PROSEA, Introduction to Stimulants

Plant Resources of South-East Asia Introduction

List of species

Introduction

Definition and species diversity

Choice of species

Most of the plant resources brought together in this volume have in common that they are cultivated or collected for the stimulating properties of certain alkaloids (e.g. caffeine) and other psycho-active compounds that are present in non-toxic concentrations in the seeds, leaves or other harvested plant parts. Based on general use, three types of stimulant plants may be distinguished:

1. plants used to make beverages: hot or cold infusions are prepared, mainly from dried and processed seeds or leaves;
2. masticatory plants: the stimulating effect is obtained by chewing plant parts, often fresh or dried seed, sometimes in combination with other plant products;
3. plants whose dried and processed leaves are smoked.

Full descriptions of 15 major stimulants (20 species) are given in Chapter 2. An overview of the origin, geographic distribution and other basic data for each of these 15 stimulants is presented in Table 1. Tea (Camellia sinensis (L.) Kuntze), coffee (Coffea arabica L., C. canephora Pierre ex Froehn. and C. liberica Bull ex Hiern), cocoa (Theobroma cacao L.) and tobacco (Nicotiana tabacum L. and N. rustica L.) are the four dominant and globally most significant stimulants. The major stimulants of regional importance are maté (Ilex paraguariensis A. St.-Hil.) and guaraná (Paullinia cupana Kunth) both in South America, areca nut (Areca catechu L.) in combination with betel pepper (Piper betle L.) in South and South-East Asia, kola (Cola nitida (Vent.) Schott & Endl. and C. acuminata (P. Beauv.) Schott & Endl.) in West Africa and kava (Piper methysticum G. Forster) in Melanesia and Polynesia. Another 5 stimulants (6 species) have also been included because of their local importance in South-East Asia as beverage plants. In certain cases a given plant resource is used as much for its medicinal or refreshing as for its stimulating properties, and its allocation to the group of stimulants is somewhat arbitrary.

Brief descriptions of 33 minor "stimulants" (34 species) are presented in Chapter 3, which also includes some species used in association with stimulants, such as plants used to wrap tobacco (Licuala pumila Blume, L. rumphii Blume and Microcos paniculata L.). Several species that are difficult to assign to Prosea commodity groups have also been included: two species that are used to obtain potable water (Desmos dumosus (Roxb.) Safford and Piper bantamense Blume) and two species that are used to improve the taste of palm wine (Garcinia amboinensis Spreng. and Garcinia picrorhiza Miquel). Other plant species whose use as a stimulant or in association with stimulants is a secondary use are listed in Chapter 4.

Origin and domestication

The majority of stimulants are of tropical origin, though a few originate from the subtropics. Asia, Africa and South and Central America have been equally important in terms of the origin and first domestication of the great beverages (tea, coffee and cocoa), tobacco and another 5 major stimulants (Figure 1).

Tea is probably the oldest domesticated stimulant. According to Chinese legend, an infusion from tea leaves was used for medicinal purposes as early as 3000 BC, but written evidence of widespread tea (Camellia sinensis (L.) Kuntze var. sinensis) cultivation in China dates back to AD 350. The local people of north-eastern India, Burma (Myanmar) and adjacent countries have long been familiar with wild tea, using the leaves for chewing or preparing infusions, but domestication of Assam tea (Camellia sinensis (L.) Kuntze var. assamica (Mast.) Kitamura) probably started after 1840 with the first British tea plantations in northern India. The cultivation of arabica coffee (Coffea arabica L.) and the preparation of a beverage from roasted and ground seeds developed in Yemen, probably in the 12th or 13th Century, and before long coffee was known as the "wine of Arabia". There may have been some form of domestication before that time in the region of origin of arabica coffee in south-western Ethiopia; early Arab travellers must have learned the stimulating properties of infusions from arabica coffee leaves, fruits or seeds from the local people. Uganda has a centuries-old tradition of deliberate planting of coffee (Coffea canephora Pierre ex Froehn.) near homesteads, and since time immemorial the inhabitants of central Africa have chewed dried seeds collected from wild or planted coffee for their stimulating effect. Cocoa and tobacco are both of South American origin, but it was Central American people like the Maya who started cultivating these crops on a large scale more than 2000 years ago. Cocoa became of major socio-economic significance to pre-Columbian civilizations of Central and South America as a nutritious and stimulating beverage and also as general currency. Tobacco found widespread use, even among indigenous North American peoples for ritualistic, medicinal and hedonistic purposes. The South American beverages maté and guaraná were domesticated only a few centuries ago close to their area of origin, as was kola in West Africa. On the other hand, the first cultivation in Asia of the areca and betel pepper, two of the essential ingredients of the "pan" (South Asia) or "sirih" (South-East Asia) quid, could have been 2500-3000 years ago. Kava, from which the Polynesians prepare their national beverage, was domesticated around AD 1000. German chamomile (Matricaria recutita L.) originates from the Mediterranean region, while the remaining 4 stimulants listed in Table 1 originate from Asia or South America. Domestication of the latter stimulants was generally restricted to backyard gardens. Certain stimulant plants (e.g. Senna occidentalis (L.) Link) are sometimes considered to be weeds. The minor stimulants are all endemic to South-East Asia, except for two species of African origin Coffea congensis Froehner and C. stenophylla G. Don. Many of these are not or only partly domesticated.

Geographic distribution

Tea, coffee, cocoa and tobacco were domesticated by Chinese, Arab and Central American civilizations respectively, long before these civilizations came into contact with the first European explorers at the beginning of the 16th Century. However, the expanding and competing colonial empires - first the Spanish and Portuguese but soon followed by the Dutch, British and French - were quick to appreciate the social and economic potential of the novel stimulants. The demand for these commodities rapidly exceeded the amount obtainable by trade with the original sources of supply, so cultivation of major stimulants was subsequently introduced into the newly colonized areas. Cocoa and tobacco were already being cultivated outside the traditional regions by the end of the 16th Century. Coffee produced in Indonesia and South America started to arrive in Europe in large quantities by the middle of the 18th century. On the other hand, China maintained a monopoly on the international tea trade until the end of the 19th Century, at which time tea produced on plantations in India, Sri Lanka and Indonesia finally started to overtake the tea from China in volume.

At present, these four stimulants have a wide distribution: cocoa and coffee are mainly confined to the tropics, tea cultivation extends into subtropical regions, but the annual tobacco is also grown during summer in temperate zones. Many other stimulants - e.g. kola, maté, guaraná and kava - have remained close to their origin and region of first domestication. An overview of present main geographic distribution of all major stimulants is given in Table 1.

Role of stimulants

Social and cultural aspects

The human interest in tea and coffee may initially have had a religious motive, e.g. tea to keep Buddhist monks awake during long hours of meditation in China and coffee as an acceptable substitute for alcoholic drinks in Islamic Arabia. However, these stimulating beverages would soon gain general popularity at all levels of societies in and around the countries of first cultivation and eventually also in Europe and the Americas. The numerous tea and coffee houses of 17th and 18th Century Europe, as places for people to meet and debate important issues, have been of great influence on major cultural, commercial and even political developments of the last 200 years. Without exaggeration it can be stated that most people in the present world drink tea, coffee or both on a daily basis to banish fatigue and refresh the mind. The two beverages continue to have considerable social functions, starting with the simple gesture of offering a guest a cup of tea or coffee to convey hospitality.

Tea is the predominant beverage in Asia. The elaborate rituals of tea ceremonies bear witness to its significance in Chinese and Japanese cultural traditions, but in most other Asian countries too, tea drinking has a role in social life. After early proliferation of coffee houses in 17th Century London, Great Britain and also Ireland became nations of tea drinkers in the 18th Century, initially for political and economic reasons. Exactly the opposite happened in the United States, after the famous Boston tea party of 1773. The Russian preference for tea as non-alcoholic beverage, dates back to 17th Century supplies of tea obtained overland from China. Tea drinking has overtaken coffee in some of the Arab and Middle-East countries, even in Yemen - the country where arabica coffee was first domesticated. Tea is generally more popular than coffee in most African countries too, except in Ethiopia, where more than half of the total coffee crop is consumed locally. Coffee drinking is a typical aspect of the societies of continental Europe, the United States and several Latin American countries, contributing as much to the quality of work as well as that of social life. Coffee consumption is increasing in the United Kingdom and, with the globalization of the Western lifestyle, coffee is becoming more popular in traditional tea countries such as Japan, South Korea and even China.

Cocoa was known in 18th Century Europe, where it continues to be consumed by the more affluent. Chocolate consumption is increasing in countries such as Japan and Singapore, but about 85% of world cocoa production is still exported to Europe and North America.

The smoking of tobacco in cigarettes, cigars or pipes, reached global proportions more than a century ago and the habit has made an impact on almost all societies of the world. The dependence on this stimulant met with little religious or cultural resistance, and smoking was universally appreciated for its socializing aspects. Tobacco consumption is still on the increase in Asia and the developing world in general. On the other hand, there is growing opposition to tobacco in view of the potential harm to public health, and the smoking of cigarettes in particular is now often considered to be an anti-social activity in Europe and North America.

Regionally important stimulants, such as maté, guaraná, kola, areca nut (with betel pepper) and kava, all have considerable cultural and social functions. The remaining species treated as major stimulants are mostly of very restricted or traditional use, sometimes with occasional additional medicinal applications.

Nutritional and health aspects

With the exception of cocoa (> 50% fat), the nutritional value of the products from major stimulants is negligible. On the other hand, the various alkaloids and phenolic substances present in these products often have considerable physiological effects on the human body. The effects on health have been extensively investigated for the universally important stimulants tea, coffee, cocoa and tobacco and some of the recent conclusions will be reviewed below. The health risk associated with maté and guaraná is considered to be low, although supporting experimental or epidemiological evidence is scarce. For the remaining stimulants the risk is unknown, or assumed to be low to considerable, depending on concentration and frequency of consumption. An overview of the main active compounds and estimated health risk for the major stimulant plants is given in Table 1.

Caffeine

The occurrence of the purine alkaloid caffeine (1,3,7-trimethyl-xanthine) appears to be restricted to six higher plant genera: Camellia L., Coffea L., Theobroma L., Ilex L., Paullinia L. and Cola Schott & Endl. (Crozier et al., 1998). Caffeine originates in the young leaves and fruits by biosynthesis from purine nucleotides. The chemical structure of caffeine is shown in Figure 2a. Theobromine (3,7-dimethylxanthine) is similar, except for the absence of one methyl group. Due to the universal consumption of tea and coffee, especially the latter, caffeine is certainly the most widely used psycho-active substance in the world. The caffeine content of the cured product and the quantity of caffeine intake per cup of tea, coffee and cocoa, are presented in Table 2 (Baumann, 1996; Chow & Kramer, 1990). A 100 ml cup of tea contains half to one-third as much caffeine as a cup of coffee. The cocoa drink has less caffeine than a cup of tea, but also contains 100-120 mg theobromine, which, however, has only one-tenth of the stimulating activity of caffeine.

Caffeine is odourless and has a bitter taste, although this contributes only partly to the perceived bitterness of the consumed stimulants. After ingestion, caffeine is rapidly absorbed from the gastro-intestinal tract, reaching maximum blood plasma levels within 30-45 minutes. About 50% of the caffeine is metabolized and excreted as methylxanthine derivatives in the urine 5-6 hours later. Low to moderate caffeine doses, not exceeding a daily total of 200-600 mg, usually result in increased alertness, energy and ability to concentrate. Higher doses may induce negative effects such as anxiety, nervousness, insomnia and rapid heart action, but individual variation of tolerance levels is considerable. In contrast to the drugs of abuse, addiction to caffeine occurs only at very high doses that are rarely reached through the regular consumption of tea, coffee or other caffeine-containing beverages. In general, caffeine is not a threat to the health of most people (Clarke & Macrae, 1988a; Nehlig, 1998; Smith, 1985).

Phenolic substances

The flavanols in tea (30-35%) and chlorogenic acids in coffee (6-15%) contribute to the characteristic colour and taste of the respective beverages (Clifford, 1985; Robertson, 1992). There is no evidence of negative biological effects; indeed, due to their anti-oxidant properties these polyphenols or their metabolites may provide some protection against carcinogenic substances (Clifford, 1998; Marks, 1992; Viani, 1993).

Nicotine

The perceived strength of tobacco is primarily determined by the nicotine content. The chemical structure of this alkaloid is presented in Figure 2b. Nicotine has a considerable addictive effect, even at low doses, and because of the hazardous carcinogenic constituents present in tobacco and its smoke this poses potential health risks to regular consumers (Shopland et al., 1991).

Economic aspects

Table 3 presents a summary of data on production and international trade for the globally most important stimulant crops, as derived from statistics (1995-98) provided by the Food and Agriculture Organization of the United Nations (FAO) and various commodity trade associations, such as the International Coffee Organization (ICO), International Cocoa Organization (ICCO), International Tea Committee (ITC) and Tobacco Merchants Association (TMA).

Compared with a single cereal crop like rice, the area cultivated with tea, coffee, cocoa and tobacco together (23.7 million ha) is less than 17% and total production (17.3 million t) only about 3%, but combined trade value of the cured products is estimated at US$ 35-44 billion. Coffee alone accounts for almost half of this value, followed by tobacco, then tea and lastly cocoa. The value of the manufactured products at retail prices is likely to be 4-10 times higher. The contribution of these four stimulants to the economies of several producer and consumer countries and to world trade is therefore considerable.

A large number of countries, the majority within the tropics, produce one or more of these stimulants. India and China are prominent producers of tea and tobacco, Brazil of coffee, cocoa and tobacco, Indonesia of coffee, tea and cocoa, Ivory Coast, Ghana and Nigeria of cocoa, Sri Lanka and Kenya of tea. Continent-wise (Table 4) Asia is predominant in tea (75%) and tobacco (59%), Africa in cocoa (66%) and South and Central America in coffee (61%). Tea and tobacco are produced outside the tropics to a considerable extent.

Table 3 also indicates the variation in export and consumption patterns. Cocoa is almost exclusively an export product for all countries except Brazil. Coffee production is also export-oriented in most countries, but in Brazil, Indonesia and Ethiopia there is also considerable local consumption. On the other hand, some of the most important producers of tea (China, India) and tobacco (China, India and the United States) have larger internal than export markets.

Data on area and production per country in the South-East Asian region for the four major stimulants are given in Table 5. The region accounts for 5-17% of world production of the four major stimulants. Indonesia is predominant in all four, Vietnam has once more become an important coffee producer and Malaysia is still a significant cocoa producer.

Maté is produced in Argentina, southern Brazil and Paraguay on 0.5 million ha. Most of the product (0.5 million t/year) is consumed in Latin America, with minor exports to the United States, Japan and the European Union. Kola is an important product (200 000 t) with considerable commercial value in West Africa (Nigeria, Ghana, Benin, Sierra Leone). The Caribbean and South America are significant export markets for kola nuts. Guaraná is produced exclusively in the Amazon region of Brazil (4000 ha) and only a small amount of the total production of 1200 t is exported to Japan and the United States. For the remaining major stimulants listed in Table 1 statistical data on production are incomplete or unavailable.

Botany

Taxonomy and morphology

An overview of the 20 major and 34 minor species with stimulant properties is presented in Table 6. These belong to 21 different families, all of them being dicotyledons except one family of monocotyledons, the Palmae, which alone contains as many as 10 stimulant species (including Areca catechu). The Rubiaceae is the next most important family with 9 species (including Coffea species), followed by the Piperaceae with 6 species, the Sterculiaceae with 4 species (including Cola species and Theobroma cacao), the Solanaceae with 3 species (including Nicotiana species) and the Chloranthaceae with 3 species. The remaining families include only one or two stimulant species, such as the Aquifoliaceae (Ilex paraguariensis), Camelliaceae (Camellia sinensis) and Sapindaceae (Paullinia cupana).

Details of type of plant, plant parts harvested and main use for each stimulant are also indicated in Table 6. The vast majority of the stimulants are perennials: woody shrubs and trees (30), palms (10) and woody or epiphytic vines (8). Only 6 stimulants are annual, biennial or perennial herbs, belonging to the Compositae, Labiatae, Leguminosae and Solanaceae.

The plant parts harvested are roots (6 species), bark (7 species), sap from the stem (4 species), leaves (23 species), flowers (2 species), infructescences (1 species) and seeds (18 species). In half of the stimulant species these plant parts are processed to prepare a beverage, in about one third they are processed into a masticatory and in about one sixth they are processed for smoking. Only in the case of Theobroma cacao can the harvested plant parts be considered to be nutritious as well as stimulant.

The remaining sections of this introductory chapter focus on three major stimulants: tea, coffee and cocoa. The wealth of published information available for these crops and the analogies in ecological requirements and crop production practices favour a comparative approach. Though tobacco is also of major economic significance it has little in common with the other three major stimulants because it is an annual crop.

Growth and development

Structures and processes

Tea, coffee and cocoa belong to the group of branched perennials (shrubs/trees) with an episodic growth rhythm. This means that the extension of the shoots is not constant but fluctuates. It is an important mechanism regulating the activity and growth of the numerous meristems of the branched species and balancing the growth of aerial and subterranean plant parts. Periods of quiescence in shoot growth also permit the initiation and development of flowers and so ease the apparent antagonism between vegetative and reproductive growth. Both coffee and cocoa show a further adaptation to cope with this problem. In coffee, shoot dimorphism spatially separates the vegetative growth of the orthotropic stem and flower development on the plagiotropic branches, while in cocoa cauliflory (i.e. flowering on the stem and older branches) fulfils the same function. Absence of such an adaptation in tea has no implications for its cultivation, because the crop is a vegetative product and periodic removal of young shoots prevents flowering.

Whereas rhythmic growth of shoots is controlled by endogenous factors of individual plants, environmental factors can synchronize this growth habit at plant and population levels. In seasonal climates this leads to distinct phenological cycles of flushing, flowering, fruiting and leaf fall. Another specific feature of the branched perennials is that the mature generative phase of development is usually preceded by a juvenile phase. In cocoa, for example, young seedlings elongate their stem at about 6-week intervals with a shoot section carrying 5-7 leaves in a spiral arrangement. When the plants reach a height of 1-2 m, vertical growth ceases and up to 5 buds at the end of the stem may grow out simultaneously into side branches with distichously alternate leaves. This is the starting point of the mature phase; the rhythmic growth of branches and flowering begins. The juvenile phase is characterized by orthotropic shoot growth and the mature phase by plagiotropic branching. In coffee seedlings the juvenile phase ends when the orthotropic stem starts to produce plagiotropic branches. As the vertical growth of the orthotropic stem continues and only the plagiotropic branches flower, it is probably better to say that in coffee the juvenile phase is followed by a stage in which juvenility of the stem and maturity of the branches co-exist.

Scope for interventions

Cultivation practices cannot change rhythmic growth, but in some cases they can influence synchronization. An example is the use of irrigation after a dry period to stimulate and synchronize flowering in coffee. Periods of immaturity, however, can be shortened by cultivation practices and also by selection for early branching and bearing. Vegetative propagation using plagiotropic, physiologically mature plant material is an obvious technique to eliminate the unproductive juvenile phase. Although proper training techniques enable plagiotropic material of cocoa to be successfully used to produce early bearing plants, this practice cannot be applied to coffee. Cuttings from the plagiotropic branches of coffee produce a horizontally growing shrub that cannot be developed into an erect plant. In tea growing it is not relevant to shorten the juvenile period, so the common application of vegetative propagation is for the large-scale multiplication of superior plants. A complex issue for agronomic interventions in the generative phase is the already mentioned antagonism between vegetative and reproductive growth. Coffee and cocoa give two contrasting models for the processes involved. In coffee, fruits are very strong sinks for assimilates. With heavy cropping these are withdrawn from leaves and branches and sometimes the reserves in stems and roots are also utilized. This may lead to a biennial cropping pattern and in extreme cases to dieback of leaves and branches. In a crop like coffee, which lacks an efficient physiological mechanism of fruit thinning, measures are needed to stimulate vegetative growth or regulate fruiting and so match the requirements for assimilates with actual supply. The reverse situation is found in cocoa, in which shoot development out-competes young fruits for assimilates. This results in the wilting of young fruits, a process which is commonly called cherelle wilt. In cocoa, measures to stimulate vegetative growth may suppress yields.

In a crop yielding a vegetative product, such as tea, cultivation measures are straightforward and should simply be directed towards maintaining a well-developed layer of photosynthetically efficient leaves that supply the reserves for continuous production of young shoots and leaves that are regularly harvested.

Ecology

Climatic factors

In the natural environment and the major production areas of tea, coffee and cocoa, light is one of the most important ecological factors affecting growth and development. Wild coffee and cocoa are both components of the understorey of tropical forests. Adaptation to conditions of low light intensity is still reflected in the current cultivars. Coffee and cocoa leaves have a low light saturation point associated with a low photosynthetic rate, which declines when leaves are exposed at above-optimal light levels. The ecology of the tea plant in its natural habitat is not known for certain. Tea leaves do not easily reach light saturation in full sunlight but still have a low photosynthetic rate. Exposure to full sunlight is associated with a rise in leaf temperature. When this exceeds 35 °C, the net assimilation rate of tea leaves starts to decline.

The temperature and rainfall requirements for commercial cultivation of tea, coffee and cocoa are given in Table 7. Whereas both tea and cocoa produce best in humid, non-seasonal climates, coffee needs a distinct dry period for synchronized flowering and successful pollination. During the dry season water stress gradually decreases the dormancy of the flower buds and a sudden relief by rainfall or a drop in temperature triggers flowering. This is why in seasonal climates well defined cropping periods are found, with implications for labour demands and processing capacity. Seasonal flushing and cropping may play an important role in crop protection because they interrupt the life cycles of pathogens and pests.

The temperature and moisture requirements of the three crops can be met by carefully selecting the growing site, while to a certain extent cultivation practices can tune the environment to the crop requirements. Matching the prevailing light conditions with the specific requirements, however, can only be achieved by agronomic interventions such as shading and plant density and to a limited extent by breeding and selection.

Soils

Generally speaking, tree crops require deep, well-drained, clayey soils with a high nutrient content and a topsoil rich in organic matter. The soil should be rootable to at least 150 cm deep. The high organic matter content is important for retaining nutrients, especially in soils with a low cation exchange capacity and also as a source of mineralizable nitrogen and phosphorus. Cocoa is among the most demanding tree crops in terms of soil conditions. A special feature of tea is its requirement of acid soils with a pH of between 4 and 6.

As chemical conditions of soils can be improved by fertilizer use but physical defects can often not be amended, the soil suitability criteria for tree crops are mainly based on the presence or absence of certain physical characteristics. In sub-optimal climates, soil conditions are more critical than under optimum climatic conditions. In areas with frequent and long dry spells, for example, tree crops can only be grown on soils with much moisture storage capacity, a property mainly depending on the texture and the potential rooting depth of the soil.

Agronomy

A special feature of the cultivation of tea, coffee and cocoa is their long economic lifetime. With adequate management and periodic "rejuvenation", tea and coffee plantations can remain highly productive for more than 50 years, whereas the economic lifetime of cocoa may vary between 20-40 years. This gives scope for long-term investments, crop monitoring for mineral nutrition and biological control of diseases and pests. Another feature in common is the large variation in production systems and cropping intensity.

Production systems

The three stimulant crops are grown in systems varying from extensively managed mixed cropping systems to intensively managed home gardens and sole crop plantations. Smallholder fields are often less than 0.5 ha, while plantation areas may be well over 2000 ha.

Smallholder production systems

Smallholders play a very important role in the production of coffee and cocoa. Apart from historical events, four reasons in particular may explain the success of the smallholder sector. These are:

1. the shade tolerance of the plants which allows mixed cropping with annual and other perennial crops;
2. the simple, non-capital-intensive establishment methods and the processing techniques giving a product which can be temporarily stored on-farm;
3. an abundance of virgin forest land and labour;
4. and the rapidly expanding demand on the world market.

The development of the entirely smallholder-based cocoa production sector in West Africa highlights these points. In the last decades of the 19th century returning plantation workers brought cocoa seeds from Fernando Po to the forest belt of the West African mainland, where cocoa growing fitted well into the existing food crop shifting cultivation system. Farmers planted cocoa seeds at close spacing among their food crops. As food crop cultivation was abandoned, cocoa seedlings were left to compete with the regenerating forest which was gradually cut back, leaving a few trees as shade. Under favourable conditions, this resulted in a dense irregular stand of cocoa as a separate new component of the farming system. Encouraged by good crop growth and economic prospects, farmers started to grow cocoa in selectively thinned forests. Processing consisted of fermentation of beans in baskets or in heaps covered by banana leaves to remove the mucilage, followed by sun-drying.

Abundance of virgin forest land, availability of labour (migrant labourers), and access through nearby ports to the expanding world market further contributed to the exponential trend of the West African cocoa production in the first half of the 20th century. Government support for the farmers' cocoa cultivation was originally export-oriented: development of infrastructure to facilitate access to cocoa-producing areas, improvement of bean quality and marketing. Only after major diseases had started to threaten the existence and continuity of cocoa growing, were major investments made in research and extension (Anonymous, 1946).

The spectacular development of smallholders' cocoa in Sulawesi, Indonesia, in the late 1970s and 1980s has features in common with cocoa development in West Africa. On this island cocoa had become an almost forgotten crop. However, when the news of the cocoa boom in Sabah (East Malaysia) reached farmers in Sulawesi, the interest in cocoa revived. The presence of forest land, surplus labour migrating from densely populated areas to the forest zone and a well-developed private trade sector provided the basic conditions for the massive farmers' response to the high cocoa prices on the world market. With additional government support cocoa production rose within less than a decade from a few tonnes to about 120 000 t dry beans in 1991 (Ruf, 1994).

Smallholder coffee in Indonesia has a long history. Shortly after coffee was introduced in Java in 1696, plants were distributed to farmers. In the 19th Century there was a period in which coffee growing was compulsory. At present, smallholder coffee is mainly found on Sumatra. Growing conditions and cultivation methods vary widely: where land is scarce it is grown in home gardens as a horticultural crop and where land is abundant it is extensively grown on "ladang" fields and partly cleared forest land. In the first years after planting it is normal practice to practice mixed cropping of coffee and food crops (Paerels, 1949).

In Indonesia, smallholder tea cultivation is a spin-off from the plantation industry. Plantation labourers and farmers in West Java were encouraged to establish tea gardens and sell their product to the estates. This outgrowers concept is currently widespread in other estate crops such as rubber and oil palm. In Kenya the smallholders' tea production is based on cooperatives that provide the farmer with planting material, inputs and management advice. The cooperatives collect the tea leaves on the farms and are responsible for the processing and marketing. This system gives a high-quality export product and a regular income to the farmers.

Generally speaking the three stimulants are "engines of development" of smallholder farming systems. They provide households with cash income for livelihood and education and for the purchase of inputs for food crops. The revenues from the stimulants are usually not invested in the crops concerned, hence the yields remain low. As long as land is not limiting, low yields can be compensated for by cultivating relatively large areas. Population growth and increasing demands for food, however, will certainly reduce the area for cash crops and bring the need to intensify production systems.

Plantations

In sharp contrast to the smallholder sector, large-scale plantation agriculture is capital-intensive and highly dependent on external inputs. Whereas farmers direct their attention and resources to the various components of their farming system and for obvious reasons give priority to food crop growing, estates have as their major objective the growing and processing of a single crop and the maximizing of profits. A striking feature of the plantation approach is the investments in high quality planting material, land preparation and processing facilities. The latter allow production of quality products: black instead of green tea and fermented and washed instead of dry-processed coffee.

By selecting suitable environments, hiring of labour and management, and attracting capital for investment, estates have created conditions in which research has contributed to a large increase in yield per unit of land and time but to a lesser extent per unit of labour.

Propagation

Given the long lifetime of perennial stimulants, it is best to use high-quality plant material. This involves the production of seeds of known parentage in seed gardens, or the use of cuttings and budwood from selected "mother" trees, as well as raising uniform plants under optimum nursery conditions. In coffee and cocoa, seedlings are mainly used. They are relatively cheap to produce, do not require early pruning, while the presence of a taproot gives them an establishment advantage over cuttings, especially in seasonal climates. In modern tea growing, cuttings have become the preferred method of propagation.

The nursery periods for cocoa and coffee are of the order of 4-6 and 6-18 months, respectively. For tea, the length of the nursery period depends on the transplanting method. Plants with adhering soil can be planted at an age of 12-24 months. Under unfavourable soil and climatic conditions, nursery seedlings are allowed to grow much longer to accumulate sufficient food reserves in the taproot and stem (stump planting method).

Land preparation and field planting

In extensive systems of smallholder production, land preparation for tree crops is no different from that for food crops. The fallow vegetation is slashed and burnt before plants are sown or planted directly without tillage. Mixed cropping of food crops and coffee or cocoa is common practice. In coffee this may last about one year, whereas in cocoa perennial food crops such as banana, plantain and cassava are kept longer as temporary shade. This is usually succeeded by permanent shade from regenerating forest trees and palms or planted fruit trees. When forest land is abundant, cocoa is also planted in selectively thinned forest.

The land preparation for plantations is more elaborate. After clearing the existing vegetation, cover crops are often sown to protect the soil and to maintain or restore soil fertility. In view of the high plant density used for the three crops, the cover crops will disappear after the early years of establishment. In future cocoa fields, shade trees are planted one or two years ahead of the cocoa, often with additional legume hedges for temporary shade and lateral protection. On sloping terrain, land preparation also involves terracing and/or contour planting of hedges and grass bunds.

In seasonal climates transplanting from the nursery into the field should take place early in the wet season to ensure proper establishment before the dry season starts. Plants are planted at final densities of 1000-1300 plants per ha for cocoa and coffee, and 11 000-14 000 plants per ha for tea. Following the development of dwarf cultivars high-density planting of coffee has been introduced (3300-5000 plants per ha). Some experimental high-density plantings of cocoa clones have been successful in Sabah (Yapp & Hadley, 1994). These systems require intensive management: a horticultural instead of an agricultural approach.

Shading

Shade requirements and effects

Tea, coffee and cocoa differ considerably in their requirements for shading (Willson, 1999). Saturating irradiances for leaf photosynthesis are much higher for tea than for coffee or cocoa, allowing most tea to be grown successfully without shade trees, except where temperatures become very high during the peak growing season (e.g. low-altitude areas in north-eastern India and Bangladesh). Arabica and robusta coffee are grown with and without shade trees, depending on the environmental conditions and crop management practices. Coffee leaves have the photosynthetic characteristics typical of shade-adapted plant species (Cannell, 1985), but cultivation without shade is often possible due to the high degree of self-shading in the deep and well structured canopies of intensively managed coffee bushes. Cocoa requires more shade than coffee. Shade is essential in the nursery and during the first years in the field to produce trees of the right shape and geometry, but it can be reduced in mature cocoa because of the self-shading effects of the closing canopies. Yield responses to fertilizers increase as shade intensity is reduced (Wessel, 1985). Nevertheless, the high yield levels of unshaded cocoa can only be sustained by high-input farming practices and under very favourable conditions of climate and soil.

Basically, shade reduces plant stress by improving climatic and soil conditions, but shade trees may also compete with the main crop for soil moisture and nutrients. Important beneficial effects to be attributed to shade in coffee and cocoa include:

1. prevention of over-bearing in coffee and excessive vegetative growth in cocoa due to partial interception of incoming light,
2. lowering of moisture stress by moderation of temperature extremes, reduction of air movement, increase of air humidity and improved availability of soil moisture
3. and amelioration of soil fertility and reduced soil erosion (Beer et al., 1998).

The effects of shade on disease and pest incidence vary. For instance, diseases such as blister blight (Exobasidium vexans) in tea and black pod (Phytophthora spp.) in cocoa are favoured by shade and so is the coffee berry borer (Stephanoderes hampei). On the other hand, the white stem borer (e.g. Xylotrechus quadripes) in coffee and mirids (e.g. Helopeltis spp.) of cocoa always cause much more damage where the shade canopy has disappeared.

From mono-shade to agroforestry systems

When environmental conditions dictate the use of permanent shade trees in tea, coffee or cocoa, large estates generally prefer to interplant with a single species. Such trees should be tall, produce a large and even canopy, be easy to prune, have feathery leaves to give a dappled shade and compete as little as possible with the main crop for soil water and nutrients. Species used as shade trees in South-East Asia are presented in Table 8. Most belong to the Leguminosae, including the very widely used lamtoro tree (Leucaena leucocephala (Lamk) de Wit), but there are also the non-leguminous Grevillea robusta A. Cunn. ex R. Br. and Cordia alliodora (Ruiz & Pavon) Oken (which also produces high quality timber). Cocoa and robusta coffee have been successfully planted under mature coconut palms in various parts of South-East Asia, as the gradually diminishing crowns of the tall palms provide just the right degree of shade (e.g. Ramadasan et al., 1978).

Coffee and cocoa can also form part of agroforestry systems. Much of the arabica and robusta coffee in South India is effectively grown in a multicropping system with pepper (Piper nigrum L.), and often also with some clove (Syzygium aromaticum (L.) Merrill & Perry) and mandarin (Citrus reticulata Blanco) trees, under a double shade canopy of Erythrina variegata L. and thinned natural forest or planted Grevillea robusta trees (Kurikanthimath et al., 1994). The combination of robusta coffee and pepper under shade is also found on Sumatra, Indonesia. Various studies have described the socio-economic and ecological advantages of smallholder coffee and cocoa production in agroforestry systems in Central America (Beer et al., 1998) and West Africa (Herzog, 1994). With few external inputs yields of coffee or cocoa will be modest but regular, while the added revenue from timber, firewood, fruits and other products will help to make farmers less financially vulnerable to fluctuations in world market prices.

Fertilizer use and mulching

The nutrient requirements and fertilizer recommendations are well established for tea, coffee and cocoa (de Geus, 1973).

For a leaf crop such as tea, the most important fertilizer nutrient is nitrogen. In unshaded tea there is a linear response to nitrogen up to a rate of 120 kg N per ha or more. Provided that other nutrients, especially P and K, and environmental conditions are not yield-limiting factors, responses are of the order of 4-8 kg processed tea per kg applied N. With regard to coffee, nitrogen and potassium are particularly required by the ripening crop. Adequate nitrogen is essential for optimum photosynthesis that has to provide the carbohydrates to the developing crop directly or through accumulated reserves. Under conditions of serious carbohydrate shortage, twigs may die back; applications of nitrogen and also of potassium have been found effective to control this physiological disorder. The low-yielding and shaded smallholders' cocoa has low nutrient requirements. When yield levels of 700 kg dry beans per ha or more have been reached with good management, a response to added P can be expected (Wessel, 1972, 1985). This is not the result of high P requirements but of the generally low availability of P in the soil. With little or no shade, cocoa will generally respond to N application, while with high yields additional K may also be needed.

The nutrient requirements of the three crops depend on light conditions and their production capacity, as determined by environmental and genetic factors.

Foliar analysis plays an important role in the assessment of fertilizer needs of perennial crops. Once critical nutrient concentrations and ratios in leaves have been established in fertilizer trials, these can be used to assess the fertilizer requirements in field plantings. By periodically sampling leaves and analyzing trends in nutrient concentrations, the effects of fertilizer use can be assessed and application rates be amended if needed. Apart from the external supply, nutrient levels are also influenced by the age and position of the leaf, light and other environmental conditions, and by cultivar. Therefore strict sampling procedures have to be followed. Although foliar analysis is widely used in tea and coffee, it is of little practical value for cocoa because large variations in light conditions occur within plantings and, as not all branches flush simultaneously, leaf age cannot be determined by leaf position. A practical problem is that the leaves to be sampled are mainly found in the upper canopy of 2-4 m tall trees.

Mulching generally helps perennial crops to establish. This is especially true in drier areas with light textured soils and a long cropping history. Mulch improves the root environment by reducing moisture losses and soil erosion, by lowering soil temperature and by improving the physical and chemical condition of the soil. The response to mulching of young, particularly unshaded coffee is often considerable (Mitchell, 1988). Mulching of young coffee is a widespread practice in Kenya, where cut elephant grass (Pennisetum purpureum Schumach.) is usually used at rates as high as 25 t dry matter per ha per year. Bottlenecks in this system are the need to set aside land for mulch production, the transport of large amounts of organic material and the labour involved. Applying a heavy mulch with a high C/N ratio tends to depress the availability of soil nitrogen, making additional nitrogen necessary. Application of mulch during the dry season increases the risk of fire.

Pruning

As the general principles of pruning of woody perennials have been dealt with in Prosea volume 2 "Edible fruits and nuts" (Verheij & Coronel, 1991), this section will highlight only the differences in pruning requirements of the three major stimulants.

It is essential to prune coffee. Over time, fruiting shifts from the inside to the edges of the shrubs, so they become too tall and the branches become too long for optimum production and management. Although checking the growth of plagiotropic branches by pruning can temporarily remedy this situation, periodic stumping is needed to rejuvenate the entire bush. This drastic replacement pruning leaves only the basal part of the main stem, encouraging dormant buds to grow out into new orthotropic shoots that later bear the plagiotropic branches. The prolific fruiting of coffee and the strong sink function of the fruits account for the positive yield response to pruning, while the strict dimorphism allows coffee to be repeatedly rejuvenated by periodic stumping.

In tea, repeated pruning of shoots and branches starts at an early stage, in order to achieve a bush of a spreading habit and of a convenient height for plucking. The harvesting of leaves is itself a form of pruning, in which a careful balance is sought between the leaves removed and the retained maintenance foliage. The removal of apical shoots by plucking encourages active growth of secondary shoots which are harvested later on. To keep the plant permanently in a productive vegetative stage and to keep the height of the plucking table within workable limits, periodic pruning is needed in which most of the foliage and some of the branches are cut. The severest form of pruning is collar pruning, in which the bush is cut near ground level. Tea's potential to recover fully from pruning explains why century-old plantings are still productive.

Cocoa seedling trees need little pruning. Once the jorquette (3-5 plagiotropic fan branches) is formed, only low and inward growing branches are cut and orthotropic shoots removed. For ease of harvesting some secondary plagiotropic branches in the centre of the tree are removed as well. More intensive pruning often reduces yield. Stimulating vegetative growth it shifts the balance between vegetative and generative development in favour of the former. Studies of the shaping and training of young plants budded with plagiotropic material have shown, however, that there are genetic differences in the responses to pruning. Some clones respond mainly with strong vegetative growth whereas others combine moderate vegetative growth with prolific early fruiting (Pang et al., 1994). This is in line with the view that favourable effects of pruning can only be expected in trees with a high fruiting capacity (Verheij & Coronel, 1991).

Crop protection

Large-scale planting of sole crops favours the development of diseases and pests. This applies also to tea, coffee and cocoa.

Disease control

The three crops suffer from diseases affecting roots, trunks and branches, foliage and fruits. Control techniques include cultural and chemical methods as well as the use of host resistance. At national and regional level, introduction and spread of diseases can be prevented by quarantine and health inspection. At field level, disease prevention starts with removal of sources of infection, e.g. the removal of roots and stumps infected by fungal root diseases and the sanitary harvest of infected fruits. Ensuring optimal growing conditions is an effective tool in limiting disease incidence, while deliberate pruning and shade management enables a microclimate to be maintained that is not conducive to the development of fungal diseases.

Chemical control is widely used in the three stimulants, but similar to current practice with insecticides, routine preventive spraying has given way to treatments applied in response to observed disease incidence. In certain cases, frequent spraying is essential, e.g. for the control of black pod disease (Phytophthora spp.) in cocoa under very humid conditions, where pods need a permanent protective coating of fungicide.

The use of resistant plant material is by far the best method of disease control. When after 1870 coffee leaf rust destroyed the arabica coffee in Asia, the only solution was to plant the resistant Coffea canephora. Since then, resistance breeding has progressed and has resulted in arabica cultivars that combine resistance to major coffee diseases with desired agronomic and quality traits (van der Vossen, 1985). Searching for disease resistance has been the major incentive for the collection of genetic material in the centres of diversity and for the establishment of germplasm collections.

Whereas on estates most diseases and pests can be effectively controlled, crop protection in extensive production systems often faces severe problems. Various factors play a role. Abandoned fields and isolated trees are sources of infection and breeding spots for pests. Furthermore, control treatments in usually small individual fields are only effective if neighbouring fields are treated as well. Another smallholder problem is that revenues of cash crops are used for direct household needs and not to purchase inputs.

Pest control

Although chemical control of certain pests has been effective, negative effects have also been observed, especially where broad spectrum insecticides have been applied and pests have proved to be less susceptible to treatments than their natural enemies.

A contrasting approach makes use of the stable environment of perennial crop fields with its diverse populations of pests and predators. Integrated Pest Management (IPM) is based on simultaneously applied methods of chemical, cultural and biological control to keep pest populations below the threshold of economic damage. IPM techniques, based on studying population dynamics of the insect pests with their natural enemies and establishing threshold values for insecticide treatments, were developed successfully for arabica coffee in East Africa (Bardner, 1985). In a leaf crop such as tea, cultural and biological control techniques are especially important to avoid insecticide residues in processed tea and negative effects on flavour, but IPM in tea is not yet well advanced.

An interesting example of a biological control system is the control of mirids (Helopeltis theivora) in cocoa in Indonesia and Malaysia, which dates back to the first decades of the 20th century (Giesberger, 1983). It involves the presence on twigs and pods of the black cocoa ant (Dolichorus thoracicus) which uses honeydew produced by the white cocoa mealybug (Cateanococcus hispidus) as its food source. The black cocoa ant disturbs Helopeltis adults to such an extent that feeding and reproduction are impeded. Black ants and mealybugs only thrive well in shaded conditions provided by shade trees or a well developed closed cocoa canopy. Ants can be introduced into fields by transferring nests and using tree branches and palm fronds as bridges between already colonized and not yet colonized cocoa trees.

Examples of recent developments in biological control in coffee include the use of an entomopathogenic fungus (Beauveria bassiana) and exotic parasitoids (e.g. Cephalomonia stephanoderis) against the coffee berry borer (Bheemaiah et al., 1996; Bustillo et al., 1995) and the use of pheromone-baited traps against the white stem borer (Hall et al., 1998).

Trends in crop production

High-density planting of dwarf trees is an important common trend for temperate fruit orchards and tropical tree crop plantations. Small trees are easier to maintain and to harvest, while smaller supporting frames have lower assimilate use for maintenance respiration, reduced competition from vegetative growth and require less pruning. Increased plant densities give higher yields per unit area. At high densities, individual trees have only limited crown space. To ensure adequate interception of light energy under these conditions there must be optimum penetration of light to leaves in the lower part of the canopy. With the successful development of compact-growing and disease-resistant cultivars, high density planting has become general practice in arabica coffee (Charrier & Eskes, 1997). Compact cultivars are not yet available in robusta coffee, but the first reports from India of really compact plant types selected in progeny of a cross between Coffea congensis and C. canephora are encouraging (Srinivasan, 1996).

Research in high-density clonal cocoa fields in Sabah, Malaysia, indicates that once the canopy has closed it is not light interception but light distribution within the canopy that becomes a yield-limiting factor (Yapp & Hadley, 1994). Vigorous clones at high density tended to produce strong, upward-growing fan branches, causing congestion of mutually shaded leaves in the top of the canopy and thus reducing incident light on the lower leaves to levels below the light compensation point. Less vigorous, semi-dwarf clones on the other hand had a crown architecture allowing light penetration in the canopy. They had lower light extinction coefficients and far out-yielded the vigorous clones. Despite these favourable results, the commercial advantage of planting these semi-dwarf clones is still doubtful (Pang et al., 1994). Contrary to the needs of plantations, the low-input smallholder field sector requires vigorous cultivars. During establishment they have to compete with weeds, and initially and often also later on with associated other crops.

The future of the smallholder cocoa in West Africa has been an important issue since the 1980s. Because the cocoa trees are getting old, very large areas are due for replanting. Although many farmers are technically capable of turning replanted areas into modern high-input cocoa holdings, they lack the resources to do this. An alternative approach is to diversify the replanted cocoa fields systematically in low-input agroforestry systems with timber and fruit trees in addition to cocoa. The fluctuating cocoa prices, the demand for timber and the shade tolerance of cocoa make the development of improved traditional multistrata systems a feasible option.

In tea, high-density planting is now common practice, and here too the importance of light distribution within the shrub has been recognized. China teas with semi-erect leaves have the potential of producing higher yields than Assam types (Hadfield, 1968). However, it is not only photosynthetic efficiency that plays a role, but also the fact that the semi-erect types produce a much higher number of shoots per unit area than the horizontal leaf types (Tanton, 1992).

Harvesting and processing

Harvesting

The plant parts harvested vary from young leaves to seeds, according to species (see Table 6), but manual harvesting is the predominant practice for all stimulant crops. Selective harvesting at the right stage of maturity is often essential for obtaining a product with the optimum composition of stimulant and other quality-determining components.

Manual harvesting is very labour-intensive for crops requiring regular harvesting rounds during several months of the year. In tea (Willson, 1992), coffee (Mitchell, 1988; Snoeck, 1988) and cocoa (Lass & Wood, 1985) the cost of harvesting alone may account for 30-50 % of total field production costs. Saving on harvesting costs by mechanization can have a direct bearing on profitability of larger estates, but not so much in the case of smallholders where the family provides most of the labour. Cocoa does not lend itself to mechanization of harvesting operations, except possibly the opening of pods after removal from the trees. However, in tea and coffee mechanical harvesting with hand-operated, tractor-mounted or self-propelled machines has found application in a few countries, where labour is in short supply or increasingly expensive and where demands for a quality product are less stringent. Examples are coffee in Brazil and tea in Japan (green tea) and Argentina. On the other hand, more than 80% of all coffee in the world is produced by smallholders, and most of the plantation tea is grown in countries where labour is abundant and still relatively inexpensive. The production of tea, coffee and cocoa will therefore continue to depend mostly on manual harvesting, just as in the case of all other stimulant crops.

Post-harvest handling and processing

For many stimulants, post-harvest handling involves little more than cleaning and drying to a lower moisture content to improve storability. A few stimulant crops are subjected to a more extensive post-harvest treatment. For instance, guaraná seeds are roasted and ground to a paste, maté leaves are heated in a furnace before further drying and grinding and areca nuts, when not consumed fresh, are sliced and boiled before drying for the purpose of preservation.

Tobacco leaves are subjected to various curing processes, from quick drying over hot air or open fire to slower sun- and air-drying, depending on whether they will be used as cigarette, pipe or cigar tobaccos. These curing processes demand technical skill and expertise, but otherwise require relatively modest and inexpensive facilities. However, the manufacture of cigarettes in particular involves a large industrial set-up with sophisticated machinery.

Curing, processing and manufacturing of tea, coffee and cocoa

The three major stimulants have to undergo several stages of curing and processing before the final products can be offered to the consumers (Hampton, 1992; Wood, 1985b, 1985c; Wrigley, 1988). Most of these are very specific to the particular commodity, but a comparative analysis as presented in Figure 3 indicates considerable analogy in a number of curing and processing stages. Green tea and unwashed coffee have been left out for reasons of simplicity, but will be referred to in the text below.

The harvested products of tea and coffee should undergo the first step of the curing process, withering for tea and pulping in the case of coffee, on the same day to avoid loss of quality. However, for cocoa an interval of 2-3 days between harvesting and opening of the pods to remove the beans may even have a beneficial effect on the fermentation process.

Fermentation in coffee and cocoa involves degradation of the mucilage surrounding the beans by the activity of naturally occurring micro-organisms, which takes about one day for coffee and 4-6 days for cocoa. Proper fermentation in cocoa and fermentation of coffee in combination with subsequent washing and soaking in water is essential to obtain a high quality product. Fermentation in tea is actually a process of enzymatic oxidation of the polyphenols, which results in the formation of several constituents giving black tea its specific flavour and brown colour. The whole curing process for tea lasts less than two days, including forced drying to about 3% moisture content (this drying process is called firing) and packing as graded tea. The same process lasts about 10-14 days for coffee and cocoa, partly because of the need for gradual drying in the sun or by artificial dryers to a moisture content of about 10% for coffee and 6% for cocoa beans. In the case of coffee there is a second stage of curing, lasting a couple of days and usually carried out in large coffee mills, to convert the parchment coffee into cleaned and graded "green" coffee.

In the curing process of green tea the leaves are fired or steamed, which destroys the enzymes and so prevents oxidation of the polyphenols. Green teas are therefore more astringent and have a different flavour. Dry-processed coffee, dried as whole fruits without prior pulping and fermentation, is called unwashed or cherry coffee. The quality is almost always inferior to washed coffees.

The curing of tea, coffee and cocoa is always carried out in the producer countries within short distance from the production fields. Further processing and manufacturing into the final products generally occurs close to main centres of consumption.

In the case of tea, the conversion from cured to final products is a relatively simple process, involving mainly blending and retail packing, which does not contribute to large increases in added value of the final product.

In contrast to tea, coffee and cocoa are largely export products (Table 3). The processing into final products requires substantial industrial investments and takes place mostly in the industrialized consumer countries. The added value of final products of coffee and cocoa is generally several times that of the imported cured product. Brazil and a few other coffee-producing countries have developed a sizable capacity for the production and export of instant coffee, but limited shelf-life prevents the development of an export-oriented local industry for roast coffee. In cocoa the situation has changed considerably during the last decade with the development of a cocoa grinding industry producing intermediate products such as cocoa liquor, butter fat and powder. At present about 30% of all cocoa produced is processed in that manner by local cocoa grinders before export to consumer countries.

Genetic resources and breeding

Genetic resources

Access to substantial genetic variability is essential to successful crop improvement. Germplasm collection, conservation and characterization is actively pursued in the four economically most important stimulants: tea, coffee, cocoa and tobacco. Other stimulants whose genetic resources are receiving some attention from national research centres are maté, guaraná, kava, areca palm, betel pepper and kola.

Tea

The germplasm collections present in the main tea research centres in Asia and Africa have generally provided sufficient genetic variation for considerable selection progress in yield and other agronomic characters, except host resistance to important diseases and pests. Recent collections of "wild" plants of Camellia sinensis var. sinensis in the Yunnan province in China have yielded very interesting genotypes. Systematic collection in major centres of genetic diversity (Assam - Burma (Myanmar) area and southern China) should receive high priority to preserve tea genetic resources for future use (Banerjee, 1992a).

Coffee

The first systematic efforts to collect Coffea arabica germplasm by a Food and Agriculture Organization of the United Nations (FAO) mission to Ethiopia in 1964 had the real intention of international collaboration and the resulting accessions were distributed worldwide to all coffee research centres (Meyer et al., 1968). The "Institut de recherche pour le développement" (IRD, formerly ORSTOM) made several collecting expeditions in the period between 1966 and 1985 to all important centres of genetic diversity for several Coffea species in Africa and most of those accessions are maintained as field collections in Ivory Coast, Cameroon and Madagascar (Berthaud & Charrier, 1988). There are also large collections of C. arabica germplasm in Ethiopia (Bellachew, 1998). Nevertheless, the free exchange of coffee germplasm is hampered by the fact that none of the collections have an international status and the International Plant Genetic Resources Institute (IPGRI) has not yet been given the opportunity to develop an effective network of collaboration on coffee genetic resources (van der Vossen, 1997).

Cocoa

There is a long and intensive history of germplasm collection from the centres of genetic diversity in the Amazon basin, commencing in 1937 with the expeditions led by F.J. Pound of Trinidad. At present, there are two large international cocoa germplasm collections in Trinidad (CRU) and in Costa Rica (CATIE) and several national collections in cocoa research centres elsewhere (End et al., 1992). The IPGRI plays an important role in promoting an international network for the optimal utilization and conservation of cocoa genetic resources (Eskes et al., 1998).

Ex situ germplasm conservation for tea, coffee and cocoa

Conservation in seed banks has not (yet) been possible for these tree crops, so instead there is reliance on expensive field collections. Cryo-preservation of seeds, embryos or meristems may eventually become a practical method of germplasm conservation, and progress in that direction has been reported recently for coffee (Dussert et al., 1998). Attempts to conserve genetic resources in situ, such as once proposed for C. arabica in Ethiopia, usually have to be abandoned under pressure from advancing deforestation and food crop production.

Tobacco

The United States Department of Agriculture (USDA) in particular, but also research centres elsewhere in the world, have made extensive collections of wild and domesticated tobacco germplasm. The exploitation of genetic diversity present in diploid Nicotiana species is also facilitated by the relative ease of introgression of useful genes into N. tabacum by interspecific hybridization (Wernsman & Rufty, 1988). The extreme longevity and small size of the seeds also allows relatively inexpensive and long-term conservation of tobacco germplasm in seed banks.

Breeding

The general objective of plant breeding is to develop crop cultivars with the potential of providing maximum economic benefits to the growers. This usually requires simultaneous selection for plant type and vigour, yield, quality and other characters. If diseases and pests have become a threat to the profitability or even survival of the crop, resistance to these may assume the highest priority in breeding. Selection progress depends on the breeding plan applied, which in turn is largely influenced by the species' life cycle (annual or perennial) and mating system (self- or cross-pollinating). Cross-pollinating species or outbreeders have in common that they are highly heterozygous and often intolerant of strong inbreeding (Simmonds, 1981, 1992).

Practically all woody perennials are outbreeders, including tea, all diploid (e.g. robusta and liberica) coffees, cocoa, kola, areca palm, maté and guaraná. The only exception is the allotetraploid arabica coffee, which is a self-pollinating species just as the annual tobacco.

There are reports on efforts of mass selection for higher yield in kola, maté, guaraná, areca palm and betel pepper, but systematic breeding programmes have been implemented only in the four main stimulants (tea, coffee, cocoa and tobacco). In the three tree crop stimulants there are interesting analogies in the opportunities and limitations of crop improvement that justify further elaboration below.

Early breeding in tea, coffee and cocoa

The long productive life-span of these tree crops and high costs of replanting have been main obstacles to quick genetic improvement. Consequently, most of the production of tea, coffee and cocoa is still based on cultivars released five to eight decades ago from relatively simple breeding and selection programmes.

The majority of the cultivars of the outbreeding tea, robusta coffee and cocoa are produced as open-pollinated seeds from phenotypically selected seedlings (open-pollinated varieties) and from poly- or biclonal gardens (composite or synthetic varieties). Clones offer the opportunity of instant fixation of superior genotypes and high degree of plant uniformity, but the logistics of mass propagation and distribution are often too complex for smallholder production systems, which dominate in coffee and cocoa. Clonal cultivars are also more widely used in tea than in robusta coffee and cocoa, largely because the propagation of tea cuttings is relatively simple and inexpensive. Cultivars of the inbreeding arabica coffee are almost always propagated from seed, as they are true-breeding lines developed from single plant selections from growers' fields or from progenies of simple crosses and backcrosses.

Yield, quality and plant vigour have always been dominant selection criteria in tea (Visser, 1969) and robusta coffee (Charrier & Berthaud, 1988), but in some early breeding programmes disease resistance was already given high priority in cocoa (witches' broom in Trinidad, swollen shoot in West Africa) and arabica coffee (leaf rust in India) to ensure long-term survival of the crop (Srinivasan, 1996; Toxopeus, 1969).

Recent advances in tea, coffee and cocoa breeding

Breeding programmes with systematic crossing designs and statistically laid out field trials of the last 30 years, in coffee and cocoa in particular, have provided opportunities for biometrical genetic analyses of yield and other agronomic characters. There is considerable evidence that the genetic variance for almost all components of yield, quality and other quantitative traits is predominantly due to additive gene effects. This should facilitate the estimation of parental breeding values and speed up selection. Hybrid vigour for yield noticed in crosses between parents of different origin appears to be the result of the accumulation of additive effects of polygenes dispersed over sub-populations. Some breeding programmes in robusta coffee (Charrier & Eskes, 1997) and cocoa (Eskes & Lanaud, 1997; Lockwood & Pang, 1994) have already adopted methods of reciprocal recurrent selection with distinct sub-populations to increase chances of producing genotypes superior in yield, quality and other important traits. However, resistance to important diseases has become a breeding objective of the highest priority in many cocoa-producing countries. There appears to be considerable scope for achieving satisfactory levels of resistance to one or more diseases according to regional priorities, notably in witches' broom, black pod, vascular-streak dieback and swollen shoot diseases. A prerequisite is the availability of efficient pre-selection tests for host resistance, and sometimes pre-breeding (recombination crosses) may be necessary to improve the resistance levels of breeding parents before these are integrated into the main breeding programme (van der Vossen, 1996).

In tea there is a general lack of biometrical data on the inheritance of yield and quality components, but hybrid vigour for yield in inter-population (e.g. China and Assam teas) crosses has been well documented. So far, no progress has been made in breeding for resistance to blister blight disease in the absence of suitable sources of host resistance (Banerjee, 1992b).

In arabica coffee, disease resistance continues to be a breeding objective of the highest priority. Efforts to obtain durable resistance to coffee leaf rust have had a long history of initial successes followed by disappointments because of repeated appearance of new virulent races of the rust fungus, but some lines of the cultivar Catimor have shown complete resistance so far (Bettencourt & Rodrigues, 1988). These results were obtained by breeding plans normally applied to self-pollinating crops, including recombination crosses followed by back crossing, inbreeding and line selection. A similar plan was initially also applied in a breeding programme in Kenya to obtain resistance to coffee berry disease. The resistance turned out to be controlled by a few major genes but was nevertheless also persistent. The switch to breeding strategies to produce F₁ hybrid (seed) cultivars instead of clones or true breeding lines was partly inspired by the confirmation of transgressive hybrid vigour in genetically divergent crosses in arabica coffee (Walyaro, 1983). Other advantages were the chances of earlier introduction of cultivars with resistance to both coffee berry disease and leaf rust, as well as several other desirable agronomic characters (van der Vossen, 1985). Interspecific hybridization has played a significant role only in coffee. Examples are crosses between C. arabica and C. canephora, with the objective of introgressing disease resistance into arabica coffees, e.g. the cultivar Icatu in Brazil (Carvalho, 1988), or improved liquor quality into robusta coffees, e.g. the cultivar Arabusta in Ivory Coast (Cambrony, 1988). Examples of interspecific hybridization leading to successful cultivars in arabica and in robusta coffee can also be found in India (Srinivasan, 1996).

Opportunities of molecular breeding in tea, coffee and cocoa

In the past decade in particular, plant biotechnology has evolved into an applied science providing powerful additional tools for plant breeders, enabling them to increase efficiency and take new approaches to hitherto unattainable objectives.

There are basically two main applications of plant biotechnology in crop improvement programmes: molecular markers and transgenic plants.

Molecular marker technology in plants has found applications in all phases of plant breeding programmes, including:

1. germplasm management,
2. measuring genetic divergence of sub-populations (e.g. to predict heterosis),
3. gene introgression (possible reduction of number of backcross generations),
4. marker assisted selection (MAS)
5. and cultivar identification and purity testing.

MAS enables early selection of a trait by selecting for a molecular marker closely linked to the gene(s) controlling the trait. For characters inherited through one or few major genes (e.g. disease resistance) closely linked markers have been identified. MAS can be particularly effective in the accumulation of resistance genes in one genotype to raise the level of resistance and increase chances of durability (gene pyramiding) and the simultaneous selection for different disease resistances. Where the trait is controlled by polygenes and the heritability is rather low, a more complex quantitative trait loci (QTL) analysis is required to determine the location on the chromosomes of markers linked to the polygenes controlling the trait. A prerequisite for such a QTL analysis is the availability of a saturated genetic linkage map (Mohan et al., 1997; Stam, 1994).

Molecular markers have been applied to germplasm characterization and identification of genetic diversity between sub-populations in tea (Paul et al., 1997), coffee (Lashermes et al., 1996) and cocoa (Lerceteau et al., 1997). The first genetic linkage maps have already been constructed for coffee (Paillard et al., 1996) and cocoa (Lanaud et al., 1995). Reports on the potential application of MAS have appeared for cocoa (Crouzillat et al., 1996), notably a QTL analysis for black pod resistance, and for arabica coffee (Agwanda et al., 1997) on molecular markers linked to one of the major genes determining resistance to coffee berry disease.

Generally, successful genetic transformation is still limited to characteristics controlled by single major genes for which gene isolation and transfer is relatively easy (e.g. herbicide, pest and disease resistances). Lack of adequate legislation for proprietary rights and biosafety in many countries is still an obstacle to the introduction and unrestricted cultivation of transgenic crop varieties. Some governments are reluctant to act, under pressure from a rather negative public perception of biotechnology in general.

Techniques of in vitro regeneration of plants (somatic embryogenesis) and genetic transformation are still in the initial stages of development in cocoa (Eskes & Lanaud, 1997), but by now these are well-established in coffee (Charrier & Eskes, 1997). Transgenic coffee plants have been produced already, e.g. with insect resistance (based on Bt genes) and with caffeine-free beans (inhibition of caffeine accumulation by anti-sense gene expression). Nevertheless, for the aforementioned reasons it is uncertain whether transgenic coffee cultivars will be widely grown within the next one or two decades.

Prospects

Supply and demand

The total demand for the universally popular stimulants tea, coffee and cocoa will continue to grow steadily, if only because of an ever increasing world population. However, due to accelerated expansion of traditional and new production areas over the last 15 years, especially for coffee (e.g. robusta coffee in Brazil and Vietnam) and cocoa (e.g. Indonesia and Ivory Coast), supplies are often in excess of demand, causing declining world market prices and consequent loss of revenues for growers and producing countries. The situation appears more stable in the case of tea due to the buffering effect of huge and still growing internal markets for this stimulant in the most important producing countries (India, China and Indonesia).

Past international coffee and cocoa agreements between producing and consuming countries, aimed at regulating supplies and stabilizing prices, have not usually endured for long periods of time, more often than not because of economic or political conflicts of interest. Recurring abiotic (frost or drought) and biotic (diseases and pests) disasters in major coffee and cocoa producing countries have been important factors contributing to a temporary restoration of the balance of supply and demand and accompanying recovery in prices. There are no indications that these large fluctuations in supplies and market prices for coffee and cocoa will diminish in the short to medium term.

The tobacco market is generally less volatile, partly because the annual nature of the crop allows quicker response to changes in demand and supply. Public awareness of health risks of tobacco smoking is still small outside Europe and North America and it has certainly not had any impact so far on the steadily increasing tobacco markets in Asia.

The areca nut in combination with betel pepper will continue to be an economically important stimulant in South and South-East Asia, but the demand may gradually decrease as its popularity with the younger generations is waning. The exotic kola, maté and guaraná are unlikely to become significant stimulants in Asia.

Research

The scope for further enhancing economic returns of tea, coffee and cocoa production, through genetic, crop physiological and agronomic research, is still considerable. Annual yields of 3 t per ha in tea and 4-5 t per ha in coffee and cocoa have already been realized experimentally and even higher yields are envisaged in the longer term by applying molecular marker techniques to exploit available genetic resources still more efficiently. Molecular biology appears to becoming a powerful tool too for achieving host resistance to crop-threatening diseases and pests that have so far eluded all efforts of plant breeders. Such resistance, sometimes in combination with biological means of control, would contribute much to reducing production costs and environmental hazards of pesticide use. Blister blight and mites in tea, leaf rust and berry borer in coffee, black pod and pod borer in cocoa are just a few examples of major crop-limiting biotic factors.

Nevertheless, the impact of innovative genetic and agronomic advances is almost always much greater in commercial plantations. Tea is mostly an estate crop, whereas more than 80% of the world coffee and cocoa is produced by small farmers, who do not usually have easy access to improved agricultural inputs to sustain economic yields. After the first few years of production - when the new planting may benefit from the high organic matter content and mineral reserves of recent forest clearings, as well as from relative absence of diseases and pests - yields decline generally to less than one quarter of that of high-input coffee or cocoa plantings. In actual fact, the tremendous growth in the world supply of coffee and more particular of cocoa has been much less the result of increased productivity than of area expansion, often at the expense of tropical forests. Research priorities in coffee and cocoa should, therefore, also include the development of more productive and sustainable smallholder production systems, which resemble forest ecosystems and help to arrest further destruction of natural resources. Of course, successful farmers' adoption of such socio-economically viable and environment-friendly production systems will very much depend on the political determination to provide effective extension services, good infrastructure and easy access to essential inputs.