Phylogeny and New Intrageneric Classification of Allium (Alliaceae) Based on Nuclear Ribosomal DNA ITS Sequences

Abstract

The internal transcribed spacer region (ITS) of nuclear ribosomal DNA was sequenced from 195 representative species of Allium, two species of Nothoscordum, and one species each of Ipheion, Dichelostemma, and Tulbaghia. Within the Allium species the lengths of the ITS regions were in a range from 612 to 661 base pairs and pairwise genetic distances reached up to 46%. The ITS data supported the inclusion of Nectaroscordum, Caloscordum, and Milula into Allium. Subgenera Rhizir-ideum and Allium, as well as sects. Reticulatobulbosa and Oreiprason were non-monophyletic taxa. Based on the phylogenetic relations, a new classification of genus Allium consisting of 15 monophy-letic subgenera is presented. Sections Microscordum, Anguinum, Porphyroprason, Vvedenskya, Bu-tomissa, Cyathophora, and Reticulatobulbosa are raised to subgeneric rank.

Full text

Aliso 22, pp. 372–395
䉷
2006, Rancho Santa Ana Botanic Garden
PHYLOGENY AND NEW INTRAGENERIC CLASSIFICATION OF ALLIUM (ALLIACEAE)
BASED ON NUCLEAR RIBOSOMAL DNA ITS SEQUENCES
N
IKOLAI
F
RIESEN
,
1,3
R
EINHARD
M. F
RITSCH
,
2
AND
F
RANK
R. B
LATTNER
2
1
Botanical Garden of the University of Osnabru¨ck, Albrechtstr. 29, 49076 Osnabru¨ck, Germany;
2
Institute of Plant
Genetics and Crop Plant Research (IPK), Corrensstr. 3, 06466 Gatersleben, Germany
3
Corresponding author (nfriesen@uni-osnabrueck.de)
ABSTRACT
The internal transcribed spacer region (ITS) of nuclear ribosomal DNA was sequenced from 195
representative species of Allium, two species of Nothoscordum, and one species each of Ipheion,
Dichelostemma, and Tulbaghia. Within the Allium species the lengths of the ITS regions were in a
range from 612 to 661 base pairs and pairwise genetic distances reached up to 46%. The ITS data
supported the inclusion of Nectaroscordum, Caloscordum, and Milula into Allium. Subgenera Rhizir-
ideum and Allium, as well as sects. Reticulatobulbosa and Oreiprason were non-monophyletic taxa.
Based on the phylogenetic relations, a new classification of genus Allium consisting of 15 monophy-
letic subgenera is presented. Sections Microscordum, Anguinum, Porphyroprason, Vvedenskya, Bu-
tomissa, Cyathophora, and Reticulatobulbosa are raised to subgeneric rank. Sections Austromontana
N. Friesen, Eduardia N. Friesen, Mediasia F. O. Khassanov, S. C. Yengalycheva et N. Friesen, Ni-
grimontana N. Friesen, Falcatifolia N. Friesen, and Condensatum N. Friesen are newly described.
Series Daghestanica, Pallasia, and Scabriscapa, as well as subsects. Eremoprasum, Longivaginata,
and Sikkimensia are raised to sectional rank. A taxonomic conspectus of Allium at sectional level is
given.
Key words: Allium, classification, evolution, internal transcribed spacer, phylogenetic analysis, tax-
onomy.
INTRODUCTION
Allium L. is probably the largest genus of the petaloid
monocotyledons, comprising some 750 species (Stearn
1992). The genus is characterized by having bulbs enclosed
in membranous (sometimes finally fibrous) tunics, free or
almost free tepals, and often a subgynobasic style. Most spe-
cies produce remarkable amounts of cysteine sulphoxides
causing the well-known characteristic odor and taste. The
genus is naturally distributed only in the Northern Hemi-
sphere, mainly in regions that are seasonally dry. It has a
main center of diversity in southwest and central Asia and
a second smaller one in North America. Allium includes
some economically important species like common onion,
garlic, chives, and leek under worldwide cultivation, and
also species with medicinal properties and others of horti-
cultural merit (Fritsch and Friesen 2002). Allium is a member
of family Alliaceae subfamily Allioideae Herb. (Fay and
Chase 1996). Takhtajan (1987, 1997) placed Alliaceae in the
order Amaryllidales close to Hyacinthaceae and Amarylli-
daceae. After Fay and Chase (1996) and Friesen et al. (2000)
subfamily Allioideae consist only of Allium (including Ca-
loscordum Herb., Milula Prain, and Nectaroscordum Lindl.).
No comprehensive monograph of the genus has been com-
piled since Regel’s in 1875 and the taxonomy is complicated,
with a proliferation of synonyms and disagreement as to the
subdivision of the genus. The history of infrageneric clas-
sification begins prior to Linnaeus (1753) who accepted 30
species in three alliances. Later studies recognized an in-
creasing number of infrageneric groups together with an en-
larged number of species: 6 sections (which trace back to
informal groups established by Don in 1832) and 262 spe-
cies (Regel 1875); 9 sections and 228 species for the former
USSR (Vvedensky 1935) alone; 3 subgenera, 36 sections
and subsections and ca. 600 species (Traub 1968); 6 sub-
genera, 44 sections and subsections (Kamelin 1973); 6 sub-
genera, 50 sections and subsections for 600–700 species
(Hanelt et al. 1992). In this last mentioned paper, subgen.
Rhizirideum was explicitly regarded as a polyphyletic assem-
blage of different phylogenetic lines, and subgen. Allium as
containing three main entities: the very diverse but insuffi-
ciently subdivided sect. Scorodon, the more homogenous
and distinctly specialized sect. Codonoprasum, and (the larg-
est among them) sect. Allium separated by other morpholog-
ical and phenetic specializations. The latter two sections
were regarded as phylogenetically young groups. Up to now
about 1400 species names have been proposed, often from
inadequate or incomplete material, which have later proven
synonymous with existing species (Gregory et al. 1998).
A first approach to structure the genus Allium by molec-
ular markers was published by Linne von Berg et al. (1996)
who conducted a chloroplast DNA RFLP analysis. They rec-
ognized the then established subgenera, but found that sub-
gen. Amerallium and Bromatorrhiza could not be clearly dis-
tinguished. A closer investigation of the Amerallium–Bro-
matorrhiza complex (Samoylov et al. 1995, 1999) proved
again the polyphyletic state of subgen. Bromatorrhiza, which
had to be integrated into subgen. Amerallium (all species
with x
⫽
7) and subgen. Rhizirideum (species with x
⫽
8).
The distribution of Amerallium species in the Old and New
World was also reflected in these phylogenetic data, as well
as in an internal transcribed spacer region (ITS) sequence
analysis of Dubouzet and Shinoda (1999). Mes et al. (1999)
included 29 species of Allium and 7 species of related genera

VOLUME 22 373Phylogeny of Allium
in a phylogenetic study using restriction fragment length
polymorphism (RFLP) data from polymerase chain reaction
(PCR) amplified chloroplast DNA. In this analysis the large
subgen. Rhizirideum and Allium, which had still remained
largely intact in the previously published studies, proved to
be polyphyletic, and Nectaroscordum siculum was placed in
Allium. Although some sections in the monophyletic subgen.
Melanocrommyum appeared then artificial, the taxonomy at
the level of sections remained more or less intact. Dubouzet
et al. (1997) proposed a first molecular phylogeny of subgen.
Rhizirideum based on nuclear DNA markers. Their results
largely confirmed the taxonomic system of Hanelt et al.
(1992). However, no species from outside the analyzed sub-
genus were included in this study, nor in some other recent
phylogenetic analyses (subgen. Melanocrommyum: Dubouz-
et and Shinoda 1998; subgen. Rhizirideum: van Raamsdonk
et al. 2000), thus preventing unambiguous circumscriptions
of these subgenera.
Another important aspect in several molecular phyloge-
netic studies is the quality of the studied plant material. Of-
ten research was conducted on seeds obtained from botanical
gardens or seed companies, mostly from free-pollinated
specimens. In our experience over 60% of such materials
were wrongly determined or had hybrid origins, thus imped-
ing phylogenetic studies (Friesen et al. 1999) when used
without further confirmation of their status.
Here we present a study where we used sequence data of
the ITS region of nuclear ribosomal DNA (nrDNA) to assess
phylogenetic relationship of the entire genus Allium. Fur-
thermore, to evaluate the proposed poly- or paraphyletic or-
igin of some Allium subgenera and sections (Mes et al. 1997,
1999; Klaas and Friesen 2002), we used a relatively dense
taxon sample, representing all major groups of the genus.
The ITS region, including the 5.8S nrDNA and the two spac-
ers ITS-1 and ITS-2, has proven to be an excellent source
of information from the nuclear genome in plants at the in-
trageneric level (e.g., Baldwin 1992; Baldwin et al. 1995;
Dubouzet and Shinoda 1999; Friesen et al. 2000; Blattner
2004). We also provide a new intrageneric classification of
Allium based on our ITS data, but also integrate earlier mo-
lecular studies to circumscribe monophyletic taxa. We do not
intend to discuss details below the section level here. Analy-
ses of the biogeography of the genus and several subgeneric
groups, as well as a discussion of the hybrid origin of Allium
taxa, will be published elsewhere.
MATERIALS AND METHODS
Plant Material
In this investigation 221 samples representing 196 species
of genus Allium, and one species each of the outgroup genera
Ipheion Raf., Tulbaghia L., and Dichelostemma Kunth, and
two species of Nothoscordum Kunth (Fay and Chase 1996;
Mes et al. 1997; Friesen et al. 2000) were studied. Most of
the accessions were selected from the Allium living collec-
tion of the Department of Taxonomy, IPK Gatersleben. This
collection comprises more than 340 species, mostly with
multiple accessions per species. Most of these accessions
were collected from naturally occurring populations. The ac-
cession numbers H600 and higher represent herbarium spec-
imens (herbaria: TASH, GAT, NS, and LE). Accession num-
ber, the taxonomical classification (Hanelt et al. 1992), and
origin of species examined are indicated in Table 1. Sub-
genera Rhizirideum and Allium were represented by 165 ac-
cessions and representatives of each of the 37 sections of
the subgen. Amerallium, Caloscordum, Nectaroscordum, and
Melanocrommyum were included. ITS sequences for some
species from subgen. Amerallium (Allium roseum L., A. cha-
maemoly L., A. neapolitanum Cirillo, A. zebdanense Boiss.
et Noe¨, A. validum S. Wats., A. sanbornii A. Wood, A. stel-
latum Ker., A. bolanderi S. Wats., A. dichlamydeum E. L.
Greene, A. unifolium Kell., A. hyalinum M. K. Curran, and
A. amplectens Torr.) were taken from Dubouzet and Shinoda
(1999). In all cases when accessions were placed in prelim-
inary trees at an unexpected position, additional accessions
of the same species were analyzed to corroborate the phy-
logenetic affiliation of the species.
Molecular Methods
DNA was isolated with the NucleoSpin Plant kit (Mach-
erey-Nagel, Du¨ren, Germany) according to the instructions
of the manufacturer. The concentration of the extracted DNA
was checked on an agarose gel. Isolated DNAs were used
directly in PCR amplifications. For most samples, the nr-
DNA ITS region (ITS-1, 5.8S nrDNA subunit, and ITS-2)
was amplified using primers ITS-A and ITS-B (Blattner
1999). ITS-1 and ITS-2 were amplified separately when
DNAs from herbarium sheets were used, in these cases un-
der inclusion of primers ITS-A together with ITS-C, and
ITS-B with ITS-D (Blattner 1999). PCR was carried out in
a Grant Autogene II thermo-cycler (Grant Instruments, Cam-
bridge, UK) programmed as: 95
⬚
C for 2 min [55
⬚
C for 30
sec, 70
⬚
C for 1 min, 95
⬚
C for 20 sec]
32
70
⬚
C for 7 min.
Amplification was carried out with 1 Unit Taq DNA poly-
merase (Boehringer, Mannheim, Germany) in the supplied
reaction buffer, 0.2 mM of each dNTP, 50 pmol of each
primer and 10–25 ng of total DNA in 50
␮
L reaction vol-
ume. The PCR products were separated on an agarose gel
and purified with the NucleoSpin Gel Extraction kit (Mach-
erey-Nagel, Du¨ren, Germany). After checking DNA concen-
tration on a 1.8% agarose gel, about 40 ng PCR product was
used in a 10
␮
L cycle sequencing reaction with the ABI
BigDye Terminator Kit (Applied Biosystems, Foster City,
California, USA) according to the instructions of the man-
ufacturer. The internal primers ITS-SF, ITS-SR (Blattner et
al. 2001), and partly also ITS-C and ITS-D were used as
sequencing primers. The sequencing reactions were analyzed
on ABI377 DNA sequencers (Applied Biosystems).
Data Analysis
Forward and reverse sequences from every individual
were manually edited with CHROMAS* vers. 1.5 software
(Technelysium Pty. Ltd., Tweantin, Queensland, Australia)
and combined in single consensus sequences. The sequences
of all 221 samples were aligned with CLUSTAL
㛮
X (Thomp-
son et al. 1997) and the alignment was manually adjusted
where necessary. Phenetic and cladistic analyses were con-
ducted in PAUP* vers. 4.0b10 (Swofford 2002), model-
based analyses with Bayesian inference (BI) in MrBayes
3.0b4 (Ronquist and Huelsenbeck 2003). Different models
of sequence evolution were first evaluated with Modeltest

374 ALISOFriesen, Fritsch, and Blattner
Table 1. New and traditional classifications for Allium species with investigated EMBL accessions and chromosome numbers. The traditional classification shown is as accepted by Hanelt
et al. (1992), or according to description if described later. TAX refers to accession numbers of the Taxonomic Allium Reference Collection of the IPK, Gatersleben, Germany. Herbarium
specimens are marked with ‘‘H’’ and herbarium acronym. Origin is given by country and locality or botanical garden (BG). If origin of BG accession is known it is noted in brackets [ ].
Nucleotide Sequence Database accession numbers of the ITS sequences were given under EMBL.
New classification
Subgenus/Section
Traditional classification
Subgenus/Section Species 2nTAX Origin EMBL
Allium Allium
Allium/Allium Allium/Allium A. ampeloprasum L. 32 2198 BULGARIA: Arkutino AJ411888
Allium/Allium Allium/Allium A. atroviolaceum Boiss. 32 5376 TURKEY: vil. Antalya, Saklikent AJ411884
Allium/Allium Allium/Allium A. dregeanum Kunth 64 5722 SOUTH AFRICA AJ411962
Allium/Allium Allium/Allium A. iranicum (Wendelbo) Wen-
delbo
32 3969 IRAN: Asara, Elburz Range, Karaj Valley AJ411961
Allium/Allium Allium/Allium A. sativum L. 16 1319 TAJIKISTAN: BG Dushanbe AJ411901
Allium/Allium Allium/Allium A. scorodoprasum L. 16 2290 BULGARIA: Plovdiv AJ412713
Allium/Allium Allium/Allium A. sphaerocephalon L. 16 5337 TURKEY: vil. Denizli, Boz Dagi AJ412717
Allium/Mediasia Allium/Allium A. turkestanicum Regel 16 3366 UZBEKISTAN: Malguzar Mts., Dzhizakh AJ411968
A. turkestanicum Regel 16 5049 KAZAKHSTAN: Karatau Range, Mt. Ul-
ken-Burul
AJ412718
Allium/Avulsea Allium/Scorodon A. griffithianum Boiss. 16 3660 UZBEKISTAN: Zaravshan Range, Pass
Takhta-Karachi
AJ411862
Allium/Avulsea Allium/Scorodon A. umbilicatum Boiss. 16 2646 IRAN: BG Tehran [Mosha] AJ412719
Allium/Brevidentia Allium/Brevidentia A. brevidens Vved. 16 5037 UZBEKISTAN: vil. Zevar, SW Hissar
Range
AJ412721
Allium/? (perhaps a new
section)
Allium/Brevidentia A. haneltii Khassanov et R. M.
Fritsch
5796 UZBEKISTAN: Chatkal Range, Kuram
Mts., Chorkesar
AJ412725
Allium/Brevispatha Allium/Brevispatha A. margaritae B. Fedtsch. H658 KAZAKHSTAN: Chu-Ili Mts., Kurdai Pass
(GAT)
AJ412732
Allium/Brevispatha Allium/Brevispatha A. cupanii Raf. subsp. hirtova-
ginatum (Kunth) Stearn
16 5336 TURKEY: vil. Denizli, Boz Dagi AJ412737
Allium/Caerulea Allium/Scorodon A. caeruleum Pall. 16 1525 RUSSIA: BG Moscow AJ411903
A. caeruleum Pall. 16 3735 KAZAKHSTAN: Chu-Ili Mts., Kurdai Pass AJ412729
Allium/Caerulea Allium/Scorodon A. caesium Schrenk 32 2561 TAJIKISTAN: Turkestan Range, Kusavlisai AJ412731
Allium/Caerulea Allium/Scorodon A. elegans Drob. 16 3688 KAZAKHSTAN: Karatau Range, Mt. Ul-
ken-tura
AJ412730
Allium/Caerulea Allium/Scorodon A. litvinovii Drob. ex Vved. 16 3339 KIRGIZSTAN: Alai Range, Pass Chigirchik AJ412727
Allium/Codonoprasum Allium/Codonoprasum A. flavum L. 16 3230 AUSTRIA: BG Linz AJ411926
Allium/Codonoprasum Allium/Codonoprasum A. kunthianum Vved. 16 2579 GEORGIA: Tbilisi, Narikala AJ412734
Allium/Codonoprasum Allium/Codonoprasum A. melanantherum Pancˇic´ 3092 UK: BG Reading AJ412739
Allium/Codonoprasum Allium/Codonoprasum A. paniculatum L. 16 3828 ITALY: Sardinia, La Corte AJ411949
Allium/Codonoprasum Allium/Codonoprasum A. rupestre Steven 5302 TURKEY: vil. Kastamonu, Ayli-Dagi Mts. AJ412733
Allium/Costulatae Allium/Allium A. filidens Regel 16 3674 KAZAKHSTAN: W foothills of Karatau
Range
AJ412723
Allium/Costulatae Allium/Allium A. filidentiforme Vved. ex
Kasht. et Nikitin
16 2573 TAJIKISTAN: Turkestan Range, Shakhris-
tan Pass
AJ412722
Allium/Crystallina Allium/Allium A. crystallinum Vved. 3662 UZBEKISTAN: SW Hissar Range, Derbent AJ412724
Allium/Kopetdagia Allium/Scorodon A. eremoprasum Vved. 16 5243 UZBEKISTAN: Aktau Mts. AJ412726
Allium/Kopetdagia Allium/Scorodon A. kopetdagense Vved. 16 5802 TURKMENISTAN: Kopetdag Mts. AJ411950
Allium/Minuta Allium/Scorodon A. parvulum Vved. 5055 KIRGIZSTAN: Talas-Alatau Range, Talas AJ412720
Allium/Pallasia Allium/Scorodon A. schoenoprasoides Regel 16 2903 KAZAKHSTAN: Trans-Ili Range, Valley
Almaatinka
AJ412728
Allium/Pallasia Allium/Scorodon A. tanguticum Regel 16 3779 CHINA: Tibet AJ411893
Polyprason/Scorodon Allium/Scorodon s.s. A. moschatum L. 16 H654 GEORGIA: Tbilissi (GAT) AJ411872

VOLUME 22 375Phylogeny of Allium
Table 1. Continued.
New classification
Subgenus/Section
Traditional classification
Subgenus/Section Species 2nTAX Origin EMBL
Polyprason/Scorodon Allium/Scorodon A. inaequale Janka 5752 S RUSSIA: Krasnodar, W Caucasus AJ412735
Polyprason/Scorodon Allium/Scorodon A. pamiricum Wendelbo H657 PAKISTAN: Karakorum (GAT) AJ412736
Allium/? (perhaps a new
section)
Allium/Scorodon s.l. A. macrostemon Bunge 16 2369 MONGOLIA: Somon Matad AJ412738
Vvedenskya/Vvedenskya Allium/Vvedenskya A. kujukense Vved. 20 3625 KAZAKHSTAN: Karatau Range, Pass
Kuyuk
AJ411947
Amerallium Amerallium
Amerallium/Amerallium Amerallium/Amerallium A. drummondii Regel 14 0200 SWEDEN: BG Uppsala AJ411908
Amerallium/Arctoprasum Amerallium/Arctoprasum A. ursinum L. 14 1350 GERMANY: Thuringia, Keula AJ412744
Amerallium/Briseis Amerallium/Briseis A. paradoxum (M. Bieb.) G.
Don
16 1657 GERMANY: BG Leipzig AJ412741
Amerallium/Briseis Amerallium/Briseis A. triquetrum L. 16 3269 N TUNESIA: Tell-Atlas Range AJ412742
Amerallium/Bromatorrhiza Bromatorrhiza/Bromatorrhiza A. hookeri Thwaites 22 2013 CHINA: Prov. Yunnan, Kunming, house
garden
AJ250297
Amerallium/Bromatorrhiza Bromatorrhiza/Bromatorrhiza A. wallichii Kunth 14 2441 GERMANY: Gesellschaft der Stauden-
freunde
AJ250294
Amerallium/Caulorhizideum Amerallium/Caulorhizideum A. brevistylum S. Watson 14 2811 SWEDEN: BG Go¨ teborg [USA: Utah] AJ412763
Amerallium/Caulorhizideum Amerallium/Caulorhizideum A. goodingii Ownbey 14 3471 USA: Pepperell, M. McDonough [Arizona] AJ411930
Amerallium/Lophioprason Amerallium/Lophioprason A. cernuum Roth 14 0682 GERMANY: ‘Wyoming Pink’, Gesellschaft
der Staudenfreunde
AJ250289
Microscordum/Microscordum Amerallium/Microscordum A. monanthum Maxim. 32 5617 RUSSIA: Far East, Khabarovsk, Ussuri
lowland
AJ411943
A. monanthum Maxim. 32 5618 RUSSIA: Far East, Vladivostok AJ412745
Amerallium/Molium Amerallium/Molium A. moly L. 14 1117 GERMANY: Gatersleben, house garden AJ412743
Amerallium/Molium Amerallium/Molium A. subhirsutum L. 14 0023 ITALY: BG Palermo [Adiacenze di Petralia] AJ411912
Amerallium/Narkissoprason Amerallium/Narkissoprason A. insubricum Boiss. et Reut. 14 0230 GERMANY: BG Marburg AJ250291
Nectaroscordum/Nectaro-
scordum
Nectaroscordum A. bulgaricum (Janka) Prodan 18 3220 UK: Wokingham, R. Dadd AJ412747
Nectaroscordum/Nectaro-
scordum
Nectaroscordum A. siculum Ucria 18 2192 BULGARIA: Distr. Burgas, Nos Emine AJ250299
Amerallium/Rhopetoprason Amerallium/Rhopetoprason A. fimbriatum S. Watson var.
purdyi Eastw.
3487 USA: D. McNeal, Stockton [California,
Lake Co.]
AJ411932
Amerallium/Rhopetoprason Amerallium/Rhopetoprason A. glandulosum Link. et Otto 28 3045 FRANCE: Montpellier, C.M. Messiaen AJ412746
Caloscordum/Caloscordum Caloscordum/Caloscordum A. neriniflorum (Herb.) Baker 16 2379 EAST MONGOLIA: Somon Chalchgol AJ411913
A. neriniflorum (Herb.) Baker 16 2797 RUSSIA: BG Novosibirsk [Dauria, Onon
River]
AJ411916
Melanocrommyum Melanocrommyum
Melanocrommyum/Acmo-
petala
Melanocrommyum/Acmo-
petala
A. backhousianum Regel 16 1502 TAJIKISTAN: BG Khorog AJ411964
Melanocrommyum/Acmo-
petala
Melanocrommyum/Acmo-
petala
A. suworowii Regel 16 3652 KAZAKHSTAN: Chu-Ili Mts., Georgievka AJ411934
Melanocrommyum/Aroidea Melanocrommyum/Thauma-
sioprason
A. aroides Popov et Vved. 16 2517 UZBEKISTAN: BG Tashkent AJ411915
Melanocrommyum/Brevicaule Melanocrommyum/Miniprason A. sergii Vved. 16 3680 KAZAKHSTAN: Karatau Range, Valley
Ikonzu
AJ411936
Melanocrommyum/Compacto-
prason
Melanocrommyum/Compacto-
prason
A. komarowii Lipsky 16 3133 TAJIKISTAN: Saravshan Range, Lake Is-
kanderkul
AJ411967
Melanocrommyum/Kalo-
prason
Melanocrommyum/Kalo-
prason
A. cristophii Trautv. 16 1920 UK: BG Manchester AJ411966
Melanocrommyum/Megalo-
prason
Melanocrommyum/Megalo-
prason
A. sarawschanicum Regel 16 3673 UZBEKISTAN: Zaravshan Range, Pass
Takhta-Karachi
AJ411935

376 ALISOFriesen, Fritsch, and Blattner
Table 1. Continued.
New classification
Subgenus/Section
Traditional classification
Subgenus/Section Species 2nTAX Origin EMBL
Melanocrommyum/Megalo-
prason
Melanocrommyum/Megalo-
prason
A. stipitatum Regel 16 2257 TAJIKISTAN: Hissar Range, Romit Reser-
vation
AJ411911
Melanocrommyum/Melano-
crommyum
Melanocrommyum/Melano-
crommyum
A. nigrum L. 16 1650 GERMANY: BG Leipzig AJ411965
Melanocrommyum/Melano-
crommyum
Melanocrommyum/Melano-
crommyum
A. noe¨anum Reut. ex Regel 16 3940 IRAN: Prov. Azerbaijan, Miyaneh AJ411970
Melanocrommyum/Miniprason Melanocrommyum/Miniprason A. karataviense Regel 18 2989 UZBEKISTAN: Chatkal Range, Chilchen-
boa Mts.
AJ411922
Melanocrommyum/Popovia Melanocrommyum/Melano-
crommyum
A. gypsaceum Popov et Vved. 16 3661 UZBEKISTAN: SW Hissar Range, Dekh-
kanabad
AJ411969
Porphyroprason/Porphyro-
prason
Melanocrommyum/Porphyro-
prason
A. oreophilum C. A. Mey. 16 0348 AUSTRIA: BG Graz AJ411931
Melanocrommyum/Pseudo-
prason
Melanocrommyum/Pseudo-
prason
A. cardiostemon Fisch. et C. A.
Mey.
16 3947 NW IRAN: Lake Orumiye, Ghushch AJ411971
Melanocrommyum/Regelo-
prason
Melanocrommyum/Regelo-
prason
A. regelii Trautv. 16 5255 TURKMENISTAN: Central Kopetdag,
Gyaurs
AJ411972
Melanocrommyum/Verticillata Melanocrommyum/Verticillata A. verticillatum Regel 16 2182 TAJIKISTAN: Gazimajlik Range, Ganjino AJ411910
Rhizirideum
Anguinum/Anguinum Rhizirideum/Anguinum A. microdictyon Prokh. 16 1300 RUSSIA: BG Kirovsk [Altai] AJ411859
Anguinum/Anguinum Rhizirideum/Anguinum A. ovalifolium Hand.-Mazz. 16 5092 CHINA: Prov. Quinghai, Beisha National
Park
AJ411882
Anguinum/Anguinum Rhizirideum/Anguinum A. tricoccum Sol. 16 2582 USA: Glencoe, J. F. Swenson AJ411917
Anguinum/Anguinum Rhizirideum/Anguinum A. victorialis L. 16 2673 GEORGIA: Caucasus AJ411919
Cepa/Annuloprason Rhizirideum/Annuloprason A. atrosanguineum Kar. et Kir. 16 2912 KAZAKHSTAN: Trans-Ili Range, Valley
Almaatinka
AJ411864
Cepa/Annuloprason Rhizirideum/Annuloprason A. fedschenkoanum Regel 16 2560 TAJIKISTAN: Turkestan Range, Kusavlisai AJ411916
A. fedschenkoanum Regel H618 KIRGYZSTAN: Alai Mts. (GAT) AJ411894
Cepa/Annuloprason Rhizirideum/Annuloprason A. monadelphum Turcz. ex Kar.
et Kir.
16 H643 RUSSIA: East Sibiria, Chentei (NSK) AJ411955
Cepa/Annuloprason Rhizirideum/Annuloprason A. semenowii Regel H628 KIRGIZSTAN: Centr. Than-Schan (TASH) AJ411897
Cepa/Annuloprason Rhizirideum/Annuloprason A. weschniakowii Regel H641 KIRGIZSTAN: Issyk-kul (TASH) AJ411946
Butomissa/Butomissa Rhizirideum/Butomissa A. ramosum L. 32 2735 KAZAKHSTAN: BG Alma-Ata AJ250295
A. ramosum L. 32 2755 RUSSIA: Bur yatia, Lake Gusinoe AJ250293
Butomissa/Butomissa Rhizirideum/Butomissa A. tuberosum Rottl. ex Spreng. 32 2454 INDIA: Kashmir, Ladakh AJ411914
Rhizirideum/Caespitosoprason Rhizirideum/Caespitosoprason A. bidentatum Fisch. ex Pro-
khorov
32 2365 MONGOLIA: Somon Erdenetsagan AJ411861
A. bidentatum Fisch. ex Pro-
khorov
32 3163 RUSSIA: Altai Mts., Czemal AJ311868
Rhizirideum/Caespitosoprason Rhizirideum/Caespitosoprason A. dentigerum Prokhorov 16 5091 CHINA: Prov. Quinghai, Xining AJ411958
Rhizirideum/Caespitosoprason Rhizirideum/Caespitosoprason A. mongolicum Regel 16 2373 MONGOLIA: Tamsarbulag AJ411883
Rhizirideum/Caespitosoprason Rhizirideum/Caespitosoprason A. polyrrhizum Turcz. ex Regel 32 2359 MONGOLIA: Aimak centre Baruun urt,
Talbulagu
AJ250296
Rhizirideum/Caespitosoprason Rhizirideum/Caespitosoprason A. subangulatum Regel H604 MONGOLIA: Gobi Altai Mts. (GAT) AJ411870
Reticulatobulbosa/Campanu-
lata
Rhizirideum/Campanulata A. drepanophyllum Vved. 16 2791 UZBEKISTAN: Chatkal Range, Bashky-
sylsai
AJ411854
Reticulatobulbosa/Campanu-
lata
Rhizirideum/Campanulata A. jodanthum Vved. 16 1330 TAJIKISTAN: Hissar Range, Kondara Val-
ley
AJ411902
Reticulatobulbosa/Campanu-
lata
Rhizirideum/Campanulata A. minorensis ined. H679 KIRGIZSTAN: Talas Alatau Range (GAT) AJ412748
Reticulatobulbosa/Campanu-
lata
Rhizirideum/Campanulata A. sordidiflorum Vved. H630 TAJIKISTAN: Sanglok Mts. (TASH) AJ411899

VOLUME 22 377Phylogeny of Allium
Table 1. Continued.
New classification
Subgenus/Section
Traditional classification
Subgenus/Section Species 2nTAX Origin EMBL
Reticulatobulbosa/Campanu-
lata
Rhizirideum/Campanulata A. tenuicaule Regel H632 TAJIKISTAN: Hissar Range (TASH) AJ411887
Reticulatobulbosa/Campanu-
lata
Rhizirideum/Campanulata A. teretifolium Regel H633 KAZAKHSTAN: Dzhungar Alatau Range
(TASH)
AJ411886
Reticulatobulbosa/Campanu-
lata
Rhizirideum/Campanulata A. xiphopetalum Aitch. et
Baker
16 1070 SWEDEN: BG Go¨ teborg AJ411858
Cepa/Cepa Rhizirideum/Cepa A. altaicum Pall. 16 0339 LITHUANIA: BG Kaunas AJ411928
A. altaicum Pall. 16 2760 RUSSIA: Sayan Mts, Tuva, Erzin AJ412749
A. altaicum Pall. 16 5561 RUSSIA: Altai Mts., Yuzhno-Chuisky
Range
AJ412750
Cepa/Cepa Rhizirideum/Cepa A. asarense R. M. Fritsch et
Matin
16 3900 IRAN: Elburz Range, Karaj Valley, Asara AJ411937
Cepa/Cepa Rhizirideum/Cepa A. cepa L. 16 4878 GERMANY: ‘Stuttgarter Riesen’ AJ411944
Cepa/Cepa Rhizirideum/Cepa A. cepa aggregatum cultivar
group
16 1810 USA: Glencoe, J. F. Swenson AJ411906
Cepa/Cepa Rhizirideum/Cepa A. fistulosum L. 16 41388 JAPAN: ‘Long White Tokyo’ AJ411918
Cepa/Cepa Rhizirideum/Cepa A. galanthum Kar. et Kir. 16 1729 KAZAKHSTAN: BG Alma-Ata AJ411905
Cepa/Cepa Rhizirideum/Cepa A. oschaninii B. Fedtsch. 16 5016 KIRGIZSTAN: Turkestan Range, Dargun AJ411940
Cepa/Cepa Rhizirideum/Cepa A. praemixtum Vved. 16 5712 TAJIKISTAN: Mogoltau Range, Spa massif AJ411873
Cepa/Cepa Rhizirideum/Cepa A. pskemense B. Fedtsch. 16 1994 DENMARK: BG Copenhagen AJ411907
Cepa/Cepa Rhizirideum/Cepa A. vavilovii Popov et Vved. 16 5238 TURKMENISTAN: C. Kopetdag Range,
Chuli
AJ411839
Cepa/Cepa Rhizirideum/Cepa A. vavilovii Popov et Vved. 16 5239 TURKMENISTAN: C. Kopetdag Range,
Chuli
AJ411840
Cyathophora/Coleoblastus Bromatorrhiza/Coleoblastus A. mairei Levl. 16 2104 SWITZERLAND: BG Zurich AJ250298
Cyathophora/Cyathophora Bromatorrhiza/Cyathophora A. cyathophorum Bur. et
Franch.
16 2824 NORWAY: BG Oslo AJ250286
A. cyathophorum Bur. et
Franch.
16 2825 NORWAY: BG Oslo AJ411889
Cyathophora/Milula Milula A. spicatum (Prain) N. Friesen 16 H680 CHINA: Tibet, Tsangpo Valley (OSBU) AJ250285
A. spicatum (Prain) N. Friesen 16 H682 CHINA: Tibet, Upper Cuamda Chu (OSBU) AJ250284
Polyprason/Falcatifolia Rhizirideum/Oreiprason A. carolinianum DC. 16 2570 TAJIKISTAN: Hissar Range, Anzob Pass AJ250290
Cepa/Condensatum Rhizirideum/Oreiprason A. condensatum Turcz. 16 5619 RUSSIA: Far East, Khabarovsk, Ussuri
lowland
AJ412752
A. condensatum Turcz. 16 H663 EAST MONGOLIA: (GAT) AJ412752
Polyprason/Daghestanica Rhizirideum/Oreiprason A. daghestanicum Grossh. 16 1741 GEORGIA: BG Bakuriani, [Daghestan] AJ411850
Polyprason/Daghestanica Rhizirideum/Oreiprason A. ericetorum Thore 16 1547 GERMANY: BG Leipzig AJ311867
Polyprason/Daghestanica Rhizirideum/Oreiprason A. gunibicum Miscz. ex
Grossh.
16 2333 GEORGIA: Caucasus, Chevsuretia, Dusheti AJ411890
Butomissa/Austromontana Rhizirideum/Oreiprason A. gilgiticum Wang et Tang H600 PAKISTAN: Karakorum Range (GAT) AJ411942
Polyprason/Falcatifolia Rhizirideum/Oreiprason A. hymenorrhizum Ledeb. 16 3135 TAJIKISTAN: Saravshan Range, Lake Is-
kanderkul
AJ411879
Polyprason/Oreiprason Rhizirideum/Oreiprason A. kaschianum Regel 16 2418 TAJIKISTAN: BG Chorog [Pamir, Shakh-
darya Mts.]
AJ412754
Polyprason/Oreiprason Rhizirideum/Oreiprason A. kurssanovii Popov 16 1612 KAZAKHSTAN: Trans-Ili Range, Medeo
Rhizirideum/Oreiprason Rhizirideum/Oreiprason A. ochroleucum Waldst. et Kit. 16 3763 SLOVENIA: Steiner Alps AJ411856
A. ochroleucum Waldst. et Kit. 5578 SLOVENIA: BG Ljubljana [Julijske Alps] AJ412755
Polyprason/Daghestanica Rhizirideum/Oreiprason A. ochroleucum subsp. pseu-
dosuaveolens Zahar.
16 2827 ROMANIA: BG Cluj-Napoca AJ411863
Polyprason/Oreiprason Rhizirideum/Oreiprason A. petraeum Kar. et Kir. 16 3040 KAZAKHSTAN: Chu-Ili Mts., Pass Kurdai AJ411952

378 ALISOFriesen, Fritsch, and Blattner
Table 1. Continued.
New classification
Subgenus/Section
Traditional classification
Subgenus/Section Species 2nTAX Origin EMBL
Polyprason/Falcatifolia Rhizirideum/Oreiprason A. platyspathum Schrenk 16 2905 KAZAKHSTAN: Trans-Ili Range, Valley
Almaatinka
AJ411878
Polyprason/Falcatifolia Rhizirideum/Oreiprason A. platyspathum subsp. ambly-
ophyllum (Kar. et Kir.) N.
Friesen
16 2396 RUSSIA: BG Novosibirsk AJ411875
Polyprason/Oreiprason Rhizirideum/Oreiprason A. roylei Stearn 16 5152 CZECH REPUBLIK: Olomouc AJ411945
Polyprason/Oreiprason Rhizirideum/Oreiprason A. setifolium Schrenk H629 UZBEKISTAN: (GAT) AJ411898
Polyprason/Daghestanica Rhizirideum/Oreiprason A. suaveolens Jacq. 16 1603 ITALY: BG Padua AJ411874
Polyprason/Oreiprason Rhizirideum/Oreiprason A. talassicum Regel 16 3376 CHINA: Prov. Xingiang AJ411865
Polyprason/Oreiprason Rhizirideum/Petroprason A. obliquum L. 16 3158 RUSSIA: Altai Mts., Lake Teletskoe AJ412753
Reticulatobulbosa/Sikkimensia Rhizirideum/Reticulatobulbosa A. beesianum W. W. Sm. 32 2211 DENMARK: BG Copenhagen, AJ411860
Reticulatobulbosa/Reticulato-
bulbosa
Rhizirideum/Reticulatobulbosa A. chamarense M. Ivanova 16 H645 MONGOLIA: Chentei (GAT) AJ411957
Reticulatobulbosa/Reticulato-
bulbosa
Rhizirideum/Reticulatobulbosa A. clathratum Ledeb. 32 3165 RUSSIA: Altai Mts., Chemal AJ411855
Reticulatobulbosa/Sikkimensia Rhizirideum/Reticulatobulbosa A. cyaneum Regel 32 3872 GERMANY: BG Tuebingen AJ411880
Reticulatobulbosa/Nigrimon-
tana
Rhizirideum/Reticulatobulbosa A. drobovii Vved. H622 UZBEKISTAN: Ugamskij Chr. (TASH) AJ411895
Rhizirideum/Eduardia Rhizirideum/Reticulatobulbosa A. eduardii Stearn 16 2761 RUSSIA: Sayan Range, Tuva, Erzin AJ411959
A. eduardii Stearn 16 2745 RUSSIA: Sayan Range, Tuva, Erzin AJ412756
Reticulatobulbosa/Reticulato-
bulbosa
Rhizirideum/Reticulatobulbosa A. flavidum Ledeb. 16 H644 RUSSIA: Altai (GAT) AJ411956
Cyathophora/Coleoblastus Rhizirideum/Reticulatobulbosa A. kingdonii Stearn H691 CHINA: Tibet (OSBU) AJ250298
Reticulatobulbosa/Reticulato-
bulbosa
Rhizirideum/Reticulatobulbosa A. kuramense F. O. Khassanov
et N. Friesen
16 5703 UZBEKISTAN: Chatkal Range, Kuram
Mts., Chorkesar
AJ411868
Reticulatobulbosa/Reticulato-
bulbosa
Rhizirideum/Reticulatobulbosa A. leucocephalum Turcz. 16 H660 RUSSIA: Burjatia, Dzhida (NSK) AJ412757
Reticulatobulbosa/Reticulato-
bulbosa
Rhizirideum/Reticulatobulbosa A. lineare L. 16 5816 N KAZAKHSTAN: vil. Kievskoe AJ411951
A. lineare L. 16 H624 KAZAKHSTAN: (GAT) AJ411951
Reticulatobulbosa/Reticulato-
bulbosa
Rhizirideum/Reticulatobulbosa A. malyschevii N. Friesen 16 H653 RUSSIA: Tuva, Sengilen (GAT) AJ412758
Rhizirideum/Reticulatobulbosa Rhizirideum/Reticulatobulbosa A. montibaicalense N. Friesen 16 3390 RUSSIA: Buryatia, NW Baikal, Muzhinai AJ411838
A. montibaicalense N. Friesen 16 H603 RUSSIA: North Baikal (GAT) AJ411871
Reticulatobulbosa/Nigrimon-
tana
Rhizirideum/Reticulatobulbosa A. oreoprasoides Vved. H627 UZBEKISTAN: West Than-Shan (TASH) AJ41189
Butomissa/Austromontana Rhizirideum/Reticulatobulbosa A. oreoprasum Schrenk 16 3643 KAZAKHSTAN: Trans-Ili Range, Turgen
Valley
AJ411867
A. oreoprasum Schrenk 16 5000 KIRGIZSTAN: Talas-Alatau Range, Pass
Otmek
AJ411933
Rhizirideum/Caespitosoprason Rhizirideum/Reticulatobulbosa A. przewalskianum Regel 32 5089 CHINA: Prov. Quinghai, Lake Koko Nor AJ411852
Reticulatobulbosa/Reticulato-
bulbosa
Rhizirideum/Reticulatobulbosa A. rupestristepposum N.
Friesen
16 5737 RUSSIA: Lake Baikal, Goloustnoe AJ411869
Reticulatobulbosa/Scabri-
scapa
Rhizirideum/Reticulatobulbosa A. scabriscapum Boiss. 16 3898 IRAN: Prov. Tehran, Merdabat AJ411881
Reticulatobulbosa/Sikkimensia Rhizirideum/Reticulatobulbosa A. sikkimense Baker H670 CHINA: SE Tibet, Lhasa (GAT) AJ411885
Reticulatobulbosa/Reticulato-
bulbosa
Rhizirideum/Reticulatobulbosa A. splendens Schult. et
Schult. f.
48 3384 RUSSIA: Buryatia, NW shore of Lake Bai-
kal
AJ411927
Reticulatobulbosa/Reticulato-
bulbosa
Rhizirideum/Reticulatobulbosa A. strictum Schrad. 48 5404 RUSSIA: upper Enisey River, Shushenskoe AJ411951
Reticulatobulbosa/Flavi-
scordum
Rhizirideum/Reticulatobulbosa A. sulphureum Vved. H648 TAJIKISTAN: Pamiro-Alai, Gasimaylik
(LE)
AJ412759

VOLUME 22 379Phylogeny of Allium
Table 1. Continued.
New classification
Subgenus/Section
Traditional classification
Subgenus/Section Species 2nTAX Origin EMBL
Reticulatobulbosa/Flavi-
scordum
Rhizirideum/Reticulatobulbosa A. trachyscordum Vved. 16 3998 KIRGIZSTAN: Talas, Talas-Alatau Range AJ411857
Reticulatobulbosa/Reticulato-
bulbosa
Rhizirideum/Reticulatobulbosa A. ubsicolum Regel 32 2769 RUSSIA: Tuva, Erzin, Sayan Range AJ411960
Rhizirideum/Rhizirideum Rhizirideum/Rhizirideum A. albidum Fisch. ex Bieb. 16 3765 AUSTRIA: BG Graz [Serbia, Deliblatska
Pescara]
AJ411841
Rhizirideum/Rhizirideum Rhizirideum/Rhizirideum A. albidum Fisch. ex Bieb. 16 1977 GEORGIA: Caucasus Range, Chevsuretia,
Mutso
AJ411892
A. albidum Fisch. ex Bieb. 16 3470 USA: Pepperell, M. McDonough [NE Tur-
key]
AJ411954
Rhizirideum/Rhizirideum Rhizirideum/Rhizirideum A. angulosum L. 16 2778 N KAZAKHSTAN: vil. Kievskoe AJ250287
Rhizirideum/Rhizirideum Rhizirideum/Rhizirideum A. austrosibiricum N. Friesen 16 2747 RUSSIA: Tuva, Ersin, Sayan Range AJ411832
Rhizirideum/Rhizirideum Rhizirideum/Rhizirideum A. flavescens Bess. 16 5736 N KAZAKHSTAN: vil. Kievskoe AJ411842
Rhizirideum/Rhizirideum Rhizirideum/Rhizirideum A. incensiodorum Radic´ 16 3764 AUSTRIA: BG Graz AJ411866
Rhizirideum/Rhizirideum Rhizirideum/Rhizirideum A. lusitanicum Lam. 32 2927 GERMANY: Harz, Benzingerode AJ411831
Rhizirideum/Rhizirideum Rhizirideum/Rhizirideum A. nutans L. 32 3161 RUSSIA: Gorno-Altaisk, Altai Mts. AJ411924
Rhizirideum/Rhizirideum Rhizirideum/Rhizirideum A. rubens Schrad. 16 1609 KAZAKHSTAN: Temirtau AJ411904
A. rubens Schrad. 16 3401 RUSSIA: Altai Mts., Sailugem AJ411891
Rhizirideum/Rhizirideum Rhizirideum/Rhizirideum A. senescens L. 32 2750 RUSSIA: Bur yatia, Lake Gusinoe AJ411834
Rhizirideum/Rhizirideum Rhizirideum/Rhizirideum A. spirale Willd. 32 1968 NORTH KOREA: BG Wonsan [Prov.
S-Hwanghe]
AJ411833
Rhizirideum/Rhizirideum Rhizirideum/Rhizirideum A. stellerianum Willd. 32 5738 RUSSIA: Muchor, Lake Baikal AJ411963
Rhizirideum/Rhizirideum Rhizirideum/Rhizirideum A. togashii Hara 16 5221 USA: Pepperell, M. McDonough AJ411843
Rhizirideum/Rhizirideum Rhizirideum/Rhizirideum A. tuvinicum (N. Friesen) N.
Friesen
16 2744 RUSSIA: Tuva, Erzin, Sayan Range AJ411853
Cepa/Sacculiferum Rhizirideum/Sacculiferum A. chinense G. Don 24 2015 CHINA: Prov. Yunnan, Kunming, market AJ411848
A. chinense G. Don 32 3407 JAPAN: Fukui, local strain AJ411948
Cepa/Sacculiferum Rhizirideum/Sacculiferum A. komarovianum Vved. H646 RUSSIA: Far East (LE) AJ412760
Cepa/Sacculiferum Rhizirideum/Sacculiferum A. thunbergii G. Don 16 1963 NORTH KOREA: BG Wonsan [Kumgang-
san Range, Prov. Kanwon]
AJ411849
Cepa/Schoenoprasum Rhizirideum/Schoenoprasum A. altyncolicum N. Friesen 32 0433 GERMANY: Quedlinburg, breeder’s strain AJ411939
Cepa/Schoenoprasum Rhizirideum/Schoenoprasum A. karelinii Poljak. 16 2592 JAPAN: Tsukuba Research Station AJ411876
Cepa/Schoenoprasum Rhizirideum/Schoenoprasum A. ledebourianum Schult. et
Schult. f.
16 3170 RUSSIA: Altai Mts., Cholzun AJ411925
Cepa/Schoenoprasum Rhizirideum/Schoenoprasum A. maximowiczii Regel 16 2772 RUSSIA: Buryatia, Romanovka AJ411877
Cepa/Schoenoprasum Rhizirideum/Schoenoprasum A. oliganthum Kar. et Kir. 16 3201 KAZAKHSTAN: Lake Zajsan AJ411835
Cepa/Schoenoprasum Rhizirideum/Schoenoprasum A. schmitzii Coutinho 16 5226 USA: Pepperell, M. McDonough AJ412761
Cepa/Schoenoprasum Rhizirideum/Schoenoprasum A. schoenoprasum subsp. la-
tiorifolium (Pau) R. Martinez
et al.
32 5432 SPAIN: Sierra de Guadarrama, Mt. Penarala AJ411837
Cepa/Schoenoprasum Rhizirideum/Schoenoprasum A. schoenoprasum L. 16 3446 RUSSIA: Far East, Peninsula Chukotka AJ411836
A. schoenoprasum L. 16 4214 GERMANY: ‘Mittelgrober’ AJ411938
Rhizirideum/Tenuissima Rhizirideum/Tenuissima A. anisopodium Ledeb. 16 2349 MONGOLIA: Aimak Chentij, Zargalant-
chan
AJ411847
Rhizirideum/Tenuissima Rhizirideum/Tenuissima A. tenuissimum L. 16 3249 NORTH KOREA: BG Pyongyang AJ411846
Rhizirideum/Tenuissima Rhizirideum/Tenuissima A. vodopjanovae N. Friesen 16 5402 RUSSIA: W Sayan Range, Khakasia AJ411942
A. vodopjanovae N. Friesen 16 3192 RUSSIA: Altai Mts., Ust-Koksa, Kyrlyk AJ411845
Ipheion Ipheion I. uniflorum (Graham) Raf. 12 3458 NETHERLAND: BG Amsterdam AJ250288
Tulbaghia Tulbaghia T. simmleri Beauv. (
⫽
T. fra-
grans Verdoorn)
12 2165 UK: London, Chelsea Physic Garden AJ250300
Nothoscordum Nothoscordum N. bivalve (L.) Britton 18 2621 UK: Wellesbourne, HIS AJ250300
N. gracile (Ait.) Stearn 18 0494 SWEDEN: BG Go¨teborg AJ250301
Dichelostemma Dichelostemma D. multiflorum (Benth.) Heller 2471 USA: California, BG Berkeley [Shasta Co.] AJ412714

380 ALISOFriesen, Fritsch, and Blattner
vers. 3.06 (Posada and Crandall 1998) and TrN
⫹⌫
dis-
tances were calculated, which were analyzed with the neigh-
bor-joining algorithm (NJ). Two parsimony analysis (MP)
algorithms were used: Fitch parsimony and generalized par-
simony, weighting transversions vs. transitions 2:1 (Swof-
ford et al. 1996). We conducted heuristic searches with sim-
ple and with 100 random addition sequences and tree bisec-
tion reconnection (TBR) branch swapping. For Bayesian in-
ference six chains were run for 1 million generations under
a GTR
⫹⌫
model of sequence evolution, sampling a tree
every 100 generations. The first 2001 (nonstationary) trees
were discarded and 8000 trees were used for the calculation
of posterior probabilities. To allow the calculation of statis-
tical support of the clades with MP a reduced data set was
used, including only representative accession from each
group found in the strict consensus tree of the first analysis.
Thus, 100 accessions were included, which resulted in the
same relative tree topology as in the analyses of the entire
data set. Support was tested with BI and bootstrap analyses
(Felsenstein 1985), involving 1000 bootstrap resamples in
neighbor-joining analysis and 200 resamples in parsimony
analysis. Decay indices (Bremer 1988) were also calculated
for this data set. Dichelostemma multiflorum was in all
analyses defined as outgroup taxon.
RESULTS
ITS Sequences
Within the 195 Allium species the lengths of the ITS re-
gion ranged from 612 base pairs (bp) in A. cyathophorum to
661 bp in A. triquetrum, though for most sequences the
length was 640
⫾
10 bp. The outgroup species had relatively
longer ITS regions than Allium:Ipheion uniflorum 658 bp,
Tulbaghia fragrans 659 bp, Nothoscordum gracile 671 bp,
N. bivalve 672 bp, and Dichelostemma multiflorum 675 bp.
GC contents of the ITS region were between 42–46% in
Allium, and between 47–51% in the outgroup genera. Align-
ing the individual sequences resulted in an alignment of 778
bp length. Of these, 116 characters were constant, 66 vari-
able characters were parsimony uninformative, and 596 were
parsimony informative. In ITS-1, 81% of sequence positions
were variable, in ITS-2 75%, and even in the 5.8S nrDNA,
23% variable positions could be found. The inclusion of out-
groups increased the amount of variable sites only by about
5%. Pairwise genetic distances were between 1% (among
species from one section) to 53% between A. haneltii (sub-
gen. Allium) and A. bulgaricum (subgen. Nectaroscordum).
Phylogenetic Analyses
Generalized parsimony analysis of the ITS data produced
78,300 equally parsimonious trees (length
⫽
5525 steps, in-
cluding parsimony uninformative characters, consistency in-
dex (CI) 0.2878, retention index (RI) 0.8036, when all 221
accessions were included. The consensus tree was generally
congruent with the result of Fitch parsimony and is available
from the authors on demand. Analyses of a reduced data set
(Fig. 1), comprising 100 representative taxa from all clades
found in the initial analyses resulted in 20 MP trees (length
⫽
4110 steps, CI
⫽
0.3348, RI
⫽
0.6940). The strict con-
sensus trees of the parsimony analyses were also congruent
with the results of model-based and phenetic analyses of the
respective data sets. Statistical support of the branches from
bootstrap (BS) and decay analyses (D), and BI posterior
probabilities (PP) were given for the reduced data set in Fig.
1, 2.
Allium is a monophyletic group, clearly separated from
the outgroup species (100% BS and PP, d
⬎
10). Two major
clades were found within Allium, comprising subgen. Nec-
taroscordum (x
⫽
9), Amerallium (x
⫽
7) together with sect.
Microscordum (x
⫽
8, A. monanthum) (Fig. 1: node 1) on
one side and all other Allium taxa on the other side (all x
⫽
8). This latter group is divided in two sister clades, subgen.
Caloscordum, Melanocrommyum and sect. Anguinum (Fig.
1: node 2) and subgen. Allium and Rhizirideum (Fig. 1: node
3). The species of former subgen. Bromatorrhiza occur at
two different positions in the tree: sect. Bromatorrhiza (A.
wallichii and A. hookeri, x
⫽
7) are clearly placed in subgen.
Amerallium and species of sects. Coleoblastus (A. kingdonii
and A. mairei, x
⫽
8) and Cyathophora (A. cyathophorum,
x
⫽
8) fall in the large clade comprising Rhizirideum and
Allium. Subgenera Rhizirideum and Allium are subdivided in
seven monophyletic groups that have different relationships:
sect. Anguinum is sister group of subgen. Melanocrommyum,
sect. Butomissa (including some species from sect. Reticu-
latobulbosa) is sister group to the rest of subgen. Rhiziri-
deum and Allium, sects. Cyathophora, Coeloblastus, and
Milula are sister group to all other sections of subgenera
Rhizirideum and Allium, sects. Rhizirideum, Caespitosopra-
son, Tenuissima, and A. eduardii (sect. Reticulatobulbosa)
are sister group to the remaining sections of subgen. Rhizir-
ideum and Allium, most species from subgen. Allium form a
monophyletic clade, excluding the species from sect. Sco-
rodon s.s. and A. turkestanicum. Allium kujukense (sect.
Vvedenskya) is clearly placed in one clade with subgen. Me-
lanocrommyum. The rest of subgen. Rhizirideum with sect.
Scorodon s.s. and A. turkestanicum form a clade that consists
of well-supported small sections with unresolved interrela-
tionships.
Phylogenetic analysis supported a monophyletic origin of
most circumscribed sections in subgen. Rhizirideum and Al-
lium, with some exceptions. The morphologically variable
sects. Scorodon, Reticulatobulbosa, and Oreiprason were
polyphyletic and some morphologically ‘‘difficult’’ species,
such as A. turkestanicum, A. condensatum, A. eduardii, and
A. przewalskianum showed unexpected relationships outside
their former sections.
Species of sect. Reticulatobulbosa appear in three groups,
(i) Reticulatobulbosa s.s., (ii) a small group of the four spe-
cies (A. trachyscordum, A. scabriscapum, A. sulphureum,
and A. drobovii), which are more closely related to sect.
Campanulata, and (iii) all Chinese-Himalayan species with
blue flowers (A. beesianum, A. forrestii, A. aciphyllum, A.
sikkimense, and A. cyaneum), which form an isolated clade
with unresolved relationships.
Species of sect. Oreiprason are split into four clades: (i)
the east Asian species A. condensatum, (ii)A. obliquum, A.
petraeum, and A. talassicum, (iii) the central Asian group
with A. platyspathum, A. filifolium, A. carolinianum, A. kas-
chianum, and A. hymenorrhizum, and (iv) the Caucasian–
European clade with A. gunibicum, A. ericetorum, A. dagh-
estanicum, A. ochroleucum, and. A. suaveolens.

VOLUME 22 381Phylogeny of Allium
Fig. 1.—Strict consensus tree of 20 most parsimonious trees from the analysis of nrDNA ITS sequences of 100 Allium accessions.
Bootstrap values (
⬎
50%) are given above the branches; BI posterior probabilities are shown below. Values of 100% are depicted by
asterisks. The basic chromosome numbers of the Allium species are shown above the branches.

382 ALISOFriesen, Fritsch, and Blattner
Fig. 2.—Modified consensus tree from Fig. 1 to represent the sects. of Allium instead of the species on the tips of the tree. Bootstrap
values (
⬎
50%) are given below the branches, decay indices above. Bootstrap values higher than 75% are represented by asterisks. The
new and old intrageneric classification of the genus is depicted to the right.

VOLUME 22 383Phylogeny of Allium
Polyploids and Hybrid Taxa
Twenty-eight Allium species (14%) included in the analy-
sis were polyploids (see Table 1). In the worst case in terms
of phylogenetic analysis, these might be allopolyploids re-
sulting from hybridization events, thus combining different
ITS types in the same nucleus. Due to their reticulate struc-
ture, hybrid taxa cannot be placed correctly in a dichotomous
tree and might even disturb parts in the phylogenetic tree
adjacent to the hybrid (Friesen and Blattner 2000). To ana-
lyze the influence of polyploid taxa on our phylogenetic re-
construction, we excluded all polyploid species from the
analysis. This resulted in only small differences within the
terminal clades (i.e., within sections), but did not change the
relationships among these clades. The ITS sequences were
mostly obtained by direct sequencing of PCR products, re-
sulting in a small number of sequence positions with poly-
morphic base composition. Only in some species like A.
hookeri (2n
⫽
22) or A. splendens (2n
⫽
48), where no clear
sequences could be obtained via direct sequencing, the PCR
fragments were cloned prior to sequencing. In these cases
the differences among the cloned sequences from one indi-
vidual were relatively small and all sequences grouped in
the same clade in the phylogenetic tree. As our aim in this
study was to define the major groups in Allium and not phy-
logenetic analyses within sections, we left all polyploid taxa
in the analysis.
Taxonomic Conclusions and Classification
The molecular results strongly suggest that a natural clas-
sification of Allium must recognize many more well sepa-
rated main groups than hitherto were accepted at the sub-
generic level. Additionally, several of the traditional sections
proved to be artificial. Thus, a modern classification that is
in accord with the above given phylogeny must necessarily
introduce new groups and new names. In order to insure
nomenclatural stability, it seemed most appropriate to use as
often as possible already known names although at differing
hierarchical levels. Following the nomenclatural rules of the
International Code of Botanical Nomenclature (ICBN)
(Greuter et al. 2000), we apply some hitherto unused but
valid names. The subgenera are listed according to their po-
sition in the phylogenetic dendrogram (Fig. 2). A nomencla-
tural conspectus of the genus Allium at sectional level is
given as Appendix 1. For representative taxa see Fig. 3–26.
A
LLIUM
subgen. N
ECTAROSCORDUM
(Lindl.) Asch. et Graebn.,
Syn. Mitteleur. Fl. 3: 96, 166 (1905).—TYPE: A. siculum
Ucria (
⫽
N. siculum (Ucria) Lindl.) (Fig. 3).
Genus Nectaroscordum Lindl., Edwards’s Bot. Reg. 9: t. 1913
(1836); subgen. Nectaroscordum (Lindl.) Traub, Pl. Life (Stan-
ford) 24: 162 (1968), comb. superfl.; sect. Nectaroscordum
(Lindl.) Gren. et Godr., Fl. France 3: 212 (1855–56).
At least two species belong to this group. The western
Mediterranean-type species and the southwest Asian A. tri-
pedale Trautv. (Kudrjashova 2003) are certainly distinct
taxa, but there is controversy as to whether A. bulgaricum,
A. dioscorides Sibth. et Sm., and A. meliophilum Juz. (en-
demic of the Crimea) might be only eastern Mediterranean
variants of the type species or good species.
The joint occurrence of several specific characters was the
main argument to regard Nectaroscordum as a genus of its
own: large and 3–7-veined tepals, the long triangular inner-
most sheath leaf surrounding the scape, a wider than long
ovary, the lower half of which is sunken into the funnel-like
widened tip of pedicels, multiovulate locules stretching in a
nearly horizontal direction (Stearn 1955, 1978), heavy seeds
(thousand kernel weight [TKW] of about 6 g) with three
sharp edges, as well as a basic chromosome number of x
⫽
9. Especially large and many-lobed nectaries mounding in a
wide secretory channel near the tip of the nectary are unique.
All these characters imply at least a very long and separate
phylogeny. Apparently only some common characters of leaf
anatomy (Fritsch 1988) support the close relationship to sub-
gen. Amerallium that is shown by molecular analyses.
A
LLIUM
subgen. Microscordum (Maxim.) N. Friesen, comb.
et stat. nov.—TYPE: A. monanthum Maxim.
Basionym: Allium sect. Microscordum Maxim., Bull. Acad. Imp.
Sci. Saint-Pe´tersbourg 31(1): 109 (1886).
This group is monotypic occurring only in eastern Asia.
Morphological characters of bulbs, bulb tunics, leaves and
flowers are similar to species of subgen. Amerallium, but
one- or two-flowered inflorescences and also the feathery
ends of stigmatic lobes are unique within the genus Allium.
The same is true for the occurrence of dioecy, which devel-
oped only at the diploid level (Noda and Kawano 1988).
The plants show also a special chromosome pattern at dif-
ferent ploidy levels based on x
⫽
8.
A
LLIUM
subgen. A
MERALLIUM
Traub, Pl. Life (Stanford) 24:
159 (1968).—TYPE: A. canadense L.
Subgen. Bromatorrhiza Ekberg, Bot. Not. 122: 58 (1969), pro
parte.—TYPE: A. wallichii Kunth; subgen. Molium (G. Don) Wen-
delbo, Bot. Not. 122: 26 (1969).—TYPE: A. neapolitanum L.; sub-
gen. Caulorhizideum (Traub) Kamelin, Florogenet. analiz est. fl.
gorn. Sredn. Azii: 245 (1973).—TYPE: A. validum S. Wats.; subgen.
Chamaeprason (F. Hermann) Garbari, Taxon 24: 541 (1975).—
TYPE: A. chamaemoly L.
Besides the sections mentioned above as synonymous
subgenera, the American sects. Lophioprason (Fig. 4), Amer-
allium, and Rhophetoprason, as well as the Old World sects.
Arctoprasum (Fig. 5), Briseis, Narkissoprason, Molium (Fig.
7) and Bromatorrhiza (Fig. 6) are included. Section Rhyn-
chocarpum possibly also belongs here. Molecular data un-
derline the existence of two sister alliances: a rather diverse
Old World and a derived New World one. Nevertheless, both
are a monophyletic unit, which agrees with a uniform elec-
trophoretic banding pattern of salt-soluble seed storage pro-
teins (Maass 1992). Bulbous and rhizomatous species occur
in both alliances. This may explain why different types of
growth form and shoot foliation exist, but irregular insertion
of cataphylls into the foliage leaf sequence was only ob-
served in subgen. Amerallium (Kruse 1992a). All species
share a special anatomy of vegetative parts (one row of vas-
cular bundles, absence of palisade parenchyma, and subepi-
dermal position of laticifers; Fritsch 1988). Furthermore, the
dominating basic chromosome number of x
⫽
7 strongly
supports its separate status, although x
⫽
8, 9, and 11 also

384 ALISOFriesen, Fritsch, and Blattner
Fig. 3–14.—Plant portraits.—3. Allium siculum (sect. Nectaroscordum).—4. A. amplectens Torr. (sect. Lophioprason).—5. A. ursinum
(sect. Arctoprason).—6. A. hookeri (sect. Bromatorrhiza).—7. A. moly (sect. Molium).—8. A. oreophilum (sect. Porphyroprason).—9. A.
neriniflorum (sect. Caloscordum).—10. A. cupuliferum Regel (sect. Regeloprason).—11. A. nigrum (sect. Melanocrommyum).—12. A.
cristophii (sect. Kaloprason).—13. A. ramosum (sect. Butomissa).—14. A. cyathophorum (sect. Cyathophora).

VOLUME 22 385Phylogeny of Allium
Fig. 15–26.—Plant portraits.—15. Allium przewalskianum (sect. Caespitosoprason).—16. A. mongolicum (sect. Caespitosoprason).—17.
A. tenuissimum (sect. Tenuissima).—18. A. incensiodorum (sect. Rhizirideum).—19. A. eduardii (sect. Eduardia).—20. A. flavum (sect.
Codonoprasum).—21. A. caeruleum (sect. Caerulea).—22. A. ericetorum (sect. Daghestanica).—23. A. scabriscapum (sect. Scabriscapa).—
24. A. carolinianum (sect. Falcatifolia).—25. A. inconspicuum Vved. (sect. Campanulata).—26. A. galanthum (sect. Cepa).

386 ALISOFriesen, Fritsch, and Blattner
occur in several morphologically derived Amerallium
groups. DNA content per genome is commonly high, rang-
ing from 45–60 pg 2C DNA content but some species of
sects. Caulorhizideum, Lophioprason, Molium, and Broma-
torrhiza range between 30–40 pg (Ohri et al. 1998). The
nectaries show a rather uniform shape and position and ex-
crete mainly through short spurs. This simple structure dif-
fers only slightly from sect. Rhizirideum. A more advanced
nectarial structure was stated for rhizomatous species only
(Fritsch 1992b).
Hypogeal seed germination, the unique A. ursinum seed-
ling type, and large chromosomes with an exceptionally high
2C DNA content of nearly 64 pg characterize sect. Arcto-
prasum as a rather old alliance with a separate position. A
very short leaf sequence beginning with a foliage leaf com-
posed of a thickened basal sheath with storage function, a
narrow petiole, and a rather wide lamina is followed by a
cataphyll and a second foliage leaf without basal storage
function. The lamina is exceptional for the whole genus Al-
lium; the upper and lower side show reversed anatomical
structures and are also reversed positioned. All these mor-
phological characters are certainly signs of a highly gener-
alized specialization. Nevertheless, molecular data show
only small differences between sect. Arctoprasum and other
sections.
Section Bromatorrhiza is characterized by knobby thick-
enings of the cortex cell walls of the roots (Fritsch 1992a),
uniovulate locules (Hanelt 1992), and by a rather low DNA
content of about 30 pg 2C DNA content (Ohri et al. 1998).
All these characters as well as the more specialized structure
of the nectary and excretory tube (Fritsch 1992b) underline
a rather advanced evolutionary state.
A
LLIUM
subgen. C
ALOSCORDUM
(Herb.) R. M. Fritsch, Kew
Bull. 49: 560 (1994).—TYPE: A. neriniflorum (Herb.)
Baker (Fig. 9).
Genus Caloscordum Herb., Edwards’s Bot. Reg 30: 64 (1844); sect.
Caloscordum (Herb.) Baker, J. Bot. 3: 290 (1874).
It is an oligotypic group with three species in east Asia.
It shares multiovulate locules, subterraneous leaf sheaths,
and the presence of relatively large inner vascular bundles
in the scapes (Friesen et al. 1986; Fritsch 1993) with subgen.
Melanocrommyum where three-lobed stigmata are rare and
such bulbs (composed of several storage leaves) do not oc-
cur. Shape and position of nectaries and excretory tubes are
similar to those hitherto found in several subgenera. How-
ever, according to our molecular data subgen. Caloscordum
is most closely related to subgen. Anguinum. Simple char-
acters of seed testa cells are an almost even and slightly
verrucose periclinal wall and straight anticlinal walls (Kruse
1984, 1988), supporting this relationship morphologically.
A
LLIUM
subgen. Anguinum (G. Don ex Koch) N. Friesen,
comb. et stat. nov.—TYPE: A. victorialis L.
Basionym: Allium sect. Anguinum G. Don ex Koch, Synopsis
Deutsch. Schweiz. Fl.: 714 (1837).
In this subgenus two distinct groups exist. One with a
Eurasian-American distribution (A. victorialis alliance, in-
cluding A. tricoccum) and the second restricted to east Asia
(A. prattii, A. ovalifolium, and others). The plants have a
prominent rhizome and develop lateral shoots of first order,
only. The leaf sequence begins with several elongated cat-
aphylls without lamina, followed without transitional types,
by two or three foliage leaves having a distinct petiole part
between basal sheath and the rather wide lamina (Kruse
1992a). Seed germination is hypogeal, and the A. victorialis-
type of seedlings is only present in this subgenus (Drusel-
mann 1992). Narrow, branched, and lengthwise twisted sep-
tal nectaries (Fritsch 1992b), a nearly identical width of the
rhizodermal and exodermal cells, and exodermal cell walls
with exceptionally high suberin content are section-specific
anatomical root characters (Fritsch 1992a). The locules are
uniovulate, a character that occurs sporadically in the genus
Allium and may reflect adaptation to mesophytic herb layers
of forest associations (Hanelt 1992). Hence, it is morpho-
logically and anatomically a rather distinct and specialized
section, but the seed testa sculpturing is very simple (Kruse
1984, 1988), sharing most characters with sect. Caloscor-
dum. Serological data point to close relationships to subgen.
Melanocrommyum and Butomissa (Hanelt et al. 1992),
which might indicate an ancient origin of the group.
A
LLIUM
subgen. Porphyroprason (Ekberg) R. M. Fritsch,
comb. et stat. nov.—TYPE: A. oreophilum C. A. Mey.
(Fig. 8).
Basionym: Allium sect. Porphyroprason Ekberg, Bot. Not. 122:
65 (1969).
Several specific morphological characters such as planar
venation of leaf blades, occurrence of up to three veins in
the outer tepals, a tripartite stigma, 3–4 ovules per locule,
and evenly granulous periclinal walls as well as only slightly
undulate anticlinal walls of the seed testa cells characterize
this monotypic group. Shape and position of nectaries and
excretory tubes do not differ from that often met within sub-
gen. Melanocrommyum, but the leaf blades do not show the
regular pairwise position of opposite vascular bundles, typ-
ical for that subgenus. Also, serological characters underline
a rather isolated position, which is closer to subgen. Melano-
crommyum s.l. than to any other alliance (Hanelt et al. 1989).
The phylogenetical distance to subgen. Vvedenskya is very
small, and both may have a common but very ancient an-
cestor.
A
LLIUM
subgen. Vvedenskya (Kamelin) R. M. Fritsch,
comb. et stat. nov.—TYPE: A. kujukense Vved.
Basionym: Allium sect. Vvedenskya Kamelin, Florogenet. analiz
est. fl. gorn. Sredn. Azii: 243 (1973).
It is also a monotypic group. The locules are multiovulate
as often found in subgen. Melanocrommyum, and also the
shape of the narrowly campanulate flowers is most similar
to A. gypsaceum of sect. Popovia. Especially the lax inflo-
rescence with rather few flowers and the small subglobose
bulbs with several stalked side bulbs and membranous tunics
do not differ much from A. oreophilum. However, the scape
and the cylindrical and tubular leaves of A. kujukense are
densely ribbed and bear short hairs differing thus consider-
ably from that species. These morphological characters, and
also nearly smooth periclinal walls, and Omega-like undu-

VOLUME 22 387Phylogeny of Allium
late anticlinal walls of the seed testa cells (Kruse unpubl.
data) are similar to subgen. Allium s.l. That relationship is,
however, not supported by molecular data, which might in-
dicate that these character states are either plesiomorphic or
evolved independently in both groups.
Three southwest Asian species included here by Khassanov
(2000) very probably belong to sect. Kopetdagia of subgen.
Allium.
A
LLIUM
subgen. M
ELANOCROMMYUM
(Webb et Berth.) Rouy,
Fl. France 2: 378 (1910).—TYPE: A. nigrum L. (Fig. 11).
Sect. Molium sensu Vved. et auct. Ross., pro parte majore, non s.s.
Although large and morphologically extremely diverse, it
is a monophyletic group. The complicated taxonomic struc-
ture may result from reticulate evolution (Mes et al. 1997,
1999). This might also explain that group-specific chromo-
somal characters are largely missing (Fritsch and Astanova
1998). However, current molecular studies did not confirm
the dominance of reticulate structures (Blattner pers. comm.)
but will only result in a new classification proposal after
completion. Therefore, the morphologically based classifi-
cation at sectional level proposed by Khassanov and Fritsch
(1994) is applied here. The relations inside the species-rich
sect. Melanocrommyum remained unclear because only very
few species could be studied as living plants. Preliminary
results of the above mentioned molecular investigation
showed that this section and also the below named large
ones are conglomerates of distantly related subgroups (Blatt-
ner pers. comm.). The sects. Acmopetala, Megaloprason, Re-
geloprason (Fig. 10) Kaloprason (Fig. 12), and Acanthopra-
son are rather large with about 15–35 species each, while
sects. Compactoprason, Pseudoprason, Miniprason, Brevi-
caule, Thaumasioprason, and Verticillata are oligotypic in-
cluding 2–8 species each. Sects. Acaule, Aroidea, and Po-
povia are monotypic.
All members show epigeal germination with seedlings of
the A. karataviense-type (Druselmann 1992) and share a reg-
ular pairwise position of opposite vascular bundles in supra-
and often also subterranean leaf parts. The basal sheath part
of the foliage leaves is commonly not visible above the soil.
Very often this sheath is only partially or not closed and
may then show connections between the margins of consec-
utive leaves (Kruse 1992a). The cell walls of the root en-
dodermis are thickened all around (‘‘O-type’’; Fritsch
1992a). None of these character states occur in any other
subgenus. A strongly unreduced, salt-soluble seed storage
protein with molecular weight of 65,000–70,000 was found
in this subgenus only (Maass 1992). Although most species
show multiovulate locules in the ovary, two ovules per loc-
ule were reported for sect. Megaloprason (Hanelt 1992), and
pseudo-uniovulate locules are characteristic for sect. Com-
pactoprason. DNA values are rather uniform and moderately
high in a range of 40–50 pg 2C DNA content although a
few taxa have only about 30 pg (Ohri et al. 1998). These
low values, as well as exceptionally low TKW (Hanelt 1992)
were found in species of the rather specialized sects. Rege-
loprason and Kaloprason. Hence also in subgen. Melano-
crommyum derived groups have lower DNA content than
more ancient ones, and the contrary conclusion by Ohri et
al. (1998) cannot be upheld. Most commonly the nectaries
are located only in the lower half of the ovaries and excrete
through spurs or rather short tubes. The excretory tubes are
most often downward bent and end in the angle between
ovary and tepals, rarely somewhat higher. These are simpler
character states than present in subgen. Reticulatobulbosa
and Cepa. However, lobed nectaries were found in a few
species but this variation was apparently not correlated with
their taxonomic position (Fritsch 1992b). Multiveined tepals,
mentioned in the description of A. koelzii (Wendelbo) K. M.
Perss. et Wendelbo of sect. Pseudoprason, apparently
evolved independently in subgen. Nectaroscordum and Por-
phyroprason and possess no taxonomic significance for these
groups.
A few aberrant characters are shown by the basal sect.
Verticillata: the leaf blades (the sheaths are completely en-
tire) are longitudinally dissected into thread-like parts. This
leaf shape is unique at least among the members of Alli-
aceae, and probably also in the entire Asparagales. Seed testa
cells show strongly crumpled periclinal walls without central
prominent sculptures and shortly but variably undulate an-
ticlinal walls. This character combination is not known from
any other species in Melanocrommyum or related subgenera
(Kruse 1994). The seed testa of A. aroides, the only species
of sect. Aroidea, is also exceptional in having flat periclinal
walls with evenly verrucose ornamentation, and nearly
straight anticlinal walls (Kruse 1994). The seed testa cells
of all other species of subgen. Melanocrommyum and Allium
(as far as studied to date) are very similar showing convex
periclinal walls with several large prominent sculptures and
verrucose ornamentation, combined with S- to Omega-like
undulate anticlinal walls (Kruse 1992b).
A
LLIUM
subgen. Butomissa (Salisb.) N. Friesen, comb. et
stat. nov.—TYPE: A. ramosum L. (
⫽
A. tataricum L. f.)
(Fig. 13).
Basionym: genus Butomissa Salisb., Gen. Pl. fragm. cont. part
Liriogamae: 90 (1866).
Butomissa is a small group. The typical sect. Butomissa
contains two genetic entities, which morphologically overlap
(Blattner and Friesen 2006). They inhabit the Siberian–Mon-
golian–North Chinese steppes. It is still being discussed
whether they represent one or two species. The growth form
(Kruse 1992a) and chromosome morphology are as simple
as in sect. Rhizirideum (Friesen 1988), but multiovulate loc-
ules (with a mean of 3.2 ovules per locule in A. ramosum)
and rather high TKW (mean 4.1 g; Hanelt 1992) as well as
serological data show relations to subgen. Melanocrommyum
and subgen. Anguinum. Molecular data suggest subgen. Bu-
tomissa occupies a position between these subgenera closer
to subgen. Cyathophora and subgen. Rhizirideum. Position,
shape, and excretory tube of the nectaries show rather simple
character states (Fritsch 1992b).
A
LLIUM
subgen. B
UTOMISSA
sect. Austromontana N. Frie-
sen, sect. nov.—TYPE: A. oreoprasum Schrenk.
Bulbi gregari rhizomati horizontali robusti insidentes, tunicis re-
ticulatis. Folia plana linearia, prope basin scapi congesta. Tepala
rosea nervo atropurpureo.
Bulbs aggregated, outer tunics reticulate. Rhizome hori-

388 ALISOFriesen, Fritsch, and Blattner
Fig. 27.—Seed surface of Allium oreoprasum. Scale bar
⫽
20
␮
m.
Fig. 28.—Seed surface of Allium ramosum. Scale bar
⫽
20
␮
m.
zontal, well developed. Leaves flat, linear, aggregated in
lower part of scape. Tepals pink with a dark purple nerve.
This section contains two species (A. oreoprasum Schrenk
and A. gilgiticum Wang & Tang) occurring in the mountains
from eastern to central Asia up to the borderline of the east-
ern Mediterranean area. The seed testa of A. oreoprasum
(Fig. 27) is very comparable to A. ramosum (Fig. 28) and
A. tuberosum of subgen. Butomissa, differing only by the
verrucose ornamentation of the periclinal walls.
A
LLIUM
subgen. Cyathophora (R. M. Fritsch) R. M. Fritsch,
comb. et stat. nov.—TYPE: A. cyathophorum Bur. et
Franch. (Fig. 14).
Basionym: Allium sect. Cyathophora R. M. Fritsch, Kew Bull. 49:
561 (1994).
Subgen. Bromatorrhiza Ekberg, Bot. Not. 122: 58 (1969), pro parte,
excl. sect. Bromatorrhiza.
The oligotypic sects. Cyathophora and Coleoblastus and
the monotypic sect. Milula belong to this small and solely
Asian (Tibet and the Himalayas) group. Though the leaf ba-
ses seem somewhat inflated, thickened parts with storage
function and storage leaves are absent (Kruse 1992a) and
also a bulb is missing. All species share only one row of
identically orientated vascular bundles in the leaf blades
combined with the presence of palisade parenchyma and
subcortical laticifers, which is perhaps the most ancient char-
acter combination in the genus (Fritsch 1988). As far as
known, all species have biovulate locules (Hanelt 1992). The
elongated inflorescence of A. spicatum (Prain) N. Friesen
attracts special attention, though all other characters tested
agree completely with the other species of this subgenus.
The roots are less specialized than those of the other sections
showing evenly and not knot-like thickened cortex cell walls
(Friesen et al. 2000). Growth form and foliation of A. cy-
athophorum seems slightly more advanced than in sect. Rhi-
zirideum showing ramification up to lateral shoots of third
order (with inflorescences on those of first and second order)
and rarely transitional leaf forms with somewhat reduced
lamina but no cataphylls (Kruse 1992a). Structures of nec-
tary and excretory tube also show a more advanced state
than in sect. Rhizirideum (Fritsch 1992b).
A
LLIUM
subgen. R
HIZIRIDEUM
(G. Don ex Koch) Wendelbo,
Bot. Not. 122: 25 (1969), s.s.—TYPE: A. senescens L.
Subgen. Steiptoprason Radic´, Razpr. Slov. Akad. Znan. Umetn.,
Razr. Nar. Vede. [SAZU] 31: 250, 251 (1990).—TYPE: A. in-
censiodorum Radic´.
Though this subgeneric name has been used for groups
with extremely different circumscriptions, it must remain in
use because of nomenclatural reasons. We regard it as a
comparatively small subgenus comprising currently about 37
species including the sects. Rhizirideum (Fig. 18), Caespi-
tosoprason (Fig. 15, 16), Tenuissima (Fig. 17), Rhizomatosa,
and the new monotypic sect. Eduardia. These are Eurasian
steppe taxa showing the most diversity in southern Siberia
and Mongolia. Only a few species are distributed in Europe,
reaching Portugal as their western-most outpost, and some
species such as A. anisopodium (sect. Tenuissima), A. spirale
(sect. Rhizirideum) occur also in Korea and far eastern Rus-
sia, as well as A. togashii (sect. Rhizirideum) in Japan. The
growth form is simple because only foliage leaves with a
complete lamina are developed (Kruse 1992a). Also, the
growth rhythm presents a rather ancestral state in these sec-
tions (Hanelt et al. 1992). The simple form of nectaries with-
out an excretory tube (Fritsch 1992b) and differing karyo-
types in every section (Friesen 1988) underline this phylo-
genetically rather ancient state. The occurrence of several
ploidy levels in the A. senescens alliance (sect. Rhizirideum)
is connected with the origin of several young species, show-
ing thus a secondary radiation in this group (Friesen 1992).
A
LLIUM
subgen. R
HIZIRIDEUM
sect. Eduardia N. Friesen,
sect. nov.—TYPE: A. eduardii Stearn (Fig. 19).
Bulbi conici gregari ad rhizoma repens insidentes. Spatha in ros-
trum longum attenuata, rostrum fere triplo basi longiorum. Umbella
hemisphaerica pauciflora laxa.
Bulbs several, conical, borne on creeping rhizome. Spathe
with long beak, nearly three times longer than base. Umbel
hemispherical, few flowered, lax.

VOLUME 22 389Phylogeny of Allium
The shape of the nectaries is identical to sect. Rhizirideum,
but the presence of a long excretory tube (Fritsch 1992b),
fibrous bulb tunics, and a deviating chromosome morphol-
ogy (Friesen 1988) support the taxonomic separation.
A
LLIUM
subgen. A
LLIUM
.—TYPE: A. sativum L.
Subgen. Codonoprasum (Rchb.) Zahariadi, Problems Balkan Flora:
230 (1976).—TYPE: A. oleraceum L.
This is the largest subgenus of Allium comprising by far
the largest number of species. Molecular data support the
division into two main groups.
One refers to the ‘‘classical’’ sect. Allium that has tripartite
inner filaments and only one thick storage cataphyll. Often
more chromosomes than in other subgenera have exception-
ally long satellites (Hanelt et al. 1992), which characterize
this group karyologically. The rather recent splitting of the
mainly oligotypic sects. Caerulea (Fig. 21), Crystallina,
Multicaulea, Spathulata, and Brevidentia by Khassanov
(1996, 2000) is widely supported by molecular data, whereas
the informal groups proposed by Mathew (1996) are not re-
flected in our molecular analyses. Inclusion of A. haneltii F.
O. Khassanov et R. M. Fritsch in sect. Brevidentia (Fritsch
et al. 1998) must be revised as Fig. 2 shows. Nevertheless,
sect. Allium remains the most species-rich in the genus. An
inflorescence with numerous long bracteoles, paper-like bulb
tunics, a symmetric karyotype with small satellites, and
straight anticlinal walls of seed testa cells characterize sect.
Spathulata as less specialized among these sections (Fritsch
et al. 1998).
The second group is morphologically more diverse and
comprises less closely related sections. The rather distinct
sects. Codonoprasum (Fig. 20), Brevispatha, and Kopetda-
gia as well as segregates of bulbous species of the ‘‘classi-
cal’’ sect. Scorodon in a broad sense belong to it. These
species often have two or more cataphylls in the bulbs and
a different shape and position of the nectaries and excretory
tubes (Fritsch 1992b). However, only small differences of
the growth form to sect. Allium exist (Kruse 1992a).
Four sections were invalidly combined by Khassanov
(2000) and will be validated here:
A
LLIUM
subgen. A
LLIUM
sect. Eremoprasum (Kamelin) F.
O. Khassanov, R. M. Fritsch et N. Friesen, comb. et stat.
nov.—TYPE: A. sabulosum Stev.
Basionym: Allium subsect. Eremoprasum Kamelin, Florogenet.
analiz est. fl. gorn. Sredn. Azii: 237 (1973).
A
LLIUM
subgen. A
LLIUM
sect. Longivaginata (Kamelin) F.
O. Khassanov, R. M. Fritsch et N. Friesen, comb. et stat.
nov.—TYPE: A. longivaginatum Wendelbo.
Basionym: Allium subsect. Longivaginata Kamelin, Florogenet.
analiz est. fl. gorn. Sredn. Azii: 238 (1973).
A
LLIUM
subgen. A
LLIUM
sect. Pallasia (Tzag.) F. O. Khas-
sanov, R. M. Fritsch et N. Friesen, comb. et stat. nov.—
TYPE: A. pallasii Murr.
Basionym: Allium ser. Pallasia Tzagolova, Novosti Sist. Vyssh.
Nizsh. Rast.: 53 (1977).
A
LLIUM
subgen. A
LLIUM
sect. Mediasia F. O. Khassanov, S.
C. Yengalycheva et N. Friesen, sect. nov.—TYPE: A. tur-
kestanicum Regel.
Scapus 50–100 cm altus. Folia in numero 4 –6, laminae lineares
1–1.5 cm latae. Inflorescentia denseque subglobosa, multiflora. Peri-
gonii late campanulati phylla obtusa. Filamenta perigonii breviora
subulata, interiora inferne denticulata. Stylus exsertus. Ovarium ro-
tundato-tripartitum, rubrum.
Plant 50–100 cm tall. Leaves 4 –6, 1–2.5 cm broad, linear.
Inflorescence subspherical, dense, many-flowered. Flowers
campanulate, tepals obtuse. Filaments slightly longer than
tepals, inner filaments basally obovate. Style exserted. Ovary
subspherical with three furrows, red.
This section is monotypic. However, A. yuchuanense Y.
Z. Zhao & J. Y. Chao from China may possibly also belong
here.
Scape length, leaf shape and number, as well as the bi-
partite spathe with a moderately long beak of A. turkestan-
icum are rather similar to the garlic alliance of sect. Allium.
However, the campanulate perianth, inner filaments with an
obovate basal lamina and very short lateral cusps, ovoid
bulbs, 3-lobed pinkish ovaries with deep pocket-like mounds
of nectarial tubes, some anatomical and embryological char-
acters (Khassanov et al. in prep.), as well as seed testa cells
having a narrow anticlinal field and slightly convex pericli-
nal walls with granulate sculptures (Kruse unpubl. data) do
not fit that section. This specific combination of characters
is known from several rhizomatous groups and may under-
line an ancient hybridogenous origin of this taxon. Section
Mediasia is included here because the bulbous character
dominates.
A
LLIUM
subgen. Reticulatobulbosa (Kamelin) N. Friesen,
comb. et stat. nov.—TYPE: A. lineare L.
Basionym: Allium sect. Reticulato-bulbosa Kamelin, Florogenet.
analiz est. fl. gorn. Sredn. Azii: 239 (1973).
Sections Reticulatobulbosa, Campanulata, and the new
sects. Scabriscapa, Nigrimontana, and Sikkimensia belong
to this subgenus. They display much similarity in vegetative
morphology, as reticulate or at least fibrous tunics and nar-
rowly linear leaf blades, and much diversity of flower char-
acters. The shoot foliation is advanced beginning with two
to five cataphylls having an extremely reduced lamina, fol-
lowed by transitional leaves with reduced lamina, and finally
foliage leaves with a complete lamina (Kruse 1992a). A
many-layered root exodermis with strongly thickened cell
walls (Fritsch 1992a), as well as the broad and phloem-rich
inner vascular bundles of the scape (Fritsch 1993) are only
typical for sect. Campanulata and a few more species from
the other sections.
A
LLIUM
subgen. R
ETICULATOBULBOSA
sect. Scabriscapa
(Tscholok.) N. Friesen, comb. et stat. nov.—TYPE: A. sca-
briscapum Boiss. (Fig. 23).
Basionym: Allium ser. Scabriscapa Tscholok, Zametki Sist. Geogr.
Rast. 31: 42 (1975).
Allium scabriscapum (including A. eriocoleum Vved.), A.
sulphureum, and A. trachyscordum represent this section.
Except for the latter species, they have bright yellow flowers.

390 ALISOFriesen, Fritsch, and Blattner
A
LLIUM
subgen. R
ETICULATOBULBOSA
sect. Nigrimontana N.
Friesen, sect. nov.—TYPE: A. drobovii Vved.
Folia plana, linearia, altitudine caulis non superantia. Tepala albo-
virescentia vel albo-roseola purpureo-nervata. Stamina integra.
Leaves flat, linear, shorter than scape. Tepals greenish-
white or pinkish-white with red nerves. Filaments entire.
Here belong A. drobovii and A. oreoprasoides, endemic
species from Karatau Mountains, Kazakhstan. They occupy
an isolated position next to sect. Campanulata. Shape of the
nectaries and excretory tubes are similar but not identical to
those of sect. Campanulata (Fritsch 1992b).
A
LLIUM
subgen. R
ETICULATOBULBOSA
sect. Sikkimensia
(Traub) N. Friesen, comb. et stat. nov.—TYPE: A. sikki-
mense Baker.
Basionym: Allium subsect. Sikkimensia Traub, Pl. Life (Stanford)
28: 136 (1972).
This section is characterized by blue flowers and occurs
mainly in southwestern and southern China.
A
LLIUM
subgen. P
OLYPRASON
Radic´, Razp. Slov. Akad. Zna-
nosti Umetn., Razr. Nar. Vede. [SAZU] 31: 250, 253
(1990).—TYPE: A. moschatum L.
This subgenus comprises mainly the former sect. Orei-
prason in the broad sense of Kamelin (1973), and sect. Sco-
rodon in the strict sense (possessing bulbs with rhizomes;
subsect. Moschata (Omelcz.) Tscholok.). We accept sect. Or-
eiprason s.s. to include sect. Petroprason (though the scape
anatomy is more similar to sect. Rhizirideum; Fritsch 1993),
and segregate the new sects. Falcatifolia and Daghestanica.
All taxa possess papery to leathery bulb tunics sometimes
breaking into strips in the upper part. The growth form is as
advanced as described above for subgen. Reticulatobulbosa,
but A. moschatum is exceptional in missing cataphylls (Kru-
se 1992a). Rather similar nectaries and excretory tubes
strongly support grouping of the above-mentioned sections
and subsections. Surprisingly, nectarial characters of the sec-
tions united now under subgen. Reticulatobulbosa do not
differ much (Fritsch 1992b).
The name proposed by Radic´ must be used because of
nomenclatural reasons although, the original Latin charac-
terization denies, incorrectly, the presence of a rhizome.
A
LLIUM
subgen. P
OLYPRASON
sect. Falcatifolia N. Friesen,
sect. nov.—TYPE: A. carolinianum DC. (Fig. 24).
Rhizoma breve, verticale. Bulbus semper fere solitarius vel inter-
dum gregarius tunicis coriaceis vel scariosis. Folia plana plerumque
falcata vel linearia.
Rhizome short, vertical. Bulbs usually single or some-
times several, with coriaceous or scarious tunics. Leaves flat,
usually falcate or linear.
The species of this section are found growing in the mon-
tane to subalpine belt of Central Asian mountains. They are
characterized by strong rhizomes, which enable the species
to survive in steep scree and rubble slopes. The more or less
falcate leaf blades may be narrow or rather broad.
A
LLIUM
subgen. P
OLYPRASON
sect. Daghestanica (Tscholok.)
N. Friesen, comb. nov.—TYPE: A. daghestanicum
Grossh.
Basionym: Allium ser. Daghestanica Tscholok., Zametki Sist.
Geogr. Rast. 25: 83 (1965).
This section consists of two geographical alliances. The
first one contains the Caucasian species A. daghestanicum
and A. gunibicum, having thin thread-like leaves and begin-
ning with anthesis only in autumn. Allium ericetorum (Fig.
22). A. ochroleucum, A. kermesinum Rchb., and A. suaveo-
lens belong to the second (European) alliance distributed
from the eastern Alps to the Pyrenees. These plants show
semicylindrical or narrowly linear leaves, are flowering in
summer, and are the only group of the subgenus having bulb
tunics splitting into longitudinal stripes (but not into fibers).
A
LLIUM
subgen. C
EPA
(Mill.) Radic´, Razp. Slov. Akad. Zna-
nosti Umetn., Razr. Nar. Vede. [SAZU] 31: 250, 251
(1990).—TYPE: A. cepa L.
This subgenus comprises taxa with fistulous leaves and
scapes although a few species are exceptional in having
more or less flat leaf blades. The sects. Cepa (Fig. 26),
Schoenoprasum s.s., Annuloprason, Sacculiferum, and the
new monotypic sect. Condensatum are included. Their close
relationship has already been stated by Hanelt et al. (1992).
Bulbs are mostly well developed, although a large size as in
sect. Cepa is exceptional. They also share a 4- to 6-fold
number of vascular bundles outside the scape sclerenchyma
compared to inside (Fritsch 1993). Reticulate or fibrous bulb
tunics do not occur in this group. Position and shape of the
nectaries is variable, but the excretory tubes are always spe-
cially shaped (Fritsch 1992b). All investigated members of
sect. Cepa (in the circumscription of Hanelt 1985) share a
specific satellite DNA sequence that had evolved already in
the progenitor forms (Pich et al. 1996).
Chromosomal and molecular characters favor a position
of A. roylei in subsect. Cepa, but morphological characters
of inflorescence, flower parts, and seed testa unequivocally
support exclusion from this section as member of sect. Or-
eiprason. Only the study of newly introduced natural acces-
sions may possibly elucidate the true taxonomic state of this
taxon (Fritsch and Friesen 2002).
A
LLIUM
subgen. C
EPA
sect. Condensatum N. Friesen, sect.
nov.—TYPE: A. condensatum Turcz.
Bulbi solitarii rhizomati brevi insidentes. Folia semicylindrica fis-
tulosa longitudinaliter profunde sulcata. Inflorescentia globosa com-
pacta multiflora. Tepala flaveola viridi-nervata.
Bulbs borne singly on a short rhizome. Leaves semicylin-
drical, deeply grooved, fistulous. Umbel globose, compact,
many flowered. Perianth pale yellow with greenish nerves.
This monotypic section has a very slender habit, thread-
like leaves, and membranous bulb tunics. It occurs from
eastern Siberia and Mongolia north to Korea and Japan.
DISCUSSION
Phylogenetic analysis of the ITS region from 221 acces-
sions representing 196 Allium and five outgroup species re-
vealed several intrageneric taxa of Allium as poly- or para-

VOLUME 22 391Phylogeny of Allium
phyletic. These findings are mostly in accord with the results
from other molecular studies as reviewed in Klaas and Frie-
sen (2002). Our taxon sample covers roughly one-quarter of
the known Allium species, assembled to represent all groups
and paying particular attention to putative or known non-
monophyletic groups in earlier classification systems. We
also included most of the ‘‘difficult’’ species with uncertain
taxonomic affiliation to provide a firm basis for a phyloge-
netic classification of the genus.
Our data showed Allium to be monophyletic when Milula
(Friesen et al. 2000) and Nectaroscordum (Fay and Chase
1996) were included in the genus. A conspicuous feature of
the ITS data were the very high genetic distances (
⬎
40%)
within Allium. Values above 40% often characterize the most
distant genera within subfamilies or even families (e.g.,
Baldwin et al. 1995; Hsiao et al. 1999; Noyes and Rieseberg
1999) in ITS analyses, while intrageneric distances in other
plant families are mostly less than 10% (Baldwin et al.
1995). Large genetic distances were not restricted to the ITS
region, but were also found in chloroplast rbcL–atpB se-
quences (Klaas and Friesen 2002). These findings suggest
that the genus Allium is of ancient origin and molecular evo-
lution was not accompanied by the rise of pronounced mor-
phological divergence and accompanying higher taxonomic
categories (Friesen et al. 2000). The pronounced molecular
differences together with the extant distribution area of Al-
lium indicate an origin of the genus early in the Tertiary
(Friesen et al. 2000; in prep.). Thus in age and genetic var-
iation the genus Allium resembles plant families in other
groups of the angiosperms. Although a split of Allium into
several genera would result in better comparability of equal
taxonomic categories among different plant families, we re-
jected this possibility as: (1) Allium in its current circum-
scription is monophyletic, and (2) a change would necessi-
tate a tremendous amount of new taxon designations. More
than 30 generic names described between 1754 and 1869
refer to Allium species and would have to be examined ac-
cording to the current rules of botanical nomenclature. Also
(current as well as former) sectional and subsectional names
would have to be proved concerning correctness of affilia-
tion and circumscription. The determination of correct types
for all these groups would become an extremely complicated
work. As only the current subgen. Allium (about 300 species)
would remain unaltered, roughly 500 new binominals and
the same high number of new synonyms would have to be
recognized, just to shift the former problems to another level.
This, we think, would not contribute to a widely accepted
and long-term stable taxonomic system for Allium.
The adequate classification of the phylogenetically com-
plicated genus Allium requires, in our opinion, about 70 in-
frageneric taxa at the subgeneric and sectional levels to cov-
er morphologically definable monophyletic groups. This
seems an inconveniently high number, with the result that
dealing with Allium may necessitate time-consuming and oc-
casionally frustrating navigation through long keys and ex-
tended comparisons of characters. However, as the majority
of our proposed changes concern the polyphyletic subgen.
Rhizirideum, the split into six new subgenera created mor-
phologically relatively homogenous groups. The presence of
a visible rhizome was the key character of former subgen.
Rhizirideum, although morphology and growth form of these
rhizomes is quite different among the different clades. All
species from sect. Anguinum have an ascending rhizome, the
species from sect. Rhizirideum s.s. have a horizontal rhi-
zome, and species from sect. Cepa have a short vertical rhi-
zome. Rhizomes were thought to be a basal character in
Allium and that within subgen. Rhizirideum long rhizomes
were the primitive character state and short or nearly reduced
rhizomes to be advanced (Hanelt et al. 1992). However, the
diversity of rhizome forms could not be correlated with other
morphological characters used as markers for phylogenetic
relationships in former subgen. Rhizirideum or the entire ge-
nus Allium. Our phylogenetic analysis indicates that rhi-
zomes evolved several times independently in the groups of
former subgen. Rhizirideum, which explains the different
rhizome forms in these groups. A correlation between the
occurrence of rhizomes and habitat preferences exists, as
well as between rhizome morphology and different life
forms (Cheremushkina 1992, 2001). Furthermore, our newly
defined sections of former subgen. Rhizirideum are also
monomorphic with regard to the bulb tunics, thus avoiding
contradicting character distribution in former subgen. Rhi-
zirideum (Hanelt et al. 1992). Generally, the new classifi-
cation will allow a sound interpretation of character states
in Allium and thus allows us to find correlations between
morphological structures and their ecological or phylogenet-
ic relevance.
The conspectus (Appendix 1) proposes to divide Allium
into 15 subgenera and 56 sections (without type sections).
The approximate number of species is given for every sub-
genus. We believe this to be helpful because one may find
very different species numbers for Allium in recent books
and papers, ranging from 450 to approximately 750. The
number given for subgen. Amerallium is based on the most
recent and excellent revision for America (McNeal and Ja-
cobsen 2002). All other numbers were concluded from our
own studies during fieldwork and with living collections, as
well as from good herbarium specimens. Unfortunately, we
were not able to study all described taxa from living plants,
and could only estimate the true state of several species
names from descriptions and incomplete dry specimens.
Thus, the sum of all species numbers given exceeds 800,
which is possibly somewhat too high, but a number of about
780 Allium species seems currently a realistic estimation.
ACKNOWLEDGMENTS
We would like to thank P. Hanelt, who initiated the Allium
work in Gatersleben, together with K. Bachmann for long-
term support of this work, the technicians of the IPK De-
partment of Taxonomy for invaluable help in maintaining
the Allium collection, and the curators of the herbaria LE,
TASH, NS, and NSK for the possibility to use their collec-
tions and facilities. We also thank the referees Michael Hav-
ey, Dale McNeal, Linda M. Prince, and Carol Annable for
critical reading the manuscript, for valuable suggestions, and
helpful comments.
LITERATURE CITED
B
ALDWIN
, B. G. 1992. Phylogenetic utility of the internal transcribed
spacers of nuclear ribosomal DNA in plants: an example from the
Compositae. Molec. Phylogen. Evol. 1: 3–16.

392 ALISOFriesen, Fritsch, and Blattner
,M.J.S
ANDERSON
,J.M.W
OJCIECHOWSKI
,C.S.C
AMPBELL
,
AND
M. J. D
ONOGHUE
. 1995. The ITS region of nuclear ribosomal
DNA: a valuable source of evidence on angiosperm phylogeny.
Ann. Missouri Bot. Gard. 82: 247–277.
B
LATTNER
, F. R. 1999. Direct amplification of the entire ITS region
from poorly preserved plant material using recombinant PCR.
BioTechniques 27: 1180–1186.
. 2004. Phylogenetic analysis of Hordeum (Poaceae) as in-
ferred by nuclear rDNA ITS sequences. Molec. Phylogenet. Evol.
33: 289–299.
,
AND
N. F
RIESEN
. 2006. Chapter 10. Relationship between
Chinese chive (Allium tuberosum) and its putative progenitor A.
ramosum as assessed by random amplified polymorphic DNA
(RAPD), pp. 133–141. In M. A. Zeder, D. Decker-Walters, E.
Emshwiller, D. Bradley, and B. D. Smith [eds.], Documenting
domestication: new genetic and archaeological paradigms. Uni-
versity of California Press, Berkeley, USA.
,
AND
J. W. K
ADEREIT
. 1999. Morphological evolution and
ecological diversification of the forest-dwelling poppies (Papav-
eraceae: Chelidonioideae) as deduced from a molecular phylogeny
of the ITS region. Pl. Syst. Evol. 219: 181–197.
,K.W
EISING
,G.B
A
¨NFER
,U.M
ASCHWITZ
,
AND
B. F
IALA
.
2001. Molecular analysis of phylogenetic relationships among
myrmecophytic Macaranga species (Euphorbiaceae). Molec. Phy-
logenet. Evol. 19: 331–344.
B
REMER
, K. 1988. The limits of amino acid sequence data in angio-
sperm phylogenetic reconstruction. Evolution 42: 795–803.
C
HEREMUSHKINA
, V. A. 1992. Evolution of life forms of species in
subgenus Rhizirideum (Koch) Wendelbo, genus Allium L., pp. 27–
34. In P. Hanelt, K. Hammer, and H. Knu¨pffer [eds.], The genus
Allium: taxonomic problems and genetic resources. Proceedings
of an international symposium held at Gatersleben, Germany, 11–
13 Jun 1991. Institut fu¨ r Pflanzengenetik und Kulturpflanzenfor-
schung, Gatersleben, Germany.
. 2001. Biomorfologija vidov roda Allium L. v Evrazii i
struktura ikh tsenopopulyatsij. Avtoreferat dissertatsii na soiskanie
uchenoi stepeni doktora biologicheskikh nauk. Novosibirsk, Rus-
sia. 32 p.
D
ON
, G. 1832. A monograph of the genus Allium. Mem. Wern. Nat.
Hist. Soc., Edinburgh 6: 1–102.
D
RUSELMANN
, S. 1992. Vergleichende Untersuchungen an Vertretern
der Alliaceae Agardh. 1. Morphologie der Keimpflanzen der Gat-
tung Allium L. Flora, Morphol. Geobot. Oekophysiol. 186: 37–
52.
D
UBOUZET
,J.G.,
AND
K. S
HINODA
. 1998. Phylogeny of Allium L.
subgenus Melanocrommyum (Webb et Berth.) Rouy based on
DNA sequence analysis of the internal transcribed spacer region
of nrDNA. Theor. Appl. Genet. 97: 541–549.
,
AND
. 1999. Relationships among Old an New World
Alliums according to ITS DNA sequence analysis. Theor. Appl.
Genet. 98: 422–433.
, ,
AND
N. M
URATA
. 1997. Phylogeny of Allium L.
subgenus Rhizirideum (G. Don ex Koch) Wendelbo according to
dot blot hybridization with randomly amplified DNA probes.
Theor. Appl. Genet. 95: 1223–1228.
F
AY
,M.F.,
AND
M. W. C
HASE
. 1996. Resurrection of Themidaceae
for the Brodiaea alliance, and recircumscription of Alliaceae,
Amaryllidaceae and Agapanthoideae. Taxon 45: 441– 451.
F
ELSENSTEIN
J. 1985. Confidence limits on phylogenies: an approach
using the bootstrap. Evolution 39: 783–791.
F
RIESEN
,[‘‘F
RIZEN
’’] N. 1988. Lukovye Sibiri: sistematika, kariol-
ogija, khorologija. Nauka–Sibirskoe Otd., Novosibirsk, Russia.
185 p.
. 1992. Systematics of the Siberian polyploid complex in
subgenus Rhizirideum (Allium), pp. 55–66. In P. Hanelt, K. Ham-
mer, and H. Knu¨pffer [eds.], The genus Allium: taxonomic prob-
lems and genetic resources. Proceedings of an international sym-
posium held at Gatersleben, Germany, 11–13 Jun 1991. Institut
fu¨ r Pflanzengenetik und Kulturpflanzenforschung, Gatersleben,
Germany.
,
AND
F. R. B
LATTNER
. 2000. Geographical isolation predom-
inates over ecological differentiation in the phylogeny of Allium
sect. Schoenoprasum (Alliaceae). Pl. Biol. (Stuttgart) 2: 297–305.
,R.M.F
RITSCH
,S.P
OLLNER
,
AND
F. R. B
LATTNER
. 2000.
Molecular and morphological evidence for an origin of the aber-
rant genus Milula within Himalayan species of Allium (Alliaceae).
Molec. Phylogenet. Evol. 17: 209–218.
,S.P
OLLNER
,K.B
ACHMANN
,
AND
F. R. B
LATTNER
. 1999.
RAPDs and noncoding chloroplast DNA reveal a single origin of
the cultivated Allium fistulosum from A. altaicum (Alliaceae).
Amer. J. Bot. 86: 554–562.
[‘‘F
RIZEN
, N. V.’’], V. V. Z
UEV
,
AND
N. S. A
LJANSKAJA
. 1986.
Krasivoluk nereidocvetnyj—Calloscordum neriniflorum Herbert,
pp. 83–92. In K. A. Sobolevskaja, Biologicheskie osobennosti ras-
tenij sibiri, nuzhdajushhikhsja v okhrane. Nauka–Sibirskoe Otd.,
Novosibirsk, Russia.
F
RITSCH
, R. 1988. Anatomische Untersuchungen an der Blattspreite
bei Allium L. (Alliaceae)—I. Arten mit einer einfachen Leitbu¨n-
delreihe. Flora, Morphol. Geobot. Oekophysiol. 181: 83–100.
. 1992a. Zur Wurzelanatomie in der Gattung Allium L. (Al-
liaceae). Beitr. Biol. Pflanzen 67: 129–160.
. 1992b. Septal nectaries in the genus Allium L., pp. 77–85.
In P. Hanelt, K. Hammer, and H. Knu¨pffer [eds.], The genus Al-
lium: taxonomic problems and genetic resources. Proceedings of
an international symposium held at Gatersleben, Germany, 11–13
Jun 1991. Institut fu¨ r Pflanzengenetik und Kulturpflanzenfor-
schung, Gatersleben, Germany.
. 1993. Anatomische Merkmale des Blu¨tenschaftes in der
Gattung Allium L. und ihre systematische Bedeutung. Bot. Jahrb.
Syst. 115: 97–131.
,
AND
S. B. A
STANOVA
. 1998. Uniform karyotypes in different
sections of Allium L. subgen. Melanocrommyum (Webb & Berth.)
Rouy from Central Asia. Feddes Repert. 109: 539–549.
,
AND
N. F
RIESEN
. 2002. Evolution, domestication and tax-
onomy, pp. 5–30. In H. D. Rabinowitch and L. Currah [eds.],
Allium crop science: recent advances. CABI Publishing, Walling-
ford, Oxfordshire, UK.
,F.O.K
HASSANOV
,
AND
N. F
RIESEN
. 1998. New taxa, new
combinations, and taxonomic remarks on Allium L. from Fergan
depression, Middle Asia. Linzer Biol. Beitr. 30: 281–292
⫹
843.
G
REGORY
, M., R. M. F
RITSCH
,N.F
RIESEN
,F.O.K
HASSANOV
,
AND
D.
W. M
C
N
EAL
(editors). 1998. Nomenclator Alliorum: Allium names
and synonyms—a world guide. Royal Botanic Gardens, Kew,
Richmond, Surrey, UK. 83 p.
G
REUTER
, W., J. M
C
N
EILL
,F.R.B
ARRIE
, H.-M. B
URDET
,V.D
EMOU
-
LIN
,T.S.F
ILGUEIRAS
,D.H.N
ICOLSON
,P.C.S
ILVA
,J.E.S
KOG
,P.
T
REHANE
,N.J.T
URLAND
,
AND
D. L. H
AWKSWORTH
(editors). 2000.
International Code of Botanical Nomenclature (St. Louis Code)
adopted by the Sixteenth International Botanical Congress St.
Louis, Missouri, Jul–Aug 1999. Koeltz Scientific Books, Ko¨ nig-
stein, Germany. 474 p.
H
ANELT
, P. 1985. Zur Taxonomie, Chorologie und O
¨kologie der
Wildarten von Allium L. sect. Cepa (Mill.) Prokh. Flora, Morphol.
Geobot. Oekophysiol. 176: 99–116.
. 1992. Ovule number and seed weight in the genus Allium
L., pp. 99–105. In P. Hanelt, K. Hammer, and H. Knu¨ pffer [eds.],
The genus Allium: taxonomic problems and genetic resources.
Proceedings of an international symposium held at Gatersleben,
Germany, 11–13 Jun 1991. Institut fu¨ r Pflanzengenetik und Kul-
turpflanzenforschung, Gatersleben, Germany.
,R.F
RITSCH
,J.K
RUSE
,H.M
AASS
,H.O
HLE
,
AND
K. P
ISTRICK
.
1989. Allium L. sect. Porphyroprason Ekberg—Merkmale und
systematische Stellung. Flora, Morphol. Geobot. Oekophysiol.
182: 69–86.

VOLUME 22 393Phylogeny of Allium
,J.S
CHULZE
-M
OTEL
,R.M.F
RITSCH
,J.K
RUSE
,H.M
AASS
,H.
O
HLE
,
AND
K. P
ISTRICK
. 1992. Infrageneric grouping of Allium—
the Gatersleben approach, pp. 107–123. In P. Hanelt, K. Hammer,
and H. Knu¨ pffer [eds., The genus Allium: taxonomic problems
and genetic resources. Proceedings of an international symposium
held at Gatersleben, Germany, 11–13 Jun 1991. Institut fu¨r Pflan-
zengenetik und Kulturpflanzenforschung, Gatersleben, Germany.
H
SIAO
, C., S. W. L. J
ACOBS
,N.J.C
HATTERTON
,
AND
K. H. A
SAY
.
1999. A molecular phylogeny of the grass family (Poaceae) based
on the sequences of nuclear ribosomal DNA (ITS). Austral. Syst.
Bot. 11: 667–688.
K
AMELIN
, R. V. 1973. Florogeneticheskij analiz estestvennoj flory
gornoj Srednej Azii. Nauka, Leningrad, Russia. 354 p.
K
HASSANOV
, F. O. 1996. Conspectus of the wild growing Allium
species of Middle Asia, pp. 141–59. In M. O
¨ztu¨ rk, O
¨. Sec¸ men,
and G. Go¨ rk [eds.], Plant life in southwest and Central Asia. EGE
University Press, Izmir, Turkey.
. 2000. Subinfrageneric grouping in genus Allium subgenus
Allium, pp. 107–112. In O. Ashurmetov, F. Khassanov, and Y.
Salieva [eds.], Plant life in southwest and Central Asia. Proceed-
ings of the 5th international symposium, 18–22 May 1998, Tash-
kent, Uzbekistan.
,
AND
R. M. F
RITSCH
. 1994. New taxa in Allium L. subgen.
Melanocrommyum (Webb & Berth.) Rouy from Central Asia. Lin-
zer Biol. Beitr. 26: 965–990.
K
LAAS
, M.,
AND
N. F
RIESEN
. 2002. Molecular markers in Allium, pp.
159–185. In H. D. Rabinowitch and L. Currah [eds.], Allium crop
science: recent advances. CABI Publishing, Wallingford, Oxford-
shire, UK.
K
RUSE
, J. 1984. Rasterelektronenmikroskopische Untersuchungen an
Samen der Gattung Allium L. Kulturpflanze 32: 89–101.
. 1988. Rasterelektronenmikroskopische Untersuchungen an
Samen der Gattung Allium L. III. Kulturpflanze 36: 355–368.
. 1992a. Growth form characters and their variation in Allium
L., pp. 173–179. In P. Hanelt, K. Hammer, and H. Knu¨pffer [eds.],
The genus Allium: taxonomic problems and genetic resources.
Proceedings of an international symposium held at Gatersleben,
Germany, 11–13 Jun 1991. Institut fu¨ r Pflanzengenetik und Kul-
turpflanzenforschung, Gatersleben, Germany.
. 1992b. Variability of testa sculptures in the genus Allium
L., pp. 181–182. In P. Hanelt, K. Hammer, and H. Knu¨pffer [eds.],
The genus Allium: taxonomic problems and genetic resources.
Proceedings of an international symposium held at Gatersleben,
Germany, 11–13 Jun 1991. Institut fu¨ r Pflanzengenetik und Kul-
turpflanzenforschung, Gatersleben, Germany.
. 1994. Rasterelektronenmikroskopische Untersuchungen an
Samen der Gattung Allium L. IV. Feddes Repert. 105: 457–471.
K
UDRJASHOVA
, G. L. 2003. The genus Nectaroscordum (Alliaceae)
in the flora of the Caucasus. Bot. Zhurn. (Moscow & Leningrad)
88: 87–92. [In Russian].
L
INNAEUS
,C.
VON
. 1753. Species plantarum, Vol. 1. Allium, pp. 294 –
302. Laurentiis Salvii, Stockholm, Sweden. [Facsimile edition,
1957–1959, Ray Society, London, UK.]
L
INNE VON
B
ERG
, G., A. S
AMOYLOV
,M.K
LAAS
,
AND
P. H
ANELT
.
1996. Chloroplast DNA restriction analysis and the infrageneric
grouping of Allium (Alliaceae). Pl. Syst. Evol. 200: 253–261.
M
C
N
EAL
,D.W.,
AND
T. D. J
ACOBSEN
. 2002. Allium L., pp. 224–276.
In Flora of North America Editorial Committee [eds.], Flora of
North America, Vol. 26. Oxford University Press, New York,
USA.
M
AASS
, H. I. 1992. Electrophoretic study of storage proteins in the
genus Allium L., pp. 183–189. In P. Hanelt, K. Hammer, and H.
Knu¨ pffer [eds.], The genus Allium: taxonomic problems and ge-
netic resources. Proceedings of an international symposium held
at Gatersleben, Germany, 11–13 Jun 1991. Institut fu¨r Pflanzen-
genetik und Kulturpflanzenforschung, Gatersleben, Germany.
M
ATHEW
, B. 1996. A review of Allium section Allium. Royal Botanic
Gardens, Kew, Richmond, Surrey, UK. 176 p.
M
ES
, T. H. M., N. F
RIESEN
,R.M.F
RITSCH
,M.K
LAAS
,
AND
K. B
ACH
-
MANN
. 1997. Criteria for sampling in Allium based on chloroplast
DNA PCR-RFLPs. Syst. Bot. 22: 701–712.
,R.M.F
RITSCH
,S.P
OLLNER
,
AND
K. B
ACHMANN
. 1999. Evo-
lution of the chloroplast genome and polymorphic ITS regions in
Allium subgenus Melanocrommyum. Genome 42: 237–247.
N
ODA
, S.,
AND
S. K
AWANO
. 1988. The biology of Allium monanthum
(Liliaceae), Vol. 1. Polyploid complex and variations in karyo-
type. Pl. Spec. Biol. 3: 13–26.
N
OYES
, R. D.,
AND
L. H. R
IESEBERG
. 1999. ITS sequence data support
a single origin for North American Astereae (Asteraceae) and re-
flect deep geographic division in Aster s.l. Amer. J. Bot. 86: 398–
412.
O
HRI
, D., R. M. F
RITSCH
,
AND
P. H
ANELT
. 1998. Evolution of genome
size in Allium (Alliaceae). Pl. Syst. Evol. 210: 57–86.
P
ICH
, U., R. F
RITSCH
,
AND
I. S
CHUBERT
. 1996. Closely related Allium
species (Alliaceae) share a very similar satellite sequence. Pl. Syst.
Evol. 202: 255–264.
P
OSADA
, D.,
AND
K. A. C
RANDALL
. 1998. MODELTEST: testing the
model of DNA substitution. Bioinformatics 14: 817–818.
R
EGEL
, E. 1875. Alliorum adhuc cognitorum monographia. Trudy
Imper. S.-Peterburgsk. Bot. Sada 3: 1–266.
R
ONQUIST
,F.,
AND
J. P. H
UELSENBECK
. 2003. MrBayes 3: Bayesian
phylogenetic inference under mixed models. Bioinformatics 19:
1572–1574.
S
AMOYLOV
, A., N. F
RIESEN
,S.P
OLLNER
,
AND
P. H
ANELT
. 1999. Use
of chloroplast polymorphisms for the phylogenetic study of Allium
subgenera Amerallium and Bromatorrhiza (Alliaceae) II. Feddes
Repert. 110: 103–109.
,H.K
LAAS
,
AND
P. H
ANELT
. 1995. Use of chloroplast poly-
morphisms for the phylogenetic study of subgenera Amerallium
and Bromatorrhiza (genus Allium). Feddes Repert. 106: 161–167.
S
TEARN
, W. T. 1955. Allium bulgaricum. Bot. Mag. 170: plate 257
and text.
. 1978. European species of Allium and related genera of
Alliaceae: a synonymic enumeration. Ann. Mus. Goulandris 4:
83–198.
. 1992. How many species of Allium are known? Kew Mag.
9: 180–182.
S
WOFFORD
, D. L. 2002. PAUP*: phylogenetic analysis using parsi-
mony (*and other methods), vers. 4.0. Sinauer Associates, Inc.,
Sunderland, Massachusetts, USA.
,G.J.O
LSEN
,P.J.W
ADDELL
,
AND
D. M. H
ILLIS
. 1996. Phy-
logenetic inference, pp. 407–514. In D. M. Hillis, C. Moritz, and
B. K. Mable [eds.], Molecular systematics, Ed. 2. Sinauer Asso-
ciates, Inc., Sunderland, Massachusetts, USA.
T
AKHTAJAN
, A. L. 1987. Systema Magnoliophytorum. Nauka, Len-
ingrad, Russia. 438 p. [In Russian.]
. 1997. Diversity and classification of flowering plants. Co-
lumbia University Press, New York, USA. 643 p.
T
HOMPSON
, J. D., T. J. G
IBSON
,F.P
LEWNIAK
,F.J
EANMOUGIN
,
AND
D.
G. H
IGGINS
. 1997. The CLUSTAL
㛮
X windows interface: flexible
strategies for multiple sequence alignment aided by quality analy-
sis tools. Nucl. Acids Res. 25: 4876–4882.
T
RAUB
, H. P. 1968. The subgenera, sections and subsections of Al-
lium L. Pl. Life (Stanford) 24: 147–163.
VAN
R
AAMSDONK
,L.W.D.,W.E
NSINK
,A.W.
VAN
H
EUSDEN
,M.
V
RIELINK
-
VAN
G
INKEL
,
AND
C. K
IK
. 2003. Biodiversity assessment
based on cpDNA and crossability analysis in selected species of
Allium subgenus Rhizirideum. Theor. Appl. Genet. 107: 1048–
1058.
,M.V
RIELINK
-
VAN
G
INKEL
,
AND
C. K
IK
. 2000. Phylogeny
reconstruction and hybrid analysis in Allium subgenus Rhiziri-
deum. Theor. Appl. Genet. 100: 1000–1009.
V
VEDENSKY
, A. I. 1935. Rod 267. Luk—Allium L., pp. 112–280. In
V. L. Komarov [ed.], Flora URSS, Vol. 4. Izd. Akad. Nauk SSSR,
Leningrad, Russia.

394 ALISOFriesen, Fritsch, and Blattner
Appendix 1. Taxonomic conspectus of the genus Allium L.
First Evolutionary Line
1. Subgen. N
ECTAROSCORDUM
(Lindl.) Asch. et Graebn.—TYPE: A. siculum Ucria (ca. 3 spp.).
1.1. Sect. N
ECTAROSCORDUM
(Lindl.) Gren. et Godr.—TYPE: A. siculum Ucria.
2. Subgen. Microscordum (Maxim.) N. Friesen.—TYPE: A. monanthum Maxim. (monotypic).
2.1. Sect. M
ICROSCORDUM
Maxim.—TYPE: A. monanthum Maxim.
3. Subgen. A
MERALLIUM
Traub.—TYPE: A. canadense L. (ca. 135 spp.).
3.1. Sect. A
MERALLIUM
Traub.—TYPE: A. canadense L. (2 subsects.).
3.2. Sect. L
OPHIOPRASON
Traub.—TYPE: A. sanbornii Wood.
3.3. Sect. R
HOPHETOPRASON
Traub.—TYPE: A. glandulosum Link et Otto.
3.4. Sect. A
RCTOPRASUM
Kirschl. (Ophioscorodon (Wallr.) Endl.).—TYPE: A. ursinum L.
3.5. Sect. B
RISEIS
(Salisb.) Stearn.—TYPE: A. triquetrum L.
3.6. Sect. B
ROMATORRHIZA
Ekberg.—TYPE: A. wallichii Kunth.
3.7. Sect. C
AULORHIZIDEUM
Traub.—TYPE: A. validum S. Watson
3.8. Sect. C
HAMAEPRASON
Herm.—TYPE: A. chamaemoly L.
3.9. Sect. M
OLIUM
G. Don ex Koch.—TYPE: A. roseum L. (2 subsects.).
3.10. Sect. N
ARKISSOPRASON
(F. Herm.) Kamelin.—TYPE: A. narcissiflorum Vill.
? 3.11. Sect. R
HYNCHOCARPUM
Brullo.—TYPE: A. ruhmerianum Asch.
? 3.12. Sect. T
RIPTERA
Kamelin et Seisums, nom. nud.—TYPE: A. tripterum Nasir.
Second Evolutionary Line
4. Subgen. C
ALOSCORDUM
(Herb.) R. M. Fritsch.—TYPE: A. neriniflorum (Herb.) Baker (ca. 3 spp.).
4.1. Sect. C
ALOSCORDUM
(Herb.) Baker.—TYPE: A. neriniflorum (Herb.) Baker.
5. Subgen. Anguinum (G. Don ex Koch) N. Friesen.—TYPE: A. victorialis L. (ca. 12 spp.).
5.1. Sect. A
NGUINUM
G. Don ex Koch.—TYPE: A. victorialis L.
6. Subgen. Porphyroprason (Ekberg) R. M. Fritsch.—TYPE: A. oreophilum C. A. Mey. (monotypic).
6.1. Sect. P
ORPHYROPRASON
Ekberg.—TYPE: A. oreophilum C. A. Mey.
7. Subgen. Vvedenskya (Kamelin) R. M. Fritsch.—TYPE: A. kujukense Vved. (monotypic).
7.1. Sect. V
VEDENSKYA
Kamelin.—TYPE: A. kujukense Vved.
8. Subgen. M
ELANOCROMMYUM
(Webb et Berth.) Rouy.—TYPE: A. nigrum L. (ca. 140 spp.).
8.1. Sect. M
ELANOCROMMYUM
Webb et Berth.—TYPE: A. nigrum L.
8.2. Sect. A
CANTHOPRASON
Wendelbo.—TYPE: A. akaka Gmel.
8.3. Sect. A
CAULE
R. M. Fritsch.—TYPE: A. hexaceras Vved.
8.4. Sect. A
CMOPETALA
R. M. Fritsch.—TYPE: A. backhousianum Regel (7 subsects.).
8.5. Sect. A
ROIDEA
F. O. Khassanov et R. M. Fritsch.—TYPE: A. aroides Vved. et Popov.
8.6. Sect. B
REVICAULE
R. M. Fritsch.—TYPE: A. sergii Vved.
8.7. Sect. C
OMPACTOPRASON
R. M. Fritsch.—TYPE: A. giganteum Regel (3 subsects.).
8.8. Sect. K
ALOPRASON
C. Koch.—TYPE: A. caspium (Pall.) M. Bieb. (4 subsects.).
8.9. Sect. M
EGALOPRASON
Wendelbo.—TYPE: A. rosenbachianum Regel (3 subsects.).
8.10. Sect. M
INIPRASON
R. M. Fritsch.—TYPE: A. karataviense Regel.
8.11. Sect. P
OPOVIA
F. O. Khassanov et R. M. Fritsch.—TYPE: A. gypsaceum Popov et Vved.
8.12. Sect. P
SEUDOPRASON
(Wendelbo) K. M. Perss. et Wendelbo.—TYPE: A. koelzii (Wendelbo) K. M. Perss.
et Wendelbo.
8.13. Sect. R
EGELOPRASON
Wendelbo.—TYPE: A. regelii Trautv. (2 subsects.).
8.14. Sect. T
HAUMASIOPRASON
Wendelbo.—TYPE: A. mirum Wendelbo.
8.15. Sect. V
ERTICILLATA
Kamelin.—TYPE: A. verticillatum Regel.
Third Evolutionary Line
9. Subgen. Butomissa (Salisb.) N. Friesen.—TYPE: A. ramosum L. (ca. 4 spp.).
9.1. Sect. B
UTOMISSA
(Salisb.) Kamelin.—TYPE: A. ramosum L.
9.2. Sect. Austromontana N. Friesen.—TYPE: A. oreoprasum Schrenk.
10. Subgen. Cyathophora (R. M. Fritsch) R. M. Fritsch.—TYPE: A. cyathophorum Bur. et Franch. (ca. 5 spp.).
10.1. Sect. C
YATHOPHORA
R. M. Fritsch.—TYPE: A. cyathophorum Bur. et Franch.
10.2. Sect. C
OLEOBLASTUS
Ekberg.—TYPE: A. mairei Le´vl.
10.3. Sect. M
ILULA
(Prain) N. Friesen.—TYPE: A. spicatum (Prain) N. Friesen.
11. Subgen. R
HIZIRIDEUM
(G. Don ex Koch) Wendelbo s.s.—TYPE: A. senescens L. (ca. 37 spp.).
11.1. Sect. R
HIZIRIDEUM
G. Don ex Koch s.s.—TYPE: A. senescens L.
11.2. Sect. C
AESPITOSOPRASON
N. Friesen.—TYPE: A. polyrrhizum Siev.
11.3. Sect. R
HIZOMATOSA
Egor.—TYPE: A. caespitosum Siev.
11.4. Sect. T
ENUISSIMA
(Tzagolova) Hanelt.—TYPE: A. tenuissimum L.
11.5. Sect. Eduardia N. Friesen.—TYPE: A. eduardii Stearn.

VOLUME 22 395Phylogeny of Allium
Appendix 1. Continued.
12. Subgen. A
LLIUM
.—TYPE: A. sativum L. (ca. 300 spp.).
12.1. Sect. A
LLIUM
(including sect. C
OSTULATAE
F. O. Khassanov et S. C. Yengalycheva).—TYPE: A. sativum L.
12.2. Sect. B
REVIDENTIA
F. O. Khassanov et S. C. Yengalycheva.—TYPE: A. brevidens Vved.
12.3. Sect. C
RYSTALLINA
F. O. Khassanov et S. C. Yengalycheva.—TYPE: A. crystallinum Vved.
12.4. Sect. M
ULTICAULEA
F. O. Khassanov et S. C. Yengalycheva.—TYPE: A. lehmannianum Merckl.
12.5. Sect. S
PATHULATA
F. O. Khassanov et R. M. Fritsch.—TYPE: A. spathulatum F. O. Khassanov et R. M.
Fritsch.
12.6. Sect. Mediasia F. O. Khassanov, S. C. Yengalycheva et N. Friesen.—TYPE: A. turkestanicum Regel.
12.7. Sect. A
VULSEA
F. O. Khassanov.—TYPE: A. rubellum M. Bieb.
12.8. Sect. B
REVISPATHA
Valsecchi.—TYPE: A. parciflorum Viv.
12.9. Sect. C
AERULEA
(Omelcz.) F. O. Khassanov.—TYPE: A. caeruleum Pall.
12.10. Sect. C
ODONOPRASUM
Rchb.—TYPE: A. oleraceum L. (2 subsects.).
12.11. Sect. Eremoprasum (Kamelin) F. O. Khassanov, R. M. Fritsch et N. Friesen.—TYPE: A. sabulosum Stev.
12.12. Sect. K
OPETDAGIA
F. O. Khassanov.—TYPE: A. kopetdagense Vved.
12.13. Sect. Longivaginata (Kamelin) F. O. Khassanov, R. M. Fritsch et N. Friesen.—TYPE: A. longivaginatum
Wendelbo.
12.14. Sect. M
INUTA
F. O. Khassanov.—TYPE: A. minutum Vved.
12.15. Sect. Pallasia (Tzagolova.) F. O. Khassanov, R. M. Fritsch et N. Friesen.—TYPE: A. pallasii Murr.
13. Subgen. Reticulatobulbosa (Kamelin) N. Friesen.—TYPE: A. lineare L. (ca. 80 spp.).
13.1. Sect. R
ETICULATOBULBOSA
Kamelin s.s.—TYPE: A. lineare L.
13.2. Sect. C
AMPANULATA
Kamelin.—TYPE: A. xiphopetalum Aitch.
13.3. Sect. Scabriscapa (Tscholok.) N. Friesen.—TYPE: A. scabriscapum Boiss.
13.4. Sect. Nigrimontana N. Friesen.—TYPE: A. drobovii Vved.
13.5. Sect. Sikkimensia (Traub) N. Friesen.—TYPE: A. sikkimense Baker.
14. Subgen. P
OLYPRASON
Radic´ .—TYPE: A. moschatum L. (ca. 50 spp.).
14.1. Sect. S
CORODON
Koch s.s.—TYPE: A. moschatum L.
14.2. Sect. O
REIPRASON
F. Herm. (including sect. Petroprason F. Herm.).—TYPE: A. saxatile M. Bieb.
14.3. Sect. Falcatifolia N. Friesen.—TYPE: A. carolinianum DC.
14.4. Sect. Daghestanica (Tscholok.) N. Friesen.—TYPE: A. daghestanicum Grossh. (2 subsects.).
15. Subgen. C
EPA
(Mill.) Radic´ .—TYPE: A. cepa L. (ca. 30 spp.).
15.1. Sect. C
EPA
(Mill.) Prokh.—TYPE: A. cepa L.
15.2. Sect. A
NNULOPRASON
T. V. Egorova.—TYPE: A. fedtschenkoanum Regel.
15.3. Sect. Condensatum N. Friesen.—TYPE: A. condensatum Turcz.
15.4. Sect. S
ACCULIFERUM
P. P. Gritz.—TYPE: A. thunbergii G. Don (A. sacculiferum Maxim.).
15.5. Sect. S
CHOENOPRASUM
Dumort.—TYPE: A. schoenoprasum L.