Monographs Details: Cavendishia engleriana var. ecuadorensis Luteyn
Authority: Luteyn, James L. 1983. Ericaceae--part I. Cavendishia. Fl. Neotrop. Monogr. 35: 1-290. (Published by NYBG Press)
Description:Latin Diagnosis - A var. engleriana foliis hispidis (non glabris) 3-nerviis (non 5-7-nerviis) basi cuneatis (non rotundatis vel obtusis), bracteis floralibus 30-36 (non 35-50) mm longis marginibus glanduliferis, pedicellis 5-8 (non 7-15) mm longis, et distri-butione geographica differt.

Variety Description - Stems glaucous, very brittle when fresh. Leaves coriaceous, very hard and brittle when fresh, (6-)10-14.5(-18) X (1.2-)2-3(3.5) cm, basally cuneate, marginally slightly revolute, moderately hispid on both surfaces with hairs 1-1.5 mm long; 3-plinerved; petioles 6-8 mm long and l(-2) mm in diam. Inflorescence bracts 25-32 X 15-24 mm; rachis 1-1.4 cm long; floral bracts 30-36 X 15-24 mm, marginally densely glandular with spherical, sessile glands 0.2 mm diam.; pedicels 5-8 mm long and 1.5 mm diam.; bracteoles 4-7 mm long. Flowers: hypanthium 2-3 mm long; limb 6-7 mm long; lobes 1.5-2.5 mm long; corolla 27-30 mm long; stamens 23-27 mm long; filaments alternately 3.2-4.2 mm and 9.2-10.2 mm long; anthers alternately 21.5-23 mm and 16-18 mm long; thecae alternately 12-13 mm and 9-10 mm long.

Discussion:Cavendishia engleriana is characterized by marginally glandular-callose calyx lobes, glandular pedicels, and linear bracteoles. Hoerold (1909) failed to mention any of these characters in the protologue and this is probably the main reason for subsequent authors’ misunderstanding of his species.

In 1932, Smith saw Hoerold’s type at Berlin and noted the glandular calyces and pedicels but neglected to mention (or did not see) the linear bracteoles. Consequently he overlooked any relationship between C. engleriana and his new C. marginata which he characterized by its glandular-callose calyx lobe margins, linear bracteoles, and eglandular pedicels.

In 1977, I had only the type photograph and Hoerold’s protologue of C. engleriana available for comparison and thus saw no relationship between that species and my new C. dodsonii. I did note that the latter was close to C. marginata, however, separable on the basis of glandular pedicels, differently shaped leaves, differently proportioned calyces, and geographical location.

After considerable field study throughout the range of these proposed species, it is obvious that only one can be recognized, C. engleriana. There is morphological variation in leaf size and shape, in the prominence of glands on the pedicels, in presence or absence of floral bract glands, and in leaf pubescence. Some pedicels on the type of C. marginata do possess glands although they are obscure. Other populations that have pedicels with obscure glands are Luteyn et al. from the type region of C. marginata, Gentry & Shupp 26538 and 26573 from Carchi Prov., Ecuador, and Luteyn & Lebrón-Luteyn 6875 from Nariño Dept., Colombia. Leaf variation may be noted as follows: Cauca Dept, and higher elevations (1800 m) in Nariflo Dept., Colombia-leaves thick coriaceous, elliptic-lanceolate, long-acuminate, 3-6 cm broad, glabrous; lower elevations in Nariño Dept., Colombia and adjacent Carchi Prov., Ecuador-leaves coriaceous, usually broadly oblong-elliptic, abruptly caudate-acuminate, 5-7 cm broad, glabrous; Esmeraldas Prov., Ecuador-leaves drying membranaceous to subcoriaceous, elliptic to broadly obovate or oblanceolate, abruptly caudate-acuminate, 4-7 cm broad, glabrous; Pichincha and Los Rios Provs., Ecuador-subcoriaceous, narrowly oblong, elliptic or oblanceolate, short-acuminate, 2-4 cm broad, glabrous; Napo and Pastaza Provs., Ecuador-coriaceous, brittle, narrowly oblong, elliptic or oblanceolate, short-acuminate, 2-3.5 cm broad, hispid. The floral bracts of var. engleriana are marginally scabrous and lack glands, whereas those of var. ecuadorensis possess spherical glands along the margins.

Luteyn & Lebrón-Luteyn 5816 (Tungurahua Prov., Ecuador) is an unusual collection and may be of hybrid origin. It has the leaves and gland-bearing twigs and pedicels of C. engleriana, but thin, elongate rachises (2-3 cm long), a narrow apophysate hypanthium, short and broad calyx lobes bearing oblong marginal glands which are not especially thickened, calyx lobes somewhat connivent after anthesis, and a style which indicates a corolla ca. 27 mm long. The nature of the calyx lobes suggests to me possible hybridization with either C. tarapotana or C. bracteata both of which are locally common (although “pure” C. engleriana has yet to be collected in Tungurahua Prov.).

In the buds of Luteyn & Lebrón-Luteyn 5640 the pedicels are densely beset with large, plump glands. However, the post anthesis-early fruiting pedicels show few, scattered glands which are shriveled and mostly deciduous. This suggests that the function of these glands is to provide moisture for the flowers needed during developmental stages (Luteyn, 1976).

Cavendishia engleriana is most closely related to C. urophylla and their similarities and differences are mentioned under the latter species. It is also probably related to the Panamanian C. santafeensis, although that species has large, spatulate, eglandular bracteoles (36-44 X 10-12 mm not 6-14 X 1-2 mm), eglandular pedicels, and apically emarginate bracteoles and floral bracts.

A more remote relationship between C. engleriana and C. nitens and adenophora may be indicated by the occurrence of slightly stalked bracteole glands on Prescott 1295 from Pichincha Province.
Distribution:Ecuador South America| Napo Ecuador South America| Pastaza Ecuador South America|