In the early 1970s, botanists surveying the inaccessible ravines of southwestern Gran Canaria began noticing dragon trees that did not look quite right. They were growing on the same island as the famous Canary Islands dragon tree, Dracaena draco, and they were assumed to be the same species — just odd individuals, perhaps stressed by their harsh cliff-face habitats. But the leaves were wrong: channeled, rigid, and glaucous blue-grey rather than flat and grey-green. The branching was wrong: the main limbs divided into three at the apex, not two. The inflorescences were wrong: slender, erect, tripinnate panicles unlike the compact, bushy clusters of Dracaena draco. And when fruit was finally collected in 1994 and seedlings were grown, even the juvenile plants were different — their leaves emerged conduplicate (folded lengthwise), not flat, and their primary roots were napiform (turnip-shaped), not cylindrical.
In 1998, Águedo Marrero Rodríguez, Rafael S. Almeida, and Manuel González-Martín published the formal description: these were not aberrant Dracaena draco at all, but a completely separate species — Dracaena tamaranae, the Gran Canaria dragon tree, one of the rarest trees on Earth. With only 76 known wild individuals clinging to some of the most inaccessible cliffs in the Canary Islands, of which only a handful have ever produced fruit, Dracaena tamaranae is a species in the genus Dracaena — a ghost of the Tertiary world, living on borrowed time, and a stark reminder that even the most intensively studied archipelagos can still harbor species unknown to science.
Quick Facts
| Scientific name | Dracaena tamaranae A.Marrero, R.S.Almeida & M.González-Martín |
| Family | Asparagaceae (subfamily Nolinoideae) |
| Origin | Gran Canaria (Canary Islands) — strict single-island endemic |
| Adult size | 6–10 m tall; sparse, irregularly branched crown |
| Hardiness | Unknown (estimated −2 to −5 °C based on habitat elevation) |
| IUCN | Critically Endangered (CR) |
| Cultivation difficulty | 5/5 — not available outside scientific conservation programs |
Taxonomy and Nomenclature
The species was described by Marrero Rodríguez, Almeida & González-Martín in 1998, published in the Botanical Journal of the Linnean Society (128: 291–314). The holotype was collected on 20 July 1997 from Barranquillo Andrés, San Bartolomé de Tirajana, Gran Canaria (LPA: 18524, with duplicates at MA, ORT, BM).
The specific epithet tamaranae honors Tamarán, the indigenous Guanche name attributed to the island of Gran Canaria — a name that carries deep pre-colonial resonance and underscores the species’ identity as a single-island endemic.
Subfamily placement. Like Dracaena draco, this species belongs to the subfamily Nolinoideae within Asparagaceae (APG IV, 2016) — not to the subfamily Agavoideae. On this site, both subfamilies are grouped under the broad horticultural term “Agavoids.”
The Phylogenetic Surprise
When Marrero et al. first described Dracaena tamaranae, they noted a striking morphological paradox: despite growing in the same archipelago as Dracaena draco, this new species looked more like the East African dragon trees (Dracaena ombet, Dracaena schizantha) than like its Canarian neighbor. The shared characters — glaucous, channeled (canaliculate) leaves and slender, tripinnate inflorescences — led the original authors to suggest that Dracaena tamaranae might represent a second, independent colonization of Macaronesia by dragon trees from an East African source, separate from the lineage that gave rise to Dracaena draco.
This “two-colonization” hypothesis was elegant and provocative: two dragon tree species on the same archipelago, arriving independently from different parts of Africa. However, molecular phylogenetics has told a different story. Plastid DNA analyses (García-Verdugo et al. 2020, using rbcL, matK, trnQ-rps16, and rpl32-trnL markers) placed Dracaena tamaranae firmly within the Macaronesian clade, as sister to all subspecies of Dracaena draco — not as a member of the East African–Arabian clade (Dracaena cinnabari, Dracaena ombet, Dracaena serrulata). The support was overwhelming: bootstrap 100, posterior probability 1.0. Dracaena tamaranae possesses a private haplotype distinct from all Dracaena draco haplotypes.
Divergence time estimates place the split between Dracaena tamaranae and Dracaena draco at approximately 2.3 million years ago (Plio-Pleistocene) — consistent with allopatric speciation driven by island isolation during Pleistocene sea-level oscillations. The morphological similarities with East African species are therefore interpreted as retained ancestral characters (plesiomorphies) rather than shared derived characters (synapomorphies) indicating recent common ancestry.
In other words: Dracaena tamaranae looks like an African dragon tree not because it came from Africa recently, but because it retained the original appearance of the common ancestor — while Dracaena draco evolved away from it. The ancestor of both Macaronesian species probably resembled Dracaena tamaranae more than Dracaena draco.
Synonyms
None. Dracaena tamaranae has no synonyms — it was misidentified as Dracaena draco in herbarium records and literature from the 1970s and 1980s, but no formal synonym was ever published.
Common Names
Spanish: drago de Gran Canaria. English: Gran Canaria dragon tree.
Morphological Description
Habit and Branching
Dracaena tamaranae is an arborescent monocotyledon reaching 6–10 m in height — somewhat smaller than Dracaena draco (which reaches 10–15 m and exceptionally exceeds 20 m). The growth pattern follows the same dracoid habitus as all arborescent Dracaena: the stem grows unbranched until the first flowering event terminates the apical meristem, then branches emerge from axillary buds. However, the branching architecture differs in a key diagnostic character: primary branching is trichotomous (three branches per node), rarely tetrachotomous, while subsequent branches are dichotomous or simple. In Dracaena draco, branching is consistently dichotomous (two branches per node). The result is a crown that is sparser, less dense, and more irregularly shaped than the massive, hemispherical umbrella canopy of Dracaena draco. The branches are shorter and stumpier.
Trunk and Bark
The bark is yellowish-grey — distinctly different from the silvery-grey bark of Dracaena draco. It is smooth, almost glossy, and barely marked by leaf scars. Like Dracaena draco, the trunk thickens through anomalous secondary growth (monocot cambium), but in wild specimens — all of which grow on exposed cliff faces — trunks remain relatively slender compared to the massive columns of old Dracaena draco.
Leaves
Leaves are 40–80 cm long (occasionally to 110 cm) and 3–4.5 cm wide, subulate (awl-shaped), distinctly canaliculate (channeled along their length) and slightly falcate (curved like a sickle). They are glaucous — a more pronounced blue-grey than the grey-green of Dracaena draco — and faintly striate on the underside. The leaf margin is entirely hyaline (translucent). At the base, each leaf has a reddish-brown pseudo-sheath that is subamplexicaul (partially clasping the stem), forming a distinctive arching collar approximately 10 cm long.
The canaliculate, rigid leaf shape is the single most obvious field diagnostic — visible at a distance and immediately separating Dracaena tamaranae from Dracaena draco, whose leaves are flat.
Seedlings
Even at the seedling stage, the two Gran Canaria Dracaena species are easily distinguished: Dracaena tamaranae seedlings have conduplicate (folded) leaves and napiform (turnip-shaped) secondary roots, while Dracaena draco seedlings have flat leaves and cylindrical roots. The roots of Dracaena tamaranae are notably highly succulent, with globose-cylindrical primary roots — an adaptation to the xeric cliff habitats where the species survives.
Inflorescence and Flowers
The inflorescence is a slender, erect, tripinnate panicle 80–100 cm long — far more elongated and open than the compact, bipinnate panicle of Dracaena draco. Branches are dispersed along the main axis, giving the inflorescence a loose, airy structure. The basal bracts resemble mature leaves; secondary bracts diminish rapidly in size from ensiform (sword-shaped) to linear. Flowers are whitish-green, with campanulate perianth segments. The flowering biology has been studied by only a handful of botanists, as so few individuals have ever flowered.
Fruits and Seeds
Fruits are globose berries with seeds approximately 6–7 mm in diameter, slightly compressed — larger than those of Dracaena draco (which averages 7.5–10 mm). The reproductive output is desperately low: of the 76 known wild individuals, only about 5 have been observed to produce fruit, and no seedlings or natural regeneration have ever been documented in the wild.
Similar Species and Frequent Confusions
| Character | Dracaena tamaranae | Dracaena draco |
|---|---|---|
| Height | 6–10 m | 10–15 m (max 20+ m) |
| Primary branching | Trichotomous (3 branches) | Dichotomous (2 branches) |
| Canopy shape | Sparse, irregular | Dense, hemispherical umbrella |
| Bark colour | Yellowish-grey, slightly glossy | Silvery-grey |
| Leaves | Canaliculate (channeled), rigid, glaucous, falcate | Flat, grey-green, straight |
| Leaf pseudo-sheath | Reddish-brown, arching, ~10 cm | Less conspicuous |
| Inflorescence | Tripinnate, slender, 80–100 cm, loose | Bipinnate, compact, shorter |
| Seedling leaves | Conduplicate (folded) | Flat |
| Seedling roots | Napiform (turnip-shaped), succulent | Cylindrical |
| Phylogenetic position | Sister to all D. draco subspecies | — |
| Divergence | ~2.3 Ma (Plio-Pleistocene) | |
| IUCN | Critically Endangered (CR) | Vulnerable (VU) |
| Wild population | ~76 individuals | Hundreds to thousands (all subspecies combined) |
Dracaena ombet Kotschy & Peyr. and Dracaena schizantha Baker
The morphological resemblance that originally prompted the “two-colonization” hypothesis. Both East African species share the canaliculate, glaucous leaves and the slender, tripinnate inflorescence architecture of Dracaena tamaranae. However, molecular data are unambiguous: Dracaena tamaranae is not closely related to these species. The shared characters are ancestral retentions (plesiomorphies), not indicators of recent common ancestry. Dracaena ombet and Dracaena schizantha belong to the East African–Arabian clade; Dracaena tamaranae belongs to the Macaronesian clade.
Distribution and Natural Habitat
Dracaena tamaranae is a strict single-island endemic, confined to the southwestern quadrant of Gran Canaria. Its range extends along the island’s oldest geological formations, from the Amurga cliffs in Barranco de Fataga in the southeast to Mesa del Junquillo in Barranco de la Aldea in the west. This is a linear distribution of roughly 25 km, but the actual area of occupancy is minuscule: all 76 known wild individuals grow on inaccessible vertical or near-vertical cliff faces between 400 and 900 m elevation.
Habitat. The species is associated with remnants of the thermo-sclerophyllous woodland (Mayteno-Juniperion canariensis alliance) that once covered the lower slopes of Gran Canaria. The cliff-face habitat is characterized by 200–500 mm annual precipitation, is generally shady and humid (north- or east-facing aspect), and supports relict plant communities including wild juniper (Juniperus turbinata subsp. canariensis), wild olive (Olea cerasiformis), and Canarian pine (Pinus canariensis) at the upper limit. The substrate is volcanic rock.
The restriction to inaccessible cliff faces is almost certainly not the species’ natural preferred habitat but rather a refuge — the last places where the trees could survive the centuries of deforestation, agriculture, grazing by introduced goats, and urban expansion that eliminated the thermo-sclerophyllous woodland from all accessible terrain on Gran Canaria. The species probably once occupied a much wider, continuous range across the southwestern slopes of the island.
Conservation — A Species on the Edge of Extinction
Dracaena tamaranae is classified as IUCN Critically Endangered (CR) and is listed in the Spanish National Catalogue of Threatened Species (Catálogo Nacional de Especies Amenazadas) as “en peligro de extinción” — a designation that legally mandates a recovery plan. It is also protected under Canarian regional legislation (Catálogo Canario de Especies Protegidas).
Population Status
The most detailed census was published by Almeida Pérez (2003). The numbers are devastating:
- 86 total individuals documented, of which 10 were already dead
- Of the 76 living plants: 1 old individual, 12 mature, 63 juvenile
- Only 12 individuals have ever flowered
- Only ~5 have been observed to produce fruit
- No seedlings and no natural regeneration have ever been observed in the wild
- At least 10 additional reproductive individuals have died in the 25 years preceding the census
The oldest known individual — on Morro de Gurbia in Barranco de Arguineguín — was photographed progressively withering between 1997 and 2000, and is presumed dead. The youngest wild individuals are estimated at over 80 years old. There is effectively zero recruitment — the wild population is aging and dying, with no replacement.
Threats
- Critically small population: With fewer than 80 living plants, of which only a handful reproduce, the species is below any viable minimum population threshold. Stochastic events (rockfalls, drought, disease) could eliminate the species at any time.
- No natural regeneration: The complete absence of wild seedlings is the most alarming symptom. Possible causes include loss of seed dispersers (as hypothesized for Dracaena draco), unsuitable microsite conditions for germination on cliff faces, seed predation, and the ongoing aridification of Gran Canaria’s southern slopes.
- Climate change: The Holocene desiccation trend, amplified by modern climate change, is progressively reducing humidity in the already xeric southwestern slopes of Gran Canaria. The species is suspected to be approaching the lower moisture limit for survival.
- Goat grazing: Feral and domestic goats can access some cliff ledges and browse seedlings — though the near-total inaccessibility of remaining individuals limits this threat to the few plants on less extreme terrain.
- Genetic contamination risk: Dracaena draco has been widely planted on Gran Canaria as an ornamental and cultural symbol. Conservation managers explicitly recommend not introducing Dracaena draco into the area likely to host or be reforested with Dracaena tamaranae, to prevent hybridization or ecological replacement.
Conservation Actions
A rescue program based at the Jardín Botánico Canario Viera y Clavijo (Las Palmas de Gran Canaria) has collected seed from the few fruiting individuals and produced ex situ seedlings. By 2011, approximately 2,900 seeds had been collected and ~350 seedlings produced, with some distributed to botanical gardens for insurance populations and others replanted into or near wild habitat. However, the results have been sobering:
- Only about 24% of seeds produce viable seedlings
- Of those seedlings, only about 12% survive transplanting into the wild
- Dracaena tamaranae seedlings are extremely drought-resistant (surviving months of bare-root desiccation) but extremely vulnerable to water on the apical bud — flooding, heavy rain, or even excessive irrigation that reaches the growing tip is rapidly fatal
- Excessive soil cover also kills seedlings by rotting the crown — a major cause of failure in reintroduction plantings on cliff slopes
The species is not available in commercial cultivation. All propagation efforts are confined to scientific conservation programs. The recommendation is unambiguous: this is a species that needs institutional protection, not private collectors.
Cultivation
| Parameter | Value |
|---|---|
| Hardiness | Unknown; estimated −2 to −5 °C based on habitat elevation (400–900 m on Gran Canaria) |
| Light | Partial shade to full sun (wild plants grow on shaded cliff faces) |
| Soil | Volcanic, extremely well-drained; mineral substrates essential |
| Watering | Very low; never wet the apical bud — water at the base only |
| Growth rate | Extremely slow (maturity requires decades; youngest wild plants estimated 80+ years old) |
| Difficulty | 5/5 |
Dracaena tamaranae is not available in the horticultural trade and should not be. The species exists in cultivation only in the Jardín Botánico Canario Viera y Clavijo and a handful of other scientific institutions holding insurance populations. The extreme vulnerability of seedlings to crown rot — where even normal rainfall landing on the apical bud can kill the plant within days — makes it a species that defies standard horticultural practice. Less water is better — but with Dracaena tamaranae, even the right amount of water delivered to the wrong spot is lethal.
If the species is ever made available for conservation-oriented cultivation (which is not the case as of 2025), the key lessons from the Viera y Clavijo program are: never allow water to pool in the leaf rosette; use purely mineral substrates (pumice, volcanic grit, coarse sand); avoid burying the root crown; and plant on slopes or raised beds that ensure total drainage away from the growing point.
What to Know Before Buying
You cannot buy this species. Dracaena tamaranae is a Critically Endangered species protected under Spanish national and Canarian regional law. No commercial trade exists. Any plant sold under this name would be either a misidentified Dracaena draco or illegally sourced — both unacceptable. The species’ survival depends entirely on institutional conservation programs, not private cultivation.
Propagation
Seed is the only viable method. Seeds should be soaked in warm water for 24 hours before sowing in a well-draining mineral substrate at 18–21 °C. Germination success is low (~24% of seeds yield viable seedlings in the Viera y Clavijo program). The critical vulnerability is the apical bud: seedlings must be protected from water on the crown at all times. Stem cuttings have been attempted but results are not well documented in the literature.
All propagation is currently conducted within the framework of the official recovery plan and is not available to private individuals.
Pests and Diseases
Crown rot: The overwhelmingly dominant cause of mortality in cultivated seedlings. Not caused by a specific pathogen but by water accumulation in the apical meristem, which triggers rapid fungal colonization and death — often within days. This extreme sensitivity distinguishes Dracaena tamaranae from Dracaena draco, which tolerates normal rainfall without issue.
Root rot: As with all dragon trees, waterlogged substrates are fatal.
Agave snout weevil (Scyphophorus acupunctatus): Not documented on Dracaena tamaranae, but the weevil’s presence in the Canary Islands and its broad host range across Agavoids make it a potential future threat, particularly if reintroduced plants are placed near infested Agave or Dracaena draco populations.
Landscape Use
This section is largely theoretical, as Dracaena tamaranae is not available for garden use. However, in the unlikely event that conservation-bred plants are eventually deployed in institutional botanical gardens or ecological restoration projects outside Gran Canaria, the species would be of extraordinary botanical and educational value:
Botanical garden display: As a living exhibit of a Critically Endangered Macaronesian relict, Dracaena tamaranae would be a powerful conservation education tool — ideally displayed alongside Dracaena draco for direct comparison of branching pattern (trichotomous vs. dichotomous), leaf form (canaliculate vs. flat), and bark color (yellow-grey vs. silver-grey).
Ecological restoration on Gran Canaria: The priority use. Reintroduction into suitable habitat on the southwestern slopes of Gran Canaria — specifically into the thermo-sclerophyllous woodland zone where remnant Juniperus and Olea populations persist — is the only path to long-term survival of the species in the wild.
Frequently Asked Questions
How can two dragon tree species exist on the same archipelago?
The original hypothesis was that they arrived independently — Dracaena draco from a western African lineage and Dracaena tamaranae from an eastern African lineage. Molecular data have overturned this: both species belong to the same Macaronesian clade and diverged from each other approximately 2.3 million years ago, probably through allopatric speciation on different islands during Pleistocene sea-level oscillations. The morphological resemblance of Dracaena tamaranae to East African species reflects retained ancestral characters, not recent shared ancestry.
Is it really that rare?
Yes. Seventy-six living wild individuals, of which only about five produce fruit, with zero natural regeneration ever documented. This makes Dracaena tamaranae one of the rarest tree species on Earth — comparable in rarity to the most critically endangered palms and conifers. The species was unknown to science until 1998, having been misidentified as Dracaena draco for decades.
Why did it take so long to discover?
Three reasons. First, all wild individuals grow on inaccessible cliffs that are difficult to survey. Second, the species was assumed to be Dracaena draco — and when you expect to see Dracaena draco on the Canary Islands, that is what you identify. Third, the critical diagnostic characters (canaliculate leaves, trichotomous branching, tripinnate inflorescence, conduplicate seedling leaves, napiform roots) require careful examination that casual field observation does not provide. It took fruit collection (1994), seedling cultivation, and years of comparative morphological study before the species’ distinctness became undeniable.
Could Dracaena tamaranae hybridize with Dracaena draco?
This is a genuine concern. The two species are sister taxa that diverged only ~2.3 million years ago. Dracaena draco has been extensively planted on Gran Canaria as an ornamental. If planted near wild Dracaena tamaranae populations and if flowering periods overlap, hybridization is theoretically possible. Conservation managers explicitly recommend not planting Dracaena draco in the area of occurrence of Dracaena tamaranae to prevent genetic contamination.
Why is it so hard to grow?
The extreme sensitivity of the apical bud to water is the core problem. In its natural cliff-face habitat, water drains instantly away from the plant. In cultivation — even in a botanical garden — normal watering or rainfall that reaches the crown can trigger fatal rot in days. This makes the species incompatible with conventional gardening practices, including normal open-air planting in Mediterranean climates where winter rainfall is inevitable. Only highly controlled, sloped, sheltered plantings with strictly basal watering have succeeded.
Reference Databases and Online Resources
- POWO — Dracaena tamaranae
- GBIF — Dracaena tamaranae
- ArbolApp Canarias — Dracaena tamaranae
- Flora de Canarias — Dracaena tamaranae
Bibliography
Marrero, Á. (2000). Dracaena tamaranae, el género Dracaena y otros afines: análisis morfológico para una aproximación filogenética. El Museo Canario, 55: 301–332.
Almeida Pérez, R.S. (2003). Censo, distribución, hábitat y estado de conservación de Dracaena tamaranae A. Marrero, R.S. Almeida & M. González-Martín. Gran Canaria, islas Canarias. Botanica Macaronésica, 24: 39–56.
Almeida Pérez, R.S. & Marrero, Á. (2011). Diáspora y rescate: acciones para la conservación del drago de Gran Canaria, Dracaena tamaranae. V Congreso de Biología de la Conservación de Plantas, Menorca.
García-Verdugo, C. et al. (2020). Iconic, threatened, but largely unknown: Biogeography of the Macaronesian dragon trees (Dracaena spp.) as inferred from plastid DNA markers. Taxon, 69(2): 217–233.
González, A.G., León, F., Hernández, J.C., Padrón, J.I., Sánchez-Pinto, L. & Barrera, J.B. (2004). Flavans of dragon’s blood from Dracaena draco and Dracaena tamaranae. Biochemical Systematics and Ecology, 32: 179–184.
Marrero Rodríguez, Á., Almeida Pérez, R.S. & González-Martín, M. (1998). A new species of the wild dragon tree, Dracaena (Dracaenaceae) from Gran Canaria and its taxonomic and biogeographic implications. Botanical Journal of the Linnean Society, 128(3): 291–314.