These are the Buddleja (Buddleia) species that I have some experience of growing, some are fully hardy and some benefit from a little winter protection. They are from China and the Himalayas, Africa and South America. All African and Asian Buddlejas are hermaphrodite (male and female in the same flower), but many Buddlejas from the Americas are dioecious (separate male and female plants). There is some confusion because the flowers appear to have both male (anthers) and female (pistil) parts but one or the other is non-functional. This is called cryptic dioecy and species that display this are described as being cryptically dioecious; in older texts it can be described as micro-dioecious.

I have found many Buddleja species currently being sold are misidentified and therefore mislabelled. Wherever possible, I have attempted to confirm, correct or comment on the labelling using the monograph of Leeuwenberg (1979).

See also:
A Comprehensive and Organised List of Buddleja Species
(with Wikipedia links)

Buddleja albiflora

Superficially, a species very like B. davidii, and quite closely related according to the latest DNA data. Despite the name, the flowers can be pale to medium lilac-pink. The leaves are large and pendulous, and the stems are usually round, rather than tetragonal (squared-off), often being pigmented and red/purple in colour. It is hardy, deciduous, and flowers slightly earlier (end of June onwards) than B. davidii, but can treated in the same way for pruning.

Buddleja alternifolia

Unlike virtually all Buddleja where the leaves are opposite, this one has alternate leaves. Sometimes called the Weeping Buddleja as it does resemble a willow. The arching stems of the previous years growth become covered in lilac flowers in late spring (May). These are in a spiral pattern all along the stem and not in a panicle or raceceme. It's deciduous so it is reduced to a bare skeleton in the winter but is really hardy, some claim even more hardy than B.davidii. Eventually it will grow into a small tree up to 4 metres across and 3 metres high, but can be pruned and shaped straight after flowering. Any later and it may not flower the next year.

Buddleja alternifolia 'Argentea'

Like the species but this cultivar has silvery-grey leaves (hence 'Argentea'). The flowers are a darker purple with reflexed petals and the scent is cleaner and stronger. It grows much slower than the species which is the only downside.

Buddleja aff. alternifolia 'Tsestangensis'

Buddleja tsestangensis, to give it the original name, is a white flowered form of B.alternifolia with grey, toothed leaves. It is also known under the name B. wardii KR4881 as some believe it may be a hybrid with B. crispa. It flowers slightly earlier than other Buddleja alternifolia and isn't fully deciduous in mild winters.

Buddleja caryopteridifolia (spurious)

Not the easiest name, but this looks a viable alternative to the finicky B.crispa. Fully hardy but deciduous, it has greyish foliage and purple flowers. The scent is a bit like B.crispa, so very pleasant. It flowers late spring (May) and can flower again in the late summer or autumn.

Up close the individual flowers resemble B. alternifolia 'Argentea' and occassionally branches have alternate leaves. The cultivar shown could be a B. alternifolia and B.crispa cross, possibly arising out of a natural hybridisation sometime in the past. I suspect it is in fact the hybrid B. wardii, but it is currently being sold as B. caryopteridifolia.

The authentic B. caryopteridifolia has been described as a variety of B. crispa. There is more on this and other B. crispa varieties and hybrids HERE.

Buddleja colvilei

An unusual Himalayan Buddleja with the largeset flowers in the genus, each one about the size of a wiegela flower (about 1.5cm diameter) but with four petals. All types form a large shrub or small tree and have large leaves. B.colvilei gets hardier with age, but needs a sheltered spot to thrive. Young plants need extra protection over winter until they are fully established.

There are quite a few cultivars around now, mostly from wild-collected seeds, in shades of red and pink. The cultivar shown left has pink flowers with a white throat. "Tregye" (below) has orange centres to the flowers and is particularly attractive.

Buddleja colvilei 'Kewensis'

The cultivar 'Kewensis', selected at Kew Gardens, is the most popular cultivar amongst the dozen or so available. It has dark red flowers and deeper green foliage - one of the finest Buddlejas you can grow and hardier than the pink form, but it still needs a sheltered spot. Cover with fleece in the winter to protect the buds from frost as it is spring flowering, anytime between April and June depending on conditions.

Buddleja crispa

Pretty pink flowers and soft grey foliage, the leaves are almost heart-shaped. This is an attractive Buddleja that can grow quite large and woody. It is fussy though and needs a very warm south-facing wall to survive. It also can't take any water-logging over winter and prefers dry conditions in summer to enhance the silvery foliage.

I have grown mine in a large pot so it can be kept under cover all winter, although it is still fully deciduous. Flowers in June-July and you can get a second flush in the autumn of you dead-head.

There is more on this and other B. crispa varieties and hybrids HERE.

Buddleja delavayi

Buddleja delavayi is a deciduous spring-flowerng species, which comes from south-western China. The flower buds are held dormant over winter on bare stems, and are both terminal and axiliary. The flowers are soft pink with a heady rose-like scent, and similar to other short-day Buddleja like B. officinalis and B.paniculata, but the leaves are much smaller than these other species. Some claim it can re-bloom in the autumn, although I've never managed to make that happen. Possibly root-hardy, but the flower-buds are easily frosted-off in all but the most sheltered situations. I grow in a large pot and drag it into the greenhouse for the winter.

The plant shown is usually sold as B. heliophila, which is synonymous with B. delavayi.

Buddleja fallowiana

One of the species more closely related to B. davidii, B. fallowiana has felted grey leaves and white felted stems, too. The species can have white, lilac, lavender or pale blue flowers. Not as hardy as B. davidii, but should survive a moderate UK winter in a reasonable sheltered garden even though it will generally lose its leaves. Winter water-logging is lethal.

This fine pale blue cultivar is P&M181, collected from a pine forest edge in Lijiang, north-west Yunnan at a altitude of 3170 metres.

Buddleja fallowiana ‘Alba’

The white, or alba, form is more commonly grown. This white form also tends to be hardier, though nowhere near as tough as B. davidii, and it usually still loses its leaves in winter.

Buddleja forrestii

B. forrestii is a highly variable Himalayan species and is closely allied with B. macrostachya. There are quite a few individual cultvars of these species and their origins are often unknown. In general, both are large shrubs to small trees with long panicles of white, lilac or purple flowers. The foliage also tends to be large and dark green, the stems tetragonal and winged,particularly in B.forrestii.

My plant is from Roseland House Garden and Nursery in Cornwall. So far it seems fairly hardy and kept its leaves over winter. A large and robust shrub, it flowers July and has off-white flowers with a mauve flush in the bud.

Many species were sunk into B. forrestii including B. pterocaulis (also called B. longifolia Gagnep.) and B. limitanea (see below), which has caused some confusion. Some plants sold as B. pterocaulis longifolia, which is not a legitimate name, are a form of B. forrestii, and some are B. salviifolia Alba (see below), a complete mislabelling.

B. macrostachya is easily mistaken for B. forrestii, as they are so similar. Only close examination of the flowers can distinguish them (see below).

Selected pages from A.J.M. Leeuwenberg (1979) The Loganiaceae of Africa XVIII: Buddleja - Revision of the African and Asian Species: B. macrostachya and B. forrestii.

Buddleja forrestii type limitanea

A summer flowering Chinese Buddleja which was once a species in its own right, but is now considered a smaller type of B. forrestii. However, it is still frequently labelled as B. limitanea in horticulture. Typically, it has pendulous panicles of mauve-purple to white flowers that appear in mid- to late-summer; the individual flowers are larger than B. davidii and have the typical short, conical corolla tube of B. forrestii . The leaves are usually smaller than the larger forms of B. forrestii, and often slightly grey-felted. Less vigorous than other forms, it grows to about three metres height and spread. It’s quite hardy, but can be cut back hard in cold winters. If pruning is required, and it can become quite straggly, it should be done in early spring like B. davidii.

There are several different cultivars (or ‘collections’) around the UK, such as the one shown left (BO-14-138) with pretty pink flowers, which was grown from seed collected in Yunnan, China. Another is Buddleja sp. ACE2522, collected in Yunnan by the Alpine Garden Society. There is a cultivar of unknown origin in the National Collection at the Longstock Nursery.

The plant with mauve flowers shown lower left and below in close up is a mystery. Labelled as B. limitanea, and looking exactly the same in habit and leaf as limitanea, the flowers are rather atypical in having longer, cylindrical corolla tubes and slightly tomentose (felted) parts. Its origins are unknown, and I suspect it is either a natural hybrid or one raised ex situ. It certainly looks to be descended from B. forrestii, and probably from the limitanea type; what other species is involved is more ambiguous, but B. davidii is possible. I have speculated it may be the hybrid Buddleja X ‘Hotblackiana’ (B. forrestii X B. davidii) currently thought lost to horticulture, although I am unable to confirm this.

Buddleja globosa

A South American Buddleja from Chile and Argentina. It is a large shrub with glossy green leaves and intriguing golf-ball flowers about 15mm in diameter. It is probably the most popular garden Buddleja after B. davidii. Many cultivars are listed: Cally Orange; Cannington Gold; Lemon Ball; to name three. There is little difference except HCM98017 decribed below. There is also a variegated form with yellow spotted leaves. It is completely hardy in the UK and more or less evergreen in most of the UK. Flowers May to June and the bees love it.

B. globosa HCM98017 has very pale yellow flowers and smaller leaves.

B. araucana, formerly called B.nappii, is a closely related species with similar orange flowers but has brown down on stems and the underside of its narrower leaves. It is relatively uncommon in cultivation and it takes an expert to tell from B. globosa. In fact it has previously been known as B. globosa var. araucana and not treated as a separate species at all.

Both B. araucana and B. globosa are cyptically dioecious (see above) like the majority of New World Buddlejas, although the flowers appear to be hermaphrodite. The plants that are in cultivation have not been sexed.

There is much more about these two species on a separate page HERE.

Buddleja glomerata

From the semi-deserts of South Africa whereit is known as Karoo Sagewood. Buddleja glomerata has puckered silver foliage and is usually evergreen. It has the tiniest fluffy yellow flowers in large plumes with a perfume of reminiscent of cloves; it usually blooms around May under cover, but flowering is unreliable when grown in the open. Although it can tolerate cold, it is very unhappy in damp conditions.

Some consider it belongs in a separate genus, Chilianthus, along with Buddleja saligna (the False or Bastard Olive) and B. dysophylla, species in which the anthers protrude beyond the corolla tube on long filaments unlike in most other Buddleja. In contrast to B.saligna, the green stigmas of B. glomerata are also held proud of the corolla.

Sometimes it sold as Buddleja 'Silver Service'.

Buddleja japonica

B.japonica, as the name suggests, is native to Japan. The plant is hardy; although deciduous and tending to die back in winter, it grows again strongly from the base the following spring.

A member of the Curvifloræ series (group) of Buddlejas (those with long, curved corolla tubes), which includes B.lindleyana and B.curviflora, it can be distinguished from other species by the lighter-green, toothed foliage and the winged-stems. In common with many of the Curvifloræ Buddlejas, the flowers are purple, although the panicle is not one uniform shade, with individual flowers ranging from a deep colour to very pale.

Buddleja lindleyana

Summer flowering Chinese Buddleja with pretty purple tubular flowers in pendulous panicles but they have absolutely no scent. Glossy small green leaves on a shrub that can grow into a small tree eventually. This species also suckers so can spread out if not kept in check. Pruning, where necesssry, should be in spring. More or less hardy but a sheltered position is better, often grown against a wall. It can be evergreen or lose its leaves in a hard winter.

Improvements have been made on the species, with a number of smaller cultivars developed. One of the best is Peter Moore's "Little Treasure", a plant not much more than a metre high with small glossy leaves and more upright flower panicles.

Buddleja loricata

From South Africa and sometimes called Mountain Sagewood. It is a smallish shrub has rough strap-like leaves with a silvery underside. It is fully hardy, evergreen and doesn't even mind snow. It is grown mainly for its foliage but has interesting white flowers that come in June or early July. Pruning is optional but should probably be done straight after flowering; alternatively it can just be trimmed to be kept tidy

Buddleja macrostachya

A very variable species with a wide range across the Himalayan region and south-east Asia; the flowers can be anything from white through to wine-red and the flowering time can vary from mid-summer to Christmas. Until it flowers this species can be virtually indistinguishable from B. forrestii (see above for more details). There are several different plants available - the one shown left is PAB4198 from Shillong, (Meghalaya India) with pink flowers. This collection flowers very late in the season, about November, meaning a mild autumn is required for it to bloom properly. In my garden, it's forming a small tree and is over four metres in height.

B. macrostachya can closely resemble B. forrestii, such as in the example shown left (RF040 from Arunachal Pradesh, India and unfortunately misidentifed as a B. forrestii). This collection forms a large semi-deciduous shrub, with large leaves and rounded branchlets, both covered in an unusual coppery-brown tomentum. Unlike the plant above, RF040 flowers in summer (around early August). The panicles are long, pendulous and many flowered, and loved by the bees. The colour is cream, but can appear slightly pale-pink at certain times. This has proved quite hardy (down to at least -5ºC) once established.

There are several other collections worth noting. SBEC360 was collected by Roy Lancaster in 1981 on the Sino British Expedition to Cangshan (SBEC), Yunnan, China. Allegedly, his personal specimen was so vigorous it had to be removed before it damaged his house.

GWJ9286 was collected in the Lachen Valley, Sikkim (northern India) by Crûg Farm Plants, North Wales. It's been labelled variously as B. paniculata, B. myriantha and species nova. I have identfied it as B. macrostachya, although GWJ9286 is atypical for the species, being smaller overall (about 2 metres tall) and with more delicate silvery foliage. However, the flowers themselves conform to botanical descriptions of B. macrostachya. Going back to C.V.B. Marquand's Revision of the Old World Species of Buddleja (1930 Bulletin of Miscellaneous Information No.5 pp177-208. Royal Botanic Gardens, Kew.), one previously recognised species was Buddleja hookeri, now synonymous with B. macrostachya . Marquand's description of B. hookeri (available here) is remarkably similar to GWJ9286 in many respects - small calyx, tomentose ovary and corolla tube, lax flower panicle and subterete (rounded) stems. Moreover, the original report cites the Lachen Valley, Sikkim (northern India) as the source of the B. hookeri holotype, the same location from which GWJ9286 was collected. Both the literature and my own investigations provide strong evidence for GWJ9286 being a form of B. macrostachya.

Buddleja macrostachya is most easily distinguished from the similar B. forrestii by dissecting the flowers. In the photo (left), a B. macrostachya RF040 flower has been cut open to reveal the pistil and ovary. Note that the ovary (and later the seed capsule) is covered in tangled hairs (stellate-tomentose) and how the base of the style is also hairy. The interior of the corolla tube has numerous hairs (villose) down to level with the ovary and the calyx is tomentose (covered in hairs). In comparison, the flower of B. forrestii does not have this tomentum: the ovary is glabrous or with very few hairs and the calyx is also glabrous; the interior of the corolla tubes usually has fewer hairs positioned more towards the throat. Another difference is how B. forrestii typically has stamens level with or just above the stigma, whereas in B. macrostachya, they are often in the throat of the flower and well above the stigma.

Due to the variabilty found in these two species gross features such as corolla tube length, panicle form, foliage, stem shape and growth habit cannot be reliably used to effect a correct identification.

Selected pages from A.J.M. Leeuwenberg (1979) The Loganiaceae of Africa XVIII: Buddleja - Revision of the African and Asian Species: B. macrostachya and B. forrestii.

Buddleja myriantha

Buddleja myriantha is also a very variable species, found in south-west China and northern Myanmar. The name is often misapplied in the horticultural trade to either to a form of B. curviflora or to the blue-flowered B.salviifolia. The authentic B.myriantha is related to both B.davidii and B.albiflora.

The plant shown upper left is H&M 1520 from Cangshan, Yunnan, China; lower left is W/O 8047 from the same location, which more closely resembles B.albiflora. In both collections, the individual flowers are tiny; the species gains its name from the huge number of these diminutive flowers in each panicle.

B.myriantha can be distinguished from B.albiflora as follows: B.albiflora has rounded or subterete stems, whereas B.myriantha has noticeably four-angled stems; the calyx and exterior of the corolla tube are smooth in B.albiflora, but in B.myriantha the calyx and outside of the corolla tube are distinctly tomentose (hairy), as can be seen in the close-up of W/O 8047 below.

See also the Illustration from Flora of China for a comparison.

Buddleja nivea

From China again, this one is distinguished by its heavily felted leaves and stems. The inflorescences are in long cats' tail panicles but the individual purple flowers are unremarkable. It is mostly grown for its large felted foliage. Fully deciduous but quite hardy and it should be pruned hard in spring to keep it a small shrub.

There is also a pink-flowered form of B. nivea. The flowers are not any more outstanding and, although a pleasant shade of pink, are individually smaller and the panicle more open. The plant is,however, more floriferous than other forms and repeat blooms if dead-headed.

Buddleja nivea type stenostachya

Another form of B. nivea, this used to be considered a separate species and is still sold as such. The leaves have smooth edges and the stem indumentum is less floccose, but otherwise the shrub is very similar to the species type. The inflorescences are a little different though, and it has very narrow panicles, for which it is named (stenostachya means narrow flower spikes). The individual flowers themselves are very similar to the species - purple, hairy throated and with a very short corolla tube. Possibly not as hardy as the other forms of B. nivea.

Buddleja nivea var. yunnanensis

Not to be confused with B. yunnanensis, a distinct and much smaller Asian species. Very like the species type in many ways, although it will grow much bigger. The leaves are longer, heavily felted and more conspicuously serrated. The purple flowers are not particularly remarkable and individually very much the same as the other types. Unlike others which flower sequentially up the panicle, the flowers open fairly randomly. Deciduous, vigorous and hardy.

There is the possibility that the Yunnan form of B. nivea has undergone a doubling of its chromosome number, from hexa- to dodecaploid (Moore 1960), and this could account for the size and vigour.
Ref: Moore R. 1960. Cytotaxonomic notes on Buddleja. American Journal of Botany 47: 511–517.

Buddleja officinalis

Unusually, this is a winter-flowering large shrub. Soft lilac-pink flowers open when the plant has few leaves. In summer it has large, long, soft and felted leaves, the stems also have brown felt so it is mostly grown as an attractive foliage plant. It loses most of its leaves for the winter. Mine flowered February, although this species can be in flower at Christmas provided it is in a particularly warm sunny position or kept under cover. Just hardy enough to grow in a sheltered spot, although the flower buds would be damaged by hard frost.

Buddleja officinalis is very similar to B. paniculata; the latter tends to have shorter leaves, which may be toothed.

Buddleja salviifolia

The leaves are very like sage, hence the name - called African Sagewood in its native South Africa. Flowers in late spring, April-May, and can be blue, lilac or off-white - mine is a fetching blue. Hardy and evergreen in southern UK, although further north it may need a sheltered spot.

It is sometimes sold mis-identified as B.myriantha, which is a totally different Asian Buddleja.

This Buddleja salviifolia Alba has unusual foliage somewhat different to the plant above. It has only tentatively been identified as a B. salviifolia Alba, although B. salviifolia is known to be a very variable species. The white flowers are similar in appearance to the blue cultivar and have a similar exquisite scent. They appear in spring at roughly the same time or a little earlier. Its hardiness is untested, but it's evergreen if kept under cover.

Buddleja speciosissima

Buddleja speciosissima is something very different indeed, a South American species with long, tubular orange-red flowers up to 2.5cm long, which emerge slowly from soft, furry buds. The attractive foliage is grey and felted. Unlike many species from the Americas, which are dioecious, B. speciosissima bears perfect (hermaphrodite) flowers and is self-fertile.

It comes from southern Brazil and has a very small range, restricted to Mount Itatiaia growing at an altiude of 2000-2500 metres. Despite the its tropical origins, it has evolved in harsh mountain conditions and proved hardy in southern UK gardens, where it forms a fairly low, scrubby shrub.

A closely related species with even longer flowers, B. longiflora, grows on the Serra Caparao in the same region.

Buddleja sterniana

Correctly described as a type of B.crispa and is perhaps more correctly called B.crispa type sterniana (B.farreri and B.agasthosma are also considered as B.crispa types). It is a spring flowering shrub with a lax habit. Leaves are grey felted and a little like B.crispa. Flowers are in early spring so are lost unless protected over the winter. In the recent harsh winters this has been cut right back and not flowered, but when it does the pale lilac flowers have a delicious scent.

There is more on this and other B. crispa varieties and hybrids HERE.