Primary tabs



Trees, shrubs, lianas, woody epiphytes or (extra-Mal.) more rarely herbs. Leaves spiral or rarely opposite or in whorls; Inflorescence terminal or more rarely lateral; Flowers hermaphrodite or heterosexual, sometimes dioecious; Petals 3 to numerous, often 5, sometimes fused into a calyptra, or forming a tube with spreading lobes (Osmoxylon), valvate or imbricate in bud, usually with a broad base but rarely narrowed below. Stamens usually as many as the petals and alternating with them, or twice as many, or indefinite; Ovary inferior, half inferior, or very rarely (extra-Mal.) superior, 1- to many-celled, the top of the ovary usually a fleshy disk; Ovules solitary, pendulous, anatropous, with the raphe ventral. Fruit baccate or drupaceous, exocarp usually fleshy, endocarp forming cartilaginous or membranaceous pyrenes around the seeds. Seeds one per pyrene, with a small embryo within smooth or ruminate endosperm.


Africa: present, Seychelles (Seychellespresent) African mainland: present America: present Americas: present Asia: present Asia-Temperate:, Taiwan (Taiwanpresent) Asia-Tropical:, Borneopresent; Jawa (Jawapresent); Laos (Laospresent); Malayapresent (Peninsular Malaysiapresent); Maluku (Malukupresent); New Guineapresent; Philippines (Philippinespresent); Sri Lanka (Sri Lankapresent); Sulawesi (Sulawesipresent); Sumatera (Sumaterapresent); Thailand (Thailandpresent) Asian mainland: present Australasia:, Queensland (Queenslandpresent) Botel Tobago: present Canary Islands: present E. Asia: present East Africa: absent East Asia: present East Java: present Lesser Sunda Is: present Madagascar: present Mascarenes: present Melanesia: present Melanesian and East Malesian: present NE. Australia: present New Caledonia: present New Hebrides: present New Zealand: present New Zealand region: present Nicobar Islands: present North America: present Northern America: Oceania: present Old World tropics: present Pacific:, Fiji (Fijipresent); Marianaspresent; Samoa (Samoapresent) Pantropical: present SE. Asia: present Society Is: present Solomon Is: present Timor: present West Malesia: present West New Guinea: present mainland Africa: present mainland Asia: present neotropics: present tropical America: present western Malesia: present
About 50 genera with a roughly estimated 1150 species, ranging mainly in the warmer parts of both hemispheres (especially in montane zones), a small number in or extending to cool-temperate regions. With the exception of SE. Asia, the family and its centres of distribution are largely found within the land masses derived from ancient Gondwanaland. In Malesia 17 genera with a total (excluding Schefflera) of 117 species in 16 genera. (The largest genus, Schefflera, with an estimated 250 species for the region, is omitted from this account.)

Three genera are endemic to Malesia (or nearly so). One of these, Anakasia (related to Polyscias) is of very local distribution in West New Guinea; the two others, Harmsiopanax (Java, Lesser Sunda Is., Celebes, to New Guinea) and Aralidium (Malay Peninsula, Sumatra and Borneo, with an extension to Thailand) are more widely distributed.

Five other genera have their main centres of distribution within Malesia. Two of these extend further eastwards: Mackinlaya ranges from the Philippines and Celebes through New Guinea to the Solomon Is. and NE. Australia; Osmoxylon occurs from Botel Tobago (Taiwan) and the Marianas to Borneo, the Philippines, Celebes, the Moluccas, New Guinea to the Solomon Is. and the New Hebrides. A third, Trevesia, is confined to western Malesia (Lesser Sunda Is., Java, Borneo, Sumatra, and Malaya), with an extension into the Asian mainland. Wider ranges are recorded for Gastonia (widespread in Malesia to the Solomon Is.) with an additional range in the Seychelles, Mascarenes, and Madagascar (but not East Africa as previously reported) and Arthrophyllum (all over Malesia with extensions to Thailand, Laos, the Nicobar Islands, and New Caledonia.

A significant element in the Malesian representation of the family consists of six genera which occur mainly in SE.-E. Asia, two of which further extending to the Americas, viz Aralia (in America south to Mexico) throughout Malesia, and Dendropanax (tropical America) in West Malesia (Sumatra, Malay Peninsula, Borneo). The remaining four genera range in Malesia as follows: Macropanax and Brassaiopsis in Malay Peninsula, Sumatra and Java, Acanthopanax in Malay Peninsula, Sumatra and the Philippines, and Pentapanax only in East Java.

Only one genus, Delarbrea, has its main centre of distribution to the east of Malesia (Melanesia, mostly New Caledonia) and Queensland, and extends through New Guinea to the Moluccas and Lesser Sunda Is. as far west as Timor.

Two other genera have wider distributions. Schefflera, including several segregate genera, is pantropical although with but few representatives (13) on the African mainland. However, the several sections of Schefflera as redefined by Frodin (1970, 1975) often show distinctive regional distributions paralleling those of several of the other genera. The type section ranges from the New Hebrides to Samoa and New Zealand, including Fiji and New Caledonia.

Polyscias is widely dispersed in the Old World tropics from the African mainland eastwards to the Society Is. and Australia, but with only a few species in SE. Asia and western Malesia. As in Schefflera, the diverse series of species comprising the genus show distinctive regional distribution patterns, with the type series being mainly Melanesian and East Malesian (in West Malesia it is only cultivated or adventive).

As said above several Malesian genera extend into, or have their main centres in, mainland Asia, but only Polyscias (in Sri Lanka) and Schefflera are shared with Africa and the same genera (with Gastonia) with Madagascar. Only Polyscias and Gastonia occur in the Mascarene Islands and Gastonia and Schefflera in the Seychelles. Two other genera are restricted to mainland Africa. Hedera of temperate Eurasia extends to the Canary Islands. The Americas have two (or three) distinctive endemic genera with recognizable affinities, as well as disjunct groups of the Australasian Pseudopanax, the Asiatic Pentapanax and Dendropanax and (in North America) Oplopanax, Aralia and Panax, all very closely related to those in East Asia; there are also distinctive sections of Schefflera in the neotropics which are nearer those in Africa than in Asia. Many distinctive endemic genera (or parts of the larger genera) occur in Oceania, the New Zealand region and Australia, some of them taxonomically very isolated.

The ratio of species: genus is here estimated as 23: 1, but if the very large genus Schefflera is not accounted for this reduces to 13:1. Some 30 genera have five or fewer species; in view of the considerable insular endemism at generic or infrageneric level, weak intercontinental links, and great distinctiveness of many genera (even though small), the family is surely of great antiquity, although much of the available palaeobotanical evidence requires re-evaluation (cf. Dilcher & Dolph, 1970). Many fossils previously referred to Oreopanax must now be placed in Platanaceae (Doyle, pers. comm.). Dendropanax has reliably been reported from Tertiary deposits in both Europe and North America where it is now absent (Dilcher & Dolph, l.c.).


Most erect Araliaceae are sparingly branched or even unbranched; their limbs are upright and the often massive, easily broken twigs generally bear rosettes of big, long- stalked leaves which leave large scars when they fall. Few form a true crown (Arthrophyllum, Gastonia, some species of Schefflera and Polyscias), while the others often look like elongated and grotesque shrubs with a candelabrum-like framework. Branches in most Araliaceae are entirely orthotropic; generally speaking, the plants would fit into the so-called Holttum, Corner, Tomlin- son, Chamberlain and Leeuwenberg models of Halle c.s. (1970, 1978), or their intermediates.

Aralia scandens (Merr.) Ha is a true climber. Epiphytic species only occur in the genus Schefflera; it is not yet recorded that any of them may appear to turn into a hemi-epiphytic habit.

Hedera (not native in Malesia) is almost the only genus with differentiated shoots and marked vegetative dimorphism.

A preliminary account of shoot-morphology in the family has been given by Philipson (1978), but much further field work is required in this area. Vegetative buds may be either proleptic or sylleptic; in the resting phase such buds may be covered by the clasping bases of foliage leaves (Osmoxylon, Schefflera) or (more rarely) specialized cataphylls (Acanthopanax). In both types of bud the primordia and young leaves may be covered by exudations of resinous slime.

There is so far little recorded evidence of the changes in leaf shape and configuration during the somatic phase of the life cycle known in many araliads in othei parts of the world; however, recent observations in New Guinea made by Frodin suggest that heteroblastism does occur, although its manifestation is not constant for a given species. Distinctive juvenile and intermediate foliage has been found in Schefflera eriocephala Harms and (to a lesser extent) in S. stolleana Harms. Other examples are seen in Brassaiopsis, Trevesia and Schefflera subg. Agalma (S. aroma- tica (Bl.) Harms; S. nervosa (King) Vig.); juvenile leaves of some of these are preserved in Herbarium Bogoriense. In Harmsiopanax, the configuration of the leaves changes abruptly just below the inflorescence, a phenomenon paralleled in some other genera although less dramatically. In Mackinlaya celebica (Harms) Philipson and M. schlechteri (Harms) Philipson, leaf polymorphism is very marked with the result that in the past several 'paper species' have been described on too limited a range of material; in this revision many reductions have been made.


General accounts of the vegetative anatomy of the ivy family are given by Gussow (1900), Viguier (1906, 1909) and Metcalfe & Chalk (1950). Secretory canals are characteristic of the stems and leaves, but are absent from Aralidium. A comparison of the wood anatomy of Araliaceae and Cornaceae is made by Philipson (1967), and an extensive account of vegetative anatomy in the context of woody Umbellales was provided by Rodriguez (1957, 1971). The xylem of the former family is characterized by fibres with small, simple pits, and the presence of scalariform and reticulate perforation plates in the vessel elements (as opposed to simple perforations) is thought to be a less advanced feature. Recent special reports bearing on taxonomy include: on stomatal development (Inamdar c.s., 1969); on sievetube plastids (Behnke, 1972), and on epidermal papillae (Bui, 1974).

Floral anatomy is discussed by Baumann-Bodenheim (1955), Philipson (1967, 1970) and especially by Eyde & Tseng (1971). Embryological characteristics of the family have been reviewed by Davis (1966) as well as by Rao (1972). The single pendulous anatropous ovule has the funiculus and ovular vascular bundle axial (Philipson, 1970). The embryo is small in a mass of endosperm (Martin, 1946; Grushvitzky, 1967).


The Araliaceae are on all grounds closely connected with the Umbelliferae, a very large but mainly temperate and tropical-montane group comprising mostly aromatic herbs with a restricted floral scheme. It seems likely that the ancestors of the ivy family were the woody tropical stock from which the herbaceous Umbelliferae evolved under the rigours of cooler climates (Corner, 1940) with the arborescent Heteromorpha of upland Africa and Myrrhidendron in Central and South America, both in the subfamily Apioideae, perhaps representing relics of the transition on account of the presence of a number of attributes primitive for the family (Rodriguez, 1957); on the other hand, the umbellifers may have originated as megaherbs on tropical mountains (Philipson, 1978). A few Araliaceous genera exhibit some features characteristic of Umbelliferae, such as Harmsiopanax, Mackinlaya, and especially Myodocarpus (from New Caledonia) and Stilbocarpa (from southern New Zealand and associated 'subantarctic' islands).

This led Harms (1898) to show three different lines leading from Araliaceae to Umbelliferae, suggesting that considerable overlap between the families existed; more recently Rodriguez (1971) again called attention to this phenomenon and Thorne (1968, 1973) has gone further by merging Umbelliferae into Araliaceae (a step also advocated by Hallier f. in 1905) and assuming the three subfamilies of the former to have arisen separately, perhaps in different parts of the world at different times, from proto-Araliaceous ancestors. However both he and Rodriguez have concluded in agreement with Corner (l.c.) and Baumann-Bodenheim (1946) that the Araliaceae sensu stricto, "because of their greater evolutionary breadth and their retention of many more primitive features, would seem closest to the proto-araliad stock" (Thorne, 1973).

Nevertheless, the concept of Araliaceae as a separate natural family appears to serve a useful purpose and has been retained for this Flora. An isolated, doubtfully included genus is the West Malesian Aralidium; it is the only genus lacking resin ducts but it would be equally anomalous in the Cornaceae to which it has also been referred. It shows some resemblance with the New Zealand genus Griselinia, usually relegated to the Cornaceae, be it as a marginal member (Philipson, 1967).

Subdivision. Since the first significant family monograph by Seemann (1868), several systems have been proposed which, taken together, are notable for their lack of consistency. This results from a lack of agreement on the relative importance of the comparatively minor structural and gross anatomical features of reproductive parts traditionally used and by conflicting claims on the relative 'antiquity' of polymery versus pentamery. Until recently, there has been for a priori reasons (cf. Eyde, 1975) little recognition of the potential value of vegetative features and their acceptance as valid evidence for a system; and information from wood anatomy, floral histology, palynology, karyology, phytochemistry, and other areas is only beginning to be utilized. While woodiness is generally accepted as a primitive feature in Araliaceae, the impact of the work of Corner and others on tree structure and growth rhythms (summarized in Halle, Oldeman & Tomlinson, 1978; see also Borchert, 1969; Hladik, 1970; Philipson, 1978) has still to be fully assimilated. Much more work is also required on inflorescences, although Frodin (1970), Philipson (1970b) and others have made a beginning. The monothetic interpretation of most attribute states usual in systems of the family was first challenged by Baumann-Bodenheim (l.c.) who considered that phyletic changes could have taken place in parallel; and Eyde & Tseng (1969) showed that at least some supposedly unidirectional sequences were reversible. This has tended to reduce the supposed importance of many of the traditional attributes, with a consequent reduction in the number of genera.

Sufficient evidence is not yet available, however, for the construction of a more balanced, polythetically based system of the family, and the long-standing subdivision proposed by Harms (1898) into three tribes, Schefflereae, Aralieae, and Mackinlayeae, based monothetically on the structure and aestivation of the perianth, is retained. In spite of its now recognized imperfections, it is more valid than the systems of Viguier (1906) and Hutchinson (1967) and remains the most widely accepted. Some steps towards the formulation of a new system on polythetic principles have been taken by Eyde & Tseng (1971: 221) who make a fundamental distinction based on basic leaf-organization (either pinnate or palmate) and recognize the heterogeneity of Harms' Aralieae; however, this system was deliberately not completely developed.


Lists of chromosome numbers for members of Araliaceae are given by Darlington & Wylie (1955), Sharma & Chatterji (1964) and Bolkovskikh c.s. (1969). The family shows considerable constancy of the basic number of x = 12, although x = 11 has been recorded for one non-Malesian group of Schefflera. The implications of chromosome data on the wider relationships of the Araliaceae are discussed by Moore (1971).


A variety of minor local uses are reported by Burkill (1966), Heyne (1927), and Ochse & Bakhuizen van den Brink (1931) as well as in the notes under individual species in this Flora. The only species that form articles of trade are the taxa of Polyscias sect. Polyscias popularly grown as foliage and hedge plants; however, many other species in a variety of genera are of actual or potential ornamental worth, with Schefflera actinophylla (Endl.) Harms and S. longifolia (Bl.) Vig. being particularly widely used. In Papua New Guinea, Gastonia spectabilis is cut for timber and the wood used for light carpentry, boxes, etc.
Monkeys are fond of the flush of some aromatic species of Schefflera, as observed in West Java and North Sumatra.


Nomina nuda and invalidly published names have only been quoted if they have been cited in Index Kewensis.


Information on the chemistry of the Araliaceae should be sought in Hegnauer (1964, 1978) where references to original sources are given. The family is characterized by the occurrence of essential oils and resins in canals and by the presence of polyacetylenic compounds (especially falcarinone-type), triterpenic sapogenins of the oleanene-, ursene- and dammarene-types, seed oils with petroselinic acid, and by the absence of true tannins. The chemistry of the family fully confirms its close relationship with the Umbelliferae and also more distantly with the Pittosporaceae and the Compositae (Hegnauer, 1969, 1971; Bohlmann, 1971). The rareness of flavones and the predominance of flavonols in Araliaceae suggests closer relationship with two of the three subfamilies of the Umbelliferae, flavones having not yet been found in Hydrocotyloideae and Saniculoideae (Harborne, 1971). The absence of iridoid substances and true tannins and the presence of polyenes, petroselinic acid and isoprenylated coumarins in the Umbellales contrasts with the Cornales (Hegnauer, 1969; Jensen c.s., 1975), and this led to a suggestion that the Umbellales and Cornales had to be separated (Bate-Smith c.s., 1975), an argument with increasing support from other lines of inquiry. However, insufficient evidence appears to be yet available for the detection of possible lines of relationship within the Araliaceae.