The Lectotypification of Coffea liberica (Rubiaceae)

Abstract

The typification of the one time economically important species Coffea liberica Bull ex Hiern has not been adequately dealt with; the two major taxonomic works on Coffea (Lebrun in Inst. Royal. Col. Belge, Sect. Sci. Med. Mem. 8' 11 (3): 153 (1941) & Chevalier, Les Caf6iers du Globe 3: 170 (1947)) do not even cite the syntypes. Lectotypification is desirable since two of the syntypes are now known to belong to a different species. Furthermore the current concept of C. liberica is very broad and includes taxa (specific or infraspecific) that may quite possibly require future recognition; it is therefore important to attempt to define the typical element. When Hiern described C. liberica (in Trans. Linn. Soc., Bot. II, 1: 171, t.24 (1876) he cited the following specimens; these are accepted as syntypes:(a) Sierra Leone, Afzelius s.n: (BM) (b) Sierra Leone, cultivated in Mr Effenhausen's farm, Daniell s.n. (BM); (c) Liberia, cultivated in Cape Palmas, Vogel s.n. (K); (d) Angola, Cazengo, Welwitsch 3181 (K) (e) Angola, Golungo Alto, Welwitsch 3182 (K). The two latter specimens from Angola have been re-identified as C. canephora Pierre ex Froehner. One of these specimens, Welwitsch 3182, was later chosen by De Wildemann as the type of C. canephora var. hiernii Pierre ex De Wild. (De Wild., Les Caf6iers:20 (1901)). No specimens of C. liberica definitely indigenous in Angola are known to me. The three remaining West African syntypes are representative of the taxon historically referred to as C liberica. The only point in Hiern's specific description (and illustration) which contradict the above species is his statement that the stipules are apiculate. Mature stipules are typically obtuse in C. liberica, but the immature apical stipule can occasionally be somewhat apiculate. In choosing the epithet C. liberica it seems that Hiern primarily had the West African element in mind, rather than the Angolan; it is from amongst the three West African specimens that the lectotype should be chosen. The Vogel specimen has inflorescences from which the corollas have all fallen. The Afzelius specimen has not been traced at the British Museum (Nat. Hist.), but five Afzelius sheets at Uppsala (isosyntypes) have been seen and are all sterile. This leaves the Daniell sheet as the most likely candidate for lectotype of C. liberica; its selection is not, however, without difficulties. Hiern claims that the plate (cit. supra) was drawn from the Daniell specimen and that flowers had been preserved in liquid. It is debatable whether this illustration was accurately or entirely based on this specimen. The mature stipules are shown as acuminate, while those on the Daniell sheet are obtuse; it is possible that this element of the drawing was based on the Angolan specimens of C. canephora. The illustration shows a stem with young buds at the