Arbutus xalapensis Kunth

TSO logo


Kindly sponsored by
a member of the International Dendrology Society


Julian Sutton (2021)

Recommended citation
Sutton, J. (2021), 'Arbutus xalapensis' from the website Trees and Shrubs Online ( Accessed 2024-05-28.


Common Names

  • Texas Madrone


  • Arbutus texana Buckley
  • Arbutus glandulosa Martens & Galeotti


Flat-topped steep-sided mountain often in arid areas.
(of a plant) Avoiding soil containing lime (i.e. usually growing on acidic soil).
Grey-blue often from superficial layer of wax (bloom).
Native to an area; not introduced.
(syn.) (botanical) An alternative or former name for a taxon usually considered to be invalid (often given in brackets). Synonyms arise when a taxon has been described more than once (the prior name usually being the one accepted as correct) or if an article of the International Code of Botanical Nomenclature has been contravened requiring the publishing of a new name. Developments in taxonomic thought may be reflected in an increasing list of synonyms as generic or specific concepts change over time.
(of habitat or site) Arid. (Cf. mesic.)


Julian Sutton (2021)

Recommended citation
Sutton, J. (2021), 'Arbutus xalapensis' from the website Trees and Shrubs Online ( Accessed 2024-05-28.

A very variable shrub or small-medium tree, to 15 m. Bark thin, red-brown, peeling off in large papery sheets, in old specimens starting to be retained from the base upwards. Leaves elliptic, (4–)5–11(–15) × 1.5–4.8(–6) cm, base and apex variable but base rarely tapered, margins entire or irregularly toothed, olive- or glaucous-green above, lighter beneath, glabrous or hairy. Inflorescence a terminal cluster of racemes, very variable in form, pedicels hairy, lengthening in fruit. Flowers held obliquely above the horizontal; corolla ovoid-urceolate, 5–6 mm, white or pinkish; flowering from November-February in the south, to March-May in the north. Fruits rounded, 6–9 mm across, ripening dark red, about 6 weeks after flowering in Mexico. (Sørensen 1995, Sørensen 2009, Grimshaw & Bayton 2009)

Distribution  El SalvadorGuatemalaHondurasMexico Throughout, except for Tabasco and the Yucatán Peninsula NicaraguaUnited States Texas, SW New Mexico

Habitat Upland terrain, (325–)2000–3000(–3400) m.

USDA Hardiness Zone 8-9

RHS Hardiness Rating H4

Conservation status Near threatened (NT)

Even more than Arbutus arizonica, A. xalapensis is worthy of any attempt to get it established (Grimshaw & Bayton 2009). As well as very fragrant white flowers in spring, followed by red fruit (sometimes copious), its bark is perhaps the most spectacular of all in the genus: it is a highlight of its native habitats (Nored 2011, Martin et al. 1998).

This is the most widespread madrone of Mexico, found in upland areas almost throughout the country, its range extending south through Central America, and north across the US border, where it is best known in the Texan hill country of the Edwards Plateau and the Trans-Pecos. A. xalapensis was described scientifically by the German botanist Carl Kunth, from a specimen collected near Xalapa, Veracruz by the 1799–1804 Humboldt & Bonpland expedition through the Americas (Kunth 1818). It is a most variable plant, distinguished from other Mexican species by the combination of peeling bark and the leaf base rarely being tapered, and from the Pacific Coast A. menziesii by its smaller leaves, less glaucous beneath, and red rather than orange fruit.

Although often found in riparian woodland, this is a tree of reasonably sunny places and grows well in exposed, xeric sites; nurse trees of other species sometimes appear to help seedling establishment in the wild (Tirmenstein 1990). In no way a calcifuge species, it is in fact associated with limestone in Texas. The current US champion grows in Brewster Co., TX; exceptionally large by northern standards, it was measured at 14 m × 411 cm, with a spread of 18.6 m in 2016 (American Forests 2021).

Beyond the wonderfully gnarled appearance of slow-growing upland trees, bark is its year-round attraction. While the mature bark is red, the colour of the underbark exposed as it peels varies from whitish and green to light red, even among wild Texan trees (Nored 2011); it seems likely that temperature and light intensity will also affect the rate or intensity of reddening. The red-brown wood is hard and dense but not durable; it has been used historically for making small turned or whittled items, as fuel and in making charcoal for gunpowder (Tirmenstein 1990). Most unusually, Moerman (2021) records no uses by indigenous North Americans; Nored (2011) suggests that this is because it was considered sacred. The fruits are eaten by many bird species (Tirmenstein 1990), while Nored (2011) illustrates a Texan tree heavily scarred by the claws of foraging bears.

A. xalapensis was introduced to Europe early in the 19th century but probably not grown in the open until reintroduced in 1938. Collected near Perote, Veracruz, by the British professional plant collector Edward Balls on an expedition to gather wild potatoes, a plant at Highdown, Sussex, first flowered in 1955, but died in the 1980s after ailing for many years (Bean 1976). A specimen at Tregrehan, Cornwall, grown from a 1994 Maurice Foster collection in the Sierra Peña Nevada, Nuevo León, is well established (if splendidly wonky), flowering each year in early summer and displaying shaggy bark which becomes red, paler on the undersides of the branches (Grimshaw & Bayton 2009), but still essentially cream coloured when seen in May (J. Sutton pers. obs. 2019). Material collected in Hidalgo (Gardner & Knees 5228 of 1992) is growing at Dunloe Castle, Co. Kerry, Ireland, within the natural range of A. unedo (Royal Botanic Garden Edinburgh 2021). Bean (1976) mentions several introductions planted by James Russell at Castle Howard, N. Yorkshire, in the 1980s, one by Keith Rushforth (KR 556, Nuevo León), J. Priest 97 collected in Coahuila, and several Russell collections from Veracruz. None of these survive today, and may well have succumbed during a period when lack of maintenance lead to ‘general jungliness’ and high humidity in Ray Wood (J. Grimshaw pers. comm. 2020). Attempts to grow more northerly genotypes in Europe seem strangely lacking, although there are three young specimens from a 2010 accession of United States provenance at Westonbirt, Gloucestershire (Forestry England 2021).

A fascinating series of autumn-flowering specimens – some very beautiful – at Ventnor Botanic Garden, Isle of Wight clearly have much to do with A. xalapensis and deserve detailed morphological and perhaps genetic study. The original two are said to have been the only plants raised from a Mexican seed introduction to the Hillier nurseries in 1969, originally labelled Arctostaphylos glandulosa (which is a Pacific Coast shrub; Arbutus glandulosa is a synonym of A. xalapensis). One, illustrated below, survives (11 m × 130 cm and 9 m × 118 cm, 2016 – The Tree Register 2021). The other had beautiful blood-red bark and proved impossible to propagate vegetatively, while neither set much seed (C. Kidd pers. comm. 2021). Seedlings from the two are grown around the garden and are immensely variable. These seedlings certainly include hybrids (J. Sutton pers. obs. 2021) and it has been suggested that the originals may be too (Kidd 2016, Architectural Plants 2021).

Far from mainstream in North American cultivation, Texas Madrone is sometimes seen in collections and private gardens in its native Texas and New Mexico, south of our area; a few nurseries around Kerrville, TX, propagate it (Natives of Texas 1998, Nored 2011). In our area, it grows in California’s Bay Area, both at San Francisco Botanic Garden (Grimshaw & Bayton 2009) and at Berkeley (from a Breedlove & Mahoney collection, Coahuila, 1994; and a Schoenfeld & Fairey collection, Nuevo León, 1997 – University of California Botanical Garden 2021). Significantly, there are several specimens well established in the semi-arid climate of Denver Botanic Gardens, CO, in the Water-Smart and Dryland Mesa Gardens (Denver Botanic Gardens 2021).

Cultivation has been plagued by the suggestion that this is a difficult tree to germinate and transplant. Most agree that like A. menziesii it is intolerant of root disturbance; containerised stock should be planted extremely carefully. Seed quickly loses viability when stored dry, so should be sown fresh; cuttings, budding and layering are all feasible (Tirmenstein 1990, Nored 2011). Cuttings from the Ventnor trees have proved difficult, but micropropagation has been successful (Architectural Plants 2021).