Pauridiantha canthiiflora Hook.

Salvator NTORE & Olivier LACHENAUD, 2019, Two new species and a new combination in the genus Pauridiantha Hook. f. (Rubiaceae) from tropical Africa, Adansonia 41 (4), pp. 29-40 : 39

Pauridiantha canthiiflora Hook. f.

In Genera Plantarum 2: 70 (1871). — Typus: Equatorial Guinea. Fernando Po [= Bioko island], X.1860, Mann 167 (lecto-, K[K000172966!], designated here; isolecto-, GH [ GH01154748 !]; P[P00219639!]) .

Pauridiantha microphylla R.D.Good , Journal of Botany, British and Foreign 64, Suppl. 2: 5 (1926). — Typus: Angola (Cabinda). Belize , Gossweiler 7006 (holo-, BM [ BM000903007 !]; iso-, LISC [ LISC000796 !, LISC000797 !]), syn. nov .

DISTRIBUTION. — Lower Guinea subcentre of endemism. Occurs from south-eastern Nigeria to Angola ( Cabinda); locally abundant.

REMARKS

Good (1926) separated Pauridiantha microphylla from P. canthiiflora on account of its smaller leaves with different venation, but these characters are not reliable. The key in Ntore (2008: 60-61) separates them based on the calyx (“cupuliforme” vs “denté ou lobé”) and type of domatia, which are not reliable either: both taxa actually have similar domatia, i.e. crypt-like and masked by a tuft of hairs. No additional differences have been found, so the two taxa are synonymised here, which extends the range of P. canthiiflora to Angola ( Cabinda).

The original description of P. canthiiflora does not cite the type specimen, but Mann 167 is the only possible candidate, and has always been considered as the type by subsequent authors (e.g. Hallé 1966; Ntore 2008). This collection is represented by three sheets in different herbaria. Ntore (2008: 82) cited the K sheet as holotype, which is an error since no herbarium of deposit was cited in the protologue; this sheet is here selected as lectotype, since it is the only one annotated by Hooker (1873).