleiostoma: A , ha bi t; B , open
corolla. Cerinthe minor subsp. minor : C, inflorescence; D,
flower; E, mericarpids. Scale bar: A = 30 mm; B & C = 6 mm;
D & E = 3 mm. Drawing by A. Maury.
16
Selvi & al. • Phylogeny and taxonomy of Cerinthe TA XON • 07 October 2009: 19 pp.
5. Cerinthe tenuiflora Bertol., Fl. Ital. 2: 325. 1836
C. glabra Mill. subsp. tenuiflora (Bertol.) Domac
in Bot. J. Linn. Soc.
65(2): 260. 1972
Lectotype
(Selvi & Cecchi, 2009a: 622): “habui ex Corsica a
Cervione a Soleirolio” (BOLO!).
Robust perennial, up to 65 cm. Basal leaves persis-
tent at anthesis, ovate-oblong, thick, dark green strongly
white-spotted in the upper surface. Stems often reddish in
the upper part, fleshy, much branched. Cymes elongated in
fruit, with linear-lanceolate bracts, 15–20 × 3.2–3.5 mm.
Sepals ca. 2 mm wide at base; corolla 13–14 mm, nar-
rowly cylindrical-clavate, with longitudinal folds between
lobes, diffusely tinged with orange-rust. Mericarpids ca.
3.3 × 2.8 mm, variegated grey-brown and blackish, shiny,
cicatrix occupying nearly all basal surface. Fig. 15.
Distribution. – Endemic to the main massifs of Cor-
sica from Capicorsu to the mountains of Cagna (Jean-
monod & Gamisans, 2007; Fig. 10).
Habitat. – Mountain slopes, damp scrubs and clear-
ings, humid woods along streams, mainly on granite, 300 –
1,500 m.
Note. – A remarkable endemic differing from C. gla-
bra by the narrow bracts, sepals and corolla, which is
narrowly cylindrical-clavate in shape and orange-rust
in colour, and the smaller mericarpids (Pignatti, 1982).
Cerinthe longiflora Viv., described from the the island
of Tavolara (Sardinia) and sometimes considered a syn-
onym of C. tenuiflora, is an ambiguous name but should
be probably refer red to C. major, based on morphological
description and habitat (Greuter & al., 1984).
6. C. glabra Mill., Gard. Dict., ed. 8: n° 2. 1768, nom.
cons. prop.
Perennial, rarely biennial. Basal and lower cauline
leaves mostly present at anthesis, rarely withering before,
broadly oblong-ovate, weakly white-spotted. Bracts ovate,
15–20 × 5–6 mm. Sepals ca. 3 mm wide; corolla 11–12
mm, cylindrical with longitudinal folds, with short lobes
recurved outside, pale yellow with reddish belt at throat.
Mericarpids variegated grey-brown and blackish, ca.
3.9 × 3.4 mm.
1. Perennial; basal leaves present at anthesis, sepals lan-
ceolate; anthers included. . . . . . . . . a. subsp. glabra
1. Biennial; basal leaves withering before anthesis; sepals
ovate; anthers partially exserted . .b. subsp. smithiae
a. subsp. glabra Type (Selvi & Cecchi, 2009b): Italy,
Piedmont, Valle Gesso (prov. Cuneo), SS. Trinità di
Entracque, margini di bosco umido nella valletta del
Rio Prer, 1080 m, Selvi 06.25 (FI, K), typ. cons. prop.
= C. alpina Kit. in Schultes, Österreichs Fl., ed. 2. 1:
353. 1814 Lectotype (designated here): “Cerinthe
alpina mihi / “nova etiam Willdenowio. Specimen
altius misit Schradero erga restitutionem”, Mauksch,
Herbar. Kitaibel no. 68 (BP!; see also Jávorka, 1926).
– Syntypes: “Cerinthe alpina mihi / Erga restitutio-
nem; nam unicum hoc specimen possideo et in horto
planta periit”, Mauksch, Herbar. Kitaibel no. 105
(BP!; Jávorka, 1926); “Cerinthe alpina foliis amplexi-
caulibus, corollae … apertis, caule erecto / habitat in
erbidis pinguis alpibus Hungariae”, Kitaibel (B-W
03376-000 and -010, B!).
= C. pyrenaica Arv.-Touv., Essai Pl. Dauphiné: 52.
1871 C. glabra Mill. subsp. pyrenaica (Ar v.-Touv.)
Kerguélen, Index Synonym. Fl. France (Coll. Patrim.
Nat., 8): 12. 1993 Described from the Pyrenèes.
Type not traced.
= C. alpina subsp. velebitica Degen & Lengy in Degen,
Fl. Veleb. 2: 578. 1937 Lectotype (designated here):
“in apertis montis Sladikovac ad Ostaria”, 23 Jun
1881, Borbás (ZA!).
= C. glabra subp. caucasica E. Hadac
in Galushko, Fl.
Severn. Kavkaza Vopr. Ist. 3: 18. 1979
Described
from the Caucaus
. T
ype not traced.
= C. glabra var. pirinica Stoj. & Acht., Izv. Prir. Inst. 12:
185. 1939 C. glabra subsp. pirinica (Stoj. & Acht.)
Fig. 15. Cerinthe tenuiflora. A, habit; B, flower with bract;
C, open corolla; D & E, mericarpids. Scale bar: A = 30 mm;
B & C = 6 mm; D & E = 3 mm. Drawing by A. Maury.
17
Selvi & al. • Phylogeny and taxonomy of Cerinthe
TAXO N • 07 October 2009: 19 pp.
N. Andreev & Peev, Fl. Nar. Republ. Bǔlgariya 9:
136. 1989
T
ype (Andreev & Peev, 1989): “Crescit
in glareis marmoreis sub cacumine El-Tepe, supra
refugium Banderica, ad cca. 1850 m s. m.” (SOM).
Diagnostic characters. Robust perennial with thick
rootstock. Basal leaves present at anthesis, bright green,
not or weakly whitish-spotted, bracts (mid-inf lorescence)
7–9 mm wide at base. Sepals lanceolate (ca. 7.0 × 2.5 mm);
corolla ca. 12 × 4 mm, cylindrical, tube pale yellow with
a purple belt at throat, limb yellow turning pale bluish
when dry. Mericarpids ca. 4.0 × 3.2 mm, variegated grey
and black; cicatrix relatively large, occupying almost all
basal surface. Fig. 16A–E.
Distribution. – Mountains of southern Europe: Pyre-
nèes (rare), Alps, Carpathians, Velebit, Bosnia, Rhodopes,
north Anatolian ranges, Pontic Alps, Caucasus (Hegi,
1975: 2186–2191; Meusel & al., 1978). An old record (FI!)
by Orsini from the central Apennines in Abruzzo (C Italy)
needs confirmation. Fig. 10.
Habitat. – Mountain pastures, nitrophilous communi-
ties of tall herbs near alpine huts, clearings and margins
of humid mountain forests, on a variety of substrates,
800–2,500 m.
Note. – The name C. alpina is here typified with a
collection (Herb. Kitaibel no. 68) well corresponding to the
posthumous description by Kitaibel himself (Kanitz, 1863).
Although it should be reinstated in place of the misapplied
name C. glabra, in agreement with important past authors
(e.g., Boissier, 1879: 148–150; Degen, 1937; Popov, 1953),
the latter has been proposed for conservation (Selvi & Cec-
chi, 2009b) in order to save its traditional and well-estab-
lished usage (Art. 14.2 of the ICBN, McNeill & al., 2006).
b. subsp. smithiae (A. Kern.) Domac in Bot. J. Linn.
Soc. 65: 260. 1972 C. smithiae A. Kern., Sched.
Fl. Exsicc. Austro-Hung. 3: 95. 1884 Lectotype
(designated here):Scoglio di S. Marco, Porto Re.
Quarnero”, 2 May 1865, Smith (FI!).
Diagnostic characters. – Biennial with parenchyma-
tous root and basal leaves withering before anthesis. Sepals
ovate (ca. 6 × 3 mm). Anthers partially exserted. Fig. 16 F.
Distribution. – Rare along the Croatian coast from
Istria near Plomin (Mt. Sisol) to Obrovac (Degen, 1937;
Meusel & al., 1978; Fig. 10). Extinct on the small rocky
islet of St. Mark, the type locality.
Habitat. – Rocky places, on limestone, 0–200 m (?).
Note. – Based on morphological characters interme-
diate between typical C. glabra and C. minor, C. glabra
subsp. smithiae may represent a homoploid hybrid origi-
nated along the Croatian coast at the foot of the Velebit
mountains, i.e., the contact between the ranges of C. gla-
bra and C. minor. Repeated field researches in the type
locality of St. Mark, a small rocky islet near Kraljeviča,
have shown that this entity is locally extinct. Also in the
few other known localities, we could not find this entity
in spite of intensive field searches, preventing karyological
and phylogenetic analyses to be made.
ACKNOWLEDGEMENTS
The authors are grateful to H. Vered Leshner (HUJ),
G. Hunt and H.H. Hilger for providing material of C. palaes-
tina, C. retorta and C. major subsp. purpurascens, respectively.
Arne Strid kindly shared relevant information on C. minor in
Greece and D. Uzunov gave useful information on type mate-
rial of C. pirinica. The curators of B, B-W, BM, BOLO, BP,
FI, FI-W, G, G-BOIS, K, LD, MPU, NAP-GUSS, NAP-TEN,
OXF, PAV, W and ZA kindly gave information and allowed the
study of relevant herbarium material. Two anonymous review-
ers helped to improve the first version of the manuscript and J.
McNeill assisted with nomenclatural aspects. Research grants
from University of Firenze and M.I.U.R. are acknowledged.
Fig. 16. Cerinthe glabra subsp. glabra: A, basal rosette;
B, inflorescence; C, flower with bract; D, open corolla; E,
mericarpids. C. glabra subsp. smithiae: F, corolla. Scale
bar: A & B = 30 mm; C, D, F = 6 mm; E = 3 mm. Drawing by
A. Maur y and L . Cecchi (F).
18
Selvi & al. • Phylogeny and taxonomy of Cerinthe TA XON • 07 October 2009: 19 pp.
LITERATURE CITED
Altamura, L., Colasante, M. & D’Amato, G. 1984. Numeri
cromosomici per la f lora Italiana: 1022–1036. Inform. Bot.
Ital. 16: 261–270.
Andreev, N. & Peev, D. 1989. Cerinthe L. Pp. 134 –136 in:
Velchev, V. & Kuzmanov, B. (eds.), Flora Republicae
Popularis Bulgaricae, vol. 9. Bulgarian Academy of Sci-
ence, Sofia.
Baltisberger, M. & Baltisberger, E. 1995. Cytological data of
Albanian plants. Candollea 50: 457–493.
Bigazzi, M. & Selvi, F. 2001. Karyotype morphology and cy-
togeography in Brunnera and Cynoglottis (Boraginaceae).
Bot. J. Linn. Soc. 136: 365–378.
Bigazzi, M. & Selvi, F. 2003. Chromosome variation in Ana-
tolian species of Nonea Medik. (Boraginaceae) with spe-
cial reference to endemics and N. persica. Caryologia 56:
509–520.
Boissier, P.E. 1879. Flora Orientalis, vol. 4. Georg, Genève,
Basel, Lyon
Bolkhovskikh, Z., Grif, V., Matvejeva, T. & Zakharyeva,
O. 1969. Chromosome Numbers of Flowering Plants.
kademiya Nauk SSSR, Leningrad.
Borchsenius, F. 2007. FastGap 1.0.8. Depa rtme nt of Biologi cal
Sciences, University of Aarhus, Aarhus. Available from
http://www.aubot.dk/f b/FastGap_home.htm.
Burdet, H.M., Charpin, A. & Jacquemoud, F. 1983. Types
nomenclaturaux des taxa ibériques décrits par Boissier ou
Reuter. Candollea 38: 401–441.
Ca nd ol le , A .L .P. P. d e. 1846 . Prodromus Systematis Universa-
lis Regni Vegetabilis, vol. 10. Masson, Parisiis.
Cecchi, L. & Selvi, F. 2009. Phylogenetic relationships of
the monotypic genera Halacsya and Paramoltkia and the
origins of serpentine adaptation in circum-mediterranean
Lithospermeae (Boraginaceae): insights from ITS and
matK DNA sequences. Taxon 58: 700–714.
Cherchi, A. & Montadert, L. 1982. Oligo-Miocene rift of Sar-
dinia and the early history of the Western Mediterranean
basin. Nature 298: 736 –739.
Church, S.A. 2003. Molecular phylogenetics of Houstonia (Ru-
biaceae): descending aneuploidy and breeding system evo-
lution in the radiation of the lineage across Nor th America.
Molec. Phylog. Evol. 27: 223–238.
Constantinidis, T. 1996. Mediterranean chromosome number
reports no. 6. Fl. Medit. 6: 308–312.
Coppi, A., Bigazzi, M. & Selvi, F. 2005. Chromosome studies
in Mediterranean Boraginaceae. Fl. Medit. 16: 253–274.
Corpet, F. 1988. Multiple sequence alignment with hierarchical
clustering. Nucleic Acids Res. 16: 10881–10890.
Degen, Á. von. 1937. Flora Velebitica, vol. 2. Un gar isc he A kad -
emie der Wissenschaften, Budapest.
Dobeš, C., Hahn, B. & Morawetz, W. 1997, Chromosomen-
zahlen zur Gefässpflanzen-Flora Österreichs. Linzer Biol.
Beitr. 29: 5–43.
Domac, R. 1972. Cerinthe L. P. 260 in: Heywood, V.H. (ed.),
Flora Europaea Notulae Systematicae ad Floram Eu-
ropeam Spectantes n. 12. Bot. J. Linn. Soc. 65: 223–269.
Domac, R. 1988. Cerinthe tristis — pr avi član Hrvatske f lore.
Acta Bot. Croat. 47: 145–148.
Doyle, J.J. & Doyle, J.L. 1990. Isolation of plant DNA from
fresh tissue. Focus 12: 13–15.
Dyer, A.F. 1979. Investigating Chromosomes. Ar nold, London
Edmondson, J. 1979. Cerinthe L. Pp. 376–378 in: Davis, P.H.
(ed .), Flora of Turkey and the East Aegean Islands, vol. 6.
Edinburgh University Press, Edinburgh.
Edmondson, J. 2004. Cerinthe maculata L. Pp. 800–801 in:
Cafferty, S. & Jarvis, C.E. (eds.), Typification of Linnaean
plant names in Boraginaceae. Taxon 53: 799–805.
Favarger, C. 1997. Notes de caryologie alpienne VI. Bull. Soc.
Neuchât. Sci. Nat.120: 19–33.
Favarger, C. 1999. Contribution à la cytogéographie du Minu-
artia glomerata (M.Bieb.) Degen (Caryophyllaceae). Bull.
Soc. Neuchateloise Sci. Nat. 122: 27–33.
Feinbrun-Dothan, N. 1978. Flora Palaestina, vol. 3. Israel
Academy of Sciences and Humanities, Jerusalem.
Felsenstein, J. 1985. Confidence limits on phylogenies: an ap-
proach using the bootstrap. Evolution 39: 783–791.
Fiori, A. 1926. Nuova Flora Analitica d’Italia, vol. 2. Ricci,
Firenze
Garnatje, T., Vallès, J., Vilatersana, R., Garcia-Jacas, N.,
Susanna, A. & Siljak-Yakovlev, S. 2004. Molecular cy-
togenetics of Xeranthemum L. and related genera (Astera-
ceae, Cardueae). Pl. Biol. 6: 140–146.
Gola, G. 1904. Osservazioni sulla Cerinthe maculata All. Mal-
pighia 18: 355–358.
Greuter, W., Burdet, H.M. & Long, G. 1984. Med-Checklist:
A Critical Inventory of Vascular Plants of the Circum-med-
iterranean Countries, vol. 1, Pteridophyta (ed. 2), Gym-
nospermae, Dicotyledones (Acanthaceae-Cneoraceae).
Conservatoire et Jardin Botaniques, Genève.
Guadagno, M. 1917. Cerinthe major var. gymnandra Gasp. (pro
sp.). P. 139 in: Fiori, A. & Béguinot, A. (eds.), Schedae ad
Floram Italicam Exsiccatam, ser. 3 (13), n° 2321. Tipogr.
Universitaria, Padova.
Gürke, R.L.A.M. 1894. Cerinthe L. P. 127 in: Engler, H.G.A.
& Prantl, K.A.E. (eds.), Die natürlichen Pflanzenfamilien,
vol. 4(3a). Engelmann, Leipzig.
Hegi, G. 1975. Illustrierte Flora von Mitteleuropa, vol. 5(3).
Parey, Berlin und Hamburg.
Hollings, E. & Stace, C.A. 1974. Karyotype variation and
evolution in the Vicia sativa aggregate. New Phytol. 73:
195–208.
Jackson, R.C. & Casey, J. 1980. Cytogenetics of polyploids.
Pp. 17–44 in: Lewis, WH. (ed.), Polyploidy: Biological
Relevance. Plenum Press, New York and London
Jávorka, S. 1926. Kitaibel Herbariuma 1. Herbarium Kitai-
belianum 1. Ann. Hist.-Nat. Mus. Natl. Hung. 24: 428–585.
Jeanmonod, D. & Gamisans, J. 2007. Flora Corsica. Edisud,
Aix-en Provence.
Johnston, I.M. 1924. Studies in the Boraginaceae. III.1. The
Old World genera of the Boraginoideae. Contr. Gray Herb.
73: 42–78.
Kanitz, A. (ed.). 1863. Pauli Kitaibelii Additamenta ad Floram
Hungaricam. Linnaea 32: 305–642.
Kliphuis, E. & Wieffering, T. 1979. IOPB ch romosome num-
ber reports LXIV. Taxon 28: 394–400.
Lacaita, C. 1913. Aggiunte alla flora del Principato Citra. Bull.
Orto Bot. Regia Univ. Napoli 3: 251–307.
ngs tröm, E. & Ch ase , M.W. 2002. Tribes of Boraginoideae
(Boraginaceae) and placement of Antiphytum, Echiochi-
lon, Ogastemma and Sericostoma: a phylogenetic analysis
based on atpB plastid DNA sequence data. Pl. Syst. Evol.
234: 137–153.
Levan, A., Fredga, K. & Sandberg, A.A. 1964. Nomenclature
19
Selvi & al. • Phylogeny and taxonomy of Cerinthe
TAXO N • 07 October 2009: 19 pp.
for centromeric position on chromosomes. Hereditas 52:
201–220.
Levin, D.A. 2002. The Role of Chromosomal Change in Plant
Evolution. Oxford University Press, New York.
Luque, T. 1990. Karyological studies on Spanish Boraginaceae.
V: Onosma L., Cerinthe L. and Alkanna Taus ch ( Lit hosp er-
meae). Bot. J. Linn. Soc. 102: 261–272.
Markova, M. & Goranova, V. 1995. Mediterranean chromo-
some number reports no. 435–473. Fl. Medit. 5: 289 –317.
McNeill, J., Barrie, F. R., Burdet, H. M., Demoulin, V.,
Hawksworth, D. L., Marhold, K., Nicolson, D. H.,
Prado, J., Silva, P. C., Skog, J. E., Wiersema, J. H. &
Turla nd, N. J. (eds .). 2006. International Code of Botani-
cal Nomenclature (Vienna Code) Adopted by the Seven-
teenth International Botanical Congress Vienna, Austria,
July 2005. Regnum Vegetabile 146. Gantner, Ruggell.
Meusel, H., Jäger, E., Rauschert, S. & Weinert, E. 1978.
Vergleichende Chorologie der Zentraleuropäischen Flora-
Karten, vol. 2. G. Fischer, Jena.
Murbeck, S. 1891. Beiträge zur Kenntniss der Flora von Süd-
Bosnien und der Hercegovina. Lunds Univ. Årsskr. 27:
84–88.
Murbeck, S. 1898. Contributions à la Connaissance des Prim-
ulacées-Labiées de la Flore du Nord-Ouest de l’Afrique et
Plus Spécialment de la Tunisie. Malmström, Lund.
Murin, A. & Majovsky, J. 1979. Karyological study of Slo-
vakian f lora I. Acta Fac. Rerum. Nat. Univ. Comenianae,
Bot. 27: 127–133.
Pignatti, S. 1982. Flora d’Italia, vol. 2. Edagricole, Bologna.
Popov, M.G. 1953. Cerinthe L. Pp. 268–271 in: Komarov V.L.
(ed .), Flora SSSR, vol. 19. Moskva-Leningrad. [Transla-
tion: Lavoott, R. 1974. Flora of the U.S.S.R., vol. 19. Keter
Publishing House, Jerusalem.]
Romero Zarco, C. 1986. A new method for estimating karyo-
type asymmetry. Taxon 35: 526–530.
Rothmaler, W. 1944. Floristische Ergebnisse einer Reise nach
dem Peloponnes. Bot. Jahrb. Syst. 73: 418–452.
Samuelsson, G. 1943. Cerinthe palaestina Eig et Sam. n.sp.
Bot. Not. 1943: 353–358.
Sañudo, A. 1979. Chromosome variability in the Genisteae
(Adans.) Benth. Webbia 34: 363–408.
Selvi, F. & Bigazzi, M. 2002. Chromosome studies in Turkish
species of Nonea (Boraginaceae): the role of polyploidy
and descending dysploidy in the evolution of the genus.
Edinburgh J. Bot. 59: 405–420.
Selvi, F. & Cecchi, L. 2009a. Typification of names of Euro-
Mediterranean taxa of Boraginaceae described by Italian
botanists. Taxon 58: 621–626.
Selvi, F. & Cecchi, L. 2009b. Proposal to conserve the name
Cerinthe glabra (Boraginaceae) with a conserved type.
Taxon 58: nnn-mmm. to follow
Selvi, F., Coppi, A. & Bigazzi, M. 2006. Karyotype variation,
evolution and phylogeny in Borago (Boraginaceae) with
emphasis on subgenus Buglossites in the Corso-Sardinian
system. Ann. Bot. 98: 857–868.
Simmons, M.P. & Ochoterena, H. 2000. Gaps as characters
in sequence-based phylogenetic analyses. Syst. Biol. 49:
369–381.
Speranza, F., Villa, I.M., Sagnotti, L., Florindo, F., Cosen-
tino, C., Cipollari, P. & Mattei, M. 2002. Age of the
Corsica-Sardinia rotation and Liguro-Provençal basin
spreading: new paleomagnetic and Ar/Ar evidence. Tec-
tonophysics 347: 231–251.
Swofford, D.L. 1998. PAUP*. Phylogenetic Analysis Using
Parsimony (*and Other Methods), vers. 4.0. Sinauer, Sun-
derland.
Teyber, A. von. 1913. Beiträge zur Flora Österreichs. Oesterr.
Bot. Z. 63: 486493.
Valdés, B. 1981. Cerinthe major L. y C. gymnandra Gasparr.
Lagascalia 10: 98–104.
Ver d court , B . 1993. Cerinthe major L. P. 32 in: Jarvis, C.E.,
Barrie, F.R., Allan, D.M. & Reveal, J.L. (eds.), A List of
Linnaean Generic Names and their Types. Regnum Veg-
etabile 127. Koeltz, Königstein.
Westphal, M., Orsini, J. & Vellutini, P. 1976. Le microconti-
nent corse-sarde, sa position initiale, données paléomagné-
tiques et raccords géologiques. Tectonophysics 30: 141–157.
White, T.J., Bruns, T., Lee, S. & Taylor, J. 1990. Amplif ica-
tion and direct sequencing of fungal ribosomal RNA genes
for phylogenetics. Pp. 315–322 in: Innis, M.A., Gelfand,
D.H., Sninsky, J.J. & White J.W. (eds.), PCR Protocols.
A Guide to Methods and Applications. Academic Press,
New York.
Appendix. List of the investigated ta xa and ac cessi ons with origin, voucher and GenBank n umber s. FI- HB: collection of Bor agina-
ceae in the Herbarium Centrale Italicum (FI).
Taxon, origin, voucher, GenBank accession
Cerinthe glabra Mill.: (1) Italy, Piedmont, Selvi (FI-HB 06.26; FJ541016); (2) Turkey, Artvin region, Altinisli 4164 (FI-HB 07.62; FJ541017); C. major
L. subsp. major: (1) Italy, Sardinia, Cecchi (FI-HB 08.01; EU919583); (2) Morocco, Rharb, Hilger 13/2004 (FI-HB 04.49; FJ541018); C. major L. subsp.
oranensis (Batt.) Selvi & Cecchi: Tunisia, Hamman-Lif, Bigazzi & Selvi (FI-HB 04.46; FJ541019); C. major L. subsp. purpurascens (Boiss.) Selvi &
Cecchi: Spain, Andalusia, Hilger (FI-HB 07.63; FJ541020); C. minor L. subsp. auriculata (Ten.) Domac: (1) Italy, Abruzzo, Bigazzi & Selvi (HB 03.06;
FJ541023); (2) Italy, Piedmont, Selvi (FI-HB 06.28; FJ541024); (3) Greece, Sterea Ellas, Cecchi & Selvi (FI-HB 08.15; FJ541025); (4) Albania, Kukes region,
Cecchi, Coppi & Selvi (FI-HB 06.17; FJ541026); C. minor L. subsp. cleiostoma (Boiss.) Selvi & Cecchi: (1) Croatia, Mt. Biokovo, Cecchi & Selvi (FI-HB
07.49; FJ541021); (2) Greece, Evvia, Cecchi & Selvi (FI-HB 08.20; FJ541022); C. minor L. subsp. minor: Italy, Tuscany, Selvi (FI-HB 06.63; EU919582);
C. palaestina Eig & Sam.: Israel, lower Galilee, Yair Ur (FI-HB 07.60; FJ541027); C. retorta Sibth. & Sm.: culta from seeds from Greece (FI-HB 07.57;
FJ541028); C. tenuiflora Bertol.: France, Corsica, Selvi (FI-HB 06.04; FJ541029).
... Cerinthe retorta Sm. (Selvi et al. 2009, Valdés 2011, Wagensommer et al. 2014. In Italy, it occurs only in Puglia, with a single population on the Gargano Promontory, Valle dell'Inferno, in the municipality of San Giovanni Rotondo (Fig. 4). ...
... Habitat and Ecology: In Italy, C. retorta grows on calcareous rocky slopes, directly on the rocks or among the rocks, at an altitude of 380-400 m a.s.l. Nevertheless, in the Aegean region this species occurs also in secondary sites, such as fields, ruderal sites, road margins, and olive groves (Selvi et al. 2009). ...
... Distribution: Countries of occurrence: Turkey, Greece, Macedonia, Albania, Italy, and doubtfully Croatia.Biology: Plant growth form: Annual (therophyte). Chromosome number: 2n = 18 (material from Greece;Selvi et al. 2009). ...
Article
Full-text available
In this contribution, the conservation status assessment of four vascular plants according to IUCN categories and criteria are presented. It includes the assessment of Epipactis maricae (Croce, Bongiorni, De Vivo & Fori) Presser & S.Hertel at global level, and the regional assessment of Cerinthe retorta Sm. (Italy), Platanthera kuenkelei H.Baumann subsp. kuenkelei (Europe) and Typha elephantina Roxb. (Egypt).
... Rapid diversifications or shifts in diversification rates have been associated with the appearance of specific morphological traits, or adaptive types, purportedly triggering adaptive radiation (Simpson, 1953). These traits have been termed "key innovations" (Schluter, 2000). Overall, Lithospermeae show a range of particular traits that could be interpreted as potential "key innovations", such as fruticose habit, bird pollination, and heterostyly (increasing outcrossing) in Lithospermum (e.g., Thomas et al., 2008;Cohen, 2011Cohen, , 2012Cohen, , 2016a, and edaphic adaptations in Onosma (e.g., Cecchi et al., 2011). ...
Article
Studies about the drivers of angiosperm clade diversifications have revealed how the environment continuously alters the species chances to adapt or to go extinct. This process depends on complex interactions between abiotic and biotic factors, conditioned to the geological and tectonic settings, the genetic variability of species and the rate at which speciation occurs. In this study, we aim to elucidate the timing of diversification of the Lithospermeae, the second largest tribe within Boraginaceae, and to identify the possible morphological and ecological characters associated with shifts in diversification rates of the most species-rich clades. Lithospermeae includes ca. 470 species and 26 genera, among which are some of the largest genera of the family such as Onosma (150 spp.), Echium (60 spp.), and Lithospermum (80 spp.). An exhaustive study of the whole clade is not available to date and its evolutionary history and diversification rates are incompletely known. In the present study, we provide the most comprehensive phylogeny of the group so far, sampling 242 species and all 26 genera. We found that crown-groups and diversification rates of Lithospermeae largely date back to the Mid-Miocene, with high diversification rates in the largest genera, though only significantly high in Onosma. Our analysis fails to associate any of the functional or morphological traits considered with significant shifts in diversification rates. The timing of the diversification of the species-rich clades corresponds with Miocene tectonic events and global climate changes increasing aridity across Eurasia and western North America. These results suggest a causal link between known ecological features of Lithospermeae (i.e., pre-adaptation to arid, open habitats, and mineral soils) and their diversification. Future studies should expand the sampling of individual subclades and detailed functional analyses to identify the contribution of adaptations to arid conditions and pollinator shifts.
... The relative importance of these processes in Boraginaceae is evident (e.g. Bigazzi & Selvi, 2003;Mártonfi et al., 2008;Selvi, Cecchi & Coppi, 2009) and most taxa in this family are reported to have undergone at least Leitch et al., 1998). A higher proportion of polyploid taxa was detected in Cynoglossoideae, although taxa belonging to Boraginoideae possessed higher GS values. ...
Article
Large-scale investigation of genome size (GS) in the family Boraginaceae remains lacking. We present the first insight into the GS variation and evolution in this family by analysing all native and naturalized Boraginaceae taxa in the Czech Republic. In total, 38 taxa consisting of 274 individuals were analysed using flow cytometry. We found 60-fold overall variation in GS, with the lowest value in Myosotis sylvatica (2C = 0.56 pg) and the highest value in Lycopsis arvensis (2C = 33.63 pg). Most of the analysed species possessed very small or small genomes (94.74 % of all analysed taxa). Monoploid GS varied 35-fold. We also focused on correlates of GS with several factors, such as length of life, residence status, selected environmental factors and phylogeny. Short-lived plants had a significantly smaller GS than that of the perennial ones, and plants of ruderal habitats possessed a smaller GS than that of plants from natural habitats. Moreover, members of the subfamily Boraginoideae had a significantly higher GS than the members of the subfamily Cynoglossoideae. The GS was not correlated with the occurrence of polyploidy; therefore, the GS variation in Boraginaceae is presumably driven by the proliferation of transposable elements.
... The family Boraginaceae Juss. is known for its considerable chromosome variation, which is a consequence of various cytological processes, such as chromosome fusion or fragmentation, polyploidy or aneuploidy (Britton 1951. These processes seem to be common in the family and play a crucial role in the evolution of many genera, such as Borago L. , Cerinthe L. (Selvi et al. 2009), Myosotis L. (Štěpánková 2001, 2006), Nonea Medik. (Selvi et al. 2002, Bigazzi & Selvi 2003, Onosma L. (Mártonfi et al. 2008), Omphalodes Mill. ...
Article
Full-text available
The Symphytum tuberosum complex is a highly polyploid and taxonomically intriguing group. At least eight ploidy levels were recorded previously within this complex. Based on flow cytometric screening of 271 central-European populations, two dominant ploidy levels were revealed: tetraploid (2n = 4x = 32) and widespread dodecaploid (2n = 12x = 96). The tetraploid cytotype is mainly distributed along the southern and south-western margins of the West Carpathians where they abut the Pannonian basin, and found only in Slovakia, the Czech Republic (south-eastern Moravia) and Hungary; our findings represent the first records of this ploidy level for the latter two countries. In contrast, the dodecaploid cytotype occurs throughout the whole area studied. In addition to their geographic distributions, differences between the cytotypes in morphology and habitat requirements were detected using a multivariate morphometric analysis and analysis of a phytosociological database, respectively. Based on this information and taking certain overlaps in morphological traits and habitat requirements into account, we propose treating the dominant cytotypes as subspecies: S. tuberosum subsp. tuberosum (dodecaploids) and S. tuberosum subsp. angustifolium (tetraploids). In some populations, aneuploids and several minority ploidy levels were also detected, including DNA-hexaploids (only within populations of tetraploids), DNAdecaploids and DNA-tetradecaploids (both only within populations of dodecaploids).
... The number of taxa cited for the Italian territory has gradually increased from 14 genera and 60 species, described in the first National Flora (Bertoloni 1835(Bertoloni -1836, to 29 genera and 115 species reported in the most recent Checklist (Conti et al. 2005; Table I). Increased knowledge also resulted from several works devoted to some critical groups, such as Pulmonaria (Puppi & Cristofolini 1996), Anchusa (Selvi & Bigazzi 1998), Symphytum (Bottega & Garbari 2003), Onosma (Peruzzi & Passalacqua 2008), Cerinthe (Selvi et al. 2009) and Cynoglossum (Selvi & Sutory 2012). However, several other groups still exist whose members require more detailed systematic analyses, starting from nomenclatural, taxonomic and distributional aspects. ...
Article
Full-text available
A synopsis of Boraginaceae subfam. Hydrophylloideae and Heliotropioideae in Italy is given, as the first contribution to the treatment of the family for the project Flora Critica d'Italia. The work is mainly based on the critical study of herbarium material conserved in most of the Italian collections. All the relevant floristic reports were examined and all the specific names of taxa reported from the National territory were typified, as well as a number of infraspecific ones. Seven species are recognized, only three of which in genus Heliotropium are autochtonous, with one represented by an endemic subspecies. A complete floristic treatment is provided, including analytical keys, distribution maps and lists of selected “specimina visa”, as well as original iconographies to help with species identification.
Article
We analysed 87 species of Onosma (Boraginaceae) from throughout its distribution range to investigate its evolutionary history. Using nrDNA ITS and two plastid (rpl32-trnL(UAG) and trnH–psbA) markers, we reconstructed phylogenetic relationships within Onosma by conducting maximum parsimony, maximum likelihood, Bayesian, and BEAST analyses. The analyses revealed that Onosma as currently circumscribed is not monophyletic. However, the vast majority of Onosma species appear to belong to a single clade, the so-called Onosma s.s. Outside of this core clade is a clade containing O. rostellata, a subclade of Sino-Indian species and Maharanga emodii. Podonosma orientalis (as O. orientalis) appear only distantly related to Onosma but is more closely related to Alkanna, as also suggested in previous molecular studies. The Onosma s.s. clade includes all representatives of O. sect. Onosma, and encompasses three subsections, i.e. Onosma, Haplotricha and Heterotricha, corresponding to asterotrichous, haplotrichous and heterotrichous groups, respectively, but none of these subsections was retrieved as monophyletic. We observed significant incongruence between nuclear and chloroplast phylogenies regarding the phylogenetic status of the heterotrichous group. A dozen of the Iranian haplotrichous species formed a lineage which may not hybridize with asterotrichous species. Divergence time estimates suggested that the early radiation of Onosma s.l. took place at the Oligocene-Miocene boundary and the diversification within Onosma s.s. occurred during middle to late Miocene and Pliocene.
Article
Full-text available
Boraginaceae s.str. is a subcosmopolitan family of 1600 to 1700 species in around 90 genera, and recent phylogenetic studies indicate that the infrafamilial classification as currently used is highly obsolete. The present study addresses the relationships of the major clades in Boraginaceae s.str. with an emphasis on monophyly of, and relationships between previously recognized clades and the position of various unplaced genera such as Afrotysonia, Anoplocaryum, Brachybotrys, Chionocharis, Craniospermum, Thyrocarpus, and Trigonocaryum using three plastid markers and a taxon sampling with four outgroup and 170 ingroup species from 73 genera. The phylogeny shows high statistical support for most nodes on the backbone and within individual clades. Echiochileae are confirmed as sister to the remainder of Boraginaceae s.str., which, in turn, fall into two well-supported clades, the Boragineae + Lithospermeae and the Cynoglosseae s.l. The latter is highly resolved and includes the Lasiocaryum-clade (Chionocharis, Lasiocaryum, Microcaryum) and the Trichodesmeae (Caccinia, Trichodesma) as sister to the remainder of the group. Rochelieae (formerly the Eritrichieae s.str., also including Eritrichium, Hackelia, and Lappula) form a poorly supported polytomy together with the Mertensia-clade (also including Anoplocaryum, Asperugo, and Memoremea) and the Omphalodes-clade. The enigmatic genus Craniospermum (Craniospermeae) is sister to an expanded Myosotideae (also including Brachybotrys, Decalepidanthus, Trigonocaryum, and Trigonotis) and these two clades are in turn sister to the Core-Cynoglosseae, in which Afrotysonia glochidiata and Thyrocarpus sampsonii are included. Core-Cynoglosseae again fall into two pairs of well-supported subclades. The majority of generic placements are now resolved satisfactorily and the remaining phylogenetic questions can be clearly delimited. Based on the extensive phylogenetic data now available we propose a new infrafamilial classification into three subfamilies and 11 tribes, representing a consensus among the participating authors, according to which major clades are renamed.
Chapter
Annual, biennial or perennial, rarely monocarpic herbs, subshrubs and rarely erect or lianescent shrubs, not aromatic; primary root usually persistent, often developed as strong taproot, rarely primary and secondary roots thickened and spindle-shaped or primary root splitting from above and forming root pleiocorm; stems terete, rarely angular, then often winged from decurrent leaf bases, erect, rarely ascending, sprawling, decumbent or procumbent, prostrate, sometimes forming rhizomes or stolon tubers; indumentum usually present on whole plant, often strongly developed, scabrid to hispid or strigose, sometimes uncinate, very rarely with stiffly deflexed trichomes for climbing or lanate or consisting of stellate trichome complexes or largely absent (trichomes reduced to basal cells), trichomes often scabrid, often mineralized with Si or Ca, often inserted on a multicellular basal tubercle or cystolith cell, rarely unicellular, simple and smooth or uniseriate and gland-tipped. Leaves alternate, rarely opposite, simple, entire, exstipulate, cauline leaves usually sessile to semiamplexicaulous, lamina linear, narrowly ovate to subcircular, usually decurrent on petiole, more rarely base distinctly rounded or cordate and distinctly petiolate, basal leaves usually forming an ephemeral or persistent rosette, sessile or tapering into a petiole. Inflorescences terminal or axillary, frondose, bracteose or ebracteose, paraclades monochasial or dichasial, lax or very dense, usually scorpioid and contracted into boragoids, these paraclades present as simple terminal inflorescence or combined into complex thyrsoids, sometimes with extensive accessory paraclades and metatopia. Inflorescences sometimes congested into terminal “heads” or strongly reduced to axillary or terminal single flowers. Flowers pentamerous, hypogynous, bisexual, often proterandrous; perianth biseriate, sepals united at base or nearly to apex, usually radially symmetrical, sometimes slightly or distinctly unequal, calyx tubular to rotate, membranaceous, usually accrescent in fruit, spreading or closing; petals usually united to form distinct tube, rarely united only basally, corolla actinomorphic, rarely zygomorphic or irregular, rotate, hypocrateriform, infundibular, campanulate or tubular, rarely curved or geniculate, corolla tube internally usually appendaged near base with ring-shaped intrusion or 5 or 10 free scales (basal scales = annulus) and near throat with 5 crescent- or scale-shaped intrusions (= faucal scales, fornices), these closing the tube and/or appressed to anthers and/or protruding to form a ring around opening of tube, erect, recurved or incurved, papillose to pubescent, usually contrastingly coloured (often yellow or white vs. blue or white corolla lobes); corolla lobes reflexed to porrect, narrowly triangular to (usually) subcircular; aestivation usually quincuncial, rarely contorted; stamens epipetalous and antesepalous, rarely unequal or some abortive, filaments free from each other, usually short and anthers included in tube or partially or completely exserted, inserted at the same or different heights in tube, filament lengths and/or point of insertion differing between morphs in taxa with stigma height polymorphism or heterostyly; anthers usually free from each other, dorsifixed, dithecous, tetrasporangiate, opening by longitudinal slits, sometimes anthers coherent, connective sometimes appendaged; gynoecium 2-carpellate, syncarpous, usually with basal nectary disc; ovary bilocular, but secondarily subdivided into four chambers by false septa, 4-lobed in flower; style overtopped by the nutlets (gynobasic), usually simple, rarely 2–4 stylodia, included or exserted, stigmas capitate to oblong; ovules anatropous-epitropous, usually basal or sub-basal, erect to pendulous, unitegmic, tenuinucellate. Fruit a dry schizocarp separating into four 1-seeded, rarely two 2-seeded, indehiscent mericarpids, sometimes fewer by abortion or fusion; mericarpids (“nutlets”, “eremocarpids”) usually equal, rarely heteromorphic (in individual fruit, in fruits on different parts of the plant or in different morphs of a population), nutlet shape and size highly variable, ovoid to subcircular, tetrahedral, lenticellate, bowl-shaped or disc-shaped, often with differentiated, lobed/glochidiate or membranaceous (“winged”) margin, surface smooth and shiny to papillose and dull, pubescent, verrucose or with complex glochidia, sometimes with base strongly modified as ring and/or with basal elaiosome. Mature seeds with scanty endosperm, cotyledons fleshy, embryo usually straight.
Article
The phylogenetic relationships of Mertensia (Boraginaceae), which comprises approximately 45 species in both Asia and North America, have been uncertain, and taxonomists have placed the genus in various tribes of subfamily Boraginoideae, with the most recent placements in Trigonotideae and Cynoglosseae. Our study applies molecular phylogenetic methods to test the monophyly and relationships of Mertensia. We used DNA sequence data from the nuclear ribosomal nrlTS region and four cpDNA regions (matK, ndhF, rbcL, trnL-trnF) to examine the placement of Mertensia among a sampling of accessions from approximately 70% of the genera of Boraginaceae s. 1. Phylogeny reconstructions using maximum parsimony, maximum likelihood, and Bayesian inference were largely congruent with previous molecular phylogenetic analyses of Boraginaceae that had applied far fewer taxa. We recovered five deep clades that correspond to Boraginaceae subfamilies Boraginoideae, Cordioideae, Heliotropioideae, Hydrophylloideae, and Ehretioideae (including Lennoa and Pholisma). In subfamily Boraginoideae, we recovered clades that correspond to the tribes Echiochilieae, Lithospermeae, Cynoglosseae, and Boragineae, although several tribes previously circumscribed on the basis of morphological data were not recovered as monophyletic in our results. Based on the sister relationship between the genus Codon and subfamily Boraginoideae found in our phylogeny reconstructions, we propose Codoneae as a new tribe of Boraginoideae. We recovered strong support for the monophyly of Mertensia and the placement of the monotypic genus Asperugo as its sister. Mertensia and Asperugo were strongly supported as members of Cynoglosseae.
Article
Utilizing 10 cpDNA regions and thorough taxon sampling, a phylogeny is reconstructed for Lithospertnum and related members of both Lithospermeae, the tribe to which it is assigned, and Boraginaceae. Lithospermum is supported as monophyletic, and the genus is hypothesized to have originated in the Old World, after which there was one colonization of the New World. The heterostylous breeding system is inferred to have originated within Lithospermum either seven times or six times with one loss, and with other independent origins within Lithospermeae. The stability of the 10-region matrix is investigated, as are the number and combination of regions necessary to accurately reconstruct phylogenies. The combination of concatenated regions is important, and the following regions are recommended for future phylogenetic studies of genera of Boraginaceae: the rpl16 intron, matK, psbA-trnH, trnL-rpl32, and trnQ-rps16. The use of these recommended regions is conservative, and it contrasts with most intrageneric studies of Boraginaceae, which often are based on the trnL-trnF spacer and nuclear ribosomal ITS.
Article
The recently-developed statistical method known as the "bootstrap" can be used to place confidence intervals on phylogenies. It involves resampling points from one's own data, with replacement, to create a series of bootstrap samples of the same size as the original data. Each of these is analyzed, and the variation among the resulting estimates taken to indicate the size of the error involved in making estimates from the original data. In the case of phylogenies, it is argued that the proper method of resampling is to keep all of the original species while sampling characters with replacement, under the assumption that the characters have been independently drawn by the systematist and have evolved independently. Majority-rule consensus trees can be used to construct a phylogeny showing all of the inferred monophyletic groups that occurred in a majority of the bootstrap samples. If a group shows up 95% of the time or more, the evidence for it is taken to be statistically significant. Existing computer programs can be used to analyze different bootstrap samples by using weights on the characters, the weight of a character being how many times it was drawn in bootstrap sampling. When all characters are perfectly compatible, as envisioned by Hennig, bootstrap sampling becomes unnecessary; the bootstrap method would show significant evidence for a group if it is defined by three or more characters.