Dracaena cubensis

The genus Dracaena is overwhelmingly an Old World group. Of its roughly 214 accepted species, the vast majority are native to Africa, Madagascar, the Arabian Peninsula, and Southeast Asia. Only two species break this pattern and occur naturally in the Americas — and both are botanical enigmas whose presence in the New World has fascinated biogeographers since their discovery. The first is Dracaena americana, a shrubby tree widely distributed from southern Mexico to Colombia. The second is Dracaena cubensis, a far more restricted and mysterious plant confined to the serpentine scrublands of northeastern Cuba.

Dracaena cubensis is not a dragon tree in the classic sense. It does not produce the legendary crimson resin, nor does it grow into a thick-trunked, umbrella-crowned tree like Dracaena draco. It is a slender, cane-stemmed shrub of tropical rainforests, growing on toxic soils that most plants cannot tolerate, in one of the wettest corners of the Caribbean. Yet its phylogenetic position is extraordinary: together with Dracaena americana, it forms the most basal clade of the entire genus — the sister group to every other Dracaena on Earth, including all the dragon trees, all the snake plants, and all the corn plants. This small, unassuming Cuban shrub sits at the root of a lineage of over 200 species spanning three continents.

For growers, Dracaena cubensis remains virtually unknown. It has never succeeded in cultivation outside of Cuba except under highly specialised conditions, and no specimens are known in European or North American private collections. This article documents what is known about its taxonomy, ecology, cultivation potential, and conservation — drawing primarily on the definitive study by Zona et al. (2014) and the phylogenetic work of Lu & Morden (2014).

Taxonomy

Dracaena cubensis was described by Brother Marie-Victorin (1885–1944), a member of the Roman Catholic congregation of the La Salle Brothers, founder of the Botanical Garden of Montreal, and one of the most important figures in the botanical history of Canada. Marie-Victorin made extensive botanical expeditions to Cuba in the 1930s and 1940s, and it was during one of these field trips — in the remote serpentine scrublands near Moa, in eastern Cuba — that he encountered this distinctive plant.

The species was formally described in 1942:

Basionym: Dracaena cubensis Vict., Contr. Inst. Bot. Univ. Montréal 43: 11 (1942).

Type: Cuba, Oriente [now Holguín province], Cananova, Cerro (Loma) de Miraflores, on serpentine scrub, 29 March 1942, Bros. Léon, Marie-Victorin & Clement 20844 [erroneously published as 20944]. Holotype: HAC. Isotype: US.

Marie-Victorin did not believe that the Cuban species could be accommodated in any of Baker’s (1875) existing sections of Dracaena and proposed a new section, Cubenses, to receive it.

Interestingly, the first collections of this species were not made by Marie-Victorin himself. John A. Shafer (1863–1918), head horticulturist at the New York Botanical Garden, collected specimens in eastern Cuba in 1910 and 1911. He labelled them “Dracaena reflexa Lam.?” — questionably assigning them to a Mascarene Islands species well known in horticulture. A subsequent collection by Roig in 1917 was similarly misidentified. It was these earlier specimens that motivated Marie-Victorin to seek out the plant in the field and give it a formal name.

In his original description, Marie-Victorin (1942) referred extensively to the Canary Islands dragon tree (Dracaena draco) and included a photograph he had taken at Icod de los Vinos, Tenerife, during a visit in 1929. He recognised the biogeographic significance of Dracaena in the Americas and wrote: “a former African–American continuity, which has left traces in the current flora of tropical America. The Cuban Dracaena is now, among examples from the plants, one of the most obvious remnants of this continuity.” In doing so, Marie-Victorin was an early supporter of Wegener’s theory of continental drift — still controversial in the 1940s.

Common name: the local name griñapo was reported by Roig (1988), but this name appears to have fallen out of use. Land managers and conservationists working in eastern Cuba today refer to the species simply as “dracaena.” The name “Moa Dragon Tree” is used by the seed trade (rarepalmseeds.com).

Family: Asparagaceae, subfamily Nolinoideae (APG IV)

Description

Dracaena cubensis is a shrub with multiple, usually unbranched stems up to 4 m tall and 3.5 cm in diameter. Only the distal quarter of each stem bears leaves; the lower portion is bare, marked by oblique leaf scars spaced approximately 7–9 mm apart. The overall silhouette is that of a cluster of slender canes, each topped by an elongated, fountain-like crown of recurving foliage — quite unlike the compact rosettes of Dracaena draco or the broad, spreading crowns of Dracaena americana.

Leaves: linear, 20–30 cm long and 1.0–1.5 cm wide at the base, leathery in texture, strongly recurved. This combination of narrow width and pronounced recurvature gives each stem a distinctive weeping appearance reminiscent of a grass tree or a slender Pandanus rather than a typical Dracaena.

Inflorescence: paniculate, terminal, and highly branched, 30–45 cm long — proportionally large relative to the plant’s overall size.

Flowers: borne on short pedicels in clusters of 2 or 3. Tepals white, approximately 4 mm long — significantly smaller than those of Dracaena americana (7 mm) or Dracaena draco (7–8 mm).

Fruit: berries 7–10 mm in diameter, sometimes lobed, containing 1–3 subglobose seeds approximately 4 mm in diameter.

Dragon’s blood resin: Dracaena cubensis does not produce the red resin known as dragon’s blood. Neither does Dracaena americana. The production of this resin is restricted to the Old World arborescent dragon trees (Dracaena draco, Dracaena cinnabari) and is not a feature of the New World species.

The original diagnostic plate of Dracaena cubensis, drawn by C. Lanouette and published by Marie-Victorin (1942), illustrates the stamen (A), pistil (B), transverse section of the ovary (C), complete flower (D), half-perianth showing the indurated base of the segments (E), floral diagram (F), fruit (G), and seed (H).

Habitat and Distribution

Dracaena cubensis has a restricted distribution in northeastern Cuba, confined to the provinces of Holguín and Guantánamo. It occurs from Moa and its vicinity westward to Baracoa, La Farola, the Río Jauco, and surrounding areas. The species is found from sea level to approximately 600 m, occasionally reaching 1,000 m elevation.

The botanist Brother Alain (1946) called the region of Moa a “paradise for botanists” and highlighted Dracaena cubensis as one of the most important elements making the region unique within Cuba.

Climate

The Moa region has one of the rainiest climates in Cuba: 1,400–1,600 mm of annual rainfall on the coast and more than 3,400 mm in the mountains. Average annual temperature ranges between 26 °C at sea level and 22 °C on the mountains. There is a pronounced wet season from May to October and a drier period from November to April, but rainfall is substantial even during the drier months owing to the northeastern exposure and orographic effects. This is a fully tropical, frost-free environment — temperatures never approach 0 °C at any time of year.

Soils

The species is strictly associated with serpentine (ultramafic) substrates — one of the most hostile soil environments for plant growth. These soils are derived from old serpentinised ultramafic rocks and are characterised by high concentrations of heavy metals (nickel, chromium, cobalt), very low calcium-to-magnesium ratios, and nutrient deficiency. Dracaena cubensis clearly prefers deep, red, ferritic soils (laterites) and areas where the serpentines are associated with brown-reddish fersialitic soils. At higher elevations (above 600 m), it occurs on thin, skeletal soils (“suelo esquelético”). Despite growing on nickel-rich soils, Dracaena cubensis is not a nickel hyperaccumulator (Reeves et al. 1999).

This strict edaphic specialisation is a defining ecological feature of the species and a major factor in its very limited distribution. Serpentine soils in northeastern Cuba support a unique and highly endemic flora, but they also constrain the species that depend on them to a small geographic footprint.

Vegetation types

Zona et al. (2014) documented four main vegetation types where Dracaena cubensis occurs:

Submontane pine forest (Pinus cubensis) on xerothermic serpentines at 0–400 m elevation. Principal accompanying species: Pinus cubensis, Jacaranda arborea, Coccothrinax orientalis, Metopium venosum, Jacquinia roigii, Coccoloba shaferi, Callicarpa oblanceolata, Eugenia pinetorum, Casearia moaense, Psidium parvifolium, Schmidtottia shaferi, and Sideroxylum cubensis.

Xeromorphic and sub-thorny scrub over serpentine (“charrascal”). Characteristic co-occurring species: Ariadne shaferi, Neobracea valenzuelana, Adenoa cubensis, Tabebuia simplicifolia, Casasia jaquinoides, Spirotecoma apiculata, Coccoloba nipensis, Guettarda crassipes, Casearia ophiticola, Phyllanthus comosus, Croton spp., Anastraphia spp., and Casearia spp.

Montane scrub over serpentine (“charrascal alto”). Principal woody species: Pinus cubensis, Cyrilla cubensis, Byrsonima biflora, Lyonia macrophylla, Lyonia glandulosa, Purdiaea spp., Rheedia polyneura, Coccoloba nervosa, Clusia spp., Buxus spp., and Ilex spp.

River margin or gallery vegetation linked to serpentines. Common associates: Tabebuia gracilipes, Cyrilla cubensis, Calyptronoma plumeriana, Neobracea ekmanii, Bactris cubensis, Calliandra enervis, Buchenavia capitata, Cubanola daphnoides, Buxus spp., Achrosynanthus revolutus, Savia clementis, Cyathea spp., and Alsophila spp.

These four plant communities frequently intermix or form ecotones, and the boundaries between them can be difficult to discern.

An important ecological observation: Dracaena cubensis is also found in disturbed areas along the edges of its habitat — along roads and paths, on landslides, and near mine sites. This resilience in the face of disturbance is unusual for a narrow endemic with strict edaphic requirements, and it gives the species potential for restoration ecology programmes.

Conservation

Dracaena cubensis is listed as Vulnerable on the IUCN Red List (Peña et al. 1998; Berazaín et al. 2005). Many populations have been affected by human activities linked primarily to open-pit mining (the Moa region is one of Cuba’s major nickel-mining areas), forest exploitation, road construction, and human settlements. These activities have caused habitat fragmentation, landslides, environmental pollution, and fires. In some areas, alien invasive species also threaten the native vegetation.

An important portion of the population of Dracaena cubensis is protected within Cuba’s national system of protected areas (Sistema Nacional de Áreas Protegidas, SNAP). Protected sites of national and international significance include:

Parque Nacional Alejandro de Humboldt — a UNESCO World Heritage Site since 2001, this is one of the most biologically diverse terrestrial sites in the Caribbean, covering approximately 71,000 hectares of mountains, forest, and coast in the Nipe-Sagua-Baracoa mountain range.

Reserva de la Biosfera Cuchillas del Toa — one of Cuba’s largest biosphere reserves, encompassing the Cuchillas del Toa mountain system.

Reserva Florística Manejada Cerro de Miraflores y Yamanigüey — the type locality for Dracaena cubensis is Cerro de Miraflores, making this reserve of special significance for the species.

Additional locally important protected areas include Reserva Ecológica Quibiján-Duaba, Reserva Florística Manejada de Alto de Cotilla, and Reserva Natural Cañón del Río Yumuri.

Phylogenetic Position and Biogeography

The phylogenetic position of Dracaena cubensis is one of the most remarkable features of this species and places it at the centre of a major biogeographic puzzle.

Lu & Morden (2014), in a comprehensive molecular phylogenetic analysis of the “Dracaenoids” (the three genera Dracaena, Sansevieria, and Pleomele) using four chloroplast DNA regions, found the following:

The Hawaiian species formerly placed in Pleomele are phylogenetically distinct and sister to all remaining Dracaenoids. They were separated into a new genus, Chrysodraco.

Dracaena americana and Dracaena cubensis together form a well-supported clade that is sister to all remaining Dracaena sensu lato — including Sansevieria, Pleomele, and all the Old World species. This means the two New World species diverged before any of the African, Arabian, or Asian species diversified.

Dracaena draco is not closely related to the New World taxa. Instead, Dracaena draco is paired with Dracaena aubryana of tropical Africa. Other dragon tree species (Dracaena cinnabari, Dracaena ombet) appear elsewhere in the phylogeny, suggesting that arborescence (the “dragon tree” growth form) arose several times independently.

The biogeographic scenario proposed by Lu & Morden (2014) is startling: the ancestors of Dracaena arose in southern or southeastern Asia. A first dispersal event brought an ancestor to Hawaii (giving rise to Chrysodraco). A second dispersal occurred from Hawaii to Mexico or Central America, where it gave rise to Dracaena americana and Dracaena cubensis. A third ancestral dispersal — from the Americas to Africa and the Arabian Peninsula — gave rise to the tremendous species diversity found in those regions today. In other words, the common ancestor of all African, Arabian, and Macaronesian Dracaena may have been an American plant.

Marie-Victorin had intuited the essence of this story in 1942, though he attributed the pattern to continental drift rather than long-distance dispersal. Modern phylogenetics, based on molecular data rather than geographic speculation, has confirmed the deep biogeographic connection between the New World and Old World Dracaena — but reversed the direction: it was not Africa that sent Dracaena to Cuba, but possibly the Americas that sent Dracaena to Africa.

These results are based on linked chloroplast DNA markers and should be confirmed with independent nuclear DNA data. Dracaena tamaranae was not included in the study. But the basal position of the Dracaena americana–Dracaena cubensis clade is robust and has been confirmed by subsequent analyses.

Ethnobotany

Unlike Dracaena draco and Dracaena cinnabari, which have been exploited for their red resin since antiquity, Dracaena cubensis has no reported ethnobotanical uses. It does not produce dragon’s blood resin, and there are no literature reports of it being used by humans for food, medicine, fibre, or any other purpose (Zona et al. 2014).

Cultivation

The fundamental challenge

Dracaena cubensis is almost certainly the most difficult Dracaena species to cultivate outside its native range. The species has two overriding requirements that are extremely difficult to replicate in a garden or greenhouse setting: serpentine soil chemistry and a hot, wet tropical climate with no dry season stress.

Serpentine soil: the species shows a strong and possibly obligate preference for soils derived from ultramafic rocks. The one successful cultivation outside the wild — at the Jardín Botánico Nacional in Havana — was achieved by transporting original serpentine soil from eastern Cuba to the botanic garden site. An attempt to cultivate the species at Fairchild Tropical Botanic Garden in Miami, Florida, from germinated seeds and juvenile plants, did not meet with long-term success — despite Miami’s warm subtropical climate being far closer to the species’ natural conditions than any European location.

Climate: Dracaena cubensis grows in a fully tropical environment with high rainfall (1,400–3,400 mm per year), high humidity, and temperatures that never fall below approximately 18–20 °C. This is a species of wet tropical forests and serpentine scrub, not of Mediterranean or subtropical gardens.

Propagation

Seeds of Dracaena cubensis are difficult to germinate. García (2004) found that the best germination response is obtained at 25 °C with light. Higher temperatures lead to a significant reduction in seed germination — an unusual pattern that may reflect the species’ adaptation to the relatively cool, cloud-covered serpentine highlands rather than the hot lowland tropics.

Seeds are not commercially available from any known source as of 2026. The species is not listed by any European or North American nursery or seed supplier.

Potential for restoration ecology

Despite its poor performance in garden cultivation, Dracaena cubensis shows surprising resilience in its native habitat. It colonises disturbed areas along roads, paths, landslides, and mine edges — habitats that are structurally and chemically similar to its natural serpentine scrub. This ecological flexibility makes it a candidate for restoration programmes in the nickel-mining areas of eastern Cuba, where large tracts of serpentine landscape have been degraded by open-pit operations. Zona et al. (2014) specifically noted this potential: “this species has potential for restoration ecology programs targeting mines, roads, deforested areas and human settlements.”

Cold hardiness

No data exist on the cold hardiness of Dracaena cubensis. Given its fully tropical, frost-free origin (minimum temperatures in its native range never approach 0 °C), the species should be assumed to be frost-intolerant. A reasonable estimate, by analogy with other tropical Dracaena species, would be a minimum of approximately +5 to +10 °C for safe cultivation — USDA zone 11 at the most optimistic.

This species is not a candidate for outdoor cultivation in any Mediterranean, subtropical, or temperate climate. In Europe or the United States, it would require a heated tropical greenhouse with high humidity and specialised serpentine-derived growing media.

Comparison with Dracaena americana

The two New World Dracaena species are sister taxa but differ in nearly every respect:

Habit: Dracaena americana is a shrub or small tree up to 10–12 m tall with trunks up to 30 cm in diameter and exfoliating bark. Dracaena cubensis is a much smaller shrub, usually under 4 m, with slender stems only 3.5 cm in diameter.

Leaves: Dracaena americana has broader (1.0–2.5 cm), softer, bright green leaves borne along the length of the stems. Dracaena cubensis has narrower (1.0–1.5 cm), leathery, strongly recurved leaves concentrated at the stem tips only.

Flowers: Dracaena americana has larger tepals (7 mm) in clusters of 2–5. Dracaena cubensis has smaller tepals (4 mm) in clusters of 2–3.

Fruit: Dracaena americana has larger berries (up to 20 mm) with larger seeds (10–12 mm). Dracaena cubensis has smaller berries (7–10 mm) with smaller seeds (4 mm).

Distribution: Dracaena americana has an extensive range from Mexico to Colombia. Dracaena cubensis is restricted to a narrow strip of northeastern Cuba.

Substrate: Dracaena americana grows on limestone karst with humus-rich rendzina soils. Dracaena cubensis grows on serpentine ultramafic substrates with ferritic laterites.

Cultivation: Dracaena americana has been successfully grown at the CICY Botanical Garden in Mérida (Mexico) and at Fairchild Tropical Botanic Garden in Miami, where a large specimen has thrived in open ground since 1953. Dracaena cubensis has succeeded only in Havana, on transported serpentine soil.

Sites and Pages of Interest

Bibliography

Marie-Victorin, Bro. (1942). Dracæna cubensis, une relique d’affinité Africaine dans la flore de Cuba. Contributions de l’Institut Botanique de l’Université de Montréal 43: 1–16.

Zona, S., Álvarez de Zayas, A., Orellana, R., Oviedo, R., Jestrow, B. & Francisco-Ortega, J. (2014). Dracaena L. (Asparagaceae) en el Nuevo Mundo: su historia y botánica. VIERAEA 42: 219–240.

Lu, P.-L. & Morden, C.W. (2014). Phylogenetic relationships among dracaenoid genera (Asparagaceae: Nolinoideae) inferred from chloroplast DNA loci. Systematic Botany 39: 90–104.

García, G.E. (2004). Características germinativas de Dracaena cubensis, un endémismo del oriente de Cuba. Trabajo de Diploma, Universidad Agraria de La Habana “Fructuoso Rodríguez Pérez.”

Berazaín, I.R., Areces Berazaín, F., Lazcano Lara, J.C. & González Torres, L.R. (2005). Lista roja de la flora vascular cubana. Gijón: Jardín Botánico Atlántico de Gijón.

Peña, E., López, P.I., Lazcano, J. & Leiva, A.T. (1998). Memorias del primer taller para la conservación, análisis y manejo planificado de plantas silvestres cubanas. CAMP 1. La Habana: Jardín Botánico Nacional de Cuba / CBSG (SSC/IUCN).

Reeves, R.D., Baker, A.J.M., Borhidi, A. & Berazaín, R. (1999). Nickel hyperaccumulation in the serpentine flora of Cuba. Annals of Botany 83: 29–38.

Roig, J.T. (1988). Diccionario botánico de nombres vulgares cubanos. Tercera reimpresión. Tomo I. La Habana: Editorial Científico-Técnica.

Baker, J.G. (1875). Revision of the genera and species of Asparagaceae. Journal of the Linnean Society, Botany 14: 508–632.

Alain, Bro. (1946). Moa, paraíso de los botánicos. Revista de la Sociedad Cubana de Botánica 3: 9–14.

Donnell Smith, J. (1905). Dracaena americana. P. 207. In: C.S. Sargent (ed.), Trees and Shrubs, Vol. 1. Boston & New York: Houghton, Mifflin and Company.

Hasdenteufel, P. (2009). Contribución de reservas de biosfera al desarrollo regional sostenible. Estudio de caso de la Reserva de Biosfera “Cuchillas del Toa” y su zona núcleo, el Parque Nacional “Alejandro de Humboldt” (Cuba). Nadir: Revista Electrónica de Geografía Austral 1(2): 4–23.

Martínez Quesada, E., Fagilde Espinosa, M.C., Oviedo Prieto, R., Foster, R.B., Alverson, W.S. & Vriesendorp, C. (2005). Appendix 4. Seed plants. Pp. 252–317 in: Fong G., A., Maceiras F., D., Alverson, W.S. & Wachter, T. (eds.), Cuba: Parque Nacional “Alejandro de Humboldt.” Chicago: The Field Museum (Rapid Biological Inventories 14).