See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/281488281

Phylogenetic relationships among Dorema,

Ferula and Leutea (Apiaceae: Scandiceae:
Ferulinae) inferred from nrDNA ITS...

Article in Taxon · August 2015

DOI: 10.12705/644.8

CITATIONS READS

5 159

6 authors, including:

Marcin Piwczyński Radosław Puchałka

Nicolaus Copernicus University Nicolaus Copernicus University
37 PUBLICATIONS 168 CITATIONS 26 PUBLICATIONS 16 CITATIONS

SEE PROFILE SEE PROFILE

Alexei A. Oskolski Krzysztof Spalik

Russian Academy of Sciences University of Warsaw
64 PUBLICATIONS 348 CITATIONS 51 PUBLICATIONS 1,245 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Next-generation systematics for challenging taxa: unravelling phylogeny and species delimitation in the
subfamily Miltogramminae (Diptera: Sarcophagidae) View project

Linking micro- and macroevolution: do genetic constraints predict phenotypic divergence? View project

All content following this page was uploaded by Radosław Puchałka on 31 December 2015.

The user has requested enhancement of the downloaded file.

Panahi & al. • Phylogenetic relationships in Ferulinae (Apiaceae) TAXON 64 (4) • August 2015: 770–783

Phylogenetic relationships among Dorema, Ferula and Leutea

(Apiaceae: Scandiceae: Ferulinae) inferred from nrDNA ITS and
cpDNA noncoding sequences
Mehrnoush Panahi,1 Łukasz Banasiak,1 Marcin Piwczyński,2 Radosław Puchałka,3 Alexei A. Oskolski4,5 & ​
Krzysztof Spalik1
1 Department of Molecular Phylogenetics and Evolution, Faculty of Biology, University of Warsaw, Aleje Ujazdowskie 4,
00-478 Warsaw, Poland
2 Chair of Ecology and Biogeography, Nicolaus Copernicus University in Toruń, Lwowska 1, 87-100 Toruń, Poland
3 Herbarium TRN, Chair of the Geobotany and Landscape Planning, Nicolaus Copernicus University in Toruń, Lwowska 1,
87-100 Toruń, Poland
4 Botanical Museum, Komarov Botanical Institute, Prof. Popov str. 2, 197376 St. Petersburg, Russia
5 Department of Botany and Plant Biotechnology, University of Johannesburg, P.O. Box 524, Auckland Park 2006, Johannesburg,
South Africa
Author for correspondence: Krzysztof Spalik, spalik@biol.uw.edu.pl
ORCID: KS, http://orcid.org/0000-0001-9603-6793

DOI  http://dx.doi.org/10.12705/644.8

Abstract  Dorema, Ferula and Leutea are genera of Apiaceae tribe Scandiceae, subtribe Ferulinae. Their placement in Scandiceae
was established based only on nuclear ribosomal DNA internal transcribed spacer (nrDNA ITS) sequence variation. Previous
molecular studies suggested that Dorema and Leutea are nested within Ferula. To confirm relationships and eventual nomen-
clatural changes, we increased taxonomic sampling and examined nrDNA ITS and three noncoding cpDNA markers: rpoB-trnC
intergenic spacer and rps16 and rpoC1 introns. Ninety-one species were included representing 68 species of Ferula, 6 species
of Dorema, 9 species of Leutea and 8 outgroup taxa. Nuclear and plastid DNA data were analysed separately and combined
using maximum likelihood and Bayesian approaches; based on the set of Bayesian trees several topological hypotheses on
relationships among these genera were tested. As nrDNA and cpDNA gene trees differed somewhat, the degree of topological
conflict was assessed using internode certainty (IC) and tree certainty (TC) measures calculated from ML bootstrap analyses.
In all analyses, Leutea was monophyletic. In nrDNA trees it was nested within Ferula (with low support), while in cpDNA it
was sister to the Ferula + Dorema clade. In all analyses, Dorema species were placed in one subclade of Ferula and intermin-
gled with its members. IC support for the sister position of Leutea to remaining Ferulinae and the TC support were higher in
combined analyses than in the analyses of cpDNA data alone, indicating that the gene trees are not significantly different with
respect to this internode. Our results do not contradict the taxonomic position of Leutea as a separate genus but suggest that
Dorema should be subsumed within Ferula.

Keywords  molecular phylogeny; nrDNA ITS; rpoB-trnC intergenic spacer; rpoC1 intron; rps16 intron; Umbelliferae

Supplementary Material  The Electronic Supplement (Figs. S1–S4) is available in the Supplementary Data section of the
online version of this article at http://www.ingentaconnect.com/content/iapt/tax; the data for this study have been deposited
with TreeBASE, study no. 16639

INTRODUCTION chemical constituents such as volatile aromatic acids and ses-

quiterpene lactones seem to be relatively good phylogenetic
Dorema D.Don, Ferula L. and Leutea Pimenov are three markers (Pimenov & al., 1978; Kurzyna-Młynik & al., 2008)
genera of Apiaceae that constitute Scandiceae Spreng. subtribe and the phylogenetic position of a species may be a good indi-
Ferulinae Engl. (Kurzyna-Młynik & al., 2008). Many species cator of its potential medicinal value. A reliable taxonomy of
of these genera are sources of aromatic resins used in tradi- Ferulinae based on well-supported phylogeny is therefore of
tional medicine including asafoetida, sagapenum, galbanum, considerable practical value.
sumbul and ammoniacum (Korovin, 1959). These resins are Ferula encompasses ca. 170 species placing it among the
also currently being investigated with respect to the presence largest genera of Apiaceae (Pimenov & Leonov, 1993). The
of biologically active compounds and potential therapeutic genus occurs in Eurasia and northern Africa with a centre of
value (e.g., Nazari & Iranshahi, 2011; Kim & al., 2012). Some diversity in Central Asia. Korovin (1947), the first monographer

Received: 18 Nov 2014 | returned for (first) revision: 9 Feb 2015 | (last) revision received: 27 Mar 2015 | accepted: 27 Mar 2015 || publication date(s):
online fast track, n/a; in print and online issues, 28 Aug 2015 || © International Association for Plant Taxonomy (IAPT) 2015

770 Version of Record

TAXON 64 (4) • August 2015: 770–783 Panahi & al. • Phylogenetic relationships in Ferulinae (Apiaceae)

of Ferula, recognized six subgenera and eight sections in the reduce the stochastic error (De Queiroz & Gatesy, 2007). How-
genus; in subsequent studies, this classification was either ever, if the markers under consideration have different evolu-
adopted with reservation (Chamberlain & Rechinger, 1987) tionary histories due to incomplete lineage sorting, hybridi-
or challenged based on fruit morphology and anatomy (Safina zation, lateral gene transfer or other processes, then the tree
& Pimenov, 1983, 1984, 1990). does not properly describe the evolutionary history of species.
Dorema includes 12–16 species (Rechinger, 1987). It resem- In plants, the phenomenon of chloroplast capture is common
bles Ferula and has long been regarded as closely related to it and in extreme cases it may result in a complete failure of spe-
(Pimenov, 1988). Their close relationship was confirmed using cies-barcoding based on plastid markers (Percy & al., 2014). A
immunological comparisons of seed-storage proteins (Shneyer discrepancy between cpDNA and nrDNA resulting from pos-
& al., 1995). In contrast, the first ITS sequence analyses placed sible reticulate evolution has also been reported for Apiaceae
Dorema, represented by a single accession of D. aucheri Boiss., (Lee & Downie, 2006; Spalik & al., 2009; Zhou & al., 2009;
as sister group to Seseli mucronatum (Schrenk) Pimenov Bone & al., 2011; Yi & al., 2015).
& Sdobnina in tribe Selineae distant from Ferula (Valiejo-​ Recent phylogenomic studies demonstrated that a large
Roman & al., 2006). This result was probably due to sample number of gene trees disagree with the species phylogeny
contamination or voucher misidentification because subsequent and there is an array of methods that capture and quantify the
molecular studies confirmed that Dorema is closely related to incongruence between characters or between trees (Salichos
Ferula (Ajani & al., 2008; Kurzyna-Młynik & al., 2008). These & al., 2014 and references therein). There is a general agreement
studies suggested that Dorema should probably be transferred that concatenation of incongruent datasets should be avoided
into Ferula but additional evidence from cpDNA data was because it results in artificial or at least less resolved phyloge-
necessary to make this transfer justified. netic trees. However, the effect of combining incongruent data-
Leutea comprises 8–10 species that are distributed in arid sets may vary for different parts of the tree: it may decrease
montane regions of Iran and Turkmenistan on stony and rocky the resolution in those subclades where the conflict occurs
slopes (Pimenov, 1987). This taxon was first recognized as but it may increase the internal support for other branches. A
Peucedanum sect. Juncea Boiss., then raised to the rank of commonly used measure of internal support is the bootstrap.
genus with suggested affinity to other peucedanoid genera, However, bootstrap analyses provide internal support only for
particularly to Ferula (Pimenov, 1987; Shneyer & al., 1995). the most commonly occurring bipartitions, which may be par-
The nomenclatural type of the genus name, L. petiolaris (DC.) ticularly misleading if topological conflicts occur. As Salichos
Pimenov was originally described in Ferula. Leutea was distin- & al. (2014) pointed out, 51% bootstrap support is very weak
guished from its presumed relatives by its mostly basal leaves when the conflicting topology has 49% support but it is strong
with rigid cylindrical ultimate segments, and by having only when the most common of the alternative bipartitions has 5%
hermaphroditic flowers (Pimenov, 1987). In phylogenetic support. Therefore, some novel measures were proposed that
analyses of nrDNA ITS sequences, Leutea was nested within quantify incongruence among phylogenetic trees based on
Ferula; therefore, Kurzyna-Młynik & al. (2008) formally syn- information theory (Salichos & Rokas, 2013; Salichos & al.,
onymized the former with the latter. 2014). Internode certainty (IC) calculates the degree of cer-
The phylogeny of Ferulinae and particularly the relation- tainty for a given internode by considering the frequency of the
ships among Ferula, Dorema and Leutea have so far been bipartition defined by the internal branch in a given set of trees
investigated using nrDNA ITS sequence variation (Ajani & al., jointly with that of the most prevalent conflicting bipartition in
2008; Kurzyna-Młynik & al., 2008). Despite some obvious the same tree set. The second measure, IC All (ICA), considers
advantages including high copy number and the availability of all conflicting bipartitions. The relative (standardized) tree
universal primers for this marker, it also has several caveats, for certainty (TC) and TC All (TCA) are mean IC and ICA for a
instance, the presence of genomic pseudogenes in various states given tree, respectively. The values for IC and ICA may vary
of decay, and/or incomplete intra- or inter-array homogeniza- between −1 (the examined partition is not present in the set of
tion (Álvarez & Wendel, 2003). For Ferulinae, ITS sequence trees) and +1 (the frequency of the partition is 100%). Values
variation is relatively low and does not provide a well-resolved at or near 0 indicate that one or more conflicting bipartitions
phylogeny for the group: the internal support for many major have almost equal support. These measures, therefore, allow
clades was weak and the relationships among them remained assessing topological conflicts across all internodes of the phy-
ambiguous (Kurzyna-Młynik & al., 2008). Therefore, evidence logenetic tree.
from other molecular markers is necessary to resolve phyloge- In this paper, we explore the phylogenetic relationships
netic relationships within the tribe. However, our preliminary among Dorema, Ferula and Leutea using nrDNA ITS and
analyses of cpDNA markers yielded phylogenetic trees that cpDNA markers and extended taxonomic sampling of these
differed from those obtained from nrDNA ITS data. genera. We consider three plastid markers: rps16 and rpoC1
Phylogenetic estimates may differ due to disparate histo- introns and the rpoB-trnC intergenic spacer. These markers
ries of genes under analysis, due to stochastic error associated have already been used in phylogenetic analyses of umbellifers
with the low phylogenetic signal contained in these markers or (Downie & al., 1996, 2000; Calviño & al., 2010). We examine
due to systematic error resulting from inadequacy of the model congruence of phylogenetic trees inferred by analyses of plastid
used for phylogenetic inference (Leigh & al., 2008). Combining and nuclear DNA sequence data and compare the results against
data from genes sharing the same evolutionary history should current taxonomic treatments of these genera.

Version of Record 771

Panahi & al. • Phylogenetic relationships in Ferulinae (Apiaceae) TAXON 64 (4) • August 2015: 770–783

MATERIALS AND METHODS from long insertions or mononucleotide repeats. For the rps16
intron, we used primers 5exonC and 3exonR following Calviño
Taxon sampling. — Ninety-one species were examined & al. (2006). Because PCR with these primers was unsuccess-
for nrDNA and cpDNA sequence variation (Appendix 1). The ful for some samples, we also designed new external prim-
ingroup comprised 83 taxa including 68 species of Ferula, ers, s16exF and s16exR, and two internal primers, s16inF and
6 species of Dorema and 9 species of Leutea. Close outgroups s16inR (Fig. 1; Table 1). These primers usually worked better
included members of two major lineages of Scandiceae: the with Mg2+ concentration increased from 1.5 mM to 3.0 mM
Glaucosciadium B.L.Burtt & P.H.Davis clade and subtribe and with an annealing temperature of 50°C–52°C. The com-
Daucinae Dumort. (Kurzyna-Młynik & al., 2008) that in our plete rpoC1 intron and parts of the flanking exon regions were
preliminary analyses of the entire tribe constituted sister taxa amplified using external primers C1exF and C1exR adapted
to Ferulinae. The Glaucosciadium clade was represented from Downie & al. (1998); internal primers C1inF and C1inR
by G. cordifolium (Boiss.) B.L.Burtt & P.H. Davis and Mozaf- were designed to aid amplification of difficult samples. PCR
fariania insignis Pimenov & Maassoumi. From among the spe- products were separated using 1% agarose gel with ethidium
cies of Daucinae, we chose some species of Laserpitium L. that bromide and purified using the QIAEX II Agarose Gel Extrac-
form a basal grade in the phylogenetic tree of the subtribe. tion Kit (Qiagen) following the manufacturer’s instructions.
Further outgroups consisted of two species of Pachypleurum The purified PCR products were sequenced using AmpliTaq
Ledeb. that in former analyses were placed in tribe Selineae DNA polymerase (Roche Molecular Systems, Alameda, Cali-
Spreng. and in the Acronema clade (Downie & al., 2010). fornia, U.S.A.) and fluorescent Big Dye terminators (Applied
DNA extraction, amplification and sequencing. — Total Biosystems, Foster City, California, U.S.A.) in both directions
genomic DNA was extracted from ca. 20 mg of dried plant using the PCR primers. DNA sequences were assembled and
material using the DNeasy Plant Mini Kit (Qiagen, Valencia, edited using SeqMan II v.4.0 (Dnastar, Madison, Wisconsin,
California, U.S.A.) following the instructions of the manufac- U.S.A.). All newly obtained sequences have been submitted to
turer. Primers N-nc18S10 and C26A (Wen & Zimmer, 1996) GenBank (Appendix 1).
were used to amplify the ITS region. For some accessions, the Sequence and phylogenetic analyses. — The sequences
ITS1 and ITS2 regions were each amplified separately using were initially aligned using the default pair-wise and multiple
primers 18S-ITS1-F and 5.8S-ITS1-R for ITS1 and ITS-3N and alignment parameters in CLUSTAL X (Larkin & al., 2007)
C26A or ITS4 for ITS2 (Spalik & Downie, 2006). The rpoB- and adjusted manually using Mesquite v.2.73 (Maddison
trnC intergenic spacer region was amplified with primers
rpoB and trnCGCAR from Shaw & al. (2005); for some difficult
samples, e.g. those containing long insertions, this region was Table 1. Primer sequences for cpDNA markers used in this study.
divided into parts and amplified using internal primers RP and Plastid region/
TP designed for this study (the position of cpDNA primers is primer name Primer sequence (5′ to 3′)
shown in Fig. 1, their sequences are given in Table 1). For some rpoB-trnC intergenic spacer
accessions, the external forward primer rpoB was unsuccess- rpoB C(GT)ACAAAA(CT)CC(CT)TC(AG)AATTG
ful and was replaced with the newly designed primer rpD; GCA
trnC R CACCC(AG)GATT(CT)GAACTGGGG
additionally, internal reverse primers rpA and 1R were used
to overcome amplification and sequencing problems resulting RP AATAAAAATAGAAGGGGAAGAC
TP GCATACGCTAAGGATTGTGAA
1R TTCATTTTTCTGGTATTC
A rpD TCTTCCGTCAAGCCCTGAT
rpoB rpoB-trnC intergenic spacer trnC
rpA CCAGATATGGAGTTGGGTTCA
rpoB TP RP trnC GCA R
rps16 intron
rpD rpA 1R
5exonC TTTGAAACGATGTGGTAGA
B 3exonR ACCCACGTTGCGAAGAT
exon 1 rps16 intron exon 2
s16exF TTTGAAACGCTGTGGTAGAA
s16exF/5exonC s16inF s16inR s16exR/3exonR s16exR AACCTACGTTCCGAAGAT
s16inF CCGAAGTAATGTCTAAACCCAC
C
exon 1 rpoC1 intron exon 2 s16inR CCCTCCTGTTTGTCTCATTTA
rpoC1 intron
C1exF C1inF C1inR C1exR
C1exF ACGTCTTCCTAGTTATATTGCG
500 bp C1exR AATAGACACAAAACCATCCAT
C1inF AAGAATTACGGGATTCATCATT
Fig. 1. Maps of the plastid regions used in this study showing the posi-
tions of primers (see Table 1): A, rpoB-trnC intergenic spacer; B, rps16 C1inR GAAGACCAAGGAGGTTTTCC
intron; C, rpoC1 intron. The position of the primers is indicated in Fig. 1.

772 Version of Record

TAXON 64 (4) • August 2015: 770–783 Panahi & al. • Phylogenetic relationships in Ferulinae (Apiaceae)

& Maddison, 2010). Gaps were positioned to minimize nucle- phylogenetic relationships among Dorema, Ferula and Leu-
otide mismatches and treated as missing data in phylogenetic tea based on our preliminary analyses of molecular data and
analyses. present taxonomic treatments. These hypotheses are given in
The topological congruence of the phylogenetic trees Table 2 while respective tree topologies are provided in Fig. S1
obtained with the different markers was examined using a hier- (Electr. Suppl.). Particularly, we tested the hypotheses regard-
archical clustering method based on likelihood-ratio testing ing monophyly of each of these genera and the sister relation-
as implemented in Concaterpillar v.1.7.2 (Leigh & al., 2008). ship between Leutea and the remaining Ferulinae. Constraining
Because this analysis suggested that nrDNA ITS and cpDNA some groups to monophyly also enforces particular relation-
topologies are incongruent, we first analysed both datasets sep- ships among the analysed genera. For instance, topology no. 4
arately using Bayesian inference (BI) and maximum likelihood constrains two groups within Ferulinae to monophyly: the spe-
(ML) with bootstrapping. Second, in the absence of signifi- cies of Leutea and the species of Ferula + Dorema. This implies
cantly supported topological conflict (≥ 0.95 PP), the data were that Leutea is sister to a clade formed by Ferula and Dorema.
combined and reanalysed using different partition schemes and The frequency of these hypotheses was inferred by filtering
the topological conflicts resulting from the concatenation of the posterior tree sample in PAUP* v.4.0a125 (Swofford, 2003)
data were examined. Based on BI analyses, posterior probabil- with a given constraint (Bergsten & al., 2013).
ities for alternative topological hypotheses on the relationship Maximum likelihood analyses were performed with
among Dorema, Ferula and Leutea were calculated. For ML RAxML v.8.0.26. One thousand searches were initiated with
analyses, the degree of topological conflict for particular nodes starting trees obtained through randomized stepwise addition
was assessed using IC and ICA measures while the trees were using maximum parsimony. The tree with the highest likeli-
compared using TC and TCA. hood score was considered as the best representation of the
The best partition scheme for the combined dataset was phylogeny. Bootstrap support was calculated with 1000 rep-
chosen with PartitionFinder v.1.1.1 (Lanfear & al., 2012, 2014) licates and summarized on the best ML tree. Based on the
with RAxML v.8.0.0 (Stamatakis, 2014) assuming linked or results of bootstrap analyses, the values of IC, ICA, TC and
unlinked branch lengths and using Bayesian (BIC) or corrected TCA were assessed.
Akaike (AICc) information criteria to choose the optimal parti- The majority of phylogenetic analyses were performed
tion. For each partition, unlinked models have more estimated using CIPRES Science Gateway (Miller & al., 2010). The
branch length parameters than linked models. Both AICc and data for this study have been deposited with TreeBASE,
BIC penalize extra partitions that add parameters but do not study no. 16639 (http://purl.org/phylo/treebase/phylows/study/
contribute much to likelihood improvement. AIC penalizes TB2:S16639).
the number of parameters less strongly than BIC. Because
the choice of criterion is rather arbitrary, we applied partition
schemes chosen by both AICc and BIC in further phylogenetic RESULTS
analyses.
In BI analyses, we applied the GTR + G + I model because Sequence characteristics. — Among the examined
the Bayesian method is more sensitive to underspecification regions, the ITS region is more variable than any of the three
of the evolutionary model than to overspecification; therefore, cpDNA regions having 42% variable characters (262 positions)
the model should be as complex as possible while still allowing as opposed to 13% (455 positions) in the combined cpDNA
parameters to be identified (Huelsenbeck & Rannala, 2004). markers (Table 3). Additionally, much of the cpDNA sequence
In ML analyses, we used the GTR + G model. GTR(+ G) is the variation included gaps of variable length (1–39 bp with mean
only nucleotide substitution model implemented in CIPRES 4.51 bp) accounting for 9.7% of the aligned sequences while
version of RAxML. The idea behind this is that GTR is the for the ITS data gaps accounted for 3.8% of aligned sequences
most common and general model for DNA analysis and it is
better to efficiently implement and optimize this model instead
of offering a plethora of distinct models which are only special Table 2. Topological hypotheses on relationships among Dorema,
cases of GTR but are programmed in a generic and thus inef- Ferula and Leutea tested in this study. The sign “+” denotes groups
ficient way (Stamatakis, 2014). that were constrained to monophyly.
The Bayesian analyses were run with MrBayes v.3.2.3 Topology no.
(Ronquist & al., 2012) with default priors. Two simultaneous Monophyletic groups 1 2 3 4 5 6 7 8 9 10
analyses were performed, each lasting 40 million generations Ferulinae + + + + + + + + + +
with one cold chain and three heated chains, starting from
Dorema D.Don + + + + + + + +
random trees and sampling trees every 1000 generations. We
checked convergence among runs by examining the average Ferula L. + + +
standard deviation of split frequencies. If the value was below Leutea Pimenov + + + + + + + +
0.01, we assumed convergence. The first 25% of trees were Ferula + Dorema + + +
discarded as burn-in. The analyses were summarized on a Ferula + Leutea + + +
50% majority-rule tree. The results of the Bayesian analyses
Dorema + Leutea +
provided posterior support for ten topological hypotheses on

Version of Record 773

Panahi & al. • Phylogenetic relationships in Ferulinae (Apiaceae) TAXON 64 (4) • August 2015: 770–783

Table 3. Characteristics of the datasets used in this study.

Datasets nrDNA rpoB-trnC spacer rps16 intron rpoC1 intron cpDNA All combined
Sequence length variation 599–610 1190–1263 833–879 1045–1058 3093–3181 3696–3782
Number of aligned positions
total 626 1443 931 1081 3455 4081
constant 364 1216 812 972 3000 3364
autapomorphic 121 141 69 68 278 399
parsimony informative 141 86 50 41 177 318
containing gaps 46 392 125 56 573 619
Percentage of gaps and missing data 3.8 15.9 8.1 2.8 9.7 8.8
Mean indel length 1.8 5.3 4.3 2.3 4.5 4.1
Maximum indel length 7 39 28 8 39 39

only and single-nucleotide indels dominated. In spite of con- and F. olivacea (Diels) H.Wolf ex Hand.-Mazz. The monophyly
siderable sequence variation, 23 ITS sequences were identical of Leutea was supported with posterior probability PP = 0.96,
to other sequences in the alignment while all combined cpDNA while the clade comprising Leutea and the eastern Asiatic clade
sequences were unique. received PP = 0.83 (Fig. 2). An alternative placement of Leu-
Hierarchical clustering of the markers based on like- tea as sister to the remaining species of Ferulinae (topologi-
lihood-ratio testing resulted in concatenation of all three cal hypothesis no. 4) occurred only in 0.05 of posterior trees
cpDNA loci while the concatenation of cpDNA and nrDNA (Table 5). The species of Dorema did not form a monophyletic
ITS was rejected with P = 0.000023. PartitionFinder with dif- group but were placed in a large clade among several species
ferent assumptions on branch lengths and information criteria of Ferula. The topology supporting the monophyly of Dorema
selected three partition schemes for the combined analyses (no. 10, Table 5) did not occur in any of the posterior trees.
(Table 4): (1) the data are not partitioned; (2) the data are parti- Monophyly of Leutea was also strongly supported in the
tioned into nrDNA ITS and cpDNA; (3) the data are partitioned cpDNA tree (PP = 1.0, Fig. 3). Additionally, the topology with
into ITS, rpoC1 intron and two remaining cpDNA markers. Leutea being sister to the remaining Ferulinae occurred in
These partition schemes were applied in successive BI and ML 0.98 trees (no. 4, Table 5). The monophyly of Dorema was not
analyses. Because we were only interested in tree topologies, supported in cpDNA trees (no. 10, Table 5) although its species
we chose the option “linked” on branch lengths. occurred in a single subclade (Fig. 3). Similar relationships
Bayesian analyses. — Bayesian analyses of separate among Dorema, Ferula and Leutea were apparent in all analy-
nuclear and plastid regions resulted in trees among which the ses of combined data (Fig. 4; Table 5). These trees supported the
placement of Dorema, Ferula and Leutea differed considera- monophyly of Leutea (topology no. 9) and its basal position in
bly. In the tree summarizing the analyses of ITS data, a basal Ferulinae (topology no. 4), whereas the monophyly of Dorema
polytomy is apparent within Ferulinae (Fig. 2) with as many as (topology no. 10) was poorly supported (PP = 0.02).
seven lineages. All these lineages comprise species of Ferula
making the genus paraphyletic with respect to Dorema and
Leutea. All examined species of Leutea formed a clade that
was sister to a group of predominantly eastern Asian species of Table 5. Frequency of Bayesian trees supporting the topological
Ferula, F. kingdon-wardii H.Wolff, F. licentiana Hand.-Mazz. hypotheses on relationships among Dorema, Ferula and Leutea (see
Table 2).
Topology no.
Table 4. Data partitions used in the analyses as inferred from Partition- Datasets 1 2 3 4 5 6 7 8 9 10
Finder analyses. nrDNA ITS
Criterion Branch lengths Partition only 0 0 0 0.05 0 0 0 0 0.96 0
(ITS, rpoB-trnC spacer, cpDNA
1 BIC unlinked
rpoC1 intron, rps16 intron) only 0 0 0 0.98 0 0 0 0 1.00 0
(ITS) (rpoB-trnC spacer, combined,
2 BIC linked
rpoC1 intron, rps16 intron) 1 partition 0 0 0 1.00 0.02 0 0.02 0 1.00 0.02
(ITS) (rpoB-trnC spacer, combined,
3 AICc unlinked
rpoC1 intron, rps16 intron) 2 partitions 0 0 0 0.98 0.02 0 0.02 0 1.00 0.02
(ITS) (rpoB-trnC spacer, combined,
4 AICc linked
rps16 intron) (rpoC1 intron) 3 partitions 0 0 0 0.99 0.02 0 0.02 0 1.00 0.02

774 Version of Record

TAXON 64 (4) • August 2015: 770–783 Panahi & al. • Phylogenetic relationships in Ferulinae (Apiaceae)

Ferula ugamica
Ferula tschimganica
Ferula rubroarenosa
Ferula pallida
Ferula ovina
Ferula lapidosa
0.93 Ferula kyzylkumica
Ferula karataviensis
Ferula karatavica
Ferula gypsacea
1.0 Ferula ferganensis
Ferula dshizakensis
Ferula dissecta
Ferula ceratophylla
Ferula leucographa
0.97 Ferula tatarica
Ferula songarica
1.0 Ferula penninervis
Ferula olgae
1.0 Ferula kirialovii
Ferula fedoroviorum
Ferula potaninii
Ferula korshinskyi
1.0 Ferula feruloides
Ferula dubjanskyi
Ferula akitschkensis
Dorema hyrcanum
Dorema glabrum
0.52
Dorema aureum
Dorema aucheri
Dorema ammoniacum
Dorema aitchisonii
1.0 Ferula sibirica
Ferula karelinii
Ferula violacea
Ferula teterrima
Ferula tadshikorum Fig. 2. Bayesian 50% majority-
Ferula samarkandica rule consensus tree of 83 repre-
Ferula nevskii
0.96 Ferula mogoltavica sentatives of Ferulinae inferred
Ferula kelifi from analyses of nrDNA ITS
Ferula foetidissima sequence data. For simplicity,
Ferula foetida
Ferula eugenii outgroup taxa are omitted. Mem-
0.94 Ferula botschantzevii bers of Dorema and Leutea are
Ferula schtschurowskiana indicated in boldface. Posterior
1.0 Ferula kokanica
Ferula kelleri probability values are given along
Ferula tuberifera branches. Branch lengths were
1.0 Ferula litwinowiana estimated using the GTR + G + I
0.93 Ferula lehmannii
Ferula szowitsiana substitution model (see scale bar).
1.0 Ferula karakalensis
0.87 Ferula undulata
1.0 Ferula gummosa
Ferula badrakema
0.95 Ferula nuratavica
Ferula moschata
Ferula diversivittata
Leutea rechingeri
0.95 Leutea petiolaris
Leutea glaucopruinosa
Leutea polyscias
1.0 Leutea nematoloba
Leutea gracillima
0.96 Leutea cupularis
1.0
Leutea avicennae
0.83 Leutea elbursensis
0.96 Ferula olivacea
1.0 Ferula kingdon-wardii
Ferula licentiana
Ferula tingitana
1.0
Ferula linkii
1.0
Ferula communis
1.0
Ferula loscosii
Ferula glauca
Ferula varia
1.0 Ferula oopoda
Ferula clematidifolia
0.83 Ferula equisetacea
0.91 Ferula decurrens
Ferula koso-poljanskyi
Ferula renardii
0.1

Version of Record 775

Panahi & al. • Phylogenetic relationships in Ferulinae (Apiaceae) TAXON 64 (4) • August 2015: 770–783

Fig. 3. Bayesian 50% majority- Dorema hyrcanum

rule consensus tree of 83 repre- 0.98 Dorema glabrum
Dorema aucheri
sentatives of Ferulinae inferred Dorema aitchisonii
from analyses of combined Dorema aureum
cpDNA data (rpoB-trnC inter- Dorema ammoniacum
Ferula tatarica
genic spacer, rps16 intron, rpoC1 Ferula lapidosa
intron). See Fig 2. for details. 0.90
Ferula kyzylkumica
Ferula kirialovii
Ferula feruloides
Ferula ferganensis
Ferula dubjanskyi
Ferula dshizakensis
Ferula teterrima
Ferula undulata
0.99 Ferula ovina
1.0
Ferula ceratophylla
Ferula gummosa
0.94
Ferula badrakema
0.85 Ferula akitschkensis
1.0 Ferula litwinowiana
0.68 Ferula lehmannii
Ferula songarica
1.0 Ferula potaninii
Ferula szowitsiana
Ferula sibirica
Ferula tschimganica
Ferula pallida
0.97
Ferula karataviensis
Ferula karatavica
1.0 Ferula gypsacea
Ferula rubroarenosa
0.98 Ferula leucographa
Ferula ugamica
0.99
Ferula dissecta
1.0 Ferula penninervis
Ferula olgae
1.0 Ferula tadshikorum
0.92 Ferula kelifi
Ferula violacea
Ferula foetida
0.98 Ferula eugenii
1.0
Ferula botschantzevii
1.0 Ferula clematidifolia
Ferula mogoltavica
Ferula karakalensis
Ferula kokanica
0.98
Ferula kelleri
0.99 Ferula foetidissima
0.98 0.98 Ferula samarkandica
0.52
Ferula nevskii
Ferula nuratavica
0.58
Ferula tuberifera
Ferula varia
Ferula korshinskyi
Ferula fedoroviorum
Ferula oopoda
Ferula karelinii
Ferula renardii
Ferula schtschurowskiana
Ferula moschata
0.95 0.94 Ferula tingitana
0.97 Ferula linkii
0.63
Ferula loscosii
Ferula communis
Ferula glauca
0.98 0.68 0.62 Ferula equisetacea
Ferula decurrens
Ferula koso-poljanskyi
Ferula diversivittata
1.0 Ferula olivacea
1.0 Ferula licentiana
Ferula kingdon-wardii
1.0
1.0 Leutea polyscias
1.0 Leutea petiolaris
Leutea gracillima
1.0 Leutea nematoloba
0.97 1.0 Leutea elbursensis
Leutea glaucopruinosa
1.0 Leutea rechingeri
Leutea avicennae
Leutea cupularis
0.1

776 Version of Record

TAXON 64 (4) • August 2015: 770–783 Panahi & al. • Phylogenetic relationships in Ferulinae (Apiaceae)

Ferula tschimganica
Ferula pallida
0.96
Ferula karataviensis
Ferula karatavica
1.0 Ferula gypsacea
Ferula rubroarenosa
1.0 Ferula leucographa
0.73
Ferula ugamica
Ferula dissecta
1.0 Ferula ovina
Ferula ceratophylla
0.99 Ferula lapidosa
Ferula kyzylkumica
0.74 Ferula ferganensis
Ferula dshizakensis
Ferula dubjanskyi
Ferula tatarica
Ferula kirialovii
0.73 Ferula feruloides
Dorema hyrcanum
0.97 Dorema glabrum
Dorema aucheri
Dorema aitchisonii
0.93 Dorema aureum
Dorema ammoniacum
0.51 Ferula potaninii
Ferula akitschkensis
Ferula songarica
1.0
1.0 Ferula penninervis
0.93 Ferula olgae
0.67 Ferula korshinskyi
Ferula fedoroviorum
1.0 Ferula sibirica
Ferula karelinii
1.0 Ferula tadshikorum
0.95 Ferula kelifi
Ferula violacea
1.0 Ferula foetida
Ferula eugenii
0.98
Ferula botschantzevii
0.69 Ferula mogoltavica
0.98 Ferula kokanica
0.98 Ferula kelleri
0.70 1.0 Ferula foetidissima
Ferula samarkandica
0.64 Ferula nevskii
1.0 Ferula teterrima Fig. 4. Bayesian 50% majority-
Ferula schtschurowskiana rule consensus tree of 83
Ferula tuberifera representatives of Ferulinae
1.0 Ferula litwinowiana
0.91 0.54 Ferula lehmannii inferred from analyses of com-
0.87 Ferula szowitsiana bined nrDNA and cpDNA data
1.0 Ferula karakalensis with two partitions, nrDNA vs.
1.0 Ferula gummosa
1.0 Ferula badrakema cpDNA. See Fig 2. for details.
Ferula undulata
0.94 Ferula nuratavica
Ferula moschata
1.0 Ferula varia
Ferula oopoda
Ferula clematidifolia
Ferula renardii
1.0 0.98 Ferula tingitana
1.0 Ferula linkii
1.0
1.0
Ferula communis
Ferula loscosii
0.90 Ferula glauca
0.98 0.95 Ferula equisetacea
0.96 Ferula decurrens
Ferula koso-poljanskyi
Ferula diversivittata
1.0 Ferula olivacea
1.0 Ferula licentiana
1.0 Ferula kingdon-wardii
1.0 Leutea polyscias
1.0 Leutea petiolaris
Leutea gracillima
1.0 Leutea nematoloba
0.96 1.0 Leutea elbursensis
Leutea glaucopruinosa
1.0 Leutea rechingeri
Leutea avicennae
Leutea cupularis
0.1

Version of Record 777

Panahi & al. • Phylogenetic relationships in Ferulinae (Apiaceae) TAXON 64 (4) • August 2015: 770–783

Maximum likelihood analyses. — The topologies of trees DISCUSSION

resulting from ML searches for all datasets were generally
similar to those obtained from BI analyses (Electr. Suppl.: Figs. Comparison of nrDNA and cpDNA data. — Molecular
S2–S4). All trees supported the monophyly of Leutea but not phylogenetic studies of broadly distributed taxa have to rely
the monophyly of Dorema; additionally, cpDNA and combined on herbarium specimens as a source of material. These spec-
data trees suggested that Leutea is sister to Ferula and Dorema. imens, however, often provide much-degraded DNA samples
Relative tree certainty (TC and TCA) and internode certainty that limit the choice of molecular markers and often preclude
(IC and ICA) values for major branches estimated based on the using single-copy nuclear genes. Nuclear ribosomal DNA inter-
results of bootstrap analysis are presented in Table 6. The IC nal transcribed spacer occurring in plant genomes in thousands
and ICA values of a given internode reflect its specific degree of copies and providing significant variation seems to be an
of incongruence, and the relative TC and TCA values describe obvious choice, although due to its idiosyncratic evolution the
the global degree of incongruence between trees in the set and results have to be treated with caution (Álvarez & Wendel,
their values may vary between −1 and +1. IC considers only the 2003). Several non-coding plastid DNA sequences were sug-
most prevailing conflicting bipartition whereas ICA takes into gested as a viable alternative (Shaw & al., 2005, 2007). Some of
account all conflicting bipartitions above a threshold of 5%. them have been useful for resolving the phylogeny of Apiaceae,
When for a given internode only one conflicting bipartition although at the level of the family rather than genera because
reaches the threshold, IC and ICA values are equal; such a most variation among the sequences comprised short (1–10 bp)
situation occurred for the internodes defining Ferulinae, Leu- deletions or duplications rather than substitutions (Downie
tea and Ferula + Dorema clades in all analyses (Table 6). For & al., 1996, 2001). Non-coding cpDNA regions such as rps16
the nrDNA ITS tree, the IC values supporting the monophyly and rpoC1 introns are a good choice for analysing phylogenetic
of Leutea (0.39) and Ferulinae (0.26) were rather low indicat- relationships among distantly related species of Apiaceae but
ing that the most common alternative bipartitions occurred their small size and abundance of length mutations confounds
frequently in the set of bootstrap trees. TC (0.12) and TCA alignment, precluding their usefulness at low taxonomic levels
(0.17) values were also low indicating considerable topological (Downie & al., 1996). In our analyses, nrDNA ITS was the
conflict among the bootstrap trees. In the trees obtained from most informative among the four regions examined but the
the cpDNA markers, the relative support for the monophyly of combined cpDNA markers contained more variable positions
Leutea was very high (0.97); IC values for Ferulinae (0.80) and and provided much better resolution at the base of the tree than
a clade uniting Ferula and Dorema (0.55) were also relatively the ITS region. However, based on TC and TCA values, the
high demonstrating that alternative topologies received little phylogenetic utility of these regions was comparable. More-
support from bootstrap analyses. However, relative TC and over, even the analysis of combined data did not provide sat-
TCA values for the best ML tree were rather low (0.18 and 0.24, isfactory resolution at the more terminal nodes, for instance
respectively) meaning that the bipartitions in this tree were not within the clade comprising Dorema. Particularly, molecular
much more frequent in the set of bootstrap trees than alternative data failed to confirm the monophyly of Dorema, a group of
topologies. The results of the three ML analyses of combined species distinguished by a unique and presumably apomorphic
data did not differ with respect to the internal support for the inflorescence architecture.
nodes and the trees. Both Ferulinae and Leutea were highly Combining datasets. — Our analyses of nrDNA and
supported as monophyletic groups as there was no topological cpDNA data using the likelihood approach (Leigh & al., 2008)
conflict in respective internodes in the set of bootstrap trees and comparisons of resulting phylogenetic trees indicated top-
(IC = 1.0). The IC value for the internode representing the clade ological differences between datasets. Incongruence between
of Ferula + Dorema was also high (0.76–0.79), therefore sup- cpDNA and nrDNA ITS data has been reported in several other
porting the sister position of Leutea in relation to this group. studies of Apiaceae (e.g., Lee & Downie, 2006; Spalik & al.,
TC and TCA values, albeit rather low, were also considerably 2009; Zhou & al., 2009; Bone & al., 2011; Yi & al., 2015). Incon-
higher than those for separate nrDNA and cpDNA analyses gruence among molecular markers may result from various
indicating that combining nrDNA and cpDNA data resulted processes including hybridisation or incomplete lineage sorting
in a decrease of topological conflict. or, alternatively, be an analytical artefact given general lack

Table 6. Internode certainty and relative tree certainty in the maximum likelihood analyses of Ferulinae.
Datasets nrDNA cpDNA comb. 1 comb. 2 comb. 3
Internode certainty (IC, ICA)
Leutea 0.39 0.97 1.0 1.0 1.0
Ferula + Dorema – 0.55 0.76 0.79 0.79
Ferulinae 0.26 0.80 1.0 1.0 1.0
Tree certainty TC/TCA 0.12/0.17 0.18/0.24 0.37/0.38 0.38/0.39 0.37/0.38
Because the values of IC and ICA for particular branches were identical, only one value is given. Numbers for com-
bined data analyses refer to partition schemes described in the text.

778 Version of Record

TAXON 64 (4) • August 2015: 770–783 Panahi & al. • Phylogenetic relationships in Ferulinae (Apiaceae)

of phylogenetic signal. Therefore, we decided to investigate did not discuss the limits between Dorema and Ferula. Gener-
the impact of combining the data using novel measures that ally, the species of Dorema are long-lived monocarpic plants
consider phylogenetic incongruence (Salichos & Rokas, 2013; that have simple umbels arranged in racemose inflorescences.
Salichos & al., 2014). Compared with other tests, these meas- These umbels are most likely derived from umbellules of a
ures can easily be applied to the study of a single internode or compound umbel. Flower and fruit pedicels are short and of
the whole tree and are not dependent on a particular phyloge- equal length; therefore, the umbels are globular in shape. This
netic optimality criterion. For Ferulinae, combining apparently distinct and unique morphology seems to separate Dorema
incongruent data from two genomes actually increased support from Ferula. The species are very similar morphologically
for the monophyly of Leutea and its sister position with respect and differ in minor characteristics including plant indumen-
to the remaining Ferulinae. Overall, tree certainty values also tum, number of flowers per umbellule, colour of petals, fruit
increased suggesting that much, if not all, of the presumed size and the width of lateral ribs (Pimenov, 1988). The results
incongruence between plastid and nuclear markers may be of our molecular analyses confirm the close relationships
attributed to low phylogenetic signal. among included species and consistently show that Dorema
The taxonomic position of Leutea. — Leutea was described is nested within Ferula. However, the phylogenetic resolution
by Pimenov (1987), who raised former Peucedanum sect. Juncea within this subclade of Ferula was too low to confirm the
to the rank of genus. Maintaining its position within Peuce- monophyly of the group. Moreover, some species of Ferula
daneae, he suggested that the genus is closer to Ferula than to that form a clade with Dorema also show certain peculiarities
Peucedanum. The affinity of the species of Leutea to Ferula in inflorescence structure suggesting a possible evolutionary
had been previously noticed because the nomenclatural type pathway from compound umbels to racemose inflorescence
of the genus, L. petiolaris, was described as Ferula petiolaris with simple umbels. Based on herbarium material from B, E,
DC. A distinctive feature of Leutea among the members of LE and W, we examined the morphology of species of Ferula
former Peucedaneae is the presence of mostly basal leaves with that are placed with Dorema in the same clade in order to com-
rigid cylindrical ultimate segments; from Ferula, it differs in pare the results of our molecular analysis with the traditional
having only hermaphroditic flowers (Pimenov, 1987). However, taxonomic treatment of these genera. Three species that form
leaf and flower characters are variable within Apiaceae and are a sister group to the Dorema inclusive clade—Ferula songa­
therefore of limited value for assessing phylogenetic relation- rica Pall. ex Schult., F. akitschkensis B.Fedtsch. ex Koso-Pol.
ships (Spalik & Downie, 2001). In the case of Leutea, morpho- and F. potaninii Korovin ex Pavlov (Fig. 4)—have typical com-
logical characters well define this small and relatively uniform pound umbels similar to F. kirialovii Pimenov and F. tatarica
phylogenetic lineage as opposed to the large and morphologi- Fisch. ex Spreng. that are part of a polytomy with the species of
cally variable Ferula, but at the family level these morpholog- Dorema. In contrast, the umbels of successive orders in F. feru-
ical characters are highly homoplastic. Therefore, they provide loides (Steud.) Korovin develop through proliferation, i.e., grow
little support for Leutea as a separate genus. Given the results from the umbels of previous orders, forming a secondarily
of nrDNA ITS analyses and poor diagnostic characters of the monoaxial inflorescence with umbellules arranged in whorls.
genus, Kurzyna-Młynik & al. (2008) formally included Leutea A similar arrangement occurs in F. dubjanskyi Korovin ex
into Ferula. Close relationships among genera of Ferulinae has Pavlov in which these whorls may be reduced to 1–3 umbel-
also been confirmed using immunochemical data: represent- lules resulting in an inflorescence resembling that of Dorema.
atives of Dorema and Ferula were serologically identical and The members of the sister clade to F. dubjanskyi have typical
distinct from Peucedanum s.str. and its relatives; two species compound umbels. One may hypothesize that upon extended
of Peucedanum sect. Juncea (i.e., prospective Leutea) grouped sampling from more variable genetic markers, F. feruloides
with the Ferula-Dorema cluster but remained distinct (Shneyer and F. dubjanskyi may form the sister group or a basal grade
& al., 2003; Shneyer, 2010). The results of the immunochemical to the species of Dorema confirming its inflorescence architec-
data agree with our phylogenetic analyses of combined data ture as a synapomorphy. Nevertheless, the position of Dorema
supporting the sister relationship of Leutea with respect to the as a separate genus seems to be untenable because it makes
remaining Ferulinae. Therefore, the inclusion of Leutea into Ferula paraphyletic. Retaining Dorema would require dividing
Ferula is optional rather than obligatory as both taxonomic Ferula into several small segregates that would hardly be dis-
treatments are congruent with the estimated phylogeny of Feru­ tinct based on morphological data. Such a treatment would add
linae. However, with many species of Ferula still not sampled, to the confusion in this already difficult genus. We therefore
a definite answer to the phylogenetic relationships of Leutea advocate sinking Dorema into synonymy of Ferula.
cannot be given. When transferring species of Dorema to Ferula, three
Relationships of Dorema. — Similarly to Leutea, Dorema nomina nova have to be established because current specific
is a relatively small genus comprising 12–16 species distributed epithets are not available and we have not found any later tax-
predominantly in the Irano-Turanian region (Rechinger, 1987; onomic synonyms that may be used instead.
Pimenov, 1988). The taxonomic history of the genus was elu-
cidated by Pimenov (1988), who discussed previous regional Ferula downieorum Spalik, M.Panahi, M.Piwczyński, R.​
treatments and provided a phylogeny based on morphological Puchałka, nom. nov. ≡ Dorema aureum Stocks in
characters. He indicated that Dorema differs from other genera Hooker’s J. Bot. Kew Gard. Misc. 4: 149. 1852, non Ferula
of umbellifers in the structure of its inflorescence; however, he aurea Link 1825.

Version of Record 779

Panahi & al. • Phylogenetic relationships in Ferulinae (Apiaceae) TAXON 64 (4) • August 2015: 770–783

The specific epithet of Dorema aureum Stocks is not avail- reviewer for comments on the manuscript. Mehrnoush Panahi was
able in Ferula because of F. aurea Link, a taxon described supported by the Erasmus Mundus Program of the European Union,
from the Canary Islands and at present synonymized grant 2007-1139/001-001 MUN-ECW; the research was supported
with F. linkii Webb (Korovin, 1947). The new name is dedicated MNiSW grant N N303 069335 to K. Spalik and NCN grant N304
to Stephen R. Downie and Deborah S. Katz-Downie for their 306740 to M. Piwczyński.
pioneering work on the molecular phylogeny of umbellifers.

Ferula glabrifolia M.Panahi, M.Piwczyński, R.Puchałka &​ LITERATURE CITED

Spalik, nom. nov. ≡ Dorema glabrum Fisch. & C.A.Mey. Ajani, Y., Ajani, A., Cordes, J.M., Watson, M.F. & Downie, S.R.
in Linnaea 10(Lit.): 88. 1836 non Ferula glabra Rech.f. 2008. Phylogenetic analysis of nrDNA ITS sequence reveals rela-
& Riedl 1963. tionships within five groups of Iranian Apiaceae subfamily Api-
Dorema glabrum Fisch. & C.A.Mey. cannot retain its oideae. Taxon 57: 383–401.
epithet in Ferula because of F. glabra Rech.f. & Riedl, an Álvarez, I. & Wendel, J.F. 2003. Ribosomal ITS sequences and plant
phylogenetic inference. Molec. Phylogen. Evol. 29: 417–434.
endemic of Afghanistan (Chamberlain & Rechinger, 1987). The
http://dx.doi.org/10.1016/S1055-7903(03)00208-2
epithet well characterizes this species of Dorema because the Bergsten, J., Nilsson, A.N. & Ronquist, F. 2013. Bayesian tests of
entire plant including terminal leaf lobes is glabrous, similar topology hypotheses with an example from diving beetles. Syst.
to D. badhysi Pimenov but in contrast to other congeners that Biol. 62: 660–673. http://dx.doi.org/10.1093/sysbio/syt029
have pubescent leaf lobes (Pimenov, 1988). In the new epithet Bone, T.S., Downie, S.R., Affolter, J.M. & Spalik, K. 2011. A phylo-
we underline this feature of the species. genetic and biogeographic study of the genus Lilaeopsis (Api­aceae
tribe Oenantheae). Syst. Bot. 36: 789–805.
http://dx.doi.org/10.1600/036364411X583745
Ferula michaelii M.Panahi, M.Piwczyński, R.Puchałka &​​ Calviño, C.I., Tilney, P.M., Van Wyk, B.E. & Downie, S.R. 2006.
Spalik, nom. nov. ≡ Dorema aitchisonii Korovin ex A molecular phylogenetic study of southern African Apiaceae.
Pimenov in Opred. Rast. Srednei Azii 7: 377. 1983, non Amer. J. Bot. 93: 1828–1847. http://dx.doi.org/10.3732/ajb.93.12.1828
Ferula aitchisonii Kosol-Pol. 1921. Calviño, C.I., Martínez, S.G. & Downie, S.R. 2010. Unraveling the
Dorema aitchi­sonii Korovin ex Pimenov cannot retain its taxonomic complexity of Eryngium L. (Apiaceae, Saniculoideae):
Phylogenetic analysis of 11 non-coding cpDNA loci corroborates
epithet because of Ferula aitchisonii Koso-Pol., recognized as rapid radiations. Pl. Diversity Evol. 128: 137–149.
a good species by Korovin (1951) but at present synonymized http://dx.doi.org/10.1127/1869-6155/2010/0128-0006
with F. karatavica Regel & Schmalh. (Czerepanov, 1995). With Chamberlain, D.F. & Rechinger, K.H. 1987. Ferula L. Pp. 387–426
the new epithet we honour Professor Michael G. Pimenov, an in: Hedge, I.C., Lamond, J.M. & Rechinger, K.H. (eds.), Flora
outstanding botanist working on umbellifers. Iranica, vol. 162, Umbelliferae. Graz: Akademische Druck- und
Verlagsanstalt.
Czerepanov, S.K. 1995. Vascular plants of Russia and adjacent states.
The following combinations are also necessary: Cambridge: Cambridge University Press.
De Queiroz, A. & Gatesy, J. 2007. The supermatrix approach to sys-
Ferula ammoniacum (D.Don) Spalik, M.Panahi, M.Piw- tematics. Trends Ecol. Evol. 22: 34–41.
czyński & R.Puchałka, comb. nov. ≡ Dorema ammonia- http://dx.doi.org/10.1016/j.tree.2006.10.002
cum D.Don in Trans. Linn. Soc. London 16(3): 601. 1833. Downie, S.R., Katz-Downie, D.S. & Cho, K.J. 1996. Phylogenetic
analysis of Apiaceae subfamily Apioideae using nucleotide
sequences from the chloroplast rpoC1 intron. Molec. Phylogen.
Ferula aucheri (Boiss.) M.Piwczyński, Spalik, M.Panahi &​ Evol. 6: 1–18. http://dx.doi.org/10.1006/mpev.1996.0053
R.Puchałka, comb. nov. ≡ Dorema aucheri Boiss. in Ann. Downie, S.R., Ramanath, S., Katz-Downie, D.S. & Llanas, E. 1998.
Sci. Nat., Bot., sér. 3, 1: 329. 1844. Molecular systematics of Apiaceae subfamily Apioideae: Phylo-
genetic analyses of nuclear ribosomal DNA internal transcribed
Ferula hyrcana (Koso-Pol.) R.Puchałka, Spalik, M.Panahi &​ spacer and plastid rpoC1 intron sequences. Amer. J. Bot. 85: 563–
591. http://dx.doi.org/10.2307/2446441
M.Piwczyński, comb. nov. ≡ Dorema hyrcanum Koso-Pol. Downie, S.R., Katz-Downie, D.S. & Watson, M.F. 2000. A phylog-
in Notul. Syst. Herb. Hort. Petrop. 2: 67. 1921. eny of the flowering plant family Apiaceae based on chloroplast
DNA rpl16 and rpoC1 intron sequences: Towards a suprageneric
Six species of Dorema recognized by Pimenov (1988) have classification of subfamily Apioideae. Amer. J. Bot. 87: 273–292.
not yet been included in the molecular analyses. Although http://dx.doi.org/10.2307/2656915
based on morphology the close relationship of these species to Downie, S.R., Plunkett, G.M., Watson, M.F., Spalik, K., Katz-
Downie, D.S., Valiejo-Roman, C.M., Terentieva, E.I., Troitsky,
other former congeners seems to be obvious, we refrain from A.V., Lee, B.Y., Lahham, J. & El-Oqlah, A. 2001. Tribes and
their formal transfer to Ferula until their placement in the genus clades within Apiaceae subfamily Apioideae: The contribution of
is confirmed with molecular data. molecular data. Edinburgh J. Bot. 58: 301–330.
http://dx.doi.org/10.1017/S0960428601000658
Downie, S.R., Spalik, K., Katz-Downie, D.S. & Reduron, J.P. 2010.
Major clades within Apiaceae subfamily Apioideae as inferred
ACKNOWLEDGMENTS by phylogenetic analysis of nrDNA ITS sequences. Pl. Diversity
Evol. 128: 111–136.
We thank the curators of the herbaria listed in Appendix 1 for http://dx.doi.org/10.1127/1869-6155/2010/0128-0005
providing plant material and Stephen R. Downie and an anonymous Huelsenbeck, J.P. & Rannala, B. 2004. Frequentist properties of

780 Version of Record

TAXON 64 (4) • August 2015: 770–783 Panahi & al. • Phylogenetic relationships in Ferulinae (Apiaceae)

Bayesian posterior probabilities of phylogenetic trees under simple pomoschchiyu yerakhicheskogo klaster-analiza. Vopr. Kibernetiki
and complex substitution models. Syst. Biol. 53: 904–913. 47: 98–113.
http://dx.doi.org/10.1080/10635150490522629 Rechinger, K.H. 1987. Dorema. Pp. 379–385 in: Hedge, I.C., Lamond,
Kim, S., Kim, K., Jeong, S., Lee, H. & Lu, J. 2012. Abstract A58: J.M. & Rechinger, K.H. (eds.), Flora Iranica, vol. 162, Umbellif-
Galbanic acid isolated from Ferula assafoetida exerts in vivo anti- erae. Graz: Akademische Druck- und Verlagsanstalt.
tumor activity in association with antiangiogenesis and antiprolif- Ronquist, F., Teslenko, M., Van der Mark, P., Ayres, D.L., Darling,
eration. Cancer Res. 72: A58. A., Höhna, S., Larget, B., Liu, L., Suchard, M.A. & Huelsen-
http://dx.doi.org/10.1158/1538-7445.prca2012-a58 beck, J.P. 2012. MrBayes 3.2: Efficient Bayesian phylogenetic
Korovin, E.P. 1947. Generis Ferula (Tourn.) L. monographia illustrata. inference and model choice across a large model space. Syst. Biol.
Tashkent: Academiae Scientiarum UzRSS. 61: 539–542. http://dx.doi.org/10.1093/sysbio/sys029
Korovin, E.P. 1951. Ferula L. Pp. 62–142 in: Schischkin, B.K. (ed.), Safina, L.K. & Pimenov, M.G. 1983. Karpoanatomicheskiye osoben-
Flora SSSR, vol. 17. Moscow, Leningrad: Izdatelstvo Akademii nosti vidov roda Ferula podroda Peucedanoides (Apiaceae) w
Nauk SSSR. sviazi s sistematikoi roda. Bot. Zhurn. (Moscow & Leningrad)
Korovin, E.P. 1959. Ferula L. Pp. 399–438 in: Vvedensky, A.I. (ed.), 68: 730–739.
Flora Uzbekistana, vol. 4. Tashkent: Izdatelstvo Akademii Nauk Safina, L.K. & Pimenov, M.G. 1984. Feruly Kazakhstana. Alma-Ata:
Uzbekskoi SSR. Izdatelstvo “Nauka” Kazakhskoi SSR.
Kurzyna-Młynik, R., Oskolski, A.A., Downie, S.R., Kopacz, R., Safina, L.K. & Pimenov, M.G. 1990. Carpology of the species of type
Wojewódzka, A. & Spalik, K. 2008. Phylogenetic position of the subgenus of the Ferula and some problems of their systematics.
genus Ferula (Apiaceae) and its placement in tribe Scandiceae as Feddes Repert. 101: 135–151.
inferred from nrDNA ITS sequence variation. Pl. Syst. Evol. 274: http://dx.doi.org/10.1002/fedr.19901010305
47–66. http://dx.doi.org/10.1007/s00606-008-0022-2 Salichos, L. & Rokas, A. 2013. Inferring ancient divergences requires
Lanfear, R., Calcott, B., Ho, S.Y.W. & Guindon, S. 2012. Partition- genes with strong phylogenetic signals. Nature 497: 327–331.
finder: Combined selection of partitioning schemes and substi- http://dx.doi.org/10.1038/nature12130
tution models for phylogenetic analyses. Molec. Biol. Evol. 29: Salichos, L., Stamatakis, A. & Rokas, A. 2014. Novel information
1695–1701. http://dx.doi.org/10.1093/molbev/mss020 theory-based measures for quantifying incongruence among phy-
Lanfear, R., Calcott, B., Kainer, D., Mayer, C. & Stamatakis, A. logenetic trees. Molec. Biol. Evol. 31: 1261–1271.
2014. Selecting optimal partitioning schemes for phylogenomic http://dx.doi.org/10.1093/molbev/msu061
datasets. B. M. C. Evol. Biol. 14: 82. Shaw, J., Lickey, E.B., Beck, J.T., Farmer, S.B., Liu, W., Miller, J.,
http://dx.doi.org/10.1186/1471-2148-14-82 Siripun, K.C., Winder, C.T., Schilling, E.E. & Small, R.L. 2005.
Larkin, M.A., Blackshields, G., Brown, N.P., Chenna, R., McGetti- The tortoise and the hare II: Relative utility of 21 noncoding chlo-
gan, P.A., McWilliam, H., Valentin, F., Wallace, I.M., Wilm, A., roplast DNA sequences for phylogenetic analyses. Amer. J. Bot.
Lopez, R., Thompson, J.D., Gibson, T.J. & Higgins, D.G. 2007. 92: 142–166. http://dx.doi.org/10.3732/ajb.92.1.142
Clustal W and Clustal X version 2.0. Bioinformatics 23: 2947–2948. Shaw, J., Lickey, E.B., Schilling, E.E. & Small, R.L. 2007. Compari-
http://dx.doi.org/10.1093/bioinformatics/btm404 son of whole chloroplast genome sequences to choose noncoding
Lee, C.-S. & Downie, S.R. 2006. Phylogenetic relationships within regions for phylogenetic studies in angiosperms: The tortoise and
Cicuta (Apiaceae tribe Oenantheae) inferred from nuclear rDNA the hare III. Amer. J. Bot. 94: 275–288.
ITS and cpDNA sequence data. Canad. J. Bot. 84: 453–468. http://dx.doi.org/10.3732/ajb.94.3.275
http://dx.doi.org/10.1139/b06-016 Shneyer, V.S. 2010. On some cases of conflict between serological
Leigh, J.W., Susko, E., Baumgartner, M. & Roger, A.J. 2008. Test- and DNA sequences in Apiaceae studies. Pl. Diversity Evol. 128:
ing congruence in phylogenomic analysis. Syst. Biol. 57: 104–115. 173–180. http://dx.doi.org/10.1127/1869-6155/2010/0128-0008
http://dx.doi.org/10.1080/10635150801910436 Shneyer, V.S., Borschtschenko, G.P. & Pimenov, M.G. 1995. Immu-
Maddison, W.P. & Maddison, D.R. 2010. Mesquite: A modular system nochemical appraisal of relationships within the tribe Peuce­daneae
for evolutionary analysis, version 2.73. http://mesquiteproject.org (Apiaceae). Pl. Syst. Evol. 198: 1–16.
Miller, M.A., Pfeiffer, W. & Schwartz, T. 2010. Creating the CIPRES http://dx.doi.org/10.1007/BF00985104
Science Gateway for inference of large phylogenetic trees. Pp. Shneyer, V.S., Kutyavina, N.G. & Pimenov, M.G. 2003. System-
45–52 in: Proceedings of the Gateway Computing Environments atic relationships within and between Peucedanum and Angelica
Workshop (GCE), New Orleans, Louisiana, 14 Nov 2010. Piscat- (Umbelliferae-Peucedaneae) inferred from immunological studies
away: IEEE. http://dx.doi.org/10.1109/GCE.2010.5676129 of seed proteins. Pl. Syst. Evol. 236: 175–194.
Nazari, Z.E. & Iranshahi, M. 2011. Biologically active sesquiterpene http://dx.doi.org/10.1007/s00606-002-0239-4
coumarins from Ferula species. Phytotherapy Res. 25: 315–323. Spalik, K. & Downie, S.R. 2001. The utility of morphological charac-
http://dx.doi.org/10.1002/ptr.3311 ters for inferring phylogeny in Scandiceae subtribe Scandicinae
Percy, D.M., Argus, G.W., Cronk, Q.C., Fazekas, A.J., Kesanakurti, (Apiaceae). Ann. Missouri Bot. Gard. 88: 270–301.
P.R., Burgess, K.S., Husband, B.C., Newmaster, S.G., Barrett, http://dx.doi.org/10.2307/2666227
S.C.H. & Graham, S.W. 2014. Understanding the spectacular Spalik, K. & Downie, S.R. 2006. The evolutionary history of Sium
failure of DNA barcoding in willows (Salix): Does this result from sensu lato (Apiaceae): Dispersal, vicariance, and domestication
a trans-specific selective sweep? Molec. Ecol. 23: 4737–4756. as inferred from ITS rDNA phylogeny. Amer. J. Bot. 93: 747–761.
http://dx.doi.org/10.1111/mec.12837 http://dx.doi.org/10.3732/ajb.93.5.747
Pimenov, M.G. 1987. Leutea. Pp. 445–450 in: Hedge, I.C., Lamond, J.M. Spalik, K., Downie, S.R. & Watson, M.F. 2009. Generic delimitations
& Rechinger, K.H. (eds.), Flora Iranica, vol. 162, Umbelliferae. within the Sium alliance (Apiaceae tribe Oenantheae) inferred
Graz: Akademische Druck- und Verlagsanstalt. from cpDNA rps16-5′trnK(UUU) and nrDNA ITS sequences. Taxon
Pimenov, M.G. 1988. Monografitcheskaya reviziya roda Dorema 58: 735–748.
D. Don (Umbelliferae). Byull. Moskovsk. Obshch. Isp. Prir. Otd. Stamatakis, A. 2014. RAxML version 8: A tool for phylogenetic anal-
Biol. 93: 76–90. ysis and post-analysis of large phylogenies. Bioinformatics 30:
Pimenov, M.G. & Leonov, M.V. 1993. The genera of the Umbelliferae. 1312–1313. http://dx.doi.org/10.1093/bioinformatics/btu033
Richmond: Royal Botanic Gardens, Kew. Swofford, D.L. 2003. PAUP*: Phylogenetic analysis using parsimony
Pimenov, M.G., Terekhin, A.T., Devyatkova, G.N. & Baranova, (*and other methods), version 4. Sunderland, Massachusetts:
Y.V. 1978. Klassifikatsiya vidov roda Ferula L. (Umbelliferae) s Sinauer.

Version of Record 781

Panahi & al. • Phylogenetic relationships in Ferulinae (Apiaceae) TAXON 64 (4) • August 2015: 770–783

Valiejo-Roman, C.M., Terentieva, E.I., Samigullin, T.H., Pimenov, Yi, T.-S., Jin, G.-H. & Wen, J. 2015. Chloroplast capture and intra- and
M.G., Ghahremani-Nejad, F. & Mozaffarian, V. 2006. Molecu- inter-continental biogeographic diversification in the Asian-New
lar data (nrITS-sequencing) reveal relationships among Iranian World disjunct plant genus Osmorhiza (Apiaceae). Molec. Phylo-
endemic taxa of Umbelliferae. Feddes Repert. 117: 367–388. gen. Evol. 85: 10–21. http://dx.doi.org/10.1016/j.ympev.2014.09.028
http://dx.doi.org/10.1002/fedr.200611106 Zhou, J., Gong, X., Downie, S.R. & Peng, H. 2009. Towards a more
Wen, J. & Zimmer, E.A. 1996. Phylogeny and biogeography of robust molecular phylogeny of Chinese Apiaceae subfamily Api-
Panax L. (the ginseng genus, Araliaceae): Inferences from ITS oideae: Additional evidence from nrDNA ITS and cpDNA intron
sequences of nuclear ribosomal DNA. Molec. Phylogen. Evol. 6: (rpl16 and rps16) sequences. Molec. Phylogen. Evol. 53: 56–68.
167–177. http://dx.doi.org/10.1006/mpev.1996.0069 http://dx.doi.org/10.1016/j.ympev.2009.05.029

Appendix 1. Accessions of Apiaceae subtribe Ferulinae and outgroups from which nrDNA ITS and cpDNA sequence data were obtained, with correspond-
ing voucher information and GenBank reference numbers (GB no.). Newly generated sequences are identified with an asterisk behind the accession number.
Taxon name, voucher information; ITS GB no., rpoB-trnC spacer GB no., rps16 intron GB no., rpoC1 intron GB no.
Dorema aitchisonii Korovin ex Pimenov, Afghanistan, prov. Heart, between Obeh and Khodja Chisht, 10 May 1969, Hedge, Wendelbo & Ekberg W 7774 (E,
EBH); KJ698410*, KJ660596*, KJ201543*, KJ660443*. Dorema ammoniacum D.Don, Iran, prov. Yazd, ca. 50 km E of Bafgh, Sendoo village, 30 May 1986,
Assadi & Bazgosha 56116 (TARI); KJ698411*, KJ660597*, KJ698349*, KJ660444*. Dorema aucheri Boiss., Iran, prov. Fars, Firuzabad, Meimand, Kuh-e
Sefidar from Tang-e Riz, 13 Jun 1992, Mozaffarian 71458 (TARI); KJ698412*, KJ660598*, KJ698350*, KJ660445*. Dorema aureum Stocks, Afghanistan,
prov. Bamian, Darrah Siakar, ca. 15 km S of Doab, 21 May 1962, Hedge & Wendelbo W 3411 (E, EBH); KJ698413*, KJ660599*, KJ698351*, KJ660446*.
Dorema glabrum Fisch. & C.A.Mey., Iran, prov. Azarbaijan, ca. 10 km from Gharaziaeddin to Marand, 14 Jun 2005, Mozaffarian 87182 (TARI); KJ698414*,
KJ660600*, KJ698352*, KJ660447*. Dorema hyrcanum Koso-Pol., Iran, prov. Gorgan, Park-e Golestan, Almeh, 5 Jun 1987, Mozaffarian & Abouhamzeh
59046 (TARI); KJ698415*, KJ660601*, KJ698353*, KJ660448*. Ferula akitschkensis B.Fedtsch. ex Koso-Pol., Kazakhstan, Dzhungar Alatau, Kopal Mtns.,
Pendzhak, 1909, Lipsky 7074 (LE); DQ379385, KJ660602*, KJ698354*, KJ660462*. Ferula badrakema Koso-Pol., Turkmenistan, Badkhyz, near Kepelia, 12
May 1976, Botschantzev 696 (LE); DQ379388, KJ660608*, KJ698360*, KJ660467*. Ferula botschantzevii Korovin, Tajikistan, Konki-tau Mtns., near Khozhadi
village, 19 May 1960, Nepli (LE); DQ379389, KJ660611*, KJ698363*, KJ660471*. Ferula ceratophylla Regel & Schmalh., Kazakhstan, Syrdaria Karatau,
Teksakkan Mtns., 7 Jul 1973, Kamelin T-25 (LE); DQ379390, KJ660614*, KJ698367*, KJ660475*. Ferula clematidifolia Koso-Pol., S Tajikistan, near Tutkaul
village, 12 Jun 1960, Botschatnzsev & Egorova 1290 (LE); DQ379391, KJ660615*, KJ698368*, KJ660476*. Ferula communis L., Spain, Almería, Vélez-Rubio,
12 Jun 2004, Sánchez-Gómez (number unknown) (Universidad de Zaragoza, Spain); DQ379392, KJ660616*, KJ698369*, KJ660477*. Ferula decurrens
Korovin, Tajikistan, South-Tadjik depression, near Sarygor post, 4 Jun 1961, Soskov & Junusov 623 (LE); DQ379393, KJ660619*, KJ698372*, KJ660480*.
Ferula dissecta (Ledeb.) Ledeb., Kazakhstan, near Balkhash lake, Arkharly Mtns., 28 May 1976, Botschantzev & Botschantzeva 1010 (LE); DQ379394,
KJ660620*, KJ698373*, KJ660481*. Ferula diversivittata Regel & Schmalh, Uzbekistan, Nuratau range, near Sintob village, 24 Jun 1971, Botschantzev &
Kamelin 447 (LE); DQ379395, KJ660621*, KJ698374*, KJ660482*. Ferula dshizakensis Korovin, Uzbekistan, prov. Samarkand, between Katta-Kurgan and
mine Ingichka, 19 May 1975, Botschantzev 280 (LE); DQ379397, KJ660622*, KJ698375*, KJ660483*. Ferula dubjanskyi Korovin ex Pavlov, Kazakhstan,
Muiun-Kum, near Ak-Zhar lake, Kotlan, 19 Jun 1970, Soskov & Bukhteeva 51 (LE); DQ379396, KJ660623*, KJ698376*, KJ660484*. Ferula equisetacea
Koso-Pol., Tajikistan, Gyssar range, basin of river Varzob, near Kabuty, 3 Aug 1973, Kamelin s.n. (LE); DQ379398, KJ660624*, KJ698377*, KJ660485*.
Ferula eugenii Kamelin, Tajikistan, Gyssar range, basin of river Varzob, Kondar gorge, 13 Jun 1970, Kamelin 86 (LE); DQ379399, KJ660625*, KJ698378*,
KJ660486*. Ferula fedoroviorum Pimenov, Kyrgyzstan, near Naryn river, 18 Jul 1985, Kluikov 201 (LE); DQ379400, KJ660627*, KJ698380*, KJ660488*.
Ferula ferganensis Lipsky ex Korovin, Uzbekistan, Koitash range, near Kiotash village, 28 Jun 1971, Botschantzev & Kamelin 608 (LE); DQ379401, KJ660628*,
KJ698381*, KJ660489*. Ferula feruloides (Steud.) Korovin, Kazakhstan, prov. Almaty, near pass Arkhanlyk, 1952, Maccagetov 3775 (LE); DQ379402,
KJ660629*, KJ698382*, KJ660490*. Ferula foetida (Bunge) Regel, Uzbekistan, Kysylkum, prov. Bukhara, near Shafrikan, 19 Apr 1975, Ikonnikov 25 (LE);
DQ379403, KJ660632*, KJ698385*, KJ660493*. Ferula foetidissima Regel & Schmalh., Tajikistan, Upper Badakhshan, near Khorog Botanical Garden, 24
Jun 1966, Kamelin s.n. (LE); DQ379404, KJ660633*, KJ698386*, KJ660494*. Ferula glauca L., Spain, Gerona, 6 Jul 1923, Augustin 4821 (KRA); DQ379406,
KJ660637*, KJ698390*, KJ660498*. Ferula gummosa Boiss., Turkmenistan, Nukhur Mtns., 24 Apr 1913, Samokish s.n. (LE); DQ379407, KJ660642*,
KJ698395*, KJ660503*. Ferula gypsacea Korovin, Kazakhstan, south part, Alym-tau, 22 Jun 1973, Kamelin 11 (LE); DQ379408, KJ660643*, KJ698396*,
KJ660504*. Ferula karakalensis Korovin, Turkmenistan, Malye Balkhany Mtns., near Akcha-Kuima, 22 May 1949, Tarasov 25 (LE); DQ379409, KJ660651*,
KJ698404*, KJ660512*. Ferula karatavica Regel & Schmalh., Kazakhstan, Karatau Mtns., near Arys river, 16 Jun 1959, Botschantzev 345 (LE); DQ379410,
KJ660652*, KJ698405*, KJ660513*. Ferula karataviensis (Regel & Schmalh.) Korovin ex Pavlov, Kazakhstan, between Karatau and Ulken-tau, 18 Jun 1959,
Pjansaeva 64 (LE); DQ379411, KJ660653*, KJ698406*, KJ660514*. Ferula karelinii Bunge, Kazakhstan, Almaty region, Altyn-Emel State National Park,
Singing Barkhan, 24 May 2008, Rae, Gardner & Ogar 72 (E, EBH); KJ660792*, KJ660655*, KJ698408*, KJ660516*. Ferula kelifi Korovin, Uzbekistan,
Guvar-Baisunski massif, near village Shurab, 16 May 1979, Kamelin & Makhmedov 185 (LE); DQ379442, KJ660657*, KJ660350*, KJ660518*. Ferula kelleri
Koso-Pol., Kyrgyzstan, prov. Dzhalalabad, Fergana range, Ermendy river, Mt. Alash-bashi, 8 Jul 1945, Fedorov & Ilina 58 (LE); DQ379405, KJ660658*,
KJ660351*, KJ660519*. Ferula kingdon-wardii H.Wolff, China, Yunnan, pref. Diqing, co. Deqin, side valley between Deqin and Mekong River, 2 Jun 1993,
Alden & al. (KEG) 711 (E, EBH); KJ660794*, KJ660659*, KJ660352*, KJ660520*. Ferula kirialovii Pimenov, Uzbekistan, Chatkal range, Akbulak river, 3
Jul 1973, Kamelin 2 a, b (LE); DQ379443, KJ660660*, KJ660353*, KJ660521*. Ferula kokanica Regel & Schmalh., Uzbekistan, Aktau Mtns., Ingichke, near
Nurat, 26 Jun 1971, Botschantzev & Kamelin 457 (LE); DQ379412, KJ660661*, KJ660354*, KJ660522*. Ferula korshinskyi Korovin, Uzbekistan, Tcharvak,
Arslanbai, 16 May 1916, Lipsky s.n. (LE); DQ379413, KJ660663*, KJ660356*, KJ660524*. Ferula koso-poljanskyi Korovin, Tajikistan, Upper Badakhshan,
near mouth of Iazgulem river, 18 Jun 1968, Ladygina & al. 18834 (LE); DQ379414, KJ660664*, KJ660357*, KJ660525*. Ferula kyzylkumica Korovin, Uzbeki-
stan, prov. Samarkand, Sangrultau Mtns., 10 Jun 1948, Rodin & Arkadiev 389 (LE); DQ379415, KJ660665*, KJ660358*, KJ660526*. Ferula lapidosa Korovin,
Kazakhstan, prov. Dzhambul, 10 km S from village Kaindy, 30 Jun 1970, Soskov & Bukhteeva 107 (LE); DQ379416, KJ660666*, KJ660359*, KJ660527*.
Ferula lehmannii Boiss., Kazakhstan, prov. Aktubinsk, Baiganin region, near Danguz-tau, 13 Jun 1956, Yunatov & Kuznetsov s.n. (LE); DQ379418, KJ660667*,
KJ660360*, KJ660528*. Ferula leucographa Korovin, Kazakhstan, Mashat-tau, Kelte-mashat, 27 Jul 1973, Kamelin 26 (LE); DQ379419, KJ660669*, KJ660362*,
KJ660529*. Ferula licentiana Hand.-Mazz., China, Yunnan, Zhongdian, N of the town, 2 Jul 1994, ACE 900 (E, EBH); KJ660796*, KJ660670*, KJ660363*,
KJ660530*. Ferula linkii Webb, cult. Conservatoire botanique de la Ville de Mulhouse, 9384, 3 Nov 2005; DQ379421, KJ660672*, KJ660365*, KJ660532*.
Ferula litwinowiana Koso-Pol., Turkmenistan, near Uzboi, kolkhoz Dzhigurdekli, 14 May 1953, Rustamov s.n. (LE); DQ379423, KJ660674*, KJ660367*,
KJ660533*. Ferula loscosii (Lange) Willk., Spain, Albacete, Hellín, Las Minas, 28 May 2004, Pérez-Collazos & al. (number unknown) (Universidad de
Zaragoza, Spain); DQ379425, KJ660676*, KJ660370*, KJ660535*. Ferula mogoltavica Lipsky ex Korovin, Tajikistan, Khudjent, Mogol-tau, Tchashma-arz-
anok, 18 May 1924, Popov & Vvedensky 763 (LE); DQ379429, KJ660679*, KJ660373*, KJ660538*. Ferula moschata (Reinsch) Koso-Pol., Tajikistan, Gyssar
range, near Iskander-kul lake, 31 Jul 1960, Egorova 2235 (LE); DQ379457, KJ660680*, KJ660374*, KJ660539*. Ferula nevskii Korovin, Uzbekistan, Kugitangh
range, near village Tengideval, 13 Jun 1971, Botschantzev 111 (LE); DQ379432, KJ660684*, KJ660378*, KJ660543*. Ferula nuratavica Pimenov, Uzbekistan,

782 Version of Record

 TAXON 64 (4) • August 2015: 770–783 Panahi & al. • Phylogenetic relationships in Ferulinae (Apiaceae)

Appendix 1. Continued.
Aktau Mtns., sai Petyr, 14 Jul 1972, Kamelin 113 (LE); DQ379433, KJ660686*, KJ660380*, KJ660545*. Ferula olgae Regel & Schmalh., Uzbekistan, Aktau
Mtns., near Nurat, Ingichke, 13 Jul 1970, Botschantzev 150 (LE); DQ379434, KJ660687*, KJ660381*, KJ660546*. Ferula olivacea (Diels) H.Wolff ex Hand.-
Mazz., China, Yunnan, Lijiang, Yulong-Shan Ganghoba, 26 May 1985, Chamberlain, Ming, Yuan & al. 229 (E); KJ660802*, KJ660688*, KJ660382*, KJ660547*.
Ferula oopoda (Boiss. & Buhse) Boiss., Turkmenistan, Kopet-Dagh, Alty-Topdon gorge, 6 Jun 1930, Bobrov 94 (LE); DQ379435, KJ660689*, KJ660383*,
KJ660548*. Ferula ovina (Boiss.) Boiss., Kazakhstan, Dzhungar Alatau, Tchulak Mtns., gorge Tchulak-Dzhigde, 6 Jun 1956, Goloskokov s.n. (LE); DQ379449,
KJ660692*, KJ660386*, KJ660551*. Ferula pallida Korovin, Kazakhstan, South Kazakhstan, Alym-tau, 3 Jun 1960, Priakhin & al. s.n. (LE); DQ379438,
KJ660697*, KJ660391*, KJ660556*. Ferula penninervis Regel & Schmalh., Kirghistan, Alaici, valley of Akbura, Alczaly, 17 Jul 1981, Pimenov, Vassiljeva
& Tomkovich 482 (E, EBH); KJ660806*, KJ660698*, KJ660392*, KJ660557*. Ferula potaninii Korovin ex Pavlov, Kazakhstan, near Balkhash lake, Arkharly
Mtns., 28 May 1976, Botschantzev & Botschantzeva 1018 (LE); DQ379440, KJ660700*, KJ660394*, KJ660559*. Ferula renardii (Regel & Schmalh.) Pimenov,
Kyrgyzstan, Dzhalalabad prov., Arkesht village, Sary-Chilek, 10 Aug 1945, Vipper 30 (LE); DQ379444, KJ660703*, KJ660397*, KJ660562*. Ferula rubroarenosa
Korovin, Uzbekistan, Alai range, Sokh-Khashi-mardan, pass Meting-bel, 16 Jul 1962, Botschantzev 186 (LE); DQ379430, KJ660704*, KJ660398*, KJ660563*.
Ferula samarkandica Korovin, Uzbekistan, Chimgan Mtns., source of Chimganka river, 29 Jun 1974, Kamelin 1420 (LE); DQ379445, KJ660707*, KJ660401*,
KJ660566*. Ferula schtschurowskiana Regel & Schmalh. ex Regel, Turkmenistan, Charshangin region, Kugitangh valley, between villages Kugitangh and
Guardok, Albasty, 12 Oct 1970, Botschantzev 103 (LE); DQ379446, KJ660708*, KJ660402*, KJ660567*. Ferula sibirica Willd., Kazakhstan, prov. Dzeskasgan,
near Zhairem, 27 Jul 1991, Pimenov & Kljuykov 133 (G); DQ379447, KJ660712*, KJ660405*, KJ660571*. Ferula songarica Pall. ex Schult., Kazakhstan, prov.
Akmolinsk [Astana], near Tengyz lake, 17 Aug 1954, Kalinina 1993 (LE); DQ379448, KJ660714*, KJ660407*, KJ660573*. Ferula szowitsiana DC., Turk-
menistan, central Kopet-Dagh, 14 Jun 1969, Meshtsheriakov s.n. (LE); DQ379427, KJ660718*, KJ660411*, KJ660577*. Ferula tadshikorum Pimenov,
Tajikistan, Saiglok range, near Sebiston village, 23 May 1979, Kamelin & Mikhailov 380 (LE); DQ379450, *KJ660722, *KJ660415, *KJ660581. Ferula
tatarica Fisch. ex Spreng., Kazakhstan, prov. Karaganda, Atasuiski region, 29 Jul 1959, Varivtseva 335 (LE); DQ379451, KJ660723*, KJ660416*, KJ660582*.
Ferula teterrima Kar. & Kir., Kazakhstan, near Balkhash lake, 5 Jul 1958, Rusliaeva s.n. (LE); DQ379452, KJ660724*, KJ660417*, KJ660583*. Ferula tin-
gitana L., cult. Conservatoire botanique de la Ville de Mulhouse, 9521, 3 Nov 2005; DQ379453, KJ660725*, KJ660418*, KJ660584*. Ferula tschimganica
Lipsky ex Korovin, Kyrgyzstan, Syrdaria Karatau, Kzyladyr Mtns., near gorge Kysh-Kapka, 2 Jul 1973, Kamelin 3-KA (LE); DQ379428, KJ660728*, KJ660421*,
KJ660587*. Ferula tuberifera Korovin, Turkmenistan, Kugitangh range, near village Kugitangh, 13 Oct 1970, Botschantzev 106 (LE); DQ379454, KJ660729*,
KJ660422*, KJ660588*. Ferula ugamica Korovin, Uzbekistan, Angren region, near village Chesh-su, 2 Jun 1954, Knorring 59 (LE); DQ379455, KJ660731*,
KJ660424*, KJ660590*. Ferula undulata Pimenov & J.V.Baranova, Turkmenistan, central Kopet-Dagh, 35 km S from Gueok-Tepe, near Novo-Kheirabad, 25
May 1963, Gubanov s.n. (LE); DQ379456, KJ660732*, KJ660425*, KJ660591*. Ferula varia (Schrenk) Trautv., Kazakhstan, Central Betpak-dala, Kok-ashik,
17 Jun 1960, Grubov s.n. (LE); DQ379386, KJ660733*, KJ660426*, KJ660592*. Ferula violacea Korovin, Afghanistan, Qataghan, Mirza Atbili pass, SE of
Samangan, 10 Jun 1962, Hedge & Wendelbo 4030 (E, EBH); KJ660827*, KJ660734*, KJ660428*, KJ660594*. Glaucosciadium cordifolium (Boiss.) B.L.Burtt
& P.H.Davis, Cons. Bot. Mulhouse 98112, 8 Dec 1999, Reduron s.n. (WA); DQ379459, KJ660745*, KJ660439*, KJ660458*. Laserpitium archangelica Wulfen,
Slovenia, Mt. Slivnica above Grahovo near Cerknica Lake, 12 Aug 1967, Mayer 63416 (KRAM); FJ415153, KJ832093*, KJ832103*, KJ832098*. Laserpitium
carduchorum Hedge & Lamond, Turkey, Bitlis/Van, 10 km SE of Pelli, 8 Jul 1954, Davis & Polunin 22551 (E); FJ415117, KJ832094*, KJ832104*, KJ832099*.
Laserpitium eliasii subsp. ordunae P.Monts., Spain, Vizcaya, Salvada, Orduña, 30 Aug 1983, Montserrat (number unknown) (JACA733483); FJ415119,
KJ832095*, KJ832105*, KJ832100*. Laserpitium nestleri P.Monts. subsp. nestleri, Italy, Villafranca, 1991, Fabregat & Udias 2007 (BCC); FJ415123, KJ832097*,
KJ832107*, KJ832102*. Leutea avicennae Mozaff., Iran, prov. Hamadan, Kabudar-Ahang, Daghdaghabad, Gholiabad Mnts., without date, Mozaffarian 65041
(TARI); KJ660830*, KJ660736*, KJ660430*, KJ660449*. Leutea cupularis (Boiss.) Pimenov, Iran, prov. Fars, Dena Mnts., Bijan ghaut, 2 Aug 1978, Assadi
& Mozaffarian 31236 (TARI); KJ660831*, KJ660737*, KJ660431*, KJ660450*. Leutea elbursensis Mozaff., Iran, prov. Tehran, Mtns. NW of Tehran, Suleg-
hun valley, 19 Jul 1980, Assadi, Mozaffarian & Jamzad 33570 (TARI); KJ660832*, KJ660738*, KJ660432*, KJ660451*. Leutea glaucopruinosa (Rech.f.)
Akhani & Salimian, Iran, prov. Mazandaran, Kelardasht, Rutbarak, Vandarbon, Kalahou, 27 Jul 1983, Termeh & Zargari 040483-E (W 0023608); KJ660833*,
KJ660739*, KJ660433*, KJ660452*. Leutea gracillima Pimenov, Iran, Golestan National Park, 7 km NEN of Tangegol, Gerieh-Sar, 23 Aug 1995, Akhani
12060 (W 1999-03655); KJ660834*, KJ660740*, KJ660434*, KJ660453*. Leutea nematoloba (Rech.f.) Pimenov, Iran, prov. Mazanderan, in valley of Calus
river, 6–7 Aug 1948, Rechinger & Rechinger 6668 (W 02835); KJ660835*, KJ660741*, KJ660435*, KJ660454*. Leutea petiolaris (DC.) Pimenov, Iran, prov.
Azarbaijan, Zanjan, Gheidar Mnts., 25 Jul 1998, Mozaffarian & Massoumi 78152 (TARI); KJ660836*, KJ660742*, KJ660436*, KJ660455*. Leutea polyscias
(Boiss.) Pimenov, Iran, prov. Gillan, Manjil, Harzvil, 25 Aug 1987, Mozaffarian 64227 (TARI); KJ660837*, KJ660743*, KJ660437*, KJ660456*. Leutea
rechingeri (Leute) Pimenov, Iraq, Irbil district (Kurdistan), Zagros Mnts., Halgurd, 10–14 Aug 1957, Rechinger 11416 (W 05825); KJ660838*, KJ660744*,
KJ660438*, KJ660457*. Mozaffariania insignis Pimenov & Maassoumi, Iran, prov. Khuzestan, Masjed Soleyman to Haftgel, 27 Jun 1996, Mozaffarian 77099
(TARI); KJ660839*, KJ660746*, KJ660440*, KJ660459*. Pachypleurum alpinum Ledeb., Mongolia, Hovsgol National Park, Dorboljen, Borhat, 24 Jul 1996,
Caddick 172 (E, EBH); KJ660840*, KJ660747*, KJ660441*, KJ660460*. Pachypleurum xizangense H.T.Chang & R.H.Shan, China, Qinghai, Madoi Xian, E
of Malayiwan, on road between Gonghe and Madoi, 10 Aug 1996, Ho, Bartholomew, Watson & Gilbert 1580 (E, EBH); KJ660841*, KJ660748*, KJ660442*,
KJ660461*.

Version of Record 783

View publication stats