Original elemets and typification of the names included in the Ononis tridentata complex (Fabaceae) Author Ferrer-Gallego, P. Pablo 0000-0001-7595-9302 Servicio de Vida Silvestre y Red Natura 2000, Centro para la Investigación y la Experimentación Forestal (CIEF), Generalitat Valenciana, Avda. Comarques del País Valencià 114, E- 46930 Quart de Poblet, Valencia, Spain. & flora. cief @ gva. es; https: // orcid. org / 0000 - 0001 - 7595 - 9302 flora.cief@gva.es text Phytotaxa 2023 2023-05-19 598 1 32 58 http://dx.doi.org/10.11646/phytotaxa.598.1.3 journal article 10.11646/phytotaxa.598.1.3 1179-3163 7958772 Ononis tridentata Linnaeus (1753: 718) subsp. tridentata Ononis tridentata Linnaeus (1753: 718) was described in the protologue with a phrase-name “Ononis foliis ternatis carnosis sublinearibus tridentatis”. Linnaeus (1753) stated about the origin of the plant “Habitat in Hispania ” and also cited one synonym “Anonis hispanica frutescens, folio tridentato carnoso” from Magnol (1697) , Ray (1704) , and Tournefort (1719) . The lectotype of this name was designated by F̂rther (in Turland & Jarvis 1997 [August]: 478) from an illustration “Anonis Hispanica frutescens folio tridentato carnoso” published by Magnol (1697 : tab. [4] unnumbered). The lectotypification excluded top right-hand plant, which belongs to another species ( Fig. 1 ). This illustration match with the traditional concept and current use of the name as a species with leaflets blade obovate-cuneate, less than six times longer than wider, and with (0–)3–5(–9) teeth in the upper half (see e.g., Losa 1958 , Ivimey-Cook 1968 , Devesa & López González 1997 , Devesa 2000 , 2009 , López González 2001 ). On the other hand, Devesa & López González (1997 [December]: 254) designated as the “ lectotype ” of the name the same illustration, as “ Lectotypus (Iconotypus): Pierre Magnol, Hortus Regius Monspeliensis, t. [4], 1697, excl. ram. ster. [Designado en este trabajo]”. Although this designation is later than the one published by F̂rther, and therefore ineffective, these authors make an exhaustive exposition about the possible origin of the material used as voucher for Magnol’s illustration (1697: tab. [4] unnumbered). These authors concluded that the origin of the plant cultivated in the Botanical Garden of Montpelier used as voucher could come from seeds collected by Tournefort in Cardona ( Catalonia , Spain ) during his trip to Spain in 1687 (see Henriques 1898 , Devesa & López 1997 ). FIGURE 1. Lectotype of Ononis tridentata L. subsp. tridentata , illustration “Anonis Hispanica frutescens folio tridentato carnoso” published by in Magnol (1697 : tab. [4] unnumbered) [excluding top right-hand plant]. On the other hand, F̂rther (in Turland & Jarvis 1997: 478 ) designates an epitype of Ononis tridentata from a specimen preserved at M, as “ Spain , Granada, in salsis et collibus argillosis La Mala prope Granatam, Jul 1837 , Boissier (M)”. Unfortunately, this epitypification of O . tridentata published by F̂rther (in Turland & Jarvis 1997: 478 ) has gone unnoticed (see e.g., Devesa 2000 ). In the herbarium M there are two specimens of the gathering indicated by F̂rther, barcoded as M0232961 and M0232962 ( Fig. 2 ). Therefore, the epitypification proposed by F̂rther (in Turland & Jarvis 1997 ) may be further narrowed to a single specimen by a “second-step” epitypification according to Art. 9.17 of the ICN ( Turland et al . 2018 ) (see more details below). Consequently, under Art. 9.20, first sentence, and Art. 9.9, the epitype selected by F̂rther (in Turland & Jarvis 1997 ) must determine the precise application of Ononis tridentata s.str. Although in the epityping it is not mentioned, we assume that F̂rther select an epitype because the lectotype “is demonstrably ambiguous and cannot be critically identified for purposes of the precise application of the name to a taxon” (see ICN Art. 9.9). However, in our opinion, the epitification was not a good choice, since the lectotype of the name, the illustration “Anonis Hispanica frutescens folio tridentato carnoso” in Magnol (1697 : tab. [4] unnumbered) can be critically identified for the purposes of its taxonomic identification as was clearly and explicitly mentioned by Devesa & López González (1997) . Once an epitype is designated in accordance with the provisions of the Shenzhen Code , the identity of the type it supports (in this case the lectotype ) is totally irrelevant for the purpose of application of the name. In addition, that although the type being supported by an epitype should be “demonstrably ambiguous and cannot be critically identified for purposes of the precise application of the name to a taxon” (Art. 9.9), even if there is evidence that this is demonstrably not the case (hardly the situation here), there is no provision in the Code for the designated epitype not to continue to determine the application of the name. Therefore, under the current rules O . tridentata must be applied to the species to which F̂rther’s epitype belongs (Art. 9.9 (“An epitype is a specimen or illustration selected to serve as an interpretative type ...”) and Art. 9.20 (“The author who first designates (Art. 7.10, 7.11, and F.5.4) an epitype must be followed.”). Only if the type that an epitype supports is displaced (lost, destroyed, or supersedable under Art. 9.19) does an epitype cease to be the element determining the application of the name. FIGURE 2. Epitype of Ononis tridentata L. subsp. tridentata , M (barcode M0232962). Image courtesy of the herbarium M, reproduced with permission. The epitype designated by F̂rther (in Turland & Jarvis 1997 ) “ Spain , Granada, in salsis et collibus argillosis La Mala prope Granatam, Jul 1837 , Boissier (M)” is undoubtedly identifiable with the traditional concept and current use to which the name O. tridentata subsp. crassifolia is generally applied (leaflets blade lanceolate–cuneate, less than six times longer than wider, with 3–9 teeth in the upper half, sepals entire, inflorescence peduncle shorter than 10 mm , usually with an inconspicuous arista, inflorescence 1–2-flowered and raceme-like, indumenta of short glandular hairs and sessile glands) (see Devesa & López González 1997 , Devesa 2000 , 2009 , Mota et al . 2011 ). However, the epitypification from the specimen at M generates an abrupt and relevant change in the nomenclature of O . tridentata complex, because the name O . tridentata subsp. tridentata must be applied to populations endemic to the southeastern Iberian Peninsula (with diagnostic characters mentioned above), and another name (in subspecies rank, see below) should be proposed to name the Iberian Peninsula populations traditionally called as O . tridentata subsp. tridentata , with leaflets blade lanceolate–cuneate, less than six times longer than wider, with (0–)3(–5) teeth in the upper half, inflorescence most times 1–2-flowered, axillary; indumenta only of short glandular hairs (see Devesa & López González 1997 , Devesa 2000 , 2009 , López González 2001 , Mota et al . 2011 , Agulló et al . 2013 ). Consequently, the names O. barrelieri , O. crassifolia var. intricata , O. tridentata var. canescens , O. tridentata f. intermedia , and O. tridentata var. intermedia f. erecta , traditionally considered as synonyms of O . tridentata subsp. tridentata , should be treated as synonyms of O . tridentata subsp. arbuscula , comb. & stat nov. (see below).