Primary tabs



Trees, scandent shrubs or woody climbers. Leaves alternate or spirally arranged, penninerved, simple or imparipinnate, the leaflets in the latter case opposite on often somewhat swollen nodes of the rachis; Flowers small, bisexual, rarely polygamo-dioecious, in terminal or axillary racemose panicles, or cymose: Sepals (3-)5, imbricate, free or ± connate at the base, equal or unequal. Petals (4-)5, mostly opposite the sepals (rarely alternate: Ophiocaryon spp., South America). Stamens (including staminodes) 5, opposite the petals, all polliniferous (Sabia) or only 2 inner ones opposite the reduced petals polliniferous and the other 3 staminodial. Ovary of 2(-3) carpels united to form a compound superior ovary, carpels very rarely free in the apical part, in that case tapering to 3 short styles with a capitate stigma; Fruit either 1-celled or 2-coccous, drupaceous or dry, indéhiscent;


Africa absent, Asia-Tropical, Australasia, Kashmir present, S. Deccan present, S. Japan present, Solomons present, neotropics present, tropical America present
Three genera: Sabia Indo-Malesian, from the S. Deccan and Kashmir to S. Japan, throughout Malesia as far as the Solomons; Meliosma with a similar range but also occurring in tropical America; Ophiocaryon in the Neotropics. The family is absent in Australia and Africa.


— Leaf anatomy. Hairs unicellular in Sabia; uniseriate nonglandular and capitate glandular in Meliosma. Stomata confined to the lower leaf surface, anomocytic or paracytic. Mesophyll dorsiventral, with arm palisade cells in Meliosma. Veins embedded in mesophyll and sheathed by sclerenchyma. Petiole in distal end with a closed vascular cylinder. Crystalliferous cells containing clusters common near the veins.

Young stems. Cork superficial. Cortex with stone cells in some species of Meliosma. Pericyclic sclerenchyma forming a composite, closed ring in Sabia, and composed of isolated fibre groups in Meliosma. Phloem with broad lignified rays in Sabia, and with non-lignified, dilatating (triangular) rays in Meliosma. Vessels with mixed simple and scalariform perforations in first formed xylem. Cluster crystals common in cortex, phloem, and pith. Secretory cells with unidentified contents noted in parenchyma of several Meliosma species.

Wood anatomy. Vessels exclusively solitary in Sabia, solitary and in radial multiples or small clusters in Meliosma; vessel perforations typically simple in Sabia; mixed simple and scalariform or exclusively scalariform to reticulate in Meliosma. Intervessel pits alternate. Vessel-ray and vessel-parenchyma pits simple, and often large. Fibres, usually thin-walled, with minutely bordered to simple pits, and mainly confined to the radial walls in Meliosma (libriform fibres); with distinctly bordered pits common in both the radial and tangential walls in Sabia; occasionally septate. Parenchyma scanty paratracheal to vasicentric with occasional lateral extensions in Meliosma, very sparse to almost absent in Sabia, usually in 8-celled strands. Rays sometimes of two different sizes, the broad ones 4-8(-15) cells wide in Meliosma, up to 20 cells wide in Sabia, usually over 2 mm high, heterogeneous (Kribs type II), often with sheath cells.

Taxonomic note based on vegetative anatomy. The above description is mainly based on early studies of a very limited number of species, so that the information is far too limited to serve in the discussion of infrageneric classification and delimitation. The two genera are anatomically quite distinct in their leaf and wood anatomy. Partly this is related to general anatomical differences between climbers (Sabia) and erect shrubs or trees (Meliosma). Thus, the anatomical evidence can be interpreted both in favour of the separation of Meliosma and Sabia into two families, or alternatively to retain their tribal position in the same family. Anatomically Sabia is quite distinct from the Menispermaceae to which it has been compared (see above, under taxonomy); affinity of Meliosma and Sabia with families of the Sapindales, especially A nacardiaceae seem to find more support in vegetative anatomy.
— P. BAAS.


There is no concensus of opinion on the affinity, hence the systematic position of Sabiaceae. Some even doubt whether Sabia and Meliosma are correctly placed in one family.

After the description of Sabia by COLEBROOKE (1818), BLUME (1851) accommodated it in a new monogeneric family, Sabiaceae, suggesting its affinity with Menispermaceae. Shortly afterwards MIERS (see LINDLEY, 1853), while working on Menispermaceae, placed Sabia between that family and Lardizabalaceae. HOOKER f. & THOMSON (1855) considered the genus intermediate between Menispermaceae and Schisandraceae.

The scandent habit and the resemblance of the drupelets of Sabia with those of Menispermaceae undoubtedly were a major argument for supposed affinity.

Subsequently BENTHAM & HOOKER (1862) extended the then monogeneric family Sabiaceae to include Meliosmaceae ENDL., adding the genera Meliosma BLUME and Ophiocaryon SCHOMB.; both are trees, the first Asian-American, the latter tropical American. They removed the family in its new concept from the Menispermaceous affinity and accommodated Sabiaceae near Sapin-daceae and Anacardiaceae. This position has been stable for a century and was adhered to by many leading botanists: WARBURG (1895), VON WETTSTEIN (1911), HUTCHINSON (1926, 1973), MELCHIOR (1964), TAKHTAJAN (1969), DAHLGREN (1975, 1983), and THORNE (1976, 1983). Some of these authors showed some doubt about the position and some made suggestions, e.g. WARBURG (l.c. 370), who believed one could possibly derive the flower of Meliosma from the Menispermaceous scheme and mentioned that RADLKOFER was not in favour of an affinity with Sapin-daceae or Anacardiaceae.

In recent years there is a tendency to return to BLUME'S opinion towards affinity with Menisper-maceae. Pollen morphology (ERDTMAN, 1952) and embryology (MAURITZON, 1936) have been interpreted in favour of a relationship with Menispermaceae. AIRY SHAW (1973) remarked that the opposition of calyx, corolla and stamens is a most unusual feature, but can probably be derived from the Menispermaceous type of flower. In his recent classification CRONQUIST (1981) tentatively placed Sabiaceae near Menispermaceae in the Ranunculales. Also FORMAN, in his treatment of the Menispermaceae (), shares this opinion.

Another matter is whether Sabia and Meliosma/Ophiocaryon should be accommodated in one family; hitherto they are represented by two tribes in Sabiaceae (WARBURG, 1890), differing in habit (climbers versus trees), the leaves, and in the androecium. Moreover, CRONQUIST (1981) mentioned in his discussion that, according to WOLFE, the leaf venation of Sabia is highly compatible with a position near Menispermaceae, but that of Meliosma more similar with some members of the Rosidae. There may be more arguments to accommodate Meliosma in a separate family Meliosmaceae ENDL., apart from Sabiaceae sensu stricto. This opinion was held by AIRY SHAW (1973).


Though the genera are extremely clearly defined, specific delimitation has in both genera been difficult, as it seems that racial segregation is common in both. VAN DE WATER has in Sabia employed a finer specific distinction than VAN BEUSEKOM did in Meliosma.


The only observations worth to be reported here are the presence of pen-tacyclic triterpenoids of the oleanene series and the absence of starch in seeds. The 3-acetates of oleanolic acid and oleanolic aldehyde were isolated from bark of Meliosma simplicifolia. Seeds of Meliosma myriantha SIEB. & ZUCC. (continental SE. Asia) were reported to give positive reactions for alkaloids and to contain 8% of protein and 10% of fatty oil but no starch.