Kindly sponsored by a generous supporter who loves hydrangeas.
Julian Sutton (2025)
Recommended citation
Sutton, J. (2025), 'Hydrangea kawakamii' from the website Trees and Shrubs Online (treesandshrubsonline.
Deciduous shrub. Branchlets dark grey, terete, hairy or not. Leaves opposite. Leaf blade papery, oblong-ovate to elliptic, 9–12 × 4.5–10 cm, with 6–9 secondary veins on each side of midvein; upper surface with short, sparse, appressed hairs; lower surface with dense, slightly crisped hairs; base broadly cuneate to rounded; margin irregularly doubly serrate; apex acute to shortly acuminate; petiole 2–9 cm, pubescent. Inflorescence a terminal corymbose cyme 10–14 cm across, flat-topped or slightly convex; branches tomentose; both fertile and sterile flowers present. Sterile flowers largely consisting of 3–4 petaloid sepals, white, greenish, pink or lilac, suborbicular, ~2 cm wide, margins toothed especially towards the apex. Fertile flowers with cupular calyx tube ~1.5 mm long, with triangular lobes ~1 mm long; petals free, purple to pink or bluish; stamens 10, unequal; ovary inferior; styles 2–3, recurved, ~1.5 mm in fruit. Fruit a hemispherical capsule with truncate apex, 3–4 mm across, dehiscing apically, between the styles. Seeds winged at both ends. Flowering July, fruiting October–November (Taiwan). (Wei & Bartholomew 2001; Li 1963).
Distribution Taiwan
Habitat Dense forests, 2200–2300 m asl.
USDA Hardiness Zone 7-8
RHS Hardiness Rating H5
Conservation status Not evaluated (NE)
Hydrangea kawakamii is the name often given to Taiwanese plants of the Hydrangea aspera complex, arguably distinct from H. aspera s.s.. At least in gardens they tend to make medium to large, often spreading shrubs with large, velvety leaves, and can be impressive in flower.
First described from Yushan, Central Taiwan, Hayata’s (1908) specific epithet commemorates Japanese botanist Takiya Kawakami, one of the collectors of the type specimen. Morphologically there is little to distinguish it from some variants of narrow-sense H. aspera; Flora of China (Wei & Bartholomew 2001) resorts to patterns of veining on the tiny seeds in the identification key, something one would be hard pressed to see without a microscope. McClintock (1957) sank it within her H. aspera subsp. aspera, although some Taiwanese and mainland Chinese botanists have continued to accept it (e.g. Li 1963). Surprisingly the name seems never to have been published as a variety or subspecies of H. aspera, and horticulturists have resorted to a cultivar group (Edwards & Marshall 2019), defined in practice more on provenance than on morphology.
However, there is evidence that this is more than a purely geographical taxon. A cytological study using numbers of living plants in the French reference collection at Angers divided H. aspera subsp. aspera into Kawakamii (Taiwanese) and Villosa groups, based on karyotype differences; significantly H. kawakamii has 2n=36 like the bulk of Hydrangea species, rather than 2n=34 as seen in the rest of the H. aspera complex (Mortreau et al. 2010). These authors suggested accepting both H. kawakamii and H. villosa, although the implication is that no plants in the study would then represent H. aspera s.s. (whatever that might be). A DNA-based study suggested that H. kawakamii is sister to H. strigosa, placing H. aspera s.s. in a different clade, and supporting the idea that it is a species in their own right (De Smet et al. 2017).
Botanical descriptions have been misleading as to habit and height. Li (1963) called it a ‘low shrub’, though Bleddyn Wynn-Jones has seen and collected from wild plants 4 m tall (pers. comm. 2023). Wei & Bartholomew (2001) describe it as a ‘climbing shrub’, perhaps a reference to ‘frutices ad truncos scandentes’ in the Latin protologue (Hayata 1908), but this is not normal in what is a bushy, free standing – if sometimes widely spreading – shrub.
The earliest introduction we can trace is the plant now known as ‘Taiwan’, grown from Taiwanese forestry authority seed by Maurice Mason, UK, in the second half of the 20th century (described under Hydrangea aspera Species Complex Cultivars). It helped establish a reputation for late flowering and strongly spreading habit. Many subsequent introductions by Crûg Farm, several of them named, the rest simply under BSWJ and other numbers (Crûg Farm Plants 2024) are beginning to show that it is more variable than had been imagined, with different clones flowering between July and October in Britain. Some are potentially very tall. Leaves tend to be broad to very broad (60 cm across is claimed for the wild parents of some!), velvety-hairy, often with quite long petioles, and sometimes with red or purple tints, especially in the veins and petioles. Sterile flowers are lilac/purple, usually from pink buds, sterile flowers paler, often with serrated sepals. Time will tell which will make the best garden plants.