Hydrangea hirta (Thunb.) Siebold

Unmistakeable once seen, this Japanese endemic is far too rarely grown in Western gardens. Individual inflorescences are small and lack showy sterile flowers, but the shrub’s growth form brings them together on flat-topped clumps which assort well with herbaceous plants in the mixed border. Their sweet smell, resembling Meadowsweet (Filipendula ulmaria) is also worthwhile in the summer garden. Light shade or – with adequate moisture – more or less full sun suits this species.

The complete lack of sterile florets, combined with thin, distinctly pubescent, coarsely toothed, usually ovate leaves with acuminate apices, is most unusual in Hydrangea. The fertile flowers are individually small, much like those of Section Chinenses, with distinctly smaller petals than those of Section Dichroa. Petals and stamens together contribute colour, typically a light blue (pinkish with low aluminium). Occasional white flowered plants have sometimes been called var. albiflora Honda or – perhaps more appropriately – f. albiflora (Honda) Okuyama, taxa rarely recognized today (Ohba 2001; Royal Botanic Gardens, Kew 2023). Leaf shape and serrations inspire the name Nettle-leaved Hydrangea, but this has given rise to the idea that it might have stinging hairs, a bizarre misconception spread by websites (e.g. Edrom Nurseries 2023). The Japanese vernacular name translates as ‘small hydrangea’.

Hydrangea hirta tends to make broad, almost flat-topped clumps 1–1.5 m high, with inflorescences terminal on erect, crowded, mostly unbranched stems. Branching is hidden away near the base. A study of shoot dynamics in a wild population makes it clear that this is in the nature of the species (Ishii & Takeda 1997). New shoots emerging at or near the base may produce several branches in their first year, but those which become thick enough to survive competition with their neighbours enter a mature phase, in which only one axillary bud per shoot comes into growth each spring, soon overtopping the dead remains of the terminal inflorescence and continuing shoot extension as a sympodium. Old stems eventually die when they become unable to produce a shoot vigorous enough to compete for light in the canopy. In the garden, it seems prudent to cut out weak old shoots before they reach this stage.

Hydrangea hirta was among the first Asiatic hydrangeas known to Western science, described by Thunberg (1784) as Viburnum hirtum. Thunberg would have met it during his 1775–6 travels in Kyushu and Honshu. The several superficial similarities between these two genera resulted in several hydrangeas being initially named as viburnums. It seems to have no particularly close living relatives. While McClintock (1957) grouped it with H. scandens and relatives, molecular evidence suggests a rather more isolated position. De Smet et al. (2015) place it alone in Section Hirtae, a branch of the Hydrangea II clade.

The species did not enter Western gardens quickly. The exact date and origin remain unclear, but it probably slipped in without fanfare during the later decades of the 20th century. Even the final (1984) edition of Michael Haworth-Booth’s once-authoritative monograph gives no clear indication that he had actually seen the plant in gardens while writing it off in four lines as having apparently ‘little or no garden value’. Bean (1981) omitted the species, Mallet, Mallet & van Trier (1992) make no mention of it in cultivation, while Lawson-Hall & Rothera (1995) who apparently grew everything they could obtain, admitted to having seen it only in the Savill Garden, Windsor. It is now seen occasionally in British, French and Belgian collections, more rarely in general horticulture. It is sometimes offered by a few European nurseries, including the wild collections BSWJ 8899 & 11022 (SE Honshu – Crûg Farm 2023). However, De Langhe (2017) warns that ‘plants in collections under the name Hydrangea hirta are often something else (most frequently H. anomala [very surprisingly!], H. arborescens, or H. scandens)’.

This is a similarly uncommon plant in North American gardens, recorded in a few Southeastern and Northwestern collections (University of Washington Botanic Gardens 2023; JC Raulston Arboretum 2023) and very occasionally offered by specialist nurseries as a ‘rarified rarity’ (Odyssey Plants 2023). Wild material collected in 1997 by Dan Hinkley and Darrel Probst under HC 970649 is grown at Heronswood, WA, although Ross Bayton admits ‘we only keep it for the sake of having the full set’ (pers. comm. 2023).

Probable natural hybrids are known between H. hirta and H. luteovenosa (= H. × amagiana Makino), H. serrata (= H. × mizushimarum H. Ohba), and H. scandens (Ohba 2001); these are occasionally passed around among the most hardcore collectors. We know of no attempts to use H. hirta in breeding programmes, although one might speculate that its distinctive habit could give rise to some useful garden novelties.