Overall view on the tradition of tapping palm trees and prospects for animal production

Palm trees have proved to be efficient converters of solar energy into biomass in most agro-ecological zones of the tropical world. Most tapped palm trees gives a sap very rich in sugar (10 to 20%). For several millennia, many species of palm trees (including coconut) have been used for sugar production. Highly sophisticated techniques of tapping were developed through the centuries in Asia, Africa and America. High yields of sugar were obtained from palms that could continue for up to a hundred years of production. One of the main constraints on production in recent times has been the increasing lack of fuel needed for processing palm sap into sugar and the price thereof. Nevertheless, since trials of feeding pigs with fresh sugar palm sap were successfully initiated in an FAO project in Cambodia, there has been renewed interest in tapping palm trees for sap to be used as feed. A thorough review of the literature has shown that intensive pig rearing based on palm sap has already been practised by the Indonesians for centuries and was found to be a very efficient system for intensifying agriculture in some highly populated islands. In today's economy, developing animal production using palm sap as the main source of energy in the diet looks very promising: the land could sustain higher population densities through the intensification of crop and animal production within sustainable integrated systems for small farmers.

Key words: sugar palm, tapping, sap, livestock, feed

Foreword

The recent documentation of the very high potential of the sugar palm (Borassus flabellifer) as a source of digestible energy for pig production (Khieu Borin and Preston 1995) has focussed attention on the palm species in general and their multi-purpose uses in particular. This review demonstrates the enormous amount of knowledge that exists on the tapping of palms and use of the sap, including the fattening of pigs done traditionally for centuries in certain islands in Indonesia. It is clear that most of the information is derived from indigenous sources and that palm cultivation has played a major role in poverty alleviation throughout the tropics. The neglect of this species by researchers, other than for monocultural export purposes (eg: coconuts, dates and palm oil), is a reflection of the reductionist thinking that has prevailed in agricultural science during the last 50 year or so years. It is hoped this review will revive interest in the palm species at a time when efficient capture of solar energy and environmental friendliness - characteristics shared by most of the palm species - are receiving priority as indicators of sustainable farming systems.

Introduction

Palms are believed to be among the oldest flowering plants in the world (Redhead 1989). For centuries, many palm species have been tapped throughout the tropical world in order to produce fresh juice (sweet toddy), fermented drinks (toddy, wine, arak), syrup ("honey"), brown sugar (jaggery) or refined sugar. One of mankind's first sources of sugar was probably Arenga pinnata (Redhead 1989). Evidence of the use of Borassus flabellifer sugar in India has been reported by the Greek historian Megasthenes, ambassador to the court of Chandragupta, in the 4 th century BC. Hindus knew how to extract it about 4,000 years ago (Ferguson 1888, cited by Fox 1977). Jaggery and treacle extracted from Caryota urens sap in Sri Lanka has been an important source of sugar from antiquity (Dissanayake 1977). In Africa, the main traditional use of palm sap is for wine production. It has been reported in Egypt (date palm) long before the birth of Christ (Barreveld 1993) and on the Guinea coast by early navigators in the 15th century (Sodah Ayernor and Matthews 1971).

Most tapped palm trees do not only produce sap but are multipurpose (edible fruits, building materials, fuel, fibres, wax, etc.) and their socio-economic importance can be critical for the rural poor: Ghandi used to call Borassus flabellifer, a remedy against poverty. A Tamil classical poem (Tala Vilasam) composed by Arunachalam in Tamil Nadu is entirely dedicated to the glory of this tree and enumerates 801 articles made from its various parts (Rangaswami 1977; Kovoor 1983). Another outstanding example is the coconut palm, for which every part is used. This tree is called in India "Tree of Heaven", "Mankind's greatest provider in the tropics", "Tree of life", etc. (Rangaswami 1977). There are at least 1,000 uses for the coconut palm (Dissanayake 1986). Nypa fruticans is also an outstanding provider of various products which are essential to everyday living and therefore is said to be the mangrove's counterpart of coconut (Quimbo 1991). Palm trees are also often associated with crops and pastures.

The scope of this paper is focused on palm trees that are traditionally tapped or with a potential for tapping in a non-destructive manner, guaranteeing sugar production in the long run within sustainable production systems. Sugar production from the fruits (as from dates of date palm) or from the starch accumulated in the pith of the stem (as for sago palm) is not reviewed here.

Rationale

Theoretically, the advantages of taking the sugars from the sap before it goes to the fruits are obvious. These sugars are intercepted before being used in the production of the non-edible parts such as husk in coconut, which represents 35% of the fruit (Rangaswami 1977), and in the production of edible material through chemical reactions which imply a loss, mainly a conversion of sugar into oil as for coconut and oil palm. It is therefore more profitable from the point of view of edible energy production to tap a palm for the sap rather than allowing the palm to produce fruits. Similarly, it was demonstrated that, in the context of harvestable energy from the coconut palm, the amount of energy harvested in the sap (through production of ethanol) could be 5 to 7 times higher than from the oil of the nuts (coconut oil is mixed with diesel fuel in a 5-10% blend as part of a national energy program in the Philippines) (Banzon 1984).

Comparing five sources of alcohol fuel, Hamilton and Murphy (1988) emphazise the interest of nipah in terms of yield (Table 1) and management: tapping is easy compared to coconut; the land is properly used and the environment protected as there are no wastes to get rid of as bagasse in the case of sugarcane.

Physiology

It is possible to obtain a sugary solution by the excision of the meristem in nearly all palms (Tuley 1965a). Basically, starch reserves from the trunk are converted to sugar and are transported upwards toward the stem apex (Fox 1977). Although this is true in the case of Corypha, other explanations are needed for palms such as coconut which does not accumulate starch in its trunk (Reijne 1948 cited by Van Die 1974). Pethiyagoda (1978) describes the upward stream as a watery liquid containing dissolved salts absorbed from the soil, and the downward stream as a comparatively rich mixture of food (principally sugars) manufactured in the leaves. The sap flow is intercepted by injuring fibro-vascular tissues of the apex or of the inflorescence. Nevertheless, this author recognizes that the large volume of exudate produced during tapping and the high sugar concentration clearly indicate that the material is drawn from stored resources and is in excess of currently synthesised sugars.

Van Die (1974) considers that the bleeding sap of Cocos nucifera and Arenga saccharifera, and probably some other palms too, may be regarded as the mobile aqueous phase of the sieve tube system of these trees, flowing to an artificial sink, the bleeding site. However, the rate of bleeding from a single inflorescence is several times higher than the rate of assimilate flow into a single bunch during fruit formation. The origin of the large flow of sap that occurs in a tapped tree is not yet clearly demonstrated. This is also the case for Borassus flabellifer where water from root absorption appears quite insufficient (Kovoor 1983). Nypa fruticans sap yield is higher in cloudy weather, and it is claimed that transpiration competes with sap yield but is partly compensated by variation in sugar content (Hinchy 1938a cited by Hamilton and Murphy 1988).

Pethiyagoda (1978) suggests that there is a steep rise in respiration which occurs whenever there is a rapid solubilisation and movement of materials from sites of storage to the points at which they are needed such as during seed germination, flower opening and fruit ripening. This phenomenon can be fostered, heightened and sustained by manipulative processes, the use of generally young growing sites (merismatic tissues) and the act of freshening the wound. Preliminary studies (not published) cited by Pethiyagoda (1978) show a considerably increased respiration by fragments of coconut inflorescence drawn from stimulated spadices.

In Cambodia, a recent study on Borassus flabellifer shows that there is no significant differences between days in the yield of juice and its Brix value, but differences for these two parameters are highly significant between farms, months and sex of the palm tree (for the Brix between male and female, the difference is significant instead of highly significant). During the production season, juice flow significantly decreases as the Brix significantly increases. Sap flow is higher during cool nights (Khieu Borin and Preston 1995; Khieu Borin1996).

Tapping is generally practised twice a day. For example, in the case of the coconut tree, the scars heal and the sap flow stops when slicing is not done for one or two days and 2 weeks would be necessary for restoring the sap flow (Levang 1988).

Location, products and tapped parts of palms

Annex 1 lists more than 30 different palm species that are traditionally tapped in parts of the tropical world. This was compiled from more than a hundred references on palms in the literature. Many references are several decades old as there is not much recent literature on this subject. The major part of the information was found on palms that are tapped in the Old World, with more or less as many different tapped species in Asia and in Africa. Very little literature seems to be available on tapping palm trees in the New World. In America and Africa, it seems that tapping palms has been practised exclusively or mainly for wine production, whereas in Asia the sap is used either as fresh juice or processed into a large array of products (wine, arak, sugar, vinegar, etc.). Annex 1 also shows that there are tapped palm species adapted to almost all agro-ecological zones of the tropical world from tidal areas and swamps to deserts and mountains.

Sap composition

Most tapped palm trees (Table 2) give a sap very rich in sugar (10 to 20% according to species and individual variation) whereas the maple tree, tapped for sugar production in North America gives a sap containing only 3% of sugar (Annett 1913). Gibbs (1911) cited by Van Die (1974) has reported a striking resemblance in dry matter, sucrose, and ash content of the bleeding sap from the palm species investigated by him (Cocos nucifera, Arenga saccharifera, Nypa fruticans and Corypha elata). Sugars from palm trees such as Phoenix sylvestris and Borassus flabellifer are reported to be more nutritious than cane sugar (Rangaswami 1977; Roy 1951 cited by Morton 1988). Some medicinal value is also often reported: it is the case for the sugar from Borassus flabellifer (Rangaswami 1977). In Madagascar, Cocos nucifera sap is used against nephritis and bladder infections.

Borassus flabellifer fresh sap is a good source of vitamin B complex (Chopra et al 1958 cited by Morton 1988) and it contains ascorbic acid (Dissanayake 1986). 100 litres of sap give 7-8 kg of sugar and 8 kg of molasses (Paulas and Muthukrishnan 1983a cited by Davis and Johnson 1987). As the sap from the morning collection has been produced at lower temperatures than the one collected in the evening, it is less fermented and therefore contains more sucrose (Romera 1968). A recent study in Cambodia showed that sucrose can range from 66 to 94% or from 51 to 81% of total solids in the juice respectively in January or in April. Glucose and fructose levels increased during the same period (Khieu Borin and Preston 1995; Khieu Borin 1996). In the case of Borassus aethiopium, sap composition of the morning collection markedly differs from the evening collection and also sap composition differs on the level of cutting of the terminal bud, losing its taste when its base is reached; the sap also contains less sucrose than Borassus flabellifer sap (Port�res 1964). In Burma, the tappers distinguish up to 6 different qualities of sap according to the tapping stage and the sex of the tree (Lubeigt 1979).

The Caryota urens sap also consists almost entirely of sucrose which is rapidly fermented and inverted (Dissanayake 1986). Its pH is 6.5 and it contains traces of acids, 0.34% of reducing sugars and a total sugar between 15-16% (Theivendirarajah et al 1977). The jaggery has the following composition: 76.6-83.5% sucrose, 0.76-0.9% reducing sugar 1.65-1.98% ash 1.79-2.27% protein and 6.6-8.34% pectin gums (Dissanayake 1977). Fresh sap of coconut has a pH equal to 6 and a Brix of 17 (Naim and Husin 1984). It contains 12 to 18% sugar (mainly sucrose and small amounts of glucose, Norris et al (1922) cited by Grimwood 1975), 0.5% ash, 2% organic solids (Nathanael (1960) cited by Grimwood 1975) and small amount of vitamins, such as 16-30 mg of ascorbic acid per 10ml (Banerjee (1935) cited by Grimwood 1975). Coconut sugar contains 0.5 to 1.3% protein, 72 to 88% of total sugar, 57 to 81% sucrose 1.3 to 2.5% total minerals and 5 to 10% humidity. The fresh sap of the African oil palm contains 9.6 to 10.6 % of sucrose. Glucose and fructose are much below 1% (Eze and Uzoechi Ogan 1988). Sprecher von Bernegg (1929) cited by Van Die (1974) reports 14% total sugars. The sap is also rich in vitamin B (particularly B12) and in sulphur-proteins (Hartley 1977; Tuley 1965a). The sap of Phoenix sylvestris is a good source of vitamins of the B group and contains in addition an appreciable amount of ascorbic acid (Rangaswami 1977).

Management of tapped palm trees and yields of sugar

The management of palm trees for sap production varies very much according to species (See Annex 2). Nypa fruticans, Phoenix sylvestris, Elaeis guineensis, Raphia hookeri and Cocos nucifera can be tapped at a rather early age, respectively when the trees are 4, 5, 6, 7 and 7 years old (Crevost and Lemari� 1913; Abedin et al 1987; Essiamah 1992; Profizi 1988; Levang 1988). On the other hand, many years are needed before tapping Caryota urens (10 to 15), Borassus flabellifer (15 to 30) or Corypha elata (20 to 100) (Redhead 1989, Fox 1977).

The number of years a palm tree can be tapped is also very different depending on the species. Corypha elata and Raphia hookeri flower just once. They will produce sap only for a few months before dying (Fox 1977; Profizi 1988). Arenga pinnata and Caryota urens will produce sap for several years, with large interruptions in the case of Caryota urens as it flowers only every two or three years (Redhead 1989; Dissanayake 1977). Other palm trees will produce sap for much longer periods: 10 to 15 years for Elaeis guineensis, more than 20 years for Cocos nucifera, 50 years for Nypa fruticans and Phoenix sylvestris and 30 to 100 years for Borassus flabellifer (Adand� 1954; Levang 1988; Magalon 1930; Abedin et al 1987; Lubeigt 1977). Some species are able to produce sap all year round: Arenga pinnata, Cocos nucifera, Elaeis guineensis and Nypa fruticans (Mogea et al 1991; Rangaswami 1977; Tuley 1965a; Kiew 1989). Borassus flabellifer and Phoenix sylvestris produce only seasonally (Crevost and Lemari� 1913; Annett 1913).

Most tapped palm trees gives a sap very rich in sugar (10 to 20% according to species and individual variation). The yields are highly variable according to the species and their management. In the case of Borassus flabellifer, the female trees can give as much as 50% more sap than male trees (Lubeigt 1979). Under proper management, the main tapped palm species (Arenga pinnata, Borassus flabellifer, Cocos nucifera and Nypa fruticans) can reach yields of about 20 tonnes of sugar per hectare (Van Die 1974; Watson cited by Kiew 1989). Compared to sugarcane production (5-15 tonnes of sugar/ha/year), the Borassus flabellifer tree can reach 18 tons/ha/year under rain-fed conditions (Khieu Borin and Preston 1995; Khieu Borin 1996) and the coconut tree 19 tons/ha/year (Jeganathan 1974). According to estimates, Elaeis guineensis produces much less sugar (1.2 tonne per hectare, Udom 1987) but, as it has never been exploited for sugar production but only for wine production, there are good prospects for obtaining much higher yields in a production system oriented towards sugar production.

Methods of palm tapping and sap preservation

The techniques for tapping palms are numerous and can vary drastically from one continent to another, as demonstrated by the case of Borassus aethiopium in Africa and Borassus flabellifer in Asia. Refined techniques of tapping the inflorescence of the latter are compatible with production in the long term. Destructive techniques are usually practised on the terminal bud of B. aethiopium and are often responsible for the death of the tree within a few months. As tapping is mainly done in Africa for wine production, the use of the Asian technique in Africa would permit sap production throughout the rainy season whereas palm wine is drunk much more during the dry season (Port�res 1964). The African oil palm is used in Africa for producing wine mainly through two different techniques: one is destructive (incision of stem apex of felled palm) and is preferred in Ghana; the other is not destructive (excision of male inflorescence and sometimes of female inflorescence as well) and has been developed where economic considerations have forced the people to preserve their palms, e.g. in eastern Nigeria (Hartley 1977).

Tapping can be practised after cutting down the trees. Otherwise, except for Nypa fruticans, which is trunkless and develops its inflorescence at a height of about 1m (Hamilton and Murphy 1988), other palm trees have to be climbed for tapping as their terminal buds and inflorescences are located at the summit of their trunk which is often over 10m high. The excision of the terminal bud of standing trees is quite harmful since tapped palms never resume vigorous growth. If the terminal bud is only perforated, then the trees will show malformation in subsequent leaves, flowers and trunk growth (Kovoor 1983). Nevertheless, it has been observed that multi-stemmed trees such as Hyphaene coriacea and Phoenix reclinata in south-eastern Africa generally recoppice after tapping, although tapped stems die unless tapping is stopped before the apical meristem is totally destroyed (Cunningham 1990). The very low yields of sap from these trees are interpreted as a result of over exploitation. Cunningham (1990) suggests that if palm size classes shifted to the extent that there was again a high proportion of mature fruit-bearing palms in the population, then inflorescence tapping could be practised as Hyphaene coriacea is an annually flowering palm.

The most advanced method of tapping is that applied to the inflorescence spadix which guarantees a high yield for long periods without affecting the well-being of the tree. It only entails a sacrifice of a bunch of fruit in the case of tapping female inflorescences. Tapping the inflorescence is practised throughout S.E. Asia on all species of tapped palm trees (Kovoor 1983). Two features are common in tapping: manipulative treatment or preparation (application of chemicals and substance of plant origin, twisting, distortion, kneading, pounding, bruising, beating or tapping) necessary as a prelude to copious and sustained sap flow, and renewing the exuding wound by shaving off a thin slice of tissue once or twice a day (Pethiyagoda 1978). Tapping is an art: sap yields depend on the skills of the tapper (Khieu Borin and Preston 1995; Khieu Borin 1996; Coconut Research Institute 1967).

Once extracted, sugarcane or sugar beet juices are immediately processed for sugar production. In the case of palm sugar, as the sap flow is slow, intermediate storage before collection is necessary: the juice is left for a minimum of 10 hours at temperatures around 30 ⁰C, with high hygrometry and high contaminations with natural yeasts, bacterial and fungal spores (Romera 1968; Pethiyagoda 1978). Usually, if the sap is used for sugar production, it will be collected twice a day as fermentations have to be avoided as far as possible. However, the inclusion of an anti-fermenting agent is usual (Kovoor 1983): traditionally lime and bark or leaves (rich in tannins) from various tree species are used. The vessels are regularly cleaned and rinsed with water and sometimes smoked. Sugar is produced by boiling the sap to evaporation until a sufficiently thick syrup is obtained out of which sugar will crystallize on cooling (Kovoor 1983). A study conducted on Borassus flabellifer and Caryota urens sap by Joachim et Kandiah (1938) cited by Dissanayake (1986) shows that limed pots inhibit fermentation and inversion of sucrose better than smoked pots; the latter technique is effective only for a period of 3 to 4 hours after collection from the tree.

Methods of palm tapping and sap preservation for the main species:

There are significant differences between the methods practised by the different tribes (Mogea et al 1991). Tapping is generally done on the male inflorescence spadix (Redhead 1989; Friedberg 1977). In fact, the sap from female inflorescence is of inferior quality and the spadix requires more effort during preparation (Mogea et al 1991). A wooden mallet is used for slowly hitting several minutes a day the trunk beneath and above the stem of the inflorescence while swinging it before incision (Friedberg 1977). The tapper cleans the male inflorescence stalk from its bracts as soon as the flowers are nearly open. The clean stalk is then slowly beaten for several minutes each day for about 2 or 3 weeks at a distance of about 45-60cm from the stalk base, until the stalk becomes swollen; later the stalk is cut at the swollen section. The sap is collected in a bamboo pipe. In every tapping, the apex of the stalk is sliced 1 to 5mm depending on the pre-treatment. Another method consists of splitting the stem of the inflorescence into 2 parts and in removing secondary stems except the last four that are grouped into a bamboo recipient where the juice will fall; the four secondary stems extremities are cut and then bruised with wood pliers twice a day during 5 days; the two following days, they are put in a bamboo recipient containing one litre of water, the next two days, they are bruised again and then the production starts (Crevost and Lemari� 1913). The simplest method consists in inserting a bamboo pipe into the base of the male flowers and to collect the sap into a gourd (Redhead 1989). The closer to the ground the male inflorescence arises, the smaller usually the amount of sap it produces (Mogea et al 1991). In some places of Sulawesi (Indonesia), Arenga pinnata and Arenga wightii are cut down, the leaves are lopped and the trunk is sliced at intervals (Kovoor 1983; Pethiyagoda 1978).

Arenga pinnata tapping is a very labour intensive activity. It is done according to a very regular working schedule for the generally twice-a-day activity. Beating and preparation of 20 inflorescence stalks (of which 12-16 will produce each for 4 months) takes some 2.5 hours each day. One healthy person can tap 12-16 trees on the basis of an 8 hour working day, or more if the trees are close enough to go from one to the other by jumping. On the average, a person taps about 10 trees and 10 workers are necessary per hectare (Mogea et al 1991). Tapping must be done daily otherwise the sap flow will diminish rapidly as tissue healing occurs. To slow down sap fermentation, the container is sterilized with the smoke of a wood fire (Mogea et al 1991).

In Banfora (Burkina Faso), the technique involves the lateral portion of the young stem and does not kill the tree (Redhead 1989). A hole (10cm diameter) is made above the leave crown in order to reach the extremity of the terminal bud which is essentially made of soft tissues of the young leaves. The terminal bud is perforated superficially and horizontally and the sap is collected in a calabash through a gutter made of a young folded leave put at the periphery of the terminal bud (Port�res 1964). A heap of leaves is put on the entry of the hole to protect the cut bud from desiccation and insects. The perforation is refreshed twice a day as long as the tree looks healthy. Once the tree looks exhausted, the perforation is left to heal and new leaves will come but the trunk will remain stunted and deformed (Bellouard 1950).

In Ivory Coast, in the past, male trees only were tapped as female trees were preserved for guaranteeing regeneration. A tree with a marked bulge and at the end of its development (at a height of 6-8m on the trunk) is preferred as it ensures regularity, good quantity and quality of the sap. A "ladder", made of rattan creeper, bamboo or rachis of Raphia sp, is tied to the north-east side of the trunk. A "safety belt" made of rattan creeper is also used by the climber while working at the top of the tree. The leaves are cut except 2 to 4 six-month old leaves on the south-west which serve for drawing the sap upwards and 2 or 3 petioles are also left to serve as a seat. One or two top young leaves of the terminal bud are preserved. One or two days later, a lateral window is made in the north-east petiole in order to assess the level of the basis of the terminal young leave. The terminal bud will be sectioned 10cm below the junction petiole-limb of the more internal terminal leave (neither too low to avoid shortening total period of production, nor too high as it will delay profuse sap flow). All young leaves above this point are delicately cut without damaging the terminal bud. The window is enlarged and the terminal bud sectioned with a slight slope towards the entrance of the hole. It is closed and protected to avoid insects and desiccation. The following day the terminal bud is cut again with more slope towards the window. A gutter made from a leave is put and goes to a calabash and the protection is put again. Then twice a day, the terminal bud is cut 2-4mm further and the sap collected (Port�res 1964).

In other African regions (as in Senegal), the trees are often tapped to death after several months. Used for the leaves during the first years, the trees are then tapped once they have reached 2-3m high. The heart is reached with a knife at the level of the terminal bud. A hole, 30cm long 15cm broad and 10cm deep is made. From 1 to 6 litres of sap (there is more collection of sap during the cold months and when there is fog) will be collected every day during 4 to 6 months (Bellouard 1950; Giffard 1967). Another technique consists in bruising and cutting bunches which is harmless for the trees (Chevalier 1930).

Various methods are used to climb the tree (six recorded by Kovoor 1983), using ankle-loops, aerial ropeways between trees, hoop-belt, riveted bamboo, mobile 4-9m long ladders and fixed ones on the upper part of the trunks, notches in the trunk, etc. Tapping techniques of Borassus flabellifer are similar to that of coconut except for slight differences. Sap is collected twice a day in all cases. In some countries (Cambodia), different tongs are used according to the sex of inflorescence and stage of tapping (Kovoor 1983).

In Sri Lanka, four different methods are used according to the sex and age of the inflorescence (Kovoor 1983):

In Upper Burma, the 8-month duration tapping season consists in 3 tapping phases which also include specific operations such as four different tapping techniques, two for both male and female trees (Aubert 1911). In Central Burma, sap collection can be continued all year round but with low yields between November and January (Lubeigt 1977). In Cambodia, if a tree starts giving many inflorescences at the same time, some of them (1 to 7 according to tree, climate and location) will receive the usual treatment while the others will be sliced and crushed for 5-8 days and then preserved (3 to 5 months) for juice collection some months later (Khieu Borin and Preston 1995; Khieu Borin 1996). Clay pots, bamboos internodes, calabashes or leaf-buckets are used for sap collection.

Another tapping technique is as follows. The inflorescence is bruised in its particular stage of growth through beating and crushing it with wooden tongs. Then it is bound into "torches" and tapped by cutting off the bandaged tips (Redhead 1989). A mallet is used for slowly hitting several minutes a day the trunk beneath and above the stem of the inflorescence while swinging it and finally pinching it before incision. A new incision is made after every collection (Friedberg 1977). The method described by Crevost and Lemari� (1913) for Arenga pinnata is also practised on Borassus flabellifer.

In order to slow down the fermentation of the sap, the vessels are smoked and coated with lime (Redhead 1989). Bark or leaves from different species are also used: Schleichera oleosa bark or leaves (Friedberg 1977; Kovoor 1983), Shorea cochinchinensis bark in Cambodia (Crevost and Lemari� 1913; Kovoor 1983), Shorea obtusa in Burma, Shorea talura in Thailand (Lubeigt 1977), Votica hermandiana bark (Magalon 1930), Launaea coromandelica dried bark, Anacardium occidentale leaves (Kovoor 1983), Vateria acuminta bark or Cyminosma pedunculata (Dissanayake 1986). These pieces of bark are finely stamped and added to each receiving bucket (Fox 1977). Tkatchenko (cited by Kovoor 1983) has found that 5 to 7 g of lime per litre conserve the sap quite satisfactorily for more than a day against fermentation. In Roti island (Indonesia), the tapper uses for every producing pair of rachillae a set of two leaf-buckets. Twice a day, he brushes the inside and cleans with water the leaf-buckets from which he has just collected the sap and which are going to dry as he places the other leaf-bucket of the pair for the following sap collection. This is reputed to slow down the fermentation of the fresh juice without changing its taste as would do lime or bark additives (Fox 1977). In Upper Burma, the earthen vessels used for sap collection receive special treatment. Before being used for the first time, they are washed and exposed, still wet, with their openings towards a fire made of dried branches and leaves for about 10 minutes. Then they are changed twice daily after every sap collection. They are brought back to the house and washed and dried and just before replacing them on the tree, a few chips of the bark of Shorea robusta are put in the vessel (Aubert 1911).

The inflorescence in its particular stage of growth is bound into a "candle" form and tapped by repeatedly slicing off the end of the candle (Redhead 1989). The bruising technique is similar to the one used for Borassus flabellifer (Magalon 1930). In Sri Lanka, the technique has been described in details by Dissanayake (1977): "When the flower appears, it is tapped before it reaches maturity. The art of tapping a Caryota urens flower is very much a village craft requiring a high degree of traditional expertise. A long bamboo pole is erected along the trunk with suitable support to reach the crown. The selected flower is prepared for tapping first by beating it with a stone or wood and then tied with a string in several places to keep it in proper shape. A concoction made of herbs, ash and salt is then applied to the end of the flower, where a fine angular slash is made. Tapping begins 3-4 days after this application; a big earthen pot is placed at the end of the flower to collect the sap."

The most common methods used in Sri Lanka to prevent fermentation of the sap are: lining the inside of the collecting pot with fresh lime; placing Vateria copallifera bark, Careya arborea bark or Achronychia laurifolia leaves in a clean pot before it is used for collecting sap (Theivendirarajah and Jeyaseelan 1977). Studies showed that lime is the most effective, next Vateria copallifera bark, followed by Careya arborea bark and Achronychia laurifolia leaves; Vateria copallifera bark could be used as effectively as lime to preserve sweet toddy for 39 hours (Theivendirarajah and Jeyaseelan 1977).

In Sri Lanka, segments of coconut husk are generally fastened all the way up the trunk for climbing the coconut tree; a loop round the ankles is used sometimes (Coconut Research Institute 1967). In south Sumatra, simple notches are carved in the trunk. The inflorescence in its particular stage of growth is tapped (Redhead 1989). The development of female flowers inside the spathe (about 60cm long) causes a swelling at the basis of the spathe. The appearance of the swelling is taken as the correct stage for tapping. The inflorescence selected for tapping is first trained, by a gentle uniform beating all over the spathe, twice a day, in the morning and in the evening, the spathe being wound around with a strong coir or coconut fibre string to prevent it from bursting. After a week's beating, about 7-10cm of the tip of the spathe is cut off. On the third day, the cut end is carefully pounded (Coconut Research Institute 1967). Beating is continued for about a week, while an earthen pot (in Sri Lanka) or a bamboo tube (Thailand) is hung from the spadix to collect the oozing liquid from the cut surface (Grimwood 1975). The free end of the spadix is gradually bent down (Coconut Research Institute 1967). If the trees are tapped for sweet juice, they are collected in the morning (Rangaswami 1977). The sap generally begins to drip after 12-15 days, or up to 35 after the beginning of the tapping process according to the tapper's skill, seasonal conditions and nature of the palm. Then slicing about 2mm and collecting sap are done twice a day while tapping and bruising operations are discontinued (Nathanael 1970, cited by Kovoor 1983; Coconut Research Institute 1967). Nevertheless, light tapping of the end of the spathe with a mallet may be continued and the sliced surface of the spathe is sometimes smeared with a mixture of bruised leaves which contains saponin and stimulates the sap flow (Child (1964) cited by Grimwood 1975). A single spathe will be tapped until it is reduced to a stump about 10-15cm in length. About three weeks before reaching this point, another spathe is prepared in order to ensure continuity of sap production without any break for a period of one year on good palms (Nathanael (1970, cited by Kovoor 1983).

According to Redhead (1989), coconut sap to be used for jaggery is collected in the mornings only, in clean vessels containing a fermentation inhibitor such as lime. In Thailand, the bark of Vateria acuminata or Shorea talura may be added. The sliced surface of the spathe might also be daubed with lime (Levang 1988). In Indonesia, the earthenware pots are sterilized in a naked flame and about a teaspoonfull (5 g) of finely cut chips of Vateria acuminata bark is added (Purnomo 1992). In Malaysia, a quarter of the container is filled with a mild solution of lime water together with dried latex from mangosteen bark (Naim and Husin 1984).

The entire growing point of the terminal inflorescence is excised just before the development of the inflorescence starts (Redhead 1989; Kovoor 1983; Van Die 1974). A shallow depression is scooped out at the top of the naked stem, and the juice is collected from the depression (Redhead 1989; Kovoor 1983). In Roti (Indonesia), the technique consists in breaching the inflorescence (Fox 1977). Incisions of the flower stalk are also reported (Pethiyagoda 1978).

The most common method in West Africa, and particularly in Ivory Coast and Ghana is destructive. Just before the fructification, the young tree is felled down by loosening the roots with a spade and pulling down the tree. For older trees, the trunk is cut at 30-60cm. All fronds in the merismatic region are removed after one or two weeks. A cavity is made in the terminal bud. It is either a rectangular well about 7.5cm deep with a round hole bored in its floor or a bowl-shaped hollow 10cm x 9cm, about two-third (15cm) the diameter of the stem in depth. A brand is put in it in order to activate the sap flow and reduce the infestation of the well by insects and their larvae, bacteria, moulds and yeasts. Collection of the sap starts 2 to 4 days later, using a bamboo gutter (15cm long) which goes into a container (calabash) under the stem. It is done twice a day: each time, the tapper cuts away from the terminal bud a 1 mm layer of tissue and flares with fire the stem surface around the tapping wound. The sap flow lasts 1 to 2 months (Blanc-Pamard 1980; Kovoor 1983; Essiamah 1992; Hartley 1977; Sodah Ayernor and Matthews 1971). A less destructive method is practised in Nigeria (Igbos) and consists in piercing the stem of the standing tree up to the bud (Tuley (1965) cited by Kovoor 1983). A detailed description is given by Essiamah (1992). Several leaves are cut off the side of the stem. A rectangular hole (8cm x 5-10cm) is cut in the stem to a depth of about 30cm or more according to the size of the palm tree. A suction-pipe made of bamboo is used to collect the sap into a jug hung on the stem. As the sap flow is very intense, the tapper must climb the tree 3 times a day to collect it. Each time, he cuts away a thin layer of tissue from the wound surface in a downward direction. The length of the rectangular hole can thus extend to 30cm or more according to the duration of sap flow. The tree dies 2 or 3 weeks after being tapped. In Madagascar (Sakalava), the terminal bud is also incised (Decary 1964).

A non destructive method consists in tapping the male inflorescences (Sprecher von Bernegg 1929, cited by Van Die 1974). It is especially practised by the Jorubas in Nigeria, but it is also found in Benin and Ivory Coast. It has been described in detail by Essiamah (1992) and Sodah Ayernor and Matthews (1971). All fronds below an unopened male inflorescence (before the differentiation of the flower spikes starts) are cut off and the leaf subtending it is removed to obtain access to the inflorescence enclosed in its spathes. About 7.5cm above the base of the male inflorescence, a neat triangular hole is made in the front spathe (about 2.5 to 6cm deep in the main stem of the spadix and 2.5 cm on the sides). The hole is then covered and the following day a funnel of small bamboo (16cm long and 2cm in diameter) is inserted in the cut in such a position that the sap flows into calabashes (vessels). To ensure continuous flow, the hole is extended downwards by removing a daily slice of about 3mm. The tapping is stopped when the base of the male inflorescence is reached. The sap is collected morning and evening and a new slice taken at each collection. Palm trees tapped in this manner set fruit one to two years after tapping (Sodah Ayernor and Matthews 1971). In Casamance (Senegal), the Diola use a similar technique but both male and female inflorescences are used. While preparing the inflorescence, its stem is covered with a mattress of leaves which is struck vigorously with the head of an axe in order to stimulate the arrival of the sap (Adand� 1954).

The technique is described in details by Cunningham (1990). Palm clumps are selected and large stems within these clumps are prepared for tapping after burning each clump to remove undergrowth and leaf spines. Selected stems are trimmed to initiate sap flow using a razor-sharp bush knife whetted against a stick kept for that purpose. The stem and young leaf bases are cut at a slight angle to guide sap onto a leaf gutter and into the clay collection container. By the end of the tapping period, these young leaf bases have been removed by trimming 2 to 3 times per day over a 5- to 7- week's period. Therefore the terminal bud is cut to death of the stem but the clump survives (Cunningham 1990).

In Djibouti and Eritrea, the terminal bud is often cut to death of the stem (Jahiel 1993; Fanshawe 1966). As Hyphaene thebaica is a multi-stemmed palm, this practice does not mean always the death of the clump. There is also a less destructive technique. The selected stem is cleaned from all leaves. Then only the bases of the future leaves are cut but not the terminal bud itself. New cuts are practised up to 4 times a day for a couple of weeks and generally stopped before reaching the terminal bud. Some tappers will reach this terminal bud in order to get more sap before having to prepare another tree (Audru 1985).

The tree is first felled, the leaves lopped and the trunk sliced daily over a period of six to eight weeks during which the bleeding sap is collected (Pethiyagoda 1978). Darwin (1860, cited by Kovoor 1983) reported yields of around 400 litres.

The palms are cut down, defoliated and scorched for increasing sap flow. The sap is tapped from the inflorescence stalks or by holes made into the trunk during the development of inflorescences (Kovoor 1983; Pethiyagoda 1978).

The stalk of the inflorescence is cut at a proper stage (Redhead 1989). Nevertheless, pre-treatment of the stalk is possible at various stages of infructescence development from flowering to mature fruiting stages. Plants at more mature stages require longer periods of pre-treatment than at young stages. Kiew (1989) characterizes the proper stage as follows: the fruiting head is well developed but has not begun to darken. Spathes and bracts at the base of the fruit head are cleaned off (Davis 1988). The fruiting head is first oscillated gently, then with increasing severity until after 3 weeks, it is subjected to violent shaking (Watson 1928, cited by Kiew 1989). According to Dennett (1927), this operation prevents the cells from hardening and atrophying as the stalk matures. The stalk keeps soft and supple compared to a non-treated stalk. Studying the labour costs, the same author found that 3 weeks was the most profitable duration for this treatment. The bruised inflorescence is tapped by light hammering and by incising the tissues above the bruised portion as close as possible to the young fruit. The almost full-grown fruit-head is cut away (Hamilton and Murphy 1988). Each day after collecting the juice, a thin slice of about 2mm is removed. Cutting tools are disinfected since flow is reduced by bacterial growth (Hamilton and Murphy 1988). Sap flow depends mainly on prior preparation of the stalk: if none is done, flow will be low and will last only a few days. The base of the stalk is kicked in most countries where this tree is tapped. In Papua New Guinea, a sophisticated protocol has been developed by P�iv�ke (1985): bending the stalk 12 times in one direction; patting the length of the stalk backwards and forwards with the bare hands 64 times; kicking the base of the stalk 4 times; and repeating this 4 times a week. Yields of 1,800 ml/day/tree over a month were reached in intensively treated younger-aged stands whereas only 155 ml/day/tree were produced when the treatment was done only once a week (P�iv�ke 1985). In the Philippines, the schedule of kicking lasts for 3 months and 10 days per palm: once a week during the first month after flowering, twice a week during the second month, once every two days during the third month and finally daily for 10 days (Melana (1980, cited by Hamilton and Murphy 1988). Tapping life depends on the length of the stalk which can vary a lot: 0.6 to 1 m in the Philippines, 1.4 m in Sumatra and 1.9 m in Papua New Guinea (Hamilton and Murphy 1988). To slow down the very rapid fermentation of the sap, the bamboo recipients (2 litres capacity) are washed with lime and water (Redhead 1989; Crevost and Lemari� 1913).

The entire growing point is cut off and a shallow depression is scooped out at the top of the naked stem. The juice is collected from the depression (Redhead 1989; Kovoor 1983). Another technique consists in cutting off the fronds: exudation of sap occurs in the cavity created by defoliation (Pethiyagoda 1978). Tapping a date palm is a severe intervention as it deprives the palm of most of its leaves and food reserves; great care is required during the twice daily scarring, because if it is carried on too far, the palm will die. Every 20 days or so, the cone, canal and spout are readjusted. After tapping is completed (up to three or four months), three to four years are needed for the tree to be able to bear a full crop of fruit again and five years before tapping it again (Barreveld 1993). Using the Indian method of tapping for Phoenix sylvestris on Phoenix dactylifera, Barreveld (1993) reported inferior daily yields but considering that annual tapping would be practised in this case, there would be higher yields of sap per tree over a period of 3 to 4 years; nevertheless, labour hours involved per unit sap harvested is definitely higher.

In Ivory Coast, the tree is cut down and uprooted, all leaves are removed and the terminal bud is cut. A cavity is made in which a brand is put in order to activate the sap flow. Collection of the sap starts 3 to 4 days later through a gutter linked to a calabash. Collection is done twice a day, each time, the terminal bud is further cut. It lasts 1 to 2 months (Blanc-Pamard 1980). Another technique consists in incising the lateral portion of the young stem (Redhead 1989). In south-eastern Africa, the same technique as described by Cunningham (1990) for Hyphaene coriacea is used.

The stem is pierced up to the terminal bud (Kovoor 1983). In Bangladesh, the oldest leaves are removed at the end of October on one side of the palm tree along about 35cm. The bases of the petioles and the sheaths are carefully removed. A triangle piece of bark (point downwards) is cut but great care must be taken not to expose the sap-supplying inner zone. A week later, as the fine covering of soft tissues gets a little hardened and begins to crack, this covering is removed without damaging the inner zone. Then comes a rest of 12 to 14 days after which a lateral incision is done in the triangle. The sap exuding from the scooped surface is run through a bamboo spout into a vessel. A new incision is made the 2 following days and then there is a 4-day period of rest and the same cycle is repeated until the heart is reached. The following year, the same treatment is done on the other side of the tree which gives a zigzag configuration to the trunk; tapping can be practised every year for several decades (average of 25 years) (Fealy 1925; Annett 1913; Barreveld 1993). In order to slow down the fermentation of the sap, the earthen vessel are coated with lime; in Bangladesh, they are well smoked every morning (Annett 1913).

In some parts of Nigeria, the palms are cut down and scorched for increasing sap flow (Kovoor 1983). The sap can also be collected from the upper stem pierced to death as it is currently practised in West Africa; the method is the same as the one described for oil palms by Igbos in Nigeria (Tuley 1965, cited by Kovoor 1983; Essiamah 1992). The tappers extract the sap just before the appearance of the spadix and tapping is done on flowering trunks (Profizi 1988). In Cameroon, the base of the tree is split before fruit production (Fyot 1973).

Multipurpose uses

As described in Annex 1, most palm trees have multipurpose uses (see also the review by Johnson 1997). These uses are not always compatible. Sap production is at its maximum just before or during fruit formation, when starches in the stem are converted into sugars which enter the sap and flow to the site of fruit formation. Tapping the tree competes with the production of the ripening fruit (Redhead 1989). Nevertheless, some species can be exploited for both fruit and sap, but the trees might die prematurely in the case of over-exploitation (Redhead 1989). Tapping Borassus aethiopium impedes fruit production and is detrimental to the quality of the wood; collecting young leaves reduces sap yields (Bellouard 1950). In Madura island (Indonesia), pruning the Borassus leaves twice a year will drastically reduce the juice-producing life of the tree from 30-40 years to 8-10 while moderate pruning will still permit 20-30 years of juice production (Fox 1977). Nypa fruticans immature leaves are cut for cigarette papers, mature leaves for thatch but harvesting leaves reduces the yield of sugar (Kiew 1989). In Ivory Coast, the African oil palm is tapped before fructification which kills the tree within one or two months (Blanc-Pamard 1980).

Tapping can also stimulate fruit production: a young coconut palm tapped during 6-12 months for sugar production will then produce more nuts (Magalon 1930; Fealy 1925). This has been clearly demonstrated by Mathes (1984): low (in regard to nut production) yielding palms, after tapping, had a fruit production three times that of the high yielding palms, and double that of the very high yielding untapped palms. A technique called sequential coconut toddy and nut production has been developed in the Philippines at the Davao Research Centre. The first half of the spathe is tapped and the second half is left for fruit production as female flowers that develop to mature nuts are situated in this lower portion. Nut and copra yields are about 50% lower than non-tapped palms; however, this technique has been demonstrated to be very feasible and highly profitable for small producers (Maravilla and Magat 1993).

Arenga pinnata can be tapped when they are between 12-15 and more than 30 years old; then they can be cut for sago production (Sumadi 1988). Nevertheless, in West Java, where sago is obtained from trees 10-12 years old, no tapping will be done previously, farmers arguing that it would reduce the quantity of starch in the trunk (Mogea et al 1991). In Eastern Nigeria, oil palms that have been abandoned as uneconomic bunch producers usually give good economic returns for wine production (using a non-destructive tapping technique on male inflorescences) before old plantings are cleared and replanted (Tuley 1965a).

Role in sustainable integrated production systems and in the protection of the environment

There are various types of palm-crop associations in Bangladesh. Phoenix sylvestris and Borassus flabellifer can both be associated with several of the following crops: rice, wheat, chickpea, mustard, jute, lentil, potato, linseed, winter vegetables and sugarcane (Abedin et al 1987). Palm trees often have advantages compared with other crops as far as sustainability is concerned: in parts of west Java where Arenga pinnata is still tended in groves, soils appear much more stable and productive of other crops than where cassava is cultivated (Dransfield 1977). Furthermore the advantages of this tree are its great ecological tolerance, its ability to grow and stabilize unproductive erosion-prone sites such as steep dryland slopes (e.g., coffee orchards on stony slopes in North Sulawesi, Mogea et al 1991), its potential to grow on almost any type of soil, to increase soil fertility and water conservation, its great tolerance of accidental burning (the only surviving tree in the Minahassa, Sulawesi, after volcanic activity), the relatively fast growth rate, the fact that it needs almost no maintenance and usually does not suffer from any serious pests or disease, and the wide range of secondary or alternate products obtainable. Arenga pinnata is considered as one of the most diverse multipurpose tree species and the most important sugar palm of the humid tropics; a further advantage is that this tree grows wild in many places (Mogea et al 1991).

Borassus flabellifer is often planted on boundaries of paddy fields in Cambodia and India. The effects of shading on understorey crops are likely to be negligible due to the small-sized crowns and to the large space (10-15m) between trees (Jambulingam and Fernandes 1986). In Cambodia, rice cultivation and Borassus flabellifer sugar production are associated in many areas: an average family owns 24-25 Borassus flabellifer trees and 2-2.5 ha rice field (Romera 1968). This tree thrives in reputedly the poorest, infertile and arid regions. In central Burma, after a year of drought, it still can produce sap during a year but the yield will decrease and the flow will finally stop (Lubeigt 1977; Lubeigt 1979). It also suffers remarkably little from prolonged flooding. It is extraordinarily pest and disease-resistant, requiring limited means of cultivation if any. As it grows in sandy plains, it is used for blocking erosion and fixing dunes, thanks to its deep root system (Kovoor 1983). It is also, like Corypha elata, a fire resistant palm that is pioneer species on regularly burnt land such as those exploited by the slash-and-burn technique (Ormeling 1956, cited by Fox 1977). The young trees (up to 15 to 40 years) are used in Burma as windbreak in areas cropped with groundnut (Lubeigt 1977; Lubeigt 1979). It plays a major role in Savu and Roti islands (Indonesia) where the soil fertility is a crucial constraint. The traditional slash-and-burn system, which is currently practised in neighbouring islands (Timor and Sumba for example), has been replaced by semi-permanent gardening through the use of large amounts of old Borassus leaves that are burnt in the fields. This permits fertile gardens to be kept in the vicinity of the houses. Animal manure is also used in selected gardens. In addition to an intensive pig-rearing, these islands are also famous for having well-developed goat and sheep herds and for having kept water buffalo whereas in neighbouring islands these have been displaced by Bali cattle. Gardens are easily fenced with palm leafstalks but in many cases in Savu, animals are penned or corralled and fields are left unfenced. Therefore these two islands which are drier, more wind-swept and more eroded that larger neighbouring islands, have based their main reliance on palm utilization integrated with wet-rice cultivation, dry-field cultivation, gardening, herding and fishing. all these components of the integrated farming systems vary in proportion according to area and years and show the great sense of adaptability that these people have reached. Whereas hunger periods and famines used to occur following droughts in neighbouring islands, calculations were always made in Savu and Roti in order that failed crops could be always replaced by other foods, Borassus sap or syrup being the ultimate one that was never missing (Fox 1977). Borassus forests possess a potentially unique pattern of nutrient cycling, which enables them to support relatively productive and stable forms of agriculture as well as to contribute to recovery of disturbed sites (Anderson 1987).

In the Peruvian Amazonia, Mauritia flexuosa constitutes dense populations in seasonal swamp forests on waterlogged soils or on sandy soils, which are generally considered as unfit for agriculture (Kahn 1988). Unlike sugarcane, Nypa fruticans does not compete with other crops for agricultural land except where total reclamation is undertaken on mangrove land (Hamilton and Murphy 1988).

Socio-economic aspects of tapping palm trees

Nowadays, the five major economic palms of the world are coconut (Cocos nucifera), African oil palm (Elaeis guineensis), date (Phoenix dactylifera), betel nut palm (Areca catechu) and pejibaye (Bactris gasipaes). Even if the three first species mentioned are currently tapped, this remains highly marginal in their economic role.

In the year 1910, 93% of the total amount of alcohol and alcoholic beverages produced in the Philippine islands was processed in 68 distilleries from 90 millions litres of the sap exuding from palm trees (mainly nipah and coconut palms); this industry was constantly increasing and nipah palm was considered the cheapest source of alcohol in the world (Gibbs 1911). Besides this, there was also a very large consumption of undistilled palm sap. In Burma, during the beginning of the seventies, up to 28 million bottle of arak made from Borassus flabellifer sap were commercialized per year through shops and bars, which means about a bottle per inhabitant (Lubeigt 1979). In 1969, palm wine produced in Nigeria was estimated at roughly 2 million metric tons (FMANR 1974, cited by Okereke 1982). In this country, the trade in palm wine provides a source of income, not only to the tappers themselves, but also to a wide range of market intermediaries (Okereke 1982). The species of Raphia (R. hookeri, R. vinifera, R. sudanica) are very important in the economy of the villages around the swamps in West Africa (Southern Benin and Nigeria): most of the people uses the fruits and the leaves, and palm wine is made from the sap which is consumed fresh or distilled (Profizi 1988; Tuley 1965b). Palm wine is also still largely produced and commercialized in Ivory Coast where it provides good opportunities of employment (Blanc-Pamard 1980). In many parts of Nigeria, and other parts of West Africa, the oil palms surrounding large towns are regularly tapped for wine production. It is an industry of considerable economic and nutritional importance which satisfies a steady market in these areas where higher returns are ensured using oil palm for wine production instead of oil production (Hartley 1977). In South-Eastern Africa, palm wine tapping of Hyphaene coriacea and Phoenix reclinata is a labour-intensive activity (high densities of small palms with low yields) providing a subsistence income together with pastoralism and gathering of wild fruits (Cunningham 1990).

As estimated by Pethiyagoda (1978), in the early decades of this century, in South-East Asian countries, hundreds of thousands of tonnes of sugar were produced annually through the tapping, mainly of Arenga sp, Borassus sp, Cocos sp, Nypa sp and Phoenix sp. Between 10 and 17% (estimate of 480,000 tonnes of palm sugar in 1911) of India's sugar was produced from palms (Annett 1913). In Bangladesh, the industry of sugar made from Phoenix sylvestris is an old one and date sugar was reported as largely manufactured and exported at the end of the eighteen century. In the middle of the last century, after slave emancipation measures gradually decreased the supplies of sugar from the West Indies, a rapid rise in date sugar production occurred. In 1849, one fifth (10,000 tonnes) of the whole annual quantity of sugar exported from India to England was date sugar (Annett 1913).

Palm tappers can be either prosperous or among the poorest: this is mainly related to the tree and land property. In Savu island (Indonesia), tapping is an activity among others and tappers own their land and trees. In Upper Burma, most tappers are landless and the palm trees are the property of a well-off minority which gets one third of the sap produced (Aubert 1911). In Sri Lanka, the art of tapping Caryota urens and making jaggery has been recognized as an important function to warrant the existence of a separate subcaste which still exists today (Dissanayake 1977). In the low-rainfall Ganges floodplain of Bangladesh, juice extraction from Phoenix sylvestris and molasses making support a special group of technicians, who are usually poor farmers (Abedin et al 1987). Arenga pinnata, through production and processing of its sap and sago ensures increased farmers' incomes and rural employment opportunities (Sumadi 1988). In West Java, it contributes widely to the daily life of many villagers for whom it is also a considerable source of income. In North Sulawesi, the dowry is still payable in the form of a number of sugar palms, indicating the great economic value placed on these trees. There, an average of 6 fully grown palms can provide a good income for one family (Mogea et al 1991). In the Philippines, most coastal farmers depend on Nypa fruticans sap production for their income (Quimbo 1991).In 1986, 118 tonnes of Nypa fruticans sugar valued at M$237,464 was exported from Malaysia (Kiew 1991). In Sarawak (Malaysia), nipah tapping is carried out as a cottage industry and provides a livelihood to a group of people who have no other means of earning a living (Pearce 1991). In many Asian countries, the use of this multipurpose palm tree continues and in some places, they have become the basis for cottage industries and commercial operation. According to Kiew (1991), coconut palm sugar is still for sale in every grocery shop in peninsular Malaysia. Recently, some large-scale commercial interest has developed: in particular, the high cost of fossil fuel and fertilizer has initiated a new look at the fuel alcohol and sugar potential of palm trees (Hamilton and Murphy 1988).

Palms of the genius Borassus come second only to the coconut in the palm population of the world and they span two continents across a good third of the equator (Kovoor 1983) where they have been tapped for centuries. Kovoor (1983) notes that the simple rationale of avoiding single-crop economies by all possible means is by itself sufficient to foster the systematic cultivation and exploitation of this palm. Generally speaking, it is the poor and the landless that benefit most from the Borassus palm (Fox 1977). In India, it is tapped by members of the lowest castes and it plays an important role in the economy of the rural poor, especially in drought-prone areas (Rangaswami 1977). In Tamil Nadu, it provides farmers with a cash income throughout the dry season when there is no crop on the fields and when demand for on-farm labour is low (Jambulingam and Fernandes 1986). Ferguson (1850, cited by Fox 1977) gives some details on the importance of this tree: "It is not the wholesomeness or nourishing qualities of the edible products of the Borassus flabellifer tree that make it so important to the inhabitants of India; but simply the fact, that thousands, perhaps millions of the people can procure these from their Borassus flabellifer groves, or purchase them for a low rate from their neighbours; whilst Rice and other articles of food, are frequently so expensive as to be placed beyond their means. The Borassus flabellifer tree is, in this respect, what the Potato has so long been to the poor Irish and Scotch". In Burma, half of the sugar consumed in the country comes from Borassus flabellifer (Lubeigt 1977). In Upper Burma, the tappers form a special class, the profession is handed down from father to son, requiring long training and good skills. The men only are tappers but all the family is involved in sap collection and processing. The life is hard, badly remunerated and attended often with sad accidents (Aubert 1911). Several hundred thousands of people live on Borassus flabellifer in Central Burma and more than one million for all Burma (Lubeigt 1977; Lubeigt 1979).This limits the drift from the land. Borassus flabellifer is also important in the Burmese mythology: the King Anoratha (1044-1077), founder of the first great Burmese Empire, had four close paladins, one of them well-known for his considerable strength demonstrated by his ability to climb one thousand sugar palm trees in a single day (Lubeigt 1979). In Indonesia, thousands of poor families in the Madura and Sunda island group make their living from this palm because of its multiple uses, the sweet sap representing its most economically valuable product (Davis 1988). In two Indonesian islands, Roti and Savu, studied by Fox (1977), the populations depend for their survival on utilization of the Borassus. There is a nearly total exploitation of this tree which is harvested for daily needs and which constitutes the pivot of a complex diverse economy. The potential is such that Fox notes that the precise carrying capacity of this economic adaptation to palm utilization remains unknown. Rotinese and Savunese live largely from the sugar of their trees: fresh sap throughout the production season or diluted syrup otherwise is consumed several times a day, often in place of any solid food. Fox (1977) notes that it can be literally said of the Rotinese and Savunese that they are fed, equipped, attired, buried, and remembered after their decease by the products of their palms. Greater specialization has accompanied increase of population density: as Borassus tapping is a labour-intensive activity, and as long as there are sufficient palms, new labour is absorbed. But in some areas in Savu, where the population densities are the highest, native palms were not sufficient and therefore people have planted clusters of palms in fenced enclosures, which in turn permits increased efficiency in tapping and further intensifies all the components of the integrated systems guaranteeing better yields. Some Rotinese and Savunese established themselves in parts of Sumba and Timor islands where they reproduced their traditional farming system based on Borassus and became economically stronger than the native people (Ormeling 1956, cited by Fox 1977). Borassus palms are essential buffers for ensuring food security of many people in Asia when droughts occur and other crops fail (Fox 1977). In Cambodia, where Borassus is associated with rice, if the rice yields are low, the harvest of Borassus sap is increased (Romera 1968). There, the areas where Borassus flabellifer trees are more numerous are also areas with higher population densities. Borassus flabellifer sugar production is essential for the rural community having reached a certain density. Once fuelwood becomes scarce, immigration starts (Romera 1968). Sugar palm syrup production is still today the main activity in most of rural Cambodia after rice growing (Khieu Borin and Preston 1995; Khieu Borin 1996). In the case of Borassus flabellifer and Caryota urens, the most important and primary material is their sap, and while other products of these palms have been replaced by substitutes, jaggery and toddy still demonstrate cost advantages to the communities who use them (Dissanayake 1986).

Origin of the decline in palm tree tapping activity

One of the main reasons for the decline of sugar production from palm trees is the increasing lack of fuelwood and its increasing price. Sugar production from Arenga pinnata requires 2-3 m³ wood for 100 to 120 kg of sugar (Mogea et al 1991). In South Sumatra, 3 to 4 kg of dry clove tree leaves or 0.02 m³ of durian or hevea wood are necessary to prepare 1 kg of sugar (Levang 1988). In Madura island (Indonesia), the cost of the firewood for producing Borassus sugar amounts to a third of the price of the sugar (Gebuis and Kadir 1928, cited by Fox 1977). In Cambodia 160 to 200 kg of fuelwood are necessary for making 35 kg of sugar (Lubeigt 1977; Romera 1968) and this is the most expensive component in the cost of sugar syrup production (Khieu Borin and Preston 1995; Khieu Borin 1996). There, Borassus flabellifer sugar making has been responsible for major forest-cutting and this sugar is not produced any more in certain rural areas because of the lack of fuelwood. On the other hand, the market for fresh juice is very limited if the area is not close to a city (Romera 1968). In Upper Burma, the sugar industry from Borassus flabellifer consumes a large amount of fuel and has for centuries led to indiscriminate forest-cutting. In some areas, fuel for domestic use is even becoming scarce (Aubert 1911). In Burma, every tapper uses about 3 tonnes of firewood per year; 4.4 kg are needed per kg of palm sugar (Lubeigt 1979). In Central Burma, the price of fuelwood is rapidly increasing and tappers can hardly afford it (Lubeigt 1977). This fuel problem started to become crucial in Burma at the end of the fifties when the government initiated measures to enforce protection of the forest (Lubeigt 1979).

In the case of wine-producing palm trees, the decline often occurred under religious or colonial pressure within the framework of a general policy aimed at preventing people from drinking alcoholic beverages. Thus, tapping of Borassus aethiopium in Africa mainly done for wine production has been disrupted wherever Muslim religion has become dominant. This happened in parts of Senegal where the catholic sereer had to leave their fields to Muslim wolof coming from the north. Before, in the catholic sereer areas, some people were fully employed as tapping professionals (Niang 1975). In Ivory Coast, the French colonial pressure tried to discourage the production of palm wine. In 1914, in the Baoul� region, the average consumption of palm wine was 135 litres/man/year but 2 years later the French governor forbade making and commercializing palm wine (which was also made from the African oil palm and Phoenix reclinata) except for family consumption and traditional events (Blanc-Pamard 1980). In Madagascar, making wine from Raphia ruffia also used to be forbidden (Decary 1964). In Sri Lanka, a 10th century rock inscription prohibits royal officers from demanding toddy from villagers or engaging in illicit trade (Dissanayake 1977). Centuries later, under colonial rule and thereafter, there was also a general ban on the production and sale of toddy at village level; the village headmen and the local constabulary who enforced these regulations were so repressive that very few villagers would have taken the risk of tapping Caryota urens even for making jaggery and treacle (Dissanayake 1977).

In Africa, some destructive techniques of tapping were responsible for the disappearance of the trees in some areas. In Ivory Coast, Port�res (1964) mentioned the high consumption of palm wine and the numerous specialized bars along the roads and in the villages, all this offering a lot of employment opportunities. The palm wine became very popular as through better means of communication, it could be made available throughout the country. The trees disappeared in entire areas within 10 years as the techniques used for tapping trees were destructive and regeneration of trees was neglected.

In many places, socio-economic changes were responsible for the decline in palm tree tapping activity. In Ivory Coast, the important moves of population in the fifties (settlers setting up coffee, cocoa, rubber trees and oil palm plantations) were also responsible for loss of traditional codes of managing the trees and less long term concerns. Thus the traditional technique of tapping only male trees and keeping females for regeneration was abandoned (Port�res 1964; Blanc-Pamard 1980). In many areas of this country, wine producing palm trees were replaced by rubber trees and oil palms under the French colonial pressure. In 1940, in Burma, the areas planted with Borassus flabellifer were reduced by more than one sixth because of the military operations consecutive to the Japanese invasion (Lubeigt 1979). In Cambodia, the decline of the production of Borassus flabellifer sugar started in 1955 with the end of the exports of sugar to Vietnam and the increasing consumption of imported cane sugar: Borassus flabellifer sugar prices collapsed (Romera 1968). In Sri Lanka, under colonial rule and thereafter, widespread cultivation of coconut as an exported-oriented crop drastically changed the local economy and imported sugar became cheaper (Dissanayake 1977). In Peninsular Malaysia, swamp areas were drained for coconut plantations where Nypa fruticans was before predominant (Kiew 1989). Fishpond developers also found great profits in various fishpond operations made possible by converting mangrove swamps, including Nypa fruticans areas, for fish production (Encendencia 1985).

Tapping sugar palms is very labour intensive. It must be done daily otherwise the sap flow rapidly diminishes as tissue healing occurs and restarting the sap flow requires long and hard work. Whenever easier and better paid jobs were available, tapping was given up. The emergence of the oil economy in Malaysia has collapsed the industry of alcohol from Nypa fruticans sap (labour too expensive). During the colonial period in India, Borassus tappers were recruited in the British plantations abroad, particularly on the rubber and oil palm estates where their skills could be easily adapted to those required for these trees (Fox 1977).

In many countries, in comparison to other crops or commodities, there is a general lack of interest shown by the decision makers about the socio-economic potential of tapping palms. None or little research, selection of higher yielding varieties or training and extension services are funded and the tappers are seldom exposed to technological innovations if they do not generate them by themselves.

Origin of the new interest for palm tree tapping activity

In today's economy, the profitability of tapping palms for sugar has sometimes improved: this is the case for coconut and Caryota urens in Sri Lanka. In the mid-seventies, with continuing foreign exchange crises, a reduction in the import of sugar occurred and was immediately followed by a sudden rise in its price. The government was then forced to relax the old colonial regulations and in many areas, palm sugar again became a low-cost source of sugar (Dissanayake 1977). In parts of South Sumatra (Sriwangi), tapping coconut for sugar production is 8 to 10 times more profitable than selling nuts. One family is rather well-off with 35 trees and in 1987, more than 40% of the families of the area were involved in tapping as their main source of income. Their average earnings were more than twice those of neighbouring areas where coconut tapping was not practised. A tapper earns per day twice to three times more than a field worker, that is to say about what earns a carpenter or a bricklayer. An important advantage for the tapper is that he has a daily income throughout the year. This gives him an easier access to credit facilities (Levang 1988). In Sri Lanka, about 0.5% (i.e. 2,025 ha) of the total surface for coconut land is tapped for toddy by at least 5,000 tappers (Abeysekera 1979). In the Philippines, a sequential coconut toddy and nut production system can provide the small scale coconut farmers with incomes nearly 10 times higher per hectare and per year compared to the traditional practice of producing nuts only (Maravilla and Magat 1993). In this country, a programme labelled as "Establishment of Buri Plantation for Technology Development and Employment Generation" is promoting the multipurpose sugar producing palm tree Corypha elata in order to stimulate the cottage industry (Abrenilla et al 1988). In Nigeria, an oil palm estate is likely to be better off devoting all its resources to the production of 9,770 litres/ha/year of oil palm wine than producing 10 tonnes of fresh fruit bunch per hectare per annum. Furthermore, as oil palm wine production is more labour-intensive than fresh fruit bunch production, tapping oil palm trees for wine is likely to create more jobs than harvesting fruit bunches (Udom 1987).

Producing sugar for palm trees that can be tapped all year round (like coconut and Nypa fruticans) is an advantage compared to the seasonal production of sugar from sugarcane. Production is not interrupted by replanting and rotation and the continuous productivity means no displaced labour, which is a major problem in cane economy. Furthermore, there is no bagasse disposal problem and there is no expensive crushing mill to maintain (Hamilton and Murphy 1988). On the other hand, the main drawback is the need for firewood or other fuel: for example, the dead leaves from Nypa fruticans used as fuel are far from being sufficient to make sugar from the sap of this tree. Palm trees that produce sugar seasonally, like Phoenix sylvestris from November to March (cold weather) and Borassus flabellifer from April to September (hot weather) would grow very well side by side as suggested by Annett (1913) in Bangladesh and would ensure continuous sugar production all year round.

In Latin America, the most abundant palm, Mauritia flexuosa, has a considerable potential from an economic development standpoint as it is a source of many different products among which palm wine should be promoted (Johnson 1997). In a tentative list of palms with development potential compiled by Johnson (1997), it is interesting to note that for about one out of two palms of the list, the sap is a major product.

Industrialized processing of the sap:

Borassus flabellifer sugar production is industrialized in Sri Lanka (Morton 1988). In India, different products from the sap of this tree are industrially processed: sap as soft drink (600,000 bottles in 1982-1983 in Madras), chocolate bars, candies, apple jam, mixed fruit jam and palm syrup (Davis and Johnson 1987). In West Africa, the sealed bottling of palm wine from the African oil palm is being undertaken (Bassir (1968, cited by Hartley 1977). In Cameroon, in 1969/1970, the consumption of locally made or imported drinks amounted to 405 million litres out of which 280 million litres were palm wine. In 1971 the first industrialized unit with a capacity of 1,000 litres/day of Raphia sp wine in sealed bottles was set up (Fyot 1973). In Ghana, the palm wine industry (large distilleries in the cities and small scale factories located in villages) was also of growing importance with the decline in the palm oil industry (Sodah Ayernor and Matthews 1971).

Prospects for increasing sugar yields

As stated before, compared to sugarcane production (5-15 tonnes of sugar/ha/year), the Borassus flabellifer tree can reach 18 tons/ha/year under rain-fed conditions (Khieu Borin and Preston 1995; Khieu Borin 1996) and the coconut tree 19 tons/ha/year (Jeganathan 1974). Nevertheless, the potential of increasing sugar production from palm trees is likely to be much higher than from sugarcane as much less research has been devoted to palm trees in this respect.

Indigenous knowledge is available in countries that have had a long experience in tapping palm trees. The tapper generally makes a selection before starting tapping: he chooses the trees that, according to his experience, should fulfil the following objectives: high sap yield; reduced time between commencement of working an inflorescence and the first flow of sap; maximum volume of sap sustained for as long as possible; health and well-being of the tree maintained during tapping (Pethiyagoda 1978).

Tall varieties of coconut trees yield twice as much sap as dwarf palms and are more resistant to pests and to droughts and winds because their root system is more developed (Jeganathan 1974). Considerable variations occur according to tapper's skills, day, season, prevailing weather, spadix, nature of the palm, locality and manure (Pethiyagoda 1978). A period of 31 days of sap flow per spathe is the average in Sri Lanka. There, coconut palms can be tapped throughout the year as long as rainfall is satisfactory and every palm is rested for 4 months a year (Coconut Research Institute 1967). Elsewhere, it is possible to encounter situations where trees have been continually tapped for many years without apparent ill-effects (Pethiyagoda 1978). There are considerably varying estimates of daily sap yields: from 0.9 to 15 litres per tree. The same occurs for sugar concentration of the sap (from 4 to 22%): the lower concentration (4%) is related with the higher yield (15 litres). The high-yielding coconut trees with regard to the production of nuts are not necessarily the best sap-yielders (Pethiyagoda 1978). Nathanael (1956) cited by Pethiyagoda (1978) lists the following criteria for promising sap yields: shiny, pliable leaflets, long internodes, uniform inflorescence production and thinner inflorescence sheaths. In Sri Lanka, through hybridization work to identify the most promising species with regard to nut production, a hybrid between a tall variety (Typica) and a dwarf one (Pumila) was found to be the best. Its sap producing capacity was studied later by Jeganathan (1974): on the basis of 158 trees to the hectare, the daily sap yield per hectare was 346 litres (with 15% sugar as for Typica sap) for the hybrid, while it was 247 for Typica and only 74 for Pumila (with 247 trees to the hectare for this latter variety). The difference between the hybrid and Typica was mainly due to the higher production of spadices in the hybrid (17.7 per year against 11.7 for Typica) which supersedes the lower and shorter production of every spadix for the hybrid (45 litres/21 days) compared to Typica (nearly 50 litres/31 days).

The impact of manuring on sap yields is reported to be high for coconuts but scientific data are scarce. The Sri Lankan Coconut Research Institute is presently doing some research and a recent survey has identified the following factors affecting toddy yields: moisture stress, failure to apply a mulch around the manure circle (2m around the palm), nutritional deficiencies, rain storms, winds, lack of experience and shortage of skilled tappers. A fertilizer mixture has been developed for toddy tapping-palms: rock phosphate (4 parts by weight), potassium chloride (18 parts), dolomite (8 parts). 2 to 3 kg of this mixture are given per palm per year according to soil type and climatic conditions; 600 g urea and 1 kg agricultural salt are also given per palm per year with an increase of 50% for the higher producing palms (Coconut Research Institute 1986).

Selection and breeding of the African oil palm for high sap yields and high concentration of sugar have not yet started on a large scale. It is likely that yield improvement research will produce varieties that will yield more than 100 litres of sap per palm and more than 14,800 litres per hectare per annum (Udom 1987). It is absolutely essential for most tapped palm species in Asia to have a sophisticated preparatory phase, sometimes continued throughout the tapping period, in order to ensure high yields of sap. It involves all sorts of techniques such as bending, kicking, hammering, kneading, bruising the inflorescence and its stalk. Such a preparatory phase has not been reported in Africa for the African oil palm and it is likely that south-south transfer of technology could permit a major increase in sap production from this tree.

Different management techniques permit increased sugar production from palm trees. Nypa fruticans produces more inflorescences (and potentially more sap) when the stands are kept thinned of old leaves. Sap production can be improved by wider spacing between trees than in wild almost pure stands of Nypa fruticans: from 2,500/ha down to 500 or less by removing the rhizomes of the others plants to prevent them from regenerating (Hamilton and Murphy 1988). In the Philippines, Quimbo (1991) developed a new, highly profitable method of tapping that increases the sap yield from less than 60,000 litres/ha to more than 100,000. This technique consists in using clenched fists as pressure applicator, which loosens the silica and crystal-like deposits in the tracheal elements of the peduncle, thus enhancing free sap flow at a similar rate but for much longer periods (111 days instead of 60 days). Treatment could also be reduced from 94 days (of which 52 contact days) to 43 days (of which 22 contact days) with total sap yield still higher on a per plant basis. Similarly in Papua New Guinea, P�iv�ke (1985) has developed a sophisticated protocol of pre-treatment that permits much higher sap yields.

For Arenga pinnata in North Sulawesi, a 10x10 m spacing produced the highest yields of sap (Sumadi 1988). High-yielding strains of Arenga pinnata have been isolated in the intensive sugar palm cultivation in Minahassa, North Sulawesi (Mogea et al 1991). The trees, with an average of 19m (up to 25m) are much higher than elsewhere. Near Tomohon, at an altitude of 700-800 m a.s.l., the sap production amounts to 12-15 litres a day, containing 12-15% of sugar with considerably longer tapping periods (6-12 months) for each inflorescence (normally more than 1m long). A superior sugar palm in this location produces all year round an average of 30 litres a day (Mogea et al 1991).

Daily Borassus flabellifer sap yields average between 6 and 10 litres per tree but can be as low as 1 litre or as high as 20 litres per tree (Paulas (1983; Tjitrosoepomo and Pudjoarinto 1983, cited by Kovoor 1983). This can be explained by genetic and environmental factors. More sap per tree can be obtained if each inflorescence produces more, over a longer period (skill of the operator), if there are more inflorescences in a given time, if flowering starts on younger trees and lasts longer (genetic factors) and if the response to tapping is higher (genetic factors) (Kovoor 1983). Borassus aethiopium is commonly tapped for wine production in west Africa but the production of sugar has never been recorded. Nevertheless, it has a good potential for sugar production and, in areas where sugarcane does not grow if not irrigated (Sudanese areas with long dry season), this tree could replace it for sugar production (Chevalier 1930).

Research is needed to assess the relative importance of genetic vs environmental vs technological factors in order to establish a priority agenda for increasing sap production and sap quality either through breeding programmes or management techniques or a combination of both.

Prospects for facilitating sap collection

For most non-destructive tapping techniques, a high degree of traditional expertise is needed and where this technique is not traditionally practised, great difficulties might be encountered in training people. Usually, the father teaches his son about palm tapping during a very long period. In the case of the high sugar producing palms, reduced height would be a much appreciated characteristic decreasing labour time, effort and risks. Unlike the coconut, dwarf mutants and races have not been reported to occur in the case of Borassus flabellifer (Kovoor 1983). This may be attributed to the lack of systematic research. An alternative would be to select the most precocious trees (that start flowering at a very low height) as precocity is a genetic trait (Kovoor 1983).

Devices for safer and more efficient ways of climbing palm trees have been invented: one by Davis (1984) cited by Davis and Johnson (1987); another was developed by the Palmyra Development Board of Sri Lanka and, using it, the tapper would be able to tap about 100 trees a day, more than twice the present average (Dissanayake 1986). The ergonomic problems of toddy tappers have been reviewed in Sri Lanka by Abeysekera (1979).

Hybridization of the African oil palm with the American species, Elaeis oleifera, which has a creeping trunk and better resistance to disease (Kahn 1988) could produce a productive variety, easy to tap because of low and stable height.

Prospects for animal production within sustainable integrated farming systems

Storage of sap at local level is not possible as fermentation rapidly occurs even if delayed by some chemical agents. Fermented sap is not suitable for the production of good quality sugar and this usually limits the expansion of palm sugar making at village level. Processing sap into good quality jaggery is also a difficult and time-consuming task: up to 16 hours per day in Cambodia (Khieu Borin and Preston 1995; Khieu Borin 1996). It also requires an experienced and skilled worker, often a woman (stirring, removing of froth, maintaining the appropriate temperature, etc.). This is also a major bottle-neck which limits sap processing (Dissanayake 1986). Furthermore, in many countries, production and sale of toddy is prohibited by regulations and some raw material is wasted (Dissanayake 1986). According to Mogea (Mogea et al 1991), Arenga pinnata in North Sulawesi (Indonesia) could considerably contribute to the increase of the local incomes and provide many job opportunities if the problems due to the enormous amounts of fuelwood required for making sugar and the limited marketing possibilities of the products could be solved. It is difficult to bring the fresh sap and the palm wine to the remote markets whereas the neighbouring ones are often saturated. If the sugar is better commercialized, then the fuel constraint acutely occurs. In Roti and Savu islands (Indonesia), the tappers claim that they could tap 20 to 30 Borassus trees, but rarely does anyone tap more than 10 to 15 at any one time as cooking the sap to produce syrup is the limiting factor (Fox 1977). There is a huge potential for capitalising on under-exploited sugar palm trees which are not used because of the lack of fuelwood for making sugar or the limited marketing possibilities. In Cambodia, two surveys conducted in 1938 and 1960 showed that respectively only 26% and 30% of tappable palmyra trees were actually tapped (Romera 1968). In Sri Lanka, only about 2% of the total area suitable for tapping is reported to be actually tapped (Sivilingam 1983, cited by Dissanayake 1986). In coastal Nigeria and Cameroon, Nypa fruticans which was introduced early in this century, has naturalized and is underutilized (Johnson 1997). Therefore, there is a niche for diversification. In these cases, the sap could be used for animal production. Present labour constraints can be overcome through the use of climbing devices that enable the tapper to tap twice as many palm trees (Dissanayake 1986).

On the other hand, meat demand is increasing in many developing countries as population grows and living standards improve: in the case of Cambodia, the pig population is increasing at a rate of 16.6% per year (Devendra 1993, cited by Khieu Borin and Preston 1995; Khieu Borin 1996). Instead of preparing sugar from the sap of sugar producing palm trees, the sap can be directly fed to the animals and provide most of the energy needed in the diet. This has been done for centuries in two Indonesian islands, Roti and Savu. They have a complex diverse economy that has Borassus as the centre and which includes a small-scale semi-intensive or intensive pig-rearing component (7-8 pigs per household). In a Borassus economy, pigs are a prime means of converting palm products to protein. Pigs are fed fresh sap throughout most of the tapping season and therefore fatten during the dry season while other livestock usually lose weight. In addition, pigs often receive the residue (froth) and spill from the syrup-cooking process. During the rainy season, they are frequently fed syrup mixed with water (nevertheless, when Corypha elata is available as in Roti, people will generally favours the use of its sago as pig feed instead of palm syrup). Fox concludes in these words: "Borassus syrup and fruit constitute the primary food for pigs; pigs in turn are a principal means by which Savu's palm economy is able to support its dense population; pigs and palms go together and one can view pigs as a reasonable indicator of palm utilization". This is further demonstrated by the strong correlations (much higher than for other livestock species) between pig and human populations in the different areas of these two islands. The areas where the population densities are highest, are the areas of most intensive pig-rearing; pigs also representing the highest proportions of the total livestock (Fox 1977). Captain James Cook, sailing west from New Guinea stopped at the Savu island from 17 to 21 September 1770, at the high point of the tapping season. He reported in his book "Voyages" detailed information on the use of Borassus. In this particular year, the crops were reported to have failed. Therefore the maximum harvest of sap was taking place in order to secure 6 to 8 months food supply. Despite this threatened food security situation, Cook witnessed that syrup was given to pigs and used even for other animal production: "I have already observed, that it is given with the husks of rice to the hogs, and that they grow enormously fat without taking any other food: we were told also, that this syrup is used to fatten their dogs and their fowls..." (Cook cited by Fox 1977).

The author of this paper spent a week in Roti in mid November 1997. It was at the end of a long and severe dry season (due to El Ni�o). As Roti is the most southern Indonesian island, it gets the shortest rainy season of the country. Borassus tapping is still practised everywhere on the island. Tappers were interviewed in two places, a Muslim fishing village on the north east coast (Papela) and a Christian village along a beach on the south west coast (Nemberala). Fresh sap is still commonly used for animal production. Pigs are fed with fresh sap twice a day, just after sap collection (early morning and at dusk). When fresh palm sap is not available (rainy season), palm syrup (one litre of syrup comes from 4 to 5 litres of sap) is diluted with water to feed the pigs. The quantity of palm sap given per pig per day varies according to the availability of other feeds. Here are two examples of a daily pig diet, respectively during the rainy season, in Papela, and during the tapping season in Nemberala: three meals, each one including one litre of syrup diluted in two litres of water and about 170g of rice; two meals per day, each one including 1.5 litre of fresh sap and chopped meat from about two and half coconuts. Apart from these feeds, the pigs generally satisfy the greatest part of their protein needs through scavenging on the beach, especially at low tide. The pigs spend all days and even part of the night, digging long ditches in the sand, going far from the sea shore and close to the corral reef and turning over the stones in every pool accessible at low tide. They get crabs, snails and various other shells, small fishes, sea cucumbers and sea weeds. The pigs are sold on the market once they are about one year old, weighing around 40 kg. The poultry also drinks the palm juice. In Nemberala, the juice is also given to the dogs. Every dog gets about a litre of sap per day shared in three meals. The other part of the diet is made of rice, fish and meat wastes and bones. Apart from guarding the house, the use of the dog is for its meat as it is still a traditional food for the non-Muslim people of the island. The males are generally slaughtered when they are around 2 years old.

Trials on feeding pigs with palm juice have been initiated recently in Cambodia within the framework of an FAO Technical Cooperation Project (FAO 1995). Pigs were reared from 20 to 80 kg, with ADG of 356g using the following daily diet: approximately 8 kg of palm juice 156 g of protein (from soya bean), lime, salt and 500g of fresh water spinach per day. Twelve farms were studied. Taking into account the price of fuelwood, the profit per tree per day was nearly 14 times higher when the juice was used for feeding pigs instead of making sugar syrup (Khieu Borin and Preston 1995; Khieu Borin 1996).

Using fresh sap for feeding animals will avoid burning large quantities of fuel. Nevertheless, as part of this fuel generally comes from the palm tree itself, it might be possible to make syrup or sugar that will be easy to preserve and that will be later fed to the animals when the sugar production season is over. If this is not possible, sap production can be entirely used as fresh juice for feeding fattening animals and the fattening cycle can coincide with the sometimes rather short tapping season. This can easily be done with pigs and ducks. Apart from giving a lot of added value to the sap, this also permits the small farmers, who usually have very low storage capacity for syrup or sugar, to avoid selling sap at low price during the tapping season. Sap, syrup or sugar could also be used as emergency feeds, replacing other feeds whose production has been compromized by droughts or other calamities, whenever necessary.

To balance monogastric diets based on sugar palm juice or syrup, a good source of protein is required. As soya bean is hardly available at a reasonable price in many tropical areas, some alternative sources of protein are needed: possible sources include cassava leaves, sweet potatoes leaves, fodder tree leaves, aquatic plants (duckweed, Azolla), whole soya plant at milky grain stage and fish wastes. Proper use and management of these different alternative sources of protein can contribute to reducing pollution, increasing carbon sinks and decreasing erosion. Animal feeding systems based on palm juice/syrup favour keeping the animals in confinement instead of grazing or scavenging systems. This protects the environment, limits the dissemination of contagious diseases and also optimizes the integration of livestock within an intensive farming system (Figure 1). Manure can be processed through a biodigestor, producing the energy for family cooking needs, and the effluent can be used as a fertilizer either for crops or for fish ponds.

Ruminants also benefit from the sugar palm trees. As mentioned by Khieu Borin and Preston during the Second FAO Electronic Conference on Tropical Feeds, Borassus sap and the scum obtained during sap cooking are traditionally used for softening rice straw used for cattle feeding (especially draught animals) in Cambodia (Speedy et al 1997). Cattle are also fed with mature fruits or parts of them that are not used for human consumption. During his stay in Roti, the author of this paper did not get any testimony related to feeding fresh sap or syrup to the ruminants. Nevertheless, the goats get the foam produced during the sap cooking. Green twigs with their leaves, taken from the tree that also provides the firewood used for cooking the sap, are slightly cooked together with the sap and used to remove the foam. Then, they are immediately given to the goats. Thanks to this feed, the goats were still in a very good body condition despite of the fact that it was the end of a long and severe dry season.

The potential of feeding goats and cows with palm sap as the main source of energy for milk production should be investigated as well as the source of nitrogen (non-protein nitrogen and by-pass proteins), minerals and fibre to complete the diet. Incidentally, tapping palm trees will also always offer an easy source of sugar for bees which will tend to spontaneously harvest all wasted sugar. Honey production is therefore increased in areas where palm trees are tapped (Fox 1977).

Conclusion

Borassus palms are the most numerous palms in the world after the coconut palm (Fox 1977). Despite this, they are among the least studied of all the palm species in the world. As for many other sugar producing palm trees such as Caryota urens, very little is known about their economic potential (Dissanayake 1977). This lack of interest can be explained during colonial history by the fact that, from the colonizer's point of view, it was much easier to set up, manage and control large sugarcane plantations to produce sugar than to use existing scattered palm trees that had been managed for centuries by the local people, often within a subsistence economy. Beside this, these trees are often associated with the poor. The fact that their juice quickly ferments and makes alcohol made tapping activities undesirable to governments, and also for the Hindus, Buddhists and Muslims orthodoxes (Fox 1977).

Nevertheless, there are many good arguments for revitalizing knowledge and research on sugar producing palm trees. Considering their multipurpose uses, they can contribute in many ways to the sustainability of integrated farming systems. For example, in Indonesia, Mogea (1991) suggests to promote and develop through research projects three particular taxa within the native palm resources. Two of them have sugar as their main product: Arenga pinnata and Borassus flabellifer. As sugar producing palm trees are often the main subsistence resource for the poorest people, improving the way these trees are used will contribute to the alleviation of poverty. Palm tapping, especially as far as wild and semi-wild species are concerned, is an activity that does not require capital to start. In highly populated rural areas, it can be a major source of self-employment for the poorest people and avoid major drifts from the land. In the case of coconut (in Sri Lanka for example) or African oil palm (in Colombia and Nigeria for example), with the low and unpredictable world prices of copra and palm oil, it has become increasingly difficult for small farmers to depend on their production. This encourages attempts to find other ways of using these trees, including diversification for better sustainability of the system. Sugar production and animal production are alternatives to consider if markets can be developed for these products.

When selecting palms suitable for small farmers, it is important to consider the amount of knowledge available about the candidate species. A good indicator is the degree to which a palm has been domesticated. Generally, the more advanced the stage of domestication, the easier the species can be cultivated successfully (Johnson 1987). Considering the high cost and duration of a domestication program, in the case of non-domesticated sap-producing palms, it is suggested to evaluate and select one for domestication (Arenga pinnata, Corypha umbraculifera and C. utan are strong candidates) whereas the other species should be considered for management improvement (Johnson 1997).

Future research on using palm tree sap for animal production should consider the following issues:

What is needed is a thorough field survey reviewing in detail all indigenous knowledge related to tapping palm trees for sugar and animal production. This would permit a major breakthrough for assessing the future potential of these trees and for sharing techniques and experiences between regions and countries. Many rural areas are likely to benefit from a new source of self-employment and sustainable income once the potential of tapping palm trees for sugar and animal production has received the full attention it deserves from decision makers through funding research, selection, technology improvement, training and extension and small credit for farmers,

References

Abedin Z, Aktar S, Haque F and Alam S 1987 Uses of Multipurpose Trees on the Small Farms of the Low-Rainfall Ganges Floodplain Soils of Bangladesh. In: Proceedings of an international workshop held in Pataya, Thailand, 2-5 November 1987, pp. 31-47.

Abeysekera J D A 1979 Some ergonomic problems among (coconut) toddy tappers in Sri Lanka. In: Proceedings of the Fifth Joint Ergonomic Symposium. Wageningen, Netherlands 14-18 May 1979. pp 112-114.

Abrenilla C G, Pablo N R and Polinag M A 1988 Buri Plantation: towards a sustainable industry Canopy International 14(6): 9-11.

Adand� A 1954 Le vin de palme chez les Diola de la Casamance - Notes Africaines 61:4-7.

Anderson A B 1987 Management of native palm forests: a comparison of case studies in Indonesia and Brazil. In: Agroforestry: realities, possibilities and potentials [edited by Gholz H.L] Dordrecht, Netherlands. 155-167.

Annett H E 1913 The date sugar industry in Bengal - An investigation into its chemistry and agriculture. Agriculture Research Institute, Pusa, India, Chemical Series II(6): 281-389.

Aubert L 1911 The manufacture of palm sugar in upper Burma. Agricultural Journal of India. VI (4): 369-376.

Audru J 1985 Restauration et meilleure utilisation des hyphaenaies d'Agna et de Galafi (R�publique de Djibouti). Paris, CIRAD-IEMVT. pp. 117.

Balick M J 1990 Production of Coyol Wine from Acrocomia mexicana (Arecaceae) in Honduras. Economic Botany, 44(1): 84-93

Banzon J A 1984 Harvestable energy from the coconut palm. Energy in Agriculture, 3: 337-344.

Barreveld W H 1993 Date palms products. FAO Agricultural Services Bulletin No 101. FAO, Rome. pp. 262.

Baumer M 1995 Arbres, arbustes et arbrisseaux nourriciers en Afrique occidentale. S�rie Etudes et Recherches, n 168-169-170. Dakar. enda tiers-monde. ENDA-Editions.

Bellouard P 1950 Le R�nier en A.O.F. Bois et For�ts des Tropiques. 14: 117-126.

Blanc-Pamard C. 1980 De l'utilisation de trois esp�ces de palmiers dans le sud du "V Baoul�"(C�te d'Ivoire). Cahiers ORSTOM, s�rie Sciences Humaines. (3-4): 247-255.

Chevalier A 1930. Le Borassus aethiopum de l'Afrique Occidentale et son utilisation Revue de Botanique Appliqu�e. pp. 649-655.

Coconut Research Institute 1967 Toddy tapping. Ceylon. Leaflet No. 48. pp. 13pp.

Coconut Research Institute 1986 Sri Lanka Advisory Circular No. A4 March. pp. 8.

Crevost C and Lemari� C 1913 Catalogue des produits de l'Indochine. I Produits Alimentaires. Gouvernement G�n�ral de l'Indochine, pp. 347-352

Cunningham A B 1990 Income, sap yield and effects of sap tapping on palms in south-eastern Africa. S. Afr. J. Bot. 56(2): 137-144.

Davis T A and Johnson D V 1987 Current Utilization and Further Development of the Palmyra Palm (Borassus flabellifer L., Arecaceae) in Tamil Nadu State, India. Economic Botany, 41(2): 247-266.

Davis T A 1988 Uses of Semi-Wild Palms in Indonesia and Elsewhere in South and Southeast Asia. Advances in Economic Botany 6: 98-118.

Decary R 1964 Les utilisations des palmiers � Madagascar. Journal d'Agriculture Tropicale et de Botanique Appliqu�e. XI (8-9): 259-266

Dennett J H 1927 Final Observations on the Nipah Palm as a Source of Alcohol. The Malayan Agricultural Journal. XV (12): 420-432.

Dissanayake B W 1977 Use of Caryota urens in Sri Lanka. In: First International Sago Symposium. The Equatorial Swamp as a Natural Resource. Ed. Tan K. Sarawak, East Malaysia, Kuching 1976. Kuala Lumpur. pp. 84-90.

Dissanayake B W 1986 Techno-economic aspects of production and utilization of two related palms kitul (Caryota urens) and palmyra. Proceedings of the Third International Sago Symposium, Tokyo, Japan. pp. 75-85.

Dransfield J 1977 Dryland Sago Palms. In: First International Sago Symposium. The Equatorial Swamp as a Natural Resource. Ed. Tan K. Sarawak, East Malaysia, Kuching 1976. Kuala Lumpur. pp. 76-83.

Encendencia E M 1985 Versatile Nipa: Potential Resource for Various Economic Needs. NSTA Technology Journal/July-Sept. pp. 4-12.

Essiamah S K 1992 Sapping of Oil Palm (Elaeis guineensis, Jacq.) in the Rain Forest Region of West Africa. Tropenlandwirt (Germany). 93(2): 123-135.

Eze M O and Uzoechi Ogan A 1988 Sugars of the unfermented sap and the wine from the oil palm, Elaeis guinensis, tree. Plant Foods for Human Nutrition 38: 121-126.

Fanshawe D B 1966 The Dum Palm - Hyphaene thebaica (Del.) Mart. East African Agricultural and Forestry Journal. October 1966: 108-116

FAO 1995 Plan d'urgence pour la sauvegarde du b�tail au Cambodge. Compte rendu final de projet. TCP/CMB/2254. FAO, Rome. pp. 14.

Fealy N E. 1925 Quelques Palmiers exploit�s pour la production de Sucre. Revue de Botanique Appliqu�e & d'Agriculture Coloniale. 45: 383-387.

Fox J F 1977 Harvest of the Palm, Ecological Change in Eastern Indonesia. Harvard University Press, Cambridge, Massachusets, and London, England. pp. 290.

Friedberg C 1977 Les palmiers � sucre et � vin dans le sud-est asiatique et en Indon�sie. Journal d'Agriculture Traditionnelle et de Botanique Appliqu�e, XXIV (4): 341-345.

Fyot R 1973 La valorisation industrielle d'une boisson traditionnelle: le vin de palme pasteuris�. Techniques et D�veloppement. Les Produits. 8(10): 27-29.

Gibbs H D 1911 The alcohol industry of the Philippine Islands. I. A study of some palms of commercial importance with special reference to the saps and their uses. Philipp. J. Sci. Sect. A, 6: 99-145

Giffard P L 1967 Le Palmier R�nier Borassus aethiopum MART. Bois et For�ts des Tropiques. 116: 3-13.

Grimwood B.E 1975 Coconut Palm Products, Their processing in developping countries. FAO Agricultural Development Paper No. 99. FAO. pp. 189-192.

Hamilton L S and Murphy D H 1988 Use and Management of Nipa Palm (Nypa fruticans, Arecaceae): a Review - Economic Botany. 42(2): 206-213.

Hartley C W S 1977 The Oil Palm (Elaeis guineensis Jacq.). 2nd ed. Longmans and Green London. pp. 600-603.

Jahiel M 1993 Le palmier doum Hyphaene thebaica (L.) (Mart.). Flamboyant. Bulletin de liaison des membres du R�seau Arbres Tropicaux. 28: 4-10.

Jambulingam R and Fernandes E C M 1986 Multipurpose trees and shrubs on farmlands in Tamil Nadu State (India). Agroforestry Systems 4: 17-32.

Jeganathan M 1974 Toddy yields from hybrid coconut palms. Ceylon Cocon. Q. 25: 139-148.

Johnson D J 1987 Palms as Multipurpose Cash and Subsistance Tree Crops. In: Proceedings of an international workshop held in Pataya, Thailand, 2-5 November 1987, pp. 222-236.

Johnson D J 1997 Non-wood forest products: tropical palms. RAP Publication 1997/10. FAO, Bangkok. pp. 166.

Kahn F 1988 Ecology of Economically Important Palms in Peruvian Amazonia. Advances in Economic Botany 6: 42-49.

Khieu Borin and Preston T R 1995 Conserving biodiversity and the environment and improving the wellbeing of poor farmers in Cambodia by promoting pig feeding systems using the juice of the sugar palm tree (Borassus flabellifer). Livestock Research for Rural Developement (7) 2: http://www.cipav.org.co/lrrd/lrrd7/2/5.htm

Khieu Borin B 1996 A study on the use of the sugar palm tree (Borassus flabellifer) for different purposes in Cambodia. M.Sc. Thesis Swedish University of Agricultural Sciences, Uppsala. pp. 67.

Kiew R 1989 Utilization of Palms in Peninsular Malaysia. Malayan Naturalist 43 (1&2): 43-67.

Kiew R 1991 Palm utilization and conservation in Peninsular Malaysia. In: Palms for human needs in Asia. Palm utilization and conservation in India, Indonesia, Malaysia and the Philippines. Johnson, D.J. (Ed). WWF Project 3325. WWF & IUCN. Rotterdam, pp.75-130.

Kovoor A 1983 The palmyrah palm: potential and perspectives. FAO Plant Production and Protection Paper No 52. FAO, Rome. pp. 77.

Levang P 1988 Le cocotier est aussi une plante sucri�re. Ol�agineux, 43(4): 159-164.

Lubeigt G 1977 Le palmier � sucre, Borassus flabellifer L., ses diff�rents produits et la technologie associ�e en Birmanie. Journal d'Agriculture Tropicale et de Botanique Appliqu�e. XXIV (4): 311-340.

Lubeigt G 1979 Le palmier � sucre (Borassus flabellifer) en Birmanie centrale. Publications du D�partement de G�ographie de l'Universit� de Paris-Sorbonne. No 8. Paris. pp. 197.

Magalon, M 1930 Contribution � l'�tude des palmiers de l'Indochine fran�aise. Paris, pp.185-191.

Maravilla J N and Magat, S S 1993 Sequential coconut toddy (sap) and nut production in Laguna tall variety and hybrid coconuts Philippine Journal of Coconut Studies XVIII(2): 7-15.

Mathes D T 1984 Effect of extraction of inflorescence sap and the removal of immature fruit bunches on the production of female flowers in coconut. Sri Lanka. COCOS, 2: 44-47.

May P H, Anderson A B, Balick M J and Fraz�o J M F 1985 Subsistence Benefits from the Babassu Palm (Orbignya martiana). Economic Botany, 39 (2): 113-129.

Mogea J 1991 Indonesia: Palm utilization and conservation. In: Palms for human needs in Asia. Palm utilization and conservation in India, Indonesia, Malaysia and the Philippines. Johnson, D.J. (Ed). WWF Project 3325. WWF & IUCN. Rotterdam, pp.37-73.

Mogea J, Seibert B and Smits W 1991 Multipurpose palms: the sugar palm. Agroforestry Systems 13: 111-129.

Morton J F 1988 Notes on Distribution, Propagation, and Products of Borassus Palms (Arecaceae). Economic Botany, 42(3): 420-441.

Naim S H and Husin A 1984 Coconut Palm Sugar. In: Cocoa and Coconuts: Progress and Outlook. Ed. E. Pushparajah & Chew Poh Soon. Kuala Lumpur 15-17 October 1984. pp 943-946.

Niang M 1975 Le R�nier dans la R�gion de Thi�s - Etude g�ographique. Notes Africaines. 147: 77-82.

Okereke O 1982 The traditional system of oil palm wine production in Igbo Eze local government area of Anambra state of Nigeria. Agricultural Systems. 9: 239-253.

P�iv�ke A E A 1985 Tapping practices and sap yields of the nipa palm (Nipa fruticans) in Papua New Guinea. Agriculture, Ecosystems and Environment. 13:59-72.

Pearce K G 1991 Palm utilization and conservation in Sarawak (Malaysia). In: Palms for human needs in Asia. Palm utilization and conservation in India, Indonesia, Malaysia and the Philippines. Johnson, D.J. (Ed). WWF Project 3325. WWF & IUCN. Rotterdam, pp.131-173.

Pethiyagoda U 1978 Coconut inflorescence sap. Perak Planters' Association Journal. Malaysia. pp. 86-91.

Port�res R 1964 Le Palmier R�nier (Borassus aethiopum Mart.) dans la province du Baoul� (C�te d'Ivoire). Journal d'Agriculture Tropicale et de Botanique Appliqu�e. XI (12):499-515.

Profizi J P 1988 Swampy Area Transformations by Exploitation of Raphia hookeri (Arecaceae) in Southern Benin (West Africa). Human Ecology 16(1): 87-94.

Purnomo H 1992 Sugar Componentss of Coconut Sugar in Indonesia. ASEAN Food Journal. 7(4): 200-201.

Quimbo L L 1991 Increasing Nipa Sap Yield Through Improved Mechanical Tapping. Technology! PCARRD. The Philippines. XIII (4): pp.16.

Rangaswami G 1977 Palm tree crops in India. Outlook-on-Agriculture (UK). 9(4): 167-173.

Redhead J 1989 Utilization of tropical foods: trees. In: FAO Food and Nutrition Paper No 47: 3. FAO, Rome, pp. 52.

Romera J P 1968 Le Borassus et le sucre de palme au Cambodge. L'Agronomie Tropicale. 8: 801-843.

Sodah Ayernor G K and Matthews J S 1971 The Sap of the Palm Elaeis guineensis Jacq. as Raw Material for Alcoholic Fermentation in Ghana. Tropical Science. XIII(1): 71-83.

Speedy A W, Dalibard C and Sansoucy R 1997 Livestock feed resources within integrated farming systems. Second FAO Electronic Conference on Tropical Feeds. http://www.fao.org/waicent/faoinfo/agricult/aga/agap/frg/

Stewart L 1994 A guide to Palms & Cycads of the World. Angus & Robertson. Sydney, pp.244.

Sumadi G 1988 Arenga pinnata: A Palm for Agroforestry. In: Proceedings of an international workshop held in Pataya, Thailand, 2-5 November 1987, pp. 149-150.

Theivendirarajah K, Dassanayake M D and Jeyaseelan K 1977 Studies on the Fermentation of Kitul (Caryota urens) Sap - Ceylon J. Sci. (Bio. Sci.) 12(2): 147-150.

Theivendirarajah K and Jeyaseelan K 1977 An Evaluation of the Methods Employed in the Preservation of Kitul Sap. Ceylon J. Sci. (Bio. Sci.) 12(2): 150-155.

Theivendirarajah K, Jeyaseelan K and Puvirajasingam V 1979 Some Aspects of Fermentation of Coconut (Cocos nucifera L.) Sap. Ceylon J. Sci. (Bio. Sci.) 13(1 & 2): 79-87.

Tuley P 1965a Studies on the production of wine from the oil palm. J. Nigerian Inst. Oil Palm Res. 4: 284-289.

Tuley P 1965b The production of Raphia wine in Nigeria. Expl agric. I : 141-144.

Udom D S 1987 Economics of oil palm wine tapping. Nigerian Journal of Palms and Oil Seeds. VIII(1): 56-77.

Van Die J 1974 The developing fruits of Cocos nucifera and Phoenix dactylifera as physiological sinks importing and assimilating the mobile aqueous phase of the sieve tube system. Acta Bot. Neerl. 23(4): 521-540.

Ven�ncio Machado J 1974 Produ��o e industrializa��o de derivados de coco em Mo�ambique. Agron. mo�amb., Louren�o Marques, 8 (1): 61-69.

Annex 1: Location, management, products and tapping techniques of tapped palms species


Latin name (Common names)	Regions and management	Products	Tapped parts

*Acrocomia aculeata*	Latin America and the Caribbean (Haiti) [32]	Wine [32]
*Acrocomia mexicana*	Honduras [37]	Wine [37]	Terminal bud on felled trees [37]
*Areca catechu (Areca palm, betel palm, arecanut)*	Tropical rain forest to 900m ; S & SE Asia: India, Indonesia; Improved cultivated palm [1]	Sugar (jaggery), fresh juice, toddy; seed as masticatory, edible heart, leaves for thatch, leaf sheaths for hats, containers, trunk for wood, dye source, etc. [1]	Same as for Arenga pinnata
*Arenga pinnata or saccharifera (Sugar palm, toddy palm, lontar); A. undulatifolia in Borneo; A. tremula in the Philippines*	Tropical rain forest into dry forest to 1,200m ; S & SE Asia: India, Thailand, Cambodia, Malaysia, Indonesia (N Sumatra, W Java, E Borneo, N Sulawesi), Philippines; Unimproved cultivated or managed palm [1]; Strong candidate for domestication [32]	Sugar (jaggery), fresh juice, toddy [1], vinegar [2]; starch from trunk (sago), seawater-resistant fibre from leaf sheath, edible heart [1], edible fruits, leaves for roofing, for making brooms, matting, baskets, cigarette papers, etc. [3]	Male (and rarely female) inflorescence spadix or secondary stems
*Arenga wightii*	S Asia: endemic to India (Kerala, Karnataka, Tamil Nadu) [32]	Toddy [32]	Peduncle [32]
*Attalea sp*	Central and S America [38]	Fresh juice, wine [38]	Terminal bud [38]
*Beccariophoenix madagascariensis*	Central Madagascar (1,000m)	Fresh juice; leaves: hats [4]	Betsimisaraka: trunk [4]
Borassus aethiopium *(African palmyra, African fan palm)*	Tropical savanna [1]; From 500 (even 200 if superficial water; also tree benefitting from dews and fogs [5]) to 1,300mm yearly rainfall with 6 to 8 months of dry season [6]; Africa: Senegal (Thi�s, Casamance), Gambia, Guinea, Ivory Coast, Burkina Faso (Banfora), Cameroon, Gabon; Semi-wild or wild palms [1]; Candidate for domestication with agroforestry potential [32]	Wine, fresh juice, vinegar [34]; potential for sugar production [7] [8] which is denied by [9]; edible immature fruit, terminal bud, young roots and germs, leaves for thatch, weaving, wood for building [1] [7]	Terminal bud, lateral portion of young stem or bunches
*Borassus flabellifer (Sugar palm tree, fan palm, palmyra palm, toddy palm, desert palm, lontar palm, brab tree, rondier)*	Tropical forest into savanna to 750m [1]; East Africa: Zanzibar; Asia: Persian Gulf, coast line of India, Sri Lanka, Bangladesh, Burma, Thailand, Malaysia, Indonesia (Borassus sundaicus (lontar palm) [10]) Laos, Cambodia, Vietnam, China; Unimproved cultivated or managed palm [1]; Candidate for domestication with agroforestry potential [32]	Sugar "candy", jaggery, toddy, palm juice, arak, vinegar, yeast for bread [11]; edible fruits and shoots, leaves for manuscripts, timber, fuel, building material, canoe, fan [12], from leaf stalk: fibre (exported from Tamil Nadu to 40 countries [13]), fences [10] leaves for thatch, weaving [1]	Inflorescence spadix, spathe of male and female trees or fruits
*Borassus madagascariensis (R�nier Malgache)*	Along rivers [4]; Madagascar	Wine	Inflorescence spadix [4]
*Calamus vanuatuensis*	Pacific: endemic to Vanuatu [32]	Sap [32]	Stem [32]
*Caryota urens (Indian sago palm, toddy palm, kitul palm, fish-tail palm, caryota palm); C. cumingii in the Philippines*	Tropical rain forest, especially primary forest [1] to 1,500m [14]; Asia: India, Sri Lanka, Indonesia, Malaysia, Cambodia; S Pacific; Unimproved cultivated or managed palm [1]; Domestication potential in agroforestry systems [32]	Honey, sugar (jaggery), fresh juice, toddy [11]; Price of sugar about 50% over Borassus flabellifer and also higher price than coconut [15]; fibre from leaf sheath, starch from trunk (sago), seeds chewed as substitute for betel nut, edible heart [1] [11]	Inflorescence spadix
*Cocos nucifera (Coconut palm)*	Tropical rain forest, coastal areas to 300m; E Africa: Mozambique; Asia: Southern India, Sri Lanka, Thailand, Cambodia, Malaysia, Indonesia (Central Java), Philippines; Pacific islands; Improved cultivated palm [1]	Sugar (jaggery), fresh juice (occasionally tapped in Mozambique [16]), wine, arak, vinegar; edible oil, fruit, heart, leaves for thatch, weaving, trunk for wood, etc. [1]	Unopened flower spathe
*Corypha elata or utan (Bajur palm, Buri palm, sabal palm)*	SE Asia: Indonesia, Philippines; Unimproved cultivated or managed palm [1]; Strong candidate for management or domestication with agroforestry potential [32]	Philippines: fermented drink, vinegar, muscovado sugar [17]; sago, clothes, hats, furniture, ropes from leaf fibre, fences from leafstalks, wood for building [10]	Immature inflorescence growing point or flower stalks
*Corypha umbraculifera (Talipot palm)*	Tropical rain forest to 600m; S & SE Asia; Unimproved cultivated or managed palm [1]; Strong candidate for management or domestication with agroforestry potential [32]	Sugar, wine [1]; starch from trunk, leaves for matting, paper [1]
Elaeis guineensis (Oil palm, African oil palm)	W Africa: Cape Verde, Senegal (Casamance), Guinea, Ghana, Ivory Coast, Nigeria; Madagascar [4]; Indonesia [6]; Improved cultivated palm [1]; also wild in Nigeria [31]	Wine, strong alcohol (over 80%) [18]; edible oil, heart, leaves for thatch, weaving, petioles for fencing, construction [1], medicinal and cosmetic uses [19]	Felled or standing trees terminal bud or male or male and female inflorescences
*Gronophyllum microcarpum*	SE Asia: Indonesia (Moluccas) [35]	Sap; starch; leaves for thatch [35]	Inflorescence [35]
*Hyphaene coriacea*	SE Africa; Madagascar [32]	Wine	Terminal bud
*Hyphaene petersiana*	SW Africa: Namibia [32]	Wine [32]	Flower bud [32]
Hyphaene shatan	Madagascar	Wine	Terminal bud or basis of the inflorescence [4]
Hyphaene thebaica *(Doom or dum palm, ginger-bread palm)*	Semi-deserts & deserts of E Africa to 600m [1]: Djibouti [20], Eritrea [21]; Unimproved cultivated or managed palm [1]; Candidate for improved management [32]	Wine; edible fruit and heart, fruit for medicinal use, leaves for weaving [1]	Terminal bud or only future leaves
*Jubaea spectabilis (Chilean Wine Palm)*	Chile	Wine	Trunk of felled trees
*Mauritia flexuosa (Swamp palm, moriche, buriti, muriti, aguaje, fan palm)*	Tropical rain forest, areas subject to floods ; Northern S America (Bolivia, Brazil, Colombia, Peru, Venezuela) [32]; Semi-wild or wild palms [1]	Wine; edible oil, fruit, heart, starch from trunk, leaf fibre for rope, petiole for cork, trunk for wood [1]	Trunk or inflorescence stalk of felled trees
*Nypa fruticans (Water coconut, nipa palm, apung palm)*	Tidal areas (salt tolerant); Asia: India, Sri Lanka, Bangladesh, Burma, Thailand, Cambodia, Malaysia, Indonesia, Philippines; Pacific; Mainly, wild almost pure stands [22]; Unimproved cultivated or managed palm [1], also managed: planted pure stands; replanting not necessary [23]; Candidate for improved management [32]	Fresh juice, toddy, arack, sugar and vinegar; edible almonds, salt from the petiole and stem ashes, frond for thatching and weaving [11] [1] [36], cigarette paper [24], leaves made into shingles for roofing and walling purposes, hats, mats, raincoats [25]	Inflorescence spadix
Orbignya martiana (Babassu Palm) *O. cohune*	Wide range of edaphic and climatic conditions; Tropical Latin America: Brazil [33] O. cohune in Honduras [37]	Wine; fruit as source of food, feed, oil and charcoal; edible heart; salt from burnt stems; leaves for thatch, weaving [33]	Stump hollowed out after felling the tree [33]
*Phloga polystachya*	Madagascar [4]	Fresh juice, wine	Trunk [4]
*Phoenix dactylifera (Date palm)*	Sub-tropical semi-desert and desert [1]; N Africa: Tunisia, Egypt; S Asia: India [34]; Improved cultivated palm [1]	Wine, arack, sugar; edible fruit, leaves for thatch, weaving, trunk for wood, etc. [1] [34]	Terminal bud or only defoliation
*Phoenix reclinata (Dwarf date palm, raffia palm)*	SE Africa near the coast [26]; W Africa: Cape Verde, Senegal (Casamance), Ivory Coast [6] [19] [27]	Palm gur, wine; edible fruit and heart; very resistant building material; thatch, weaving bags and mats, sponges, toys [19] [34]	Terminal bud or side of young stem or inflorescence [32]
*Phoenix sylvestris (Wild date palm, toddy palm, silver date palm)*	Tropical rain forest to 1,500m; S Asia: India, Bangladesh; W Africa: Ivory Coast; Unimproved cultivated or managed palm [1]; Bangladesh: plantations [28]; Domestication potential within agroforestry systems [32]	Wine, sugar [32]; edible fruit, fuel, mat, building material, fencing [12]; leaves for weaving [1]	Terminal bud or side of young stem
*Pseudophoenix ekmanii*	Caribbean: Dominican Republic [32]	Wine [32]	Terminal bud on felled trees [32]
*Pseudophoenix vinifera*	Caribbean: Dominican Republic, Haiti [32]	Wine [32]	Terminal bud on felled trees [32]
Raphia hookeri, R. vinifera, R. palma-pinus, R. regalis, *R. sudanica, R. ruffia (Raffia palm, African piassava, African wine palm, Rattan Palm)*	Tropical rain forest, swamps, areas subject to floods [1]; W Africa: Ivory Coast, S Mali, S Burkina Faso, Benin, Nigeria, Eastern Cameroun [8] [29]; R. hookeri: S Benin: wild stands that do not need replantation when adult trees are tapped; cultivation also practised [30]; R. ruffia: Wet areas of Madagascar [4];Semi-wild or wild palms [1]; Candidates for improved management [32]	Fresh juice, wine, strong alcohol [30] [32]; edible fruit, oil, leaves for fibre, thatch, petiole, leaf rachis for building material [1]	Terminal bud on felled trees or terminal inflorescence on standing trees

References: [1] Johnson 1987; [2] Davis 1988; [3] Sumadi 1988; [4] Decary 1964; [5] Niang 1975; [6] Giffard 1967; [7] Bellouard 1950; [8] Chevalier 1930; [9] Port�res 1964; [10] Fox 1977; [11] Redhead 1989; [12] Abedin et al 1987; [13] Jambulingam and Fernandes 1986; [14] Rangaswami 1977; [15] Dissanayake 1986; [16] Ven�ncio Machado 1974; [17] Abrenilla et al 1988; [18] Essiamah 1992; [19] Blanc-Pamard 1980; [20] Jahiel 1993; [21] Fanshawe 1966; [22] Hamilton and Murphy 1988; [23] Encendencia 1985; [24] Kiew 1989; [25] Quimbo 1991; [26] Cunningham 1990; [27] Adand� 1954; [28] Annett 1913; [29] Fyot 1973; [30] Profizi 1988; [31] Okereke 1982; [32] Johnson 1997; [33] May et al 1985; [34] Baumer 1995; [35] Mogea 1991; [36] Pearce 1991; [37] Balick 1990; [38] Stewart 1994.


Latin name	Litres of sap/tree - Sugar % - Sugar/tree	Age of first/last tapping (years) - Height of tree	Frequency of tapping - Season - Duration	Sugar/ha - Tree/ha - Work needed

*Arenga pinnata or saccharifera*	3-6l/day per peduncle (2 to 4 can be tapped in a tree, each for 1 to 4 months [1]) - 11 to 16.5% - 0.4kg/day [2] [3] [4]; North Sulawesi, good trees: 35 to 40l/day [5]	From 5-12 to 2 to 5 years later [2] [4] [1] - 7 to 15-20m [1]	Once to twice a day [2] [1] - 2 to 9 with average 5 months [2] [4]; all year round in East Kalimantan and North Sulawesi [1]	Indonesia: 3kg/day/ha (256 trees/ha) [5] Pure plantation in North Sulawesi: 20,000kg/year (80kg/tree/year) [6]
Borassus aethiopium	5 to 6l/day 2 months/year through bruising and cutting bunches [7] 100l/tree/year [8] - Ivory Coast and Senegal: 200 to 300l/tree only during one month with 9-12l peak on the 20est day [9] [10] - Up to 20% [50]	10 to 20m high [11] [9]	Twice (Ivory Coast) to three times (Senegal) a day [11] - 4 to 6 months before death [11]; 8 months if clayey or clay-silted soils [11]; Senegal: all year round but on different areas: plateau in rainy season; low areas close to forest in dry season [9]; better yields at the end of the dry season: February and March [10]	500 trees/ha [12]
*Borassus flabellifer*	3l/day the first 3 months and 1.5 the 3 last - 8 to 20% - 0.5 to 0.25kg/day [13] [14] [12] [15] [16]; 150l/tree/year[3]; 60-65kg sugar/tree/year [8]; 370-400l/female tree/year [12]; 5 to 10l/day average with minimum of 1 and maximum of 20 (Paulas 1983, Tjitrosoepomo et Pudjoarinto 1983 cited by [17], [18]); 11 to 20l/day in Indonesia [51] Borassus madagascariensis: 1l/day during one week [19]	20-30 to 50-60 years old [2]; 15-17 to 45 with optimum 25-33 in Bangladesh [20]; Burma: up to 80 to 100 years of production [21] - From 10-12m to 20-30m [14] [22]; Burma: from 4-5m when 15-20 years old [21]	Once or twice a day [2] [13] (night collection = twice the day collection [23]) 5 months, every 3rd year rested [2] - Mar. to Oct. in Upper Burma [24]; Nov. to April-May in Cambodia [13]; Jan. to May in Tamil Nadu [22]; Dec. to July [16] - Outer arc of the lesser Sunda islands (Indonesia): dry season (Apr. to Nov. with 2 peaks, one in April-May and a higher one in Sep.-Oct.) [23]	1,500kg with 250-300 trees/ha [7]; 100 trees/ha in Tamil Nadu [22]; Madura (Indonesia): 70kg/year per good tree planted at 6m interval [23]; Cambodia: 70kg/tree/year (42kg sugar & 28kg molasses)[16]; Cambodia: if 200 trees/ha: 18,000kg/year [18] - Upper Burma: average tapper: 40-50 trees twice/day: males during first part of 8 month season, females during second part [24]; Cambodia: good tapper: 25-30 trees twice/day (4x2hrs/day) [16]; 20 trees twice/day if assistant at the base of the tree [18]
*Caryota urens*	5 to 27l/day (up to 60) [2] [25] [3] [26] - 15-16% including 9.8 to 13.6% sucrose [15] [14] [26] [27]	10-15 to 2 -10 years later [2] [14] (flowers only every 2-3 years [26]) - 15-20m [14]; much shorter / Borassus flabellifer when starts flowering [15]	Twice a day [26] 3-6 months [26]	110kg of sugar/tree/year [26]
*Cocos nucifera*	1 to 1.5l/day [14]; 1 l/day/spadix [25]; Sri Lanka: 0.9 to 1.5l/day all year round (Nathanael 1955 cited by [6]) [28] ; South Sumatra: 5 to 15l - 4 to 6% - 0.44kg/day [29] - 9.5 to 18% [25] [14] (Norris et al. (1922) cited by [28]); Sri Lanka: 2.2l/day during 14 months from hybrid (Typica x Pumila) [30] Malaysia: 17% sucrose [31]; Sri Lanka: 16-22% sugar [32]; Assumption [33]: 30 kg/year	South Sumatra: 7 years; East Java: during more than 20 years [29] - 20-30m	Throughout the year [14]; average of 2 spathes (up to 3) tapped simultaneously; a new flowering axis produced each month [29] [33]	Assumption [33]: 4,448kg/ha (148 trees); Sri Lanka: Tall variety (Typica): 13,590kg/ha (158 trees/ha); Dwarf variety (Pumila): 3,950kg sugar/ha (247 trees); hybrid of 2: 19,027kg/ha [30] - 25-30 trees/tapper/working day if palms not connected; 75-90 if palms connected (Nathanael 1970 cited by [17]); about 5hrs/day/35 trees [29]
*Corypha elata or utan*	20l/day for 3-4 months [49]; Up to 45.2l/day [34] - 17% sugar - Several hundreds of kg of sucrose to death [6]; 2,699l during 132 days (Gibbs 1911 cited by [34])	Flowers once when 20 - over 100 years old and dies [23] - 20-40m [35]	Twice daily [2]
*Elaeis guineensis*	Felled tree: 3-4l/day [10] - 14% - Sprecher von Bernegg 1929 cited by [6]; depends on number of male inflorescences available: Senegal (Casamance): 5l/day for a tree with 8 inflorescences (4 females and 4 males) [36] ; 149 l/month (Hartley 1977 cited by [34]) - 11-12% [37] Male Inflorescence of palms that do not produce enough fruits for oil: 26l/tree/year (about 0.2l/day) [38]	6 to 10 or more years for felled tree technique [34]; Senegal (Casamance): 10 to 15 years of production [36] - 7-10m [36]	Twice (Ivory Coast) to three times/day [10] [34] - 2 weeks to 4 months on felled trees [34]; Nigeria: yields highest at the beginning of rains (increased physiology activity in March-April) and end of rains (more male inflorescences for tapping in Oct.-Nov.) ([38], Hartley 1977 cited by [34]); Senegal (Casamance): October to June (dry season); sap less concentrated during rainy season [36]; all year round (4 to 5 months/tree) [38]; 2-3 weeks per male inflorescence [47]	31-34kg sucrose/semi-wild oil palm/year (Sprecher von Bernegg 1929 cited by [6]); >450kg/year (150 trees) [38]; Nigeria: 1,200kg/year when male and female inflorescences tapped (148 trees) [39]; Ghana: 20 to 100 trees tapped to death per farmer per season; Nigeria: estimate of 14/22 trees tapped per tapper per working day of 5/8hours respectively (male and female inflorescences): one hectare of 150 trees requires 2 tappers as only one third of the palms are tapped at any one time; average of 11mn/tree twice a day [39]
*Hyphaene thebaica*	Up to 4l/day [40]	Up to 30m [40]	Up to 4 times/day; 14-25 days and then 2.5 year rest before next tapping [40]	From 20-25 to 40-50 trees/tapper [40]
*Jubaea spectabilis*	340 to 450l to death [17] [33]		Once/day - Several weeks [17]
*Nypa fruticans*	0.6 to 1.8l/day [3]; 0.5l/day/fruiting head for about 1 month; up to 26 fruiting heads/tree, of which 2 can be tapped at any one time [41]; 43l/year - 12% [25]; 0.5 to 3l/day - 14-15% - 36 to 40l/tree/year [42]; 14-17% (Halos 1981 cited by [43])	4-5 to 55 [13] [25] Trunkless: the inflorescence is carried on a stalk borne on the root stock and sticks out of the ground to a height of about 1m [43]	Once to twice (for high yielding trees) daily [2] [41] [43] - 6 months (3 months/tree between July & Dec.) [13]; throughout the year [41]; Malaysia: all year round (Dennett 1927a cited by [43])	3,000kg/year ([2] [25]) to 10,000kg/year (2,000 plants) [13] Gibbs cited by [4]; 3,800 to 4,500kg/year (2,500 plants, 700-750 sap producing) [42] [43]; Peninsular Malaysia: 20,300kg/year (500 plants with 2 stalks tapped per plant/340 days/year) Watson cited by [41]; Sumatra estate: 22,400kg/year as a conservative estimate (Johnston cited by [43]); Papua New Guinea: 28,000kg/year if 50% of palms flower yearly and mean tapping period of 100 days [48] - Sumatra estate: 38 workers/10ha plot: 30 on tapping/collecting, 5 on maintenance and preparation of stalks through kicking, etc., 2 on syrup transport and one overseer (Johnston cited by [43])
*Phoenix dactylifera*	19l/day from a notch in the stem (Bose 1927 cited by [34]); 9l/night during 50 nights [50] 8-10l/day (500 to more than 1,200l/tree) - 10% - 1kg/day [52]	40 to 50 years of production [50]	Twice daily [2] - Autumn in Mediterranean regions [50]; Warm season, once every five years for 3 to 4 months [52]
*Phoenix reclinata*	Sap rich in sugar [10]	10m [10]
*Phoenix sylvestris*	10 to 40kg sugar/tree/year [3] [42] [44] - 10 to 14%	Bangladesh: from 5-6 to 55 years, optimum 15-25 [20] [44] 0.5 to 10-16m [44] [14]	Once daily [44] - Bangladesh: from Nov. to March with greater yields in mid winter [44]	Bangladesh: > 8,000kg/year (1,250 trees) [44] - 20 trees prepared in 1h20' [42]; 50-60 trees/day/tapper; 300-400/year [44]
*Raphia hookeri*	2l/day [45]	Flowers once when 7-10 years old and dies; 4-7m up to 8m (or less) [46]	Twice daily [45]
*Raphia ruffia*	2l/day [19]			Madagascar: Trunk incised (4-5 cm deep) [19]

References: [1] Mogea et al 1991; [2] Redhead 1989; [3] Johnson 1987; [4] Davis 1988; [5] Sumadi 1988; [6] Van Die 1974; [7] Chevalier 1930; [8] Bellouard 1950; [9] Port�res 1964; [10] Blanc-Pamard 1980; [11] Niang 1975; [12] Giffard 1967; [13] Crevost and Lemari� 1913; [14] Rangaswami 1977; [15] Dissanayake 1986; [16] Romera 1968; [17] Kovoor 1983; [18] Khieu Borin and Preston 1995; Khieu Borin 1996; [19] Decary 1964; [20] Abedin et al 1987; [21] Lubeigt 1977; [22] Jambulingam and Fernandes 1986; [23] Fox 1977; [24] Aubert 1911; [25] Magalon 1930; [26] Dissanayake 1977; [27] Theivendirarajah et al 1977; [28] Grimwood 1975; [29] Levang 1988; [30] Jeganathan 1974; [31] Naim and Husin 1984; [32] Theivendirarajah et al 1979; [33] Pethiyagoda 1978; [34] Essiamah 1992; [35] Abrenilla et al 1988; [36] Adand� 1954; [37] Eze and Uzoechi Ogan 1988; [38] Tuley 1965a; [39] Udom 1987; [40] Audru 1985; [41] Kiew 1989; [42] Fealy 1925; [43] Hamilton and Murphy 1988; [44] Annett 1913; [45] Fyot 1973; [46] Profizi 1988; [47] Okereke 1982; [48] P�iv�ke 1985; [49] Johnson 1997; [50] Baumer 1995; [51] Mogea 1991; [52] Barreveld 1993.

Table 1: Yields of alcohol fuel from different sources (Hamilton and Murphy 1988)

Source of alcohol fuel	Annual yield (litres/ha)

Sweet potato	6,750 - 18,000
Tapped nipah	6,480 - 15,600
Tapped coconut	5,000
Tapioca	3,240 - 8,640
Sugarcane	3,350 - 6,700

Table 2: Data on the composition of sap and jaggery from Borassus flabellifer

Product and reference	Fresh sap [1]	Fresh sap before cooking [2]	Fresh sap stabilized with sodium benzoate [2]	Jaggery [3]

Specific gravity	1.07
pH	6.7-6.9	6.00 to 4.00	6.45
Nitrogen	0.056 g/100 cc
Protein	0.35 g/100 cc	0.10 to 0.20%	0.10%	1.04%
Total sugar	10.93 g/100 cc	14.00 to 9.40%	15.78%
Sucrose		12.30 to 8.50%	15.04%	76.86%
Reduced sugar	0.96 g/100 cc	0.80 to 3.50%	0.62%	1.66%
Fat				0.19%
Minerals as ash	0.54 g/100 cc	0.10 to 0.30%	0.17%	3.15%
Calcium	Trace			0.861%
Phosphorus	0.14 g/100 cc			0.052%
Iron	0.4 g/100 cc			11.01 mg per 100g
Copper				0.767 mg per 100g
Vitamin C	13.25 mg/100 cc
Vitamin B₁	3.9 IU
Vitamin B complex	Negligible

References: [1] Paulas and Muthukrishnan 1983a cited by Davis and Johnson 1987; [2] Romera 1968; [3] Roy 1951 cited by Morton 1988.