In the 1860s, the German-Austrian explorer Theodor von Heuglin ascended the Red Sea Hills of what is now eastern Sudan and encountered a sight that moved him to ecstasy. “I had to shout for joy,” he wrote, “when, upon approaching, all the thousands of dots with which the wide valley floor appeared speckled from afar resolved themselves into colossal Ombet and euphorbias, arm in arm, bursting with sap and gazing down with contempt upon the surrounding sand-drought and the wretched vegetation of the lower flora.” He estimated the trees in their millions. Their typical dimensions were 12–15 feet (3.6–4.5 m) in height with an equal crown width.
Today, at the same latitude — in the Gebel Elba National Park of southeastern Egypt — surveys have recorded 353 trees, of which only 161 are alive, only 96 are healthy, and only 2 are young. The populations at nearby Erkwit in Sudan have completely vanished — dead to the last tree. Dracaena ombet, the Nubian dragon tree, is a species in the genus Dracaena — the continental African member of the dragon tree group, a once-abundant flagship of the Afromontane drylands from Egypt to Ethiopia, and one of the Agavoids most rapidly approaching extinction.
Quick Facts
| Scientific name | Dracaena ombet Heuglin ex Kotschy & Peyr. |
| Family | Asparagaceae (subfamily Nolinoideae) |
| Origin | Northeast Africa (Egypt, Sudan, Eritrea, Ethiopia, Djibouti, Somalia) and western Arabian Peninsula (Saudi Arabia, Yemen) |
| Adult size | 2–8 m tall; forked trunk with densely packed crown |
| Hardiness | Unknown (estimated frost-sensitive; minimum ~0 to −3 °C based on montane habitat) |
| IUCN | Endangered (EN) — some authors propose Critically Endangered |
| Cultivation difficulty | 5/5 — virtually unavailable; conservation propagation only |
Taxonomy and Nomenclature
The species was described by Kotschy & Peyritsch in 1867 (Plantae Tinneanae: 47), based on material and observations by Theodor von Heuglin during his expeditions to the Red Sea coast of Sudan and Eritrea. The epithet ombet is the local Beja (Bidhaawyeet) name for the tree — a direct borrowing from the indigenous language of the nomadic pastoralists of the Red Sea Hills, for whom the tree has been a culturally significant species for centuries.
Subfamily placement. Asparagaceae, subfamily Nolinoideae (APG IV, 2016). As with all dragon trees, treated on this site under the broad horticultural term “Agavoids.”
Two Subspecies
POWO recognizes two subspecies:
| Subspecies | Range | Described by | Key distinction |
|---|---|---|---|
| Dracaena ombet subsp. ombet | Egypt (Gebel Elba), Sudan (Erkwit), Eritrea, northern Ethiopia (Desa’a), Djibouti | Kotschy & Peyr. (1867) | Glabrous or pubescent inflorescence; northern dry Afromontane forests |
| Dracaena ombet subsp. schizantha | Eastern & southern Ethiopia (east of the Rift Valley), Somalia (Somaliland) | (Baker) Bos | More open habitat; often associated with Buxus hildebrandtii; tomentose inflorescence in some forms |
The subspecies schizantha was originally described as a separate species (Dracaena schizantha Baker, 1877) and is still treated as such by some authors. POWO sinks it into Dracaena ombet as a subspecies. Dracaena serrulata Baker, found on the Arabian Peninsula (Yemen, Oman, Saudi Arabia), is sometimes considered conspecific with Dracaena ombet but is currently accepted as a separate species by POWO.
Phylogenetic Position
Dracaena ombet belongs to the East African–Arabian clade of the dragon tree group, alongside Dracaena cinnabari (Socotra), Dracaena serrulata (Arabian Peninsula), and Dracaena ellenbeckiana (Ethiopia, Kenya, Uganda). This clade is sister to the Macaronesian clade (Dracaena draco + Dracaena tamaranae), with a distribution disjunction across the Sahara that exemplifies the Rand Flora pattern — a biogeographic signature of former Tethyan subtropical forests fragmented by Neogene and Quaternary aridification.
Morphologically, Dracaena ombet shares with Dracaena tamaranae (Gran Canaria) the characters of glaucous, canaliculate leaves and tripinnate inflorescences — the ancestral dragon tree morphotype. These shared characters were initially interpreted as evidence of close relationship, but molecular data show they are plesiomorphies (retained ancestral traits), not synapomorphies.
Synonyms
- Draco ombet (Kotschy & Peyr.) Kuntze (1891)
- Dracaena schizantha Baker (1877) — now subsp. schizantha
- Dracaena rhabdophylla Chiov. (1951) — synonym of subsp. schizantha
Common Names
English: Nubian dragon tree, Gabal Elba dragon tree. Arabic: أومبت. Beja (Bidhaawyeet): tombat, ombet. Tigrinya (Ethiopia/Eritrea): ዕረ (era). French: dragonnier d’Ombet, dragonnier nubien.
Morphological Description
Habit
Dracaena ombet is an arborescent, evergreen monocotyledon reaching 2–8 m in height — significantly smaller than Dracaena draco (10–15 m) and Dracaena cinnabari (up to 10–12 m). The trunk is forked and divided into several branches, producing a densely packed, umbrella-shaped crown following the characteristic dracoid habitus. The overall impression is of a smaller, more compact, more rugged tree than the Canarian or Socotran dragon trees — a species adapted to harsher, drier, more exposed montane habitats. Like other arborescent Dracaena, branching follows each flowering event when the terminal bud dies, and the sympodial branching pattern produces successive “orders” that can be used for indirect age estimation.
Trunk
The trunk produces a red resin when cut — dragon’s blood, chemically similar to but distinct from the resins of Dracaena draco and Dracaena cinnabari. The trunk thickens through anomalous secondary growth (monocot cambium). No annual growth rings are produced, making age estimation difficult. The first age-estimation model for Dracaena ombet was published by Lengálová et al. (2020) — a crown age model based on the number of branch orders, analogous to models developed for Dracaena cinnabari.
Leaves
Leaves are linear with a broad base, 40–60 cm long and up to 3 cm wide, gradually tapering to a sharp tip. They are thick, rigid, with smooth margins, flat to concave on the upper surface — reducing the transpiration surface in the species’ intensely xeric habitat. Leaves form dense terminal rosettes at the branch tips. Their erect posture further reduces water loss by minimizing solar exposure.
Inflorescence, Flowers, and Fruits
The inflorescence is a highly branched panicle approximately 0.5 m long, glabrous or pubescent (variable between populations), with tiny ovate-lanceolate bracts. Tepals are whitish, 4–6 mm long, linear. Stamens are somewhat shorter than the tepals, with flattened filaments. Fruits are spherical berries 10–12 mm in diameter, each containing 1–3 seeds. Seeds have long dormancy and are dispersed by birds, water, and wind. The long seed dormancy, combined with very low germination rates in the wild, contributes to the catastrophic recruitment failure observed across the species’ range.
Similar Species and Frequent Confusions
| Character | Dracaena ombet | Dracaena cinnabari | Dracaena serrulata | Dracaena draco |
|---|---|---|---|---|
| Range | NE Africa + W Arabia | Socotra only | S Arabian Peninsula | Macaronesia + Morocco |
| Height | 2–8 m | Up to 10–12 m | 2–6 m | 10–15 m (max 20+) |
| Canopy | Compact, irregular umbrella | Dense inverted mushroom | Sparse, irregular | Large hemispherical umbrella |
| Leaf length | 40–60 cm | 30–60 cm | 30–50 cm | 60–120 cm |
| Inflorescence | Glabrous to pubescent | Glabrous | Tomentose | Glabrous |
| Clade | E African–Arabian | E African–Arabian | E African–Arabian | Macaronesian |
| IUCN | Endangered | Vulnerable | Not assessed / LC | Vulnerable |
Dracaena serrulata is the closest geographic relative (Arabian Peninsula). It is distinguished by its tomentose (hairy) inflorescence, which separates it from the typically glabrous inflorescence of subsp. ombet. Some authorities treat serrulata as conspecific with ombet; POWO keeps them separate.
Distribution and Natural Habitat
Dracaena ombet has the widest geographic range of any dragon tree species, spanning from southeastern Egypt to Ethiopia, Eritrea, Djibouti, Somalia, Sudan, and (for subsp. schizantha) extending into the Horn of Africa and possibly western Saudi Arabia. However, this vast nominal range is profoundly misleading: the species exists as tiny, scattered, isolated fragments on mountain slopes facing the Red Sea, separated by hundreds of kilometers of uninhabitable desert.
Key populations:
- Gebel Elba, Egypt (22°N): The best-studied population. Located in the Gebel Elba National Park, southeastern Egypt, near the Sudanese border. Elevation 450–950 m. A total of 353 trees recorded in 8 wadis on Elba Mountain (surveys 2007–2009). Of these, only 161 (46%) were alive, only 96 (27%) healthy, and only 2 (1%) were young. The population has lost 45.9% of its occupied area since 1997, now occupying only 0.719 km².
- Erkwit (Arkawiit), Sudan (~19°N): The historical type locality area, where Heuglin described millions of trees in the 1860s. Populations have completely vanished — dead to the last tree — due to prolonged drought and dieback. However, recent reports (Andersen et al. 2022) document the surprising discovery of saplings in a historically core area, suggesting potential recovery after an abrupt ecological change — a rare positive signal.
- Desa’a dry Afromontane forest, northern Ethiopia (~14°N): Subsp. ombet occurs in the northern dry Afromontane forests of Tigray. Populations are characterized by low abundance, unstable age structure dominated by old trees, and severe impacts from overgrazing, overexploitation (rope-making, fodder, utensils, beehives), and encroachment by pioneer shrubs. Climate change models predict a 48–83% decline in suitable habitat.
- Eritrean escarpment: Scattered populations on Red Sea–facing mountain slopes.
- Djibouti: Goda and Hemed Mountains — status poorly documented.
- Somalia/Somaliland: Hargesia region — subsp. schizantha; status poorly known.
- Bale Mountains, Ethiopia: Subsp. schizantha reported; needs updated survey.
Habitat. Bushland and woodland on mountain slopes facing the Red Sea, generally at 1,000–1,800 m elevation (lower at Gebel Elba: 450–950 m), with 200–500 mm annual rainfall. The species depends on orographic precipitation — moisture from Red Sea air masses forced upward by coastal mountains. Like Dracaena cinnabari, it is a nebulophyte: its crown captures fog and horizontal moisture. Recent data suggest an upward elevational shift in Sudan, possibly as lower populations die from progressive aridification and survivors regenerate at higher, cooler altitudes.
Conservation — A Species Disappearing in Real Time
Dracaena ombet is classified as IUCN Endangered (EN). However, Kamel et al. (2014) have formally suggested the species should be upgraded to Critically Endangered based on the Gebel Elba data: 80% of the Egyptian population is facing imminent extinction, and the species has lost nearly half its range in under two decades. The overall pattern across the species’ range is consistent: declining populations, aging trees, catastrophic recruitment failure, and progressive habitat loss.
The Erkwit Paradox
The story of Erkwit illustrates both the horror and the fragile hope of Dracaena ombet conservation. In the 1860s, Heuglin described millions of trees. By the late 20th century, mass death events had been reported, and the population was feared extirpated. The species was considered locally extinct at Erkwit. Then, in the early 2020s, saplings were discovered in a historically core distribution area — the first evidence of natural regeneration after decades of apparent extinction. Andersen et al. (2022) interpret this as a sign that Dracaena ombet, as a long-lived species with remnant population dynamics, can recover from severe disturbance events if grazing pressure is reduced and if rare recruitment windows (favorable rainfall years) occur. This does not mean the species is safe — far from it — but it demonstrates that the biological capacity for recovery has not been lost.
Threats
- Climate change and aridification: The progressive drying of the Red Sea Hills over centuries is the fundamental driver of decline. Annual rainfall and fog inputs are decreasing, pushing the species toward extinction at lower elevations. MaxEnt models predict suitable habitat reductions of 48–83% in northern Ethiopia alone.
- Overgrazing: Livestock browse seedlings, preventing regeneration. Flowers and fruits are fed to animals. Fencing exclosures in Ethiopia have demonstrated that seedling establishment is possible when grazing is excluded — proof that regeneration failure is partly anthropogenic, not purely climatic.
- Overexploitation: Local people harvest leaves for rope-making, wood for beehives and utensils, bark and resin for medicine. While traditional use levels were sustainable in large populations, in tiny remnant populations every tree lost is catastrophic.
- Pathogen attack: Disease-related dieback has been reported in some populations, though the specific pathogens are poorly characterized.
- Encroachment by pioneer shrubs: In degraded habitats, light-demanding pioneer species (Carissa spinarum, Tarchonanthus camphoratus) invade the understory and outcompete Dracaena ombet seedlings.
Ethnobotanical Value
The Afar and Tigrinya people of Ethiopia and Eritrea, and the Beja people of Sudan and Egypt, use Dracaena ombet for a wide range of purposes: household utensils, beehives, construction materials, traditional medicine (resin and root gum), rope (from leaves), food (fruits consumed in some areas), and livestock fodder. This deep ethnobotanical integration means the species’ decline has direct socioeconomic consequences for pastoral and agropastoral communities — a dimension absent from the conservation of island-endemic dragon trees like Dracaena cinnabari or Dracaena tamaranae.
Cultivation
| Parameter | Value |
|---|---|
| Hardiness | Unknown; estimated ~0 to −3 °C (montane habitat with occasional frost) |
| Light | Full sun |
| Soil | Extremely well-drained; rocky, mineral substrates |
| Watering | Very low; less water is better |
| Growth rate | Extremely slow |
| Difficulty | 5/5 |
Dracaena ombet is not available in the commercial horticultural trade. The species exists in cultivation only in a handful of botanical gardens holding ex situ conservation collections — most notably, micropropagation protocols have been developed at King Saud University (Dewir et al. 2019) specifically for conservation purposes.
If the species were ever available for conservation-oriented cultivation, the key requirements would parallel those of other dragon trees: perfect drainage, mineral substrate, minimal watering, full sun, and extreme patience. The montane origin (1,000–1,800 m in the Red Sea Hills) suggests marginally better frost tolerance than the lowland-tropical Dracaena cinnabari, but no hardiness data exist.
What to Know Before Buying
You cannot buy this species. Dracaena ombet is an IUCN Endangered species with no commercial trade. Any plant offered under this name would be a misidentification (most likely Dracaena draco) or illegally sourced. The species’ survival depends on in situ habitat protection and ex situ conservation breeding programs, not private horticulture.
Propagation
Seed: Seeds have long dormancy and low germination rates in the wild — a major factor in recruitment failure. In controlled conditions, germination has been achieved but remains difficult and slow.
Micropropagation: Dewir et al. (2019) developed an in vitro protocol for Dracaena ombet using shoot tip explants, achieving successful multiplication and rooting under controlled conditions. This technique offers the most promising path for large-scale production of conservation stock, bypassing the bottleneck of seed dormancy and low germination.
Pests and Diseases
Drought-related dieback: The primary cause of mortality across the species’ range. Not a disease per se, but a progressive physiological collapse driven by declining moisture inputs over decades.
Uncharacterized pathogens: Disease-related dieback has been observed but the causal agents are not well studied.
Agave snout weevil (Scyphophorus acupunctatus): Not documented on Dracaena ombet, but the weevil’s broad host range across the Asparagaceae makes it a theoretical risk in any future garden or arboretum setting.
Landscape Use
Purely theoretical — the species is not in cultivation outside conservation programs. However, Dracaena ombet would be of exceptional value for:
Arid-lands botanical gardens: As a representative of the East African–Arabian dragon tree clade, it fills a biogeographic niche that no other cultivated dragon tree occupies. A comparative display of Dracaena draco (Macaronesian clade), Dracaena cinnabari (Socotra), and Dracaena ombet (continental Africa) would illustrate the entire geographic arc of dragon tree evolution.
Ecological restoration in the Horn of Africa: The most important use. Community-based exclosure programs in northern Ethiopia have demonstrated that Dracaena ombet can regenerate when grazing is excluded. Integrating dragon tree restoration with alternative livelihood programs (honey production, non-timber forest products, ecotourism) offers the most sustainable path to species survival.
Frequently Asked Questions
How many dragon tree species exist?
The “dragon tree group” currently includes approximately 10 arborescent Dracaena species, all growing in seasonally arid climates with 200–500 mm annual rainfall and mean temperatures of 18–20 °C. The group splits into two major clades: the Macaronesian/western clade (Dracaena draco, Dracaena tamaranae) and the East African–Arabian/eastern clade (Dracaena ombet, Dracaena cinnabari, Dracaena serrulata, Dracaena ellenbeckiana, and others). Dracaena ombet is the continental African representative of the eastern clade.
Is the Nubian dragon tree really dying out?
Yes. The data are unambiguous. In Egypt (Gebel Elba), only 27% of recorded trees are healthy, only 1% are young, and the occupied area has shrunk by 46% in two decades. In Sudan (Erkwit), the population was feared completely extirpated before surprising sapling discoveries in 2022. In Ethiopia, populations are old, overexploited, and declining, with climate models predicting 48–83% habitat loss. Multiple research groups have suggested that the IUCN status should be upgraded from Endangered to Critically Endangered.
Is there any hope?
Yes — fragile, conditional hope. The discovery of saplings at Erkwit (Sudan) after decades of apparent local extinction demonstrates that the species can regenerate when conditions align. Fencing exclosures in Ethiopia prove that grazing exclusion allows seedlings to establish. Micropropagation protocols exist for ex situ production. But all of these require sustained institutional commitment in a region characterized by political instability, poverty, and competing land-use pressures. Dracaena ombet will survive only if its conservation is integrated into the livelihoods of the communities that share its habitat.
Reference Databases and Online Resources
Bibliography
- Andersen, G.L. et al. (2022). Endangered Dracaena ombet Population in the Red Sea Hills, Sudan, Recovers After Abrupt Change. Frontiers in Environmental Science, 10: 793583.
- Birhane, E. et al. (2023). Impact of land-use and climate change on the population structure and distribution range of the rare and endangered Dracaena ombet and Dobera glabra in northern Ethiopia. Journal for Nature Conservation, 76: 126506.
- Dewir, Y.H. et al. (2019). Micropropagation to conserve the endangered Gabal Elba dragon tree (Dracaena ombet Heuglin ex Kotschy & Peyr). HortScience, 54: 162–166.
- Gidey, T. et al. (2023). Prioritizing forest conservation strategies using a multi-attribute decision model to address concerns with the survival of the endangered dragon tree (Dracaena ombet Kotschy and Peyr.). Journal for Nature Conservation, 73: 126404.
- Gidey, T. et al. (2024). Population and conservation status of the endangered Dracaena ombet tree in dry Afromontane forests. Global Ecology and Conservation, 50: e02809.
- Kamel, M., Ghazaly, U.M. & Callmander, M.W.M. (2015). Conservation status of the Endangered Nubian dragon tree Dracaena ombet in Gebel Elba National Park, Egypt. Oryx, 49(4): 704–709.
- Kotschy, T. & Peyritsch, J. (1867). Plantae Tinneanae: 47.
- Lengálová, K. et al. (2020). First age-estimation model for Dracaena ombet and Dracaena draco subsp. caboverdeana. Forests, 11(10): 1–14.
- Maděra, P. et al. (2020). What We Know and What We Do Not Know about Dragon Trees? Forests, 11(2): 236.
- Shumbahri, M. et al. (2025). Understanding the ethnobotany of a native tree Dracaena ombet (Kotschy and Peyr.) and associated indigenous knowledge of the local people in the Afar drylands, Ethiopia. Ethnobotany Research and Applications, 31: 1–13.