Hydrangea macrophylla (Thunb.) Ser.

This medium sized deciduous or semi-evergreen shrub from coastal Japan has the dubious distinction of being the principal wild ancestor of one of the world’s most widely grown and easily recognized garden plant groups. We first discuss the wild plant, occasionally grown in collections, and contrast it to its close relative H. serrata. Then follows a general discussion of the cultivars so dominant in gardens, their origins, diversity, cultivation and breeding. These have traditionally been listed under H. macrophylla even though a still-uncertain proportion of them have genetic input from H. serrata. About 300 cultivars – a mere sample – of Section Macrophyllae (H. macrophylla, H. serrata, and hybrids between them) are described in a series of alphabetical articles, ‘Hydrangea Section Macrophyllae Cultivars A’ etc.

Wild-type Hydrangea macrophylla is a bulky, thick stemmed plant which shoots freely from the base. Its leaves are thick textured, glossy and essentially glabrous (Ohba 2001), looking absolutely the coastal plant that it is. Inflorescences (typically lacecap form) are terminal; the sterile flowers usually have 4 roughly equal sepals, most often with entire margins, but having 3 sepals or toothed margins is not rare. It is not a very hardy plant, though tolerant of coastal exposure and bright light. Hirasawa (2017) gives a fascinating, illustrated insight into the habitats and natural variation of the species on a small section of the Izu Peninsula coast where H. serrata also occurs, sometimes hybridizing with it. Flower colour in H. macrophylla is particularly susceptible to aluminium availability, the classic blue versus pink issue (see below).

H. serrata is the closest living relative of H. macrophylla (McClintock 1957; De Smet et al. 2015), and the two can hybridize. H. serrata is a lower growing, more finely textured plant, rarely much above 1 m tall, with slender branches and smaller, thinner, matt-textured leaves (5–10 cm long versus 10–15 cm) which often have red or purple tints especially when young and again in autumn. Its inflorescences are smaller, terminal but sometimes also on short lateral branches along the previous year’s growth; at least in gardens it tends to flower earlier. A more widespread species, its main habitat is montane woods; it is much hardier but less tolerant of sun and exposure (Ohba 2001; Foster 2023; van Gelderen & van Gelderen 2004).

Mophead mutants (temari to Japanese enthusiasts) in which the inflorescence is covered in sterile flowers, hiding any fertile ones, are known in both species, as are double-flowered (yae) plants, with many extra sepals in the sterile flowers. While intermediate inflorescence forms are occasionally seen, most plants are clearly either lacecaps or mopheads. At least in H. macrophylla-based cultivars, both mophead inflorescence and double flower are single gene recessive characters (Waki et al. 2018).

Some of the cultivars usually attributed to H. macrophylla conform to the species description above. The early mophead introductions ‘Joseph Banks’ and ‘Otaksa’ (see below) are probably pure macrophylla (Association Shamrock 2022), thriving in maritime gardens but defoliated by the first hard frost. ‘Sea Foam’ is a lacecap sport on ‘Joseph Banks’ and for decades was probably the closest thing to a normal, wild-type H. macrophylla in European gardens (Bean 1981; Foster 2023). More recent wild introductions, especially by Corinne Mallet, France, are occasionally seen in collectors’ gardens; they have at least curiosity value. One such in Maurice Foster’s sheltered woodland in Kent, UK, has extremely glossy, scarcely toothed leaves, is semi-evergreen and has not flowered (M. Foster pers. comm. 2023).

The taxonomic and garden histories of Bigleaf Hydrangea are intertwined. Western science first engaged with it in the late 18th century. Carl Thunberg, working in Japan for the Dutch East India Company, recognized both these species, mistook them for viburnums (understandably given their inflorescence form and opposite leaves) and described Viburnum macrophyllum and V. serratum (Thunberg 1784). Both were eventually transferred to Hydrangea by De Candolle (1830). Thunberg’s mopheaded type specimen of V. macrophyllum has been equated with the cultivar ‘Otaksa’, later introduced to European gardens from Japan by Philipp von Siebold around 1860. But living plants were reaching Europe from Chinese cultivation as early as the 1780s. Jussieu (1789) based the genus Hortensia on one of these, spawning a common name for the mopheads; inconsequential speculation surrounds the identity of its namesake Hortense. In 1789 Sir Joseph Banks at Kew was one of several to obtain a mophead now known as ‘Joseph Banks’, most likely through merchants based in Canton (present day Guangzhou); within 10 years it was ‘very common’ in Britain (Curtis 1799). Western botanists seem to have been unaware of H. macrophylla in its wild state until Wilson (1923) drew attention to it. From a garden standpoint, three more very significant introductions from Japanese cultivation were made around 1880 by the Veitch nursery, UK. These were ‘Mariesii’ and ‘Rosea’ (both mopheads) followed by a lacecap, ‘Veitchii’; Veitch’s collector Charles Maries was responsible for the first two, perhaps all three. They are now often considered hybrids with at least some genetic input from H. serrata (Association Shamrock 2022; Foster 2023), although molecular evidence is incomplete. Finally, two 20th century workers further complicated naming in this group. Firstly, Elizabeth McClintock took exceptionally broad views of Hydrangea species in her 1957 revision of the genus, treating H. serrata as a subspecies of H. macrophylla; one of the more patchily adopted of her taxonomic positions, it still should be borne in mind when interpreting mid to late 20th century work, and given that at least one molecular study places H. macrophylla within rather than sister to the H. serrata complex (Hirota et al. 2022), may yet find favour. Secondly, while nurseryman Michael Haworth-Booth did much to increase knowledge and appreciation of hydrangeas in Britain, he muddied the water by using the novel, perhaps invalidly published, and entirely unnecessary specific name H. maritima for wild H. macrophylla, as well as H. acuminata Siebold & Zucc. for H. serrata (Haworth-Booth 1950, 1984) Fortunately, few followed his lead.

20th century breeding from these founders of the European stock began in France around 1903 with nurseryman Victor Lemoine raising seedlings from ‘Mariesii’; his lacecap selections (now called ‘Mariesii Grandiflora’, ‘Mariesii Lilacina’ and ‘Mariesii Perfecta’) are still grown. Early experiments in mophead breeding by Émile Mouillere, again in France, yielded such enduring varieties as ‘Générale Vicomtesse de Vibraye’ and ‘Madame Émile Mouillère’, both introduced 1908. Serious breeding work spread to Belgium, the Netherlands, Germany and Switzerland, feeding a market which was driven by the potted plant trade in areas too cold for them to be reliable open-garden plants. Showy mopheads in strong colours, which suited pot culture and could easily be forced into flower, extending the selling season, were favoured. Cultivars were named in vast numbers, and if some proved worthwhile long-term garden plants it was more by accident than design. Mophead hydrangeas are also important cut flowers: some bred for this market have also proved gardenworthy. The Swiss-bred Teller Series, introduced from the 1960s, brought lacecaps back into the commercial mix; these were tested in the open garden at several sites in Europe, and many (e.g. ‘Blaumeise’) have proved excellent garden shrubs. British, North American and New Zealand markets in the early to mid 20th century were largely fed by continental European breeding (Foster 2023; Haworth-Booth 1984).

It is still unclear to what extent H. serrata influenced this first body of European cultivars. With a few exceptions which look intermediate (‘Preziosa’ is one), the great bulk resemble H. macrophylla in habit, leaves, inflorescence size and position, although some are hardier than others, some have less glossy leaves, and so on. They are usually classified as such. On the other hand Haworth-Booth (1984) considered the whole lot a hybrid race, surely an overstatement, but given the possibly hybrid status of some of the founders, it seems fair to suggest that taken together the cultivars at this stage were an introgressed population, heavily weighted towards H. macrophylla, but at least some with a small input of H. serrata genes. Several molecular studies on the cultivars (Reed & Rinehart 2007; Sukhikh et al. 2018; Wu & Alexander 2019) have revealed interesting groupings, but without wide sampling of the wild species and all the important ‘foundation’ cultivars, they do not answer the question definitively.

Hydrangea breeding is still big business in continental Europe, with the focus still on compact potted plants for the conservatory, terrace or house. In recent decades North American breeding has taken off (Dirr 2021), and new Japanese- and Chinese-bred plants have begun to enter Western markets more directly. There has been a serious push towards extended flowering period, partly by selecting for long-lasting blooms, but also for remontancy, the hydrangea world’s favoured term for repeat-flowering (see below). This second goal has brought a new influx of H. serrata genes into the pool. The business models behind commercial hydrangea production are increasingly complex. While some companies breed, produce and market their own plants, it is every bit as common to see new varieties selected by a breeding company, protected and licenced through another specialist business, produced under licence by one or more nurseries (sometimes independently in different parts of the world), and marketed under plant brands which are often nation- or region-specific. McClellan (2024) gives a contemporary insight into plant branding in North America. Successful varieties are often marketed under multiple Trade Designations across time and space. For example, the plant whose formal cultivar name is the ‘nonsense’ breeder’s reference Hydrangea ‘Horwack’ has been marketed in Europe as both SCHLOSS WACKERBARTH (in both the Saxon® and Royalty® Series) and GLAM ROCK, but in North America as PISTACHIO®.

Remontancy has two components. Firstly, plants which produce terminal heads followed by others on short side shoots inevitably flower for longer. Crosses with H. serrata can sometimes achieve this, e.g. H. ‘Mak20’ (TUFF STUFF™ / BLUEBERRY CHEESECAKE), bringing greater hardiness at the same time. Secondly, an ability to flower on the current year’s basal shoots is valuable in cold climates when most or all woody growth dies back in winter, but may extend flowering even where old wood routinely survives. A number of such plants turned up in the 1990s, many of them otherwise rather ordinary ‘finds’ spotted flowering after severe winters; ‘Bailmer’ (ENDLESS SUMMER®) and ‘All Summer Beauty’ are well known examples, but there are more. Molecular work apparently inspired by commercial rivalry established that most were closely related, but distinct (Lindstrom, Pelto & Dirr 2003). These original finds have been used in breeding; their descendants are typically launched as remontant, but in the real world some rebloom better than others (Boebel 2011; Boebel 2018; Dirr 2021).

Flower initiation and its triggering factors are at the heart of remontancy. In non-remontant H. macrophylla flowers are initiated in the terminal bud during autumn in shoots which have not flowered that year (Harraß, Gundermann & Tränkner 2023; Joeright 2009). Because shoots which have flowered terminated in an inflorescence, growth continues the following year from axillary buds immediately behind the inflorescence, and it is in these that flowers are initiated, again in autumn. This is why pruning at this stage, or winter shoot death, reduces or prevents flowering. Both cooler temperatures and shorter daylength seem to be involved. A 7 or 8 hour photoperiod encourages initiation; following winter dormancy both higher temperature and longer days accelerate subsequent development of the inflorescence (Shanks, Mityga & Douglass 1986). Commercial liner producers will supply pre-chilled plants, with flowers already induced and winter dormancy mimicked, to allow growers to produce flowering plants quickly over a long season (Joeright 2009). In remontant cultivars which bloom on the current year’s shoots, flower initiation is independent of daylength and temperature (Harraß, Gundermann & Tränkner 2023). The underlying molecular biology is under investigation (Liu, Lyu & Lyu 2023; Li, Lyu & Lyu 2024). The reasons behind some cultivars flowering on side shoots as well as terminally are less studied.

The cold and short day requirement in non-remontant cultivars is clearly not absolute. A few late summer flowers may be often seen on new shoots within an established plant. In the coastal west of Scotland with a mild, maritime climate at above 55°N (a latitude roughly equivalent to much colder Moscow, Novosibirsk, southern Hudson Bay and the Alaska Panhandle) Anne Greenall finds that some non-remontant cultivars including ‘Blaumeise’ will flower properly late in autumn despite having been pruned to near the base the previous winter (pers. comm. 2023). Perhaps something about the long growing season combined with high-latitude rate of daylength change allows what seems not to happen in Georgia, USA (Dirr 2021), a warm but southerly area. Constant moisture and nutrient availability certainly help (Foster 2023).

Tim Boebel, gardening in the USDA Zone 6 Finger Lakes region of northern New York, USA, where shoots are routinely killed to near the base in winter, has catalogued the flowering performance of many cultivars as well as developing a method to encourage cold-climate flowering (Boebel 2011, 2018). His regime aims to produce buds which will go on to flower as close to the ground as possible, where they can be protected over winter, and to avoid the overcrowded thicket of weak stems which often follows winter dieback; it involves carefully timed pruning, feeding and mulching. There are five elements: 1) cut out about one third of non-blooming stems, leaving 5–20 cm, to prevent overcroding next year; 2) just before the first hard frost is expected shear away the upper third of the clump leaving only what can be mulched using the cut-away material; 3) before temperatures fall below –12°C mulch the entire clump with leaves or evergreen branches, the higher the better, and cover the entire mound in fine bark; 4) around mid-April (in his own area) remove the mulch, taking care not to break off buds; 5) apply high nitrogen fertilizer at intervals between the last frost and early August (Boebel 2011).

In milder areas pruning is not necessary; anything beyond cutting out a few of the very oldest stems at the base in late winter to allow room for strong new growth, or cutting out winter-killed wood in spring, is usually unhelpful. Old, dry mophead inflorescences left in place over winter will give a degree of frost protection to the buds. To prevent seeding, dead lacecap flowers may be cut off just above the second pair of leaves from the top (Haworth-Booth 1984). Where blue flowers are the aim, it is wise to leave any prunings under the bush to decompose in situ, letting the aluminium accumulated in them return to the soil (Foster 2023).

A quite different frost issue can affect garden plants and nursery stock in maritime areas such as southwest England. Mild winters followed by ‘start-stop’ springs lead to early bud break. A late frost of only –5°C may then destroy the soft young growth and all the year’s flowers (pers. obs.).

The blue / pink colour conundrum in Hydrangea is centred on H. macrophylla; some other species are changeable, but few to the same extent. It is well known – if not always understood – among gardeners, and spills out into wider culture. For example Rilke’s poems Blaue Hortensie and Rosa Hortensie, written 1906/7 in France, the centre of early enthusiasm for potted hydrangeas, muse on the origins, mutability and perception of colour (Rilke 2011). Aluminium availability is the key (Chalker-Scott 2018 reviews literature and cultural advice). Flowers tend toward blue when available aluminium is abundant (typically on acidic soils, or with supplements), pink when at low levels (as on alkaline soil, where aluminium becomes bound to clay particles). Soil pH itself is not the cause: acidic but aluminium deficient soils grow pink hydrangeas. Nor is iron involved as many gardeners believe: pH certainly affects iron availability, but this is not the mechanism here, and treatment with chelated iron (let alone rusty nails as sometimes suggested!) will not affect colour. Most unusually, Hydrangea macrophylla actively accumulates aluminium which is generally toxic to plants at higher levels, perhaps part of a defence mechanism. The same anthocyanin pigment gives blue and pink, blue if part of a complex with aluminium and a phenolic co-pigment, otherwise pink. Some cultivars blue readily, some never get more than purplish. Some vary from deep blue to deep pink, some change much less. The biology underlying this variation is not fully understood, but may involve pH within the pigment-bearing cells themselves (independent of soil pH): high rather than low intracellular pH may favour complex formation and hence blueing. With potted plants and nursery stock it is relatively easy to manipulate soil pH and hence flower colour, adding lime for pink or using an acidic medium for blue. Blueing can be enhanced by aluminium supplements, usually as potassium aluminium sulphate (alum). In the open garden, topdressing with limestone chips can be use to intensify reds where they appear ‘muddy’, but cautiously to avoid lime-induced chlorosis. For blues, however, aluminium supplements in the open ground might be less than beneficial to the health of other plants and soil organisms. Hydrangea enthusiasts use the word blue as a verb, especially in the form ‘blueing’ to denote a cultivar’s tendency to develop blue flowers in suitable conditions. We follow this usage here.

Variation in hardiness among the cultivars makes it hard to generalize about where they can be grown, and more importantly where they will flower. Climate change and the advent of remontancy have changed the image of what were once considered quite tender plants. Near the limits, it is interesting to note that the only H. macrophylla-based cultivar in the catalogue of Gothenburg Botanical Garden, Sweden, is the remontant ‘Bailmer’ (ENDLESS SUMMER®) (Gothenburg Botanical Garden 2024). Light shade with consistent moisture availability is ideal, although some cultivars tolerate full sun and dry spells well. Flower colour is usually more intense with brighter light; green notes are accentuated in shade, while most whites will scorch in full sun, especially at lower latitudes.

Many gardens have excellent displays of cultivars; few have comprehensive collections which allow thorough comparisons. One of the Western world’s best and most complete is at Jardin Shamrock, Varengeville-sur-Mer, on the northern French coast (Collection Mondiale d’Hydrangea Shamrock 2024). Thorough comparative trials seem to have been strangely few. In Southern England a Royal Horticultural Society trial in 2015–2018, with 172 entries planted across two sites resulted in 25 Award of Garden Merit selections (Royal Horticultural Society 2018), although the degree of variation in conditions across the sites may have limited its value (Foster 2023).

The alphabetical articles, ‘Hydrangea Section Macrophyllae Cultivars A’ etc, include pure H. macrophylla, pure H. serrata, known hybrids between them, and the great mass whose exact genetic heritage is unclear; anything else would introduce guesswork and controversy. We have tried to include cultivars on the basis of proven garden value, historical significance, unique features or importance in today’s market. So many new cultivars enter the potted plant market, only to be swept away by their successors, that it would be foolish to describe too many. This is not to denigrate what are often impressive cultivars, the serious breeding programmes behind them, or their intended use; they are simply marginal to our focus on trees and shrubs for the open garden. We will continue to monitor new selections for garden merit and periodically add worthwhile cultivars.