PROSEA, Introduction to Rattans

Introduction to Rattans

General

What is a rattan?

Rattans are spiny climbing palms occurring in the Old World tropics and subtropics. They are the source of cane for the cane furniture industry, while at the same time being used for a wealth of minor purposes locally. Most cane entering world trade is collected from the wild, and throughout much of South East Asia rattan represents the most important forest product after timber. At a local level, rattan may be of great social significance in providing a not always sustainable source of income for the poorer societies living near the forest. These people may use rattan growing in the forest as a source of income for tiding over difficult periods in the agricultural cycle. Recently there has been great interest in the possibility of cultivating rattan.

The most important product of rattan is cane, that is the rattan stem stripped of its leaf sheaths; canes are sometimes confused with bamboo and, when processed into strips, may be difficult to identify as such. Bamboo is almost always hollow, and even in the few solid species it is not easily bent. Rattan is always solid and can usually be bent easily without gross deformation. The greatest diversity of rattans occurs within South East Asia.

Choice of species

The selection of species to be discussed in this volume has presented several problems. Almost all species of rattan are used for some purpose somewhere within their range of distribution. Rattans may be used by hunter gatherers for binding or thatching small temporary shelters in the forest, and their ripe fruits may be eaten. For example, Penan people living near the Mulu National Park in Sarawak could suggest uses for almost all the 71 taxa occurring within the National Park (Dransfield, 1984a). However, local species are, for the most part, not included in this volume. Most attention is given to those species which are cultivated, show great silvicultural potential, or are extensively and selectively harvested for the commercial trade. Also included are species (Calamus andamanicus Kurz, C. egregius Burr., C. ovoideus Thw. ex Tr., C. simplicifolius Wei, C. tetradactylus Hance, C. wailong Pei & Chen, and Daemonorops margaritae (Hance) Becc.) which occur naturally outside the region covered by Prosea and which are cultivated and show potential, or which may serve as silvicultural models. Species which are fairly consistently used for particular purposes, even though these may be of minor importance, are included in this volume.

Inevitably, species of little significance will have been omitted from this account. In Table 1, certain species previously recorded by Heyne (1927) and others as important have been excluded for the following reasons: unsound taxonomic basis of the data, the species are known to be extremely rare, the uses ascribed to the species are questionable or insignificant, and the lack of authentic information.

Origin and geographic distribution

True rattans are strictly Old World palms belonging to subfamily Calamoideae. Two genera of palms in the New World have climbing members: Desmoncus Mart. (all species) in the tribe Cocoeae subfamily Arecoideae, and Chamaedorea Willd. (one only of about 100 species) in tribe Hyophorbeae subfamily Ceroxyloideae. Stems of Desmoncus are used for weaving and furniture, but they are of inferior quality and little commercial significance except locally. Stems of Chamaedorea elatior Mart. are too soft to be of use.

In the Old World, rattans are distributed in equatorial Africa, the Indian subcontinent, Sri Lanka, the foothills of the Himalayas, southern China through the Malay Archipelago to Australia and the western Pacific as far as Fiji. The greatest diversity of genera and species is in the western part of Malesia.

Three of the four genera recorded for Africa are endemic. A fourth genus, Calamus L., is represented by a single, very variable species (C. deeratus Mann & H. Wendl.). Calamus, the largest rattan genus with between 370 400 species, occurs throughout the geographical range of rattans. Although overwhelmingly the greatest diversity at both generic and specific levels occurs in western Malesia, the least specialized rattans in terms of inflorescence and floral morphology are the African genera Laccosperma (G. Mann & H. Wendl.) Drude and Eremospatha (G. Mann & H. Wendl.) H. Wendl.

The biogeography of the Calamoideae is in need of further study. The distribution of the rattan genera is summarized in Table 2.

Some rattans are widespread: for example, Calamus scipionum Lour. is found from Vietnam southwards to Borneo, Sumatra and Palawan, and C. ornatus Bl. and its varieties are found in Thailand, Peninsular Malaysia, Sumatra, Java, Borneo, the Philippines and Sulawesi. In contrast, other species seem to be very narrow endemics: for example, Daemonorops oblata J. Dransf. is found only in kerangas (Bornean heath forest) forest in north west Borneo, and D. unijuga J. Dransf. is known from a single limestone hill in western Sarawak.

A striking feature of rattans is the abundance of species which occur sympatrically; as many as 30 species may be found in one locality in what is apparently rather uniform vegetation. However, there are probably microhabitat differences and subtle breeding barriers between rattan species which are not understood. Nevertheless, the abundance of rattan species in many forest types in Malesia is one of the distinctive and remarkable features of tropical rain forest in the region.

Uses

Because of their strength, flexibility and uniformity, the bare stems of rattans are used commercially for cane furniture and matting. The cane generally varies in diameter from 3-60(-70) mm or more, depending on the species. Perhaps 20% of the species are used commercially, either in whole or round form, especially for furniture frames, or in splits, peels and cores for matting and basketry. Other species may not be used because of rarity, shortness or poor mechanical properties.

In rural areas, many rattan species have been used for centuries for numerous purposes such as cordage, construction, basketry, thatching and matting. As Corner (1966) noted, long before the Portuguese brought rattan commerce to Europe with the opening of the Orient, rattans were so invaluable to village life that one could speak of a rattan civilization in South East Asia.

Heyne (1927), Burkill (1935), Brown (1941-1943), Corner (1966) and Dransfield (1979) have listed the various local uses of rattans. The uses are so numerous that a complete enumeration is impossible. Rattans are used in the making of baskets, mats, furniture, broom handles, carpet beaters, walking sticks, fish traps, animal traps, sun blinds, bird cages and for almost any other purpose requiring strength and elasticity combined with lightness. Houses, fences, bridges and even boats are bound together with rattan, often without the use of a single nail. Ropes for tethering buffaloes, mooring ropes, anchor and bridge cables are also made from rattans. Old rattan leaflets are woven for thatching, young leaflets are used as cigarette papers, young shoots or "cabbage" are eaten, rattan fruits variously used as fruit and medicine, and "Dragon's blood" obtained from the fruit scales of a few species was previously used as a dye, varnish and in local medicine.

History of the rattan trade

International trade in rattan dates to the mid 19th Century (Corner, 1966; Whitmore, 1973). However, village level utilization in the Asian region spans many centuries. Some early trade information is given in Burkill (1935).

Singapore was the clearing house for practically the entire rattan output of the South East Asia and the western Pacific at the turn of the 20th Century. From 1922 1927 it exported from 27 500 16 000 t, mainly to Hong Kong, the United States and France, in that order. During this period, exports from Kalimantan and Sulawesi increased from 9 400 19 300 t and 10 300 21 800 t respectively. Much of the raw material from Kalimantan was re exported through Singapore and Sulawesi. However, processing and further conversion were mainly done elsewhere.

By the 1970s Indonesia had become the supplier of about 90% of the world's requirements of raw rattan. In 1977, Singapore, which has no commercially harvestable rattan resources, earned more than US$ 21 million from processing and converting rattan into semi-finished products, with about 90% of its supplies coming from Indonesia (Menon, 1980). In the same year, Hong Kong, also without raw rattan of its own, imported more than US$ 26 million worth of rattan and rattan products which, after conversion and manufacture, was worth US$ 68 million in export value. By comparison, Indonesia's share of the trade, mainly of unprocessed canes, was only US$ 15 million.

External trade in rattan and rattan products has undergone great expansion in recent decades. Details of world trade in rattan and rattan products are provided by Manokaran (1990a). The increases in the value of exports from the main producer countries are striking: 250 fold over 17 years in Indonesia, 75 fold over 15 years in the Philippines, 23 fold over 9 years in Thailand, and 12 fold over 8 years in Malaysia. The combined value of exports for these four countries had risen to an annual figure of almost US$ 400 million by the late 1980s, with Indonesia accounting for almost US$ 200 million. The entire export of the Philippines and Thailand, and almost half of that of Malaysia, was of furniture. Net revenue derived from the sale of rattan goods by the two "middle men" countries, Taiwan and Hong Kong, together totalled about US$ 200 million in the late 1980s. Much of the partially processed and semi finished rattan material exported to European countries was converted to high value added products, mostly furniture.

During the 1980s, Thailand, the Philippines, Indonesia and Malaysia banned the export of rattan except as finished products. These bans have been imposed in order to stimulate the development of rattan based industries and to protect the wild resource. Indonesia, with 75 80% of the world's present production, has targeted export earnings of about US$ 600 million in the near future. Malaysia's target for export earnings from rattan furniture is about US$ 60 million by 1995.

Problems of interpretation of vernacular names

The development of extensive local rattan taxonomies reflects the social significance of rattans. Classifications have evolved to deal with rattan as it grows in the forest and to serve the product in the trade. Widespread species may be referred to by many names, and where people from different language groups live near to each other, several names may be used even for local rattan species. Serious confusion has arisen from the uncritical use of vernacular names.

The following three examples illustrate the problems of interpreting local names. Calamus caesius Bl. is commonly referred to by Malay speakers as "sega", yet in the rattan growing area of Kalimantan Tengah, it is known as "taman"; throughout Sarawak it is known as "leutik", while "sega" is used for the related species Calamus optimus Becc. In Palawan, the presence of true Calamus caesius remained unrecorded until 1979 because "sika", the name by which it is called in Palawan, is said to be the vernacular name of an unrelated and rather rare species (Calamus spinifolius Becc.) in Luzon (Merrill, 1922). Uncritical use of the lexicon of Philippine vernacular plant names obscured the presence of one of the best small diameter furniture canes, for which a cultivation procedure had already been developed that could have been transferred earlier to the Philippines (Dransfield, 1980). Incidentally, it was also responsible for the incorrect recording of Calamus spinifolius in Palawan.

Another example is provided by the Malay name "rotan batu": in Peninsular Malaysia this is used fairly consistently for Calamus insignis Becc., a species with a stem diameter of 8 10 mm; in Sabah it is used for an unrelated species, Calamus subinermis Becc., which has a stem diameter of 18 25 mm.

The final example is one of the most serious confusions: Heyne (1927) cites "tohiti" as the vernacular name of a rare, poorly known species, Calamus inops Becc. ex Heyne; herbarium specimens of this species have a stem diameter of about 12 mm. "Tohiti" is currently also used as a trade name for the best quality large diameter (22 30 mm) canes in Sulawesi. There is evidence of the use of "tohiti" for at least 4 species, yet the trade name "tohiti" has uncritically been referred to as Calamus inops, which is unlikely to be correct because of the much smaller stem diameter of this species. Accounts of various aspects of the cane(s) known as "tohiti" have been published under the scientific name Calamus inops, yet the results cannot refer to this species, and, in the absence of voucher specimens, the cane identity cannot be established; the experimental results are thus irreproducable and useless.

Therefore, great caution is required in the interpretation of vernacular names. In experimental work it is important to prepare herbarium specimens as vouchers, to provide an essential reference.

Botany

Morphology

Habit

Although most members of the 13 rattan genera are climbers, some species are short stemmed or acaulescent undergrowth palms that scarcely fit into the definition of rattan. Nevertheless, they belong to rattan genera, and those that are used are mentioned in this volume. Those species which have subterranean stems, such as Calamus minutus J. Dransf., occur as rosettes on the forest floor. Another group of non climbing rattan relatives are short stemmed and tend to scramble weakly in the undergrowth, such as Calamus myriacanthus Becc.. In a third group the stems are stiff and erect; good examples are provided by Calamus nanodendron J. Dransf. and C. arborescens Griff.; their stems are generally too stiff and heavy to be used for furniture.

Rattans may be single stemmed as in Calamus manan Miq., or multiple stemmed as in C. caesius. The consequences for harvesting are obvious: single stemmed species provide a single harvest and do not regenerate from cut stumps, whereas multiple stemmed species can be harvested continually. Clumps are built up by the development of sucker shoots produced from axillary or leaf opposed buds at the base of the stem. Such buds usually develop as short rhizomes that then grow into aerial stems; in C. trachycoleus Becc. sucker shoots develop into stolons that may reach 3 m or more in length before metamorphosing into aerial stems. Usually vegetative buds are confined to the few basal nodes of the stem, and clumps are built up by the production of shoots from successive orders of suckers. Occasionally, vegetative shoots are produced some way up the aerial stem; the significance of such aerial shoots is not obvious as they rarely, if ever, seem to grow into new aerial stems. However, in most Korthalsia species and in Laccosperma opacum (G. Mann & H. Wendl.) Drude the stems regularly branch in the forest canopy, resulting in vast aerial entanglements that make harvesting these rattans extremely difficult. In a few unusual acaulescent rattans such as Calamus pygmaeus Becc. and Daemonorops ingens J. Dransf. vegetative reproduction can occur by the production of vegetative shoots from the tips of the inflorescences.

Root

Little is known about the root system of rattans. Casual observations indicate that the root system of Calamus caesius may be complex, with widely radiating, horizontally growing roots, and short vertical root systems, some growing geotropically (i.e. downwards) and some growing apogeotropically (i.e. upwards) (Dransfield, 1979). The apogeotropic roots may be concentrated in the leaf litter layer, and frequently bear patches of loose corky tissue, usually associated with gas exchange. Nur Supardi has reported that roots of Calamus manan can radiate from the base of the plant as far as 8 m.

Stem

Rattan stems, the canes of the rattan trade are, in the young state, covered by tight, usually densely spiny, leaf sheaths; as the stem ages and the lower leaves die and drop off, the stem becomes exposed. In most genera the leaf sheaths eventually erode away to leave a clean and smooth stem surface, but in the endemic African genera and Korthalsia Bl., remnants of the sheaths remain adhering closely to the stem surface.

Stems can vary from a few mm to over 10 cm in diameter. The longest cane recorded was 175 m long (Burkill, 1935). If left unharvested, canes longer than 100 m would occur relatively frequently. Rattan stem does not increase in diameter with age. The seedling first grows in stem diameter and then increases in length, the early establishment growth being responsible for the diameter of the aerial stem. There is, however, some variation in diameter along the stem, the base usually being thicker and then decreasing in diameter upward, the diameter usually attaining a maximum when the rattan crown reaches the forest canopy or reproductive maturity. There may also be variations in diameter between nodes that produce inflorescences and those that do not, the former usually being thinner. All such variations in diameter may affect the quality of the harvested cane.

Most canes are more or less circular in cross section, but in those species of Calamus which have flagella (see below at climbing organ) the position of the flagellum is marked by a vertical ridge on the internode, the position of the ridge being in line with the axil of the leaf that subtends it. In most species of Plectocomiopsis Becc. and Eremospatha the stem is more or less triangular in cross section, thus restricting the uses to which the canes can be put. The growing point of the stem is concealed within the tightly enclosing upper leaf sheaths; it is thus positioned some distance below the apparent stem tip as represented by the tip of the uppermost leaf sheath. In almost all species this vital well protected growing point is edible and is sometimes harvested and sold in markets. Just below the growing point, the stem is usually soft and poorly lignified, and thus cannot be used as cane.

Various studies have been done on the anatomy of rattan stems, starting with Tomlinson (1961). This was followed by two pilot studies, one by Siripatanadilok (1974) on the canes of Java, and Teoh (1978) on the canes of Peninsular Malaysia, aimed at providing a basis for the identification of bare harvested canes. More detailed research has been done by Liese & Weiner (1988) and Weiner & Liese (1990) and Weiner (1992), aimed at studying all genera and species of rattan. Recently there have been numerous anatomical studies of Indian canes (Bhat, 1991, 1992; Bhat & Thulasidas, 1989; Bhat & Varghese, 1991; Bhat et al., 1989, 1990, 1993). All these studies have shown that there are significant anatomical differences among most of the rattan genera, and that species or groups of species are anatomically distinct, and that the quality of cane can be correlated with anatomy to some extent. Generally, good quality cane has an even distribution of vascular bundles throughout the stem, and the ground parenchymatous tissue is uniformly lignified. Cane with little lignification and uneven distribution of vascular bundles is usually poor in quality.

Leaf

Leaves are produced sequentially one at a time; they consist of a tubular sheathing base, the leaf sheath, which arises from the node of the stem. At its upper end, the sheath narrows into the petiole that continues into the rachis or leaflet bearing portion of the leaf. Although a petiole is usually present, it is sometimes very short or absent. In many species the rachis is extended beyond the terminal leaflets into a barbed whip (cirrus) which acts as a climbing organ (see below at climbing organ).

The leaf sheath develops from a soft meristematic (i.e. growing) area at its base; thus, the upper part of the sheath matures before its base. Only 1/4 1/3 of the length of the sheath is exposed beyond the shelter of the preceding (i.e. older) leaf. This length of exposed sheath is assumed to correspond with the length of the internode of the stem itself, and is usually densely spiny. Spine arrangement is extraordinarily diverse and frequently of diagnostic importance. In a few species, the leaf sheaths lack spines; for example, in some forms of Calamus laevigatus Mart. and C. ornatus the sheaths are devoid of spines; these species are not totally spineless the leaf rachis and cirri and flagella are still heavily armed. A wide range of hairs, scales and/or wax may occur between the spines. At the mouth of the sheath (i.e. the tip) where the sheath narrows into the petiole or leaf rachis, there may be a prolongation of the main part of the sheath. Often this prolongation is tubular and encloses the sheath of the subsequent leaf. This is usually referred to as an ocrea. Sometimes the ocrea is split and the margins rolled. It may be leathery or papery in texture. All species of the genus Korthalsia bear ocreas, and in some species provide a home for ants. In Pogonotium J. Dransf. the prolongation consists of two lobes, one on each side of the petiole, referred to as auricles.

In many rattans there is a marked swelling on the leaf sheath just below the petiole or leaf rachis, known as the knee. It is only present in climbing members of the genera Calamus, Ceratolobus Bl., Calospatha Becc., Daemonorops Bl. and Pogonotium; in acaulescent members of these genera it is absent.

The petiole is variable in length; it is usually much longer in juvenile than in mature leaves, and, as indicated above, it may be absent altogether. The petiole is often characteristically armed with spines. In some species, there may be neat rows of very long, slender but tough spines on either side of the petiole that appear to function as organs for trapping litter from the forest canopy.

The leaf rachis is defined as the part of the leaf axis which bears the leaflets. Frequently it bears reflexed grapnel like spines on the lower surface and these contribute to the climbing process by locking onto support.

The rattan leaf is basically pinnate. In almost all rattans the leaflets consist of a single ⋀ shaped fold. Only in the endemic African genera and in the seedlings of Plectocomiopsis and Myrialepis Becc. are the leaflets sometimes composed of more than one fold. Leaflet shape, armature, hairiness and arrangement are all of diagnostic importance.

Climbing organ

There are two whip like organs associated with climbing in rattans, superficially very similar, but not homologous. Climbing organs usually develop only when the aerial stem begins to develop. The cirrus is an extension of the leaf rachis beyond the terminal leaflets, while the flagellum is a sterile inflorescence borne on the leaf sheath near the knee. Both are whip like and bear groups of short reflexed spines. In a cirrus the spines are borne directly on the extension of the leaf rachis; in the flagellum an axis is covered with very tightly sheathing bracts that bear spines. Flagella are found only in the genus Calamus but not all species of Calamus have flagella many, including the best commercial species, are cirrate. Usually, flagella and cirri are mutually exclusive, but in a few species of Calamus (e.g. C. semoi Becc.) there may be short flagella as well as well developed cirri. In all Asiatic rattan genera the cirrus develops as a extension of the rachis beyond the terminal leaflets. In the African genera Laccosperma, Oncocalamus (G. Mann & H. Wendl.) G. Mann & H. Wendl. and Eremospatha there is a clearly defined cirrus bearing very large paired reflexed spines on either side, and also reflexed grapnel like spines on the undersurface. The paired spines are in fact modified leaflets (known as acanthophylls). The reflexed grapnel spines serve to anchor the whips to potential support trees when they come in contact and further movement tends to strengthen the attachment. Species of Calamus which bear flagella have internodes which are uneven in cross section; this is due to the adnation (joining) of the base of the flagellum to the surface of the internode above the leaf which subtends it (Fisher & Dransfield, 1977). This means that, even though it may otherwise have a good internal structure, the cane is marred by the uneven cross section. Canes such as those of Calamus ornatus are usually decorticated to remove the irregularities before they are used.

Not all rattans have cirri or flagella. Members of the genera Pogonotium, Retispatha J. Dransf. and Calospatha lack climbing whips but climb weakly in the forest with the aid of the reflexed spines on the lower surface of the leaf rachis.

Inflorescence and flower

Inflorescences are produced singly at the node, borne in the leaf axil, and usually the lower part of the axis of the inflorescence is joined to the internode, and also to the leaf sheath of the following leaf (Fisher & Dransfield, 1977). The result of the "adnation" is that the internode of the cane above the leaf which subtends the inflorescence will be marked by a ridge representing the vascular supply to the inflorescence.

Inflorescences vary greatly in size and overall structure, and inflorescence structure offers the basis for separating the genera. Details of the inflorescences and their taxonomic significance may be found in Dransfield (1979, 1984) and Uhl & Dransfield (1987) (Section 1.2.2).

The basic branching pattern of the inflorescences of rattans is more or less the same. Differences between rattan genera are found in the elaboration or reduction of the bracts and differences in the degree of persistence, and also in the number of orders of branching. The main axis bears a basal bract or prophyll which may be short and tubular, or large, and which encloses the entire inflorescence. Branches are borne in the axils of subsequent bracts. The branches in turn bear bracts, the lowermost of which is usually empty, subsequent bracts subtending branches, and so on. The ultimate flower bearing branches are termed "rachillae". In most species of Calamus and Daemonorops male flowers are borne on branches of the third order, whereas in the female inflorescence flowers are borne on branches of the second order.

All species of Korthalsia, Laccosperma and Eremospatha have hermaphroditic flowers. Oncocalamus is monoecious, bearing male and female flowers in tight clusters. All other rattan genera are dioecious that is, male and female flowers are borne on separate plants. Details of flower structure may be found in Uhl & Dransfield (1987).

Fruit

The fruits of all rattans are covered with vertical rows of overlapping reflexed scales. Beneath the scales lies the rest of the pericarp. In Korthalsia, Laccosperma, Eremospatha and Oncocalamus the mesocarp is thick and fleshy at maturity and the seed coat is dry. In all other genera, the pericarp is thin and dry at maturity, and the seed is covered with a fleshy outer layer, the sarcotesta. Fruit maturity is usually indicated by a slight change of colour of the scales. There is usually only one seed in each fruit but some rattans may regularly have up to three seeds. Within the integuments lies the endosperm which can be variously shaped. The endosperm is homogeneous or ruminate and the embryo lies in a shallow pit basally or laterally.

Seedling

Germination of rattan seeds is adjacent ligular (Uhl & Dransfield, 1987). The first sign of germination is the emergence of a short plug from the embryo pit. From this emerge roots and then an irregular swelling, out of which emerges the shoot. The first foliage organ to emerge from the plug is a small, bladeless structure. The first leaf which bears a blade (eophyll) is usually the next foliar organ to emerge. The shape of the eophyll and the number of leaflets it bears vary from species to species and are of importance in identifying rattans at the seedling stage.

Taxonomy

In the most recent classification of the palms (Uhl & Dransfield, 1987) the rattan genera are included in subfamily Calamoideae in tribe Calameae (Table 3). However, there are remarkable differences in inflorescence and flower structure between groups of rattan genera. Laccosperma (synonym Ancistrophyllum (G. Mann & H. Wendl.) H. Wendl.) and Eremospatha from equatorial Africa resemble each other in vegetative and inflorescence morphology and the flowers are hermaphroditic and borne in pairs. Because of this, the two genera are grouped together in subtribe Ancistrophyllinae. The third endemic African genus Oncocalamus is vegetatively very similar to the other two genera, but its flowers are quite different: they are unisexual and are borne in tight complex clusters in the hollows of the bracts. The structure and development of these unique flower clusters have yet to be elucidated. Nevertheless, they are so peculiar that Oncocalamus has been included in its own subtribe Oncocalaminae. Asiatic Korthalsia has hermaphroditic flowers, but borne singly in the axils of bracts on cylindrical branches reminiscent of those of the non climbing sago palms, Metroxylon Rottb. Korthalsia is thus included in the same subtribe as Metroxylon - Metroxylinae.

Calamus, Calospatha, Ceratolobus, Daemonorops and Pogonotium all share unusual characters in the inflorescences with the non climbing genera Salacca Reinw. and Eleiodoxa (Becc.) Burr. The plants are dioecious; in the male inflorescence the flowers may be borne singly or in pairs, whereas in the female inflorescence, each female flower is borne together with a sterile male flower, the anthers of which do not contain pollen. Retispatha lacks these sterile male flowers but is otherwise very similar to these other genera. They are all included in the subtribe Calaminae. The remaining rattan genera Myrialepis, Plectocomia and Plectocomiopsis share vegetative and inflorescence characters; they are dioecous, the flowering is terminal and is followed by the death of the flowering stem, and there are no sterile male flowers in the female inflorescence. These three genera are included together in their own subtribe Plectocomiinae.

The striking differences in reproductive characteristics between the groups of genera suggest that the climbing habit may have arisen more than once during the evolution of the Calameae.

Despite their importance and the recent upsurge of interest in the rattan resource, the rattan flora of much of the rattan growing areas of South East Asia remains poorly known. Some areas are now provided with taxonomic guides, e.g. Peninsular Malaysia (Dransfield, 1979), Sabah (Dransfield, 1984b), Sarawak (Dransfield, 1992) and Sri Lanka (de Zoysa). Introductory guides are available for Papua New Guinea (Johns & Taurereko, 1989a, 1989b; Johns & Zibe, 1989) but a complete taxonomic account of the rattans of the whole island of New Guinea will require much further taxonomic research. A guide to the rattans of India has recently been published (Basu, 1992). Although there has been recent rattan taxonomic survey work in Indonesia, a taxonomic guide will require a great deal more inventory and herbarium work. Rattans of Thailand have been studied by Vongkaluang and Dransfield but an identification manual has yet to be completed. Wei and Pei have both been active in working on the taxonomy of Chinese rattans. Fernando, Lapis and Madulid have worked on the rattans of the Philippines and the rattan flora is now relatively well known there. Taxonomic inventory work provides an essential base for developing rattan, both as wild crop and in plantation. As there are so many species it is important that species delimitation is well understood. It is essential that the economically important species can be distinguished by their fruits and seedlings as well as at maturity, as there are many instances where seedlings of commercially unimportant species may at first sight be closely similar to those of the best species. Among the six hundred or so species there may well be species of excellent quality and great silvicultural potential which remain taxonomically unknown, particularly in the Moluccas and New Guinea.

With so many species and complex synonymy, there are bound to be many nomenclatural problems. As the taxonomy of the rattans becomes better known, name changes will be inevitable (Dransfield, 1985). In order to catalogue the large number of names, a comprehensive nomenclatural data base of rattans is being developed at the Royal Botanic Gardens Kew.

Growth and development

Vegetative growth

Most rattans growing naturally produce abundant seedlings, but high mortality, presumably through competition for light, water and nutrients, and through predation, leads to only a few seedlings attaining maturity. For high climbing rattans, earlier stem production from the rosette stage and greater production of stems in a clump are initiated by exposure to adequate light; light also enhances stem elongation (Manokaran, 1985). Stem elongation is continuous but variable from period to period.

Whereas there is no published information on the growth rates of rattans growing in the wild, such information is available for commercial species undergoing silvicultural trials. Some information is provided in Table 4.

Flowering and fruiting

Rattans display two main modes of flowering. In one (hapaxanthy), a period of vegetative growth is followed by the simultaneous production of inflorescences from the axils of the uppermost leaves. Flowering (in males) and fruiting is followed by the death of the stem which is usually replaced by sucker shoots. In the other flowering mode (pleonanthy), inflorescences are produced continually after a period of juvenile growth, and flowering and fruiting do not result in the death of the stem. All species of Korthalsia, Plectocomia Mart., Plectocomiopsis, Myrialepis, Laccosperma and Oncocalamus and a few species of Daemonorops are hapaxanthic; all other rattans are pleonanthic. The significance of the different methods of flowering is discussed in Dransfield (1978).

Information on time and periodicity of flowering, age at first flowering, fruit production and fruit maturation period have all been reviewed by Manokaran (1985).

For many rattan species, development of inflorescences and production of fruits are seasonal, triggered by some external influences. C. caesius, C. manan, C. tumidus Furt., C. scipionum, C. speciosissimus Furt., C. trachycoleus and Plectocomiopsis wrayi Becc. have been shown to belong to this category. The cue for flowering for many of these species may be a period of relative dryness, and hence of higher temperature, followed by a period of higher rainfall (Manokaran, 1989). In C. caesius it has been shown that certain clumps and individual stems flower almost every year. In other rattan species, flowering is independent of external influences. Daemonorops angustifolia (Griff.) Mart. and Calamus filipendulus Becc., for example, produce inflorescences continuously along the stem. Age at first flowering has been ascertained for species from one or other of these two categories. C. caesius flowered at 5.3 years from planting, C. manan at about 5.5 years from planting, C. trachycoleus almost 4 years from planting, and C. filipendulus at about 4.5 years from planting (Manokaran, 1985).

Anthesis in C. caesius has been shown to be 1.5 2 months from inflorescence emergence, and the period from inflorescence emergence to fruit production is 16-18 months for C. caesius and 16 months for C. speciosissimus (Manokaran, 1989). Fruits of C. manan have been shown to mature 15 months after flowering (Darus & Abdul Rasip, 1989) whereas fruits of C. trachycoleus have been observed to mature 14 months after flowering (Tan & Raja Barizan). Elsewhere in the seasonal climate of Bangladesh, fruits of C. viminalis Willd. have been observed to mature within 160-170 days from flower initiation.

Manokaran (1979) has reported that the clustering species C. caesius can have more than 2000 fruits maturing on a stem at any one time; on one occasion about 3000 fruits were obtained from one stem. C. scipionum (clustering species) and C. manan (solitary species) can likewise have respectively 2000-3000 and 3000-5000 fruits or more on a stem. Thus, during the lifetime of each stem of these pleonanthic species, the number of fruits produced by each stem probably runs into the tens of thousands. In the clustering species, the total number of fruits increases according to the number of mature stems produced. At the other end of the scale, the rare Pogonotium divaricatum J. Dransf. produces only 1 or 2 fruits per inflorescence, and Calamus gonospermus Becc. produces only 5 (Dransfield, 1981). The solitary C. laevigatus may produce only 400-500 fruits on any one occasion, and the solitary C. tumidus and the clustering C. ornatus may produce 3000-4000 and up to 1000 fruits per stem respectively.

Ecology

The large number of rattan species and their wide geographical range is matched by great ecological diversity; consequently, generalizations concerning the ecology of rattans frequently need to be qualified and it is imprudent to draw inferences from observations which do not have a firm taxonomic base.

Crude ecological preferences for rattan species have been obtained from observations made during taxonomic inventory work. They are summarized for Peninsular Malaysia, Sabah and Sarawak in Dransfield (1979, 1984, 1992). Such broad ecological summaries are invaluable as a basis for establishing cultivation procedures, even though they are not based on detailed ecological experimentation. A major gap in the knowledge of the ecology of rattans is an understanding of the population dynamics (demography); such knowledge is a prerequisite for developing strategies for sustainable rattan harvesting.

The restriction of rattan species to different climatic zones suggests that these species may have precise climatic requirements. At the extreme north of the range of rattans, it is possible that they may occasionally be subjected to temperatures below O°C. In altitudinal range they occur from sea level up to 3000 m, the highest altitude record being held by Calamus gibbsianus Becc. at 3000 m on Mount Kinabalu, Sabah. There are usually differences in the rattan flora at different altitudes.

Most rattans are vigorous climbers. Species such as Calamus manan and C. ornatus, which grow in primary forest, require light gaps for further growth and development. Those growing in secondary or disturbed forests such as Plectocomiopsis geminiflora (Griff.) Becc. and Myrialepis paradoxa (Kurz) J. Dransf. receive ample light for much of their lifespan. These species appear to need very high light intensities for maximum growth. Certain rattan species slowly grow to maturity in shade in the forest undergrowth. These include non climbing short stemmed species such as Daemonorops calicarpa (Griff.) Mart. and also climbers such as D. didymophylla Becc. and D. collarifera Becc.

It has been demonstrated that adequate light is an important factor for enhanced growth in C. caesius and C. scipionum, and has been observed to be so in C. manan, C. trachycoleus, C. tumidus Furt. and other commercial species. Despite the importance of light, however, seedlings of C. scipionum, C. caesius and C. manan cannot withstand full sunlight throughout the day because the leaves will become scorched and the plants appear unhealthy. This response may be true for most climbing rattan species.

In the forest there may be a wealth of different species of rattan growing sympatrically, occupying different niches, from dense undergrowth to large light gaps formed by the fall of large trees. There is a wide range of different light regimes in the forest and there appears to be a corresponding range of rattans adapted to different light regimes (Dransfield, 1979). At one extreme are rattan species adapted to grow in the low light regimes of the forest undergrowth; at the other extreme are species adapted to the high light regimes of major gaps and land slips. The single most significant ecological factor affecting the growth of commercially important rattan species in plantation is light; unless sufficient light reaches the rattan plant, it may remain on the forest floor, stunted and in a state of perpetual juvenility.

Within the natural distribution range of rattans, rattan species can be found in most forest types and on most soils and rock types. Mangrove, however, is generally devoid of rattans except on the landward fringe. Some species of rattan are restricted to forest on certain rock or soil types. For example, limestone and ultramafic rock carry distinctive rattan floras.

There is a range of rattans adapted to different soil moisture regimes from swamps to dry ridge tops. In cultivation, all species tested so far in trials have performed best in soils which are not subjected to severe drying out, but few rattans can withstand permanent waterlogging or prolonged flooding. For example, although C. trachycoleus is adapted to seasonally severely flooded habitats, it also grows quite well on higher and drier ground, provided total annual rainfall is as high as in its natural habitat. The water and light requirements for the growth of some species of rattans have been amply demonstrated and reviewed by Manokaran (1985).

Despite being well protected by the tightly tubular sheaths that are themselves usually fiercely armed, the rattan growing point is not infrequently attacked by herbivores. Such attacks usually result in the death of the growing point and hence of the whole stem. The animals involved include pigs (when the rattan is accessible at ground level), rats and squirrels and, perhaps most destructive, elephants.

Animals seem to be the main agents of dispersal in rattans. The fleshy layer in the fruit wall or the sarcotesta appears to be attractive to birds and mammals. Fruit may be ingested whole or sucked and spat out. Sometimes the first signs of fruiting are twisted broken rachillae with fruit remains on the forest floor, presumably the work of apes or monkeys.

Several species of Daemonorops, Korthalsia and Calamus have morphological adaptions that provide nesting sites for ants; these adaptations include interlocking spines that form galleries and swollen ocreas or recurved proximal leaflets that enclose the sheathed stem. Details of ant/rattan relationships are given in Dransfield (1979) and Uhl & Dransfield (1987). This subject is of considerable interest and has been neglected until recently in ant/plant studies. The relationship usually also involves scale insects. It has been suggested (Dransfield, 1979) that the presence of ants may be of adaptive significance to the rattan in providing extra protection against herbivores, and circumstantial evidence seems to suggest that rattan species lacking ants may be preferentially eaten before those with ants (Dransfield, 1981). However, protection may not be the only significance. Rickson & Rickson (1986) have demonstrated an increase in nutrients available to Daemonorops verticillaris (Griff.) Mart. brought to the rattan by the foraging activities of the symbiotic ants. They suggest that this enhanced availability of nutrients may be of adaptive significance.

Exploitation and cultivation

History of rattan cultivation

Rattan has been cultivated at three scales: plantation scale for commercial use, village scale for domestic use and as a cash crop, and experimentally in small plots. Cultivation in botanic gardens is not reported here.

Indonesia

The first and most successful rattan plantations are those that were established in the areas around Barito, Kapuas and Kaharjan in Kalimantan about the year 1850 (van Tuil, 1929). Christian missionaries are said to have encouraged the planting of the two small diameter clustering species Calamus caesius and C. trachycoleus by villagers on smallholdings. Since then, the area of the smallholdings along the alluvial flats of the Barito River and its tributaries in Central and South Kalimantan provinces has increased to 15 000 ha, mainly planted with C. trachycoleus. By the latter part of the 1980s these village level plantations were contributing about 10% of Indonesia's raw rattan supplies.

In East Kalimantan province, shifting cultivators have, for a long time, planted C. caesius in forest land left fallow after clearance and production of food crops (Weinstock, 1983). Within a period of up to 15 years, the rattan is harvested for sale and the land cleared again for food crop production.

Cultivation trials of a few commercial species including C. manan were begun in the 1980s, mainly in Java. From 1988 1993 the state forestry corporations have planted several thousand ha mainly with C. caesius, and to a small extent C. trachycoleus, in both Java and East Kalimantan.

Malaysia

Cultivation of C. caesius was reported along the Pahang River in Peninsular Malaysia early this century (Brown, 1913) but this no longer exists today. In 1910 an attempt by the Forest Administration of Perak to plant C. caesius was unsuccessful. In Sabah, C. caesius was planted in several ha of secondary forest on the lower Labuk River (Meijer, 1965). This species is commonly cultivated by the Iban people along the Rejang River in Sarawak (Browne, 1955). Village level cultivation in Sarawak of species such as C. caesius, C. optimus, C. ornatus, C. javensis, C. scipionum, C. semoi, C. erioacanthus Becc. and Korthalsia cheb Becc. has also been reported by Dransfield (1992). These species are cultivated in orchards, in secondary forest behind longhouses, or in rubber holdings, mainly for domestic use.

The earliest experimental planting of rattan in Peninsular Malaysia was undertaken in 1960 by the Forest Department of Selangor when C. manan was planted in small plots in the Ulu Langat Forest Reserve (Manokaran, 1977) and in Sungei Buluh Forest Reserve (Johari Bin Baharudin & Che'Aziz Bin Ali, 1982). A cultivation trial was later initiated in 1972 in Pahang. Beginning in 1975, experimental cultivation of C. scipionum and C. caesius was undertaken by the Forest Research Institute of Malaysia (FRIM) (Manokaran, 1985). This institute undertook trials with C. manan beginning in 1978 (Nur Supardi & Wan Razali Mohd, 1989) and C. manan was also interplanted with rubber trees (Salleh & Aminuddin, 1986). C. trachycoleus, which is endemic to Kalimantan (Indonesia), has also been planted in trials by FRIM in recent years. The forest departments of some states in Peninsular Malaysia have also established rattan trial plots in recent years. During this period, a C. caesius plantation of a few hundred ha was established in Selangor. In all, over 1100 ha have been planted with rattan in Peninsular Malaysia.

The first commercial rattan plantation was developed over 4 years by the Sabah Forestry Development Authority (SAFODA) beginning in 1980 1981 in the Sandakan District. The 4000 ha plantation in logged forest is subject to severe annual flooding and consists of 70% C. trachycoleus and 30% C. caesius. Subsequently, a total of about 3000 ha of logged forest, abandoned rubber plantations and Acacia mangium Willd. plantations in various parts of Sabah have been planted with C. manan, C. caesius and C. merrillii Becc. by SAFODA. A further 2000 ha or more in Lahad Datu and Sandakan districts in Sabah have been planted with C. manan and C. caesius by two private companies in recent years. In Sarawak, two experimental plantings of C. optimus were carried out in small plots in 1982-1983. In 1989 about 100 ha were planted with rattan. In 1990 a private company began planting on a large scale in Ulu Bintulu, Sarawak, using C. caesius, C. trachycoleus, C. optimus and C. manan.

The Philippines

In the Philippines the Forest Research Institute (now the Ecosystems Research and Development Bureau) initiated cultivation trials of C. merrillii and C. ornatus var. philippinensis Becc. in Pagbilao, Quezon in 1977 (Pollisco, 1989). A plantation of 5000 ha, mainly of C. merrillii, was started in 1983 in Bislig in Mindanao by the Paper Industries Corporation of the Philippines. So far, plantings have been carried out in nearly 4000 ha of logged forest. Elsewhere, a private company (San Teodoro, Mindoro Oriental) planted 50 ha of logged forest with various rattan species in 1983. By early 1990, the same corporation had planted 500 ha of C. merrillii and 280 ha of C. ornatus var. philippinensis under a matchwood tree plantation of Endospermum peltatum Merr. in Agusan del Sur, Mindanao.

Papua New Guinea

In Papua New Guinea, village scale plantings of small diameter canes have been carried out for many generations in several areas (Zieck, 1972).

Thailand

In Thailand rattan cultivation first began in 1968 with the planting of C. caesius in Rangea District in Narathiwat Province (Bhodthipuks & Ramyarangsi, 1989); a total of 213 ha had been planted by 1978. Beginning in 1979 the Royal Forest Department in Thailand initiated cultivation trials of C. caesius of 16 ha each in Ranong, Surathani and Chumporn districts. Between 1980 1987 C. manan and C. caesius were planted on an area of 930 ha in Sukirin district of Narathiwat Province. In recent years, other cultivation trials of various species have been initiated in several regions.

China

For generations, clumps of rattan have been planted in forest or village fringes by villagers in southern Yunnan for domestic use. Decreasing domestic supplies from the wild, and increased demand for the raw material, led to large scale cultivation of rattan in the 1970s. By 1980, 30 000 ha of forest on Hainan Island, Guangdong Province had been planted with 20 million seedlings, mainly of C. tetradactylus Hance and, to a lesser extent, of Daemonorops margaritae (Hance) Becc. (Xu, 1985, 1989). A rattan plantation of these species as well as of C. egregius Burret and C. simplicifolius Wei was also established in Lundao, Guangdong Province. Small scale plantations have now been established in other areas in Guangdong and Fujian Provinces. Cultivation trials on various species have also been initiated within the last decade.

Other countries

In India various species have been cultivated (Badhwar et al., 1957). In recent years the Kerala Forest Research Institute started cultivation trials on various species of the genus Calamus (mainly C. gamblei Becc., C. hookerianus Becc., C. pseudotenuis Becc., C. rotang L. and C. thwaitesii Becc.).

In Sri Lanka C. ovoideus Thwaites ex Trimen, C. zeylanicus Becc. and C. thwaitesii Becc. have been planted experimentally in recent years (de Zoysa & Vivekanandan, 1989). In Bangladesh cultivation trials, especially of Daemonorops jenkinsiana (Griff.) Mart. were carried out in the early 1980s (Wong, 1984).

Propagation

Aziah & Manokaran (1985) have reviewed the various ways in which rattans are propagated. Propagation is usually by seed; vegetative methods through the use of offsets and rhizome transplants, and tissue culture are rarely employed.

Propagation by seed

Generally the sarcotesta adheres closely to the seed and is difficult to separate. It is recommended that the outer scaly pericarp of the fruit and the fleshy sarcotesta be removed before the seed is sown, as seeds sown with the pericarp and sarcotesta intact show poor germination rates (Manokaran, 1978).

Direct sowing

Direct sowing of seed in soil at the planting site is unlikely to be successful as heavy losses of seeds and newly germinated plants will result from predation by animals and birds, from drying out in hot weather, and disease.

Cultivation of rattans is at present based entirely on seedlings raised in nurseries. The germinated seeds and seedlings in a nursery are protected from pests and are shielded from sun scorch by overhead shelters of palm frond thatching or by other crop plants in village backyard nurseries.

Where seeds are procured from the wild, seed supply is a problem because of the rapidly depleting rattan resources. This is compounded by the fact that almost all the commercially important rattans are dioecious, with only a proportion of the rattan plants of any species in any area bearing fruits. In mature plantations, however, vast amounts of seeds become available during the fruiting season.

Planting of wildings

The use of wildings obtained from the forest floor allows more rapid establishment of a plantation by circumventing the nursery stage. However, wildings of economic species are generally found in too low a density to be the source of propagating material for large scale cultivation. Furthermore, survival rates tend to be low, and there may be problems in identification of wildings, resulting in the possibility of mixed plantings.

Vegetative propagation

Rattans can be propagated vegetatively by suckers, whole rhizomes, and by tissue culture. Suckers from clustering rattans can be separated from the clump with some roots intact, potted immediately in a suitable soil mixture and transplanted to the field after a period of stabilization in the nursery. This method of propagation is feasible only when it is difficult to obtain seeds, and is applicable only to freely clustering species, and only when very small scale or backyard planting is undertaken. As the treatment of growth regulating substances induced better rooting of the suckers, observations are being made on the field performance of such treated suckers in India (Bhat et al., 1989).

The rhizome systems of clustering species support the numerous suckers that develop into stems. For propagation purposes, stems of a clump can be cut off flush with the ground, and the rhizome dug up and transplanted to the planting site; the clump then develops very quickly from the suckers left intact. Rhizome planting of Daemonorops jenkinsiana has only been carried out in Bangladesh (Wong, 1984).

The value of tissue culture for rattans would be in mass production of elite strains, and in the conservation of genetic variability. Tissue culture programmes in Malaysia and Thailand have yielded plantlets, but only in small quantities. Tissue culture methods must be developed further if they are to be used on a large commercial scale.

Management

At present wild rattan resources are rarely managed. Licences valid for variable periods of 6 months to a few years allow harvests from forested areas, often on an unsustainable basis.

Village level rattan plantations have been managed on a sustainable basis in Central and South Kalimantan since the 1850s. During the last decade, rattan plantations have been developed and managed in Sabah (Malaysia) and in southern China.

In general, management of plantations begins soon after planting. It involves replacing planted seedlings that have died and those growing very poorly. This process should be completed within 3 6 months after planting. During the early stages, weeding is carried out to remove climbers that twine around the seedlings. Access paths are also kept passable by slashing weeds with a "parang" or machete. This process is best carried out once in 6 months in the early years after planting.

As about 50% overhead light appears to be optimum for good growth of rattan seedlings, selective tree felling and removal of overhead branches may be necessary. This needs to be done soon after transplanting has been carried out.

In Kalimantan, the mother stems of C. caesius and C. trachycoleus are often pruned to stimulate earlier and more abundant suckering. In this area, no maintenance work is carried out in mature rattan gardens. At the time of harvest, paths are cut for access and to bring out the canes.

Diseases and pests

The plantations in Kalimantan appear to be relatively free from diseases and pests. In other plantations in southern China and Sabah and Sarawak (Malaysia), there have been few reports of any significant disease or pest problems. In Sabah, elephants have been reported occasionally to uproot young plants, and they also feed on mature apices. Rats, squirrels and porcupines can locally be very destructive to young plants. Small mammals are best controlled by a programme of systematic trapping. Rhinoceros beetle has been reported to destroy C. manan in plantations in Sabah (Lee); if such attacks become widespread there could be serious implications for rattan cultivation.

The only serious outbreak of a disease was that of leaf blight of C. trachycoleus in nurseries in Peninsular Malaysia (Norani et al., 1985). The various potential diseases and pests of rattans have been reviewed by Maziah et al. (1992).

Harvesting

Rattan is harvested mainly from the wild. Only a very small proportion is harvested from plantations in Central and South Kalimantan and in Sabah. There may be little control over the collection of rattans from the forest in many countries. In Malaysia, rattan collection licences for a particular forest area need to be obtained from the relevant authority.

Groups of 3 5 villagers go some distance into the forest on rattan collection expeditions. Collection of high climbing rattans is laborious, unpleasant, and sometimes dangerous because of the falling of dead tree branches dislodged in the process of tugging at the rattan. It is also wasteful, as the top portions of the cut stem may have to be left behind if they are still entangled in the forest canopy after the collector has climbed a neighbouring tree to try to free them. The mature stem, cut above the ground, is normally twisted around a tree trunk as it is dragged down, to rid it of the spiny leaf sheaths. The immature uppermost several metres of the stem are discarded, the stem cut into lengths of 2-3 m for large diameter canes, and of 5-7 m for small diameter canes. Then they are bundled and carried out of the forest to be transported to the processing site.

It is much less laborious and difficult to harvest slender canes.

Yield

The mature plants of most commercial rattan species in the wild are found widely dispersed and no reliable information is available on yield per unit area of forest. Neither is there any sound information on yield from plantations of large diameter canes, because none has yet reached harvestable age.

Variable yield estimates have been reported for smallholder cultivation of Calamus trachycoleus and C. caesius in Central and South Kalimantan. These have been summarized by Tan & Woon (1992). Because of the lack of standardization in reporting, the figures are difficult to compare. In general, C. trachycoleus is said to be ready for harvest 7 10 years after planting. Annual yields may be as low as 1 3.5 t/ha of green canes (Godoy & Tan, 1989), 2.2 3.9 t/ha (Priasukmana, 1989) or as high as 7 t/ha (Menon, 1980). The lower figures are probably the more reliable indicators of yield. C. caesius is said to be harvestable 9 10 years after planting, with annual yields varying from 3.5 t/ha (Menon, 1980), 5 7.5 t/ha of green cane (Tardjo, 1986) and 2.3 3.1 t/ha (Priasukmana, 1989).

Post harvest handling

In general, large diameter canes have to be cured with a hot oil mixture within 1-2 days after harvesting to prevent deterioration; this treatment is said to make the canes durable by removing gums, resins and water. The oil mixture may be of diesel, kerosene or coconut oil and the proportion varies from place to place. The curing bath tube is usually a trough made either of galvanized soft iron sheeting or of longitudinal halves of empty oil drums welded together. The trough of oil is heated by burning wood underneath and the canes are immersed in the hot oil (100-250°C) for 30 minutes or more. After curing, the canes are removed and rinsed or scrubbed with sawdust or gunny sacking to remove the oil from the surface. They are then placed upright against wooden frames in the open, or bundled and loosely tied at one end before being placed in wigwam like fashion to dry in the sun. The period of drying varies from 1-3 weeks, depending on the species and weather conditions. The dried canes are then bundled and stored on wooden racks in covered sheds to prevent contact with moisture before transport to buyers. Good quality canes may be fumigated with sulphur dioxide before sale. Poor quality large diameter canes are decorticated and stained before sale.

Small diameter canes are also dried soon after collection by laying them out on racks in the sun, and then sold whole. They may also be stored under water before being processed and dried. Processing involves scraping the nodes, fumigation with sulphur dioxide, drying in the sun, and then splitting or coring (usually by machine). From a length of cane 3-4 splits may be obtained. In coring, the outer layers of the cane are removed to provide the core. The outer layers, called "peel", are also sold. Canes of C. caesius and C. trachycoleus, which possess a highly silicified epidermis, usually have the silica removed before being split. In this process (known as "runti" or "lunti"), the canes are rubbed with sand or pulled through a series of bamboo bars or wooden pulleys so that the silica layer snaps off in glassy flakes.

Breeding and genetic resources

As yet no breeding of rattans has been carried out, and even primary selection of species for cultivation is in a very preliminary stage. Of the large number of rattan species, less than 25 have been cultivated for their stems, and only Calamus caesius and C. trachycoleus have been intensively cultivated. Yet there is a wealth of rattan species growing wild in the forest, some of which may be just as suitable for intensive cultivation as C. caesius.

There has been much research on the cultivation of C. manan in recent years. As this large diameter cane has stems of unsurpassed quality, it was an obvious choice for cultivation. Yet this species is single stemmed. There may be multiple stemmed large diameter species more suited to cultivation. Clearly there is a need to screen other wild rattans. Only recently has research begun to investigate differences between different provenances of the commercially more important species. Such work could become the basis for an active selection and breeding programme.

Prospects

Management of the wild resource

Unfortunately it is extremely difficult to control rattan harvesting. Even in well policed and controlled areas, high quality rattan stems continue to be harvested illegally. The value of rattan is now so high that any stem of a commercially important species is worth harvesting. Where logging roads have opened up areas previously difficult to reach, rattan pullers can go in and harvest rattan from a huge area. Even where licences are issued and royalties paid to forest departments, there is evidence to suggest that harvesting is carried out with little thought for sustainability. As the area of forest decreases due to logging activities, so pressure increases on the remaining rattan populations. This has been further exacerbated in parts of South East Asia by the introduction of export controls in some countries. Within the controlling countries, rates of harvesting may decrease initially, but elsewhere pressure has been enormously increased. Already much of the legally harvestable rattan in some parts of the region has been exhausted. Viable populations may survive in national parks, but even here they are collected, either legally (by forest dwelling people who may have customary rights to collect) or illegally. It would seem essential that rattans be strictly protected within nature reserves, and that more attention be given by forest departments to the control of rattan harvesting. One promising approach is the granting of long term rattan harvesting leases to give incentives to sustainable harvesting. It is essential to involve local people in developing rational harvesting strategies. The recently started demographic work on wild populations of rattan (Stockdale) may provide the basic data required for understanding possible rates of harvest.

As a conservation measure in countries like India, the following harvesting regulations have been stipulated to control the indiscriminate cutting of rattans:

• Only mature canes should be extracted from the clumps; immature or tender ones should not be collected or damaged.
• Digging of rhizomes or roots will not be permitted.
• No canes shall be extracted from outside the specified blocks.
• All the one year old canes plus six stems of the second year shall be retained in the clump.
• Clumps consisting of less than six stems shall not be worked.
• Felling should be done not less than 15 cm and no more than 30 cm from the ground level.

Cultivation

The successful cultivation of Calamus trachycoleus and C. caesius by smallholders in Central Kalimantan for over 100 years is an indication that cultivation of these two species can be carried out intensively on a smallholder basis, given suitable land and climate. The planting methods used by the smallholders were used and modified when setting up the first commercial estates in Sabah. Although growth rates have been astonishingly good in the new plantations, several important aspects of the cultivation, such as precise light and fertilizer requirements, spacing and line maintenance are not yet fully understood. Until large estates have reached maturity and have been harvested over several years, the commercial viability of such plantations will not be known, and rattan planting on such a large scale will be a risky, though promising, venture.

The cultivation of large diameter canes is less well understood than that of the two small diameter species mentioned above, and there are no records of successful smallholder cultivation of large diameter canes. At present the only large diameter canes to have received much research attention are the single stemmed Calamus manan and the multiple stemmed C. ornatus, C. merrillii, C. ovoideus and a few other more local species. No new commercial plantation of large diameter cane has reached harvestable age. Smallholdings of C. manan under rubber in Peninsular Malaysia show promise, but there are several aspects of the cultivation procedure that remain contentious, in particular the ability of rubber to support densely planted mature C. manan.

It has often been said that there are few pests of rattan in cultivation, but this may only be a reflection of the fact that large rattan plantations have only recently been established. A recent outbreak of rhinoceros beetle in plantations of C. manan in Sabah (Lee) has serious implications for all would be cultivators of this single stemmed species. Despite this, rattan cultivation is certainly attractive enough for rigorously managed companies to risk investing capital in the establishment of new large plantations. Although some rattan cultivation is undoubtedly successful, it cannot be assumed that the cultivation methods developed for one species in one habitat and climate can be applied in the cultivation of other species elsewhere. Much research needs to be done before untried species can be introduced into intensive cultivation.

Great potential exists for the development of rattan as a smallholder crop as part of social forestry schemes. Rattan requires tree support. It is already cultivated by villagers under fruit trees or in secondary forest near villages. Canes do not have to be harvested at fixed seasons; harvest times can be deferred if the selling price of cane is unattractive. Such cultivation augments villagers' income. Most importantly, rattan cultivation provides an incentive to leave forest cover intact and there are implications for the conservation of other forest organisms. SAFODA's Batu Putih Estate on the Sungei Kinabatangan in Sabah has large concentrations of wildlife including one of the densest populations of orang utan in Sabah (Shim) and Commonwealth Development Corporations's plantation in Sarawak contains about half the entire rattan flora of Sarawak in the wild state. For the smallholder, forest cover maintained for rattan support provides the habitat for a wealth of other minor forest products from "jungle meat" to medicine, and is essential for the maintenance of water catchments.

Trade

Rattan enjoys a still expanding international trade in furniture, matting and other manufactured goods. This external trade is estimated to be US$ 4 billion annually (World Resources Institute et al., 1985). A very conservative estimate of the total domestic trade is another US$ 2.5 billion (Manokaran, 1990a); this includes value of goods in urban markets and rural trade, and the value of rural usage of the material and products. In some way or another, of 0.7 billion of the world's 5 billion people use, or are involved in the trade of, rattan and rattan products.

The reduction in the forest area is resulting in loss of the resource in several producer countries. The resource base in some of the main producer countries is partially protected by bans on the export of the raw material; this also encourages the expansion of the domestic manufacturing industry. The increase in world population, expected to reach 8.2 billion by the year 2025 (World Resources Institute & International Institute for Environment and Development, 1988), is expected to lead to increased demand for the resource and finished goods. Research and development activities, especially into cultivation, have increased significantly during the last decade and are likely to increase further. The rattan trade seems set to continue to expand both domestically in producer countries and globally.

Research priorities and development

Although pioneering research on rattans, essentially taxonomic, goes back several centuries, it is only during the last 15 years that in response to periodic shortages of supply research on rattan has been initiated and pursued vigorously in several producer countries in the Asia Pacific region. During the last 15 years, with financial support mainly from Overseas Development Administration (UK), the International Development Research Centre (IDRC) (Canada), and the Food and Agriculture Organization of the United Nations (FAO), studies on taxonomy, propagation and utilization have been begun as national projects by research institutes, forest departments and universities in these countries. Today, these projects form a loose informal research network.

While a great deal has been accomplished over the past 15 years, much remains to be achieved. Research needs and priorities, discussed and summarized at recent international meetings, have been further elaborated on by Dhanarajan & Sastry (1989), International Development Research Centre (1989), Manokaran (1990b), and the International Fund for Agricultural Research (IFAR; see Williams, 1991) in an effort to gain support for the development of the rattan resource.

The priorities for rattan research and development are as follows:

1. Survey existing resources:

to establish the taxonomic and resource base and the rate of resource depletion;

to document and use indigenous knowledge about rattan;

to identify critical areas and under utilized species that could be brought into use.

1. Germplasm collection, storage, exchange and characterization:

to expand greatly the extent of living collections of rattan;

to explore the existing natural genetic diversity which is already at risk of depletion;

to screen lines for adaptability to various ecological conditions, suitability for cultivation, and utility of diversified products.

1. Development of propagation techniques:

to permit the large scale production of superior planting material for establishing plantations;

to overcome the acknowledged difficulty of obtaining adequate supplies of seed.

1. Investigation of technologies for plantation cultivation:

to identify and test cultivation and management techniques for cultivating rattan economically at village level and on a commercial scale.

1. Evaluation of domestic use:

to quantify the value of domestic (urban and rural) use and of employment generated.

1. Improved harvesting system, use and marketing:

to explore opportunities for developing appropriate techniques for harvesting and processing including post harvest protection, for improved use of added value products for domestic and international markets.

1. National policies:

to examine national policies covering the harvesting, use, marketing and development of the resources;

to examine quarantine laws for possible solutions to allow the exchange of propagules and germplasm.

The list of priority fields on which future rattan research and development could be focused may be expanded further. Supported by adequate funds for research and training, substantial gains may be expected in the development of the resource.