Cercis is a small genus of 7–10 species distributed in temperate forests across the northern hemisphere in four disjunct regions: the eastern United States to northeast Mexico, western North America, the Mediterranean region east to central Asia, and eastern Asia (Davis et al. 2002). Cercis species are unarmed, deciduous trees or shrubs. The leaves are simple, entire and may be heart-shaped or kidney-shaped; there is a long, distinct petiole. Stipules are caducous and may be small and scale-like or membranous. All Cercis species are dioecious and the flowers are produced directly on the surface of mature stems (cauliflory), prior to leaf expansion. The flowers are 5-merous and arranged in fascicles or clustered racemes, and are typically pink. The floral structure is similar to that in many other legumes, with a small uppermost petal (‘standard’) enclosed by two lateral ‘wing’ petals, subtended by two ‘keel’ petals. The 10 stamens are free, but somewhat enclosed by the keel petals. The leguminous fruit is flat, linear and acute at both ends; the fruits persist on the tree for many months, often well into winter, and can be very attractive in their own right. They are dehiscent, though slowly, shedding their many seeds over a long period (Robertson 1976, Allen & Allen 1981).
The Judas Tree, Cercis siliquastrum, has long been a popular small tree in the milder parts of Europe, and in the current period of warming temperatures is escaping from south-facing walls to more open sites in more northern areas. With the exception of the purple-leaved C. canadensis ‘Forest Pansy’, one of the most striking of foliage plants, the eastern Asian and North American species remain comparatively poorly known in northern Europe. The diversity of cultivars of C. canadensis in North America is remarkable and expanding, including many superior selections that should be sought out if planting this species (Dirr 1998, Roethling 2006). It is curious that the Western Redbud, C. occidentalis, has apparently not produced horticultural selections. Among the Asian species, C. chinensis has given rise in Auckland, New Zealand to the very striking cultivar ‘Avondale’, which has bright clear pink flowers borne in great abundance. This has been available for some time, but is currently (2007) being heavily promoted as small grafted specimens in garden centres across the United Kingdom, and is popular in the United States. The genus as a whole responds to summer heat, and all species should be given as warm and open a site as possible, although some (for example, C. canadensis, C. chinensis) can withstand very low winter temperatures if the wood is well ripened. Cercis are often short-lived – incurable canker and Verticillium wilt being important diseases – but as they grow rapidly, flower profusely and harmonise with most garden plants it is well worth giving them even a comparatively temporary space in the garden.
An important review of Cercis in cultivation was given by J.C. Raulston (Raulston 1986), after which, in the remaining decade of his life, he continued to build up a magnificent collection of species and cultivars at the arboretum that now bears his name, and contributed to another review of the taxa (Burns & Raulston 1994). Unfortunately a new building has since been placed in the area where the majority of the collection grew, but the arboretum still has a very wide and interesting assortment to study. The Asian taxa described here do require study, as they are a confusing group, and the fundamental similarities between all Cercis do not help. Neither are they well described in the literature or frequent in herbaria, although the recently available draft Flora of China account (Dezhao et al. 2008) should help clarify the situation. This and the admirable webpage on Cercis by Laurence J. Hatch (1998–2006) make it clear that the Chinese species fall into two groups: those with clusters of flowers, sometimes cauliflorous (C. chinensis, C. chingii), and those with racemose inflorescences (C. chuniana, C. glabra, C. racemosa). Cultivated plants of C. gigantea have clustered flowers, but this mysterious taxon is not treated in the draft Flora of China account.
All material of Asian Cercis needs reconsideration in the light of this new account, as it is clear that there are many mislabelled plants in circulation. The notes below reflect observations made on material cultivated under these names. The American taxa are equally muddled, and a really thorough modern analysis of the variation between populations is required. As Cercis species will hybridise in cultivation, seedlings from horticultural sources should be regarded with particular suspicion.
From Bean's Trees & Shrubs
The members of this genus, seven in number, constitute a very distinct and homogeneous group of hardy leguminous trees and shrubs whose resemblances to each other are as marked as are their differences from the other genera of the family Leguminosae. Their leaves furnish the most distinctive feature of the genus, being alternate, simple, entire, prominently five- or seven-nerved, broad and rounded, with a heart-shaped base, and from 2 to 6 in. long. The flowers in most of the species come in fasciculate clusters on wood one to many years old; but one Chinese species (C. racemosa) has them in racemes. The petals are nearly equal, but arranged somewhat after the fashion of a pea-shaped flower.
Few shrubs or small trees are more beautiful than the hardy species of Cercis at their best. They enjoy and merit generous conditions at the root, and succeed best in a deep, sandy loam, and should have as sunny a position as possible. Plants should be given a permanent position whilst still young, as the long, thick roots are liable to decay after the inevitable injury involved in transplanting old trees by ordinary means. Whatever transplanting is necessary should be done in May, and not until the expanding buds give some indication that active growth has recommenced. The most insidious enemy of these trees in my experience is the coral-spot fungus, for which drastic surgery is the only remedy; the affected branches should be cut back to undoubtedly healthy wood, and the wounds thoroughly covered with a protective dressing. The older and well-known species are propagated by seed, and this, of course, is preferable for all; but the newer species may be grafted on roots of C. siliquastrum or C. canadensis.