Dioscorea haitiensis Knuth ( 1926: 6)

Raz, Lauren, 2016, Untangling the West Indian Dioscoreaceae: New combinations, lectotypification and synonymy, Phytotaxa 258 (1), pp. 26-48 : 34-35

7. Dioscorea haitiensis Knuth ( 1926: 6) . Lectotype (designated here):— HAITI. Massif de la Selle, Morne Fort Jacques between Fortes Jacques and Cadets, ca. 800 m, 7 Sept. 1924, E.L. Ekman H1821 ♂ (lectotype S!, isolectotypes B! (2 sheets: B_10_0250057, B_10_0250058), US!). The S sheet bears a holotype label by T. Clayton (1954, unpublished), however this cannot be interpreted from the protologue. The S sheet and one of the B sheets (100250058) were both annotated by Knuth as “Typus”

Rajania ovata Swartz (1788: 59) . Lectotype (designated here):— DOMINICAN REPUBLIC.“Sto. Domingo”,no date, Swartz s.n. [lectotype S! (the sheet with no fragment packet), isolectotypes BM!, LD “A”! (pistillate fragment), S! (sheet with fragment packet)].

Rajania mucronata auct non Willd. Griseb (1866: 251). Lectotype (designated here):— CUBA. Oriente, 1860-64, Wright 691 (lectotype K!, isolectotypes BM!, BREM!, HAC! (p.p.), GH! (barcode 00030440) “MO1A”!, NY! (three sheets, one pistillate, two staminate). A second Wright specimen at K corresponding to this species (but without a collection number) from “Monte Verde” can also be considered an isolectotype. At GH there are two duplicates of Wright 691. The second one has mixed pistillate elements of D. haitensis and D. baracoensis .

Note: Dioscorea haitiensis occurs in Eastern Cuba and Hispaniola. It was described by Knuth from staminate material only. It is the male of R. ovata , described 138 years earlier from pistillate material by Swartz.

The epithet “ovata ” is not available for this taxon, because it was applied by Vellozo (1827: Tab. 117) to a South American yam species. The interpretation of the WI species has been somewhat confounded by the mixed nature of the original Swartz material. Both S sheets and the BM sheet each have a single element: a fruiting specimen of what is here called D. haitensis . The LD isotype includes two elements: a fruiting specimen on the left hand side of the sheet (here designated “A”), and a staminate specimen on the right (here designated “B”). Only the “A” portion corresponds to D. haitensis ; the “B” portion is Dioscorea baracoensis . The latter has coriaceous, generally acute leaves, and tiny flowers with erect tepals, while the former has membranaceous, acute to acuminate leaves, and larger, rotate flowers. Swartz’s description “leaf ovate, acuminate, three nerved” does not resolve the differences between the two taxa.

The Swartz name is also a homonym of a Walter name (1788: 247) published between the months of April and June of 1788; Swartz’s Prodromus was published between 20 June and 29 July of the same year. It is difficult to determine priority in this case, although Swartz’s name has been in use since its publication, while the Walter name was subsequently transferred to Polygonaceae as Brunnichia ovata (Walter) Shinners.

The Grisebach name, cited here in synonymy is also based on a similarly mixed collection, Wright 691. The K sheet, designated here as the lectotype, is true D. haitensis . Both sheets at MO also contain elements of D. baracoensis (MO1B and MO2A&B). Additional sheets at NY and HAC are D. baracoensis .