The Natural and Economic History of the Coconut in Zanzibar

The Palm Enthusiast (Journal of the South African Palm Society)

Vol 19 No. 2 pp. 7-20 (2002)

E. Krain¹, J.A. Issa², A. Kullaya³, M. Schuiling⁴ & H.C. Harries^5*

ABSTRACT

Natural stands of coconuts encouraged human settlement of islands like Zanzibar by providing fresh, sweet, uncontaminated drinking water. Coir cord, twisted from coconut husk fibres, stitched together the hulls of Indian Ocean boats. Decorated coconut shells from the region reached Europe long before the Portuguese sailed to Zanzibar. Four centuries later coconut oil from copra, the dried kernel of the nut, became a major source of income in Zanzibar and desirable varieties were taken as far afield as the Philippines by 1912. Yet this was overlooked when similar varieties were reintroduced from West Africa in 1980. Here is the history behind these fascinating facts.

Keywords: Cocos nucifera; coconut varieties; copra; coconut oil; Tanzania; East Asfrica.

INTRODUCTION

Coconut is a food and cash crop in Zanzibar as it is on the coast of the mainland of East Africa and elsewhere in the humid tropics. It became the second most important foreign exchange earner in Zanzibar, after cloves, and it held this position for a long time. The coconut palm is the ‘Tree of Life’, because of its many different uses. These include: coconut sap (toddy) as a source of sugar, vinegar or alcohol; coconut water as a delicious, non-alcoholic beverage and substitute for blood plasma in emergency surgical operations; coconut oil for cooking, as milk and cream, and for making soap, as diesel fuel, for lighting, for making candles; coir fibre for ropes and mats, cocopeat for horticulture; shell for buttons, decorative carving, burnt as a fuel and for charcoal; wood from the coconut stem for furniture and construction purposes; leaves for decoration and as a thatching material; finally the heart of the palm as vegetable salad (millionaire's salad).

The coconut palm has long been an integral part of the social and cultural life of the people of Zanzibar. Coconut palms identify property land boundaries. In many rural parts of Zanzibar the tradition continues to this day to plant a coconut seedling over the placenta when a child is born. The coconut palm has the same life expectancy as a human being and thus accompanies the new individual from the cradle to the grave. This is one reason why it is very difficult to persuade farmers to cut down under-productive, over-aged coconut palms.

The value of coconuts is known and appreciated in Zanzibar. Many rulers and governments in Zanzibar and East Africa tried to foster its development. The most recent project is the National Coconut Development Programme (NCDP). It is a development programme run in the coastal areas of the Tanzanian mainland and on Zanzibar. It focuses on producing improved coconut planting material, controlling pests and diseases, testing and adapting husbandry techniques, and advising coconut farmers.

During the operations of NCDP it became necessary to look into the history of the palm. To begin with, the lethal coconut disease on the mainland was especially rampant in the south of Tanzania. The plantations there mostly date from the period of German colonial government. In contrast, the disease occurs to a much lesser extent in the north, the traditional area of coconut growing since the earliest Arab/Omani/Shirazi times. Literature surveys and inquiries with old and experienced farmers indicated the coconut planting dates in the various areas and the source of seed. This knowledge may help to explain the different degrees of lethal disease susceptibility within the local palm population[⁶]. In addition, some of the varieties that NCDP imported for hybrid seed production are very similar to local coconut varieties. If historical evidence establishes a co-identity between local and exotic varieties there will be more confidence in their use[⁷,⁸].

This paper summarises the findings made so far, on the history of the coconut palm and the coconut industry in Zanzibar.

HISTORY

Names and Descriptions of Coconut Varieties

There is only one species of Cocos nucifera L but many varieties. The simplest classification is on habit, which recognises tall and dwarf forms. Another system based on habit, precocity and fruit size distinguishes typica, javanica, aurantiaca and nana .By using a whole range of characteristics, from fruit components to disease reaction, the varieties resolve into predominately wild or domestic types and a multitude of intermediates originating by introgressive hybridisation[⁹].

The oval shape of the fruit of the East African Tall coconut variety is a visual indication of a higher proportion of husk and shell, low water and thick endosperm. This leads to the suggestion that the East African Tall is predominantly a wild type. Originally, the coconut palms in East Africa were the wild type tall. However, at some time domestic type tall material was introduced. Most of today's coconut palm population is predominantly wild type tall with some introgression from domestic tall types. This is demonstrated by studies of germination rates for the Tanzanian Tall[¹⁰] and fruit component analysis and other characteristics for the Moçambique Tall[¹¹]. Dwarf varieties, which are highly domesticated, need to be cultivated to survive.

In Zanzibar, there are two distinct local palm populations. One is the local tall, called East African Tall and the other a dwarf, called Pemba Red Dwarf or simply Pemba Dwarf.

The local tall variety grows to a height of 20-30 m, is slow maturing, first flowering within 6-10 years of planting, bearing comparatively few nuts, producing good copra and generally being tolerant to marginal growing conditions. The local dwarf variety reaches a height of 8-12 m, is earlier to mature, first flowering after 3-4 years, bearing many fruit with good drinking (madafu) qualities. Tall varieties may live for 60-100 years while dwarf varieties may only achieve 40-60 years.

Tall varieties, and the East African Tall in particular, are very heterogeneous because they are generally, but not always, cross-pollinated. In contrast, dwarf varieties are highly, but not always, self-pollinated. The East African Tall shows a wide, but not exhaustive, range, of fruit colours and shapes. The colours range from a light yellow to red, brown or green, though most of the East African Tall palms bear brown or green fruit since these colours are dominant to red or yellow. Most of the fruit are oval shaped. The wide variation within the East African Tall shows it is a population with numerous sub-types.

Pemba Dwarf palms have a close resemblance to certain exotic varieties, such as the Cameroon Red Dwarf (CRD) from West Africa, the Madang Red Dwarf from Papua New Guinea, and others. The trunk is very short and the leaves are comparatively short and very erect. The fruit are oval-shaped and the fruit colour is a light yellowish orange (golden). The bunch stalks are very long, holding the fruit of all ages at some distance. The similarity to other varieties was not realised until after the Cameroon Red Dwarf had been introduced for hybrid seed production purposes. Most Pemba Dwarf palms are the CRD-sub-type. Recent investigation showed that about 7% of Pemba Dwarf, show characteristics of another exotic variety. This is the MRD-sub-type, with some resemblance to the Malayan Red Dwarf (MRD). It is significantly taller than the CRD-sub-type and the leaves are longer and curve down. The fruit are round-shaped, the fruit colour is more reddish (apricot), and the bunch stalks are shorter[¹²].

Records show that an introduction of dwarf varieties from St Lucia was not successful, since seednuts did not germinate after arrival[¹³]. An introduction of 1,000 coconut seed from Malaya was made in 1955, purchased by the Copra Board[¹⁴]. Although the seed were called ‘dwarf coconuts variety “King” ex Malaya’ according to many reports [¹⁵] they were in fact yellow, red and green forms of the ‘ordinary’ Malayan Dwarf. The original plot planted was finally uprooted in about 1986 due to pencilling, wilting and death of many palms, but progenies had already been established at Kizimbani[¹⁶]. Malayan red, yellow and green dwarf seen in some government plots come from here. A fourth variety in these plots, resembling Cameroon Red Dwarf, are now considered to be local Kitamli from a one acre field planted in 1954 at Chukwani (south of Zanzibar town) [¹⁷].

The most recent major importation of varieties, made by the National Coconut Development Programme between 1980 and 1983, largely came from a germplasm collection held in the Ivory Coast. These included the nine dwarf forms, 14 tall and 22 different hybrid varieties, most of which are dwarf x tall crosses. The original palms grow on NCDP breeding sites and their progenies are on farmers' land.

People in Zanzibar readily distinguish between local tall (mnazi mrefu) and dwarf varieties (mnazi mfupi). The tall varieties are often further differentiated according to the colour of the fruit. The mnazi mweusi is the green coconut (although mweusi literally means black). The mnazi mwekundu is the brown (or red) coconut. The mnazi mweupe is the light yellow coconut (mweupe literally means white).

The mnazi mweupe is often also called kipemba or the mnazi wa Pemba, but this only in Zanzibar where it refers to a semi-tall. There are persistent reports of a King Coconut variety that occurs in Zanzibar (see below), and that it is supposedly either identical, or looks similar to the mnazi mweupe.

The Pemba Dwarf is Kitamli in Zanzibar, on the coast of Kenya and in the Tanga Region. In the latter area it is also Kisamli. Elsewhere, in most parts of the Tanzanian mainland south of Bagamoyo, including Mafia Island the name is Kipemba.

The terms Mnazi wa Unguja or Mnazi wa Bahari as mentioned by Voeltzkow [¹⁸], or African Nut and Indian Nut mentioned by Craster [¹⁹] are not in current use.

In East Africa, coconuts grow predominantly along the shore of the Indian Ocean and on the off-shore islands. The coastal strip from Somalia, Kenya, through Tanzania to Moçambique is narrower in the north, broader in the south and about 150km wide in Tanzania. Coconut palms also grow along the shores of Lake Nyassa and Lake Victoria and inland at places like Mbeya and Tabora.

The Pemba Dwarf occurs everywhere along the Tanzanian and Kenyan coastal coconut growing area, probably also in Moçambique[²⁰] and Madagascar[²¹]. It even reached the Philippines[²², ²³]. The Pemba Dwarf occurs in the Comoros and Mauritius under the same name. Very many grow in Zanzibar. Since the dwarf needs better soils, more rainfall and good husbandry, it predominantly grows around houses, where it receives more attention. Estimates suggest that dwarf varieties account for about 1.2% of the total palm population in Zanzibar[²⁴]. The percentage of dwarf varieties is about five to ten times higher in Zanzibar than on the mainland.

The larger number of local dwarf palms in Zanzibar could be due to a combination of factors. It may be the original site of introduction. Coconuts have greater agricultural and economic importance on the islands. The climate and growing conditions are more suitable for coconut cultivation and more survive than on the mainland. Pemba is better than Unguja.

It seems that Pemba has been the primary source of local dwarf palms for the coast of Tanzania and Kenya. The name of the dwarf shows this and it is Kipemba in many parts of the East African mainland. On Zanzibar the name Kipemba may be applied to another type of coconut that is a yellow coloured tall or semi-tall rather than a red coloured dwarf (see below). Colours and plant habit used to identify coconut varieties are usually subjective descriptions, not based on any measured or standardised criteria.

Earlier Names and Descriptions of Coconut Varieties

There seem to be no descriptions of coconut varieties in Zanzibar or East Africa until after the turn of the 19th century. Attempts to find descriptions of different coconut varieties in East Africa and Zanzibar has yielded the following:

Baumann (1896, 1897, 1899)

As far as the tall coconut palms are concerned Baumann reports that they were perhaps introduced by the Wadebuli[²⁵]. Writing at a time when copra had recently become the most important product of the coconut palm, Baumann discounts the drinking qualities of the dwarf variety by saying: ‘This variety [the Pemba coconut] has no value as a copra palm . . . and is cultivated for drinking nut production only’ [²⁶]. Baumann also reports that he had seen the Pemba coconut palms in Mafia[²⁷] . Baumann summarised his findings thus:

The coconut palm[²⁸] is planted in large numbers and provides a high yield of copra. There is one special variety on Pemba, which has been disseminated from this island to Zanzibar an the mainland, it is called Nazi ya Pemba a short variety with short leaves and yellow-red nuts containing a very delicious juice palm[²⁹].

Lyne (1905)

Lyne, who was the first Director of the newly formed Department of Agriculture in 1895, reported that.

Several varieties of coconuts are cultivated, which may be distinguished by the colour of the nuts, some of which are brown, some green, others a rich cream colour, but no attention is paid to their selection. A small species, called Pemba cocoanut (though as it is common in both Zanzibar and Pemba the reason of the name is not obvious), is grown only for the milk of the young nuts, which provides a very refreshing drink[³⁰]

Stuhlmann (1909)

Stuhlmann refers to three different tall varieties, but he does not describe them:

According to the shape of the nuts there are at least three varieties known of the common Cocos. Their exact description has still to be done Additionally there is the mnazi ya Pemba, a shorter variety, which normally does not have a bole, with egg-shaped, orange-yellow nuts which are only used as drink nuts. This seems to be the variety which is known as Tembili, or King Coconut, coming from Ceylon. Also in Zanzibar there occurs rarely and as a curiosity a dwarf, known as mnazi ya kitamli[³¹].

Craster (1913)

The natives recognize three different kinds of coconut: the Pemba nut, the African nut, and the Indian nut. The Pemba cocoa-nut palm grows to a height of sixty or eighty feet, and bears nuts with light brown, oblong-shaped husks. The African palm grows to one hundred feet or more, and the nuts have very thick, green, oblong-shaped husks. The kernel and milk of the African nut are not so sweet as those of the other two. The Indian palm does not grow as tall as the others, and begins to bear much earlier - after about six years: the husk of the nut is round in shape, yellow and thin, and the kernel is very sweet[³²].

Voeltzkow (1920) [³³]

The natives differentiate between three different types of coconut palms; the common Zanzibar Palm Mnazi wa Unguja with bright yellow inflorescences and nuts of the same colour; the Mnazi wa Bahari (the sea palm) with very big green nuts and inflorescences, and finally the Mnazi wa Pemba which can be recognized by its short growth, the very yellow midribs of the leaves, many dark-brown yellow inflorescences with egg-shaped beautiful orange-yellow nuts, which very often mature only three to four meters above the ground. This type is often cultivated near the homes and along the roads. It bears nuts only after five years, but which normally are only used as drinking nuts. This type can be found not only in Pemba but also in Zanzibar (Unguja) and along the coast. Occasionally it is also called Indian Coconut and might perhaps as Stuhlmann[³⁴] thinks, be identical with a variety grown in Ceylon which is called Tembili or King Coconut. Also according to Stuhlmann there can occur a special dwarf type which is called Mnazi wa Kitamli[³⁵].

Pearce (1920)

There are two varieties of coco-nut palms grown in Zanzibar. The first is the ordinary species, while the second is a diminutive variety known as the Pemba coco-nut. This latter palm is very much smaller than the ordinary species, and with its clusters of gold-coloured nuts has a most pleasing and graceful appearance. It is planted to mark boundaries, and its milk is esteemed for drinking[³⁶].

Tidbury, in 1942, enumerates the Mnazi wa Kiswahili, the Mnazi wa Pemba, the Mnazi wa Kitamli and the Mnazi wa Kifunzi[³⁷]. He explains that the first two varieties are used for copra making, cooking, thatching, firewood, brooms etc. while the third one only for drinking. One of his staff told him that the last one occurs in Pemba having very small nuts that are very sweet[³⁸].

In his book on the useful and ornamental plants in Zanzibar and Pemba[³⁹], Williams includes a photograph of a local tall and dwarf. From the picture, the dwarf can clearly be identified as a CRD-sub-type Pemba Dwarf.

In 1961 there was highly interesting correspondence between Child and Selby[⁴⁰]; Child was researching the first edition of the standard English coconut textbook (q.v.); Selby was Director of the Department of Agriculture. Child, who was looking for information about coconut varieties, quotes O.W. Barret[⁴¹] who wrote.

PEMBA - This variety seems to be native to the Island of Pemba in the British Protectorate of Zanzibar . . . The tree is comparatively small, though the writer has noted 15 meter specimen in the Zambezi Valley. It is exceedingly prolific. The husk is of a distinctly pale yellow colour with a more or less shining surface and without the prominent angles of the common sorts; the husk is comparatively thin, while the meat itself is of ordinary thickness; in other words, the percentage of meat is very high in the total weight of the nut.

Child continues ‘I do not recall having seen any other reference to this, though in Mozambique a dwarf palm is sometimes referred to as a Zanzibar variety.’ Child requests information on this and Selby responds

This coconut may be one called Kitamli in Ki-Swahili. Some people also refer to it as the Pemba coconut. It corresponds to the description given in your letter in some particulars but not in others. Thus it is ‘comparatively small’ (though not ‘exceedingly’ so). The nut is yellow (though hardly ‘pale yellow’, in fact it is slightly orange-yellow). It has ‘a shiny surface without the prominent angles of the common sorts’. The husk is ‘comparatively thin’. But the meat is not of ordinary thickness - in fact it is almost completely lacking and the kitamli variety is used exclusively for drinking.

There is, however, another coconut which is known in Ki-Swahili as the razi ya kiPemba or Pemba coconut[⁴²]. This also is a yellow coconut, paler in colour than the kitamli. The nut is triangular with the normal prominent angles. Medium size (9½ x 6½ inches, compared with 11 x 7 inches of the bigger varieties), the nut has flesh of average thickness (0.5") and mesocarp (fibrous part) of normal or slightly greater than normal thickness (1.2" compared with 1.1" or 1.0" of the commoner varieties). The tree is of a fair height - probably up to 60 feet but not many trees have been noted and this characteristic is not very certain. It is doubtful, however, whether it can be called ‘comparatively small'.’

Contemporary Sources [⁴³]

Herz-Schweizer assumed that the local dwarf ‘arrived during the increased trade movements between Zanzibar and India, Sri Lanka and the Far East - after the Oman Arabs declared Zanzibar a Sultanate of their own in 1848’[⁴⁴], but she did not provide any supporting reference. She was the first to describe the Pemba Dwarf as closely resembling the Cameroon Red Dwarf[⁴⁵]. It is interesting also to note one quotation of Herz-Schweizer's report ‘In the very variable EAT population, a few palms can be found with light yellow fruit colour (Kipemba or Kineupe) and a few palms with red fruit colour. The nuts are pear-shaped and small, the growth in height of the palm is slower than others. It could be that this palm is a King Coconut’[⁴⁶].

For the coast of Kenya, Adams gives an interesting account[⁴⁷]:

A dwarf variety with bright yellow nuts is often referred to as the King Coconut whilst the names of Pemba and Kitamli also seem to belong here. Trees of this variety are uncommon except in gardens and I know of no commercial plantings. Apparently it is not favoured as it is a poor producer of copra and the nuts are easily stolen.

A tall variety which has no particular name being simply called mnazi (Swahili: coconut). Personally I have dubbed it Kenya Tall or East African Tall for want of something better. All commercial plantings are of this variety and it appears to be something of a mixed bag genetically. Nuts are normally green or brownish-green to orange-brown. I have come across trees with yellow nuts but they are decidedly uncommon. Nut shape ranges from almost spherical to very much elongated (up to twice as long as wide) with something in between being the norm. Nothing is known of the origin of this variety it certainly being of considerable antiquity and has probably been cultivated since the first permanent settlement of this coast. I suspect a possibility, however, that its diversity has been added to by introductions of material from other areas, as trade between this coast and other ports of the Indian Ocean developed.

Until recently the word Kitamli seemed to have no particular meaning, other than just being a name for the local coconut variety. In a personal communication, Mr. M.A. Ghassan[⁴⁸], conjectured that tamli is a corruption of tambili[⁴⁹], having learnt from Tremlett[⁵⁰], that tamli was a word coming from Sri Lanka (Ceylon) and that the Tambili, or King Coconut, is a distinct variety. It is said to have originated in Ceylon, bears a yellow, ovoid fruit, is distinguished by its sweet juice and esteemed for culinary purposes, but of little commercial value for copra[⁵¹]. This identification, which matches that of Voeltzkow (q.v.), supports Jacob’s[⁵²] visual observations in Sri Lanka and Zanzibar.

Mr. Abdulrahman Rashid[⁵³] was told that the Kitamli came from the Tamil region of India and that the word Tamil was corrupted to Tamli, Ki being the Kiswahili prefix meaning ‘type of’.

We also asked many coconut farmers. Only one, Mr. Omar Masoud Hamad of Kipumbwi Village[⁵⁴], Pangani District, told us that he had heard from his late grandfather that the Kitamli came from Ceylon. Some farmers guessed that the word Ki-tamli had something to do with ‘sweet’ = tamu.

According to Purseglove[⁵⁵] the King Coconut of Sri Lanka is ‘Rath Thembili’, Cocos nucifera var. aurantiaca. He describes it as ‘a semi-tall cultivar late flowering, self-pollinating, heavy-setting, with fruits of medium size, but little endosperm, and of little use for copra, but providing a popular drinking nut in India and Sri Lanka’.

On Pemba there is a village with the name Mtambile. However, inquiries there did not yield in a connection between tambili and Mtambile[⁵⁶].

Recent efforts to identify King Coconut palms on Zanzibar did not result in clear-cut success. The examples found were not really convincing. They could have been yellow-fruited tall varieties or natural hybrids.

Discussion and Conclusions on Varieties

The most important local coconut variety is the East African Tall. It is indigenous to East Africa, probably for some thousands of years. This variety is mnazi mrefu in Kiswahili. According to its colour it includes mnazi mweusi, mnazi mwekundu and perhaps includes of those called mnazi mweupe. It seems the same as mnazi wa Unguja and mnazi wa Bahari of Voeltzkow and the African Nut of Craster.

At first sight, it is difficult to reach a conclusion about the Pemba Dwarf. Obviously two varieties that are the same in English (the Pemba coconut, the Pemba Dwarf) and in Kiswahili (Kipemba or Mnazi wa Pemba) cause the confusion.

Baumann describes the Nazi ya Pemba as a dwarf. Voeltzkow like Stuhlmann describes the Mnazi wa Pemba also as a dwarf, but says that there is another dwarf variety called Mnazi wa Kitamli (Kitamli would be understood as the Pemba Dwarf today). Craster describes the Indian Palm as the shortest and the Pemba Cocoa-nut Palm as in between it and the typical tall African Nut. Pearce only describes two local varieties and so does Williams. From their descriptions the local dwarf seems to be identical with the Pemba Dwarf. From Selby's description it becomes evident that there are two short varieties. The Kitamli being the shorter one, while the Razi ya Kipemba is of fair height, and the fruit colour is red-orange rather than pale yellow. According to Herz-Schweizer's description the Kipemba is a bit shorter than the tall varieties and the Pemba Dwarf rather similar to the Cameroon Red Dwarf.

In discussing coconut varieties introduced to Madagascar from the Comoro Islands, Prudhomme [⁵⁷] mentions two sub-varieties of Coco Sultan, one giving dwarf coconuts and the other giving coconut palms of tall habit. Coco Sultan is clearly another name for King coconut and says that ‘ . . . The Pemba dwarf coconut, or Merassi . . . doubtless represents the dwarf sub-variety of the Sultan Coconut named Irassi or Mourassi’[⁵⁸].

However, the clue may lie in the name King Coconut, if this is a transliteration from Indonesian of Kalapa raja[⁵⁹]. A variety of the name, with a description matching the King Coconut, the Thembili or the Kitamli has been known for at least 250 years[⁶⁰]. This might help to explain why the King Coconut, not mentioned by the earlier authors on Zanzibar, has apparently been absorbed into the local coconut population to such an extent that it is no longer possible to clearly identify individual King Coconut palms.

A likely explanation is that the King Coconut or Rath Thembili was introduced to Pemba as the Tambili (from either Sri Lanka or India) and the name became corrupted to Kitamli. When taken from Pemba to Unguja (Zanzibar) the name was retained but on the mainland the name kipemba was adopted. The introduction may have been comparatively recent, maybe 150 to 250 years ago, and from India since the name Indian Coconut is still well established in the source references and in the memories of a few people. Once established in Pemba it then became a source for other parts of East Africa.

One important link for the first part of the two words Rath Thembili might be established through Selby's letter. He names the second and taller dwarf variety as razi ya kiPemba. However, razi is not Kiswahili. The word was not known to anyone asked, nor is it in the English-Kiswahili dictionary. A z is softly spoken in Kiswahili like rose. It could be that the word was corrupted from rath (the th spoken as in English) into raz-i, the i commonly added into Kiswahili words ending with a consonant. The ya of razi ya indicates that the word belongs to the i-zi-class, into which are put most of the Kiswahili words borrowed from other languages.

A Pemba coconut, described as a prolific variety, palm comparatively small, husk pale-yellow and thin, kernel of usual thickness was reported by Barrett from the Philippines in 1912 [⁶¹]. At that time it must have been sufficiently well established though it is not apparently known today[⁶²].

There is no specific mention in the literature about an early introduction of dwarf varieties to Pemba, but Pemba is well accepted as the place from where dwarf coconuts were distributed to elsewhere of East Africa.

The close resemblance with the Cameroon Red Dwarf is not likely to be a coincidence. In this case the link appears to be the German colonial period about 100 years ago. Red dwarf coconuts have been found in other territories with a similar colonial history - Papua New Guinea, the Caroline Islands, Samoa and elsewhere in the Pacific.

A red dwarf coconut that has been identified in countries in West Africa and the Pacific could share similar origins. One hundred years ago, under colonial influence, these could even have originated from Pemba. In which case, the recent introduction of Cameroon Dwarf to Tanzania, might be a reintroduction.

The Development of the Coconut Industry in Zanzibar

The origin and distribution of the coconut was once a heated topics of botanical controversy. The school of thought that favoured the Pacific coast of South America is now abandoned. Other suggestions include the Indian Ocean, the New Guinea coast, the southwest Pacific, a link between South America and southeast Africa via Antarctica, or Gondwanaland[⁶³]. Currently, considered opinion suggests a Gondwanaland origin and a Malesian centre of domestication[⁶⁴].

Until now it cannot be certain that the coconut occurred indigenously in East Africa or whether it was brought by man. The first theory is based on the fact that very close relatives of the coconut palm occur in southeast Africa[⁶⁵, ⁶⁶] and Madagascar[⁶⁷]. The second theory expands on the possibility that Malaysian sea-rovers, who reached Madagascar in the first centuries AD, brought with them the banana and the coconut. From there coconuts could have reached East Africa. Alternatively, they might have been an early introduction to East Africa by Arabs, as there was interchange of crops between this region and India at least 3,000 years ago[⁶⁸].

The earliest known written source giving some evidence, dates back to the first century AD, the Periplus Maris Erythraei (Circumnavigation of the Erythraean Sea). The Periplus reports that coconut oil is exported from the ancient town of Raphta, believed to have been situated at the mouth of the Pangani river[⁶⁹] on present-day Tanzania mainland. However, there is some debate over the reference since the word taken to mean coconut oil has also been read as ‘cuttle-fish’ and ‘pearly sea-shells’[⁷⁰]. Interpretations vary between ‘a little coconut’, ‘coconut oil’ and even ‘copra’. The explanation may lie in the fact that when the translations were published in the late 19th century coconut oil and copra were important articles of trade. But at the time the Greek text was written the coconut palm was hardly known in Europe, copra was not traded and oil was produced and consumed locally, where the coconuts grew. If the coconut was known at all in Europe it was as the empty shell of the nut possibly decorated and ornamented. Coconut shells, together with those of the Coco de Mer, were called Nux indica. They were carried overland by Arab traders, known in Greek temples[⁷¹]. and reached as far as English cathedrals[⁷²] long before the Portuguese sailed to East Africa.

The domestication of the coconut palm is thought to have occurred in the East Indies from where selected varieties where carried into the Pacific and Indian Oceans [⁷³]. This predates the persistent traditions on the coast of Tanzania mainland, Zanzibar and Mafia the arrival of the coconut is attributed to the Wadebuli, whose ships reputedly had sails of palm matting[⁷⁴, ⁷⁵, ⁷⁶, ⁷⁷]. The Wadebuli are thought to have come from Dabnal, formerly Dabul, a port on the west coast of India, connecting Persia, the Red Sea and the coast of East Africa[⁷⁸]. Pemba tradition credits the introduction of the coconut palm to the Wadiba, who according to Gray[⁷⁹] hailed from the Maldives islands, which were known to 14th century Arab geographers as the Diba islands. According to the Arab traveller Ibn Battuta[⁸⁰], great quantities of cowries and coconut products were exported from these islands. Both the Maldives and the Laccadives were the scene of remarkable ship building activities. The ships, including hulls, masts, ropes, caulking, bailers and even sails, could be built and provisioned entirely from the various products of the coconut. The Arabs and Persians from the Gulf used to import coconut products from these islands or go there to have their ships built on the spot. There is evidence that the Maldives were first settled by Singhalese Buddhists who planted coconuts and dug wells[⁸¹, ⁸²].

Two Arab geographers during the early Middle Ages have referred to the cultivation of the coconut in East Africa. Al-Masoudi made several visits to Kanbalu, which is generally considered to be the present-day Pemba or Zanzibar. From Kanbalu he reported the extensively cultivated coconut palm [⁸³, ⁸⁴, ⁸⁵]. Ibn Said reported in the 13th century that the Mand islands near Mombasa were celebrated for their coconuts. Among these islands he described Kilwa as the most important and, amongst others, Zanzibar[⁸⁶]

By the time the Portuguese reached East Africa in 1498 the coconut was well established in the settlements along the coast of East Africa. The multitude of its uses feature in many of their chronicles since it was a great novelty to them. The Portuguese observed the role of coconuts in ship-building: ‘The vessels of this country are of a good size and decked. There are no nails and the planks are held together by cords. The sails are made of palm matting’[⁸⁷]. Duarte Barbosa observed the same boats on the islands of Zanzibar, Pemba and Mafia in 1517[⁸⁸]

From the 1820s Seyyid Said, Sultan of Oman from 1806 to 1856 who transferred his capital from Oman to Zanzibar in 1832, encouraged the settlement of Oman Arabs on Zanzibar and the coast of East Africa. These immigrants together with the local Swahili began to grow cloves, grains and coconuts on a plantation scale. The Oman Arabs took an interest in coconut production at a critical period when the earlier interest in coir fibre for maritime purposes was replaced by the demand for copra and oil for industrial purposes[⁸⁹].

In the 19th century there were substantial ups and downs in coconut cultivation mainly due to price fluctuations in cloves, the restrictions to the slave trade and finally the abolition of slavery. In the 1830s the price for cloves was so attractive that the French naval officer Captain Loarer talked about a ‘clove mania’. Coconut palms were cleared away to make room for clove trees. Sultan Seyyid Said, who was anxious not to become too dependant on one crop, compelled his subjects, under threat of confiscation of land, to plant a certain proportion of clove to coconut palms[⁹⁰, ⁹¹]. And indeed in the 1840s the price for cloves collapsed and coconuts became much more attractive. The export of coconut products amounted to about MT$ 50,000 [⁹²] in the late 1840s and increased to about MT$ 200,000 in the 1860s.

In 1859 coconuts and coconut products ranked as seventh most important export commodity. As much as 1,750,000 pounds weight of coconuts, 2,450,000 pounds of copra and 252,000 pounds of coconut oil were exported[⁹³]. Beside the decline of clove price, restrictions in the slave trade[⁹⁴] are also held responsible for the rise in coconut production[⁹⁵]. During the period of the 1840s to the 1860s the Zanzibar agriculture became so oriented towards cloves and coconuts that Zanzibar turned from an exporter of foodstuffs to an importer[⁹⁶] The opening of the Suez canal in 1869 must have had a positive effect on Zanzibar copra production by shortening the distance to European markets. In contrast, in 1872, a hurricane hit Unguja Island destroying two thirds of the clove trees and coconut palms. Pemba Island was hardly touched and since ever then became the major producer of Zanzibar cloves. Unguja Island slowly recovered.

Around 1900 the first estimates were made as far as the yield of coconuts and the coconut population in Zanzibar was concerned[⁹⁷]. An anonymous writer estimated that a coconut palm would yield between twenty-five and thirty nuts[⁹⁸] and that Pemba palms yield less than those from Zanzibar. He assumed that home consumption was approximately equal to the copra export in 1901 (6,519,216 lbs) and that one may assume a coconut population of about 1 million palms on about 28,571 acres. During that time also the first coir fibre factory was established in Zanzibar though it did not prove viable[⁹⁹]. Coir fibre would have been produced domestically from the time when it was first used for boat construction and rigging.

At the turn of the century everyone was very much concerned about the impact of the abolition of slavery. The British General Consul Hardinge estimated that the production of the Sultanate would be reduced by two-thirds. This encouraged other colonial powers, especially the Germans in Tanganyika, to speculate about possibilities of moving into these production gaps[¹⁰⁰]. Some scholars are of the opinion that indeed the less labour intensive coconut cultivation was favoured instead of clove cultivation[¹⁰¹]. But this has been convincingly refuted by Ferguson who shows that the clove and copra production constantly increased in the years from 1895 to 1935 in approximately the same proportion. The five years' averages of cloves and copra exports of domestic production increased from 1896 to 1935 from 5,700 tonnes of cloves to 10,000 tonnes and from 5,000 tonnes of copra to 12,000. Ferguson concludes that ‘The transition from slave to wage labour in Zanzibar was accompanied by some dislocations in the labour supply, but this made no adverse impact on the volume of clove and copra exports following abolition in 1897’[¹⁰²]. He adds that the period from 1915 to 1930 was a prosperous one because the prices for cloves and copra rose on the world market substantially. During this period the coconut palm population was estimated at around 3.85 million palms on about 45,000 acres in Unguja and 10,000 in Pemba[¹⁰³].

The trend changed in the 1930s due to the Depression. The prices of copra and cloves plunged: copra from 22,000 to 10,000 (Shillings) per ton and cloves from 81,000 to 53,000 (Shillings) (average of 1926-1930 against 1931-1935). Copra export production decreased from 12,000 tonnes in 1931-35 to 9,000 in 1936-40. The development from 1941 to 1945 was also disturbed by the Second World War. Copra produced for export reached only about 8,800 tonnes in that period[¹⁰⁴].

In 1935 the first attempt was made to systematically select high yielding local tall coconuts. Initially 300 were phenotypically selected from about 45,000 coconut palms of a government plantation in Kidichi[¹⁰⁵]. In the following three years fourteen palms with an average of 120 nuts/palm and a copra output of 66.31 lbs. per palm were selected as ‘permanent seed-bearers’ while thirteen palms with an average of 109 nuts/palm and a copra output of 57.77 lbs. per palm were selected as a seed source for the general public. From 1937 in Kizimbani a block of coconuts was planted with seednuts from the permanent seed-bearers[¹⁰⁶]. The block was later abandoned[¹⁰⁷]. Reasons are not entirely clear, though there is a note from 1944 deploring high losses due to neglect and a shortage of labour[¹⁰⁸].

After 1945 copra production for export recovered again. The five years' average for the years 1946 to 1950 showed again 12,000 tons and 11,800 for 1951 to 1955[¹⁰⁹] For 1946 a coconut palm population of 4.5 million palms was estimated[¹¹⁰].

A major event of the year 1949 was the creation of the Copra Board. This was a public body that administered the funds obtained from a small cess on exported copra targeted at fostering and encouraging the production of coconuts and coconut bye-products. The Copra Board produced coconut seedlings reaching in some years more than 200,000 seedlings, provided skilled masons for improved copra kiln construction free-of-charge, offered building material for copra kilns at cost price, set up a copra production advisory service, and ran a coconut products and by-products demonstration factory in Mtoni. The Copra Board was especially successful in enhancing the coconut coir production in Zanzibar and operated from 1949 to 1959[¹¹¹].

In the early 1950s one of the major coconut pests, the Coconut Bug, Pseudotheraptus wayi, was identified. Since the pest could only be controlled efficiently by spraying insecticides the Department of Agriculture soon began to introduce dwarf coconut varieties in Zanzibar[¹¹²] (on the misapprehension that these would be easier to spray).

Reports after 1955 are scanty. It is not clear if this was due to the turmoil of the pre-independence period or that records simply got lost. Nevertheless, one interesting report in 1958 can be mentioned,. The District Agricultural Officer of Pemba reported a coconut palm survey in which 1,176,440 bearing coconuts were counted[¹¹³]. Unfortunately there is no mention that such a survey had been undertaken for Unguja as well.

In 1963 Zanzibar gained independence. All British personnel of the Department of Agriculture left. In 1964 the Zanzibar Revolution shook the islands. Factories and large land-holdings were nationalised. Where formerly the copra buying was in the hands of private, mainly Indian and Arab traders, this shifted to the state organisation Bizanje in April 1964[¹¹⁴]. This task was transferred in June 1970 to the Zanzibar State Trade Organisation (ZSTC), which is responsible for all exports of commodities from Zanzibar[¹¹⁵]. From 1970 data on copra production are again available.

In the 1970s and 1980s there was a general decline in coconut production on Zanzibar. The five years' average copra production dropped from 10,554 tons in 1970-74, to 9,477 in 1975-79, to 9,360 in 1980/81-84/85 and to 2,042 in 1985/86-89/90[¹¹⁶]. Reasons were manifold. On the one hand there were technical problems, such as pests and diseases, unproductive overaged palms, and on the other hand marketing problems such as a low producer price for copra and a ban on the export of fresh nuts.

Additionally, many of the former coconut plantations (about 66,000 acres) in Zanzibar had been nationalised and shared out to some 22,000 smallholders[¹¹⁷]. Lack of knowledge on improved coconut husbandry and insufficient land tenure security added to the problems in coconut production. Last, but not least, the Zanzibar population increased annually by 2.9%. Thus more of the coconuts went into domestic consumption instead of for export. The coconut turned from a prime cash crop to a prime subsistence crop. The coconut palm population for Zanzibar was estimated on air photos taken in 1977 at around 5.7 million. palms on about 61,000 hectares, 4.6 million coconut palms in Unguja and 1.1 million in Pemba[¹¹⁸] with a total production of approximately 164 million nuts with 113 million for home consumption and 51 million for copra production[¹¹⁹]

In order to halt and reverse these negative trends the National Coconut Development Programme was launched in 1979 on Tanzania Mainland and Zanzibar. The project is externally supported by technical assistance from the Federal Republic of Germany and by financial assistance from the International Development Agency. The project is primarily oriented in the agricultural development of the coconut industry. It began with a major thrust in research of coconut pests and establishing hybrid coconut seed farms. While much useful information was generated and made available through the extension service, the hybrid seed production was suspended in 1991. This was due to unsatisfactory performance of the new hybrids when exposed to drought and lethal disease. Since 1988 the project has put more effort into selecting improved genetic material in the local East African Tall population and breeding from the lethal disease survivors in local and introduced populations.

ACKNOWLEDGEMENTS

This paper was presented at the International Conference on the History and Culture of Zanzibar in December 1992 by Eberhard Krain but post-conference Proceedings were not published. The authors wish to acknowledge the permission given by Professor Abdul Sheriff and Dr. Hamadi H. Omar , two of the conference organizers, to allow it to be published separately. At the time of writing the authors all worked for the Tanzanian Ministry of Agriculture or for the German Agency for Technical Cooperation (GTZ ) in the National Coconut Development Project (NCDP) in Zanzibar and Dar es Salaam, and they would specifically like to thank Mwatima Juma, Commissioner of Research and Extension in the Zanzibar Ministry of Agriculture, and Lothar Diehl, the GTZ/NCDP Team Leader, for their help and encouragement. All translations from German texts were made by E. Krain. Additional bibliographic material is accessible through the Coconut Time Line at <http://www.geocities.com/harriesjh/cocotimeline.html>.