Hydrangea aspera Buch.-Ham. ex D.Don

Roughleaf hydrangea

Synonyms
  • Hortensia aspera Ham. ex Dippel
  • Hydrangea alba Reinw. ex Miq.
  • Hydrangea aspera var. angustifolia Diels [nom. illeg. homonym. post.]
  • Hydrangea aspera var. cordata Pamp.
  • Hydrangea aspera f. emasculata Chun
  • Hydrangea aspera var. scabra Rehder
  • Hydrangea coacta C.F.Wei
  • Hydrangea glabripes Rehder
  • Hydrangea oblongifolia Blume
  • Hydrangea pubescens Decne.
  • Hydrangea vestita var. fimbriata Wall.
  • Hydrangea vestita var. integriloba DC.

Introduction

Roughleaf hydrangea (Hydrangea aspera Buch.-Ham. ex D.Don) is a deciduous woody shrub, occasionally approaching the stature of a small tree, and the taxonomic center of one of the most difficult species complexes in Hydrangea. It belongs to Subgenus Hydrangea, Section Asperae, an Asian lineage notable for hairy shoots and foliage, inferior ovaries, terminal flower clusters, and extraordinary variation in the type, density, and distribution of its indumentum. In its broadest historical sense, H. aspera has been stretched from the Himalaya through China and Indochina to Taiwan and used to absorb much of the morphological diversity now divided among several species. Current Kew treatment still uses a comparatively broad circumscription.

Tree TSAR adopts a narrower interpretation. Hydrangea kawakamii, H. villosa, and true H. strigosa are recognized as independently evolving species, while the robust Himalayan-to-Chinese morphotype is retained within H. aspera as ssp. robusta. This treatment follows the integrative work of De Smet and colleagues, who combined traditional plastid and low-copy nuclear markers, RADseq data, morphology, geography, and cytological evidence rather than allowing any single conspicuous hair type or database treatment to define the species.

The resulting species remains variable. De Smet's treatment describes shrubs or small trees usually 1–4 m tall but occasionally considerably larger, with leaves ranging from lanceolate to broadly elliptic or oblong, petioles from short to long, and lower leaf surfaces bearing either gray-white appressed hairs or more erect yellowish-brown hairs. Much of this variation intergrades rather than falling into neatly separated morphotypes.

Character

H. aspera

H. aspera ssp. robusta

H. kawakamii

H. strigosa

H. villosa

Core geography

Himalaya and mainland Asian complex

Eastern Himalaya through southern China and Indochina

Taiwan

Type-linked Hubei lineage

Central China; exact range needs refinement

Leaf underside

Variable appressed to more erect hairs; no diagnostic white papillae

Usually robustly hairy, often broad-leaved

Long erect hairs; small inconspicuous papillae

Short stiff appressed hairs plus conspicuous white papillae

Long erect/villous hairs, especially brownish on major veins

Petiole/stem tendency

Variable

Often stout, long-petioled, robust

Erectly hairy

Densely strigose

Villous, with long erect hairs

Cytology

2n = 34 or 36 reported

2n = 34 reported in studied material

2n = 36

2n = 34

2n = 34

Molecular signal

Shares a genetically permeable complex with robusta and Sichuan strigosa-like plants

Not sufficiently independent for species rank

Strongly supported lineage

Hubei lineage strongly supported

Strongly supported lineage

Tree TSAR rank

Species

Subspecies

Species

Species

Species

The hardest boundary is between H. aspera and ssp. robusta. RADseq and earlier multilocus analyses could not reliably distinguish nominal aspera, robusta, and Sichuan plants resembling strigosa when they occurred in geographic proximity. Intermediate plants bridge differences in leaf shape and pubescence, suggesting either heavy introgression among formerly differentiated populations or environmentally structured variation within a genetically permeable lineage. De Smet therefore recommended representing the most persistent aspera–robusta differentiation at subspecies rather than species rank.

True H. strigosa represents a different situation. Plants linked geographically to the Hubei type population form a discrete evolutionary lineage, while superficially strigose plants from Sichuan can cluster with H. aspera. The lesson is important: hair orientation alone does not reveal evolutionary history. In Section Asperae, remarkably similar-looking indumentum can occur in populations belonging to different genetic lineages.

The opposite problem occurs with H. kawakamii and H. villosa. Their gross morphology can look frustratingly close to H. aspera, which led major twentieth-century revisions to synonymize them. Yet both are independently supported by molecular species delimitation and RADseq data, while H. kawakamii also differs cytologically. Their restoration illustrates why subtle morphology should not automatically outweigh multiple independent lines of evolutionary evidence.

The meaning of ssp. robusta

Tree TSAR recognizes one non-nominotypical subspecies, H. aspera ssp. robusta. The subspecies preserves a geographically and morphologically meaningful robust, often long-petioled expression of the species without claiming the level of genomic independence expected of a separate species.

That rank is deliberately provisional rather than a declaration that the problem is finished. De Smet emphasized incomplete sampling from the respective type regions of H. aspera and H. robusta. Dense nuclear population sampling from Nepal, the eastern Himalaya, southwestern China, and intervening populations remains the clearest test of whether the current subspecies structure accurately represents the evolutionary history.

Horticultural and Agricultural Uses

The H. aspera complex has produced some of the most texturally dramatic hydrangeas in cultivation, but horticultural naming is considerably less precise than botanical naming. Large hairy leaves, purple or blue fertile flowers, pink to lilac sterile flowers, dark foliage, and imposing woodland-shrub architecture have made members of the complex prized collector plants. A lightly shaded, moisture-retentive setting generally suits them better than exposed heat or drought.

The complication is that many garden plants called simply H. aspera cannot confidently be assigned to narrow-sense H. aspera. Much established material more closely resembles H. villosa, while other cultivars are selections, seedlings, or hybrids of uncertain ancestry. Garden progeny within the complex can show substantial recombination in leaf shape, pubescence, habit, and inflorescence morphology. Consequently, the horticultural label “H. aspera” is often best interpreted as a complex-level identification unless the plant has documented wild provenance and matches a defensible taxonomic concept.

Breeding has remained comparatively limited relative to bigleaf, panicle, and smooth hydrangeas, although selection for dark foliage and distinctive flowers has produced notable modern cultivars. Hybrid fertility among close members of the complex reinforces the need to keep cultivar classification separate from the taxonomy of wild evolutionary lineages. Experimental hybrids have even been produced across the chromosome-number difference between H. sargentiana and H. kawakamii.

Conservation Concerns

Hydrangea aspera is currently listed as Least Concern by IUCN, reflecting its broad distribution and the absence of recognized species-wide threats.

That assessment must be interpreted taxonomically. The conservation concept underlying broad H. aspera can include populations that Tree TSAR recognizes as H. kawakamii, H. villosa, or other more narrowly circumscribed entities. A Least Concern assessment of the broad complex therefore cannot automatically establish security for each segregate lineage.

The principal conservation need for narrow-sense H. aspera is better population-level resolution. Genomic work that distinguishes introgression, environmental plasticity, and independently structured populations would allow conservation attention to focus on genuine evolutionary units rather than historically named morphotypes.

Additional Information

References and Further Reading

De Smet Y (2020) Through the fog: Evolutionary insights provide novel genus- and species-level boundaries in tribe Hydrangeeae and genus Hydrangea. PhD thesis, Ghent University, Faculty of Sciences, Ghent, Belgium. https://biblio.ugent.be/publication/8716125 (opens in a new tab)

De Smet Y, De Clerck O, Uemachi T, Granados Mendoza C, Wanke S, Goetghebeur P, Samain M-S (2017) Multilocus coalescent species delimitation to evaluate traditionally defined morphotypes in Hydrangea sect. Asperae (Hydrangeaceae). Molecular Phylogenetics and Evolution 114: 415–425. https://doi.org/10.1016/j.ympev.2017.05.021 (opens in a new tab)

Mortreau E, Siljak-Yakovlev S, Cerbah M, Brown SC, Bertrand H, Lambert C (2010) Cytogenetic characterization of Hydrangea involucrata Sieb. and H. aspera D. Don complex (Hydrangeaceae): Genetic, evolutional, and taxonomic implications. Tree Genetics & Genomes 6(1): 137–148. https://doi.org/10.1007/s11295-009-0235-8 (opens in a new tab)

Sutton J (2025) Hydrangea aspera. Trees and Shrubs Online. International Dendrology Society. https://www.treesandshrubsonline.org/articles/hydrangea/hydrangea-aspera/ (opens in a new tab)