PROSEA, Introduction to Fruits and nuts

Introduction

Definition and species diversity

The fruits and nuts form a large and diverse commodity group: 120 species are dealt with more or less extensively in Chapter 2, and summary data on about 275 minor species are listed in Chapter 3. Arbitrary decisions had to be made regarding the species to be included; it is not possible to define the edible fruits and nuts in a way that clearly sets them apart from species in other commodity groups. This volume deals with perennial species which are primarily grown for their edible fruit - fleshy or dry - which, as a rule, is consumed raw. Annuals (tomato, melon) and species with fruit which is not usually consumed raw (e.g. Moringa oleifera Lamk) are included under the vegetables. A number of herbaceous (e.g. Hornstedtia spp.) and woody (e.g. Ochrosia akkeringae (Teysm. & Binnend.) Miq. and Lepisanthes fruticosa (Roxb.) Leenh.) perennials are grown primarily for the seed, which is consumed raw or - more often - roasted or boiled. These species have been treated as edible fruits and nuts in so far as they are not covered by other commodity groups such as the dye and tannin producing plants (e.g. Bixa orellana L.) or the pulses (e.g. Padbruggea dasyphylla Miq.).

Thus, compared with the commonly accepted delimitation of "edible fruits and nuts", the commodity group is here narrowed by the exclusion of annuals and fruits which are hardly fit to be eaten raw. On the other hand, the group has been extended to include plants grown for their edible seeds and which do not easily fit in other commodity groups.

About 400 species bearing edible fruits, nuts or seeds are included in other commodity groups since the use of the fruit is of only secondary importance. The coconut, for instance, is primarily used for its vegetable oil and kei apple is important as a hedging plant, but the plants also produce an edible nut or fruit. These species are listed in chapter 4, indicating the commodity group in which they have been classified.

The following subdivisions of the commodity group (about 400 species) give some idea of the diversity of the plants:

• habit: herbaceous plants 10%, woody plants 90%;
• product: dry fruits, nuts or seeds 12%, fleshy fruits 88%;
• domestication: cultivated 45%, growing wild 55%.

The herbaceous group comprises important crop plants such as banana, pineapple and strawberry; the ginger family comprises quite a few minor fruits (making it hard to demarcate fruits from the commodity group "Spices and Condiments"). The woody species divide into 250 tree species, about 60 shrubs or treelets and nearly 50 vines and scramblers.

It may come as a surprise that more than half the species are not cultivated, even though "cultivation" here is used in a minimal sense. Moreover, most cultivated species appear to be just seedling populations; cultivars which propagate true to type are mentioned for only one third of the cultivated species. The latter group is agronomically of major importance, of course.

Indigenous species, including lesser known ones, have been given precedence when selecting species for textual treatment in this volume. Introduced species have been included too, not only crops such as pineapple, soursop and avocado which have long been naturalized in South East Asia, but also more recent promising introductions, such as the pejibaye palm and the acerola. Fruits from higher latitudes which have found a niche in South East Asia - e.g. loquat, apple, peach, strawberry, grape - are also described on the basis of experience in South East Asia and with reference to their role worldwide.

Hopefully, all the major and most of the minor species of edible fruits and nuts (as defined here) of importance in or for South East Asia have been covered.

Role of fruits and fruit growing

The many ways in which fruits and fruit trees are used

Fruits and nuts are consumed because they are tasty and healthy. Fruits mainly contribute vitamins and minerals to balance the diet. This contribution is comparable to that of vegetables, the main difference being that fruit is more often consumed raw, precluding loss of vitamins through boiling, etc. Some fruits, nuts and seeds are rich in protein and/or energy. Banana and breadfruit form part of the staple diet in certain areas, and the same applies to jackfruit, durian and avocado in their season.

Most fruits and nuts grow on trees. Trees are much valued for a number of reasons apart from the fruit they bear. As large perennial plants they play an important role in shaping and protecting the environment. Trees ameliorate the micro climate, casting their shade over man, beast and companion crops. They reduce erosion hazards and explore deep soil layers, recycling water and nutrients which are beyond the reach of the roots of annuals. Moreover, fruit trees produce a wide range of by products, the most obvious one being wood. The medicinal uses of various tree parts reported in this volume add up to a sizeable pharmacopoeia. The use of leaves as a vegetable, of tree bark, fruit rind or husk as a tannin or dye, of the flowers for perfume (citrus), of the fruit for oil extraction (avocado) or to tap papain (papaya) applies to far fewer species, but these subsidiary uses may assume considerable importance in certain species.

As one moves from high latitudes towards the equator, trees become more important as sources of food (Cannell, 1989). Annual staple foods dominate in areas with cold winters, but in the tropics - and particularly in the humid tropics - many people rely largely on trees for their food. Thus it is not surprising that fruit trees and the fruit they produce figure prominently in the cultural traditions of South East Asia.

Most fruit in South East Asia is produced in home gardens. Presumably these gardens are as old as settled agriculture. Through the ages South East Asians have grown up under the shade of fruit trees around the house. This long tradition of familiarity with trees and their fruits has given rise to the notions that

• the fruits hold no secrets for the consumer, and
• the trees hold no secrets for the grower.

In keeping with theses notions, South East Asians are excellent judges of fruit quality; they are familiar with the many uses of the fruit, both green and ripe, and with the beneficial effects on health according to traditional medicine. In short, they have a great appreciation of fruit. This means that they are eager and discerning buyers, providing fruit growers with a large market which generously rewards them for better quality.

This description suggests an ideal situation, and that impression is further strengthened by the lush vegetation in home gardens, and the seeming abundance of fruit in the marketplace. One might easily be led to believe that all is well with fruit growing in South East Asia, and that there is little need for research and development to improve on the achievements of a long and rich tradition. Indeed, the emphasis in fruit research on quality aspects of the fruit and on packing, storage and transport does reflect the notion that the grower knows as much about the trees as the consumer knows about the fruit; the research worker only needs to streamline the flow of fruit from grower to consumer. Development efforts are often directed towards the export markets where, again, the post harvest problems demand most research attention.

Unfortunately this idyllic picture crumbles under close scrutiny. All indications are that consumers in South East Asia cannot afford to eat as much fruit as they would like, not even as much as nutritionists recommend. Fruit is generally too expensive for large sections of the community!

Where eager consumers face scarcity rather than abundance, there must be serious constraints on the production side. The simple truth is that most fruit trees are shy bearers. Young trees take a number of years to come into bearing, and factors such as biennial bearing also depress yield levels. Much as he may know about his trees, the grower in effect has little control over yield. He has to wait patiently until his trees come into bearing - and when they do, he has to be prepared for crop failures as well as good years.

Low yield levels tend to make a commodity expensive. Growing one's own fruit in the home garden was in fact the natural solution, enabling rural societies to enjoy fruit without having to pay. The solution was all the more effective since surplus fruit could be sold. Presumably that is how the idyll of people surrounded by fruit trees and enjoying their plentiful fruit came into being. However, population growth is exiling ever more people from the "Garden of Eden"" to the burgeoning towns where for many, most fruits are prohibitively expensive. The yield levels of tropical fruit crops will be discussed in paragraph 1.3.1 Growth and development. In the following section, to form a more quantitative picture of fruit production and the availability of fruit, statistics on production and consumption are considered in relation to information on nutrition.

Data on production and consumption

It is not easy to quantify production and consumption since there are few facts and figures (Asian Productivity Organisation, 1985). Because most of the fruit comes from trees scattered in home gardens and field borders, it is next to impossible to compile reliable statistical data. Thus it is not surprising that in a survey giving mean annual yield per ha over a two year period, there is much disparity in mean yield between countries (Table 1). After all, mean yield is the quotient of estimates of production and estimates of areas, each with a large margin of error. Nevertheless, it is clear that the figures are generally low, the majority being well below 10 t/ha.

If yields are low, the next question is whether growers produce enough to satisfy the consumers. Estimates of fruit production in the region show surprisingly large differences between countries, including in fruit production per capita (Table 2). Nutritionists recommend a daily intake of 50-100 g fresh fruit, equivalent to 18-36 kg per year. Assuming an edible portion of somewhat more than 50% of fruit weight, this is about 30-60 kg fresh fruit per person per year. Taking into account that many minor fruits are not included in the table, fruit production in Indonesia may just reach the minimum level of 30 kg per person. Peninsular Malaysia and Thailand easily reach the upper level of 60 kg, and the Philippines even produces nearly twice as much per person.

Unfortunately there is a large discrepancy between production and consumption within a country, the obvious reasons being post harvest losses (for many fruits in the order of 20-40%) and export of fruit and fruit products (which plays an important role in the Philippines). An analysis of 32 nutrition and household budget surveys in the Philippines (Regalado, 1984) shows a gradual decline in annual per capita use of fresh fruit from about 50 kg in 1970 to 30 kg in 1980. Similar surveys in Thailand revealed surprisingly low figures: 14.5, 5.7 and 15.2 kg per head in 1963, 1970/71 and 1978 respectively (Apichart Pongsrihadulchai, 1985).

It is hard to explain the very large discrepancy between the figures in Table 2 and the outcome of food consumption studies, but it is clear that the production estimates cannot be taken at face value. The prevalence of deficiency diseases also points to a shortage of fruit: many people in South East Asia cannot afford to eat as much fruit as nutritionists recommend. Vitamin (A, C) and mineral (Fe) deficiencies still belong to the most widespread and most serious nutritional disorders in the region (Kyoden Yasumoto et al., 1988).

High market prices are the strongest indication that fruit is too expensive for large sections of the community. However, although studies indicate that in South East Asia, as elsewhere, more fruit is consumed by higher income groups and when prices are low (Kobayashi, 1985a), there is not enough information to relate the cost of a standard basket of fruit to the standard of living in various countries. For some fruits in several Asian countries, an attempt has been made to assess the cost price - a complicated exercise in view of the long investment period and erratic yields. Comparison with market prices reveals that the margins are small (Kobayashi, 1985b), which explains why the growers are not too keen to expand production.

Table 2 shows that production of most fruit in most countries has increased over the years, but the increments are not spectacular and on the whole, fruit production does not seem to expand much faster than population. According to Table 2, pineapple, durian, mangosteen and cashew are clearly gaining importance in the countries listed.

The evidence indicates that there is ample room for expansion of fruit production in South East Asia, provided the price level comes down. Increased production has to come primarily from increased productivity, since much higher yields are a prerequisite to balance supply and demand at prices which people can afford and which are remunerative for the fruit growers. That is why this volume stresses the study of tree growth and development: a better understanding of how the tree functions should provide the clues for a breakthrough in yield levels.

The export trade

Governments in the region have high expectations of increased exports of fruits and nuts. So far exports cover only a modest fraction of fruit production. Table 3 gives an indication of the volume of exports. Since the data do not allow analysis for individual crops, total quantities are presented. This means that high and low value crops and fresh and preserved produce are all grouped together. Hence the figures give only a rough indication. The volume of exports shows a sharp increase from a very low level for Indonesia, substantial quantities and a healthy growth for Thailand, and stagnation for the Philippines (where small setbacks in the vast banana export cannot quite be made good by diversification in exports).

With the exception of the Philippines, the exports of fruits and nuts from South East Asia are increasing rapidly - but from a very low level and with much more volatile ups and downs for individual crops than is indicated by the overall figures. The erratic fluctuations are often attributed to the difficulties of offering these perishable tropical fruits for sale in distant markets in prime condition. In fact low and erratic yields may be as great a handicap, since nothing undermines international trade more than failure to deliver the agreed quantities on time. Thus higher and more predictable yields are not only important to give fruits and nuts their due role in South East Asia, they also are the key to lasting success in the export market.

The value of exported fruit ranges from less than US$ 20 million for Indonesia (1988) to over US$ 300 million for Thailand (1987), the latter country being closely followed by the Philippines (1988) (Table 4). Thailand largely owes its leading position to the added value of preservation and processing; only 13% of the value is contributed by the export of fresh, perishable produce. Although in quantity no export product comes near the banana from the Philippines, in value the Philippines banana is surpassed by pineapple exports from Thailand (nearly US$ 220 million), because most of the latter is exported after processing, which adds greatly to the value. The strength of the Philippines lies in its long established export of some major products - banana, pineapple and mango. This is also its weakness, in view of the stagnating demand (for banana in particular). The other countries have a more diverse range of export products, although for Thailand this is obscured since several of the crops listed in Table 4 are included in "others".

More than half of Indonesia's earnings derive from cashew nuts, which is also an up and coming export product in Thailand and the Philippines. It may come as a surprise that another new export product, the papaya, already earns more than the banana in 3 of the 4 countries, despite its short shelf life. In Malaysia the durian accounts for nearly half the value of fruit exports, and durian also ranks high in Thailand. The table hides the rapidly rising star of the carambola, but in Malaysia "other fruits" stands largely for this fruit. Notice that the table also lists some fruits which might not come to mind immediately as potential export crops: jackfruit, tamarind and sugarpalm fruit.

In promoting exports it is important to remember that there are millions of customers in the domestic market who are eager to buy, provided the price is right, whereas the export market is extremely demanding with respect to timing, quality and presentation. On the other hand, there is no conflict between domestic and export markets because a well supplied, competitive domestic market forms the best possible basis for sustained export trade.

Summing up, it would be a shame if, in spite of the great appreciation of fruit in South East Asia, the intake of fruit by many people remains too low to prevent malnutrition. There is a need to expand fruit production and to raise the productivity of fruit growing to bring prices down. This will also strengthen the position in the export markets. There is an urgent need for more, reliable quantitative data to support development policies.

Botany

Growth and development

It has been argued that low productivity limits the role of tropical fruits and nuts. Growth and development determine productivity. A better understanding of how fruits and nuts function will reveal whether yield potential is much higher than the actual yield or not.

Very little is known about growth and development of most species. There is a striking discrepancy between the botanical description of every plant part in the articles in this volume, and the lack of information on the dynamics of growth and development, in particular on tree phenology. For a large majority of species the processes leading up to the harvest remain obscure: it is not clear to what extent flowering and pollination or fruit set limit yield, nor when floral initiation takes place; how flower differentiation proceeds, and whether or not many initials abort before anthesis. Even leaf longevity and the pattern of leaf change are only known for a handful of species.

Thus, although mankind in a very early stage of settled agriculture domesticated fruit trees and developed the art of grafting and budding to propagate superior trees, to this day those same fruit trees remain ill understood crop plants. The grower can only attribute bumper crops and crop failures to the whims of nature. The information in this volume is largely based on experience accumulated by generations of growers and recorded by scientists; it is not clear how much traditional knowledge remains to be brought to light. The contribution of science has been quite modest up until now, certainly in comparison with its contribution to the knowledge of field crops, for many of which growth and development can be simulated in computer models.

What can be achieved in the way of biological productivity is shown by some plantation crops and a few fruit crops of the high latitudes. A realistic target - approached by coconut (Ouvrier, 1984) and coffee (Cannell, 1971), and surpassed by oil palm (Corley, 1983), apple (Verheij, 1972) and peach (Miller & Walsh, 1988) -is that half the energy fixed annually by the tree is recovered in the fruit. It is by no means certain that such yield levels can be attained for all tropical fruits - most plantation crops and fruit crops from high latitudes have much lower harvest indices too - but it is an objective yardstick, and it would be very interesting to know how the major tropical fruit crops are using their energy.

If trees in a well laid out plantation or orchard allocate 50% of their energy to fruit growth, this corresponds to yields in the order of 50 t/ha per year for the common fruits with a high moisture content; for energy rich fruit such as durian and avocado the potential yield may only be half as much, and for nuts it is much lower still. Compared with the actual yield levels reported in this volume, these targets imply a several fold increase in productivity, even for most of the leading fruit crops in the region.

A breakthrough in fruitfulness of this magnitude obviously requires a much better understanding of the way in which fruit trees function. The emphasis in research work will have to be shifted from the fruit, the product - to the tree, the producer. To assist in this effort an attempt is made here to review crucial aspects of growth and development.

Architectural models

To start with, not all trees follow the same strategy for growth and development. Since form reflects function, attempts to distinguish between strategies are based on differences in tree form, particularly in branching habit. The coconut is a single giant shoot, the kapok tree a structure of orderly ramifying branches and shoots, the sweet orange tree a complex structure without an obvious pattern of shoot arrangement. Comparison of many tree species suggests that the infinite variety in tree form can be reduced to a limited number of architectural models. These models have been described by Hallé, Oldeman & Tomlinson (1978) and named after the persons who first drew attention to that pattern of growth and development. The models are dynamic in the sense that they describe tree habit as a result of ongoing growth and development. The models are based on ramification, flowering sites (e.g. terminal/axillary) and growth rhythm (e.g. continuous/episodic). Actual tree habit is modified by environmental pressures; a model is best expressed by a solitary tree growing under ideal conditions. Corner's model, for instance, describes trees consisting of a single giant shoot that grows continuously and produces an inflorescence in every leaf axil, e.g. papaya and coconut. The banana represents a form of Tomlinson's model; this form is characterized by a succession of equivalent shoots produced by a continuously growing underground stem, but the shoots do not grow indefinitely since instead of bearing axillary inflorescences - their growth terminates in an inflorescence.

Most models describe more complex and hierarchical branching habits such as Aubreville's model, which is most typically expressed by Terminalia species. A monopodial trunk bears whorled branch tiers; rhythmic growth is instrumental in the whorled branching of the trunk as well as in the repetitive, highly organized, plagiotropic extension of the branches in a tier. Shoots turn upwards near the compressed tip which bears axillary flowers, and laterals emerge below the tip with long internodes before turning upwards. The sapodilla conforms to this model, although in a far less obvious way, particularly in clonally propagated trees.

Branching in Scaronne's model is much more diffuse. The architecture is shown by mango, cashew and mombin; it is determined by an orthotropic rhythmically active terminal meristem which produces an indeterminate trunk bearing tiers of branches. Each branch complex is orthotropic and sympodially branched as a result of terminal flowering.

The most common model is that of Troll: 20-30% of tree species conform more or less closely to this model, including a number of fruit crops (e.g. guava, caimito, carambola, capulin). The architecture allows for great plasticity: all axes are plagiotropic and the tree is shaped by their superposition. The trunk is formed by the proximal part of successive branches becoming erect, usually after leaf fall. Clonally propagated trees often have a short trunk and several limbs contribute to tree stature as branch bases become more or less erect to extend these limbs.

In this volume architectural models, if known, are mentioned under botanical information because they help to characterize the way in which tree form evolves, including the role of flowering and fruiting.

Branching habit and fruitfulness: "single stemmed" plants

Discussing the basis of productivity of fruit crops, Verheij (1986) showed that fruitfulness is related to branching, if so called single stemmed plants such as pineapple, banana, papaya and coconut are contrasted with freely branching fruit trees. Fruiting has a well defined and secure place in the development of single stemmed plants. The growth of a pineapple or banana stem inevitably culminates in flowering and fruiting if growing conditions are not too bad; likewise, in each leaf axil of papaya, coconut or pejibaye an inflorescence is found which is normally carried to fruition. Branching plays a subordinate role; floral development precedes the emergence of suckers or side shoots, so that branching comes too late to interfere strongly with the fruit bearing stem. Plants in this group grow continuously and can do so because the leaf area soon becomes constant, emerging leaves replacing withering leaves. Nearly all plants in this group are monocotyledons, in which replacement of leaves is matched by replacement of adventitious roots, a growth habit which maintains a more or less constant top : root ratio.

The coupling of fruiting to growth puts the grower firmly in control. The faster the plant grows, the sooner it bears its fruit; moreover, fruiting benefits more from improved growing conditions than growth. Thus these species are very responsive to crop care and attain high yields. In fact the top producers among fruit crops (pineapple, banana, papaya) and oil crops (oilpalm, coconut) are found in this category. Small wonder that these crops are extensively grown within their ecological limits; they are of major importance to rich and poor, also because the fruit ripens throughout the year.

The habit of growth in which a single meristem produces all areal parts in orderly succession makes it easy to describe growth and development. Indeed, in the following articles these species are the exception to the rule that information on this topic is lacking. Even if information is incomplete, as in the case of pejibaye, it can be fitted into a simple model that draws attention to the aspects of growth and development which need to be clarified.

Ecological requirements are straightforward, as they are a direct consequence of the growth habit. In terminally flowering plants such as pineapple, stress mainly delays fruiting; the crop may still be quite good. Where the inflorescences are axillary, as in the papaya, stress sets back flowering and fruiting much more than growth. Hence plants such as the papaya are designed for undisturbed growth, both with respect to time (no cold or dry seasons) and space (no fierce interplant competition).

Since under good management yields are already high, no spectacular yield increment is to be expected. Breeding offers the best prospect for raising ceiling yield. In view of the large acreages under these crops, even small gains in yield assume great importance. Moreover, average yield lags far behind the attainable yield level. Agronomic research should concentrate on refinements in growing techniques based, for instance, on water or nutrient response curves and crop protection, to maintain high growth rates.

Branching: flexibility at the expense of flowering

There is a sharp contrast between the single stemmed plants and the great majority of fruit crops in which branching comes to the fore and precedes floral development. This is the problem group, both with respect to yield and to our understanding of growth and development. The most obvious difference with the first group is that fruiting no longer has a well defined, secure place in the development of these branched trees, shrubs and vines: whereas some shoots or twigs flower, others do not. Low fruitfulness is a common feature of most species in this group. The plant stakes out its territorial claims during a juvenile phase which may last 5-10 years (if not longer), and reaches an impressive size before the first flowers are formed. Even in the mature tree, flowering is erratic and frequently limits yield.

The yield levels for this group are generally much below those of the single stemmed plants, and they vary so much from tree to tree and year to year that it is quite difficult to set normative figures. This is the main reason why yield indications in this volume are often vague; if precise, they may be based on incidental observations that cannot be generalized. Variable yields and the associated high risk also greatly complicate cost/price analyses. On the whole it has not been possible to elaborate the economics of the crops in this volume. Since growth does not necessarily culminate in flowering/fruiting in the branched species, growth promotion does not normally improve fruiting. On the contrary, vigorous growth is usually associated with failure of flowering. Thus the normal growing techniques cultivation, irrigation, fertilizing and crop protection are safe only on trees which happen to be fruitful; on unfruitful trees they may do more harm than good.

Branching and episodic growth

Another striking difference with the single stemmed group is that in most freely branching species, flowers and fruits are produced seasonally instead of throughout the year. The underlying reason is that branching is generally associated with episodic growth: the rate of extension of the shoots is not constant, but fluctuates. In many species these fluctuations give rise to shoot growth in flushes; during a period of quiescence leaf initials are laid down in the buds, which unfold rapidly during a flush. Buds may also initiate flowers and if the quiescent period lasts long enough to complete floral differentiation, such buds break into bloom during the next flush. Most seedlings start by growing continuously and the growth rhythm changes as branching becomes more complex (Borchert, 1978), a strong indication that rhythmic growth is indeed linked with branching.

The rhythmic growth of most branched fruit crops is more or less synchronous, i.e. a large proportion of the terminal buds of the entire tree population breaks almost simultaneously. The degree of synchronization varies, both with respect to the timing and with respect to the percentage of buds that partake in a flush. It is not clear why not all buds partake, nor why not all buds produce an inflorescence. Whereas rhythmic growth is a characteristic of the shoot and is controlled endogenously (Borchert, 1978), synchronization is a characteristic of the tree population and imposed by environmental factors.

Some tropical species are evidently synchronized by quite subtle environmental cues. Striking examples are rubber and ambarella trees which even shed their leaves simultaneously in response to a change in the weather which most people hardly notice. Other species, including those introduced from high latitudes, require strong environmental triggers. If the triggers are too weak, the plants revert to an asynchronous growth rhythm, which as a rule lowers the yield level. Thus, where the single stemmed species thrive in stress free environments, the branched species generally produce better crops in seasonal climates which periodically impose a degree of stress. Sapodilla and soursop are examples of the few fruit crops for which asynchronous growth seems to be normal and which are not convincingly more productive when their rhythm is synchronized.

Fitting flowering and fruiting in with branching

Evidently in freely branching trees there is an antagonism between vegetative and reproductive processes (Browning, 1985). Adaptations have evolved to ease this antagonism, based on separation - either spatially or temporally - of extension growth and floral development (Verheij, 1986). Spatial separation finds expression in cauliflory and shoot dimorphism. Cauliflory, flowering on the trunk and main branches, sets shoot growth free; the growth rhythm of cauliflorous trees indeed ranges from continuous shoot extension (as in jackfruit) to frequent flushing (as in cocoa). Shoot dimorphism of coffee, where only the plagiotropic shoots flower, is more extreme than shoot polymorphism of apple and other rosaceous fruits, in which there are transitional forms between short shoots with a rosette of leaves (spurs) and long shoots. The spurs extend for the shortest period and are most likely to flower; the long shoots have the shortest quiescent period and are least likely to flower.

Rhythmic growth makes it easy to separate extension growth and floral development in time: flowers are differentiated during a period of quiescence in shoot growth. A large majority of freely branching tree species has adopted this strategy to lessen the antagonism between vegetative and reproductive processes. It is especially important to understand the tree phenology of these species, in order to time operations to have a maximum impact on either growth or flowering/fruiting.

Phenology of branched trees

Tree phenology - the complex annual course of flushing, quiescence, flowering, fruiting and leaf fall in a given environment - must be studied to understand how the tree functions in the course of a year and how variations from year to year affect the crop.

The effect of different environments on phenology is much more dramatic in the tropics than at high latitudes, where the winter confines growth and development largely to one season. In the tropics it helps to limit the study of phenology initially to seasonal climates which impose a synchronous growth rhythm, the more so since yields are usually best where the growth rhythm is synchronous. It is, for instance, more relevant to study the phenology of mango in a monsoon climate than in an equably humid climate where frequent, erratic flushing aggravates the imbalance between growth and fruiting.

Daylength in the tropics varies too little to noticeably affect the annual course of radiation or temperature. Thus most climatic cues are not as firmly tied to calendar months as is the case at higher latitudes, and the phenological cycle may shift a few months from one year to the next. This complicates phenological studies, but should make it easier to detect which environmental factor (e.g. a dry spell) triggers a certain growth or development response (e.g. flowering). Most of the information on minor fruits and nuts has been gathered by botanists, who record the calendar months in which their observations have been made. The authors and editors of this volume have tried to translate calender months into climatic seasons, e.g. "flowers in May-June in East Java, that is early in the dry season". Whereas "flowers in May-June" may not apply outside East Java, "flowers early in the dry season" presumably applies generally in regions with a monsoon climate.

The year-to-year variation in phenology may also distort harvest seasons as recorded by marketing agencies. For example, if durians are harvested one year in March-April and the next year in May-June, statistics on average supplies over the 2 years will show supplies over March-June, a 4 month period, although the fruit is only available for 2 months in any one year. This kind of "statistical lie" is quite common and may also have crept into some of the following articles. Year to year variations in tree phenology are caused not only by the environment, but also by carry over effects, which inevitably occur in perennials. A well known example is alternate or biennial bearing, in which an excessively heavy crop is followed by a very light crop the following year. Although fickle climatic triggers and carry over effects complicate the study of phenology, understanding their impact offers insight into growth and development.

Phenological studies may be directed to the tree as a whole or - in much greater detail - to the population of meristems whose activities determine phenology. Looking at a tree from a distance suffices to see whether it is flushing, flowering, etc. A closer look is needed to detect the first signs of bud break, and regular counts of fallen leaves may be needed to establish the pattern of leaf change. This information may add up to a lengthy description, even for trees which follow a synchronous growth rhythm. Presumably the major phenological events are linked with cambial activity and thus affect the rate of trunk thickening. This suggests that the trunk growth curve in the course of a year might help to concisely characterize whole tree phenology.

A closer look at a flushing tree shows that not all twigs produce new shoots; likewise, on a tree in full bloom a high percentage of the twigs may not be flowering. Thus it is important to study the activity of the population of meristems. It has, for instance, been suggested (Lal Singh & Abdul Aziz Khan, 1939) that in some mango cultivars the shoots on a twig which has flowered this year will not flower next year. In that situation flowering of all terminals in one year would inevitably lead to complete crop failure the next year. Thus biennial bearing at the twig level may be an important cause of erratic yields. In a study of 3 rambutan cultivars Valmayor et al. (1970) observed that in all 3 cultivars the shoots on a fruit bearing twig are capable of flowering and fruiting for the next crop. Nevertheless there was an element of biennial bearing, since non flowering twigs were more likely to flower and fruit the next year. Moreover, side shoots on flowering twigs were the main mode of branching of these trees, most non flowering twigs only making extension growth. If at harvest the grower cuts not just the fruit bunch but the entire subtending twig, he removes wood which is likely to produce relatively much vegetative growth and relatively little fruit next year, thereby sparing the twigs which are expected to produce fruit rather than side shoots. In other words, he is shifting the balance between growth and development, curtailing growth in favour of fruiting. These examples of mango and rambutan show the importance of studying phenology at the twig level if one wants to understand and manipulate growth and development. Such a study is based on labelling a representative sample of shoots and recording foliation, flowering and fruiting of the ramifications of these shoots over a period of at least 2, but preferably 3 or 4 years.

Tree phenology changes with age, not only because the juvenile plant does not flower and fruit, but also because the growth rhythm changes as the tree grows up. These changes are greatly reduced if vegetatively propagated trees are considered. If mature wood is used in propagation, juvenile features are virtually eliminated. Moreover, clonal material is genetically identical, so that only a few trees need to be observed to get a representative picture of phenology. Although this advantage may be affected by the need to study each cultivar separately, comparison of phenological differences between cultivars may throw light on the major strategies in the functioning of the species.

Towards balanced growth and development

Insight into tree phenology should help the grower to manipulate the tree and its environment so as to achieve a favourable balance between growth and development (Browning, 1985). The ultimate objective for the group of branched fruit trees is to somehow induce the trees to become excessively fruitful! In that situation the grower can safely stimulate vegetative growth as much as he can; if tree vigour still remains inadequate to support the heavy crop, balance can be achieved by thinning the inflorescences or fruits. The main instruments to tip the balance towards excessive fruitfulness are vegetative propagation and control over the growth rhythm.

Vegetative propagation (paragraph 1.4.2) and pruning and bending (paragraph 1.4.4) are discussed under Agronomy (section 1.4). The numerous techniques both traditional and modern - to control the growth rhythm are mentioned briefly under other cropping techniques (paragraph 1.4.6). Unfortunately, little practical information can be given. The problem is that to regulate the growth rhythm, the prevailing rhythm must be thoroughly understood. Apparently this is not the case, for the arsenal of time honoured methods to control the growth rhythm is hardly being used. This suggests that the grower indeed does not know his trees well enough to use these methods to advantage. Thus it is up to science to investigate how tropical fruit trees function and how they can be made more fruitful.

Cultivars

The development of cultivars of edible fruits and nuts, as in other cultivated plant species, can be said to have started when prehistoric people first brought wild plant species from the forest under cultivation near their abodes. They must have followed certain criteria (e.g. fruit taste, colour, size) to determine what kinds of fruit of a certain species they would collect and bring home for food. The seeds contained in the fruit and discarded near the dwellings, would germinate and grow into new plants. Initially, of course, people were not aware, that seedlings of most fruit bearing plant species produce variable offspring, most seedling trees bearing fruit inferior to that selected and collected in the forest.

As knowledge of agriculture improved, people discovered - initially accidentally, but later purposefully - that they could propagate some fruit species by means other than using the seeds, and that by doing so, they could perpetuate the exact plant characters of the parent trees. Thus they were able to multiply several fold the desirable mother plant, which by this time they may have identified by a name. This cultivar name was later used to identify all the plants which were produced asexually thereafter. This gave rise to the term clonal propagation. When early people used certain criteria to identify desirable plants from which they gathered fruits, they were essentially practising what modern plant breeding is doing to develop crop cultivars: selection. In South East Asia and elsewhere, most fruit and nut cultivars, including those of the major fruit species such as mango, durian and rambutan, are selections from seedling trees. Indeed, selection will continue to play a major role in improving South East Asian fruits and nuts. Exchanges of plants and fruits between South East Asian countries have played a major role in improving the fruit industry within the region. The banana industry, for example, has benefited greatly from this interchange. The cultivar Lakatan, the most popular dessert cultivar of the Philippines, is also important in Indonesia (known as "Pisang Barangan") and Malaysia (known as "Pisang Berangan"). The durian and rambutan industries in the Philippines owe their development to the introduction of superior cultivars from Thailand (durian cultivars "Chanee" and "Mon Tong") and Indonesia (rambutan cultivars "Lebakbulus" and "Simacan").

Plant material has also been introduced from other tropical regions. The pineapple and papaya industries of South East Asia, for example, started with the introduction of these species from Central America. Later, superior cultivars of papaya (e.g. "Solo") and pineapple (e.g. "Smooth Cayenne") were introduced in South East Asia and became its major cultivars for export.

The preference for superior fruit quality in South East Asia is so strong that in many instances fruitfulness may have been a less important selection criterion than fruit quality. The preferred mango and longan cultivars, for instance, are not as productive as the cultivars of lesser standing. In such cases the search for cultivars which combine good quality with high yield remains a major challenge.

Controlled hybridization has contributed very little to the improvement of fruits and nuts in South East Asia and elsewhere. One major reason is that it takes many years before hybrid seedlings start bearing fruit, and an equal length of time before such hybrid trees are finally fully evaluated. The other factors that complicate controlled hybridization in fruits and nuts are poor fruit set (e.g. rambutan), sterility (e.g. banana), and apomixis (e.g. mango, langsat, mangosteen). One of the very few commercial fruit cultivars that has been developed by controlled hybridization in South East Asia is the "Bangkok Golden Apple" guava of Thailand, a cross between the leading Thai cultivar and an Indonesian seedless type.

Investigations on varietal improvement in South East Asia have been focused on the scion cultivars. Very little work has been done on stock scion relations with the view of solving specific problems. The use of calamandarin rootstocks in propagating various Citrus species in the Philippines is standard commercial practice; this rootstock has an excellent root system, is fairly resistant to Phytophthora foot rot, and is tolerant of virus diseases. Some scion rootstock relations studies have been conducted to improve growth and fruiting of the mangosteen, to obtain dwarfed mango trees, and to avoid Phytophthora foot rot infection in durian and rambutan. To date, however, these studies have not made many concrete contributions towards solving these problems.

The concise treatment of the crops in this volume did not allow detailed information on individual cultivars. On the other hand, the available information is very limited, too. Published data usually refer only to quality attributes of the fruit; it appears that far too little is known about other important characteristics such as tree vigour, productivity, harvest season and susceptibility to diseases and pests.

Agronomy

Fruit growing in farm systems

The evolution of cropping systems for fruit

Within South East Asia fruits are harvested in an extremely wide range of cropping systems, i.e.

• collecting wild fruit
• fruit enriched fallows in shifting cultivation
• fruit grown in home gardens
• fruit grown in orchards
• fruit grown in corporate plantations.

Fruit can be found in virtually all farm systems; the list shows a progression towards more intensive and market orientated cropping systems. In Table 5 a few characteristics of the cropping systems are compared.

Of the approximately 400 species primarily used for their edible fruits or nuts in South East Asia, less than half are cultivated. This leaves more than 200 wild species, many of which are classified merely as "edible" or - even less inviting - as "non poisonous". However, a good number are eaten readily by people visiting the forest, and quite a few are gathered to be sold in local markets. It is generally believed that cultivation of some of these wild fruits, particularly from remote areas, may prove worthwhile.

The simplest and most natural form of fruit growing is found in shifting cultivation. After clearing a plot, farmers plant seeds of fruit trees to enrich the succeeding fallow vegetation. This is common practice in South East Asia. If wild species hold promise for cultivation, one would expect them to be tried first by shifting cultivators. This is what the Amazon Indians in South America do; "wild" fruits play an important role in their swidden culture (Denevan et al., 1984). Lists of species found in fallow vegetations in South East Asia, however, only record well known fruit species, often including introduced species such as soursop and avocado (Bodner & Gereau, 1988). This may be an indication that in South East Asia there is not much scope for bringing more wild fruit species into cultivation. If that is so, their main potential in fruit development is as rootstocks and in breeding programmes for cultivated relatives. From the more permanent home plot in shifting cultivation to the home garden in settled agriculture is only one step. Home gardens are the most important source of fruit in South East Asia. Comparison of home gardens with the more commercial cropping systems shows that the choice of fruit crops is closely related to the cropping system: all crops are grown in home gardens, but as the cropping system becomes more commercial, the choice rapidly narrows and only pineapple and banana are plantation crops (Fig. 1). The evolution of the cropping systems offers a plausible explanation.

In a traditional self sufficient rural society the staple food supplied by field crops was supplemented largely by a wide variety of products from the home garden, the only other source of products being natural vegetation. Households could even out surpluses and shortages of garden products in the local market. Flourishing home gardens on which nearby towns depended for their supply of horticultural produce developed into market gardens, i.e. they simply became the main source of income and supplanted the field crops. Such oversized home gardens can still be found in the vicinity of towns such as Jakarta and Bangkok. Further specialization gave rise to professional vegetable growers, floriculturists, nurserymen and fruit growers who depended on these crops for their income. They could compete with the home gardeners because of economics of scale in both production and marketing.

Whereas vegetables, flowers and nursery plants readily found their way to more commercial cropping systems, this has not been the case for most fruit crops. The slow return on investment and the erratic yield made dependence on them too risky. That is why there are very few examples of fruit crops that have made the transition to more commercial forms of production. In other words: late bearing and the high risk associated with production still confine most fruit species to the home garden. The illustration also shows what properties are required to qualify for commercial production: a quick return on investment, high and predictable yield and small plant size to facilitate management. Pineapple and banana - the only fruits grown on a plantation scale - exhibit these properties in the extreme.

After a period of unqualified praise for home gardens it is now fashionable to stress the need for improvements. There is no scarcity of home garden surveys describing species diversity and assessing the contribution to the diet or to family income, etc. (Brownrigg, 1985). However, it is not easy to achieve significant results by intervening in cropping systems whose strength lies in stability rather than peak performance. Thus it is not surprising that research on home gardens rarely goes beyond the descriptive stage. The slow development of research methods for mixed intercropping systems and the often uninspiring results of experiments with simple, well designed crop mixtures, do not augur well for research work on the complex, more or less random crop mixtures in home gardens.

However important, the gardens are not the mainstay of life, of course; nor can farmers be expected to master all the skills - and find the time - to treat each garden plant to its best advantage. Thus, although all fruit of some importance can be propagated vegetatively, most fruit trees in home gardens are raised from seed. Girdling and various root treatments to improve flowering (withholding irrigation, root pruning, laying roots bare, applying salt in a ditch around the tree) are seen so infrequently in home gardens that it is difficult to form an opinion about the value of these methods. The striking observation that known skills are not widely applied in home gardens provides food for thought. In a typical rural home gardening situation, where everybody grows fruit but nobody is a fruit grower, the traditional expertise is dispersed in the community and much of it is latent, i.e. not put to use. One of the main reasons that more commercial growers can so quickly make the home gardener look backward, is that they specialize in a few crops and integrate all this scattered information to put it to use. It is in fact difficult to take stock of local knowledge as long as it has not yet been accumulated by professional growers.

Trees, in particular fruit trees, dominate home gardens almost everywhere, for the very reason that they have been less successful in migrating to more commercial cropping systems than herbaceous garden crops. Where many of these tree crops are confined to home gardens because of low productivity, it appears more appropriate to say that the trees need to be improved rather than the gardens! This is a much more tangible problem and, in spite of the shortage of researchers on tree crops, it stands a much better chance of being solved. The conclusion that the majority of the fruit crops in South East Asia are confined to home gardens and come into bearing too late and too erratically for cultivation in more commercial cropping systems has serious implications. It means that these fruits play only a marginal role: they either do not get as far as the major markets or - if they do - they are too expensive for most people.

Towards more commercial cropping systems

It is clear that if the "confined" fruit crops are to be popularized, their productivity has to be improved. The key to a breakthrough in productivity in tree crops has always been vegetative propagation of superior trees. Vegetative propagation is also a prerequisite for production in orchards, as shown in Table 1; the long juvenile phase and the fact that the product lacks the identity of a named cultivar in the market virtually rule out planting orchards with seedling trees. In this volume the exceptions are limited to palms (salak, pejibaye), papaya and cashew. Where orchards are found they nearly always consist of precocious trees of small sized and relatively fruitful species: papaya, tangerine, guava, rambutan, cashew.

In Thailand the use of clonal material is standard practice and it is no accident that Thailand is also the only country in South East Asia where fruit for the market comes mainly from orchards rather than home gardens! In Thailand mango, durian, mangosteen, langsat, tamarind and other species which elsewhere in South East Asia are almost exclusively found in home gardens or on field borders, have found their way to orchards. This is a development of the last few decades. Presumably it started in the vicinity of Bangkok, where the high water table in the central delta limits rooting depth, providing an important additional instrument to control tree size. The accelerating pace of developments in Thailand, also outside the central delta, suggests that general adoption of clonal propagation may be the first step towards a complete transformation of fruit growing.

Home gardening need not suffer from the shift of fruit growing for the market to orchards. On the contrary, home gardeners benefit from the clonal material available in nurseries. This means, for instance, that instead of one large seedling durian tree, several smaller durian cultivars can be accommodated in the garden.

Nurseries fulfil a key role in the development of fruit growing. They can also be valuable indicators of development trends in two respects. Firstly, increasing sales of trees show that fruit growing is expanding and the assortment indicates which crops and cultivars are in the lead. Where national statistics are hard to interpret, particularly in the case of fruits with their wildly fluctuating yields, nursery surveys can give a much more direct and predictive assessment of development. Secondly, the level of nursery work, the specialization in a limited number of fruit crops and the scale of operations with respect to the leading cultivars, are a very good indication of the role of commercial production in orchards as compared with home production. Fruit growers require large quantities of uniform plant material at competitive prices to establish orchards; once nurserymen begin to respond to that demand, the character of the nurseries changes markedly.

It seems unlikely that many tropical fruits will join pineapple and banana as plantation crops because - in addition to the required properties of the trees - the fruit should lend itself to trade, i.e. it should have a long supply season and a long shelf life. How critical these requirements are is clear from the example of papaya, a crop that is as fruitful as the banana and pineapple, with a strong demand and a long supply season, but whose fruit is too perishable to grow the crop on a large scale. The cashew, on the other hand, is now being tested on plantation scale in Indonesia despite its low productivity. Its main assets are a short juvenile phase, a fairly predictable yield and the fact that the nuts can be traded throughout the year.

Summing up, it is to be expected that orchards will soon take over from home gardens as the largest supplier of fruit in South East Asia, provided that professional nurserymen can supply superior clonal material in large enough numbers at a reasonable cost, as in Thailand. Higher yields and more efficient production should bring prices down so that the common man can afford a better choice of fruits. Households with a home garden will be able to eat more produce but will sell less, which is not altogether a bad development. After all, home gardens are better suited to feed the family than to cater to the millions. These small units with their intricate mixture of crop plants are not really compatible with production for large markets.

Few fruit crops lend themselves to production in corporate plantations. As more fruit is more efficiently produced in orchards, it may in fact become increasingly difficult for corporations to compete.

Vegetative propagation

In South East Asia many fruit trees are still propagated from seed, even though virtually all fruit crops can be propagated vegetatively. Experience shows that vegetative propagation is the crucial step towards improving productivity of fruit crops. Two factors are involved:

• the most fruitful trees can be selected for propagation;
• the juvenile phase is eliminated or greatly reduced.

Genetic variation in fruit tree populations is generally great, but for a few fruit crops in this volume it is suggested that populations are very uniform (e.g. soursop), so that there is limited scope for selection.

Elimination of the juvenile phase means that it is no longer necessary to raise a sizeable tree before the first fruit can be harvested: clonal trees may flower in the nursery already! If mature parts of the mother tree are cloned, juvenility is completely eliminated, but a seedling rootstock introduces an element of juvenility into the stock scion combination. Lychees in Thailand and sapodilla in the Philippines are usually air layered because such trees come into production one year or several years earlier than trees budded on seedling rootstocks. These examples show how important early bearing is to the common grower. Early bearing not only shortens the investment period but also siphons energy into fruit which otherwise would have gone into growth. The reduction of tree growth through early fruiting is a great bonus: it means that more trees can be planted per ha, further increasing early crops. Moreover, the reduction of tree size greatly facilitates orchard management, from pruning and crop protection to harvesting. Thus, vegetative propagation sets a number of processes in motion leading to intensification: fewer unproductive years, more trees per hectare, higher maximum yield per ha, much higher mean yield over the orchard's lifespan. Given the small farm size in South East Asia, such a development is extremely beneficial.

A final advantage of clonal propagation which is often overlooked is that it gives the product a name in the market. Clonal material has a definite identity; in a market where quality is extremely variable and hard to judge, it is a great leap forward if a "Mon Tong"" durian can be offered rather than just a durian. Table 6 shows the relationship between the major vegetative propagation methods; typical examples are given for each method. It is generally assumed that most of these methods originated in Asia, presumably already several thousand years ago. Present day refinements are largely a consequence of the many applications of plastic materials; the use of mist installations has made the propagation of leafy material easier. The major modern contribution to clonal propagation is the development of tissue culture techniques. The impact on fruit tree propagation in South East Asia is still very limited, but this technique will no doubt find wider application.

Table 6 is split into two sections, the first presenting methods in which plants are propagated on their own roots; the second section presents methods in which a rootstock provides the root system. Starting with the natural propagation methods, the methodology becomes generally more complex in the lower echelons of the table; it takes far less time to set 100 stem cuttings than to prepare 100 air layers; preparing 100 approach grafts requires even more time and skill. On the whole budding and grafting demand more skill than inducing plant parts to form adventitious roots and - if the time required to raise the rootstocks is taken into account - a longer nursery period is needed. Therefore, budding and grafting are only used for plants which do not form adventitious roots or where the rootstock offers important advantages such as restricting tree size (apple), salt tolerance (avocado), better fruit quality (citrus) or tolerance to diseases (citrus).

There is a confusing array of names for the artificial (non natural) propagation methods. However, if the terms used to describe the various grafting methods for established trees are excluded as in the diagram (which only refers to the raising of young plants), the terminology is fairly clear. Some general terms are explained here and more specific terms used in the articles are defined in the glossary.

Shoots or branches used for cuttings or as budwood or graftwood are divided into softwood (young, herbaceous stems, the primary tissues - epidermis and cortex still intact), semi hardwood (partly lignified stems) and hardwood (mature, woody stems, secondary thickening having obliterated the primary tissues). Definitions of approach grafting and inarching vary; sometimes inarching is considered as a form of approach grafting. Inarching commonly refers to a decapitated rootstock being inserted into an incision made in the scion plant. In approach grafting both partners are intact plants, their stems spliced together to effect the graft union. One of the best known forms of inarching is intended to rescue a poorly anchored tree: one or several rootstocks are planted nearby, decapitated and inserted into the trunk. This method is employed for established longan trees in Thailand and occasionally for other trees when the roots are damaged by rodents or soil borne diseases.

A form of inarching used to produce young plants in substantial numbers is called suckle grafting after the Indonesian name "susuan". Surprisingly, this method appears to be hardly known outside South East Asia; it is not mentioned in the leading textbooks on tree propagation. Suckle grafting seems to be practised throughout South East Asia; in Thailand it is used commercially to propagate large numbers of mango, durian, langsat and jackfruit trees. Rooted stocks are potted in polythene bags and taken to the mother tree which is to be propagated. There each rootstock is decapitated, the bag tied to a sturdy twig and the tip of the decapitated rootstock is inserted into a cleft made in the twig. The graft union is wrapped in polythene tape. Because the soil ball is completely enclosed in the bag, the rootstock needs no watering; in fact it gets no attention until the successful graft is severed from the mother tree and transferred to the nursery to be raised to the required size. Suckle grafting is only a shade less subtle than approach grafting (in which the rootstock top continues to function while union is being achieved), but suckle grafting is not nearly as cumbersome and labour intensive. Hence it deserves much wider recognition as a valuable propagation method for plants which are hard to propagate by simple forms of budding or grafting.

Most fruit and nut species can be propagated in a variety of ways. Broadly speaking, simple methods require more attention to external conditions (e.g. shade, high humidity), whereas the sophisticated methods demand more time and skill in execution. As a consequence, the simple methods are more suited to mass propagation since they require little labour per plant and the cost of ameliorating the environment is shared by a large number of plants.

The small plant size of nursery stock and the short duration of the trials make propagation a popular topic in fruit tree research work. Large numbers of plants permit the use of experimental designs with a high power of tests. Unfortunately even extremely significant results are often not reproducible. Ill defined factors such as the condition of both the stock and scion material, the skillfulness of the grafter, and the more or less controlled nursery conditions all add up to making the results less applicable elsewhere. An example of practical consequences is that Garner & Chaudhri (1976) reviewed nearly 400 references of reports on propagation experiments with mango without arriving at clearcut recommendations on what method(s) of propagation to use! The time may have come to reduce the number of propagation experiments and to sharpen the hypotheses to be tested. Better definition of materials and methods and inclusion of check treatments (e.g. using the rootstock itself as one of the scions along with the cultivars to be compared) should also make the results more reproducible.

Orchard layout

What the factory is to the industrialist the orchard is to the fruit grower; it is his means of production. The lay out is of major importance for orchard efficiency, both in the biological sense - the conversion of radiant energy into quality fruit - and in the managerial sense: the use of man and machines. Tree arrangement in the field is determined by the size, the shape and the orientation of the area needed per tree. Both the size and shape of the area are defined by tree spacing, e.g. 7 m × 5 m: area 35 m2/tree, shape a rectangle. Orientation is only of slight importance in the case of row cropping. At high latitudes rows running north south are preferred for fruit crops. The reason is that the sun moves low through the sky; if an east west orientation was used, there would be a sunny and a shady flank to each row which would lead to uneven growth and fruiting, and particularly to later ripening and poor fruit quality on the unexposed flank. In the tropics the sun is overhead most of the day and rows may be aligned to run across the direction of the prevailing wind or on sloping land along the contour.

Tropical tree crops are almost always planted in equidistant planting patterns (see Fig. 2):

• the hexagonal (or triangular) pattern
• the square pattern.

The quincunx, another equidistant pattern, is a modification of the square pattern obtained by moving every second row up half a space. It is a common planting pattern for vegetables but not for trees. The square pattern is much more popular than the hexagonal pattern as it is easy to lay out and is accessible to man and machines from two directions. The hexagonal pattern gives the most even distribution of the trees over an area, each plant being surrounded by 6 neighbours, all at the same distance, as compared with 4 neighbours in the square pattern. Because of the ideal tree distribution over the area, at equal distances from tree to tree 15.5% more trees can be planted in a hexagonal pattern than in a square. However, the hexagonal pattern is difficult to lay out and row width is less than tree spacing, making the orchard less accessible. Its use is limited; tea and oil palm are sometimes planted to this pattern.

The popularity of the equidistant planting patterns is largely based on the notion that all trees will grow equally well and will ultimately reach the size that was foreseen at planting. In fact irregular and unpredictable tree growth makes a mockery of the geometrical perfection of the equidistant planting patterns. The tree to tree variation and especially the impossibility to predict at planting the growth rate and mature size of the trees makes tree spacing guesswork and favours flexible tree arrangements. Where vegetative parts are harvested, as in tea, one can plant so close that full canopy cover is attained even if growth is weaker than anticipated. Fruiting trees, on the other hand, do not tolerate fierce inter tree competition and it is safer to err on the side of wide spacings. A striking example is the coconut: because overcrowding is deleterious, the palms are often spaced so far apart that intercropping becomes necessary because the area per tree increases quadratrically with increasing equidistant spacing.

Rectangular arrangements are much more flexible since they are based on two parameters and they cope much better with the uncertainties of tree growth than the single parameter equidistant patterns. The open fringe around each tree in a 7 m × 7 m spacing can be concentrated into lanes between tree rows in a 8 m × 6 m or 9 m × 5.5 m rectangular planting, giving nearly the same number of trees per hectare. If growth is disappointing, the trees may still fill the rows and the poor area utilization by the trees can be supplemented by prolonged intercropping in the wide lanes. If tree vigour is excessive, the wide interrow spacing offers some outlet. Moreover, tree variability is to some extent absorbed in the rows, since a more vigorous tree can compensate for a weaker neighbour. For these reasons rectangular planting patterns - in other words, row cropping - deserve much wider application in tropical fruit growing. Tree arrangement assumes greater importance when trees are planted closer together, and it is no accident that row cropping comes to the fore wherever fruit growing becomes more intensive.

Another tree arrangement that deserves more attention is planting in double rows. This is commonly done for small plants such as pineapple and strawberry. However, it is suitable whenever plants become small in relation to the required lane width. If, for instance, the lanes should be wide enough for a tractor to pass and the cropping volume per unit row is small (e.g. papaya, lime, apple) planting double rows makes it possible to widen the lanes without sacrificing cropping volume. In Figure 3 a comparison is made of single and double row arrangements, both at 2000 trees per ha with the same tree size. The single rows become closed hedges, but the trees in the double rows are only touching, for - in spite of the more spacious lanes - the minimum distance between trees is larger than in the single rows. The reduction of row length per ha implies that the tractor has to run only 2000 m instead of 3333 m to treat a hectare with a fungicide; the same reduction applies to the length of irrigation ditches or pipes! Moreover, the water supply can be installed within the double row, where it does not interfere with traffic in the lanes.

The main limitation of the use of double rows is that not too much cropping volume should be heaped in a double row, since flowering and fruiting suffer more from inter tree competition than does growth. Consequently, double rows are only suited for very small trees and these should be arranged so that an open canopy structure is maintained. Examples of double row planting patterns are given in Fig. 4. It is clear that the equidistant arrangements on the left hand side (A C) result in rather compact double rows. On the right hand side (E), the open structure has been overdone: the double row begins to look like a badly aligned single row! The double row next to it (D) is meant to strike a balance, facilitating light and spray penetration and access for pruning, harvesting, etc.

Finally, a remark should be made on the correct notation of spacings. It is customary to let the inter row spacing precede the intra row spacing. The inter row spacing is the larger of the two values, but for trellised plants this is not always the case: 3 m × 4 m indicates an interrow space of 3 m; the vines are trained in the direction of the row to fill the space of 4 m. Note also the arithmetically correct presentation of double row spacings in Fig. 3. By first adding the large and the small inter row spacings and then multiplying by the intra row spacing, the reader can reconstruct the layout. Moreover, the result of the calculation is the correct area per tree pair: (3.5 m + 1.5 m) 2 m = 10 m2 per 2 plants or 2000 plants per ha.

Training: pruning and bending

Outline

Pruning can be defined as the removal of unwanted growth in order to stimulate desired growth. This definition contains two clauses: something is removed (1), to elicit a certain response (2). It follows that pruning cannot be measured solely by the amount of prunings lying under the tree, or by the looks of the tree immediately after pruning; the quality of pruning can only be judged after the plant has had the time to respond. Deshooting of tomatoes to improve fruit set is covered by this definition and so is the trimming of hedges to increase shoot density. Root pruning, e.g. in tree nurseries to induce a more extensively branched root system in a confined soil volume, is also included. However, the following discussion centres on the pruning of aerial parts of woody plants: because of the perennial habit, the effects of pruning accumulate over the years.

A shoot or branch can be placed in a desired position by bending (see Fig. 5). This is an attractive alternative to pruning, particularly for young plants which still have to fill the allotted space. To shape an open centre tree, near vertical limbs can be bent down and lax limbs tied up to achieve a more balanced shape. In this way growth can be retained which otherwise would have to be removed by pruning (with the risk that the new shoots would still not be disposed as desired). The trend to replace pruning by bending has led to a much more rapid increase in crop volume, for instance in tea (pegging down) and apple.

Creeping, winding and climbing plants have to be trained on a support structure to obtain the desired canopy architecture. In the tropics live stakes such as Lannea coromandelica (Houtt.) Merr., Moringa oleifera Lamk and Leucaena spp. and - in dry areas - Commiphora spp. are often used in combination with bamboo crossbars and wires. Training is a combination of pruning and bending.

Pruning and bending as a rational system of man's action and the tree's reaction has evolved in the deciduous fruits at high latitudes (grape, apple, pear, peach). The role of pruning in tropical tree crops is comparatively small and the recommendations in the articles in this volume are generally rather superficial. In fact, by far the most experiments in tropical fruit trees (e.g. in citrus) show yield reductions following pruning, which are not compensated for by clearcut advantages. This suggests that pruning may be counterproductive. On the other hand, pruning is essential in coffee, and grapes in the tropics require pruning just as much as at high latitudes. Moreover, in crops such as tea and rambutan, harvesting itself is a form of pruning. Hence the scope for this typical horticultural growing technique is briefly considered here.

Factors affecting the response to pruning

The response to pruning largely depends on:

• which portion of a shoot or branch is cut (1);
• the timing in relation to the tree's growth rhythm (2);
• the fruitfulness of the tree (3).

(1) According to which portion of a shoot or branch is cut, the severity of pruning can be classified as follows (Fig. 6):

• tipping or pinching: removal of the shoot tip;
• cutting back: removal of a substantial part of the shoot/branch;
• stubbing: cutting a shoot/branch near its base, leaving a stub;
• cutting out or thinning: removing the entire shoot/branch by cutting at the point of attachment.

Tipping - also called pinching if done without the use of a tool - sounds like the most delicate of pruning methods, but it elicits the most dramatic response. Many axillary buds sprout and grow into fairly weak shoots. The result is a sharp increase in branching as shown in the reaction of tea to harvesting and that of hedges to trimming. An important side effect, also obvious in the examples given, is that flowering is suppressed. The violent response to tipping is explained by the elimination of correlative inhibition; the apical dominance of the shoot tip no longer inhibits the axillary buds.

When more is removed than the shoot tip only, the treatment is called cutting back. As a larger part of the shoot is removed, the response changes: fewer lateral shoots grow out, they arise from the nodes just below the cut and they are more vigorous, especially the uppermost ones. These upper shoots grow at a small angle with the branch; further down, the angle of emergence gets wider as the shoots get weaker. Cutting back is the principal treatment in the shaping of young trees. Often trees are cut back at planting, or even in the nursery, to obtain a well branched tree frame (Fig. 7). Later, cutting back is practised to improve the balance between these branches, which will form the tree's main limbs. The most vigorous branch is cut back farthest, allowing the weaker branches to catch up. Hence the weakest limb determines the rate of expansion of the tree frame! Cutting back may have to be followed by another pruning treatment to thin out the new shoots, leaving only one to extend the limb plus a few small laterals. Cutting back reduces tree size but induces compensating regrowth at the expense of flowering and fruiting, which may be set back several years.

Stubbing is a drastic form of cutting back, leaving only the basal part of the shoot or limb. The principal difference with cutting back is that on the basal part of a shoot only underdeveloped, dormant buds are found, whereas on stubs of older wood no buds may be present at all and adventitious buds have to form before new shoots can sprout. This means that regrowth following stubbing is slower than following cutting back. More important, there is no clear hierarchy among the sprouting buds, so that the general response to stubbing is the growth of several to many shoots of near equal vigour. Stubbing is practised in ornamental shrubs (Hibiscus spp., Euphorbia pulcherrima Willd. ex Klotzsch) where the regrowth of a number of equivalent shoots is exactly what is needed. Occasionally trees are drastically rejuvenated through stubbing, e.g. coffee, citrus, peach. This form of stubbing is also called stumping; another pruning treatment (thinning out) is needed, to leave only the required number of shoots for the new framework.

Cutting out or thinning is the complete removal of shoots or branches. This is the most drastic cut that can be made, yet the plant's response is rather mild. Often there is no regrowth at all near the cut, implying that the response must be dispersed over the remaining tree structure. Because of its obvious direct effect and the moderate reaction of the tree, thinning is the principal pruning method. The pruner can see immediately what he has done and need not worry too much about the after effects.

In the tropics pruning is too often equated with cutting back and cutting out is not given its due role. If the canopy is too dense, as evidenced by inferior fruit quality, poor flowering and fruit set, or early leaf fall in the tree interior, cutting out is the way to relieve the overcrowding (if the stand is too dense, uprooting part of the trees is better than trying to contain each tree). Thinning is also practised in the case of excessive flowering (coffee) or excessive fruit set (citrus). In these cases, thinning of twigs is the simplest way to prevent exhaustive flowering or fruiting. Removal of ageing, sagging branches to the point where a younger, more vigorous shoot emerges (usually on the dorsal side of the sagging branch) is an essential form of thinning in pome and stone fruits. It is a way to gradually rejuvenate the fruiting wood to prevent a decline in fruit quality. Thinning can be stretched to include desuckering of crops such as banana.

In some of the following articles thinning is also suggested as a possible means of restricting tree size and complexity. This applies particularly in the case of biennial bearing at the twig level. If the side shoots on twigs which have flowered this year are unlikely to flower next year, flowering twigs which have not set fruit may be cut out. This helps to contain tree size and simplifies tree structure; moreover it eliminates wood which is unlikely to flower next year in favour of wood which, hopefully, is more disposed to flower. For similar reasons it may be advantageous when harvesting trees such as rambutan, litchi or longan, not to cut only the fruit panicle (cutting back), but to cut out the entire fruiting twig.

(2) Methodical pruning as evolved in fruit growing at high latitudes is largely winter pruning. Summer pruning plays a modest role and is mainly done after the termination of extension growth. The advantages of pruning leafless trees with dormant buds are:

• in the absence of leaves the competitive relations in the tree can be better judged;
• no leaves are removed, so the amputation does not directly reduce photosynthesis;
• the quiescent tree has the time to attune vascular connections and hormone flows to the new situation; perhaps this is why the repones of the tree is more predictable than when trees are pruned during a flush.

The above reasoning largely explains why pruning plays only a modest role in evergreen tropical crops. It also makes clear that pruning during a flush should generally be avoided. Pruning during active growth has very limited applications in fruit growing, e.g. removing shoots which threaten the dominance of the leader in trees with a central leader, eliminating sylleptic shoots in grape to improve budbreak of the proleptic shoots.

(3) Fruiting moderates the response to pruning. A top heavy tree needs corrective pruning in the upper portion. If the tree is unfruitful, vigorous regrowth ensues so that next time the tree again has to be pruned hard; pruning engenders more pruning. In a fruitful tree pruning results in moderate regrowth, better distributed over the entire tree. Low yield levels of tropical fruit crops are the main reason for the modest role of pruning in the tropics!

Pruning, a balancing act

The general purpose of pruning and bending is to improve the balance within the plant, as for instance:

• between top and root: e.g. trimming the shoots before transplanting to match the reduced root system.
• between old and young growth: e.g. replacement pruning which is the cutting out of ageing wood to rejuvenate the branch.
• between branches;
• between growth and fruiting.

The last two forms of balance are most important and are considered here. Training vines is basically a matter of balancing shoots. The grape is a good example; the horizontal arms, the evenly spaced horizontally trained shoots and more severe cutting back of stronger shoots all aim at a perfectly uniform shoot population. The passionfruit, with its curtain of shoots hanging down from the wire and trimmed near the ground, shows another application of the same principle. Trees are not so pliable; the pruner must strive to reconcile balance in the branching structure with dominance of the shoots expressing tree architecture. If the dominant shoots are too strong, vegetative growth gets the upper hand. If dominance is lost or if several growing points compete for dominance (Fig. 8 left), growth coordination is weakened with undesirable consequences for the distribution of growth over the tree. Drastic corrective pruning may be required to eliminate the competitors (Fig. 8 right).

Many tropical fruit trees grow rhythmically and flower on the shoots. Floral differentiation takes place on the quiescent shoots, and opportunities for floral development are best if there are many, sturdy shoots, which flush simultaneously over a limited period. If apical dominance is strong, the vigorous shoots enter quiescence too late for floral differentiation and the suppressed shoots may be too weak to differentiate viable inflorescences. Thus a favourable balance between branches consists of a limited number of dominant shoots and a large population of uniform shoots of medium vigour.

The apple in the tropics presents a striking example. Tipping, bending and defoliation set back the dominant terminals and causes many axillary buds to break. The horizontal branch puts them all in the same position and because of their large number, the roots cannot sustain their extension growth for very long. Thus there is time to form flower buds which break into a prolific bloom when the trees are defoliated again. This approach has made the ill adapted apple one of the more prolific fruits in the tropics! A less obvious example is the pruning of the very vigorous shoots in young sugarapple or plum trees which continue to grow for a long time. These so called whips benefit from tipping or moderate cutting back, as this dissipates their vigour over a number of laterals which - because of their more restricted growth - are more inclined to initiate flowers.

The balance between growth and fruiting is of crucial importance for fruit crops. Growth and fruiting compete, as shown for instance by the effect of deshooting of tomato: pinching out the young side shoots improves fruit set and fruit growth. Likewise, fruit set in grape and apple can be improved by pinching out the shoot tips. In both cases a sink which competes with the fruitlets is removed. However, whereas the pinching out of shoots may improve fruit set, the subsequent leafing out of laterals greatly reduces the practicality of this method.

These examples are exceptional in that pruning generally sets back fruiting more than growth. Repeated pinching out (as in harvesting tea and trimming of hedges), for instance, can suppress flowering completely. Formative pruning of young trees postpones the onset of fruiting, and even replacement pruning - the removal of ageing wood which has borne several crops - reduces the next crop in favour of growth.

Fruiting benefits from pruning mainly because fruit size and quality are improved and because pruning maintains tree vitality to the benefit of future crops. However, these favourable effects can only be expected in prolifically bearing trees; in the case of light crops, fruit size and tree vigour should not pose problems. Thus this train of thought leads to the same conclusion as was reached before: the low yields of many tropical fruit crops limit the scope for pruning. This also means that a breakthrough in fruitfulness may greatly enhance the role of pruning in tropical fruit crops.

Crop protection

Crop protection has evolved along with the crops and the cropping systems. The use of commercial insecticides, fungicides, etc. is virtually limited to commercial cropping systems, i.e. orchards and corporate plantations. In home gardens many traditional crop protection measures are occasionally practised, but these are largely limited to small plants and nursery work (e.g. use of wood ash and plant extracts); protection of tall trees is limited to trunk collars (which stop rodents from climbing the tree) and the like. The home gardener relies largely upon the garden being such a diverse biotope that it is not easy for diseases and pests to get the upper hand. The buffering capacity of the biotope does not provide adequate protection, but the home gardener accepts substantial losses as a fact of life.

The use of commercial biocides is limited by tree size too: spraying of tall trees is impossible with the equipment a small grower can afford. The scattered planting in backyards, on field borders and along water courses impedes access, and orchards usually comprise a mixture of fruit trees. Moreover, the tree population may consist of different cultivars (or even seedlings) which should ideally be treated separately, and factors such as biennial bearing make it difficult to establish an annual spraying routine (e.g. stink bugs can ruin longan blossom, leading to crop failure, but in a year with prolific bloom they may be welcome thinning agents). For these and other reasons, contractors and cooperatives play a very modest role in crop protection.

In peninsular Thailand many rambutan orchards receive a treatment against powdery mildew, and in the northern Philippines spraying of mango trees with potassium nitrate to induce bloom implies that biocides (e.g. against leafhoppers or anthracnose) can be sprayed too. However, these examples are exceptional; as a rule tall fruit trees are not sprayed.

The situation is quite different for small fruit trees such as tangerines, guavas, apples and grapes. Small tree size is linked with more intensive husbandry; the high inputs and improved control over yield make the grower keen to minimize the risk of substantial crop losses from pests and diseases too. Thus there is a tendency to try to protect the crop with whatever commercial biocide is recommended. Orchards of this kind are often sprayed frequently and routinely, without regard for the level of infestation. In this way the natural checks and balances in the orchard are eliminated, there are health risks to growers when they apply the materials and to consumers when they eat the fruit, and the soil and water become polluted (as argued in the article on Citrus). The western world, which was the first to rely heavily on commercial biocides in fruit growing, is mending its ways and adopting integrated crop protection, greatly reducing the use of these products. In South East Asia there is also a rapidly growing awareness of the dangers of toxic substances, but the use of biocides in commercial fruit growing in the region is still increasing.

To turn the tide the first step is to stop routine spraying and to apply treatments only on the basis of observed infestation. To do so the grower must know what to look for, which means that he must be familiar with the life cycles of disease or pest organisms. Depending on the host parasite relationship, the life cycle of the parasite will be linked to the growth rhythm of the tree. Hence, improving efficiency in crop protection is another - and important reason for clarifying the growth rhythm.

Other cropping techniques

A fairly clear distinction can be made between the husbandry employed for crops grown in orchards and plantations and that for fruits and nuts which are largely confined to home gardens. Much is known about water and nutrient requirements and other cropping techniques for orchard and plantation crops, but information is rather incidental for home garden crops. Since these latter crops are a large majority, the range of existing cropping techniques is briefly reviewed in an attempt to draw up a possible framework for the husbandry of these crops. Because of the generally poor fruitfulness of home garden crops there is limited scope for growth promoting husbandry. In fact most cropping techniques aim at checking growth by imposing stress! The intended result is to improve flowering. The best known example of this approach is interruption of irrigation during the dry season in order to induce a flush with abundant bloom after irrigation is resumed. Many crops respond favourably to this treatment, e.g. mango, cashew, rambutan, durian, mangosteen, longan, citrus and grape.

Where growing conditions rule out a controlled dry period, the root system may be tackled directly, for instance by root pruning, removal of the topsoil to lay part of the roots bare for a few weeks, or the application of salts in a furrow under the drip line of the tree to increase the osmotic pressure of the soil solution. The main limitation of these methods is their crudeness: it is difficult to apply the required stress for the required length of time. With the increasing interest in mulching for the amelioration of soil conditions, scraping the mulch away for a short period may offer a fairly simple alternative to the traditional root treatments. Removal of the mulch in spring is practised in fruit growing at high latitudes; it enables the soil to warm up more during the day and to release more heat during the night to diminish the risk of damage by late frosts.

Another method to achieve more or less the same end is girdling or cincturing. This technique is practised worldwide on vigorous, poorly flowering trees and may be the oldest method to improve the fruitfulness of tree crops. A narrow ring of bark is removed from the trunk or some or all of the main limbs to stem the flow of carbohydrates from the top to the roots. The main application is on quiescent trees during the period that floral differentiation is supposed to take place. Where flowering is associated with a flush of shoot growth, the trees may also be girdled when in bloom, in order to improve fruit set and fruit retention by checking the competitive sink strength of the new shoots. Girdling is not without risk: if the ring is too wide or cut too deeply to be overgrown in 1 2 months, the treatment may kill the tree. In spite of this risk girdling is a fairly common cropping technique, a strong indication that growers have faith in its usefulness. There are refinements to the technique such as:

• metal bands which can be tightened and loosened in relation to the growth rhythm of the trees;
• instruments which girdle accurately to a pre set depth;
• "chemical girdling" by painting a ring of morphactin, a growth regulating substance, on the trunk or limb.

These methods have led to renewed scientific interest in the scope of girdling. All these cropping techniques can be said to accentuate or prolong quiescence of the shoots by imposing stress on the root system. This should give floral development time to run its course. The methods also suppress erratic shoot growth and therefore have a synchronizing effect on the growth rhythm.

The use of growth retardants is a modern technique to impose quiescence. This has produced few practical applications in tropical fruit growing so far. Results with paclobutrazol, however, are very promising, partly because of its powerful effect on flowering. Paclobutrazol might therefore become an important tool in balancing vegetative and reproductive processes.

Growth regulators have an ancient forerunner in smudging, which is used in the Philippines to induce flowering in mango. It has been shown that ethylene is the active substance in the smoke of the trash fires which are kept burning for weeks until the twigs clearly bear flower buds. Smudging of mango trees has been superseded by spraying with potassium nitrate which is both simpler and more effective, but not everywhere and not on all cultivars. The mode of action of potassium nitrate, which is also used as a dormancy breaking agent for temperate zone fruits grown in the tropics, remains inadequately explained.

Where imposition of root stress, girdling or application of growth retardants synchronize the growth rhythm of the trees and improve flowering, cropping techniques which improve pollination and fruit set come to the fore. The common, well known methods are the planting of pollinator trees to facilitate cross pollination, bee keeping, and hand pollination. It may be equally important to pay attention to the quality of the flowers. The general condition of the trees appears to strongly affect factors such as the percentage of female flowers in polygamous flowering plants and the length of the fertile period of the pistils. Unfortunately, the appropriate way to improve tree condition is growth promoting husbandry and, as this tends to be at the expense of flowering itself, attention to flower quality is largely restricted to trees which are already fairly fruitful.

Where measures to promote flowering and to manipulate floral biology result in good fruit set, the fruit load plays an important role in stabilizing the growth rhythm. The growing fruits absorb so much energy that shoot growth is controlled. The growth of the flush which brought on flowering soon terminates and there is little chance of further flushing of shoots as long as the fruits are on the tree. This sets the stage for a post harvest flush, which should make up for the loss of leaves incurred during the preceding months.

Thus a sequence of cause and effect emerges, in which application of stress should stimulate floral development, a more synchronous growth rhythm being an important side effect. Better flowering makes it worthwhile to pay attention to floral biology; where this leads to good fruit set, the synchronization of the growth rhythm is maintained throughout the crop cycle.

This opens up new prospects. The well defined succession of phenological stages in a synchronous growth rhythm makes it possible to time all husbandry interventions in accordance with these phenological stages. In this situation even growth promoting techniques are more safe; water and nitrogen applications, for instance, can be timed to have a maximum effect on the impending flush, minimizing the disturbing effects on floral development or fruit ripening.

The above reasoning leads to the following outline for the approach to the husbandry of unfruitful trees:

• vegetative propagation to eliminate the juvenile phase;
• growth promoting husbandry during the first few years for a rapid expansion of tree size;
• as soon as the trees are big enough to produce a worthwhile crop, shift the emphasis in husbandry from growth promotion to application of stress and synchronization of the growth rhythm;
• stimulate fruit set through measures which enhance the quality of the flowers and improve cross pollination;
• time all cropping techniques in relation to tree phenology for maximum impact on either the growth or the flowering/fruiting stage, with minimal interference with the other component.

If this approach is successful, cropping techniques designed for more fruitful trees, such as pruning and flower or fruit thinning, come into their own (for instance, to improve fruit quality and to prevent biennial bearing).

The basis for success is a synchronous growth rhythm with a measure of stress at the right time. This is generally more easily achieved in a seasonal climate, e.g. with a wet season to promote vegetative growth and a dry season to stimulate floral development. In a monsoon climate growers control irrigation to advance or delay flowering by a few months in order to obtain an out of season crop.

The crop cycle can be shifted almost irrespective of the season by a very drastic cropping technique that is gaining importance in the tropics: defoliation. Leaves, particularly older leaves, inhibit the axillary buds and the sudden release of inhibition following defoliation leads to a general, simultaneous flush. Defoliation is mainly practised on fruit crops from the temperate zone to force budbreak before dormancy has been firmly established. Moreover, it shortens the crop cycle so that, for instance, apples produce 2 crops per year and some grape cultivars achieve 5 crops in 2 years. This makes year round production possible for crops which by their nature are strictly seasonal! In principle, defoliation works on all fruit trees, but flower buds must be ready to break into bloom following defoliation, and the grower should have adequate control over growing conditions for successful completion of the crop cycle.

Harvesting and post harvest handling

Harvesting is the last stage in the life of the fruits on a tree. It involves knowledge of the state of maturity of the fruit, and the proper way to pick the fruits so that they reach the market in excellent condition. Post harvest handling involves practices such as sorting and grading, packaging, transport, disinfection, after ripening and marketing.

Maturity indices

Fruits are harvested at a stage so that when they attain optimum ripeness, they have the acceptable colour, flavour, aroma, texture and other attributes characteristic of the species or cultivar. Fruits grown in South East Asia are harvested either at the mature stage (climacteric fruits) or ripe stage (non climacteric fruits). There are many fruit crops with climacteric fruits; the more common ones include the banana, mango, papaya, durian, sapodilla and jackfruit. These fruits ripen after picking. They are sold either at the mature or ripe stage. Species with non climacteric fruits include such common fruit crops as the langsat, rambutan, pummelo, mandarin and santol. These fruits are harvested at the ripe stage because they do not ripen any further once they are detached from the tree.

The date of picking of some non climacteric fruits is easy to determine. These fruits usually change colour on the tree as an indication that they are ripe. Pummelo, mandarin and sweet orange do not usually change peel colour in the tropics, so another maturity index is used to determine the date of picking. In the Philippines pummelos and mandarins are considered ripe when they attain soluble solids to titratable acid ratios of 10 : 1 and 7 : 1-10 : 1, respectively. Many citrus growers, however, do not have the wherewithal to measure total soluble solids and titratable acidity of their fruits. They determine the date of picking by testing the fruit periodically.

There are many ways of determining maturity and consequently the date of harvest of climacteric fruits. Days from full bloom is used with mango and this varies per cultivar: 84 days for the Philippine "Carabao", 90 days for the Indonesian "Arumanis", 100 days for the Thai "Nam Doc Mai" and 105 days for the Malaysian "Tok Boon" mango.

In the banana, roundness or fullness of fingers is used as the index of maturity. For home consumption or when the fruits are to be sold in local markets, the fruits are allowed to attain full size (fully rounded in cross section). For the "Giant Cavendish"" banana for the export market, the bunch is harvested when the middle finger of the outer row of the second hand attains a diameter of 1¼-1½ inches (32-38 mm ).

Other maturity indices used for climacteric fruits are change in skin colour (sapodilla, papaya, pineapple), distance between spines (soursop, jackfruit), and a hollow sound when the fruit is tapped with the fingers (jackfruit, durian). Measurements of specific gravity, total soluble solids and titratable acid require instruments, and are not usually used by fruit growers.

Harvesting

The quality of the harvested fruits is greatly influenced by two factors: the ability of the picker to choose mature fruit, and the method of harvesting employed. To ensure that only mature fruits are picked, priming or repeated harvesting of a tree is ideal. Although this is done for many fruits including jackfruit, sapodilla and caimito, the cost of this recommended practice is not always considered worthwhile: for example, all the mango fruits on a tree are usually harvested in a single picking.

When it is no longer possible to reach the fruits from the ground, the picker either climbs the tree or a ladder, or he uses a long pole fitted with a small basket. Handpicking is used for some fruits such as durian and langsat. In Thailand durian fruits are individually tapped to determine if they are mature. In other countries the fruits are allowed to mature until they fall to the ground. Some durian growers tie the fruit stalk to the branch with a string so that when it matures and detaches, it does not fall to the ground. Then the fruit is collected by climbing the tree. This method is as laborious as it is ingenious and shows to what lengths people will go to harvest a fruit when it is at its best. Langsat fruit bunches are harvested by cutting the stalk with a knife. Fruits from tall trees (e.g. mango, santol, caimito) are picked from the ground or after climbing the tree, using a half elliptical basket made of cloth or net attached to a long bamboo pole. A V shaped notch in the top of the pole is used to twist the fruit stalk until it breaks, or the pole is simply pushed up until the wire loop holding the basket forces the fruit from its stalk. Fruits are then transferred to a hanging basket which is lowered to the ground when full. In this way many stalks are pulled from the fruit ends, leaving wounds in which rotting begins. This can be prevented by fitting the pole with a sharp cutting edge.

Post harvesting handling

Once on the ground, the fruits are sorted according to size and appearance. The fruits are then packed into containers. Proper packing preserves the quality of the produce prior to and during distribution to the wholesale and retail markets. In South East Asia, bamboo baskets of various shapes, strengths and capacities are the most commonly used packing material. Wooden crates, corrugated cartons and rigid plastic (PVC) containers are also used for specific purposes because bamboo baskets do not provide much protection since they are flexible and easily distorted.

In the Philippines, bananas and pineapples for export are packed in corrugated carton boxes. Air shipment of rambutan, pummelo and mangosteen from Mindanao to Metro Manila is by wooden crate. From the production site, the fruits are transported to urban centres, processing factories and fumigation plants. Transport to nearby destinations is usually by road or rail. Inter island shipment in the Philippines is done by air or water, as is the case of fruits produced in Mindanao and marketed in Metro Manila. Fresh bananas, pineapples and other fruits are exported by air. Mangoes, papayas, and other fresh fruits that may be infested with fruit flies are fumigated prior to shipment to fruit fly free countries such as Japan. Mangoes from the Philippines, for example, are subjected to vapour heat treatment before they are shipped by air to Japan. Other post harvest treatments are degreening of citrus fruits by application of ethylene gas in a confined chamber. Pummelos are sometimes waxed to minimize shrivelling of fruits and to add gloss.

Processing and utilization

There are many instances when fruits and nuts are processed. By the very nature of their edible parts, nuts such as cashew, macadamia and pili have to undergo processing before they are consumed. Some fruits have cultivars that are specifically suited for processing, e.g. grape cultivars for wine or raisins, and strawberry cultivars for jam. There are commercial fruit cultivars which are consumed both raw and processed (e.g. pineapple, sweet orange). Fruits of other trees are either too sour (e.g. purple passionfruit, acerola) or too small (e.g. blueberry, bignay) to be eaten raw, and are more suited for processing than for raw consumption. Fruits are also processed when they are cheap because of a glut in production. Finally, fruits that are not marketable because of their small size, poor appearance, etc. are made more useful by processing.

Fruit processing in South East Asia is not yet a fully developed industry, despite the fact that various processing methods are known. The main reason for this is that South East Asians prefer to eat fresh fruit. Furthermore, processed fruits are more expensive and people therefore would rather wait for the next fruiting season. However, there appears to be an increasing demand for processed tropical fruits outside South East Asia, particularly in temperate countries, as evidenced by the rapid growth of processing companies that cater to the export market.

Processing of cashew nuts is complicated because the cashew shell liquid, an oil in the shell to which some people are allergic, has to be removed by roasting before the kernel can be extracted. Macadamia and pili nuts are processed by first removing the husk and then drying the nuts. Dried nuts may be stored for future use or shelled thereafter. Macadamia nuts are machine decorticated, but pili nuts are shelled by hand.

There are many ways to process fleshy fruits. The methods used in South East Asia include fermentation, pickling, drying and dehydration, juice extraction, glacéing, preservation and canning. Fermentation includes the production of wine (e.g. grape, cashew, jambolan, guava, pineapple) and vinegar (e.g. apple, banana). In the Philippines a form of fermentation involves the production of a white solid substance from coconut water ("nata de coco") or pineapple juice ("nata de piñ"") with the aid of cellulose forming bacteria (Acetobacter xylinum) in the presence of sugar and other substances. Cooked with sugar, the "nata" produces a delicacy which is well liked for dessert and as an ingredient of fruit salad.

Pickling is the preservation of green mature fruits in brine or vinegar with or without bacterial fermentation. Mango and papaya are fruits that are commonly pickled. Mango and santol are also made into chutneys.

Drying is one of mankind's oldest methods of preserving food by lowering its moisture content below that at which microorganisms can grow and reproduce. Fruit slices are dried by subjecting them to controlled temperature and humidity. Banana chips, for example, are prepared by drying previously sulphited (to prevent browning) unripe fruit slices (from cooking cultivars) in an oven at 70°C for 3 4 h until the slices are dried to a moisture level of 3 5%. Banana chips are also prepared by frying banana slices in vegetable oil. Dried mangoes may be prepared by oven drying rare ripe fruit slices. Banana chips and dried mangoes are important export products of the Philippines.

Glazed fruits are prepared by allowing whole fruits or slices to absorb sugar (from syrup) until the sugar concentration in the fruit is high enough to prevent spoilage. Sun or oven drying completes the process. Tamarind, bilimbi, carambola, santol and guava are common fruits that are processed by this method. The production of fruit into juice is a very important form of processing in South East Asia, and involves many fruit species, e.g. pineapple, mango, passionfruit, calamondin, guava and soursop.

Jellies are fruit products prepared by extracting the juice from boiled fruit, boiling it with sugar, and cooking it to a consistency at which gelatinization takes places upon cooling. Jellies can be made only from fruits rich in pectin and acid, although commercial pectin or juice from fruits rich in pectin may be added to the juice of low pectin fruits. Guava, bignay, jackfruit, passionfruit, santol, and tamarind are made into jellies.

Marmalades are clear jellies in which slices of fruit or peel are evenly suspended. Calamondin and santol are currently processed into marmalades. Jams can be made from practically all fleshy fruits. They are prepared by boiling whole fruit or pieces of fruit pulp with sugar to a moderately thick consistency. Combinations of fruits can be used to produce exciting blends. Jams are commercially made from mango, pineapple, guava, jackfruit, papaya pineapple, and soursop.

Fruit preserves are whole fruits or large pieces of fruit cooked in a thick, clear syrup. Many fruits are currently processed into preserves, the more common ones being banana, calamondin, bilimbi, soursop, mangosteen, jackfruit, breadfruit (seedless), santol and guava.

Whole fruits, halves or pieces can be preserved without previously cooking them. They are packed in a light syrup and canned. Mango, pineapple, lychee and rambutan may be processed in this way. This method retains the texture, flavour and aroma of the fresh fruit.

Breeding and genetic resources

Breeding work does not have high priority in tropical fruit growing. For one thing it requires a long term research commitment, the benefits of which accrue in the distant future. Fortunately, superior genotypes can be cloned and grafting makes it possible to combine favourable features of rootstock and scion; in this way, a selection from a progeny can reach the fruit grower within a reasonable number of years. A breeding programme - particularly for tree crops - requires that the breeder has a good understanding of the selection criteria, so that the right priorities can be set. Although selection in tropical fruit crops must have gone on for a very long time, it appears to have been based mainly on fruit quality, with yield often taking second place.

The limited experience with fruit tree breeding makes clear that selection should be based on a much wider range of criteria. Indications are that precocity - an important selection criterion in itself - is closely correlated with the productivity of the mature tree. Moreover, precocity is related to the branching habit of the young tree: trees with compact shoots ("spur types"") or with early development of laterals emerging at a wide angle to the seedling axis, are generally more precocious and much higher yielding.

These relationships have largely been established in breeding work with the leading fruit crops of the subtropics and the temperate zones; they may not apply to tropical fruit crops in which the nature and the position of the inflorescences (large terminal inflorescences, cauliflory, flowering on specialized shoots) are often quite different. Nevertheless, this example shows that a much better understanding of variations in tree architecture and in their expression in tree habit and phenology is needed before efficient breeding programmes can be designed. The same applies to breeding for disease and pest resistance: breeding can only be undertaken on a strong footing when the relation between the life cycle of the pathogen and tree phenology is understood. Breeding techniques may also need adaptation to the special features of floral biology (dioecy, polygamy, parthenocarpy, different forms of apomixis including polyembryony) of the tropical fruit crops. It is therefore not surprising that so far breeding work has been virtually limited to short duration fruit crops such as pineapple, banana, papaya and passionfruit. However, gradually the woody tropical fruit crops will also be tackled by breeders. Encouraging results have already been achieved with, for instance, guava and macadamia.

Both direct selection and breeding work depend on the available genetic diversity of the species. The seed of most tropical fruit species loses its viability quickly, and germplasm collections consist largely of living trees. The large number of species, the number of specimens required per species, and the specific ecological requirements of some of them make it impossible to maintain adequate germplasm collections. The genetic diversity of cultivated fruit crops is narrowed by clonal propagation of a limited number of cultivars. Hence, the wild trees of these species are essential as genetic resources; the same applies, of course, to the majority of the species dealt with in this volume, which are only found in the wild. Conservation of the natural habitats of the species is therefore essential for the long term future of tropical fruit growing.

Prospects

This volume testifies to the great wealth and diversity of edible fruits and nuts in South East Asia. The appreciation of these fruits and nuts is reflected in the many ways they are put to use and in their role in cultural traditions. Unfortunately, fruits and nuts generally yield erratically rather than prolifically. This seriously limits the scope for commercial cropping systems; most fruit in South East Asia still comes from home gardens, field borders, and the like. Erratic yields and low efficiency of fruit production and marketing in these cropping systems make fruit expensive, so that most people cannot eat as much fruit as they need or as they would like. It follows that a breakthrough in productivity is required to produce much more at much lower prices; this will restore fruits and nuts to the position they traditionally held in South East Asia. National policy tends to emphasize the development of exports of fruits and nuts rather than domestic consumption. Larger supplies at lower prices are more competitive in export markets as well, and a well supplied domestic market is the best possible basis for export. Thus a breakthrough in productivity strengthens the role of fruit growing in both the domestic and export markets.

So far, it appears that in the tropics relatively little research work has been devoted to fruit trees themselves, in comparison with the research attention given to nursery stock and to the fruit. In South East Asia the notion that the growers know all there is to know about their trees is fairly common. This may have contributed to the stagnation in yield levels of fruit crops, whereas the yields of most tropical field crops have risen steadily through sustained research efforts. Where low productivity is the root of the problem there is much to say for a drastic shift in emphasis in research work from the fruit to the tree. The dramatic rise in productivity which has transformed some plantation crops and temperate fruits, still has to reach tropical fruits. The study of the basis of fruit tree productivity is much more demanding than experimenting with the fruit or with propagation methods and requires a strong long term commitment. Recent developments in South-East Asia, particularly in Thailand, suggest that the integration of traditional skills by professional fruit growers is already leading to a breakthrough in yield levels. Working with these growers, scientists do not face a superhuman task in unraveling how fruit trees function and reaping the rich rewards which the fruits and nuts of South East Asia have to offer.