Impatiens lyallii Baker, 1883

Abrahamcyk, Stefan & Fischer, Eberhard, 2015, Impatiens elianae (Balsaminaceae), a new species from central Madagascar, with notes on the taxonomic relationship of I. lyallii and I. trichoceras, Phytotaxa 226 (1), pp. 83-91 : 86-88

Identity of Impatiens lyallii Baker and I. trichoceras Baker

Impatiens lyallii Baker (1883: 114) was described from central Madagascar based on a collection made by Lyall as Balsamina glandulifera ( Baker 1883) . In the protologue, additional specimens collected by Meller, Baron and Parker were cited. Baker states that it “is one of the finest species of the whole genus, and would be very suitable to introduce for horticultural purposes” ( Baker 1883). Specimen Lyall 49 ( Fig. 3 ) fits well into the description of the protologue; it has been extensively studied and annotated by J.D. Hooker, and is designated here as lectotype. A diagnostic feature, which makes I. lyallii easy to recognize, is the almost glabrous oblong-lanceolate leaf ( Figs. 3 , 4 ).

Impatiens trichoceras Baker (1883: 116) was described in the same publication, also based on a specimen collected by Lyall (Lyall 46, Fig. 5 ). It differs from I. lyallii in numerous characters, i.e. shape of the leaf-lamina with the long petiole, shape of the extrafloral nectaries, shape of the dorsal petal and the lateral united petals, densely pilose stems, leaves and lower sepal with spur, and a pilose ovary. Both species have either been confused or misinterpreted for a long time. Costantin & Poisson (1907) considered I. lyallii to have opposite leaves and thus approaching I. vilersii Costantin & Poisson (1907: 469) . This observation, however, is erroneous as the type clearly has alternate leaves.

Perrier de la Bâthie (1934) included I. trichoceras into a variable I. lyallii . He confirmed that I. lyallii differs from I. vilersii in having an alternate leaf arrangement. Impatiens lyallii sensu Perrier de la Bâthie (1934) is said to be variable concerning the leaf dimensions, the pilosity and the dimensions of the floral parts. Forms with a pubescent ovary are distinguished as f. trichogyna Perrier de la Bâthie (1934: 51) from the typical form ( f. typica Perrier de la Bâthie 1934: 52 , nom. illeg.).

During field work in Madagascar, plants with oblong-lanceolate leaves and almost glabrous flowers with spur and ovary were encountered near Mandraka (east of Antananarivo; Fig. 4 ). These plants exactly match the type of I. lyalli . On the other hand, specimens matching typical I. trichoceras have been observed repeatedly, e.g. at Ambalamanakana, south of Fianarantsoa. However, no intergradations between both taxa could be detected. Field observations from different localities suggest that more than three species are involved in the Impatiens lyallii -complex. This will be the subject of a further publication.

Figures

FIGURE 3. Impatiens lyallii. A–E, lectotype (K). Scale bar = 10 mm.

FIGURE 4. Impatiens lyallii. A, habit. B, flower; C, flower and young fruit. Plants at Mandraka, Madagascar. Scale bar = 10 mm. Photographs: E. Fischer.

FIGURE 5. Impatiens trichoceras. A–F, holotype (K). Scale bars:A= 20 mm, B, E = 10 mm.