Commelina aquatica Morton (1956: 515–517

Hassemer, Gustavo, 2020, Further cleaning of the name pool in the New World Commelina (Commelinaceae), and notes on the African C. aquatica, Phytotaxa 435 (2), pp. 101-132 : 126

4.1. Commelina aquatica Morton (1956: 515–517 , fig. 12)

≡ Commelina diffusa subsp. aquatica (J.K.Morton) Ogwal (1994: 417)

Type: — GHANA. Gold Coast , around standing water in Accra Plains near Ada rd., ± 15 mls from Accra, 4 October 1953, J. K. Morton A30 (lectotype [designated here] BM-013399140! [ Figure 17 ]; isolectotypes K-000345527!, WAG-0000496!) .

Notes: — The protologue of C. aquatica ( Morton 1956: 515–517) provided the following information on the type: “Morton A30 ( B. M., G. C.)”. Therefore, the J. K. Morton A 30 specimens kept at BM and GC are syntypes (Art. 9.6 of the ICN). I was not able to search for specimens conserved at GC ; I could locate only one syntype at BM, BM-013399140 ( Figure 17 ), which I designate here as lectotype. I was able to locate two duplicates (therefore, isolectotypes): K-000345527 and WAG-0000496 .

In the context of an ongoing study of the C. diffusa species complex in the Americas—which can possibly be understood to include C. caroliniana Walter (1788: 68) , C. diffusa , C. gigas Small (1933: 264 , 1503), C. longicaulis , C. pallida and C. texcocana —it caught my attention the similarity of the original specimens of the African C. aquatica to the recently re-established South American C. longicaulis (see Hassemer 2018a, 2019). More recently, C. aquatica has been synonymised under C. diffusa (see Faden 2012: 140–144). Nevertheless, the information presented by Morton in the protologue of C. aquatica suggests that this name is not a synonym of C. diffusa ; of critical importance is the following: staminodes 2, with white antherodes (staminodes normally 2, rarely 3, with yellow antherodes in C. diffusa ) and chromosome number 2 n = 28 (2 n = 30 in C. diffusa ). It is possible that the chromosome count for C. aquatica is incorrect, a question that could only be clarified with new counts for this species, based on specimens from the type locality and matching the morphology of the type. The antherodes of C. aquatica are also differently shaped to those of C. diffusa (see Figure 12 in Morton 1956). For these reasons, I consider that the inclusion of C. aquatica within the circumscription of C. diffusa is probably incorrect, as was the case of C. longicaulis until Hassemer (2018a). It is evident that the concept of C. diffusa used in the 20 th century included a number of species lumped under it—this is further evidenced by new species still being dismembered from C. diffusa ( Nandikar & Naik 2019) . I hope that the information presented here encourages further study of C. aquatica .

Figures

FIGURE 12. Lectotype of Commelina diffusa var. cordispatha (O. Rohweder 882, HBG-514345). Copyright: Universität Hamburg.

FIGURE 17. Lectotype of Commelina aquatica (J.K. Morton A30, BM-013399140). Copyright: Natural History Museum (London). Available online at: https://data.nhm.ac.uk/object/a0086b0d-578b-43b3-b155-5c4c94960105/1567036800000