The palm genus Raphia P.Beauv. contains 22 species (Mogue Kamga et al. 2018) mainly distributed in continental Africa, with one species reaching Madagascar and one endemic species in Central and South America (Dransfield et al. 2008, Stauffer et al. 2014). This genus is also one of the most widely used and socio-economically important across Africa (Obahiagbon 2009, Tuley 1995, Burkill 1985).

Raphia was erected by Palisot de Beauvois (1804), who recognized the monoecious nature of the Raphia inflorescences, thus separating it from the lianescent and mainly Asian genus Calamus L. (rattans, dioecious) to which it was tentatively associated by Jussieu in Jussieu and Usteri (1791). Palisot de Beauvois (1804) described two new species: Raphia pedunculata P.Beauv. from Madagascar (now R. farinifera (Gaertn.) Hylander) and Raphia vinifera P.Beauv. from West Africa. The latter was suggested to be common along the rivers of the Oware (now in Delta State) and Benin (now in Edo State) provinces in nowadays south-eastern Nigeria. No collection number was mentioned in the protologue. However, the new name was accompagnied by a short description and two illustrations with analysis (code Art. 38.7, Turland et al. 2018) and is thus validely published (Art. 38.8): one of a flowering partial inflorescence and one of a fruiting partial inflorescence (Fig. 1). These drawings, however, appear to belong to two different species, being quite different in their morphology. Indeed, following the terminology of Otedoh (1982), the drawing of the partial fruiting inflorescence (Fig. 1B) is “raphiate”, being non-planar with widely interspaced rachillae (lax) and inconspicuous bracts, while the drawing of the flowering inflorescence (Fig. 1A) is distinctly “flabellate”, being planar with tightly compressed rachillae and large conspicuous bracts. In addition, R. vinifera was accompanied by a description of its morphology and uses, which provide vital pieces of information about how Palisot de Beauvois actually saw this species. He describes R. vinifera as a “tree of average size” that is “not distinguished by the length its trunk” but by the size of the infructescence (page 77). He suggests that the locals prefer to tap R. vinifera (in contrast to Elaeis guineensis, "palmiers à vins") because of the "great facility they have to collect it without danger" (page 78). These descriptions suggest that the R. vinifera he saw had a very short stem. He goes on to describe the multitude of uses of this species such as the leaves for weaving and thatching, and especially the wine that is tapped from the trunks and even the fruits.

Figure 1.  

Illustrations of Raphia vinifera from Palisot de Beauvois (1804), taken from “Flore d’Oware et de Benin”. A Inflorescence with young flowers (https://www.biodiversitylibrary.org/item/181617#page/183/mode/1up); B Infructescence (https://www.biodiversitylibrary.org/item/181617#page/185/mode/1up). Note the difference in inflorescence structure.

A few years later, in their revision of African palms, Mann and Wendland (1864) suggested that R. vinifera was closely related to the South American R. taedigera Mart. (von Martius 1823) “in the shape of fruit and scales” (page 423). However, as noted above, the illustration of the fruits (Fig. 1B) are certainly from a different species than the inflorescences. In their description of R. vinifera (page 437), Mann and Wendland (1864) cited a specimen collected by Mann s.n. from the “Banks of the Old Calabar” as being R. vinifera. We were not able to locate this particular specimen (neither in K nor GOET).

Beccari (1910), in his monograph of Raphia (in Italian), described the inflorescences and flowers of R. vinifera based on the type specimen he saw in the Candolle Herbarium (now merged in the G general collection) citing the “coplanar” nature of the partial inflorescences (page 88). However, he used the fruiting illustration of Palisot de Beauvois (1804) as a reference to describe the fruits (page 90; translated from Italian: “I have not seen the fruits of the authentic specimens of R. vinifera of Palisot de Beauvais [Beauvois]; but those figured in the “Flore d'Oware et de Benin", t. 46, f. 1”). He also indicated he had no precise idea about the general aspect of the species (size or trunk). In addition, he recognized that the specimen cited by Mann and Wendland (1864) as R. vinifera (Mann s.n., see above) was wrongly identified, and subsequently created a new name to accommodate it (R. mannii Becc.). Finally, Beccari (1910) described the species R. wendlandii Becc. from a collection of Mann s.n. collected in "Fernando Poo", currently the Island of Bioko (Equatorial Guinea). The type of this species name is located at K, in the economic botany section under the number 38686.

Chevalier (1932), based on the flowering inflorescence description of Beccari (1910), cited a Raphia species occurring in several countries across West Africa (Benin, Burkina Faso, Ivory Coast, Mali), which he tentatively associated to R. vinifera. Chevalier (1932) noted that this species was common in valleys of small rivers, was characterized by an absent or very short stem, and was tapped for wine. We saw one of his collections in Paris (P01794200, Fig. 2) from Mali (or Guinée), which contains four partial inflorescences closely matching the type of R. vinifera in G (see bleow).

Figure 2.  

Specimen of Raphia vinifera collected by A. Chevalier from Soudan Français or Guinée Française in 1908. [P01794200; https://science.mnhn.fr/institution/mnhn/collection/p/item/p01794200?listIndex=113&listCount=223]. Scanned by the Muséum Nationale d'Histoire Naturelle, Paris.

In the first edition of the Flora of Tropical West Africa, Hutchinson and Dalziel (1936) described R. vinifera as a "medium sized tree" with a stem longer than 3 m tall (page 387), which contrasts with Palisot de Beauvois’s description (see above). They reproduced the drawing of the flowering inflorescence of Palisot de Beauvois (their Figure 317, page 389) and suggested that a portion of the inflorescence is present in Kew, though we did not locate it. In fact, no collection (collector/number) is cited for this species by Hutchinson and Dalziel (1936).

Confusion really settled in when, in his review of west African Raphia’s, Russell (1965) published a long description of R. vinifera indicating it has a stem up to 5 m tall, apparently agreeing with Hutchinson and Dalziel (1936) but in contrast with Palisot de Beauvois (1804). He also described the inflorescences: “branchlets are clearly seen to be in 4 ranks extending right round the partial inflorescence” (page 180, accompanied by an illustration of the inflorescence, Figure 7A). The description and illustration clearly correspond to the raphiate type of Otedoh (1982) and do not match the type specimen or the description provided by Palisot de Beauvois (1804). The illustration and descriptions provided by Russell (1965), and in subsequent taxonomic works by the same author (Russell 1968, Tuley and Russell 1966), actually correspond to the name R. mannii (see above), which he placed in synonymy with R. vinifera in addition to R. wendlandii (Russell 1968). This publication led most authors, mainly working in Nigeria, into error about the morphology of R. vinifera.

In the latest revision of the genus, Otedoh (1982) placed the name R. vinifera in the “raphiate” section, probably following the fruiting description (and illustration) of R. vinifera (Palisot de Beauvois 1804). Interestingly, he associated very few uses to the species (“would yield good thatch and bamboos” page 162), suggesting it was sacred and generally protected. In the same publication, Otedoh (1982) described the new species R. mambillensis Otedoh (Otedoh 7401; Fig. 3) in the flabellate section (planar shaped partial inflorescences). This species was described as a small Raphia palm with a subterranean stem and leaves 5–8 m long arising from the ground. It is a common species mainly growing between 1200 and 2000 m in the Cameroonian Volcanic Line and reported from Nigeria, Cameroon, Central African Republic and South Sudan (Chevalier 1932, Letouzey 1978). Interestingly, Otedoh (1982) also noted that R. mambillensis grows alongside streams or in swamps at lower altitudes (page 164). Raphia mambillensis is a widely-used palm for wine, thatching and as a source for grubs. Prior to Otedoh (1982), R. mambillensis used to be confused with R. farinifera (e.g. Russell 1965). In addition, Otedoh (1982) also described a new variety: R. vinifera var. nigerica Otedoh, distinct by the symmetrically opposite leafets.

Figure 3.  

Isotype of Raphia mambillensis Otedoh (Otedoh 7401; K000209303)

Finally, in an overview of African palms, Tuley (1995), following Tuley and Russell (1966) and Otedoh (1982), placed R. vinifera in the raphiate section. He went further and suggested that Palisot de Beauvois (1804) (page 82) got confused with the different palms he saw and that the description of the uses of R. vinifera were in fact those of R. hookeri, which he refers to as the "true wine palm".

As indicated above, no holotype was explicitly mentioned in the protologue (Palisot de Beauvois 1804). In his revision of the genus, Beccari (1910) mentions studying a "type exemplar" (page 88) of R. vinifera from Palisot de Beauvois deposited at the “De Candolle Herbarium” in Geneva. He does not, however, provide more details about the specimen. In an unpublished PhD thesis, Otedoh (1976) identified Palisot de Beauvois s.n. from the districts of Warri and Benin as the type (page 245). Finally, Stauffer et al. (2017), in an overview of taxonomic knowledge on African palms, provided a scan of a Palisot de Beauvois s.n. collection from G and suggested it was the holotype of R. vinifera (barcode: G00301631, Fig. 4). We have now identified seven specimens belonging to the collection Palisot de Beauvois s.n., with isotypes deposited in five herbaria (see below). This specimen is composed of a partial inflorescence with young flowers and large bracts, belonging to the flabellate type, and is clearly what is illustrated in Fig. 1A of Palisot de Beauvois (1804). To date, we did not find any Raphia specimens collected by Palisot de Beauvois with fruits.

Figure 4.  

Scan of the lectotype (here designated) of Raphia vinifera, stored at the G herbarium [G00301631]. Image scanned by Conservatoire et Jardin Botanique (Geneva).

A precise understanding of R. vinifera remains unclear. From the protologue and the type specimen (Fig. 4), it is clear that the name R. vinifera refers to a species with an acaulescent or very short stem, flabellate inflorescences and having multiple uses. It is not a species with a distinct stem, raphiate inflorescences and few uses as suggested by Russell (1965), Otedoh (1982) or Tuley (1995). The flabellate inflorescences as seen on the type specimen of R. vinifera (Fig. 4) strongly resemble in shape the morphology of the isotype of R. mambillensis (Fig. 3). We thus suggest that the latter is a synonym of the former. In addition, we officially lectotypify the Palisot de Beauvois s.n. as the type of the name R. vinifera. We place the other specimens as isolectotypes.

The original illustrations of one partial inflorescence and one fruiting inflorescence from two different species have led to confusion (Fig. 1, Palisot de Beauvois 1804). Prior to Russell (1965), most authors agreed that R. vinifera had coplanar inflorescences (Beccari 1910, Chevalier 1932) and thus agreed with the description and type of Palisot de Beauvois (1804). It is interesting that Otedoh (1982) published the name R. mambillensis having seen the type specimen of R. vinifera, despite clear resemblances. Russell (1965), by describing and illustrating R. vinifera with a raphiate type inflorescence, created confusion around the proper identity of R. vinifera, and the application of the name has since been unclear and inconsistent in local floras or treatments (Akoègninou et al. 2006, Letouzey 1978, Tuley 1995, Stauffer et al. 2014, Otedoh 1982).

However, it remains unclear to what species the illustration of the fruiting partial inflorescence of Palisot de Beauvois (1804) belongs, as well as what Russell (1965) and Otedoh (1982) called R. vinifera. The descriptions, illustrations and photographs provided by Russell (1965) and Tuley and Russell (1966) correspond to R. mannii, a widespread West African species. Otedoh (1982), nevertheless, considered R. mannii as a valid species (including the name R. wendlandii, which is a synonym of R. mannii). For Otedoh (1982), R. vinifera resembles R. africana Otedoh, another little-known species. Thus, the morphological concepts of R. vinifera of Otedoh (1982) and Russell (1965) are not the same. Knowledge of Nigerian Raphias remains complicated and more field work is needed to properly sort out the different species in this region. In addition, we tentatively associate the specimens identified as R. vinifera var. nigerica by Otedoh (1982) with R. mannii.