(3116) Proposal to conserve the name <i>Pandanus amaryllifolius</i> Roxb. against <i>P. amaryllidifolius</i> F. Voigt and <i>P. amaryllifolius</i> Roxb. ex Lindl. (<i>Pandanaceae</i>)

(3116) Pandanus amaryllifolius Roxb., Fl. Ind., ed. 1832, 3: 743. Oct–Dec 1832 [Angiosp.: Pandan.], nom. cons. prop. Neotypus (hic designatus): Singapore, Singapore Botanic Gardens, Tyersall-Gallop Core, Keppel discovery Wetlands, not far from Tyersall Gate Visitor Centre, cultivated [living collection No. 20170327*A], 9 Aug 2024, Low & Schoenenberger 1471 (G barcode G00341958; isoneotypi: BKF, BO, K, KEP, KLU, MO, SING barcode 0390822, US). (H) Pandanus amaryllidifolius F. Voigt in Syll. Pl. Nov. 2: 52. Mar–Dec 1826, nom. rej. prop. Typus: non designatus. (H) Pandanus amaryllifolius Roxb. ex Lindl. in Loudon, Encycl. Pl.: 820, 821. 1829, nom. rej. prop. Typus: non designatus. Pandan wangi, or more commonly Pandan, in Malay refers to a plant widely grown in South and Southeast Asia for its fragrant leaves with a distinctive aroma. Leaves tied in a knot and juice extracted from leaves are used to flavour foods such as rice, dishes, desserts, drinks and jellies, and as a natural colouring agent to make jams and bakery products green (Burkill, Dict. Econ. Prod. Malay Penins. 2: 1676–1678. 1966; Stone in Econ. Bot. 32: 285–293. 1978). It is also used in traditional medicine (Wakte & al. in Genet. Resources Crop. Evol. 56: 735–740. 2009). This species, currently generally known as Pandanus amaryllifolius Roxb. (Fl. Ind., ed. 1832, 3: 743. 1832) (Warburg in Engler, Pflanzenr. IV. 9 (Heft 3): 87. 1900; Stone in Leroy, Fl. Cambodge, Laos Viet-Nam 20: 45. 1983; Setyowati & Siemonsma in de Guzman & Siemonsma, Pl. Resources S.-E. Asia 13: 164. 1999; Seidemann, World Spice Pl.: 271. 2005; Erhardt & al., Groβe Zander 2: 1600. 2008; Kun & DeFilipps in Wu & Raven, Fl. China 23: 128. 2010; Beentje & Callmander in Kiew & al., Fl. Penins. Malaysia 10: 211. 2023), originated from the Moluccas [Maluku], but is cultivated from India to New Guinea. It is only known in cultivation and is typically encountered as a small shrubby growth form, propagated vegetatively and never known to flower. If allowed to grow without harvesting, eventually a broad-leaved mature form develops. One collection of staminate flowers from such a plant is known (Stone, l.c. 1978). Pandanus amaryllifolius was first published by Roxburgh in a tabular presentation of the species in the Calcutta Botanical Garden as a nomen nudum (Roxburgh, Hort. Bengal.: 71. 1814), with a statement that it came from “Amboyna” [Ambon] in 1798, and the name had no other descriptive information than a symbol indicating that it was a shrub and the abbreviation “Cr.”, indicating that it had stolons. These features did not distinguish it from the following species in the table, called P. inermis Roxb., that had the same features. Like P. amaryllifolius Roxb., P. inermis Roxb. was validly published in Roxburgh, l.c. 1832: 744. Pandanus amaryllidifolius F. Voigt (in Syll. Pl. Nov. 2: 52. 1826) was validly published six years earlier than the valid publication of P. amaryllifolius Roxb. for a plant cultivated at Hortus Belvederensi in Weimar (Germany). Friedrich Siegmund Voigt (1781–1850) published the name accompanied by a brief description sufficient to validate it. A collection is extant at JE but it does not bear Voigt's handwriting. It is further labelled Pandanus amaryllifolius and not P. amaryllidifolius and cannot therefore be confirmed as original material. Lindley included Pandanus amaryllifolius in the tabular presentation of species in Loudon (Encycl. Pl.: 820. 1829). On the following page under the species number from the table (13687) appeared the description “Leaves quite entire”. While this, on its own, may not be considered sufficient to validly publish the name, it contrasts with the descriptions presented for five of the six other Pandanus species, which have either leaf margins spiny-toothed or serrate-spiny. Presumably this potential diagnosis is why P. amaryllifolius Roxb. ex Lindl. has recently been taken up by Mabberley (Mabberley's Pl.-Book, ed. 4: 674. 2017), Plants of the World Online, and World Flora Online (the latter two both accessed 23 Jun 2025). Stone (l.c. 1983: 46) referred to a Roxburgh specimen in C as the holotype of Pandanus amaryllifolius Roxb., with an “isotype” in B. There are no Roxburgh specimens of P. amaryllifolius Roxb. in these herbaria, or any other as far as we can trace. Judging from his annotations on specimens in C and B, Stone (l.c. 1983) had in mind a collection made by Joachim Otto Voigt (1798–1843) in Serampore at an unknown date. Voigt was a Danish botanist and surgeon, who mostly studied seed plants and pteridophytes. He was appointed regimental surgeon in the Danish colony of Frederiksnagore (now Serampore, India) in 1826 and was appointed manager of its Botanical Garden in 1834. He later served as the superintendent of the Botanical Garden in Calcutta [Kolkata] from 1841 to 1843. Voigt (Hort. Suburb. Calcutt.: 683. 1845) added to the information given by Roxburgh in 1814 that the species had been grown in the Serampore Botanical Garden for 30 years without flowering. Stone annotated the C specimen collected by J.O. Voigt in Serampore as the type of “Pandanus amaryllidifolius [Roxb. ex] Voigt”, evidently confusing the two botanical Voigts. However, as this typification was not published by Stone it was not effective. As Stone (l.c. 1983) had also failed to typify P. amaryllifolius Roxb., since there is no Roxburgh specimen of the species in C, we therefore here propose a neotype for Roxburgh's name, a specimen made from a plant in cultivation in Singapore. Although Pandanus amaryllidifolius F. Voigt and P. amaryllifolius Roxb. ex Lindl. have nomenclatural priority over P. amaryllifolius Roxb., and F. Voigt has a grammatically correct epithet for “Amaryllis-leaved”, these names have generally been ignored in the Pandanaceae literature and have never been seriously taken up after their publication. The absence of any original material may explain this reluctance to bring these names into use. Roxburgh (l.c.) supplied no derivation for the epithet ‘amaryllifolius’, so it is not a correctable error in spelling. Lindley's questionable earlier validation of Roxburgh's name has been used in only a small proportion of recent publications on the species. A search in Google Scholar (12 May 2025) indicated that of the 1820 publications that used “Pandanus amaryllifolius Roxb.” from 2021 onwards only 67 used “Pandanus amaryllifolius Roxb. ex Lindl.” In an effort to promote nomenclatural stability, we therefore propose to conserve, under ICN Art. 14 (Turland & al. in Regnum Veg. 162. 2025), the unambiguous name Pandanus amaryllifolius Roxb., which has been used almost exclusively for nearly two centuries. Given the cultural and economic importance of pandan wangi and the widespread use of P. amaryllifolius Roxb. as the name for it, we propose that Roxburgh's name is conserved against the competing names P. amaryllidifolius F. Voigt and P. amaryllifolius Roxb. ex Lindl. If the proposal is rejected and rules of priority followed, it seems likely that an unhelpful spelling change to the epithet will be introduced that will doubtless cause confusion to general users of plant nomenclature. We thank Nick Turland (B) for his advice. The authors are also indebted to William Baker (K), Ann Bogaerts (BR), Anna Donatelli (FI), Frank Hellwig and Kirstin Victor (JE), and Timothy Utteridge (SING) for their support. We finally thank Lua Hock Keong (National Parks Board, Singapore) for fruitful discussion on Pandanus amaryllifolius growing wild in Singapore, Nicola Schoenenberger (G) for collecting the neotype in the Singapore Botanic Gardens, and Melanie Schori and John Wiersema for improving an earlier version of this proposal.