Hydrangea aspera Buch.-Ham. ex D.Don

Hydrangea aspera is the central species in a variable and incompletely understood species complex, with a range stretching from the Himalaya to Indochina. The complex as a whole comprises medium to large deciduous shrubs, some with considerable garden presence. Young shoots and the undersides of the rather thick leaves are notably – though quite variably – hairy. Inflorescences are primarily terminal, sometimes with a few secondary flower heads in the uppermost leaf axils. They open from distinctively knobbly, rather cauliflower-like buds. Fertile flowers are usually pink to purple, and are distinguished by the emphatically inferior ovary; sterile flowers are white to pink or purple. Leaf size, shape and type of indumentum vary tremendously. Flowering time also varies a good deal within the group, mostly between early July and September in southern English gardens. Patterns of variation are so complex that neither botanists nor garden hydrangea enthusiasts have ever agreed on quite how the group should be subdivided. We reflect this by discussing the complex as a whole on this page, but also devoting separate articles to the six questionably distinct segregate species most likely to be met with, either as living plants or in name.

Careful reading of the description above (modified from Flora of China – Wei & Bartholomew 2001) suggests a very variable plant, even when several segregate species are recognised. It is in fact easier to define segregates than to pin down quite what a narrow-sense H. aspera might be. In practice, any form which does not comfortably fit the user’s idea of any segregate tends to be labelled H. aspera. This is conceptually messy, but reflects a messy reality.

Three of these segregates are defined by their indumentum. H. strigosa (C and S China, perhaps also Vietnam) is named for the short, stiff, appressed, white hairs on stems, leaf undersides and inflorescence branches, giving a rough feel. Leaf shape varies tremendously, and it is perhaps more a morphospecies than a biological one. Nevertheless is it widely recognised by botanists (e.g. Wei & Bartholomew 2001). A lower altitude plant than most asperas, it tends to be less hardy and later flowering in gardens (Bean 1981). H. villosa (C China) has short but less stiff, spreading white hairs thickly covering every vein on the leaf underside, right down to the very smallest, giving a coarsely velvety feel. Some botanists have entirely rejected it as distinct from H. aspera (McClintock 1957; Wei & Bartholomew 2001) but molecular evidence gives some support to its distinctness (Mortreau et al. 2010; De Smet et al. 2017). The possible artificiality of these two taxa is illustrated by Foster (2023) who found both types of indumentum, plus intermediates, on plants within 10 m of one another in scrub beside a mountain track in Sichuan. The bulk of material in general cultivation as H. aspera seems more or less to fit H. villosa. Perhaps the most distinct and least controversial segregate is H. sargentiana, a narrow endemic from western Hubei, which has become a garden classic. First year stems, petioles and leaf midribs are thickly covered in long, fleshy, transparent trichomes which taper from a thick base to a fine point; there is nothing else in Hydrangea quite like them (Sargent 1913; De Smet et al. 2017). With its very large ovate-oblong leaves and freely suckering habit, it makes a unique foliage effect.

A fourth species is defined primarily on geography. Taiwanese and some mainland Chinese botanists name material from Taiwan H. kawakamii (e.g. Li 1963; Wei & Bartholomew 2001). Those we have in gardens tend to be substantial, large-leaved plants.

Finally, two segregate species have strikingly long petioles, sometimes approaching or equalling the length of the leaf blade. H. longipes (C & W China) has thin leaves, rough beneath with sparse but stiff appressed hairs, and white petals and stamens in the fertile flowers (Sargent 1913; Wei & Bartholomew 2001). It is usually the earliest to flower of the complex. H. robusta (Himalaya, China, Vietnam) has thicker, much hairier leaves, and pigmented fertile flowers (Wei & Bartholomew 2001). There is ongoing debate over which garden forms might belong in H. robusta; many have red-tinted petioles and veins, and some can make magnificent specimens if well sited.

The unhappy alternative to this untidy situation is to call everything H. aspera and be done with it. This avoids the risk of error and ambiguity, but it fails completely to describe and name important, more or less distinct wild and garden variants. McClintock (1957) did something approaching this in her influential revision of the genus, accepting just four subspecies, aspera, robusta, sargentiana, and strigosa. The effect on naming and curation of garden plants, as well as on horticultural literature, has been chaotic.

Hydrangea aspera itself was the first of the complex to be described scientifically, from Nepal (Don 1825), but there is no solid evidence for 19th century introductions of any of these taxa. All the earliest documented garden introductions were by Ernest Wilson in the early 1900s (see HH. longipes, sargentiana, strigosa, villosa). Bean (1981), with his great experience of woody plant introductions through the first half of the 20th century, understood that the origin of plants distributed early on as simply H. aspera, or as H. aspera var. macrophylla (today’s ‘Macrophylla’) was unclear. Most of the variety, as opposed to bulk of material, we see in gardens is down to collections made in the last few decades.

Turning to garden performance, it is crucial to take account of variation in size, vigour and hardiness between and within the various taxa. Individual cultivars (see below) need to be taken on their own merits. The safe starting point is to recommend a sheltered, lightly shaded woodland situation in milder, maritime areas of Western Europe or America’s Pacific Northwest, but many can be grown well in much more diverse settings (Foster 2023; Bean 1981). Late spring frosts are particularly damaging, and freezing winds can cause shoot death even in the hardier forms. In smaller gardens where the less vigorous forms are in scale, large, tougher shrubs and walls can provide shelter. As with other hydrangeas, reasonable water availability through the growing season is ideal, although the tougher asperas can grow well even on thin chalk soils, given good soil preparation with organic matter incorporated, and ongoing mulching. Waterlogging is most unhelpful. Soil pH is not critical for growth; colour of sterile flowers seems unaffected by pH / aluminium availability, but fertile flowers do change somewhat. Hardier variants, especially garden forms close to H. villosa, can be successful in significantly colder areas: attention to microclimate is critical. Of 108 accessions listed in Belgian collections, the vast majority were of H. villosa and allied cultivars, other well tested modern cultivars, and H. sargentiana (Botanical Collections 2025).

The H. aspera complex has been largely ignored in North America until recent decades, and remains far from the mainstream (Dirr 2021). They have been most tried, and are probably most at home, on the Pacific coast from California’s Bay Area to British Columbia. Dirr (2021) describes acquiring a range of asperas and losing every one in Athens, Georgia, but goes on to list ‘respectable’ plants (probably mostly of the tougher sorts) in Georgia, the Carolinas, Pennsylvania and Illinois. The Arnold Arboretum lists only three accessions currently alive, the oldest from only 2018 (Arnold Arboretum 2025), emphasizing a general unsuitability for the Northeast, although H. sargentiana has flowered in the Coastal Maine Botanical Gardens (Dirr 2021). Again, microsite awareness is critical.

Most members of the complex have a chromosome number 2n=34, while H. kawakamii has 2n=36 like most other hydrangeas (Mortreau et al. 2010). Garden experience is that members of the complex appear to be interfertile, given the intermediate seedlings which appear. Certainly, fertile hybrids have been made experimentally even between H. sargentiana and H. kawakamii with different chromosome numbers (Crespel, Morel & Galopin 2012; Crespel & Morel 2014). Members of the H. aspera complex will also hybridize with the related H. involucrata (e.g. Bowen 2011).

This group has been largely ignored by commercial breeders, probably because of perceived low demand. Those we have are a mixture of choice wild finds such as the unusual mophead ‘Elegant Sound Pavilion’, selected chance seedlings from gardens (e.g. ‘Velvet and Lace’, a superb villosa), or sports such as the yellow-leaved sargentiana ‘Giel’ (GOLD RUSH); others are the work of a few less commercial breeders. Maurice Foster (UK) has put much effort into breeding dark-leaved cultivars including ‘Hpopr012’ (HOT CHOCOLATE). For easy comparison we list them all together under Hydrangea aspera Species Complex Cultivars.