SCIENTIFIC NAME: Nephelium lappaceum
FAMILY: Sapindaceae

This document is not a conventional crop summary as prepared for other established economic crops in Queensland. Rather, it is a summary of very little evaluation and experience in northern Australia, coupled with information from Asian centres of production. The paper poses many unanswered questions and hopefully, will stimulate thinking by those people attempting to grow the crop, and thus accelerate the evaluation of its potential in what is probably a marginal climatic situation.

We in the. D.P. I. are not currently recommending planting rambutan on other than trial basis, and this paper should not be seen as to be encouraging widescale development. There is no doubt that some cultivars would be well accepted in the market place and the crop could have a future in northern Australia. However, whether or not production will ever be economic yet remains to be proven.

Rambutan (Nephelium lappaceum) is indigenous to Malaysia. Clonal trees are of medium size, from 7 to 12 metres high, but seedlings often reach up to 25 metres.

The leaves are petioled, alternate and pinnate, with 2 to 4 pairs of leaflets. The leaflets are short-stalked and vary considerably in shape from elliptic to obovate. When young, leaflets are soft, light green or yellow to red and pubescent along the veins.

The inflorescences are both axillary and terminal, erect, widely-branched, rusty pubescent and bear numerous greenish/white flowers. The cultivated rambutan is usually monoecious, flowers of both sexes, borne on the same inflorescence. However, the individual flowers are effectively unisexual. With seedling populations, an average of 50% of trees will develop male functional flowers and thus never bear fruit. The balance will have predominantly female (pistillate) functional flowers with 0.1 to 2% of male (staminate) functional flowers, or occasionally be entirely functionally female.

The commercially-selected cultivars are predominantly of the second type (98 to 98.9% functionally female) but with the odd ones (eg. Seematjaan from Indonesia) producing 100% functionally female flowers. Some cultivars, - eg. Rongrien and Chompoo from Thailand, obviously have a very low number of functionally male flower.s and chemical treatment with S. N.A. is used to promote development of male flowers and thus increase pollen production.

The fruit is a drupe, with only one of the two ovules developing and ripening to various shades of red or yellow as per cultivar. Fruit range from 30 to 80 mm in length and weigh between 20 - 60 g. The pericarp is thick and covered with closely fitted tubercles, each terminated by a soft pliable spine of variable length which gives the fruit its 'hairy' appearance. Colour of spines may vary from green to yellow to red. Beneath the pericarp and completely surrounding the seed is the aril - the edible portion which is translucent white to yellow-white in colour. The aril is usually juicy and sweet, but thickness and eating quality vary according to the cultivar. Aril recovery varies from 30 to 58% of total fruit weight. Seeds are soft and may be either elliptic, ovoid or oblong, ranging from 15 to 30 mm in length. The aril may or may not (according to cultivar) peel cleanly from the seed, with testa attached. Unfortunately, whereas a large number of cultivars are 'freestone', in the majority of these the testa clings to the aril and is highly objectionable - in fact the main limitation to the popularity of the fruit for European tastes. Cultivars with large fruit size and thick aril are generally the preferred canning types.

The pulasan (Nephelium mutabile) is similar to rambutan in most respects, but the fruit tubercles are much shorter. Generally there is less problem with testa, but yields are not as good as with rambutan.

Rambutan appears tolerant of a range of soil types from heavy clays to sandy loams. However, the lighter the soil, the more particular are the requirements for satisfactory irrigation and fertilizing. High organic matter levels are very beneficial to growth in the first 3 to 4 years and also tend to lessen problems associated with trace element deficiencies. Thus, mulching with bagasse, straw, peanut shells, cane trash etc. is recommended. A pH of 5 to 6.5 appears preferable, and liming should be avoided or minimal - at least up to the beginning of cropping.

Rambutan is often quoted as suffering in particular from iron deficiency, and chelated forms are recommended to be applied in solution to the soil. However, if it is a problem, then it is most evident in nursery stock or in young trees in the field, and high organic matter levels and moderately low pH are usually sufficient to prevent deficiency development.

Drainage requirements are not as critical as for most other tree crops, but areas where water lies for some time after heavy rain should be avoided, or the very least, mounded.

Under grass or cover crop situations, rambutan can be grown on quite steep slopes, but in north Queensland, such sites are more difficult to irrigate, and provision of windbreaks is even more critical. Gently sloping sites on deep (preferably colluvial or alluvial) soils are preferred. Alluvial soils on flat land are preferred in S.E. Asia, but in north Queensland, some such sites may be too cold for satisfactory production. However it is only from use of a range of sites that we will get sufficient data to make precise recommendations.


Whereas in the Asian tropics rambutan is often grown at altitudes up to 650 metres, the preferred range is below 300 metres and within 15 degrees of the equator. The tree has been selected in hot wet environments - with few months of the year having less than 100 mm of rain and generally 2,000 to 3,000 mm per. year total precipitation. In northern Australia irrigation is a necessity and factors such as low humidity and cool winds obviously have a deleterious effect on growth - if not on yield. We suspect rambutan has poor leaf stomatal closing capacity because of evolution in wet climates, and leaf burn is probably directly associated with stress caused by low humidity and constant air movement (wind) in winter and spring. Our selection for preferred cultivars may arise from these with relatively small leaves and more effective stomatal closure. Some cultivars eg. R137 and Jit Lee are obviously in this category, whereas R4 and R7 and R99 are just as obviously easily stressed. Stress susceptibility and stomatal closure characteristics have yet to be studied in Australia. However, since yield in fruit quality determination are the most important factors, those cultivars with adaptability for this environment will be selected as a matter of course. On the other hand, understanding of the trees physiology is important in relation to management practices - which can be modified to improve yield.

Windbreaks are obviously essential - to reduce air flow and to create a more humid microclimate. Rambutans grown alone in a bare exposed field do not perform nearly as well as when interplanted with windbreak trees such as Leucaena leucocephala.

Rambutan cultivars vary in reaction to cyclonic winds. Generally, trees which lose leaves readily in strong winds suffer less damage to limbs than do trees which retain leaves (eg. mango). Cultivars which appear to retain leaves strongly are Chompoo, Rongrien, R9, Rapiah, R165 and Binjai, whereas at the other end of the scale, R4, R99, R7 and R170 defoliate readily. However, many cultivars have not yet been studied. Loss of leaves obviously delays flowering - at least until the tree has matured one or two new flushes.

We do not get significant growth in northern Australia in months with mean monthly temperature normals (average of both minimum and maximum temperatures) below 22° C. This on the average, provides 9 growth months for Cairns but only 5 months for Brisbane. We do not know how critical sustained growth is for yield, but obviously at least 8 months are required. At Nambour, trees flower and fruit, but defoliation in winter coupled with a long period of fruit development (up to 6 months) precludes satisfactory fruit-filling and fruit quality. Some cultivars may prove to require less 'heat units' than others, but the most southerly limit for even mediocre production will probably be Sarina. Even in Cairns (17° S) the climate is definitely marginal. The period from anthesis to fruit maturity in north Queensland varies from 3½ to 4 months for spring flowering (Sept/Oct.) to 5 months for autumn flowering (Feb./ March).

Rambutan is intolerant of frost - but relatively hardy when compared with other tropicals such as mangosteen, and may possibly tolerate brief periods of temperature down to 3 or 4° C However, growth is dependent on satisfactory soil temperature, which is a reflection of ambient temperature over a number of weeks and is not stimulated by one or two 'warm' days.

Odd cultivars such as R134 appear to provide new leaf flushes throughout the year in Cairns, although there is a definite slowing down in July to September.

At this stage, few cultivars introduced into northern Australia have fruited, and suitable recommendations can only be based on 'tasting' during overseas trips and information from overseas references. The main cultivars in Asia are as follows:
N.B. (*= not introduced or not well established in Australia), (y = yellow pericarp).

Malaysia: Cultivated clones are mainly R3 (Peng Thng Bee), R4 (Ya Tow), R6* (Ayer Mas), R7 (Kepala Besar), R9 (Tau Po Cheng), R86*, R99 (Triang), R134, R137, R139, R154*, R156(y), R160 (Khaw Tow Bak), R161, R162 (Oh Heok) , R163 (P.J.), R165(y) R1-6:7, R168 Chee Embong), R169* (Lychee), R170 (Deli Cheng), Gading (y), and Hew.

The current recommendations in Malaysia for fresh fruit cultivars are R3, R134, R156, R161 and R170. For canning, the range is the same plus R7, R9, and R99. The recommendation is based on a combination of fruit quality, yield and field performance. Of the fresh fruit varieties, only R3 has yet been evaluated at Kamerunga,. It is a slow-growing, precocious-fruiting cultivar with good fruit characteristics and is relatively freestone with only a minor tendency for the seed testa to adhere to the aril. However, the tree is very susceptible to canker caused by Botryodiplodia theobromae.

The cultivar R4 is well-established in north Queensland, is vigorous and easy to propagate; fruit quality is good but flower and fruit-drop common. R168 is also well established, but fruit quality disappointing. Stephens 1, a cultivar raised from seed introduced by Ernie Stephens from Keravat (P.N.G.), in 1937 has excellent flavour but is slightly stringy and very much clingstone. It flowers irregularly, but on the other hand, sets good crops. One or two other local seedlings look interesting but are not evaluated.

Thailand: Only 2 cultivars are widely planted in southern Thailand - Chompoo (y), (Seechompoo) and Rongrien. Because the Thais prefer sweet crisp flavour and very little juice, these two cultivars are preferred by them. However, in both, and particularly, Rongrien, testa adherance to the aril is a bad feature. Other cultivars are Bang Yi Khan*, See Tong*, Nam Tan Kruad (y)*, Jit Mong (Jet Mong = R4), Penang No.4 and Gulah Batu. Bang Yi Khan, See Tong and Nam Tan Kruad are for European preference, possibly the best Thai cultivars. Sadao*, Jeh Drabaru*, Tawee*, Arkorn*, Nga Kulaw*, Colan* and Patavia* are also grown on a minor scale.

Indonesia: Rambutan has been grown in Indonesia for centuries, but many of the various cultivars selected are not easy to locate and not very well described.

The main cultivars are Atjeh Kering, Manis*, Lebakbulus, Silengkeng, Pao Pao*, Sinyonya (Seenjonja), Aceh Garing*, Aceh Kuning*, Sitandkue*, Simacan (Seematjaan), Binjai, Aceh Gendut*, Aceh Rapiah, (Rapiah). Only Lebakbulus (this cultivar is possibly the same as Maharlika). Silengkeng, Binjai, and Rapiah are freely available as nursery trees have been imported. Simacan (Seematjaan) and Sinyonya (Seenjonja) have been imported via the Philippines - but not yet authenticated.

Most Indonesian cultivars have a tendency for the aril to detach together with the testa, although Aceh Gendut is reputed to be freestone and testa-free. One other cultivar, 'Bogor', selected by Mr. J. Anderson, is highly recommended and has been imported.

Singapore: The industry in Singapore is very small, but has one very good cultivar Jit Lee (=Deli) which is one of Indonesian origin. Other cultivars are R37, Ah Kow Cheng* and Kim San Cheng*.

Philippines: A few Indonesian cultivars were imported into the Philippines in the 1930s - Seematjaan (Simacan) Seenjonja (Sinyonya) and Maharlika*. These plus Maharlika 'Belen' (a seedling from Maharlika grown by the late Pedro Belen) are the main cultivars, and very few of the seedling progeny selected from them (apart from Maharlika Belen) appear to be highly regarded. The indigenous 'Bulala' is an interesting fruit and grown commercially, but not of the quality of introduced cultivars.

N.B. There has been interchange of cultivars between all the countries listed and some have been re-named locally. For example, it has been found that the Thai cultivar Jit (Jeh) Mong is identical to the Malaysian R4. Hence it is likely that during evaluation in Australia more duplications will be identified. Further confusion arises when people name seedlings after the parent tree. This should be avoided, since progeny are never identical to the parent.

Flower Induction
Initiation of flowering in rambutan is not thought to be a photo period response, and in north Queensland it can occur both in periods of increasing and shortening day length. In fact, it may occur at any time between July and March. In Malaysia, there are two main flowering periods - March/May and August/October (although not 2 every year), but variation in flowering time from district to district throughout the relatively small country is substantial. Differentiation (the laying down of flower formation cells within the branch terminal) is said to precede flower panicle emergence by only 3 to 4 weeks.

A dry spell or growth check is reputed to stimulate flowering, but in Cairns, February or March flowering does not usually fit this pattern.

In southern Thailand (12 deg. S), there is one substantial flowering in March each year - and why trees behave this way there and not as in Malaysia (2 main flowerings) or north Queensland (variable flowering over a long period - July to February) is difficult to fathom. However, for a substantial flowering in our conditions, it is obvious that a good vegetative canopy is required. If trees come out of winter with a poor cover, they tend to have at least one leaf flush in the spring, prior to flowering.

Also, a late crop (harvesting April to June) tends to restrict development of flowering for the next crop until at least the following December/January. Why, however, the cultivar R3 (and possibly others to be determined), tends to bear a number of small crops successively on different branches over the summer and autumn is difficult to determine. Cincturing of main limbs (saw cut) has proved to induce flowering, but not necessarily a heavy one on all terminals above the cut. Whether one cinctures during a leaf flush or after leaf hardening for best effect remains to be determined.

Two types of flower panicles on any one cultivar have been observed the best where there are strongly-developed terminals with a good length of foliage along the branch. This type appears to develop a long flower panicle - both terminally and from axillary buds (although panicle length and form also varies between cultivars. On weaker terminals, the flower panicles tend to be shorter and thus have less fruit.

Root pruning may also induce flowering. Whereas in Malaysia trees are not cultivated, the Thais tend to work up the ground under the canopy to a depth of 50 to 100 mm approximately 2 months prior to normal flowering time.

In summary, it is as yet not well-understood what factors cause flower differentiation (except that stressing trees can assist), and what precise cultural procedures would be required to bring a crop in at a planned time in northern Australia.

Some people may suggest that a long period of cropping is desirable (particularly for home garden trees) - but on the other hand, if flowering is very erratic and spread over different parts of the trees at different times, then pest and disease control, harvesting and pruning operations become more difficult. In our climate, it appears unlikely that we will achieve more than one crop on any one terminal during anyone year.

Flower and Fruit Development
Flower panicle emergence may start as early as July, but satisfactory fruit set appears to occur only from September or later development. Satisfactory 'set' is apparently dependent on the number of male functional flowers in each panicle, adequate insect pollinators (our native bees seem to be satisfactory, if present). Warm weather and adequate wind protection. Trees in windy exposed situations tend to set poorly, but this may also be due to water stress being accentuated (particularly rapid transpiration - leaf stomatal water loss by air movement). It is obviously important to irrigate well during flowering if a water deficit exists. The Thais start irrigating and fertilizing at the first sign of panicle development. Also, two Thai cultivars - Chompoo and Rongrien (particularly Chompoo) have a very low production of male functional flowers. They have found the chemical A.N.A. (S.N.A.) - the Sodium Salt of Napthalene Acetic Acid useful in changing female functional flowers to male functional and thus increase the amount of pollen available.

A solution of 10 to 50 p.p.m. active constituent A.N.A. is sprayed on to individual panicles one metre apart on the tree, when at 25% of flowers open, again when 50% of flowers open, and again when 75% of flowers open.

This, in actual fact, is three applications about a week apart. Some growers use only one application when 75% of flowers are open; 1 metre apart if plenty of pollinating insects are present, and 400 to 600 mm apart if insects are in low numbers. This treatment has been used on Chompoo at Kamerunga with good results. Care is required to spray only individual panicles. Also, some growers graft some branches of individual trees over to a cultivar which is 100% male functional. It appears that in Malaysia, Singapore and Indonesia, cultivars with no significant production of male functional flowers have been screened out (on the basis of low yields) however, Seematjaan (Indonesia) is of this type, and male functional trees are sometimes planted to assist in pollination.

In Australia, we will obviously select for both yield and quality, and thus poor pollinating cultivars will probably be excluded from recommendations unless fruit quality is exceptional and adequate set can be bought about by chemical means. Lack of pollinating insects, low humidity and cool winds at flowering time may also present problems in northern Australia but the first impediment may be overcome by introducing bee hives. Overhead misting to increase humidity may be beneficial, but also may be unnecessary, if adequate windbreaks are present and a suitable microclimate formed.

The period from anthesis to fruit maturity in north Queensland is from 110 to 160 days. November/December flowering provides the shortest period of fruit development whilst February/March flowering may result in up to 161 days. We can have a fruit maturity range from December to August. There are, however, obviously cultivar differences both in time of peak flowering, but most of this data is not yet available.

Winter fruit in Cairns appear to mature with satisfactory sugar levels (up to 21° brix in June/July for Chompoo and R162 etc.). Thus there appears to be no real disadvantage in having late fruit, and there may actually be a market advantage. However, whether this crop gives maximum yield potential and whether we can actually control cropping remains to be worked out.

Fruit Maturity
The various rambutan cultivars all have distinctive fruit shape, tubercle length and pericarp and tubercle colour. Colour is a valuable guide to maturity, but can only be gauged by experience with each cultivar. It is easy to judge a fruit mature when in fact the flavour and juiciness can be improved by hanging far several more days. Brix levels (total soluble solids - which are in effect approximate sugar levels) should reach at least 18° and are probably best at 21 to 22° for most cultivars. Yellow-skinned cultivars tend to turn slightly orange when very mature.

Crop Control
Some aspects of crop control have been covered. However, in essence, it is, together with pollination and fruit set, the major 'unknown area' in rambutan cultivation in Australia.

We should look for maximum growth in our trees in the first 3 to 4 years - to get tree bulk to provide a satisfactory cropping framework capable of good yields as from year 5 or 6. Currently, the methods available to us to control cropping are, water relations, nutrition and pruning.

We require a tree with good leaf cover prior to flowering. If water stress is a major factor to induce flowering, no under-tree mulch should be provided after year 4, except perhaps a light application after flower panicle emergence. Water should be withheld and no fertilizer added in the period approaching expected flowering time. Water shoots, poor growth and excessive growth should be pruned out. Cincturing (one saw cut) on major limbs or the single trunk area may assist in promoting flowering; timing has yet to be resolved.

If and when a satisfactory crop is set and harvested; all past fruiting terminals should be pruned back 150 to 250 mm below the lowest axillary fruit panicle. This will induce new terminal growth of suitable vigour. Leaving fruit panicle remains on the tree induces a number of weak growths.

If, at harvest time, whole panicles are removed, then in fact the pruning operation can be done then.

Moderate soil acidity, pH of 5 to 6.5, is recommended, and does help prevent induced trace element deficiencies. Zinc and iron are probably most suspect.

For the major elements, crop removal is probably in the order of the ratio of:


For trees up to the stage of first fruiting, a 'growth' mixture is required: At planting, in and around the planting hole, apply 100 g of single superphosphate.

Thereafter, in August, December and April for the first four years, apply split dressings from an annual total of 50g N, 15g P and 30g K per tree per year of age.

From year 5 to year 10 ( and constant after year 10) apply 80g N, 10g P and 50g K per tree per year of age - split to 2/3 immediately after harvest (or April, if no flowering) and 1/3 immediately after flower panicle emergence (or November, if no flowering).

In addition, in August each year apply 0.5 kg of Dolomite per tree per year of age up to a maximum at year 10.

All fertilizers and lime should be broadcast in a one-metre-wide band centred on the drip line. Required rates after commencement of cropping may, in fact, exceed those suggested above. Trace elements should be applied as required, the basic requirement as indicated by leaf deficiency symptoms or by past results. Critical leaf nutrient levels for major and minor elements have not been established for rambutan. However, leaf analysis will be necessary for rationalization of fertilizer use and maximising yield and fruit quality.

Animal and poultry manures are used in some countries. In Thailand, approximately 35 to 50 kg of chicken manure is applied per tree after harvest.

N.B. Fertilizer recommendations are in terms of element. These must be adjusted in terms of actual fertilizer: eg. 80g N = approximately 170 g Urea; or 400g of Sulphate of Ammonia. 60g K = approximately 150 g Muriate of Potash.

Following planting, the objective is to maintain maximum growth rates and thus there is little necessity to prune in the first 3 to 4 years, other than encouraging a strong framework. Bud-grafted trees tend to produce main branches with wide angled crotches which are very desirable. However, other propagation techniques usually result in a tree form with more acute crotch angles.

Acutely-angled branches should be avoided and young trees pruned to promote wide crotches, and an open vase shape. However all other growth should be allowed to develop unimpeded in the first four years. Subsequently at bearing age, low-lying branchlets, water shoots and weak growths should be pruned back flush to main limbs and the cut surfaces painted with copper oxychloride slurry or proprietary tree wound materials. After each harvest, panicle remains must be removed entirely - preferably the terminal cut back to 150 to 250 mm below the bottom axillary panicle.

Crossing or excessively close branches should also be removed. Individual cultivars vary in natural growth habit from lax (low-spreading) to upright crowns, but generally all provide a dome-shaped canopy if light is not restricted through competition from neighbouring trees. Density should be maintained to allow permanent trees unimpeded lateral development - otherwise, all growth and fruit is pushed up and harvesting and pest control and disease control made more difficult.

Natural grass cover between trees is satisfactory providing the dripline to tree trunk area is kept weeded. However a legume sward is preferred. Centrosema pubescens, Stylosanthes guianensis cultivars, Desmodium heterophyllum, Vigna hosei or Desmodium intortum are all suitable legume covers, but may be difficult to maintain in good condition if low splashing is attempted or regular traffic by machinery is necessary. The appropriate innoculum and adequate phosphatic fertilizer is necessary to establish and maintain legumes. Vine-type legumes such as Pueraria javanica or Macrottilium atropurpureum (Siratro) should be avoided since it is difficult to prevent them climbing the trees.

Mulching is particularly useful in the first 3 to 4 years of tree growth - and any organic waste material is beneficial. Main mulch application should be made just after each wet season.

Tree blocks should have perimeter windbreaks - particularly on the south-eastern and southerly sides. Suitable species are Eugenia cuminii (jambolan), Casuarina cunninghamii , Casuarina equisetifolia (but avoid susceptible forms of Leucaena leucocephala).

In addition, interplanting temporary windbreak trees within rows of rambutan in advance of field planting provides a very good microclimate effect. These inter-plants should be removed in the third or fourth year, and at no time allowed to substantially overgrow the rambutan trees. Suitable interplants are bananas, Leucaena leucocephala (moderately vigorous strains), Gliricidia maculata, and bana grass.

Irrigation is essential throughout northern Australia even in what are considered 'very wet' areas such as Innisfail/Tully. Expected peak requirement is 90% of pan evaporation, or up to 60 mm per week. The critical period is from the first sign of flower panicle emergence until after the first growth flush following harvesting. However for maximum growth of juvenile trees, watering at all times of soil deficit is necessary. Mulching will reduce watering requirement. In-ground lines with under-tree sprinklers are recommended. Preferably the whole of the under canopy area plus a minimum of 1 metre outside the drip line should be watered. Trickle irrigation has been suggested by some authorities but could have problems, particularly if there are not enough drip positions around each bearing tree.

Exclusion of all weeds and vegetative cover under the tree canopy and extending at least one metre outside the drip line is desirable. Contact weedicides such as paraquat, glyphosate and sodium arsenate are preferred.

Once a maturity index for a particular cultivar is established (either based on changes in fruit colour, but tested by taste or Brix level) and the fruit reach that stage, then picking can commence. Some panicles on some cultivars tend to mature fruits uniformly, but others do not. Fruit may have to be taken off in sub-panicle lots. The inflorescence stalk is cut and the branches removed to the packing shed. In Malaysia and Indonesia, most fruit are marketed in bunches intact on the panicle, and deterioration in fruit colour is slower than when fruit are cut off individually. Under our marketing system, however, it is desirable (as for litchis) to separate fruit. Exposed fruit tend to dehydrate and lose colour rapidly and are only attractive for 2 - 3 days after harvest.

However storage in polythene bags at 10° C will give a shelf life of 12 days. A cold dip in benomyl (1000 p.p.m.) will prevent storage rots at temperatures even as high as 26 to 30° C.

Studies have yet to be carried out to test hot water dips (eg. litchis, 52° for 2 minutes in hot water benomyl dip) which could well provide a very satisfactory shelf life even at relatively high storage temperatures.

The rambutan fruit is by virtue of its thick pericarp and tubercles not readily damaged by handling, although the appearance of the tubercles can suffer if compressed during transit.

Rambutan may possibly be best presented in a 4.5 kg polythene-lined carton, or, prepackaged in 'vita film'-covered strawberry punnets as per the recent development with litchis.

Fortunately, whereas we do not have the major pests and diseases affecting rambutan in the countries of origin, Dolabra nepheliae (stem canker), Oidium nephelii (fruit, flower and leaf mildew) etc., we do not have a number of major problems.

The worst fungus in north Queensland is Botryodiplodia theobromae which causes a trunk canker - particularly in the region of the graft union. Approach-grafted trees are most susceptible, and the graft union should be coated with a 50/50 slurry of burnt lime and copper oxychloride at regular intervals for the first 2 years after grafting.

The various species of Rhyparida (black beetles) give constant problems with leaf, twig and flower damage during summer and autumn. Control with trichlorphon or carbaryl (0.1%) is satisfactory, but not persistent. Carbaryl should not be used during flowering. Various species of lepidopterous caterpillars are a particular nuisance in autumn, both on foliage and flower panicles. Synthetic pyrethroids such as permethrin give fair control, but occasionally the organophosphates are necessary. Fortunately there are few insect pests on fruit but mealy bugs are sometimes troublesome. The fruit fly species existing in northern Australia do not appear to sting fruit. In the Philippines some species do sting, but Maharlika 'Belen' is reputed to be resistant.

Overall, rhyparida beetles and caterpillars are the most serious pests due to the massive defoliation which can occur in few flushes.

Flying foxes are a problem on larger trees, and once a few fruit are taken, the rest are in jeopardy. A recent development in flashing Strobe lights (which scare the bats away) looks promising.

Rambutan propagation under our conditions is costly, but may become much more simple as we come to understand limiting climatic conditions and perhaps develop new techniques.

Airlayering is now seldom practised anywhere due to unreliability. Root development after cincturing and peat-balling is often slow. Larger branches are preferred. After roots are apparent and the outer ones turn brownish, the airlayer should be potted and placed under mist or in a humid environment until established. Airlayering in November/December provides quickest results.

Approach Grafting (Sometimes referred to as Inarching)
This is best attempted by matching up seedlings and scion branches on the ground or on platforms in the canopy. Mr. H. Bosworth of Ingham has devised a technique of surrounding the parent tree with heavy arc mesh and sliding supporting boards through the mesh at desired heights.

Young vigorous seedlings and scion branches are preferred since callusing is rapid and the grafted tree can be detached in as little as 6 weeks. Older wood combinations may require up to 10 weeks to complete the process. The best season is January to April when sap flow is strong and frequent rain obviates hand-watering. However, it is desirable to provide overhead spray misting systems. The scion and stock surfaces are matched - preferably low on the stock and high up on green/brown wood on the scion. Cut no deeper than one-half the thickness of each branch, and for a maximum length of 30 mm. Long cuts (up to 60 mm) callus slowly and often poorly and allow invasion of fungi. Ties should be tight, but it does not appear to provide a water proof seal and mastic over the tie. If growth of both stock and scion is vigorous, then the stock above the union can be cut off at 4 weeks and the scion below the union cut (notched) to 1/3 scion thickness. At 5 weeks cut to 2/3 and at 6 weeks, detach the scion. If growth is slow or the wood very mature at the union, it is safer to delay detachment to 8 to 10 weeks. Then move the grafted plant to a closed, high-humidity bed or mist system, or alternatively, enclose the scion in a polythene bag for 2 weeks. Do not remove the ties until the scion has made 300 mm of new growth. Then, carefully cut the stubs and paint with cuprox/burnt lime (50/50) slurry. Approach grafting in potted trees is also useful, and very young scion growths can be used with quick turnover.

Wedge Grafting
This method has only been successfully developed by a few people and results are not consistent - although there is room for much more experimenting. In view of variable results to date, it is preferable to use approach grafting or modified forkert budding.

Side Veneer Grafts
Again results have not been consistent; although promising, there are obvious differences in success with different cultivars.

Modified Forkert Bud Grafting
This is a favoured technique at present, and it allows the maximum use of scion material, particularly when one does not have access to trees for approach grafting.

The technique is relatively simple, but the ease with which buds can be taken from the scion stocks, the lifting of bark on stocks and temperature at time of forcing buds complicates the issue.

Rootstocks used in Asia appear not to be selected for performance and "sour" or poor quality seedling fruit are used. Also, freestone cultivars are easy to seed-clean and thus these types often end up as stocks. It is unfortunate that rootstocks are not cloned, or at least one cultivar recommended for providing even-growth vigorous rootstocks. Perhaps we should use seed from vigorous cultivars such as R168. In Malaysia, many 'rogue' trees are often thought to result from using mixed stocks from various sources.

Seedlings of the Queensland selection 'Stephens l' appear to make poor rootstocks.

Seeds should be freshly-extracted, soaked in a solution of captafol and laid on their sides and barely covered. Best results appear to be from sowing individual seeds in a peat/sand (50/50) mix in 50 mm peat pots. Once established, these seedlings can be repotted with minimum set-back. Seed sowing in summer is much preferable to winter or early spring, even when bottom heat is provided for the latter.

A preferred potting mix is 2 parts coarse sand plus 1 part peat or bagasse or compost plus one of humic free-drainage loamy top soil.

Seedling growth is best in a hot humid environment. Well-grown seedlings are big enough to bud graft at 10 to 15 months of age, when stem diameter at about 100 mm above the potting medium is 8 to 10 mm.

Frequently applied (weekly) foliar fertilizers alone or in combination with slow-release, soil-applied fertilizers are best for seedling growth. However, both seedlings and newly-grafted trees are very susceptible to leaf burn and defoliation, both from excessive fertilizer application and a number of the insecticides.

Humidity does appear to favour bud 'sweating'.

The bud should be kept as dry as possible and only hand-watering attempted for 2 to 3 weeks after grafting. Sunny conditions appear to favour 'takes'.

A new patch bud technique which we are just starting to investigate is 'punch budding', as used in macadamias. If successful, it should prove a boon since it is much easier to detach buds from the scion stock and stock with the hollow punch, and it may be possible to graft in Spring and Summer when it is normally difficult to remove scion buds.

With the normal modified forkert bud, it would be very desirable to perform this in the months September to December. However, usually it is difficult to satisfactorily remove buds from scion sticks in these months. On the other hand, conditions are variable, and in August/September 1981 it has been found possible to lift buds.

It would be very desirable to grow plants from cuttings - being cheaper and possibly giving more uniform trees. However few successes have been achieved - and only when the material is cinctured. Again, some cultivars may be easier than others. Further, it may be possible to 'stool' rambutan as is used in providing deciduous fruit rootstocks, however results to date are not encouraging.

Bud-grafted trees appear to produce wide-angled crotches which are desirable to give tree strength and resistance to cyclonic winds. For trees produced from airlayers, approach grafts and side veneers, wedges, growth is more upright and careful formative pruning is necessary to modify crotch angle. Trees for field planting should be 'hardened off' and at least covered with temporary shade (tree branches etc.) for a few weeks after setting out.

Sites should be deep-ripped and preferably provided with compost or fowl manure some months before planting. Take care not to plant trees too deep, and in fact, elevate slightly on a low mound. Water in, rather than exert pressure around the plant, after filling the hole.

Tree guards made with wind break 'sarlon'-type material or hessian are beneficial - both to reduce airflow and provide shade for newly-planted trees. Trees should be staked for the first two years.

At planting - 2 additional seedlings - (to give a triangle effect) can be planted 150 to 300 mm from the grafted tree - and later inarched to give increased tree stability.

Due to insect damage which can severely restrict development of young trees, it may be of advantage to hold trees in the nursery until they are relatively well-advanced (8 to 12 months after grafting) before planting out. However, containers should be sufficiently large to prevent a 'root-bound' condition. September to February are the preferred months for planting out, but care must be taken to prevent sunburn when planting October to January.

Various cultivars have different growth rates and tree forms. For example R4 is very vigorous and has a wide crown and should be planted at least 13 metres apart. R168 is also vigorous, but more upright. On the other hand, R3 is relatively slow-growing and could have a permanent spacing of 10 metres.

It may be practicable to double-plant within the rows in order to increase early yields, followed by tree removal. Alternatively, the inter-spaces can be used for other short term productive crops such as carambola, or perhaps preferably used for windbreak trees which will grow much faster than rambutan. Data for mature tree dimensions for all cultivars is not available.

The real question as to a 'make or break' situation for rambutan in northern Australia is yield. Whether yield and market prices can together provide a satisfactory return on capital and operating costs remains to be seen, and probably at this stage must be regarded as at least very marginal. We have not yet seen any cultivar perform in a satisfactory manner, giving yearly production which would indicate profitability.

However, few cultivars have been evaluated, and then only at an early life stage. Hopefully, a few cultivars giving satisfactory fruit quality may be adaptable to our environment and provide satisfactory yields. Also, our management practises are not yet well defined.

Malaysian yields at year 12 from planting appear to average 85 kg in comparison with 300 kg in Thailand. If under our conditions 150 to 200 kg could be achieved, then prospects would be reasonable. However, at this stage it is not even worth speculating.

Our advice is either to wait until the data is available - or to trial a small number of trees in order to assist with evaluation and gain experience in the management of this crop.

N.B. Fruit drop at nearly maturity is a common occurrence and is partially due to fruit pushing each other off the panicle when set is excessive - or due to stress or unknown factors. A spray of 1 to 4% S.N.A. active constituent applied 2 to 3 weeks prior to harvest is reputed to reduce drop - in Thailand.

B.J.Watson, September 1981

DATE: November 1981

* * * * * * * * * * * * *