Introduction

During the initial editing of the draft Flora of Iraq Compositae account, written by several authors some 35 or more years ago, the present author was faced with updating much taxonomy and sorting out many nomenclatural issues with the various scripts. One of these concerned Francis Davies’s original draft account of the genus Acroptilon Cass. (Compositae: Cardueae: Centaureinae), and the sole species in the Flora, A. repens (L.) DC.

Acroptilon repens is a widely distributed taxon, and a noxious weed, possibly best known as the Russian knapweed (Maddox et al. 1985; Moore 1972; Quattrocchi 2000: 38; USDA 2015; also known as creeping knapweed, hardheads, and Turkestan thistle — Klein 2011). Its toxicity is reported as causing Chewing Disease (nigropallidal encephalomalacia), a neurological disorder in horses, a response also seen to be caused by Centaurea solstitialis L. in the same subtribe (Anon 2015; Vetstream 2014).

It became apparent to the present author that the generic placement of Acroptilon was questionable, and brought with it nomenclatural implications. Further study indicated that Acroptilon repens belonged to what Susanna & Garcia-Jacas (2006) had termed the ‘III.5.B. Rhaponticum Group’, that in turn formed part of the subtribe Centaureinae. This group, of perhaps some 40 species, included the genera Acroptilon, Leuzea DC., ‘Rhaponticum Vaill.’ and Stemmacantha Cass. (Susanna & Garcia-Jacas 2006: 142 – 144; Hidalgo et al. 2006). Earlier, in his treatment of the subtribe Centaureinae, Bremer (1994: 149 – 156) recognised a monospecific Acroptilon (A. repens) and Stemmacantha (containing about 20 species); ‘Rhaponticum Hill’ was placed into the synonymy of the latter; two literature references were provided, the first to Holub (1973) and, the second, to Dittrich (1984).

Holub (1973) had long-recognised that within the Centaureinae there were taxonomic issues with both Rhaponticum sensu auct. and Leuzea, and that the work of Soskov (1959) and Dittrich (1968a, b) had scarcely resolved them, together with the clear nomenclatural issues that still existed. Holub concluded that Leuzea (based on L. conifera (L.) DC.) was congeneric with Cestrinus Cass., Stemmacantha, Fornicium Cass. and Malacophyllum Tausch along with several species belonging to Rhaponticum sensu auct., synonymising all, reducing the genera to subgenera of Leuzea and providing many new combinations. In his discussions on the nomenclature, he highlighted and summarised the many issues with the generic name Rhaponticum, concluding that it should only be ascribed to Ludwig, and now considered synonymous with Centaurea L.; the present author concurs with this view. Holub did not treat Acroptilon repens within his ‘Contribution’, mentioning only the common basic chromosome number of x= 13 amongst several genera within the Centaureinae, including Acroptilon.

The taxonomy of Holub’s and Dittrich’s (Dittrich 1968a, 1975, 1984) concepts of this group of species was subjected to analysis (of one plastid and two nuclear marker/s) in an attempt to resolve the delimitation of the ‘Rhaponticum group’ (Hidalgo et al. 2006). Hidalgo et al. provided a ‘DNA phylogeny of Rhaponticum and related genera’ from which a quite reasonable taxonomic resolution was presented. The ‘Rhaponticum group’ was clearly comprised of the concepts of Acroptilon, Leuzea and ‘Rhaponticum Vaill.’ (≡ Stemmacantha) and usefully divided into two clades that Hidalgo et al. informally recognised as the ‘Oriental clade’ and the ‘Occidental clade’; it was in the former that A. repens (and A. australe Iljin) was placed. Furthermore, since the authors could not entirely resolve the relationship between A. repens and A. australe they merely provided a new combination of the former under Rhaponticum. The implications surrounding A. australe are discussed below.

Ascription of the authorship of Rhaponticum to Vaillant (e.g. Greuter 2003; Greuter et al. 2005; Zhu & Martins 2011) fails to recognise the issue with Vaillant’s many names (see Brummitt 2008; Greuter 2008; Greuter & McNeill 2008; Hind et al. 2007; Susanna 2009). This can be summarised by essentially discounting a post-Linnaean translation (i.e. post-1753 — ‘Vaillant’ 1754) (of a pre-Linnaean work of Vaillant, dating from 1719) that many authors considered provided validation of Vaillant’s names. Holub (1973) had correctly recognised that Leuzea had priority and made many of the required combinations — apart from considering A. repens. Since Hidalgo et al. (2006: 709, 711) clearly placed Acroptilon within the concept of Rhaponticum, here recognised as Leuzea, a new combination is required This is provided below.

New combination

Following the new combination, a fairly exhaustive synonymy of the taxon is provided. After the citation of each name and its place of publication, the ‘type citation’ in a traditional sense is provided (as a direct quote from the protologue, regardless of whether a specimen was cited). The location of type material is cited, where known and, when available, the barcode numbers present with the material given. Many of the digital images of the ‘types’ can be seen through JStor Global Plants (JSTOR-GP continuously updated), institutional subscriptions permitting.

Leuzea repens (L.) D.J.N.Hind, comb. nov.

http://www.ipni.org/urn:lsid:ipni.org:names:60478474-2

Centaurea repens L., Sp. Pl., ed. 2. 2: 1293 (Linnaeus 1763). Type: ‘Habitat in oriente.’, LINN-LH 1030.34.

Centaurea picris Pall. ex Willd. (Willdenow 1803: 2302). Type: ‘Habitat ad mare Caspium. (v. s.)’. Type material: B-W(16601-01 0, 16601-02 0, 16601-03 0), BM(000996201). Note: Of the material in B-W none indicate the locality — B-W(16601-01 0) is marked ‘C. Picris 1.’ on the top right hand corner of the sheet, and was most probably determined by Marschall von Bieberstein as ‘Serratula picris MB’ on a label to the lower left of the specimen, and subsequently redetermined as Rhaponticum repens (L.) Hidalgo by Martins (Mai 2009); B-W(16601-02 0) is marked ‘C. Picris 2’ on the top right hand corner of the sheet, and has a small label at the bottom of the sheet — ‘Cent. Picris Pall./(W.)’ The sheet bears an annotation label by Cronquist (1989) suggesting this is ‘presumably the type of Centaurea picris Pallas ex Will. = C. repens L.’. On neither the first or second sheets are there any lower leaves with anything but entire basal margins. The third sheet, B-W(16601-03 0), is annotated ‘C. Picris 3’ on the top right hand corner of the sheet. It is labelled as ‘Centaurea pycris Pall./(Adams)’. This sheet bears some lower leaves that are ‘subdentatis’, as indicated in the protologue, but the leaves in general are much narrower than on the other two sheets. The material from BM is clearly ex Herb. Pallas, with the top left hand label indicating the locality as ‘Caspica’, and it is quite clear that the basal margins of many leaves are ‘subdentatis’, although more frequently conspicuously toothed to lobed. Since the Pallas herbarium was sold in 1808, and part purchased by Lambert and deposited in BM (Miller 1970 — Pallas material was discussed at length and was sold in the auction of Lambert’s herbarium in three lots, #94, #96 and #285.), it is highly probable that Willdenow had seen the material in Pallas’s herbarium in preparing his Species Plantarum (Willdenow 1803); the BM material should therefore not be discounted if lectotypification of the name is considered necessary.

Serratula picris (Pall. ex Willd.) M.Bieb. (Marschall von Bieberstein Dec. 1819 or early 1820: 546).

Acroptilon obtusifolium Cass. (Cassini 1827: 465). Type: ‘Nous avons fait cette description sur un échantillon sec de l’herbier de M. Desfontaines, étiqueté Centaurea picris, ...’ Holotype: ?P or FI.

Acroptilon subdentatum Cass. (Cassini 1827: 465). Type: see citation for A. serratum. Holotype: ?P.

Acroptilon serratum Cass. (Cassini 1827: 466). Type: ‘Nous avos décrit cette espèce et la précédente [=Acroptilon subdentatum] sur des échantillons secs, innommés, dont nous ignorons l’origine.’ Holotype: ?P.

Acroptilon angustifolium Cass. (Cassini 1827: 466). Type: ‘Nous avons faite cette description sur un échantillon sec qui appartient problement à la Centaurea repens de Linné.’ Holotype: ?P.

Carduus picris (Pall. ex Willd.) Sweet (1830: 281).

Acroptilon picris (Pall. ex Willd.) C.A.Mey. (Meyer 1831: 67).

Acroptilon picris (Pall. ex Willd.) DC. (de Candolle 1838: 662), comb. superfl.

Acroptilon picris (Pall. ex Willd.) C.A.Mey. var. α obtusifolia (Cass.) DC. (de Candolle 1838: 662).

Acroptilon picris (Pall. ex Willd.) C.A.Mey. var. β angustifolia (Cass.) DC. (de Candolle 1838: 662).

Acroptilon picris (Pall. ex Willd.) C.A.Mey. var. γ subdentata (Cass.) DC. (de Candolle 1838: 662).

Acroptilon picris (Pall. ex Willd.) C.A.Mey. var. δ oxylepis DC. (de Candolle 1838: 662). Type: [Iran:] ‘ad agrorum margines circà Seidchadschi prov. Aderbeischan Persiae legit cl. Szowitz. Acr. Picris var. Fisch. et Mey.! pl. Szow. ined. (v. s.)’ Holotype: G-DC(G-00486874).

Acroptilon picris (Pall. ex Willd.) C.A.Mey. var. ε serrata (Cass.) DC. (de Candolle 1838: 663).

Acroptilon repens (L.) DC. (de Candolle 1838: 663).

Acroptilon australe Iljin (1937: 59). Type: ‘Turcomania, Kisil-Imam ad fl. Czan-dar, 12 VI 1912 n. 3919, leg. V. Lipsky et A. Michelson (Herb. Inst. Bot. Ac .Sc. URSS).’ Holotype: LE.

Acroptilon repens (L.) DC. f. sphaerica Bornm. (Bornmüller 1940: 209), nom. inval. — see Discussion. [Based on: ‘Iter Persico-turcicum 1892-93. No 4069. Persiae austro-orient. prov. Kerman: Kerman, in cultis. 1900 m. s. m. 1892. V. 13. legit et determ.: J. Bornmüller.’ Original material: JE(00015673).]

Acroptilon repens (L.) DC. f. cylindrica Bornm. (Bornmüller 1940: 209), nom. inval. — see Discussion. [Based on: ‘Iter Persico-turcicum 1892-93. No. 4067. Persiae austro-orient. prov. Kerman: Kerman, m. Kuh-tagh-Ali. 2100 m. s. m. 1892. V. 10. legit et determ.: J. Bornmüller.’ Original material: JE(00015672).]

Acroptilon repens (L.) DC. f. macrocephala Bornm. (Bornmüller 1940: 209), nom. inval. — see Discussion. [Based on: ‘[Iran:] Iter Persico-turcicum 1892-93. No. 4066. Persiae austro-orient. prov. Kerman: Kerman, ad Dschupar. 2000 m. s. m. 1892. VI. 8. legit et determ. J. Bornmüller.’ Original material: JE(00015678).]

Acroptilon repens (L.) DC. f. macrolepis Bornm. (Bornmüller 1940: 209), nom. inval. — see Discussion. [Based on: Iter Persico-turcicum 1892-93. No. 4064. Persiae austro-orient. prov. Kerman: Kerman, in incultis. 1900 m. s. m. 1892. 19. V. legit et determ.: J. Bornmüller — Original material: JE(00015677); ‘Iter Persico-turcicum 1892-93. No. 4065. Persiae austro-orient. prov. Kerman: in incultis. 2000 m. s. m. 1892. V. 19. legit et determ.: J. Bornmüller. — Original material: JE(00015686).]

Acroptilon repens (L.) DC. f. tincta Bornm. (Bornmüller 1940: 209), nom. inval. — see Discussion. [Probably based on: ‘am Lalesar, 300 m (11.VII.1892; [Bornmüller] no. 4071’. No original material found.]

Acroptilon repens (L.) DC. subsp. australe (Iljin) Rech.f. (Rechinger 1980: 309).

Acroptilon repens (L.) DC. subsp. australe (Iljin) Gubanov (1996), isonym.

Rhaponticum repens (L.) Hidalgo (in Hidalgo et al. 2006: 714).

distribution. S Russia, Turkey, Iran, N Iraq, Afghanistan, Pakistan, Kashmir, C Asia (Turkmenia to Uzbekistan), W Mongolia, China (Sinkiang), W Siberia. Introduced and widespread as a noxious weed in the USA (introduced late 19th century as an alfalfa contaminant) and Canada (Moore 1972), and now recognised as a weed in Argentina (Zuloaga & Morrone 1999).

discussion. Bornmüller (1940: 209), as part of his considerable output on his Iter Persico-turcicum 1892 – 1893, provided a short commentary on the variation he had observed in Acroptilon repens. He mentioned several formas in this discussion, and this appears to be the earliest record of these names, even following exhaustive searches of the works by Bornmüller mentioned by Rechinger (1955). As all the diagnostic prose is in German these are not considered to be validly published as they appeared at least five years after the relevant date for the requirement of Latin diagnoses in the Code; they are thus contrary to Art. 39.1 (Turland et al. 2018). If separately published labels of Bornmüller’s Iter Persico-turcicum were made available these would constitute valid publication of Bornmüller’s taxa as there is invariably a short Latin phrase under Bornmüller’s trinomial, the infraspecific epithet often with a penned rank added between ‘Pall.’ and the infraspecific name. Duplicates of the relevant Bornmüller material are widely distributed (e.g., K, P, etc.).

Although Iljin (1937) considered Acroptilon australe distinct from A. repens, as did Soskov (1959), later authors (e.g. Rechinger 1980; Gubanov 1996) considered it a variant of A. repens, as subspecies australe (Iljin) Rech.f. Hidalgo et al.’s (2006) commentary on ‘The Rhaponticum oriental clade’ indicated that their results precluded coming to a ‘verdict’ over whether the two taxa were distinct. If, however, research supports the continued recognition of a separate A. australe, a new name will have to be provided within Leuzea; Leuzea australis Gaudich., an Australian endemic, already exists.