Tabebuia heterophylla (DC.) Britton

Authority

Gentry, Alwyn H. 1992. Bignoniaceae--part II (Tribe Tecomeae). Fl. Neotrop. Monogr. 25: 1-370. (Published by NYBG Press)

Family

Bignoniaceae

Scientific Name

Tabebuia heterophylla (DC.) Britton

Type

Type. Puerto Rico. Bertero s.n. (G-DC).

Synonyms

Bignonia leucoxylon L., Bignonia pentaphylla L., Bignonia leucoxylon L., Raputia heterophylla DC., Leucoxylon riparia Raf., Leucoxylon acuminata Raf., Tabebuia triphylla A.DC., Tecoma triphylla Mart. ex DC., Tecoma pentaphylla Juss., Tabebuia pentaphylla (L.) Hemsl., Tabebuia riparia (Raf.) Sandwith, Tabebuia arenicola Britton, Tabebuia geronensis Britton, Tabebuia curtissii Britton, Tabebuia lucida Britton, Tabebuia camagueyensis Britton & P.Wilson, Tecoma eggersii F.Kränzl., Tabebuia gonavensis Urb., Tabebuia leptopoda Urb., Tabebuia beyeri Urb. & Ekman, Tabebuia dictyophylla Urb., Tabebuia lindahlii Urb. & Ekman, Tabebuia brigandina Urb. & Ekman, Tabebuia pallida subsp. heterophylla (DC.) Stehlé, Tabebuia pallida subsp. pentaphylla (L.) Stehlé, Tabebuia heterophylla (DC.) Britton, Handroanthus pentaphyllus (L.) Mattos, Tabebuia capotei Borhidi

Description

Species Description - Shrub or small to large tree to 20 m or more, dichotomously branched, the branchlets terete, lepidote with uniformly small, mostly whitish sessile (occasionally in part subsessile) scales. Leaves mostly 3-5-foliolate, frequently with the basal leaves of some branchlets 1 -foliolate, rarely (Tortue Island: T. lindahlii) mostly unifoliolate, the leaflets highly variable in size, shape, and texture, the terminal mostly more or less obovate or obovate-elliptic, the laterals usually elliptic or oblong-elliptic, the apex and base obtuse to rounded, the apex occasionally obtusely apiculate but never acuminate, occasionally slightly retuse (T. camaguayensis, also sometimes as result of individual injury), the terminal 1-16 cm long, 0.4-7.5 cm wide, the basals 0.7-12 cm long, 0.3-6 cm wide, more or less coriaceous, densely lepidote above and below, the scales all whitish or with a few scattered sessile reddish ones below (sometimes the undersurface distinctly whitish but only in Jamaica and the Cayman Islands), the secondary veins relatively inconspicuous and usually not strongly brochidodromous, plane above, raised below, the tertiary venation more or less plane above, varying from subplane to distinctly prominulous below, the margin usually entire, sometimes very slightly erose (=T. lucida, T. dictyophylla, T. arenicola); the terminal petiolule 0.2-5.5 cm long, the basals usually asymmetrically subsessile (occasionally the petiolule to 0.5-1 cm long), the petiole 0.5-8(-l 4) cm long, lepidote, usually drying tannish or grayish. Inflorescence a few- to several(-many)-flowered terminal panicle, often reduced to one or two flowers, lepidote. Flowers with the calyx cupular, irregularly 2-3(-4)-labiate, 7-12 mm long, 5-8 mm wide, lepidote with sessile trichomes, frequently drying grayish; corolla lavender or pale magenta to almost white (white to faintly pinkish on Jamaica: T. riparia), the throat yellow when fresh, fading to white, tubular-infundibuliform, 3.5-7 cm long, 1-2 cm wide at mouth of tube, the tube 3-5.5 cm long, the lobes 0.8-2 cm long, glabrous outside, scurfy puberulous in throat inside, strongly villous at level of stamen insertion, the lobes more or less ciliate; anthers included in lower part of tube, the thecae divaricate, 3 mm long; ovary linear, lepidote, 4 mm long, 1 mm wide, the ovules 2-seriate in each locule; disk annular-pulvinate, 1-1.5 mm long, 2 mm wide. Fruit a linear-cylindric capsule, attenuate toward base and apex, 7-20 cm long, 6-10 mm wide, the valves coriaceous, slightly to not at all striate-costate, densely lepidote, the calyx persistent; seeds thin, bialate, 7-9 mm long, 20-30 mm wide, the hyaline-membranaceous wings sharply demarcated from the seed body.

Discussion

Uses. Widely cultivated as an ornamental; also used for timber. According to Hodge and Taylor (1957) it lasts especially well in salt water and is thus used for the ribs and sides of dugouts on Dominica, the wood also being employed in the construction of mortars, wooden bowls, and floors. It is also used medicinally. For example, on Dominica a plaster of the bark is used to cure corns and the leaves as a poultice on sores (Hodge & Taylor, 1957), and the leaves and bark are boiled as a cure for colds on St. Christopher (Cooley 8797). This is probably the most variable of all species of Tabebuia. Its patterns of variation are extremely complex and I have here included extremes that would not seem possibly to be conspecific, but that are connected by a complete series of intergradations. Even many of the related species that are recognized intergrade with T. heterophylla, and many specimens cannot be assigned with certainty to a specific taxon. In adopting a rather broad concept of T. heterophylla, I have been strongly influenced by the fact that on Puerto Rico, where the flora is extremely well known, all botanists (Woodbury, pers. comm.; Vivaldi, pers. comm.; Liogier, pers. comm.) are in agreement that there is a single highly variable pink-flowered species, i.e., T. heterophylla (along with 3 red-flowered ones). In dry coastal scrub on Puerto Rico this species may have uniformly small leaves with terminal leaflets ca. 3 x 1 cm, while in moist interior forests terminal leaflets may be around 15x7 cm. The total variation in the complex is not much greater on Cuba, but it is partitioned much more discretely, often associated with ecotypic specialization, e.g., on serpentine. Hispaniola actually has more morphological variation in Tabebuia than does Cuba, although fewer species have been recognized; in part this is because the Hispaniolan species tend to be better defined so that there has been less temptation to describe minor variants. Stehle (1946) has discussed the taxonomic problems entailed by such a polymorphic species, which he referred to as T. pallida. He suggests recognizing three subspecies—ssp. genuina for the small-leaved coastal scrub form, ssp. pallida for the widespread 3-5-foliolate form, and ssp. dominicensis for the unifoliolate form here treated as T. pallida. Stehle’s concept of ssp. pallida cannot be given that name since the type illustration of T. pallida shows a unifoliolate plant. A new combination at the subspecific level based on the basionym of T. riparia (or on T. triphylla if the Jamaican population is considered subspecifially distinct) would then be needed. I have tried very hard to recognize more species in this complex but find that not only is there instability in all the potential taxonomic characters and numerous intermediates between local morphotypes, but also that the variation is differently partitioned on different islands, so that the complex of characters that might define a species on one isiand varies in an overlapping manner from island to island. In such a situation it is patently impossible to construct a serviceable key except by lumping many of the more poorly differentiated entities. This species also has nomenclatorial problems. The nomenclatural confusion associated with T. heterophylla has been clearly elucidated by Sandwith (1954, 1955). Linnaeus himself began the confusion by splitting his Bignonia leucoxylon of Species Plantarum, edition 1, in two in edition 2, unfortunately applying the new name, B. pentaphylla to most of the citations he had originally cited under B. leucoxylon and applying B. leucoxylon to a Miller (Gardeners Dictionary, ed. 7, 1759, Bignonia No. 10) citation “Bignonia foliis imiis digitatis, superioribus sim-plicibus vel tematis caule erecto arboreo” that I here typify by Houston s.n., Jamaica 1731 (BM) instead; Miller cited Plukenet’s “Leucoxylon arbor siliquosa, quinis foliis, floribus Nerii, alato semine,” also included in Linnaeus’ Species Plantarum ed. 1 (sphalm Plum.) citation and no doubt the source of the epithet “leucoxylon.” Thus it might be argued that Linnaeus in effect lectotypified his original Bignonia leucoxylon with the Plukenet reference, as indicated by his retention of the epithet “leucoxylon” when he decided that Plukenet’s plant was different from the plants of Plumier, Catesby, Sloane, and Ray that were also cited in the original description and in the Hortus Cliffortianus from which the diagnostic phrase was taken. Linnaeus’s (1763) amended B. leucoxylon was said to have ovate acuminate leaflets while B. pentaphylla had obovate leaflets. However, Plukenet’s (1692) figure shows a 5-foliolate leaf with leaflets that are elliptic and acutish (not acuminate). It is indeed true that references pertinent to several Tabebuia species were included under B. leucoxylon in Species Plantarum, ed. 1 (and before that in the Hortus Cliffortianus). Catesby’s plant, excellently illustrated, is clearly T. bahamensis. Plumier (1703) referred to two plants, “Bignonia arbor, pentaphylla flores roseo, major, siliquis planis” and “Bignonia arbor, pentaphylla, flore rosea, minor, siliquis angulatis.” Thanks to the assistance of Dr. Alicia Lourteig, I have examined tracings of Plumier’s original, never published, drawings of these two Tabebuia species. Although not indicated in the publication, the large-flowered smooth-fruited Plumier species is additionally characterized on the plates as “In Ins. Antillanis. ‘poirier’ ” and is clearly T. heterophylla as here used. The second Plumier species with angulate fruit is indicated as “Ins. Sto. Domingo” and appears to be T. calcicola Britt. Although Linnaeus included both “flore ... major & minor” in his Plumier citation, he excluded “siliquis angularis” effectively excluding the Santo Domingo plant as a possible lectotype for the second B. leucoxylon. The Sloane and Ray citations refer to yet another plant, the Jamaican “white wood,” the species with obtuse leaflet apices that was treated by Sandwith (1955) and in the Flora of Jamaica (Adams, 1972) as T. riparia. Linnaeus evidently intended to associate B. leucoxylon with Plukenet’s plant from whence its epithet had come. Thus one might argue that the Plukenet plate should be considered the type of B. leucoxylon, leaving B. pentaphylla as a perfectly valid name, with several discordant elements, one of which should be selected as lectotype. However, Sandwith (1954) considered that Bignonia leucoxylon of Species Plantarum edition 1 should be typified with the Hortus Cliffortianus collection at LINN, making B. pentaphylla an illegitimate name change. In the interest of nomenclatural stability, I opt to accept Sandwith’s lectotypification, especially since the name Tabebuia pentaphylla (L.) Hemsl. has been widely used for the quite different continental species correctly known as T. rosea; resurrection of T. pentaphylla at this late date for an Antillean species would cause tremendous confusion. There is a taxonomic problem associated with the Jamaican “white wood,” as well as the nomenclatural one. Sandwith (1954) considered Tabebuia heterophylla “wholly distinct” from the Jamaican “white wood.” Yet I am unable to discern any firm characters to differentiate it from the widespread Antillean plant. The only potential differences are the white flowers and generally narrower leaflets with more densely whitish lepidote leaf undersurfaces in the Jamaican population. However, white flowers occur in most normally lavender-flowered species of Bignoniaceae and very pale almost white flowers are known in some extra-Jamaican populations of T. heterophylla, some Jamaican collections are indicated by their collectors as having a slight pink tint (e.g., Stearn 743). Outside Jamaica, collections of T. heterophylla with the leaflets strongly whitish beneath are known only from the nearby Cayman Islands, where they are associated with the normal pink to lavender flower color. The character of whitish leaf undersurface is very inconstant, even in Jamaica, and occurs in less than half the Jamaican collections. Moreover, T. heterophylla is exceedingly variable. I conclude that the Jamaican population is best included in T. heterophylla as well. Other problem forms within T. heterophylla include T. lindahlii from Haiti which has predominantly unifoliate leaves with raised secondary veins and prominulously raised-reticulate tertiary vein network below. This form approaches T. berteroi, and especially the Cuban form of that species (T. anisophylla). Tabebuia beyeri, known only from the type from Matanzas, is supposed to be characterized by a raised midrib and sparsely lepidote leaves with the laterals having strongly asymmetric bases (Alain, 1957). However, the leaves of the type that I examined are strongly and conspicuously lepidote, especially the younger ones, the midribs of some leaflets are not raised at all, and some lateral leaflets have very weakly asymmetric bases. Since none of the characters used to define the species are consistent, even on the type, it can hardly be maintained. Tabebuia arenicola, a beach form from Oriente with rather small narrow leaves, is supposed to be distinguished by having a shinier surface and less conspicuous upper surface venation but the leaf surface of the type is actually duller than in most material of the heterophylla complex, while Matanzas material referred to this species has typically developed upper surface leaf venation. Tabebuia dictyophylla, from the karst area of Matanzas Province has the characteristic thin, strongly erose-margined intricately prominulously veined leaflets typical of karst plants but does not seem worthy of specific recognition; the Pinar del Rio collections made by Leon & Alain and identified as this species are quite different, with thickly coriaceous, strongly emarginate leaflets, and belong in T. lepidota as here interpreted. Tabebuia arimaoensis which I have recognized as specifically distinct, is close to this form. Tabebuia capotei from the Guanahacabibes Península of westernmost Cuba is apparently an ecotype of T. heterophylla characterized by unusually small leaflets with the venation below unusually prominulous. The large-leafleted forms (T. leptoneura, T. elongata, T. shaferi sensu Leon & Alain) from Cuba pose even worse problems. These plants seem distinct from other populations of T. heterophylla on Cuba, especially on the basis of the longer basal petiolule and generally larger and sometimes more numerous leaflets, and I have treated them here as specifically distinct. Nevertheless, they are within the range of variation shown by the complex outside of Cuba. Alain (1957) recognized 10 species of this complex in Cuba, whereas I accept only four. He differentiated T. leptoneura as more densely scaly on the leaf undersides, T. heterophylla and T. elongata as having sessile basal leaflets, the latter distinguished by being long acuminate with thick petiolules, T. shaferi by basal leaflets with long narrow petiolules, and T. geronensis, T. arenicola, T. leptopoda, T. beyeri, and T. dictyophylla as having mostly 3-foliolate leaves separated by minor differences in leaflet size, texture and shape. The most widespread of these was supposed to be T. shaferi but much of the material identified with this name by Alain (in herb.) has very short basal petiolules. It is possible that extensive field work would show that more of these entities should be recognized, but, if so, many of the intermediate collections would have to be accounted for by hybridization. In the absence of such field study, I am unable to justify retention of these species even though many of them seem to represent reasonably discrete morphological clusters. Several of the species described subsequent to the Flora of Cuba treatment probably represent hybrids between T. heterophylla and other Cuban Tabebuia species (see excluded species).

Common Names

whitewood, yoke wood, whitewood, roble, roble de olor, roble bianco, roble, roble prieto, poirier, poirier des Antilles, poirier rouge, poirier blanc, poirier marbre, white cedar

Distribution

Extremely widespread and often very common throughout the Antilles; from sea level to 1000 m. Occurring on many different substrates, including limestone, serpentine, beaches, and palm savannas.

Exuma Bahamas South America| Little Cayman Cayman Islands South America| Grand Cayman Cayman Islands South America| Little Cayman Cayman Islands South America| Camagüey Cuba South America| Cienfuegos Cuba South America| Granma Cuba South America| Guantánamo Cuba South America| La Habana Cuba South America| Holguín Cuba South America| Isla de la Juventud Cuba South America| Matanzas Cuba South America| Piñar del Río Cuba South America| Sancti Spíritus Cuba South America| Santiago de Cuba Cuba South America| Villa Clara Cuba South America| Barahona Dominican Republic South America| Distrito Nacional Dominican Republic South America| El Seibo Dominican Republic South America| La Romana Dominican Republic South America| Haiti South America| Islas de la Bahía Honduras Central America| Clarendon Jamaica South America| Manchester Jamaica South America| Saint Andrew Jamaica South America| Saint Ann Jamaica South America| Saint Catherine Jamaica South America| Saint Elizabeth Jamaica South America| Saint Thomas Jamaica South America| Anguilla South America| Antigua and Barbuda South America| Antigua and Barbuda South America| Saint Kitts Saint Kitts and Nevis South America| Saint Kitts Saint Kitts and Nevis South America| Saint Martin South America| Puerto Rico South America| Curaçao South America| Trinidad and Tobago South America| Florida United States of America North America| Saint Croix Virgin Islands of the United States South America| Saint John Virgin Islands of the United States South America| Saint Thomas Virgin Islands of the United States South America| Virgin Gorda Virgin Islands South America| Dominica South America| Grenada South America| Guadeloupe South America| Pointe-à-Pitre Guadeloupe South America| Martinique South America| Montserrat South America| Saint Lucia South America| Saint Vincent and the Grenadines South America| Saint Vincent and the Grenadines South America| Santa Cruz Bolivia South America|