(2987) Proposal to conserve the name Vanilla (Orchidaceae) with a conserved type

(2987) Vanilla Mill., Gard. Dict. abr. ed. 4: [1432]. 28 Jan 1754 [Orchid.], nom. cons. prop. Typus: Vanilla planifolia Andrews (in Bot. Repos.: t. 538. Nov 1808), typ. cons. prop. Upon publishing the name Vanilla, Miller (Gard. Dict. abr. ed. 4. 1754) included three species detailed as “1. Vanilla flore viridi & albo, fructu nigrescente”, “2. Vanilla flore violaceo, fructu breviori rubro” and “3. Vanilla flore albo, fructu breviori corallino”. Of the first Miller notes that “The Fruit of these Plants is call'd by the Spaniards, in America, Vanilla, or Vinello; and is much used by them to scent their Chocolate. It is the first Species here mention'd, which is chiefly esteem'd. This grows plentifully in the Bay of Campechy, in the West-Indies; where they are usually sold for about Three-pence each Fruit, English Money.” Upon adopting binomial nomenclature, Miller (Gard. Dict. ed. 8: Vanilla no. 1. 1768) treated these 1754 species as one, formally naming it as Vanilla mexicana Mill. The first typification of the generic name Vanilla was made by Britton & Millspaugh (Bahama Fl.: 83. 1920), who proposed Epidendrum vanilla L. as the type. In accordance with Art. 10.5 and 10.6 of the Code, however, their selection must be regarded as having been made using a largely mechanical method, which requires an endorsement to avoid being superseded by a later typification. Britton & Wilson (Bot. Porto Rico 5: 185. 1924) again cite E. vanilla as type, but that is also a publication following a largely mechanical method of typification. Mansfeld (in Kulturpflanze, Beih. 2: 587. 1959) typified Vanilla with V. mexicana, which he considered, at least in part, to be a synonym of V. planifolia Andrews (in Bot. Repos.: t. 538. Nov 1808). Since Mansfeld's text does not include any discussion on the method of selection, nor refer to any special set of rules employed for his selection, this may be regarded as superseding the typifications by Britton & Millspaugh (l.c.) and Britton & Wilson (l.c.). Moreover, Mansfeld's selection has been cited by numerous subsequent authors. The typification of Vanilla mexicana itself has long been unclear. In the protologue of V. mexicana, Miller (l.c. 1768) cited a single element (“Volubilis siliquosa Mexicana, plantaginis folio. Cat. Car. 3. p. 7”) referring to Catesby (Nat. Hist. Carolina 2, App.: 7. 1747). Despite Miller (l.c. 1768) not referring to Linnaeus's name in any way, Garay & Sweet (in Howard, Fl. Lesser Antilles 1: 40. 1974) considered that Epidendrum vanilla L. (Sp. Pl.: 952. 1753) was the “basionym” of V. mexicana Mill. which, in turn, was said to be based on the phrase name of the first species listed by Plumier (Nov. Pl. Amer.: 25. 1703) and on his two illustrations of Vanilla (Plumier, l.c. 1703: t. 28; Pl. Amer. t. 188. 1758). The latter reproduction is identified by Burman as Epidendrum vanilla L. but Linnaeus (l.c.) did not refer to Plumier or his illustration. Ackerman (in Mem. New York Bot. Gard. 73: 174. 1995) wrote with respect to V. mexicana “Type: Plumier s.n. from Haiti (original illustration at P, not seen, reproduction: Plumier, Nov. Pl. Amer. pl. 28, t. 188. 1703)”. It does not seem as though Ackerman was implying that there was a Plumier specimen, but reproduced illustrations. Plumier's illustrations, which are kept at the Muséum National d'Histoire Naturelle in Paris, are evidently (see Karremans & al. in Lankesteriana 20: 431. 2020) the source for the illustration from Catesby (l.c.). Given that no records or specimens of Vanilla were located among Miller's materials at BM (including the Banks, Sloane and Miller herbaria, the catalogue of Miller's library and the catalogue of Miller's specimens sent to Sloane) and that the Plumier illustrations are not cited in the protologue of V. mexicana, the illustration in Catesby (“Cat. Car. 3. p. 7”) seems to be the only original material eligible for selection as lectotype. The taxon was formally typified by this illustration by Reveal (in Rhodora 111: 359. 2009). Despite its specific epithet chosen by Miller, the plant depicted by Plumier and reproduced by Catesby was not of Mexican origin, but from Haiti. Catesby (l.c.) himself contributed to the confusion, citing “Hist. Jam. 180. Vol. 1” (Sloane, Voy. Jamaica 1: 180. 1707), an account of the “Lobus oblongus aromaticus” of Jamaica, also referring to its use together with chocolate. Among the references cited by Sloane there is a single illustration, that was published by Leonard Plukenet (Plukenet, Phytographia 4: t. 320, fig. 4. 1694), which appears to be an improved copy of the “Araco aromatico” depicted from Mexico by Francisco Hernández (Rerum Med. Novae Hispan. Thesaur.: 38. 1651). Meticulous inspection by Soto Arenas & Dressler (in Lankesteriana 9: 285–354. 2010) revealed that the illustration reproduced by Catesby is not the species of Vanilla that was cultivated in Mexico and much used by the Spaniards to scent their chocolate, as early authors such as Miller and Linnaeus (l.c.) believed. In fact, the Antillean taxon is not known to occur in Mexico, nor is it in cultivation on account of its non-aromatic fruits (Soto Arenas & Cribb in Lankesteriana 9: 355–398. 2010; Karremans & al., l.c.: 395–497). In fact, the two are now known to belong to separate subgenera and be quite distantly related (Bouétard & al. in Molec. Phylogen. Evol. 55: 621–630. 2010). As clearly stated in the protologue, Miller had the aromatic, cultivated Vanilla in mind when describing the genus Vanilla, mistakenly believing it was the same species illustrated by Plumier, and reproduced by Catesby. This brings into question the typification of Vanilla with the non-aromatic V. mexicana. Vanilla is commonly accepted as a monophyletic genus, subdivided into two subgenera, within one of which two sections are recognized (Soto Arenas & Cribb, l.c.; Karremans & al., l.c.). In its most common circumscription, only 18 species belong to V. subg. Vanilla, with the other 100 species belonging to V. subg. Xanata Soto Arenas & Cribb (l.c.). The latter includes all of the fragrant species with commercial interest. Members of V. subg. Vanilla (typified by V. mexicana) differ from members of V. subg. Xanata (type V. planifolia), in the plants having membranaceous leaves, inflorescences scarcely or not differentiated from the vegetative axis, lip without a penicillate callus, column united to the lip only at the base, and shorter, bent column, and odorless fruits (Soto Arenas & Cribb, l.c.). Attempts to remove certain species of Vanilla into new genera have already been made, one of them quite recently (Averyanov in Turczaninowia 14(2): 15–100. 2011; Averyanov & Vuong in Taiwania 60: 33–38. 2015). However undesirable, the two subgenera of Vanilla are well separated, and their recognition as distinct genera would find support from genetic, morphological, geographical and even ecological data (Bouétard & al., l.c.). Unfortunately, the eventual recognition of the two subgenera of Vanilla as distinct genera would entail the exclusion of V. planifolia and its relatives, and a transfer of one hundred taxa to the reinstated genus Myrobroma Salisb. (Parad. Lond.: t. 82. 1807), the oldest generic name that applies to V. subg. Xanata. This would not only cause the transfer of ca. 100 names to Myrobroma, but also significant issues among vanilla growers and commerce, as the world famous and renowned spice V. planifolia would require a transfer under a different generic name. We consider such a move highly unfavorable and believe that an aromatic species should be designated as type of the generic name Vanilla given that this appears to have been the original intent of the author and on account of its broad use in literature and horticulture. We propose selecting, under Art. 14.9, the well-known, widespread, commercially cultivated, aromatic Vanilla planifolia as conserved type of Vanilla even though not originally mentioned by Miller. Our proposal would conserve the name Vanilla Mill. (l.c.) with the conserved type V. planifolia, itself typified by Garay & Sweet (l.c.: 44) using the illustration reproduced by Andrews (l.c.). Under the proposed scenario, the generic name Vanilla of Miller becomes tied to the aromatic vanilla of commerce as intended by its author. This would prevent the aromatic species of Vanilla, including those of commercial interest, from being segregated from the genus. APK, https://orcid.org/0000-0001-5987-7710 FP, https://orcid.org/0000-0001-6780-8395