Malus kansuensis (Batal.) Schneid.

TSO logo


Kindly sponsored by
Francine: 'after many informative Tours and Study Days with the IDS I feel it only fitting to help and promote such a wonderful organisation'


Julian Sutton (species), Nick Dunn (cultivars) (2021)

Recommended citation
Sutton, J. & Dunn, N. (2021), 'Malus kansuensis' from the website Trees and Shrubs Online ( Accessed 2024-04-21.


Common Names

  • long dong hai tang


  • Pyrus kansuensis Batal.
  • Eriolobus kansuensis (Batal.) Schneid.
  • Sinomalus kansuensis (Batal.) Rushforth


(pl. calyces) Outer whorl of the perianth. Composed of several sepals.
Lacking hairs smooth. glabrescent Becoming hairless.
Flower-bearing part of a plant; arrangement of flowers on the floral axis.
Egg-shaped; broadest towards the stem.



Julian Sutton (species), Nick Dunn (cultivars) (2021)

Recommended citation
Sutton, J. & Dunn, N. (2021), 'Malus kansuensis' from the website Trees and Shrubs Online ( Accessed 2024-04-21.

Shrub or small tree, 3–5 m, sometimes strongly armed with stout thorns. Branchlets purplish and finely hairy when young, glabrous later. Buds dark purple, ovoid; scales tomentose at margin. Leaf blade ovate, 5–8 × 4–6 cm, dark green and glabrous above, paler and sometimes sparsely hairy beneath; base rounded or truncate; apex acute or acuminate; margin doubly toothed, often 3-lobed; petiole 1.5–4 cm. Inflorescence a 4–10-flowered corymb, 5–6.5 cm across; pedicels 2.5–3.5 cm, sometimes sparsely hairy. Flowers 1.5–2 cm diameter, in late spring (May-June in the wild). Sepals triangular-ovate or triangular-lanceolate, 5–6 mm, shaggy above, often falling before fruit ripe; petals white, broadly obovate, about 1cm; stamens 20, unequal, about half the length of the petals; styles 3, slightly exceeding the stamens. Fruit yellow to red or purple, with pale dots, July to August in the wild, ellipsoid or obovoid, 1–1.5 cm diameter, with few stone cells. (Gu et al. 2003; Cullen et al. 2011; Bean 1981).

Distribution  China Gansu, Henan, Hubei, Shaanxi, Sichuan, perhaps Qinghai.

Habitat Mixed forests, scrub; 1500–3300 m asl.

USDA Hardiness Zone 5-9

RHS Hardiness Rating H6

Conservation status Data deficient (DD)

Malus kansuensis belongs to the informal group of East Asian species with hawthorn-like flowers and an elongated inflorescence axis, but has a more or less deciduous calyx. Pretty enough with its small white flowers, elongated red-flushed but ultimately brown fruits, and sometimes good autumn colour (Bean 1981; Edwards & Marshall 2019; Jacobson 1996), it is largely eclipsed in gardens by showier species such as M. transitoria. Taxonomists have tried to divide this group in diverse ways (Phipps et al. 1990; Juniper & Mabberley 2019; Rushforth 2018). The evolutionary relationships of M. kansuensis remain unclear, but it is worth noting that several molecular studies group it and M. honanensis with M. yunnanensis and relatives, which have persistent calyces and 5 styles (Li et al. 2012; Nikiforova et al. 2013; Yousefzadeh et al. 2019). It is certainly more like these species in general appearance than M. toringo and relatives.

Together, the often broadly ovate leaves, 3-veined at the base, sharply toothed, most often with 3 shallow lobes, more or less glabrous at maturity, and the elongated reddish fruit with deciduous calyx help distinguish this species (Sargent 1916; Gu et al. 2003). Perhaps the most similar-looking species is M. honanensis, which differs in its more deeply lobed leaf margin and more rounded fruit, often with a persistent calyx. Quite widespread in central China, M. kansuensis was described (as a Pyrus) by Batalin in 1893, from specimens collected in Sichuan by Grigory Potanin, and in Hubei by Augustine Henry. The specific epithet derives from Kansu, an alternative romanization to Gansu, where it is also native.

The first certain introductions were by Ernest Wilson, collecting for the Arnold Arboretum in western Hubei (W 264 of 1907) and in western Sichuan (W 4115 of 1910), but the idiosyncracies of Wilson’s records leave open the possibility of an earlier collection for Veitch (Bean 1981; Sargent 1916). One of at least two trees from W 4115 had flowered at St Clere House, Kent by 1919 (Bean 1981); two were still standing in 1999, measured then at 5 m × 110 cm and 5 m × 79 cm (The Tree Register 2020); still extant at the time of writing, they are far from shapely, but characterful (S. Harvey pers. comm. 2020). There is a beautiful specimen with a broad, shallow crown, derived from W 264, labelled ‘Calva’ (see below) and apparently dating from 1938 at the Royal Botanic Garden Edinburgh, measured at 5 m × 73 cm in 2014 (Royal Botanic Garden Edinburgh 2020, The Tree Register 2020). Another example at Leicester University Botanic Garden, dating from 1979 had made a dense 6 m tall bush by 2015 (The Tree Register 2020). More recent introductions include HOMC 22 and 2170, from Sichuan; specimens from both are grown at the Yorkshire Arboretum where they form multi-stemmed, upright shrubs with abundant stout thorns (J. Grimshaw pers. comm. 2020).

Elsewhere in western Europe, M. kansuensis is grown at both Arboretum Wespelaar and Meise Botanic Garden, Belgium (Arboretum Wespelaar 2020; Meise Botanic Garden 2020). Despite an early introduction to North America, it remains extremely rare (Jacobson 1996). There are several at the Arnold Arboretum, MA including a grafted tree from the original W 264 introduction (14 cm basal diameter in 2018) and one labelled var. calva from SABE 893 (Hubei, 1980; 14 cm basal diameter, 2018 – Arnold Arboretum 2020). Another specimen from SABE 893 is recorded at Berkeley Botanical Garden (University of California Botanical Garden 2020). At the Morton Arboretum, IL, there are three of Gansu provenance (Morton Arboretum 2020).

f. calva Rehd.

Malus kansuensis var. calva (Rehd.) T.C. Ku & Spongberg
Malus kansuensis 'Calva'

Pedicel and calyx glabrous, leaves downy only on the veins beneath (Bean 1981; Sargent 1916). Probably of very little botanical or horticultural significance, this variant has constantly been present in collections and literature, at several levels of the taxonomic hierarchy. Both variants can result from a single seed collection (W 4115, see above) so it scarcely warrants varietal status, but neither is it a clonal cultivar (trees from W 264 have also been identified as ‘calva’); ‘form’ seems the most appropriate level.