(3124) Proposal to conserve the name <i>Allium moschatum</i> ( <i>Amaryllidaceae</i> ) with a conserved type from a later publication

(3124) Allium moschatum L., Sp. Pl., ed. 2: 427. Sep 1762 [Angiosp.: Lil. / Amaryllid.], nom. cons. prop. Typus: Herb. Burser III: 111 [right-hand specimen] (UPS No. V-173025), typ. cons. prop. (H) Allium moschatum L., Sp. Pl.: 298. 1 May 1753, nom. rej. prop. Typus: “Hispania”, 1751–1752, “Loefling 164?” [not located]. Linnaeus (Sp. Pl.: 298. 1753) described Allium moschatum providing the nomen specificum legitimum “ALLIUM caule teretifolio subnudo umbellifero, umbella fastigiata, petalis truncatis, staminibus tricuspidatis”, followed by the name “Loefl.” [Loefling] (namely, that Linnaeus described the plant with a diagnosis attributed to Loefling), and two synonyms. The first synonym “Allium foliis teretibus caulinis, floribus umbellatis” was cited from Haller (Allii Gen. Nat. Libell.: 32 , No. 14. 1745), the second “Moly moschatum, capillaceo folio” was cited from Bauhin (Prodr.: 28, text-fig. on p. 28. 1620; Pinax: 76. 1623), Rudbeck (Campi Elysii 2: 166, fig. 14*. 1701), and Sauvages (Meth. Fol.: 10. 1751). The protologue also includes the localities “Habitat in GalloProvinciae, Narbonae, Hispaniae apricis elevatis”. According to Wilde-Duyfjes (in Taxon 22: 70. 1973), the reference to Loefling (“Loefl.”) refers to the Spanish list that Loefling sent to Linnaeus, and which Linnaeus had at his disposal when he wrote the first edition of Species plantarum in 1753. On page 14 of this handwritten list appears the following entry: “-253a Allium caule teretifolio, umbella erecta, petalis truncatis. Staminib. 3 fid. Allium moschatum capillaceo folio C. Bauh. pin. in campestribus elevatis et collib. vulgaris”. Bauhin (l.c.: 1620: text-fig. on p. 28) and Rudbeck (l.c.: 165, fig. XIV and 166, fig. XIV) include illustrations titled “Moly moschatum capillaceo folio”, “XIV Moly moschatum capillaceo folio” (on p. 165) and “Moly moschatum capillaceo folio minimum. Burs. [Burser]” (on p. 166), respectively, which are original elements used by Linnaeus to describe his species. According to Stearn (Bot. Latin, ed. 2: 365. 1973 (chapter “Linnaean Signs in Species Plantarum, 1753”), the asterisk (*) indicates that a satisfactory description can be found at the cited source. Curiously, this does not hold true for Rudbeck's work, which provides only a brief polynomial and a vague, imprecise illustration. Indeed, it is not clear which taxon is represented in Rudbeck's (l.c.) figures labeled “XIV Moly moschatum capillaceo folio” (on p. 165) and “Moly moschatum capillaceo folio minimum. Burs. [Burser]” (on p. 166). On the other hand, the illustration published by Bauhin (l.c.: text-fig. on p. 28) depicts two plants, one in a vegetative state and the other with a developed inflorescence, as well as the details of two flowers. However, this illustration does not match Loefling's diagnosis, since the protologue states “caule teretifolio subnudo”, whereas the illustration shows a flowering scape with several leaves extending up to the middle of the scape; the tepals are acute rather than “petalis truncatis”, and the stamens are simple rather than tricuspidate (“staminibus tricuspidatis”). Both drawings are cited in the protologue and cannot therefore be in serious conflict with it (see ICN Art. 9 Note 10; Turland & al in Regnum Veg. 162. 2025); however, they do not reflect the characters included in Linnaeus's (Loefling's) 1753 diagnosis. Linnaeus (Sp. Pl., ed. 2: 427. 1762) published again the name Allium moschatum, accompanied by the same elements presented in the 1753 protologue, but with significant modifications. The name “ALLIUM moschatum”, numbered “17” (formerly “16” in Linnaeus, l.c. 1753), was accompanied by the revised diagnosis “caule teretifolio umbellifero, umbella fastigiata subsexflora, petalis acutis, staminibus simplicibus, foliis setaceis”. This replaced “petalis truncatis” (1753) with “petalis acutis” (1762), and “staminibus tricuspidatis” (1753) with “staminibus simplicibus” (1762). It also omitted the word “subnudo” (i.e., “caule teretifolio subnudo umbellifero” in 1753) and added “subsexflora” and “foliis setaceis” in 1762. Although Linnaeus no longer cited Loefling's name after the diagnosis, the rest of the entry is identical to that of 1753, i.e., including the first synonym “Allium foliis teretibus caulinis, floribus umbellatis” from Haller (l.c.), the second synonym “Moly moschatum, capillaceo folio” from Bauhin (l.c.: 1620: 28, text-fig. on p. 28; l.c. 1623: 76), Rudbeck (l.c. [also marked with an asterisk]), and Sauvages (l.c.), as well as the geographical provenance. Among the potential herbarium material, according to Jarvis (Order out of Chaos: 226. 2007), are specimens from Sauvages, who gave his herbarium of plants from southern France to Linnaeus in 1749. François Boissier de la Croix de Sauvages (1706–1767) was a French naturalist and physician, professor of medicine (1734), later of botany (1740) at the University of Montpellier, staunch supporter of Linnaeus and the Linnaean ideas in France. Unfortunately, no specimens collected by Sauvages, or annotated as “Monsp.”, have been found in any Linnaean and Linnaean-linked herbaria for this species. As discussed by Wilde-Duyfjes (l.c.), in the Linnaean herbarium at LINN there is a relevant sheet, Herb. Linnaeus No. 419.18, which bears four similar plants. The sheet is annotated “16 pallens”, although the number “16”, which corresponds to the species number in Species plantarum (1753), has been crossed out. On the reverse of the sheet, there is a note “Allium moschatum Hispania 253.a. Loefl.”. All annotations, including the crossed-out “16”, appear to be in Linnaeus's hand. Wilde-Duyfjes (l.c.: 71) concluded that this specimen, lacking the tricuspidate stamens described by Loefling, does not constitute original material of Linnaeus. He cited the German translation of Loefling's letter of 20–31 July (see below) to Linnaeus, in which is written: “Ich habe die mit einer Linie unterstrichen, die ich aufs genaueste beschrieben habe” (I have underlined those with a line which I have described in the greatest detail) (no. 253.a. Allium; the word Allium is underlined), as evidence that Loefling's diagnosis of the stamens as tricuspidate, an important characteristics in segregating Allium species, would have accurately matched what he had observed. Linnaeus appears to have misidentified LINN 419.18, which did not match Loefling's diagnosis, as A. moschatum sent by Loefling, and accordingly added the annotation “Hispania 253.a. Loefl.”. The number, copied from Loefling's list, was added by Linnaeus, not by Loefling himself. The specimen was identified by Linnaeus as A. pallens L. (l.c. 1762: 427). In any case, it is highly questionable for LINN 419.18 to be regarded as original material for A. moschatum of 1753. Additionally, there are two sheets in the Linnaean herbarium at S that bear specimens of Allium moschatum collected by Loefling in “Hispania”. This material is labelled as “isotypes” on a revision label affixed by Wilde-Duyfjes in 1969: S Nos. S09-16980 and S09-33660 (S-LINN No. 138.9), who at that time had considered LINN 419.18 as the holotype. However, this material is post-protologue (1 May 1753), so it cannot be original material. Loefling sent these plants to Linnaeus in October 1753, that is, after the publication of Species plantarum, as recorded in Loefling (Iter. Hispan.: 290. 1758: “Allium moschatum in campestribus elevatis & collibus, vulgare”; l.c.: 288: “Plantae Hispanicae. Missae 1753 mense Octobri, de quibus pag. 92” – Plants from Spain. Sent in the month of October 1753, about which [more is said] on page 92). The specimens most likely come from Madrid (see also Loefling, Reise Span. Länd.: 61. 1766; and his letter of 20–31 July 1752, sent from Madrid by Loefling to Linnaeus). Although this material clearly contradicts Linnaeus's 1753 diagnosis, due to its flowers with simple, non-tricuspidate staminal filaments, it corresponds well with the diagnosis provided by Linnaeus for this species in 1762 (Linnaeus, l.c. 1762). Following examination of the above candidate specimens potentially representing original material of Linnaeus for this name, Wilde-Duyfjes (l.c.: 70, 73) designated as lectotype a specimen housed in the Bauhin Herbarium at BAS, bearing the label “Moly moschatum capillaceo folio”. However, although the specimen is linked to Bauhin's synonym cited in the 1753 Linnaean protologue, this material was not seen by Linnaeus and thus cannot be considered as original material for the name. Pastor & Valdés (in Rev. Gen. Allium Penins. Iber. Islas Balear.: 98. 1983) therefore rejected this choice and designated instead the Burser specimen preserved at Herb. Burser III: 111 (UPS No. V-173025), specifically “el ejemplar de la derecha, más completo, con túnica en el bulbo y restos de una hoja” (the plant on the right, more complete, which has a tunic on the bulb and remains of a leaf) as the lectotype. However, according to López González (in Anales Jard. Bot. Madrid 57: 195. 1999), Burser III: 111 (UPS No. V-173025) “is in serious conflict with [Linnaeus's 1753] protologue” and therefore cannot be accepted as the lectotype. The specimen bears flowers with stamens that have simple, non-tricuspidate filaments, contradicting the phrase “staminibus tricuspidatis”, and tepals that are narrowly lanceolate and sharply pointed, rather than “petalis truncatis”. Based on correspondence between Loefling and Linnaeus, López González (l.c.: 196) argues that the name was primarily based on a now-lost specimen: Loefling 164?. It is possible, however, that Loefling made a mistake in describing the stamens as tricuspidate, an error that was later transcribed by Linnaeus in the 1753 diagnosis. According to Wilde-Duyfjes (l.c.: 73), Linnaeus (l.c. 1762) did not alter his 1753 concept but simply corrected a mistaken diagnosis. In reality, however, there is a serious conflict between the diagnosis published in 1753 and the current accepted lectotype of the name Allium moschatum, Herb. Burser III: 111 (UPS No. V-173025). In contrast, the 1762 diagnosis corresponds well with that specimen. López González (l.c.) suggested that continued use of the name Allium moschatum in its current sense (e.g., by Wilde-Duyfjes, l.c.; Pastor & Valdés, l.c.; Aedo in Talavera & al., Fl. Iberica 20: 244–245. 2014) could be preserved through a formal conservation proposal. Therefore, I propose, according to Art. 14.9(b), to conserve A. moschatum from its later 1762 publication over its earlier 1753 homonym. Accepting this proposal would not result in any disruptive nomenclatural change and eliminate the need to determine whether Linnaeus or Loefling made a mistake in the original diagnosis or whether the selected lectotype conflicts with the protologue. The same name would continue to be used for the same taxon, attributed to the same author, but anchored to a diagnosis that correctly matches the proposed type specimen. If this proposal is rejected, the name Allium moschatum would still be applied based on a lectotype that is in serious conflict with Linnaeus's original diagnosis, effectively accepting that Linnaeus misapplied the diagnosis (which he transcribed from Loefling) to original material, i.e., Rudbeck's and Bauhin's illustrations, the specimen at UPS-Burser, and possibly the specimen at LINN (if considered original), all of which contradict the original diagnosis published by Linnaeus in 1753. This situation could potentially lead to a future proposal to conserve the name with a conserved type that agrees with Loefling's diagnosis (i.e., tricuspidate stamens and truncated tepals), thus creating further instability in an already complex group of plants. We are grateful to Dr. Mats Hjertson (UPS), Dr. John Wiersema, and Dr. Melanie Schori for their advice, assistance, and valuable comments that improved this proposal.