Dioscorea bahamensis (R.Knuth) Raz, 2016

Raz, Lauren, 2016, Untangling the West Indian Dioscoreaceae: New combinations, lectotypification and synonymy, Phytotaxa 258 (1), pp. 26-48 : 29-30

2. Dioscorea bahamensis (R.Knuth) Raz , comb. nov. Rajania bahamensis Knuth (1917: 220) . Type:— BAHAMAS. New Providence, 25 January 1890, J. I. & A. R. Northrop 203 pro parte (holotype B! (♂, ♀), isotypes F! (♂, ♀), K! (♀), NY! (♀). The K sheet consists of three detached leaves, a portion of a stem and a detached fruit, collectively assigned barcode number K000099309, and affixed to the same sheet as the isotype of R. urbaniana , (see below). The NY sheet was annotated as the lectotype in 1975 by W.T. Gillis, who believed, mistakenly, that the B sheet had been destroyed

Rajania microphylla Kunth (1850: 451) . Type:— CUBA. Pinar del Río, In savannis Cubae ad sumidero, Oct. 1823, Poeppig s.n. (holotype B!, isotypes JE! (2 sheets), K!, MO!).

Rajania urbaniana Knuth (1917: 220) . Type:— BAHAMAS. Andros Island (“ Purser Point , Andros Island ” on labels at K and NY), 17 June 1890, J. I. & A. R. Northrop 203 pro parte ♂ (holotype B!, isotypes G!, GH!, K!, NY!). The K sheet includes two separate fragments from Andros Island, each with its own specimen label, assigned respective barcode numbers K000099308 and K000524433 (see also R. bahamensis , above for description of third fragment); a fourth fragment on the same sheet is a non-type specimen of the same species: Eggers 4287, barcode number K000099310 . The NY sheet was annotated as the lectotype in 1975 by W. T. Gillis, who thought, mistakenly, that the B sheet had been destroyed .

Rajania prestoniensis Knuth (1924: 337) . Lectotype (designated here):— CUBA. Oriente (Holguin): Prope Preston in litore maris, 14 November , 1914, E.L. Ekman 3448 ♂ (lectotype S! ( S05-5314 ), isolectotypes B!, NY!, S! ( S05-5309 )). The protologue states “Typus in herb. Holm”, without reference to one sheet or the other.

R. hastata auct. non L. Bentham in Hooker (1882: 72, pl. 1392), Northrop & Northrop (1902), and Correll & Correll (1982).

Notes: This species is restricted to Cuba and the Bahamas. Rajania microphylla is the earliest name, but the specific epithet was not available in Dioscorea and therefore it was necessary to select from among the available synonyms.

The three species subsequently described by Knuth all reflect variation in leaf outline, as well as minor differences in the length and number of staminate inflorescences, characters that vary within this and many other Dioscorea species. Knuth did not describe pistillate flowers or fruits for any of the three species listed here in synonymy, even though the holotype of R. bahamensis does in fact include a branch with developing fruits and two detached mature samaras. Counting both the 1917 and 1924 publications, Knuth cited just a single staminate specimen each to represent, respectively, R. bahamensis , R. urbaniana and R. prestoniensis : an insufficient sample to adequately characterize the plasticity in this taxon.

For centuries, confusion has prevailed about what to call this species, and most commonly R. hastata has been misapplied. Bentham (1882), in his description of “ R. hastata L.”, cited a mishmash of elements that include the Plumier illustration from “San Domingo” (Lectotype of R. hastata L.), “Poeppig” from Cuba (probably in reference to the type of R. microphylla , although other Poeppig s.n. specimens are also housed at K, where Bentham was based), “L. Brace” ( D. bahamensis from the Bahamas) and “Wright 1712” from Cuba. Among the Wright 1712 duplicates (see D. cephalocarpa , below), Bentham is most likely to have consulted the two sheets at K: these include a total of five elements, however none correspond to D. bahamensis (the K elements are K1: D. introrsa , K “2A”: D.cephalocarpa ♂, K “2B”: D. confusa ♂, K “2C”: D. scorpioidea ♀, and K “2D”: D. cephalocarpa ♀; this last element is at the top of the “K2” sheet- I neglected to label it when I annotated the specimen). The plate 1392 (Hooker 1882) that accompanies Bentham’s description, corresponds to D. bahamensis .

Norththrop and Northrop (1902), also misapplied the name “ R. hastata L.” in their Flora , citing only “Wright 1712” from Cuba (without specifying which elements of which duplicates they examined), and their own number 203 (the latter comprising material from both Andros and New Providence Islands). They made no reference to any original material cited by Linnaeus, nor did they cite Bentham (1882) or even Grisebach (1866), who described his own “ R. hastata ” in part, based on Wright 1712 (see D. psilostachya below). It is likely that the Northrops’ use of R. hastata can be traced directly to Wright’s own determination “ R. hastata ” written on the herbarium labels of his no. 1712, but it is difficult to determine which sheets they examined. The B sheet of Northrop 203 from Andros Island was subsequently designated by Knuth as the type of R. urbaniana , and the B sheet from New Providence Island, as the type of R. bahamensis .

Britton & Millspaugh (1920) in their Bahama Flora criticized “Mrs. Northrop’s” use of “ R. hastata ” in connection with the Bahamian plants that they called R. microphylla . They wrote that R. hastata was by that time understood to be a species endemic to Hispaniola, but they too failed to cite any type elements of either species.

The Corrells (1982) concept of R. hastata is the same as that employed by the Northrops, but unlike their predecessors, the Corrells cited R. microphylla in synonymy, (ignoring Britton & Millspaugh’s comment). Nevertheless, they did not cite any original material for either name.

During the years 1975-6, W.T. Gillis, who was working towards a revised Flora of The Bahamas, annotated nearly all of the type material of R. bahamensis , R. microphylla and R. urbaniana as “ R. hastata L.”, consistent with the Northrops’ misapplication of that name.

Ironically, there is no evidence to suggest that the Northrops, the Corrells, Gillis, Wright, Grisebach or Bentham consulted Clifford 458 (BM) in formulating their respective concepts of R. hastata , as the sheet was never cited by any of them. (See discussion under D. alainii .).