Hydrangea sikokiana Maxim.

Not the easiest of woodland hydrangeas to please, H. sikokiana stands out for its large, sharply lobed leaves, which make a lovely foliage effect if well grown. It is still a rare and experimental subject, though commercially available in a small way on both sides of the Atlantic.

This local endemic of southern Japan (Péan 2020) was first described scientifically by Maximowicz (1887). Its status as a distinct species seems never to have been in doubt. The fully inferior ovary is a clue to its relationship to H. involucrata, H. platyarguta and the rest of Section Asperae (McClintock 1957), confirmed in molecular studies (De Smet et al. 2015; Yang et al. 2024). The only other woody hydrangea with lobed leaves is H. quercifolia, but they are unlikely to be confused: Oakleaf Hydrangea leaves are much thicker and often more deeply, coarsely lobed. The overall ‘look’ of the H. sikokiana leaf is closer to the herbaceous Kirengeshoma palmata (another member of the Hydrangeaceae, most closely related to Deutzia – Kim et al. 2015).

Flowers are not this species’ greatest garden feature. The inflorescences are loose and open, with few, rather small sterile flowers. In one study removing the sterile flowers made little difference to pollinating insect visits – they are best regarded as vestigial (Wong Sato & Kato 2019). The primary attractant seems to be the fertile flowers themselves; they are scentless, produce no nectar, and are always white.

A plant grown at the Sir Harold Hillier Gardens, Hampshire, UK, from the 1970s until the early 2000s is the first we can trace in Western gardens; Harold Hillier’s records do not reveal its origin (B. Clarke pers. comm. 2024). There have been several subsequent introductions including one by Darrell Probst (Kii Peninsula, before 1998, still grown at the Arnold Arboretum – Arnold Arboretum 2024); several by Crûg Farm in the 1990s and 2000s, including BSWJ 5601 (Honshu), BSWJ 5910 & 11436 (Kyushu), and BSWJ 11174 & 11385 (Shikoku); and at least one Dan Hinkley collection made alongside the Wynn-Joneses (Crûg Farm Plants 2023).

Growing mainly in brighter parts of the forest (Péan 2020), light shade in the woodland garden seems ideal. Some find it tricky, perhaps ‘fastidious’ (Foster 2023). Joe Sime describes a plant growing satisfactorily in his woodland garden on the English-Welsh border for 15 years before suddenly dying at the same time as one of its seedlings, reason unknown (pers. comm. 2024). The Hillier plant died following overenthusiastic pruning, others propagated from it failing to establish (B. Clarke pers. comm. 2024). We have not heard of specimens grown in Europe beyond Britain and Northern France.

It is too little tried in North America to be sure where it will thrive. In the Pacific Northwest it grows well in Dan Hinkley’s Windcliff garden, coastal Washington (J. Grimshaw pers. comm. 2023), and is recorded from University of British Columbia Botanical Garden, Vancouver (University of British Columbia 2024). It failed to thrive in the summer heat and humidity of the Southeast for Dirr (2021). In the Northeast it grows well at the Zone 7 Arnold Arboretum, MA (Arnold Arboretum 2024), and at Cornell University in Zone 6 Ithaca, NY (Hinkley in Roach 2017).

Cuttings seem difficult to root (J. Sime pers. comm.), maybe explaining why it has never been easy to find in the nursery trade. Seedlings from garden plants often appear to be hybrids with nearby plants of the H. aspera complex (Dirr 2021). One such raised by Joe and Wendy Sime, probably with a nearby H. kawakamii as the pollen parent, has ‘ovate leaves with two or three tips at the end’, and flowers intermediate between the parents (pers. comm. 2024). In a failed attempt to combine its leaf form with purple tints, Dan Hinkley has crossed H. sikokiana with ‘Plum Passion’ (H. aspera complex); the resulting plants have intermediate form (see image below) but no sign of purple (Hinkley 2024). Another deliberate cross with H. aspera, ‘Smoked Salmon’ raised by Rhode Island nurseryman Ed Bowen, has achieved purple leaves but not yet a significant distribution; however, ‘we’re still playing the numbers game with this combo’ (Issima 2024).