The precise taxonomy of the genusAllium is still poorly understood with incorrect descriptions being widespread. With over 850 species distributed over the Northern hemisphereAllium is the sole genus in the Allieae, one of four tribes of subfamilyAllioideae (Amaryllidaceae). New species continue to be described[1] and Allium is both highly variable and one of the largest monocotyledonous genera,[2] but the precise taxonomy of Allium is poorly understood,[2][1] with incorrect descriptions being widespread. The difficulties arise from the fact that the genus displays considerable polymorphism and has adapted to a wide variety of habitats. Furthermore, traditional classications had been based on homoplasious characteristics (what turn out to be independently evolved similar features in species from different lineages). However, the genus has been shown to be monophyletic, containing three major clades, although some proposed subgenera are not.[2] Some progress is being made using molecular phylogenetic methods, and the internal transcribed spacer (ITS) region, including the 5.8S rDNA and the two spacers ITS1 and ITS2, is one of the more commonly used markers in the study of the differentiation of the Allium species.[1]
Allium includes a number of taxonomic groupings previously considered separate genera (Caloscordum Herb., Milula Prain and Nectaroscordum Lindl.) Allium spicatum had been treated by many authors as Milula spicata, the only species in the monospecific genus Milula. In 2000, it was shown to be embedded in Allium.[3]
Descriptions of Allium taxonomy date back at least as far as Carolus Clusius' Rariarum plantarum historia (1601).[6] When Linnaeus[7] formerly described the genus Allium in his Species Plantarum (1753), there were thirty species with this name. He placed Allium in a grouping he referred to as Hexandria monogynia[8] (i.e. six stamens and one pistil) containing 51 genera in all.[9] In 1763, Michel Adanson, who proposed the concept of families of plants, included Allium and related genera as a grouping within 'Liliaceae'[10] as Section IV, Les Oignons (Onions), or Cepae in Latin.[11]De Jussieu is officially recognised as the first formal establishment of the suprageneric grouping into families (Ordo) in 1789. In this system Allium was one of fourteen genera in Ordo VI, Asphodeles (Asphodeli), of the third class (Stamina epigyna) of Monocots.[12]Jean Henri Jaume Saint-Hilaire (1805), who developed the concept of Amaryllidaceae, continued Jussieu's treatment of Allium under Asphodeli (which he considered synonymous with Adanson's Liliaceae and Jussieu's Asphodeli).[13] He placed Allium in an unnamed monotypic section of Asphodeli defined as Fleurs en ombelle, racine bulbeuse. Calice à six parties egales (umbellate flowers, bulbous, calyx of six equal parts).[14]
Subsequently, de Candolle reverted the family name back to Liliaceae from Asphodeli.[15] He divided the Liliaceae into a series of Ordres, and the second ordre was named Asphodèles, based on Jussieus' family of that name,[16] in which he placed Allium.[16] The term 'Alliaceae' then reappeared in its subfamilial form, Allieae, in Dumortier's Florula Belgica (1827),[17] with six genera. The 'Alliaceae' have been treated as Allieae within the family Liliaceae (or Aspholecaceae, a partial synonym) by most authorities since.
Regel produced a major monograph of the genus in 1875,[18] and this remained the major reference work for over 100 years till the molecularly based study of Friesen and colleagues in 2006.[19] Despite recent advances the precise taxonomy of Allium remains still poorly understood with incorrect descriptions being widespread.[20][2][1]
Linnaeus originally grouped his 30 species into three alliances, e.g.Foliis caulinis planis and as the number of recognised species increased, so did the number of subgroups. Since then, many attempts have been made to divide the growing number of recognised species into infrageneric subgroupings, initially as sections, and then as subgenera further divided into sections. For a brief history, see Friesen et al. (2006)[19] and Li et al. (2010)[2] Regel's 1875 treatise on Allium divided his 262 species between the six sections proposed by Don, in his 1832 monograph on the genus.[21]Stearn (1944) described 14 subgenera.[22]Traub (1968) described 3 subgenera, 36 sections and subsections and about 600 species.[23] By 1992 there were 6 sub-genera, 50 sections and subsections and 600–700 species.[24] The situation was further confused by the presence of over 1,000 taxonomic names, many of which turned out to be synonyms.[25][20]
The modern era of phylogenetic analysis dates to 1996.[26] In 2006 Friesen, Fritsch, and Blattner[19] described a new classification with 15 subgenera, 56 sections, and about 780 species based on the nuclearribosomalgeneinternal transcribed spacers. Some of the subgenera correspond to the once separate genera (Caloscordum, Milula, Nectaroscordum) included in the Gilliesieae.[2][27] The terminology has varied with some authors subdividing subgenera into Sections and others Alliances.
The term alliance has been used for both subgroupings within species, e.g.Allium nigrum, as well as infrageneric subsections. These alliances are informal groupings based on morphological similarity and reflecting hypotheses of evolutionary relationship. and can be used between any two formal ranks.[28] For instance the some 70 North American species were divided into nine well-defined species alliances, of which the largest was the Allium falcifolium alliance with 31 taxa.[29] These alliances are usually referred to as the Ownbey alliances, after Marion Ownbey[20] and were also used by Traub.[23] A number of classification schemes have chosen to retain these, the Traub system not being universally accepted.[19][5]
Subsequent molecular phylogenetic studies have shown the 2006 classification is a considerable improvement over previous classifications, but some of its subgenera and sections are probably not monophyletic. Meanwhile, the number of new species continued to increase, reaching 800 by 2009, 900 by 2016[30] and the pace of discovery has not decreased. Detailed studies have focused on a number of subgenera, including Amerallium. Amerallium is strongly supported as monophyletic.[31] Subgenus Melanocrommyum has also been the subject of considerable study (see below), while work on subgenus Allium has focused on section Allium, including Allium ampeloprasum, although sampling was not sufficient to test the monophyly of the section.[32]
The major evolutionary lineages or lines correspond to the three major clades. Line one (the oldest) with three subgenera is predominantly bulbous, the second, with five subgenera and the third with seven subgenera contain both bulbous and rhizomatous taxa.[2] Banfi and colleagues (2011) have suggested that the phylogenetic trichotomy of this genus Alliumsensu lato is sufficiently distinct as to warrant splitting it into three separate genera. Banfi's scheme thus proposes the restoring the three originally separate genera Nectaroscordum Lindl. (type: N. siculum), Caloscordum Herb. (type: C. neriniflorum) and Allium L. sensu stricto (type: A. sativum) to correspond to lines 1-3.[33]
Evolutionary lines and subgenera
The three evolutionary lineages and 15 subgenera here represent the classification schemes of Friesen et al. (2006)[19] and Li (2010),[2] and subsequent additional species[34][31][35][30] and revisions.[36]
Although this lineage consists of three subgenera, nearly all the species are attributed to subgenus Amerallium, the third largest subgenus of Allium. The lineage is considered to represent the most ancient line within Allium, and to be the only lineage that is predominantly bulbous, the other two having both bulbous and rhizomatous taxa. Nectaroscordum and Microscordum are bulbous, but Amerallium contains some rhizomatous elements. Within this lineage Amerallium is a sister group to the other two subgenera (Microscordum+Nectaroscordum).[2]
Subgenus Nectaroscordum
Type species: Allium siculum
Disjunct distribution, involving the western Mediterranean (type species) and southwest Asia
Section Nectaroscordum (Lindl.) Gren. & Godr.
Allium siculum Ucria (Syn. Nectaroscordum siculum (Ucria) Lindl.) Type
This large monophyletic subgenus is extremely diverse, both morphologically and ecologically and is characterised by leaves with one row of vascular bundles, absence of palisade parenchyma and a subepidermal position of laticifers, with a predominant base chromosome number x=7.[2]
Taxonomy
Amerallium is a relatively large subgenus with about 120–140 species. Under the alliance system of classification proposed by Ownbey (1966), species north of Mexico and two Mexican endemics were treated as eight informal alliances: the A. acuminatum, A. campanulatum, A. canadense, A. cernuum, A. falcifolium, A. kunthii, A. sanbornii, and A. validum alliances.[20] Traub (1968) then arranged the New World alliances into four sections: Amerallium Traub, Caulorhizideum Traub, Lophioprason Traub, and Rhopetoprason Traub. In addition he arranged the Old World species into 6 sections.[23]
Since Traub's revision of the subgenus, two biogeographical sister clades (or alliances) have been recognised. The Old World clade is represented by two relatively small groups from the Mediterranean and East Asia. The larger New World clade by all North American species of Allium. The New World sections are Lophioprason, Amerallium, and Rhophetoprason, while the Old World is represented by sections Arctoprasum, Briseis, Narkissoprason, Molium, Bromatorrhiza and Rhynchocarpum.[19]
The subgenus is thought to originate in the Old World, with a later split, and to have its origin in the higher latitudes of East Asia, at the time of transition from Cretaceous to Tertiary, dispersing to western North America. Twelve sections were subsequently recognized, with sections Amerallium and Molium further split into two subsections.[19][2][5]
Distribution
Amerallium is widely distributed within North America, Europe, north Africa, Ethiopia, the Caucasus, northern Iran, southeast Tibet, and southwest China.[5] The greatest species diversity occurs in North America with 81 species recognized in the 2002 Flora of North America (north of Mexico)[38] and a further 13 are unique to Mexico,[39] and a total of 26 species recognised there.[5] Within N America, the genus covers most of the area south of the 53rd parallel, including the oak hillsides of California and Oregon, deserts of Nevada and Texas, alpine meadows of Utah and Idaho, prairies of Nebraska and Manitoba, and forest glades of Missouri and Arkansas.[5]
Nearly all the species in this lineage of five subgenera are accounted for by subgenus Melanocrommyum, which is most closely associated with subgenera Vvedenskya and Porphyroprason, phylogenetically. These three genera are late-branching whereas the remaining two subgenera, Caloscordum and Anguinum are early branching.[2] Of the five subgenera, the large Melanocrommymum and the oligo- or monotypic Caloscordum, Vvedenskya and Porphyroprason are bulbous and the remaining small subgenus Anguinum is rhizomatous.[28]
This Eurasian subgenus, the second largest, is complex and has had a confusing taxonomic history and is extremely diverse, morphologically. It is distributed from the Mediterranean to the Near and Middle East, to north-western China and Pakistan in the east, and southern Siberia in the north. The centre of diversity is Central Asia where it evolved, but its ancestry is located in East Asia.[2] The 2006 classification of Fritsch and colleagues included 150 species but this has continued to grow. An extensive molecular based study in 2010 confirmed its monophyly but showed that the traditional sections were either para- or polyphyletic. On the other hand, a number of monophyletic subgroups were recognised, with about 40 clades, although their exact relationships remained not fully resolved. Consequently, traditional sections required considerable re-alignment. Eventually 160 species and subspecies were recognised in 20 sections and 22 subsections.[1][28][40]
Description: The subgenus is characterised by true tunicated bulbs, annual roots, leaves that are mostly broad and flat with subterranean sheath parts that are barely visible above the ground, scapes that are strong and most often strictly upright and of varying length, and large, fasciculate to globular inflorescences. The latter are composed of many moderately small to large, often star-like, flowers, and some of which have a sweet or noticeable odor.[28][2]
History: Early (prior to 1950) classifications of Allium included many of the members of this subgenus within the bulbous section, Mollium based on morphological characteristics. Mollium was later raised to subgenus level (and then again reduced to a section of Amerallium after transferring many species to Melanocrommyum. The subgenus was then divided into sections in 1969. Molecular methods in the 1990s confirmed the identity of Melanocrommyum as a distinct monophyletic group, together with the presence of several subgroups, but the deeper relationships remained inconsistent.[28]
Subdivision: Subdivision of the subgenus was first proposed by Wendelbo in 1966,[41] proposing section Regeloprason Wendelbo, followed in 1969 by Melanocrommyum Webb & Berthel., Kaloprason K. Koch, Acanthoprason Wendelbo, Megaloprason Wendelbo, and Thaumasioprason Wendelbo.[42] Kamelin (1973) provided an alternative arrangement of sections,[43] which was supplanted by the Gatersleben Allium Group classification (1992) which used a broad range of variables.[24]
The use of molecular markers to develop phylogeny began in the 1990s and showed that the subgenus was a well separated taxon with a number of subgeneric groupings. Friesen and colleagues (2006) carried out an extensive molecular phylogenetic study [19] resulting in a taxonomy based on 15 sections. These were then further subdivided into five of the sections to create 17 subsections. While Melanocrommyum itself appeared monophyletic, most of the sections were either para- or polyphyletic, favouring the formation of a larger number of smaller subgroups. In their study there were a number of larger sections with 15–35 species, Acmopetala, Megaloprason, Regeloprason, Kaloprason, and Acanthoprason. The remaining sections are either oligogotypic with 2–8 species (Compactoprason, Pseudoprason, Miniprason, Brevicaule, Thaumasioprason, Verticillata) or monotypic (Acaule, Aroidea, Popovia).
In a more focused study in 2010 this was expanded to 20 sections and 22 subsections, or in some cases, e.g. section Melanocrommyum (type: A. nigrum) into nine alliances and Acanthoprason into seven.[28] This section is the most diverse one within the subgenus in which subgroups differ according to the relationships of the lengths of leaves and scapes (leaves shorter, equal or longer than scapes) and inflorescences (fasciculate, umbellate or subglobose).[44] The increased number of sections resulted from the splitting of some of the earlier sections, such as Acmopetala. The two species in the resulting section Longibidentata are sister to all the remaining sections (core clade). This section, together with another new section, Decipientia form the basal clade. Although Li et al. (2010) included three sections, their study was confined to species endemic to China.[2]
Uses: The subgenus also contains many species grown as ornamentals, such as A. giganteum, A. cristophii, A. schubertii, A. aflatunense, A. atropurpureum, A. nigrum and A. karataviense. These species are predominantly from Southwestern and Central Asia, where they are used for both culinary and medicinal purposes. The latter usage is associated with the presence of cysteinesulphoxides and also radical scavenger activity, although many members of the subgenus possess less of these compounds and lack the distinctive taste and smell of garlic and onion, their properties appear to be associated with dithiodipyrroles and sulfur-pyridins. These substances also occur in the ornamental species, that were introduced into European and North American gardens in the early 19th century, and now are represented by an increasing number of named cultivars and hybrids. Cysteine sulphoxides are also largely responsible for the flavour and spicy taste of these species, predominantly the isomeric cysteine derivatives alliin and isoalliin.[44]
The third evolutionary line contains the most number of subgenera (seven) and also the largest subgenus of the genus Allium, subgenus Allium which includes the type species of the genus, Allium sativum. This subgenus also contains the majority of the species in the line. Within the lineage the phylogeny is complex. Two small subgenera Butomissa and Cyathophora form a sister clade to the remaining five subgenera, with Butomissa as the first branching group. Amongst the remaining five subgenera, Rhizirideum forms a medium-sized subgenus that is the sister to the other four larger subgenera. However, they may not be monophyletic.[2] Of the seven subgenera, the large subgenus Allium, represents the bulbous elelement.[28]
~ 37 species. Eurasian steppes, with greatest diversity in southern Siberia and Mongolia. Only a few species distributed in Europe, with Portugal as most western point. Some species occur also in Korea and far eastern Russia, and one in Japan.
Subgenus Allium, the youngest of the subgenera, is predominantly Mediterranean but its distribution extends east towards Central Asia. This very large subgenus is divided into 15 - 16 sections[a] and demonstrates two main groups. One has been referred to as classicalAllium with tripartite inner filaments and only one thick storage cataphyll. The other is more diverse morphologically reflected in less closely related sections. A number of sections appear to be non-monophyletic, including Avulsea, Pallasia, Brevispatha and Kopetdagia.[32] It includes both ornamentals, such as A. sphaerocephalon, A. caeruleum, A. carinatum and A. flavum as well as food crops such as A. sativum and A. ampeloprasum.
16 Sections
Section Allium Type A. sativum
Section Avulsea Type A. rubellum
Section Brevidentia Type A. brevidens
Section Brevispatha Type A. parciflorum
Section Caerulea Type A. caeruleum
Section Codonoprasum Type A. oleraceum
Section Costulatae Type A. filidens
Section Crystallina Type A. crystallinum
Section Eremoprasum Type A. sabulosum
Section Kopetdagia Type A. kopedeganse
Section Longivaginata Type A. longivaginatum
Section Minuta Type A. minutum
Section Mediasia Type A. turkestanicum
Section Multicaulea Type A. lehmannianum
Section Pallasia Type A. pallasii
Section Spathulata Type A. spathulatum
Sections
Section Allium
This is the largest section with about 114 species, a number of which are economically important, such as A. sativum (garlic) and A. ampeloprasum (leek). This section also expresses frequent polyploidy and contains a number of species whose boundaries have been difficult to establish, notably A. ampeloprasum which includes a number of subspecies and varieties, as well as synonymous species, which have been labelled the "A. ampeloprasum complex". Horticulturally, it is represented by at least four groups, including leeks, whose exact ancestry has been considered uncertain. In the molecular phylogenetic study by Hirschegger and colleagues (2010) showed section Allium to be a well supported clade with two main subclades, one of which included two smaller clades. All of the tetraploid forms of A. ampeloprasum were resolved in a single clade, and leeks appeared to be more closely allied to A. iranicum and A. atroviolaceum than A. ampeloprasum. Restoration of A. porrum L. was therefore proposed for the tertraploid forms, reserving A. ampeloprasum for the forms known horticulturally as great headed garlic and A. ampeloprasum var. babingtonii.[32]
Section Codonoprason is strongly monophyletic[19][45] and has its centre of diversity in the Mediterranean region, particularly Greece and Turkey, but extends to other areas of Europe, North Africa and the Middle East. It was originally conceived of as a separate genus, Codonoprasum by Reichenbach in 1828.[46] The taxonomy of the section is complicated with inconsistent speciation. In 2005 the section was considered to consist of 58 species and 7 subspecies.[47] The section is characterised as large plants with multiflowered inflorescences, long pedicels, very long spathe valves and a cylindrical-campanulate perigon, with unequal and long-caudate spathe leaves.[48][45]
Historically the section has been considered to have a number of subsections. Friesen recognised 2,[19] while others have described 3, e.g.;[49]
Codonoprasum (Rchb.) Kamelin
Longistamineum Cheshm. ex Omelczuk
Haemoprason (F. Herm.) Cheshm.
Many species were previously included in a grouping referred to as the Paniculatum complex. Molecular studies demonstrate the presence of two clades within the section. Clade A contains the two autumn flowering species, Allium tardans and Allium parciflorum as a subclade. Clade B contains the smaller taxa within the section.
The name Allium is ancient, and the plant was known to both the Romans and the Greeks.[51] The name is thought to be Celtic in origin, meaning "to burn", in reference to its taste and smell.[52] One of the earliest uses of the name in botany, was by Joseph Pitton de Tournefort (1656- 1708).
Notes
^Friesen et al. include section Costulatae within section Allium, while others treat it separately
^The history of A. macrostemon is complex. Historically the species was included in subgenus Allium section Scorodons.l.. When the species in the section were redistributed by Friesen et al. 2006[19] they were unable to allocate this species to a section and questioned whether a new section should be created. Nguyen et al. (2008)[31] found it most closely related to species within section Codonoprason. In contrast Li et al. (2010)[2] placed it in section Allium and Choi et al. (2012)[50] believed it belonged in section Caerulea, but did not recommend moving it on the basis of available information. It is therefore unplaced within subgenus Allium or considered the sole member of section Scorodons.l.
^In older classifications, such as Hanelt (1992)[24] section Scorodon was placed in subgenus Allium. It was a relatively large section with several subsections. In the study of Friesen et al. 2006[19] the section was shown not to be monophyletic, and most species were allocated to other sections. One subsection segregated with subgenus Polyprason and was elevated to section Scorodon. To distinguish between the two, the older section is designated Scorodons.l. and the new section Scorodons.s.
^Whereas Friesen et al. 2006[19] treat A. taquetii as synonymous with the type species A. thunbergii, Choi and Oh 2011 present evidence as to why they should be treated separately[36]
Hanelt, P; Schulze-Motel, J; Fritsch, R M; Kruse, J; Maass, H; Ohle, H; Pistrick, K. Infrageneric grouping of Allium— the Gatersleben approach. pp. 107–123., in Hanelt et al. (1992)
Friesen, Nikolai; Fritsch, Reinhard M.; Pollner, Sven; Blattner, Frank R. (2000). "Molecular and Morphological Evidence for an Origin of the Aberrant Genus Milula within Himalayan Species of Allium (Alliacae)". Molecular Phylogenetics and Evolution. 17 (2): 209–218. doi:10.1006/mpev.2000.0844. PMID11083935.
Gurushidze, Maia; Fuchs, Jörg; Blattner, Frank R. (1 March 2012). "The Evolution of Genome Size Variation in Drumstick Onions (Allium subgenus Melanocrommyum)". Systematic Botany. 37 (1): 96–104. doi:10.1600/036364412X616675. S2CID85838253.
Rieseberg, Loren H.; Peterson, Paul M.; Soltis, Douglas E.; Annable, Carol R. (November 1987). "Genetic Divergence and Isozyme Number Variation Among Four Varieties of Allium douglasii (Alliaceae)". American Journal of Botany. 74 (11): 1614. doi:10.2307/2444130. JSTOR2444130.
Saghir, Abdur Rahman B.; Mann, Louis K.; Ownbey, Marion; Berg, Rolf Y. (May 1966). "Composition of Volatiles in Relation to Taxonomy of American Alliums". American Journal of Botany. 53 (5): 477–484. doi:10.2307/2440346. JSTOR2440346.
Stearn, W. T. (1944). "Notes on the genus Allium in the Old World; its distribution, names, literature, classification and garden-worthy species". Herbertia. 11: 11–34.
Huang, De-Qing; Yang, Jing-Tian; Zhou, Chun-Jing; Zhou, Song-Dong; He, Xing-Jin (21 December 2013). "Phylogenetic reappraisal of Allium subgenus Cyathophora (Amaryllidaceae) and related taxa, with a proposal of two new sections". Journal of Plant Research. 127 (2): 275–286. doi:10.1007/s10265-013-0617-8. PMID24362461. S2CID15531719.
Peterson, Paul M.; Annable, Carol R.; Rieseberg, Loren H. (April 1988). "Systematic Relationships and Nomenclatural Changes in the Allium douglasii Complex (Alliaceae)". Systematic Botany. 13 (2): 207. doi:10.2307/2419099. JSTOR2419099.
Samoylov, A.; Friesen, N.; Pollner, S.; Hanelt, P. (1999). "Use of chloroplast DNA polymorphisms for the phylogenetic study of Allium subgenus Amerallium and subgenus Bromatorrhiza (Alliaceae) II". Feddes Repertorium. 110 (1–2): 103–109. doi:10.1002/fedr.19991100118.
Wheeler, E. J.; Mashayekhi, S.; McNeal, D. W.; Columbus, J. T.; Pires, J. C. (26 March 2013). "Molecular systematics of Allium subgenus Amerallium (Amaryllidaceae) in North America". American Journal of Botany. 100 (4): 701–711. doi:10.3732/ajb.1200641. PMID23535771.
Line 2
Li, R. J.; Shang, Z. Y.; Cui, T. C.; Xu, J. M. (1996). "Studies on karyotypes and phylogenetic relationship of Allium sect. Caloscordum (Liliaceae) from China". Acta Phytotax. Sin. (in Chinese). 34: 288–295.
Gurushidze, Maia; Fritsch, Reinhard M; Blattner, Frank R (2010). "Species-level phylogeny of Allium subgenus Melanocrommyum: Incomplete lineage sorting, hybridization and trnF gene duplication". Taxon. 59 (3): 829–840. doi:10.1002/tax.593012.
Line 3
Dubouzet, J. G.; Shinoda, K.; Murata, N. (17 December 1997). "Phylogeny of Allium L. subgenus Rhizirideum (G. Don ex Koch) Wendelbo according to dot blot hybridization with randomly amplified DNA probes". Theoretical and Applied Genetics. 95 (8): 1223–1228. doi:10.1007/s001220050685. S2CID33083704.
Brullo, Salvatore; Guglielmo, Anna; Pavone, Pietro; Salmeri, Cristina; Terrasi, M. Carmen (January 2003). "Three new species of Allium sect. Codonoprasum from Greece". Plant Biosystems. 137 (2): 131–140. Bibcode:2003PBios.137..131B. doi:10.1080/11263500312331351391. S2CID86821437.
Salmeri, Cristina; Brullo, Cristian; Brullo, Salvatore; Galdo, Giampietro Giusso Del; Moysiyenko, Ivan I. (March 2016). "What is Allium paniculatum? Establishing taxonomic and molecular phylogenetic relationships within A. sect. Codonoprasum". Journal of Systematics and Evolution. 54 (2): 123–135. doi:10.1111/jse.12170. S2CID83671442.
Tzanoudakis, Dimitris; Trigas, Panayiotis (12 March 2015). "Allium occultum, a new species of A. sect. Codonoprasum (Amaryllidaceae) from Skiros Island (W Aegean, Greece)". Phytotaxa. 202 (2): 135. doi:10.11646/phytotaxa.202.2.5.
.jpg)
.jpg)
