Tabebuia heterophylla (DC.) Britton

Roble Blanco, White-Cedar

Bignoniaceae -- Bignonia family

P. L. Weaver

Roble blanco or white-cedar (Tabebuia heterophyl1a) is a small- to medium-size, mostly deciduous tree with showy pink flowers. It grows on any soil type and will adapt to poor or degraded soils if moisture is available. Valued as a timber tree, it has been widely planted for both reforestation and ornamentation. The tough strong wood is used for many products and is favored for boat building in the Lesser Antilles.


Native Range

Roble is native to Puerto Rico and widely distributed through the West Indies from Hispaniola to Grenada and Barbados. It is also naturalized in Bermuda and planted in southern Florida (16).

In Puerto Rico, it is widespread in abandoned pastures and secondary forests and found in dry or wet natural forests, except for the highest elevations in the Luquillo Mountains and the Cordillera Central. Elsewhere in the Caribbean, roble is particularly common in dry, coastal woodlands and in secondary forests.


In Puerto Rico, roble is found principally in the Subtropical Dry, Subtropical Moist, and Subtropical Wet life zones (12,15) where the annual rainfall varies from about 850 to 2500 mm (33 to 98 in). Temperature ranges from a mean minimum in January of 16° C (61° F) to a mean maximum of 31° C (88° F) in August (5). Potential evapotranspiration over the same regions varies between 1400 and 1900 mm (55 and 75 in) annually, with the lowest measurements in the mountainous interior.

Throughout the West Indies, roble is found predominately in areas where the annual rainfall varies between about 1000 and 2500 mm (39 and 98 in) (table 1). All sites are frost free.

Table 1- Presence of roble blanco (Tabebuia heterophylla) in tropical forests of the Western Hemisphere.

Islands Forest types¹

Puerto Rico Dry Evergreen Forest
Lower Montane Rain Forest
Nevis Dry Evergreen Forest
St. Kitts Dry Evergreen Forest
Dominica Dry Scrub Woodlands
Fire grassland and standards
St. Lucia Littoral Woodland
Dry Scrub Woodlands
Secondary Woodlands
St. Vincent Dry Scrub Woodlands
Grenadines Dry Scrub Woodlands
Grenada Dry Scrub Woodlands
Antigua Secondary Woodlands
Barbuda Bush land
Anguilla Bush land
Barbados Dry Scrub Woodlands
Martinique Seasonal Forests
Dry Scrub Woodlands
Guadeloupe Dry Scrub Woodlands
Littoral Woodland
British Virgin Islands Dry Scrub Woodland
Xerophytic Rain Fore

¹Roble is found throughout the Windward and Leeward Islands as a component of the Dry Zone Flora with rainfall between 900 to 1700 mm/yr (35 to 65 in/yr). In all instances, classification is according to Beard (1,2,3).

Soils and Topography

In Puerto Rico, roble is found on sand, limestone, and heavy clay soils, acid or alkaline in reaction, and residual, alluvial, or colluvial in origin. It appears to grow best, however, on deep clays. Roble is a cosmopolitan species and is found on all soils presently identified in Puerto Rico. The most common soil order on the island is Inceptisols. Physiographically, it is most common on slopes and ridges (19) but is also found on flats adjacent to river beds (9). In general, it is tolerant of degraded sites and abandoned farm lands where it tends to form nearly pure stands.

In Puerto Rico, roble is planted on poor sites to provide cover and to improve the soil. It is recommended for planting on uniform and convex slopes and ridges, where trials have shown it to be a promising species for reforestation (20). It has also done well on humid, waterlogged sites.

Associated Forest Cover

In Puerto Rico, roble is associated with algarrobo (Hymenaea courbaril), laurel avispillo (Nectandra coriacea), guamá (Inga fagifolia), and laurel geo (Ocotea leucoxylon) in the Dry Evergreen Forest (classification according to Beard, 1,2,3). In the Lower Montane Rain Forest of the Luquillo Mountains, it is found associated with guamá, yagrumo macho (Didymopanax morototoni), palo de matos (Ormosia krugii), achiotillo (Alchornea latifolia) and various composites, all of which are constituents of the secondary vegetation (9).

In the Windward and Leeward Islands, roble is frequently found with the same species listed for the Dry Evergreen Forest of Puerto Rico. Beard (2) called this the dry zone flora, of which the Dry Evergreen Forests, Dry Scrub Woodland, and Littoral Woodland are the principal forest types.

Life History

The mature roble in Puerto Rico is easily identified by its opposite, palmately compound leaves, furrowed bark, and narrow, columnar crown. It is a small- to medium-size tree attaining a height of 18 In (60 ft) and a diameter of 60 cm (25 in). In the seedling and sapling stages, roble is an aggressive pioneer.

Reproduction and Early Growth

Flowering and Fruiting- Large white to light purple perfect flowers are borne few to several in terminal and lateral clusters, or occasionally as individuals. In Puerto Rico, flowering is chiefly in the spring, or dry season, and is accompanied by complete leaf drop (11,16). Sporadic flowering occurs at other times. Fruits are borne in May and June with fruit fall varying from July through September. Mature fruits, dark brown cigarlike pods, may be found on the tree during most of the year (16).

At 55 randomly placed collection stations comprised of 0.5 m² (5.4 ft²) screen baskets in the Subtropical Wet Forest of Puerto Rico, roble dropped 39 fruits in 39 months. Of the 38 species observed, roble ranked 37th in the number of fruits collected (11).

Seed Production and Dissemination- The fruits are pods, about 8 to 20 cm (3 to 8 in) long and 6.5 mm (0.25 in) in diameter. The pods contain many winged seeds each about 2 cm (0.79 in) long. The capsule splits along two lines and seeds are dispersed varying distances from the parent tree, ranging up to 100 m (330 ft) or more, depending upon weather conditions. Dispersal is by wind. The seeds germinate in open areas and form dense stands of seedlings.

Several seed experiments were conducted at the Institute of Tropical Forestry during the mid-1940's. About 70,000 air-dried seeds were counted per kilogram (31,750/lb), and these had a mean moisture content of 31 percent, based on the dry weight of the seeds. Seeds sown directly in seedbeds after collection in the field showed germination rates of 90 percent within 2 weeks. A 3-week delay in sowing seeds reduced viability to about 55 percent and after 5 weeks, no seeds germinated. Attempts were made to store seeds for long periods using seed moisture contents of 100, 75, 50, and 25 percent at room temperature and at 4' C (40' F). The best germination after 25 months, nearly 55 percent, was attained with the lowest moisture content and temperature combination.

Seedling Development- Germination of roble is epigeal. Experiments by staff of the Institute of Tropical Forestry established roble in two different regions by means of broadcast seeding, spot planting of seeds, and planting, on lands that had been burned, cleared in a swath 1 m (3.3 ft) wide, or planted without site treatment. Direct seeding proved unsuccessful. The nursery stock survived, although the seedlings suffered dieback and did not recover for 6 to 8 months. Site treatment did not influence survival because grass grew quickly on all areas under study and competed with the transplants. The seedlings, after recovery, grew slowly.

Transplanting of wildlings was found preferable to nursery stock because they are abundant and have better root systems (21). In some instances, however, dieback of the leader was observed. Pruned wildlings and shelterwood plantings of wildlings were then tested, but neither gave better results. Survival remained good, but growth was not improved.

The lesson learned from testing of roble wildlings was that survival is high, even on waterlogged soils and exposed ridges. Leaves are lost after transplanting and the wildlings require about 6 to 8 months to recover, if rainfall is adequate. Of the size classes tested ranging through 60 cm (24 in), the best results were attained with the largest wildlings. Subsequent growth in all instances was slow and averaged about 1.8 m (6 ft) in 2 years.

Vegetative Reproduction- Cuttings were tested on degraded heavy soils in Luquillo Forest and Carite, but only a few survived (19,20). Roble fence posts have been observed to sprout (26), but vegetative reproduction cannot be relied on for reforestation.

Sapling and Pole Stages to Maturity

Growth and Yield- Roble regenerates well in open fields and develops into a dense stand of seedlings, after which it appears to stagnate. This phenomenon may be partially attributable to the shallow, infertile soils and to exposure. The density of the seedling stands may also be a contributing factor.

Plantations established in Puerto Rico show that the dominant and codominant stems averaged about 1 in (3.3 ft) in height growth and I cm (0.4 in) in diameter growth annually over a period of 11 to 14 years (table 2). Annual basal area growth was about 1.5 m² /ha (6.5 ft² /acre). Height growth in Hawaii was less, but the measurements were for smaller trees over a shorter period of time.

Within natural forest, diameter increment varies between 0.28 to 0.39 cm (0.11 to 0.15 in) annually for all sites with the exception of limestone ridges where growth was only 0.13 cm (0.05 in) (table 2). In a study of several crop trees within the Sabana compartment of the Luquillo Forest, roble was found to grow significantly slower than the remaining species (10). Differences by crown class were evident. On more than 435 trees within the Sabana 8 compartment, annual diameter growth for dominants was 0.38 cm (0.15 in), codominants 0.32 cm (0.13 in), intermediates 0.21 cm (0.08 in), and suppressed stems 0.09 em (0.03 in). Moreover, diameter growth increased with increasing diameter class, perhaps due to a more favorable competitive position within the canopy (10).

Table 2- Growth information for roble blanco (Tabebuia heterophylla) in the Western Hemisphere

Site characteristics Stand Mean annual increment

Location Elevation Annual rainfall Soil Age¹ Density Height D.b.h. Basal area

m mm yr stems/ha m cm m²/ha
Puerto Rico                
Luquillo² (25) 300 3050 residual clay 11 NA³ 1.3 1.18 1.82
Luquillo² (25) 250 2550 residual clay 14 1000 1 0.93 1.29
Luquillo (19) 360 2700 eroded ridge 5 400 0.5 0.71 0.32
Hawaii (28) 30 to 625 2250 to 5600 stoney muck 5.3 NA 0.6 to 0.7 NA NA
Hawaii (28) 180 700 stoney clay 5.6 NA 0.3 NA NA
Natural forest                
Puerto Rico                
Sabana (10) 180 to 360 2300 deep acid clay 17 NA NA 0.28 NA
Rio Grande (10) 420 to 600 3300 deep acid clay 17 NA NA 0.35 NA
Cubuy 300 to 550 2000 clay loam 17 NA NA 0.3 NA
St. Just (27) 60 1900 shallow clay 2 2150 NA 0.38 NA
Cambalache (27) 60 1400 limestone ridge 25 4350 NA 0.13 NA
El Verde (24) 450 3000 acid clay 2 NA NA 0.38 NA
Luquillo Foothills (24) 200 2500 acid clay 11 2420 NA 0.39 NA
Luquillo Foothills (24) 200 2500 acid clay 11 2700 NA 0.28 NA
  ft  in    yr  stems/acre  ft  in  ft²/acre 
Puerto Rico                
Luquillo² 984 120 residual clay 11 NA 4.26 0.46 7.93
Luquillo² 820 100 residual clay 14 405 3.28 0.37 5.62
Luquillo 1,130 106 eroded ridge 5 162 1.64 0.28 1.4
Hawaii 98 to 623 98 to 220 stoney muck 5.3 NA 1.97 to 2.30 NA NA
Hawaii 590 28 stoney clay 5.6 NA 0.98 NA NA
Natural forest                
Puerto Rico                
Sabana 590 to 1,180 90 deep acid clay 17 NA NA 0.11 NA
Rio Grande 1,378 to 1,968 130 deep acid clay 17 NA NA 0.14 NA
Cubuy 984 to 1,804 78 clay loam 17 NA NA 0.12 NA
St. Just 197 75 shallow clay 2 870 NA 0.15 NA
Cambalache 197 55 limestone ridge 25 1,760 NA 0.05 NA
El Verde 1,476 118 acid clay 2 NA NA 0.15 NA
Luquillo Foothills 256 98 acid clay 11 1,093 NA 0.15 NA
Luquillo Foothills 256 98 acid clay 11 1,093 NA 0.11 NA

¹As used in natural forests, age refers to the duration of measurements.
²Growth increment recorded for dominant and codominant stems.
³Not available.

From a sample of 360 trees ranging in diameter from 9 to 40 cm (3.5 to 15.7 in) growing within a secondary, thinned stand, it was estimated that roble would attain the 40 cm (16 in) diameter class in about 100 years.

Rooting Habit- The use of wildlings as planting stock revealed that young roble develops a thick stem and well-developed root system at an early age (21).

Reaction to Competition- The silviculture of roble was also investigated by the staff of the Institute during the mid-1940's. Roble wildlings underplanted in an Australian beefwood (Casuarina equisetifolia) stand, a species used to provide a light shade, showed 80 percent survival after 18 months, but growth was very slow. In another experiment with nursery seedlings raised in sun vs. shade conditions, 40 percent greater height growth was ob served in the exposed conditions after 5 months. Shaded seedlings grew very little. In natural conditions, wildlings are capable of surviving shade for years with no appreciable growth (21).

Roble regenerates and forms pure stands on grasslands and degraded soils, in particular on exposed upper slopes and ridges, where competition from faster growing, larger, and more tolerant trees is lacking (19). Plantations of roble wildlings usually require weedings where grass is dense, one at 6 months and a second at about 18 months. Plantations should have close spacing, not greater than 1.8 by 1.8 m (6 by 6 ft), so that ground cover is provided rapidly (21).

Within the Lower Montane Rain Forest (1,2,3) of the Luquillo Mountains, roble was found on four of six permanent plots totaling 2.1 ha (5.2 acres), measured since the mid-1940's. Of the 30 species studied, it ranked 25th in density, 14th in basal area dominance, and 15th in volume (4). Moreover, on a scale of 1 (most tolerant) to 29 (most pioneer), roble ranked 20th in shade tolerance among tree species in the Luquillo Forest (23). The scale considered the presence of seed, seedlings, and understory trees within the forest. Overall, roble blanco is classed as intolerant of shade.

Roble's persistence in the natural forest, despite its slow growth, is largely attributable to its capacity to survive on poor sites where competition is minimized.

Damaging Agents- In the natural forest, pathogens do not appear to be of any consequence. However, branches of city and roadside trees are often deformed into a witches' broom appearance, apparently by a virus possibly transmitted by the leaf hopper Protalebra tabebuiae (8). The insect also defoliates the tree or causes the leaves to turn yellow and fall prematurely (16,22). A similar disease on a closely related species, Tabebuia pentaphylla, was observed on trees grown for cacao shade on the Paria peninsula of Venezuela (7). Because of the numerous problems with pathogens, some authorities have recommended that closely related members of the same genus be used as substitutes in ornamental plantings.

A dieback disease was observed in 3 percent of potted trees in the Cambalache nursery on the north coast of Puerto Rico and was attributed to Botryodiplodia spp. (13). Transplants from a nearby wooded area to a golf course near the town of Dorado were infested by a shoot borer, probably Pachymorphus subductellus (14).

The heartwood is rated as moderately durable in contact with the ground, but susceptible to Cryptotermes brevis, the dry wood termite (6,29) and marine borers (16). Moreover, the wood rates only fair in weathering characteristics. Unpainted wood loses its smooth surface and develops considerable checking (17).

Special Uses

The heartwood is light brown or golden and not easily separated from the sapwood. The grain is straight to interlocked, and the specific gravity is about 0.55. The wood seasons rapidly with little warping and is fairly easy to work, rating fair for planing, excellent for boring, mortising, and sanding, and good for turning. Penetration and absorption of preservatives is low, even in the sapwood (6,16,17,18). The wood is tough and strong for its weight.

Roble's appearance and technical properties resemble both oak and ash. The wood is widely used for flooring, furniture, cabinetwork, interior trim, tool handles, decorative veneers, boatbuilding, ox yokes, millwork, and sporting goods. Less valuable grades are suitable for boxes, crates, concrete forms and similar items, and occasionally for posts and poles (16,17,18).

Roble's large flowers and narrow, columnar crown have made it a favorite ornamental in yards and along roadsides throughout Puerto Rico. Flowering in many instances has been observed a few years after planting (22).

The tree comes in readily on abandoned farm soils and is particularly adapted to degraded sites. Foresters have planted it on abandoned farmlands where its growth has been slow, but satisfactory.

Roble is also classed as a honey plant.


Tabebuia heterophylla is a variable species that has been classified into subspecies, or related species, by several authors. Synonyms considered by some as varieties include T. pallida (Lindl.) Miers and T. dominguensis Urban (18).

Roble in Puerto Rico typically has five leaflets. In dry areas and coastal thickets in the Lesser Antilles, trees are shorter, fruits and seeds smaller, and leaflet number declines to three, or at times to a single leaflet. Another variation found in Guadeloupe, Dominica, and Martinique has a single, broadly elliptic leaf (16).

Literature Cited

  1. Beard, J. S. 1944. Climax vegetation in tropical America. Ecology 25(2):127-158.
  2. Beard, J. S. 1949. The natural vegetation of the Windward and Leeward Islands. Oxford Forestry Memoirs 21. Clarendon Press, Oxford. 192 p.
  3. Beard, J. S. 1955. The classification of tropical American vegetation-types. Ecology 36(l):89-100.
  4. Briscoe, C. B., and F. H. Wadsworth. 1970. Stand structure and yield in tabonuco forests of Puerto Rico. In A tropical rain forest. H. T. Odum, and R. F. Pigeon, eds. B79-89. U.S. Atomic Energy Commission, TID-24270. Washington, DC. (Available from National Technical Information Service, Springfield, VA.)
  5. Calvesbert, R. J. 1970. Climate of Puerto Rico and U.S. Virgin Islands. Rev. U.S. Department of Commerce, Environmental Science Services Administration, Washington, DC. 29 p.
  6. Chudnoff, Martin. 1984. Tropical timbers of the world. U.S. Department of Agriculture, Agriculture Handbook 607. Washington, DC. 464 p.
  7. Ciferri, R. 1949. La escoba de bruja de Algunos arboles de sombrio del cacao (Erythrina y Tabebuia) en Venezuela. Una enfermedad de origen no criptogámico. Revista de la Facultad Nacional de Agronomía 10(34):143-147. Medellin, Colombia.
  8. Cook, M. T. 1938. The witches' broom of Tabebuia pallida in Puerto Rico. Journal of Agriculture of University of Puerto Rico 22:441-442.
  9. Crow, T. R., and D. F. Grigal. 1979. A numerical analysis of arborescent communities in the rain forest of the Luquillo Mountains, Puerto Rico. Vegetatio 40(3):135-146.
  10. Crow, T. R., and P. L. Weaver. 1977. Tree growth in a moist tropical forest of Puerto Rico. USDA Forest Service, Research Paper ITF-22. Institute of Tropical Forestry, Rio Piedras, PR. 17 p.
  11. Estrada Pinto, Alejo. 1970. Phenological studies of trees at El Verde. In A tropical rain forest. H. T. Odum, and R. F. Pigeon, eds. D237-269. U.S. Atomic Energy Commission, TID-24270. Washington, DC.
  12. Ewel, J. J., and J. L. Whitmore. 1973. The ecological life zones of Puerto Rico and the U.S. Virgin Islands. USDA Forest Service, Research Paper ITF-18. Institute of Tropical Forestry, Rio Piedras, PR. 72 p.
  13. Flake, H. W. 1980. Tropical Report-Puerto Rico, January 13-19, 1980. USDA Forest Service, Forest Insect and Disease Management, Asheville, NC. 21 p.
  14. Flavell, T. H., and W. R. Phelps. 1973. Evaluation of tree insect and disease pests in Puerto Rico and the Virgin Islands. USDA Forest Service, Forest Pest Management Group Report 74. Southeastern Area State and Private Forestry, Atlanta, GA. 19 p.
  15. Holdridge, L. R. 1967. Life zone ecology. Rev. Tropical Science Center, San José, Costa Rica. 206 p.
  16. Little, Elbert L., Jr., and Frank H. Wadsworth. 1964. Common trees of Puerto Rico and the Virgin Islands. U.S. Department of Agriculture, Agriculture Handbook 249. Washington, DC. 548 p.
  17. Longwood, Franklin R. 1961. Puerto Rican woods-their machining, seasoning, and related characteristics. U.S. Department of Agriculture, Agriculture Handbook 205. Washington, DC. 98 p.
  18. Longwood, Franklin R. 1962. Present and potential timbers of the Caribbean. U.S. Department of Agriculture, Agriculture Handbook 207. Washington, DC. 167 p.
  19. Marrero, José. 1947. A survey of the forest plantations in the Caribbean National Forest. Thesis (M.S.), University of Michigan, Ann Arbor. 167 p.
  20. Marrero, José. 1950. Results of forest planting in the insular forests of Puerto Rico. Caribbean Forester 11:107-147.
  21. Marrero, José. 1950. Reforestation of degraded lands in Puerto Rico. Caribbean Forester 11:3-15.
  22. Schubert, Thomas H. 1979. Trees for urban use in Puerto Rico and the Virgin Islands. USDA Forest Service, General Technical Report SO-27. Southern Forest Experiment Station, New Orleans, IA. (Institute of Tropical Forestry, Rio Piedras, PR.) 91 p.
  23. Smith, Robert Ford. 1970. The vegetation structure of a Puerto Rican rain forest before and after short-term irradiation. In A tropical rain forest. H. T. Odum, and R. F. Pigeon, eds. D103-140. U.S. Atomic Energy Commission, TID-24270. Washington, DC.
  24. Tropical Forest Experiment Station. 1950. Tenth annual report. Caribbean Forester 11(2):59-80.
  25. Wadsworth, Frank H. 1960. Datos de crecimiento de plantaciones forestales en México Indias Occidentales y Centro y Sur América. Comité Regional sobre Investigación Forestal, Comisión Forestal y la Alimentación. Departamento de Agricultura de los E. U., Rio Piedras, PR. 272 p.
  26. Wadsworth, Frank H. 1981. Personal communication. Institute of Tropical Forestry, Rio Piedras, PR.
  27. Weaver, Peter L. 1979. Tree growth in several tropical forests of Puerto Rico. USDA Forest Service, Research Paper SO-152. Southern Forest Experiment Station, New Orleans, LA. (Institute of Tropical Forestry, Rio Piedras, PR.) 15 p.
  28. Whitesell, Craig D., and G. A. Walters. 1976. Species adaptability trials for man-made forests in Hawaii. USDA Forest Service, Research Paper PSW-118. Pacific Southwest Forest and Range Experiment Station, Berkeley, CA. 30 p.
  29. Wolcott, G. N. 1957. Inherent natural resistance of woods to the attack of the West Indian dry-wood termite, Cryptotermes brevis Walker. Journal of Agriculture of the University of Puerto Rico 41:259-311.