Monographs Details:
Authority:
Luteyn, James L. 1983. Ericaceae--part I. Cavendishia. Fl. Neotrop. Monogr. 35: 1-290. (Published by NYBG Press)
Luteyn, James L. 1983. Ericaceae--part I. Cavendishia. Fl. Neotrop. Monogr. 35: 1-290. (Published by NYBG Press)
Family:
Ericaceae
Ericaceae
Synonyms:
Cavendishia gilgiana Hoerold
Cavendishia gilgiana Hoerold
Description:
Variety Description - Inflorescence 5-6-flowered; rachis 0.5-2.5 cm long; floral bracts 35-53 X 2235 mm, apically rounded or rarely shallowly emarginate; pedicels 10-20 mm long; bracteoles 5-7 mm long. Calyx 7-11.5 mm long; limb 4-7.5 mm long; corolla 33-40 mm long and 8-10 mm in diam.; stamens 21-26 mm long.
Variety Description - Inflorescence 5-6-flowered; rachis 0.5-2.5 cm long; floral bracts 35-53 X 2235 mm, apically rounded or rarely shallowly emarginate; pedicels 10-20 mm long; bracteoles 5-7 mm long. Calyx 7-11.5 mm long; limb 4-7.5 mm long; corolla 33-40 mm long and 8-10 mm in diam.; stamens 21-26 mm long.
Discussion:
The fruits are edible and are quite juicy although not very sweet (pers. obs.).Cavendishia tarapotana is a very showy species easily recognized in the field by its white corolla with a red band around the constricted throat and its dark pink to glossy red floral bracts. It would make a very handsome ornamental shrub. The following combination of characters also serves to distinguish the species: 1) leaves often with a pinkish hue when dry; 2) thinly apophysate hypanthium which extends below the articulation and encircles the pedicel; 3) calyx limb longer than hypanthium at anthesis (although once the corollas have fallen the ovary quickly expands); 4) ovary (hypanthium) broader than long after anthesis with the pericarp appearing loosely attached; 5) ovary (hypanthium) usually translucent pale green or whitish to pinkish after anthesis but before fruit maturation; 6) calyx lobes often strongly deltate and short relative to overall limb length; and 7) calyx lobes connivent-twisting around the base of the style.Within the species significant variation occurs in the degree of fusion (or lack thereof) of the glandular fimbriae along the calyx lobe margins. The type of C. tarapotana from northeastern Peru displays mostly distinct fimbriae although a few may coalesce towards the lobe apex to form short oblong glands. Cavendishia weberbaueri, also from northern Peru, is similar to C. tarapotana in calyx glands and differs only in a somewhat longer acuminate leaf apex. In 1937, Smith described C. mexiae from eastern Ecuador but failed, at that time, to note its relationship to C. tarapotana or C. weberbaueri. It differed from these two in having leaves abruptly short-caudate-acuminate apically, and in the almost completely fused fimbriae on the calyx lobes. In 1950, Smith annotated the US isotype (of C. mexiae) “why not C. tarapotana?' Cavendishia gilgiana, described from western Ecuador, was characterized by its short, few-flowered racemes, long floral bracts and corollas, and glandular-callose calyx lobes.With the exception of Smith’s 1950 manuscript query on that of C. mexiae, the four types mentioned above have not previously been considered related to each other. Each was collected from a different location and each species was described only from a single gathering, none of which had been seen in the field by its describer.My extensive fieldwork and the accumulation of many collections from various localities throughout the geographical range have convinced me that only one species can be recognized. As mentioned earlier, the variation in the calyx lobe glands is considerable, involving progressive lateral fusion of the fimbriae into short or long, usually thin supramarginal glands which do not continue to the lobe apex (see Fig. 56). This fusion may rarely be noticed in a single calyx, one lobe of which may have distinct fimbriae while an adjacent one has them distally coalescent. Geographically, free fimbriae are most prevalent in northern Peru. The eastern Ecuadorian-Colombian collections show higher degrees of fusion, with the continuous marginal glands being thin and often undulate or irregular where fusion has occurred. In western Ecuador (Pichincha Prov.) the fusion is usually complete and the glands seem somewhat thicker than those from the eastern Andes. This stage may represent a phylogenetic link to sect. Engleriana where the marginal glands are conspicuously thickened. In a few unusual collections from Carchi Prov., Ecuador (Gentry & Shupp 26397, 26455; Madison 3928) and adjacent Nariño Dept., Colombia (Ewan 16046; López-Palacios 3797; Luteyn & Lebrón-Luteyn 6811) the calyx lobes either lack fimbriae or possess only a few which are early deciduous.The varieties of C. tarapotana are easily separated by the characters mentioned in the key. Two collections (López-Palacios 3791 and Luteyn & Lebrón-Luteyn 6811), however, from Nariño Dept., Colombia, appear somewhat intermediate between var. tarapotana and var. gilgiana in corollas 26-29 mm, floral bracts 22-27 mm, and rachises varying on the same plant from 1.6 cm long (7-flowered) to 5.5 cm (12-flowered). Their superficial appearance and habit is that of var. gilgiana.Within ser. Cavendishiae, C. tarapotana is most closely related to C. colombiana. They have in common a relatively short calyx limb which, with the lobes, is connivent-twisting after anthesis; short, deltate or broadly triangular calyx lobes usually broader than long; and apophysate hypanthia. The differences are mentioned in the key.Cavendishia tarapotana var. tarapotana may sometimes be confused with C. bracteata when only non-flowering or early fruiting material is available. In these cases the following features may be helpful in differentiating the taxa: 1) the hypanthium of C. tarapotana is basally apophysate with a smooth rim 0.5-1 mm long produced beyond (and below) the articulation and encircling the distal portion of the pedicel, whereas only very rarely does the hypanthium of C. bracteata have a rim which is at best very short and quite inconspicuous; 2) the “loose” pericarp and broader than long immature fruit of C. tarapotana-, 3) the connivent-twisting calyx limb/lobes of C. tarapotana-, and 4) the inflorescence of C. tarapotana which often has the proximal flowers in fruiting stages or already fallen when the distal ones are still in bud and the rachis is still elongating, in contrast to that of C. bracteata in which all the flowers mature simultaneously.In the collections available to me there exist three series of unusual plants from different geographical areas somewhat outside of the “normal” range of C. tarapotana. These plants display features which suggest hybrid origin.The first series is from Santander and Norte Santander Depts., Colombia. I suspect they represent hybrids between C. tarapotana var. tarapotana and some form of C. bracteata. The C. tarapotana features include: 1) long rachises often still elongating and in bud distally while in post-anthesis proximally; 2) calyx limbs about equal to or slightly longer than the hypanthium; 3) hypanthia basally slightly apophysate and “loose”; and 4) maturing ovaries (hypanthia) which are broader than long. Features which may be attributed to C. bracteata are: 1) leaves smaller and of a texture thicker than normal for C. tarapotana; 2) corollas red in color; 3) calyx lobes glandular-callose to the tips; 4) habitats which are open, exposed, and at higher elevations (2100-3000 m).The second series is found in northwest Antioquia Dept., Colombia and corresponds to C. kalbreyeri, described from a single collection from 2896 m elevation. Smith (1932) saw only the type collection and characterized the species by an elongate calyx limb, short rachis, relatively short corolla, and marginally thickened calyx lobes. I recollected the species in flower (Luteyn et al. 7055) and fruit (Luteyn et al. 7047 and 7053) and noted that, as in the type, the corolla was white. It also has the following features: 1) small leaves to 10 cm long of thin texture; 2) rachis thin and to 3 cm long with no more than 12 flowers; 3) floral bracts ca. 25-32 mm long; 4) elongate calyx limb with very short and broad lobes which are glandular-callose; and 5) corollas 23-30 mm long. It is a poorly defined species allied to C. tarapotana. This population may have arisen by hybridization between C. tarapotana var. tarapotana and C. bracteata, or possibly even between the two varieties of C. tarapotana.The third series of collections is from central Chocó and adjacent Risaralda Depts., Colombia at elevations of 500-1700 m (e.g. Forero et al. 3184. Luteyn & Lebrón-Luteyn 7207 and 7231). They feature: 1) elongate rachises (6-12 cm long); 2) calyx limbs longer than hypanthium; 3) corollas varying from 20 mm to 32 mm long; and 4) small to large leaves. The corollas in Luteyn collections were orangish-red with white tips and the fruits seemed very large.These collections, although somewhat intermediate in their morphologies, may allow some speculation as to the past and/or present distribution of C. tarapotana. They may indicate that in early tertiary times C. tarapotana had a much broader distribution in northwestern South America, but as the Andes reached their present altitudes populations were isolated on either side of the mountains. These populations, now differentiated into var. tarapotana on the eastern slopes and var. gilgiana on the western, may in turn, have suffered further reduction in range, come into secondary contact and hybridized with other species (e.g. C. bracteata), thereby giving rise to some of the variants.
The fruits are edible and are quite juicy although not very sweet (pers. obs.).Cavendishia tarapotana is a very showy species easily recognized in the field by its white corolla with a red band around the constricted throat and its dark pink to glossy red floral bracts. It would make a very handsome ornamental shrub. The following combination of characters also serves to distinguish the species: 1) leaves often with a pinkish hue when dry; 2) thinly apophysate hypanthium which extends below the articulation and encircles the pedicel; 3) calyx limb longer than hypanthium at anthesis (although once the corollas have fallen the ovary quickly expands); 4) ovary (hypanthium) broader than long after anthesis with the pericarp appearing loosely attached; 5) ovary (hypanthium) usually translucent pale green or whitish to pinkish after anthesis but before fruit maturation; 6) calyx lobes often strongly deltate and short relative to overall limb length; and 7) calyx lobes connivent-twisting around the base of the style.Within the species significant variation occurs in the degree of fusion (or lack thereof) of the glandular fimbriae along the calyx lobe margins. The type of C. tarapotana from northeastern Peru displays mostly distinct fimbriae although a few may coalesce towards the lobe apex to form short oblong glands. Cavendishia weberbaueri, also from northern Peru, is similar to C. tarapotana in calyx glands and differs only in a somewhat longer acuminate leaf apex. In 1937, Smith described C. mexiae from eastern Ecuador but failed, at that time, to note its relationship to C. tarapotana or C. weberbaueri. It differed from these two in having leaves abruptly short-caudate-acuminate apically, and in the almost completely fused fimbriae on the calyx lobes. In 1950, Smith annotated the US isotype (of C. mexiae) “why not C. tarapotana?' Cavendishia gilgiana, described from western Ecuador, was characterized by its short, few-flowered racemes, long floral bracts and corollas, and glandular-callose calyx lobes.With the exception of Smith’s 1950 manuscript query on that of C. mexiae, the four types mentioned above have not previously been considered related to each other. Each was collected from a different location and each species was described only from a single gathering, none of which had been seen in the field by its describer.My extensive fieldwork and the accumulation of many collections from various localities throughout the geographical range have convinced me that only one species can be recognized. As mentioned earlier, the variation in the calyx lobe glands is considerable, involving progressive lateral fusion of the fimbriae into short or long, usually thin supramarginal glands which do not continue to the lobe apex (see Fig. 56). This fusion may rarely be noticed in a single calyx, one lobe of which may have distinct fimbriae while an adjacent one has them distally coalescent. Geographically, free fimbriae are most prevalent in northern Peru. The eastern Ecuadorian-Colombian collections show higher degrees of fusion, with the continuous marginal glands being thin and often undulate or irregular where fusion has occurred. In western Ecuador (Pichincha Prov.) the fusion is usually complete and the glands seem somewhat thicker than those from the eastern Andes. This stage may represent a phylogenetic link to sect. Engleriana where the marginal glands are conspicuously thickened. In a few unusual collections from Carchi Prov., Ecuador (Gentry & Shupp 26397, 26455; Madison 3928) and adjacent Nariño Dept., Colombia (Ewan 16046; López-Palacios 3797; Luteyn & Lebrón-Luteyn 6811) the calyx lobes either lack fimbriae or possess only a few which are early deciduous.The varieties of C. tarapotana are easily separated by the characters mentioned in the key. Two collections (López-Palacios 3791 and Luteyn & Lebrón-Luteyn 6811), however, from Nariño Dept., Colombia, appear somewhat intermediate between var. tarapotana and var. gilgiana in corollas 26-29 mm, floral bracts 22-27 mm, and rachises varying on the same plant from 1.6 cm long (7-flowered) to 5.5 cm (12-flowered). Their superficial appearance and habit is that of var. gilgiana.Within ser. Cavendishiae, C. tarapotana is most closely related to C. colombiana. They have in common a relatively short calyx limb which, with the lobes, is connivent-twisting after anthesis; short, deltate or broadly triangular calyx lobes usually broader than long; and apophysate hypanthia. The differences are mentioned in the key.Cavendishia tarapotana var. tarapotana may sometimes be confused with C. bracteata when only non-flowering or early fruiting material is available. In these cases the following features may be helpful in differentiating the taxa: 1) the hypanthium of C. tarapotana is basally apophysate with a smooth rim 0.5-1 mm long produced beyond (and below) the articulation and encircling the distal portion of the pedicel, whereas only very rarely does the hypanthium of C. bracteata have a rim which is at best very short and quite inconspicuous; 2) the “loose” pericarp and broader than long immature fruit of C. tarapotana-, 3) the connivent-twisting calyx limb/lobes of C. tarapotana-, and 4) the inflorescence of C. tarapotana which often has the proximal flowers in fruiting stages or already fallen when the distal ones are still in bud and the rachis is still elongating, in contrast to that of C. bracteata in which all the flowers mature simultaneously.In the collections available to me there exist three series of unusual plants from different geographical areas somewhat outside of the “normal” range of C. tarapotana. These plants display features which suggest hybrid origin.The first series is from Santander and Norte Santander Depts., Colombia. I suspect they represent hybrids between C. tarapotana var. tarapotana and some form of C. bracteata. The C. tarapotana features include: 1) long rachises often still elongating and in bud distally while in post-anthesis proximally; 2) calyx limbs about equal to or slightly longer than the hypanthium; 3) hypanthia basally slightly apophysate and “loose”; and 4) maturing ovaries (hypanthia) which are broader than long. Features which may be attributed to C. bracteata are: 1) leaves smaller and of a texture thicker than normal for C. tarapotana; 2) corollas red in color; 3) calyx lobes glandular-callose to the tips; 4) habitats which are open, exposed, and at higher elevations (2100-3000 m).The second series is found in northwest Antioquia Dept., Colombia and corresponds to C. kalbreyeri, described from a single collection from 2896 m elevation. Smith (1932) saw only the type collection and characterized the species by an elongate calyx limb, short rachis, relatively short corolla, and marginally thickened calyx lobes. I recollected the species in flower (Luteyn et al. 7055) and fruit (Luteyn et al. 7047 and 7053) and noted that, as in the type, the corolla was white. It also has the following features: 1) small leaves to 10 cm long of thin texture; 2) rachis thin and to 3 cm long with no more than 12 flowers; 3) floral bracts ca. 25-32 mm long; 4) elongate calyx limb with very short and broad lobes which are glandular-callose; and 5) corollas 23-30 mm long. It is a poorly defined species allied to C. tarapotana. This population may have arisen by hybridization between C. tarapotana var. tarapotana and C. bracteata, or possibly even between the two varieties of C. tarapotana.The third series of collections is from central Chocó and adjacent Risaralda Depts., Colombia at elevations of 500-1700 m (e.g. Forero et al. 3184. Luteyn & Lebrón-Luteyn 7207 and 7231). They feature: 1) elongate rachises (6-12 cm long); 2) calyx limbs longer than hypanthium; 3) corollas varying from 20 mm to 32 mm long; and 4) small to large leaves. The corollas in Luteyn collections were orangish-red with white tips and the fruits seemed very large.These collections, although somewhat intermediate in their morphologies, may allow some speculation as to the past and/or present distribution of C. tarapotana. They may indicate that in early tertiary times C. tarapotana had a much broader distribution in northwestern South America, but as the Andes reached their present altitudes populations were isolated on either side of the mountains. These populations, now differentiated into var. tarapotana on the eastern slopes and var. gilgiana on the western, may in turn, have suffered further reduction in range, come into secondary contact and hybridized with other species (e.g. C. bracteata), thereby giving rise to some of the variants.
Distribution:
Colombia South America| Nariño Colombia South America| Ecuador South America| Carchi Ecuador South America| Pichincha Ecuador South America|
Colombia South America| Nariño Colombia South America| Ecuador South America| Carchi Ecuador South America| Pichincha Ecuador South America|
Common Names:
espelma
espelma
Multimedia:
