Dioscorea bulbifera L.

Tuber (as produced in the East) globose to pyri-form, rarely somewhat lobed, usually covered densely with harsh short roots, nauseous in the wild plant and in quantity poisonous; but with tubers fit to eat in various cultigens selected by man, weighing up to 1 kg. In certain cultigens the tuber is suppressed in favour of rather large bulbils. In the wild plant bulbils numerous; a seedling only a few months old can be made to produce one by a checking of its growth; bulbils usually in the leaf-axils but at times displacing ♂ flowers at the base of flowering axes; small bulbils are as a rule warted; when large they may be smooth; in Africa curiously angled bulbils are produced. Flesh usually pale yellow, perhaps tinted with violet, and it oxydizes when cut to orange; very mucilaginous. Leaves usually broadly cordate, but sometimes long-cordate [var. heterophylla (ROXB.) PRAIN & BURK. (1936)], up to 20 by 20 cm and even to 32 by 32 on unusual plants, 5-nerved, secondary nerves very conspicuously ladder-like; upper surface shining and slightly bullate between the secondary nerves, lower duller, with the nerves prominent; petiole from half as long to as long as the blade, sometimes marginally winged with wings that ascend to the backs of the outer primary nerves; these wings descend on the stem having between them a wing from the centre of the base of the petiole. Sub-foliaceous auricles (Fig. 4b) may occur at the base of the petiole and partly embrace the stem; they are larger than in any other sp., but they are commonly altogether absent. Male flowering axes pendulous, 1-4 from the axil of a bract or rarely of a leaf; those arising from axils of bracts may make large pendulous inflorescences (Fig. 4a) even to 100 cm; axis carrying the flowers to 14 cm, though usually about 4 cm; when large with perhaps 100 flowers which open from the lowest upwards in a succession that is often by no means as obvious in the genus; flowers facing forwards which, the axis being pendulous, means towards the earth, pleasantly scented, sometimes tinted with a rose coloured pigment and generally with the green so diluted that collectors have called them white. Torus very small, tepals and stamens crowded together. Outer tepals linear-lanceolate, 1¼-4 mm long, but nothing like the large flowers of the Himalayan var. simbha PRAIN & BURK. (1914) have been met with in Malaysia; inner ones not so acute. Stamens half as long or a little longer, anthers as long as the filaments. Female flowering axes directed downwards but not rigid, solitary or 2 or more together from a leaf-axil, carrying about 40 flowers which appear sessile but after pollination show that they have enough pedicel to direct the ovary upwards. Bracts long-acuminate. Outer tepals rather broader than those of the male and greener, rarely more than 1½ mm long, inner ones a trifle shorter and blunter than in the male. Capsules bright chestnut and usually formed in large numbers, imbricating; wings rounded at both ends (Fig. 5f) sometimes rather broader towards the apex than below, 20-22 by 8-9 mm. Seeds as a rule not quite as broad as their wing, but nearly so.

From a rather extensive study of D. bulbifera in India a climatic limit has been determined towards the dry NW where its existence ceases ( PRAIN & BURK. Ann. R. Gard. Cale. 14 1938 433-4 ). This limit is where the rainfall of the wetter six months diminishes to less than 700 mm, an amount seemingly just adequate, though 1000 mm serve better. This observation in India leads to an expectation that there is little land in Malaysia insufficiently humid to provide a home. D. bulbifera crosses Torres Straits into Australia and passes down the east coast in the coastal margin to 21° S; but here, though the summer rains of Brisbane exceed 700 mm, it does not reach that city. Towards the W. coast of Australia it does not overpass Port Darwin. It ascends the Himalaya to 1800 m and is recorded in Yunnan at 2700 m, heights which cause the records of it on the mountains of Malaysia to be unsatisfying; there are specimens in herbaria from 1000 m in Sumatra, Borneo and Java; it probably exceeds this. Mountains however, are not where it is most abundant; this is always near sea-level and it seems to find a home readily in small islands. It was found in Verlaten Island and Sebesi near Krakatoa when the returning vegetation was investigated ( DOCTERS VAN LEEU-WEN Krakatoa 1936 435 ). It gets some dispersal by floating bulbils. The high forest chokes it out; disturbed wooded land is suitable; and in coastal woodland which is always exposed to storm damage as well as to interference by man; storms and man working together produce conditions favourable to it, and so aid in creating a greater abundance of it near the sea than in little inhabited country.

Uses. The tubers of wild plants become increasingly unpalatable as the time of new growth approaches. They are bitter and acrid and the backward tribes who fall back on them slice and cook them with lime and wood ashes; the product is rather like starch paste with acridity introduced. 'Vile in taste', wrote RUMPHIUS, adding humourously, 'but suited to the Amboinese stomach'. The bulbils of the varieties suavior PRAIN & BURK. (1914) and sativa PRAIN (1903) are much pleasanter and the best even pleasant to eat. Until recently the latter existed on the edge of Singapore from former cultivation and the names ubi china (Chinese yam) and ubi kastéla (Portuguese yam, literally yam of Castile) lingered but without providing proof that the Chinese or Portuguese had brought it either from China or elsewhere. The same variety, cultivated near Jakarta, passed as huwi singapura (Singapore yam), as if Singapore had supplied it. It has been found elsewhere in the Malay Peninsula to which it is certainly exotic. Var. suavior has been recognized in specimens from Java, Madura, Buru, and Halmaheira. Both varieties have been met with among collections from the SE of New Guinea and from the islands to the eastward. It is interesting that var. sativa, as it grows in Singapore, produces fertile female flowers at the base of male axes.It may be well to remark that PRAIN did not use the name 'sativa1 for his variety in reference to any previous application of that adjective to this species. The earlier applications of the adjective in the form of Dioscorea sativa from LINNAEUS to BENTHAM were extremely confused (PRAIN & BURK. in Kew Bull. 1919, 339). BENTHAM in 1861 accepted it as for D. bulbifera and for a time many botanists followed him; this was an error for there is no doubt regarding the earlier use of the name D. bulbifera. THUNBERG'S D. sativa was D. bulbifera as cultivated in Japan: the type of D. bulbifera var. sativa PRAIN is a plant brought to India from the Pacific and cultivated in Jail gardens under the name Otaheite potato. The writer has eaten its bulbils and found them pleasant, as cooked.

Either the bulbils off the vine, or the underground tubers are planted. Because the vines are long, training them up trees is convenient. The bulbils have to be stored for a period of time before they will sprout. The plant is annual and leaves die off for 1-4 months each year before re-sprouting from the tuber. Bulbils only grow shoots from one end unless the bulbil is cut into pieces. If the larger bulbils are cut, the cut surfaces should be dried and healed in a shady place for 2-3 days before planting. Bulbils are planted 8-12 cm below the surface and spacing can be 100 cm by 100 cm or variations of this. Normally nitrogen and potassium fertilisers give greater responses than phosphorus. Friable well drained soils are most suitable. Often very little cultivation or mounding is done. A high level of organic matter improves yield. Staking is normally required but often trees or living stakes are used. Branched stakes 2 m high are suitable. They need to be strongly erected as vine and tuber growth can be extensive and heavy.