Cornus capitata Wall.

Once viewed as a species for mild, sheltered gardens, Cornus capitata has undergone something of a reincarnation in recent decades, as new, hardier introductions and a changing climate have brought it within reach of a much wider range of growers. The tree is one of the ‘strawberry-fruited’ members of subgenus Syncarpea, a group of rather similar – indeed, all too easily confused – species with a distribution throughout East Asia and westwards along the Himalaya. Cornus capitata shares with them the large ornamental bracts (in this case typically yellowish creamy-white) that give a spectacular spring display, while the small globose heads of flowers ripen into compound fruits that are edible when ripe, often seen in Chinese markets.

The species was first described in 1820 by the superintendent of the Calcutta Botanic Garden, Nathaniel Wallich, from specimens collected by British diplomats and East India Company functionaries in northern India (Kashmir, Uttarakhand and Himachal Pradesh) and Nepal (Roxburgh & Wallich 1820; Wallich 1829–1849). Wallich (1832) reported finding it himself at the top of Chandagiri (2550 m asl) overlooking Kathmandu, Nepal, ‘covered with yellow heads of scentless flowers in the month of June’, the yellowish inflorescences nicely figured in the accompanying plate. A few years later, Lindley found it impossible to believe that the syncarpous fruits could belong in Cornus, so assigned the tree to a new genus, as Benthamia fragifera (Lindley 1833). Lindley’s name – a valid synonym, though the genus Benthamia Lindl. is illegitimate since the name was already in use for a species of orchid – gained broad acceptance, and it was as the ‘Strawberry-Fruited Benthamia’ that the plant was known for much of the nineteenth century.

Cornus capitata was first introduced to Europe in 1825, when Anthony Buller, a Calcutta judge, sent seed to his relation J. H. Tremayne at Heligan. The provenance of this seed is not known, but may have been obtained through the Calcutta Botanic Garden – even a rough indication of its ultimate origin is unavailable, given that Wallich’s several types came from an area spanning several hundred miles. From Buller’s seed, Tremayne’s gardener managed to grow a plant, unprotected, on the stiff clay of this coastal garden. By 1833 this tree was 5 m tall and fruiting heavily (Loudon 1844), enabling the planting of huge numbers of seedlings along Heligan House’s newly laid out Long Drive. By the end of the century, the spectacular dogwood-lined approach to the house was acknowledged as one of the finest in the country, and the species became synonymous with Heligan. (The modern re-emergence of the ‘lost’ gardens is a celebrated horticultural success story, and the famous Benthamia avenue is being restored.)

The importance of provenance

Cornus capitata is noteworthy as a very early Himalayan woody introduction to Western cultivation. While botanical exploration in the southern Himalaya had been underway since the travels of Francis Buchanan-Hamilton in 1802, it took several decades – and growing familiarity with the northern Indian flora on the part of British colonists – for the suitability of newly discovered Himalayan species for the British climate to be seriously contemplated. Cornus capitata is historically significant as the subject of an important paper by John Forbes Royle, in which the climatic similarities between the Indian western Himalaya (Uttarakhand) and the British Isles were investigated systematically and in detail for the first time (Royle 1835).

It soon became clear that plants from this early Himalayan introduction were tender in all but the most benign conditions, and in Britain the species gradually consolidated its reputation as a tree for mild gardens (Arnold-Forster 1948). Trees with this provenance were (and still are) seen in their pomp in mild western localities such as Cornwall, and – perhaps above all – the west of Ireland where they best exhibit their evergreen character, in mild winters at least. But Bean, writing in 1914, remarked it ‘hopeless’ to attempt to grow the species without protection in London, citing a miserable plant that survived on a wall at Kew, where it rarely even flowered (Bean 1914). Introduced at a remarkably early date to the New York Botanic Garden – then located on Broadway – Cornus capitata was treated as a ‘greenhouse and hothouse plant’ (Hogg 1834).

While Bean’s remarks are undoubtedly true of older Himalayan stock, subsequent introductions throughout the twentieth century have made available new material from hardier populations in southwestern China and the Sino-Himalayan region, which show great promise. The earliest British introduction from China grew at Highdown, West Sussex, at one time considered to be the easternmost garden where the species was cultivated in Britain (Bean 1976). It was raised from seed collected by Forrest’s team in Yunnan on the expedition sponsored by Henry McLaren in 1932. A plant from this collection (McL C40) was planted at Highdown by Sir Frederick Stern (Stern 1974). Other material from this expedition (McL C221 – seeds obtained in China ‘by native collectors’) was introduced to the United States from Bodnant in 1933 under PI 103122 (United States Department of Agriculture 1935).

Earlier collections from northwest Yunnan were made by Joseph Rock in 1922–1923, namely Rock 6684 gathered at 2400 m asl, two days’ travel west of Yangpi in September 1922; Rock 6791 from the lower slopes of the Yulong (Jade Dragon Snow) Range near Lijiang, at 2700 m asl; and Rock 7008 from the mountains north of Tengchong (Tengyueh) at 2100 m asl. This material was introduced to the United States under PI 56085, PI 56301 and PI 56302, respectively (United States Department of Agriculture 1925).

Many subsequent introductions followed when China ‘reopened’ to Western botanists in the 1980s and the species was collected from its upper altitudinal range. As had been the case earlier in the century, northwest Yunnan proved a rich hunting ground: seed collected on the Alpine Garden Society Expedition to Yunnan in 1994 (ACE 2033) was shared widely through the Society’s distribution list: one seedling has made an impressive, free-flowering 8 m tree at the Sir Harold Hillier Gardens, Hampshire (Tree Register 2023). Also at Hillier’s is a Sichuanese collection by Mikinori Ogisu (Ogisu 93333), an excellent, free-flowering clone from a collection near Leibo, proven hardy at well below –10°C in the extended British freezing spell of 2010 (Pan Global Plants 2026) – its spherical fruits are atypical of the species (but see below). Howick and Macnamara collected the species at 2830 m asl near Lijiang, northwest Yunnan in 1992 (H&M 1563), plants from this raised at Howick Arboretum, Northumberland first flowered more than two decades after planting, in 2014. Plants from B&L 12175, also collected in Yunnan, grow at Ness Botanic Garden and as young plants at the Castle Howard Arboretum where they have not been affected by frosts of –8°C. At Edinburgh ENEP 142, collected in Nepal in 2001 grows on the Chinese Hillside, and – useful to note – thrives particularly well on the north side of the East Gate Lodge, where it has formed an attractive small tree with pinkish-white, occasionally pluribracteate involucres (Acc: 20011064).

By and large, Cornus capitata remains a European preoccupation, its easy-going character exploited by growers along the Atlantic seaboard from Northern Spain to Ireland. Rock’s high-altitude introductions from China appear to have made little or no impression in the United States, where the species is treated with kid gloves and not much grown. Recommended for the West Coast climate by Dirr (2009), it does not have much representation outside Northern California (the botanical gardens at Sonoma, San Francisco and Berkeley) and Oregon (Hoyt Arboretum, Portland). It also has a presence in cooler Australasian gardens.

A handful of cultivars are commercially available; others, such as ‘Legacy’ are known by name only and do not circulate. It is to be hoped that increased familiarity with the diversity of material now available will allow for the propagation of clones of proven hardiness, identifiable by clonal names if necessary. A purported Bhutanese cultivar, ‘Mountain Moon’, commonly listed under C. capitata, turns out not to belong to this species, though its taxonomic position is uncertain (see Cornus ‘Mountain Moon’).

Identification

In appearance, Cornus capitata has been compared to an evergreen version of C. kousa, from which it is, however, easily distinguished by its slender, fewer-veined, leathery leaves, obovate bracts, much more crowded inflorescences, and compressed fruits (the exposed buds are a further useful character in winter); in leaf shape, C. capitata recalls the more easterly species C. elliptica, which has more leathery leaves, with glossy (not matt dull green) upper surfaces, and globose fruits: where these two species are grown together in cultivation, C. elliptica flowers several weeks later (M. Wenzel, pers. comm. 2026). Cornus elliptica was formerly treated (as subsp. angustata) as an easterly subspecies of C. capitata (still the position reflected in POWO 5/2/2026) but is recognised as a distinct species in Flora of China (Xiang & Boufford 2005). The Flora’s approach is broadly congruent with available molecular phylogenies (e.g. Du, Cheng & Xiang 2024; Guan et al. 2024), but the distinction between Cornus capitata and C. elliptica is not crisply delineated, either morphologically or geographically. Where the two species overlap, intermediates and/or possible hybrids occur, in which the distinguishing characters are mixed. In the typical form of C. capitata the syncarps are compressed, held on thicker, shorter peduncles, but in western Guizhou specimens have been found with the narrower leaves of C. capitata but globose fruits, and others with compressed fruits and elliptic leaves; likewise trees with globose fruits on thick peduncles and vice versa (Xiang & Boufford 2005). (Such anomalies throw interesting light on the atypical fruits of Ogisu 93333 of southern Sichuanese provenance.) In a similar vein, the identity of E. H. Wilson’s collection with narrower, kousa-like bracts from western Hubei (Wilson 1851) appears questionable, occurring outside the species’ documented range; indeed W 1850 from the same area is the type of Rehder’s C. capitata var. mollis, placed by Flora of China in synonymy with C. elliptica (W 1851 comprised collections from two trees, the other closer to typical C. capitata) (Rehder in Sargent 1916; Xiang & Boufford 2005).

Cultivation

Though it grows as an understorey species in its native range, in our area of study Cornus capitata flowers best in more open settings. Plants raised from seed can be slow to flower, making it worthwhile to seek out younger-flowering grafted plants offered by a few nurseries (Junker Nursery 2009). Once questions of hardiness are accounted for, Cornus capitata is a surprisingly undemanding subject, wind-hardy and tolerant of coastal exposure, said to endure full sun ‘in all but the hottest, driest climates’ (Hogan 2008), even to thrive on chalk (Stern 1974), and to ‘sow itself by the hundred [in] Cornish woodland gardens’ (Arnold-Forster 1948). As indicated by such reports, germination is straightforward and prolific as long the fruits are not allowed to dry. McMillan Browse (2017) recommends maceration then two days’ fermentation of the fruit, followed by washing and immediate cold stratification at 3°C for 42 days. Germination then proceeds naturally at 20°C.