The Development Of Oranges As A Cash Crop In Ndijani, Unguja

MINISTRY OF AGRICULTURE, LIVESTOCK AND NATURAL RESOURCES

ZANZIBAR CASH CROPS FARMING SYSTEMS PROJECT (ZCCFSP)

WORKING PAPER No.: WP 95/22

AUGUST 1995

THE DEVELOPMENT OF ORANGES AS A CASH CROP IN NDIJANI, UNGUJA

BY: Abdallah Kombo Soud Hemed Peter Oldham Martin Walsh (ed)

ZCCFSP P.O.BOX 2283 Zanzibar

Telephone / fax: (054) 33121

CONTENTS

Page Preface

ii

Introduction: Purpose of the Study

1

Research Methods

3

Sweet Oranges: Origin and Introduction

4

Research during the Colonial Period

4

Sweet Orange Varieties in Ndijani

6

The Expansion of Production in Ndijani

7

Farmer Case Studies

8

The Availability of Land

12

Land Improvement

14

The Availability of Labour

16

Social and Economic Differentiation

18

Large and small Farmers

18

Squatters and Agricultural Labourers

20

Gender Differentiation

21

The Evolution of the Marketing System

22

Leasing Arrangements

23

Sale to Middlemen

23

Sale Direct to the Market

24

The Rotating Marketing System

25

Conclusion: Lesson Learned from the Study

27

References

29

PREFACE

This is one of a series of five cash crop case studies undertaken by the socioeconomics section of ZCCFSP in Zanzibar. The main aim of these studies is to analyse the different factors which have led to local cash crop development, with a view to forming policy recommendations and devising practical interventions which might further facilitate and promote this kind of development in the future. We hope that the lessons which are drawn from these studies will be of relevance to the Ministry as a whole, and not just to projects like ZCCFSP whose primary interest is in encouraging sustainable cash crop production and marketing.

The present study examines the development of sweet orange production in Ndijani, in central Unguja. It is based upon fieldwork and library research carried out by the staff of the socio-economics section of ZCCFSP on Unguja. Here we would like to take the opportunity to thank all of the farmers in Ndijani who discussed the history of sweet orange production with us, as well as colleagues in ZCCFSP Unguja who provided additional background information on Ndijani and its environs.

Martin Walsh (ed) Socio-economics Section ZCCFSP Pemba

Wete, August 1995

Other titles in this series: The Development of Turmeric as a Cash Crop in Mwambe, Pemba (WP 95/23) The Development of Sweet Potatoes as a Cash Crop in Makangale, Pemba (WP 95/24) The Development of Pineapples as a Cash Crop in Machui, Unguja (WP 95/25) The Development of Mangoes as a Cahs Crop in Muyuni, Unguja (WD 95/26)

THE DEVELOPMENT OF ORANGES AS A CASH CROP IN NDIJANI, UNGUJA Introduction: Purpose of the Study

1. One of the principal objectives of ZCCFSP is to foster the development of cash crop production and marketing in Zanzibar and to enhance the capacity of farmers, traders, and government to participate effectively in this development.

The history of government-directed efforts to

promote cash crop diversification in Zanzibar has not always been a happy one. Agricultural research and extension have more often than not followed the top-down model, where researchers and policymakers decide what is good for farmers and extensionists impose it upon them with a minimum of consultation and consideration for what farmers themselves might think.

As long as clove production

dominated the economy of Zanzibar the recurrement failure of this approach seemed to make relatively little difference. The recent and drastic decline in clove prices, however, has given agricultural diversification a new urgency. The adoption and dissemination of a participatory approach to research and extension has been a key feature of ZCCFSP’s response to this situation, and the present study (one of a series) was conceived in this light, and not as an academic exercise. 2. If participatory research and extension are to be effective it is essential for researchers and extensionists to understand how farmers and traders think and act and why they do so. The traditional approach to this question, however, is to ignore it until it starts to have a negative impact upon project or programme objectives (when farmers’ and traders’ actions do not match expectations) and it is often too late to do anything about it (especially if a lot of time and resources have already been expended in reaching this point).

Zanzibar provides a good

example of this: before ZCCFSP began work (in November 1991) almost nothing had been written about farming and marketing practices

on the islands aside from the little that could be gleaned from the results of questionnaire-based surveys.

Agricultural projects and

programmes were therefore devised and implemented in a virtual void, filled only by the assumptions of planners and the incomplete or informal knowledge of implementers about the indigenous practices relevant to these. 3. As

ZCCFSP’s

own

experience

has

shown,

developing

and

institutionalising an alternative approach is not something that can be done overnight. The orthodox model of cash crop diversification and development is, as might be expected, crop-oriented and often focuses upon the introduction of new germplasm (new at least to a particular group of farmers and/or a particular location). The basic method is simple: it is the researchers’ task to find a crop or variety which has a promising market, and then to work on ways to introduce it, expand production and ensure that it reaches the market. It is also an old method, and one which was employed extensively during the colonial period. In some cases it succeeds, and a number of the crops being sold in Zanzibar’s markets are improved varieties which were introduced through the agricultural research station at Kizimbani. However, many of them were not developed in this way, and the records of MALNR are replete with time-consuming and expensive failures, of introduced crops and varieties which have all but sunk into oblivion. 4. ZCCFSP has continued to work with this model, albeit with a number of important refinements. These include an emphasis upon on-farm trials, careful assessments of export markets, and of the potential for expanding production in terms of what is known about the land available and the likelihood that farmers’ will grow the crop or switch to the new variety in the first place.

To the extent that this work is

conducted participatively with farmers and traders, it might be thought of as a compromise between old and new approaches.

It is not,

however, entirely free from the problems which afflict the inherited mode. On-farm trials in a range of crops and varieties on both Unguja and Pemba have been abandoned following their failure for one reason

or another, and while important lessons have been learned in this process, it is quite likely that similar failures will occur in the future, in other MALNR projects which operate with this model as well as ZCCFSP. 5. The present study of the development of oranges as a cash crop in Ndijani, in central Unguja, is one of a series designed to address this problem. These approach the issue of cash crop development from a quite different angle, by looking in detail at selected crops which have already become important commodities without any direct input at all from government researchers and extensionists, and in some cases in spite of their efforts. 6. The philosophy behind this approach to the problem is quite straightforward.

Rural Zanzibar comprises more than 100,000 farm

households, most of them with two or more members involved (in varying degrees) in farming and (to a lesser extent) in the sale of farm produce.

Every year they make innumerable decisions about the

cultivation and harvesting of a wide variety of farm plots, including major decisions about what to plant, what to sell, and how. In so far as they are striving to solve agricultural problems (as a means to meet the requirements of domestic survival and subsistence), they are conducting agricultural research. From this point of view the number of formal experiments which government researchers can undertake pales into insignificance.

And while the proportion of farmers’

experiments which have an uninteresting design or result might seem inordinately high to a formal researcher, the cumulative effects of farmers’ research can be very impressive, and have consequences which many government researchers and extensionists can only dream of. In short, a lot can be learned from looking at what farmers (and traders) are already doing. 7. By looking at past and present examples of successful cash crop development it is possible to examine and assess the contribution of a wide range of factors in each case. This is rather more difficult to do when retrospectively analysing the failure of on-farm trials, especially when they have failed at an early stage. Although it may be possible to

isolate the cause or causes of failure, there is no guarantee that this will suggest ways in which these and other constraints may be overcome, including constraints which emerge at a later stage of development. For example, if a trial fails before a crop is harvested, then there is obviously nothing which can be learned from this about marketing. The study of ‘real-life’ cases, however, offers a lot more information and should make it possible to provide much clearer guidelines to researchers who are experimenting with crops, varieties, and techniques which have not been known to farmers before. It also makes it easier to assess the viability of crops, especially tree crops, which cannot be developed in a short period of time (for example within the life-cycle of a project) and provides some perspective on the length of time which the development of any cash crop might reasonably be expected to take. 8. By comparing a series of similar studies common patterns should emerge, and this has already begun to happen in the case of the ZCCFSP studies.

The most important application of this is in the

design of more appropriate strategies for fostering the development of cash crop production and marketing, as well as in predicting what some of its impacts might be upon different sectors of the farming and trading community.

Readers of this report are asked to consider

carefully what they think the implications of it might be, and what kinds of intervention might be effective in replicating the kind of development which has taken place in Ndijani, or how it might be improved upon. In the final section below we discuss some of the lessons which we think emerge from this study. This should not, however, be taken as the final word on the matter, and in a future report we will present an expanded comparison of all of the case studies and their implications, which would no doubt benefit from any constructive comment and criticism which is supplied to us in the interim.

Research Methods 9. This study is based primarily upon fieldwork undertaken in Ndijani by the socio-economics sections section of ZCCFSP Unguja in December 1994 and January 1995. Ndijani is located in central Unguja on and to the south of the Zanzibar-Chwaka road, to the east and south-east of Dunga.

It comprises a collection of hamlets straddling two distinct

types of land.

The original settlement of Ndijani is sited on old

plantation land, while many of its newer satellites are on the neighbouring uwanda or semi-coral rag land. Administratively Ndijani forms a single shehia (formerly CCM branch) within Chwaka constituency (jimbo) in Unguja’s Central district.

In this report the

name Ndijani is applied rather more loosely to include the neighbouring shehia of Binguni, which is also a source of ‘Ndijani’ oranges. Ndijani was selected as the location of this study because of the fame of its sweet oranges in Zanzibar town market and the fact that it is the primary orange-producing area on Unguja. 10. A checklist of research questions was drawn up prior to conducting informal interviews with farmers in the field. After a preliminary round of group interviews the checklist was expanded and refined and then used as the basis for interviewing sixteen individual orange growers in the Ndijani area. The results of this work were written up in the form of interview notes (retained on file in the ZCCFSP Unguja office) and subjected to initial analysis by the socio-economics section in early 1995. Further information on orange production and marketing was collected in the course of a study of Ndijani (Mseweni) Farmers’ Research Group (FRG) conducted by a multidisciplinary team of researchers (from ZCCFSP and other MALNR departments) in May 1995.

This study provided a much more detailed overview of the

farming system as a whole and its development over time, and the relevant results (which have been written up in a separate report) have been incorporated into the present study. 11. A brief literature review has provided additional background data on sweet oranges and their production in Unguja. No attempt has been made, however, to make a thorough search of the archives or interview

other key informants either within or outside of MALNR. The full history of orange production and marketing on Unguja therefore remains to be written, as indeed it does for most other crops on the island. 12. In writing the report we have tried to highlight the main stages in the development of oranges as a cash crop in Ndijani, singling out the principal factors which have contributed towards it. The logic behind this form of presentation has already been outlined in the introduction above.

Again, we invite anyone with additional information an/or

alternative interpretations to share them with us, and so add to the potential value of our research.

Sweet Oranges: Origin and Introduction 13. The sweet orange, Citrus sinesis, originated in southern China or thereabouts, apparently as a hybrid between the mandarin (Citrus reticulata) and the pummelo or shaddock (Citrus grandis, syn. C.maxima). A number of Citrus species, including the sweet orange, are believed to have reached the Arabian peninsula more than two thousand years ago, and its first appearance in East Africa could conceivably date from the same period. The likely etymology of the Swahili name of the tree (mchungwa) and its fruit (chungwa) suggests that a bitter-tasting Citrus species was introduced first. Whatever the case, the sweet orange has been cultivated on the Swahili-speaking coast for at least the past 500 years. Ibn Battuta found oranges, of unspecified sweetness, growing in Mombasa in 1331, while the first Portugese visitors in 1498 marvelled at the “fine oranges, better than those of Portugal” which were the main cargo of a small boat they met in the Pemba Channel.

These were presumably botanical sweet

oranges, which had only recently become known in their home country following their reintroduction to Europe circa 1425 (sweet oranges had been grown by the Romans but were apparently lost after the fall of the Roman Empire). 14. The subsequent history of sweet orange production in Zanzibar, prior to the establishment of the British Protectorate, is obscure. It appears to have been a relatively minor crop which nonetheless always found a

ready local market, especially in the towns and on visiting ships. Older people in Ndijani can only say that oranges were introduced there more than a hundred years ago, and that they were being sold in Chwaka, Uroa, Bambi and even Zanzibar town for as long as they can remember (since at least the 1930s when our oldest informants were in school).

Research during the colonial Period 15. Rather more information is available on sweet orange production during the period of British rule, and especially on the research undertaken by the Agricultural Department. The Annual Report of the Agricultural Department for 1902 records that a (?) European plantation owner at Mtoni had imported and planted oranges from Spain and Portugal, though it is not stated whether these were ‘Valencia’ and/or other cultivars.

Presumably this landowner was not alone in

experimenting with imported varieties. However, there is little evidence to suggest that plantation owners’ experiments had any long-lasting impact upon the production and marketing of sweet oranges, especially that already being undertaken by smallholders. The only exception to this is the local cultivar called ‘Binti Juma’, which is said to have originated on an Arab estate at Kwambani (Binti Juma being the owner of the estate and mother of Seyyid Hafidh bin Muhammed el Busaidi). The fruits of ‘Binti Juma’ bear a close resemblance to those of the commercial cultivar ‘Washington Navel’ (which probably originated from a bud mutation in Brazil in the early nineteenth century), suggesting that it was originally an importation rather than a local mutation. 16. The Agricultural Department itself did not begin serious work on sweet oranges and other Cirtus species until more than a quarter of a century after this early report. In November 1928 a number of young Citrus trees were imported from Sough Africa and planted on research plots at Kinooni and Dunga. These included eight varieties of sweet orange: ‘Du Roi’, ‘Jaffa’, ‘Joppa’, ‘Lue Gim Gong’, ‘Mediterranean Sweet’, ‘Pineapple’, ‘Valencia’, and ‘Washington Navel’. Rough lemon (Citrus

jambhiri) rootstocks were found to be superior to sour orange (Citrus aurantium) stocks for budding. In 1933 it was reported that ‘Pineapple’ and ‘Valencia’ had yielded well at Dunga, but that the fruits were not as sweet as those of local orange varieties. Research on sweet orange varieties continued at the new Kizimbani Experimental Station after work at Kinooni and Dunga was discontinued. In 1934 at least four varieties were being grown in the orchard: ‘Binti Juma’, ‘Joppa’, Mediterranean Sweet’ and ‘Washington Navel’. 17. By 1936 it had become evident that the local varieties of oranges were superior to the imports. Research was therefore begun at Kizimbani into the indigenous cultivars, and selected local scions were budded on rough lemon rootstocks in a trial plot. In 1945 the fruits from these, as well as from budded trees at Selem Government Plantation, were subjected to detailed examination. Testing of quality for flavour, size and shape of fruit, number of seeds per fruit, thickness and colour of skin, and juiciness continued in the following year, but seasonal variation produced conflicting results.

Although budding appears to

have continued, systematic research into fruit quality seems to have been abandoned without being brought to any definite conclusion. 18. In addition to the search for better varieties, the Agricultural Department conducted research into other aspects of orange production and processing during this period. This included work on pests and diseases, the use of fertilisers, and the artificial colouring of fruits to give them a more attractive appearance.

By 1939 it had

become stated government policy to establish the international export of oranges. Needless to say, none of the research undertaken by the Agricultural Department brought this objective any nearer to realisation than it already was. The production and marketing of local seedling varieties remained largely unaffected by this research. In 1949 the Director of Agriculture noted that these oranges had a long-standing reputation for quality in the nearby mainland markets, and were already a cash crop of some importance. 19. Nonetheless, this early research had had some influence upon more recent developments.

Budded trees were already being sold at

subsidised prices to farmers by the late 1940s.

This supply has

continued through to the present, and one variety in particular forms an important component in the orange-growing industry of Ndijani, though it is still less important than the indigenous seedling.

Sweet Orange Varieties in Ndijani 20. Three main types of sweet orange are cultivated in Ndijani. These are (a) ‘Original Ndijani’, Ndijani asili, a true cultivar which is grown from seed; (b)

‘Budded Ndijani’, which is budded in the Forestry

Department nursery in Ndijani using an ‘Original Ndijani’ scion and rough lemon rootstock; and (c)

‘Budded Binti Juma’, also called

Mpemba, which is said to derive (or to have originally been derived) from Kizimbani and to be propagated using a ‘Binti Juma’ scion and rough lemon rootstock. There is, however, some uncertainty among informants on the propagative origin of the two budded types and it may well be that their history is often more mixed than this account implies. The characteristics of ‘Original Ndijani’ and the early fame of sweet oranges on the East African coast strongly suggest that it has developed directly out of the old local varieties. 21. The typical characteristics of the three recognised types are more certain than their origins. ‘Original Ndijani’ is preferred by most farmers despite the fact that it is a thorny tree and the fruit therefore difficult to harvest. Its favourable characteristics are reported to be (i) the ease with which it can be propagated (at zero cost); ((ii) good yields; (iii) the fact that fruits can be left on the tree for some time after they mature without any decline in quality and with less damage from pests than other varieties (which favours delayed marketing strategies); and (iv) the attractiveness of the fruit to customers in terms of its average size, yellow peel colour, and especially its sweetness and juiciness. These last two characteristics override the fact that the fruit contains more seeds than any of the budded oranges. 22. While the fruit of ‘Budded Ndijani’ does not match up to that of ‘Original Ndijani’ and the tree is equally thorny it has other advantages: (i) the rootstock makes it more resistant to disease and drought; (ii) the tree

matures early; and (iii) it produces the highest yields, bearing fruit which are larger than ‘Original Ndijani’. High yield is its most favoured characteristic, and individual trees are reported to produce up to 20 basket-worth of fruit per season (around 3,000 oranges calculated on an average of 150 per susu). 23. By comparison ‘Budded Binti Juma’ has relatively little in her favour. Although the trees are thornless and the fruits large and seedless they contain very little juice and sometimes none at all when overripe. As a result the market demand is minimal. It is ironic that this offspring of a ‘commercial’ cultivar (assuming that ‘Binti Juma’ is ‘Washington Navel’ or something very close to it) should be the least successful. It is common knowledge that ‘Washington Navel’ produces very large fruits in wet climates, and presumably no one would attempt such an introduction now in a place with conditions similar to those of Zanzibar. 24. The principal factor in the local demand for sweet oranges is clearly juiciness and taste, and in this respect ‘Original Ndijani’ comes out well on top. Many and probably most of the oranges sold in the local and town markets and roadside stalls are sucked by consumers after being skilfully peeled and sliced in half with a knife. The pips in the oranges are a minor inconvenience for these consumers and are readily spat out.

The Expansion of Production in Ndijani 25. It seems that prior to the Zanzibar Revolution in 1964 oranges were grown in and traded from Ndijani, but only on a small scale. In this respect oranges were no different from many other tree crops, including mangoes, which were planted in small numbers around homes in the plantation area, providing fresh fruit for domestic consumption and a useful source of seasonal and in most cases petty income for the farmers who had enough oranges to sell and wanted to do so. A relatively small number of farmers probably earned more income from buying and trading oranges further afield. This, at least, is the pattern still found throughout most of the plantation areas of both Unguja and Pemba, and all the indications are that Ndijani was once

not much different. Certainly sweet oranges were not the major cash crop which they have since become. Local farmers derived the main part of their income from coconuts and cloves and/or working on the larger plantations of these. 26. Some informants stated that Ndijani became famous for its oranges just after the 1964 Revolution, though the production histories we collected suggest that significant expansion did not take place quite so soon after this major transformation in the political and social economy of Zanzibar.

In

the absence of more detailed information about

orange-growing in this period, we can only guess at what may have happened.

The flight and eviction of most large landowners, the

nationalisation of their land and its redistribution to squatters, farm labourers and other had far-reaching consequences for smallholder production, creating a whole new set of agricultural opportunities for farmers once they had oriented themselves to the changed social and economic environment.

At the same time the Revolution and the

redistribution of wealth which it engendered seems to have had a considerable impact upon the Zanzibar town market. No doubt many trading links were severed in this process. It also appears, however, that new opportunities emerged as a different class of urban consumers came to the fore. Precisely what effects this may have had upon the demand for Ndijani oranges is not clear, but it can be presumed to have had some. 27. Individual case histories indicate that orange production began to expand significantly in Ndijani in the early 1970s.

By 1975 large

numbers of farmers had entered into production, many of them planting trees in large numbers, and oranges had become the most important economic crop in the Ndijani area. Thereafter production appears to have expanded steadily year by year, as existing growers continued to plant oranges and others, including young and newly-independent farmers, established their own farms. At the same time some farmers discontinued planting, especially those who were unable to obtain more land on which to expand production. Nonetheless the overall trend has been one of expansion, though there are some signs that this trend

may be abating with the recent emergence of fears that the market may be reaching saturation point. 28. This quarter-century history of increasing production can be attribute to a number of factors. These include (a) population growth in Ndijani and the need for farmers, especially first-time farmers, to open up new land; (b) the availability of this land on the previously underexploited uwanda bordering the original centres of settlement on plantation land; (c) the evolution of an efficient mixed farming system on the uwanda which increases its fertility through extensive manuring; (d)

the

availability of agricultural labour , in the form of a large local population of mainlanders, able to sustain this development; (e) the decline in income from other sources, both on-farm and off-farm, which has pushed farmers into developing alternative cash crops; (f)

the

agronomic suitability of orange trees for this purpose, including their adaptability to a mixed farming system on uwanda soils; (g)

the

presence of a good local cultivar and the availability of budded trees, all of which farmers (or at least some farmers) were familiar with; (h) the existence of a well established market for Ndijani oranges and the experience of some farmers in marketing them; (i) the capacity of this market to expand considerably (which it clearly has, though we can only speculate on the evolving relationship between supply and demand in this case); and (j)

the development of new marketing

strategies by orange growers and traders to match this expansion. 29. This is a long list, and other factors might be added (for example the improvement of transport links between Ndijani and Zanzibar town). Clearly these have not all contributed equally to the development of oranges as a cash crop, and some of them belong to quite different causal categories. In later sections we will examine a number of these factors in more detail, and attempt to pinpoint the most important in the concluding section.

First, however, we will present a series of

individual case histories which illustrate the expansion of production and development of marketing strategies in more detail and provide a starting point for the analysis which follows.

Farmer Case Studies 30. The following individual case studies are abstracted from our field notes, and presented in sequence according to the year in which each farmer started to grow oranges. All of the farmers interviewed were men, reflecting the fact that women play very little direct role in the production and marketing of oranges in Ndijani, except as farm labourers. The figures they quoted are not always exact, but should be treated as estimates or approximations. This is particularly so in the case of large numbers of orange trees, and likewise in the measurement of farm acreages.

1971 Farmer A, currently living at Ndijani Kwa Baniani.

Originally he grew

maize, tomatoes and pumpkins as his main cash crops. He started to grow oranges in 1971. In this year he planted a total of 40 trees, 37 ‘Original Ndijani’ from seed and the remaining 3 ‘Budded Ndijani’. In 1981 he planted another 25 trees on a different farm in Ndijani, but was unable to grow more because of he had insufficient land on which to do so. In 1985 he was given a plot by his father and so planted a further 25 trees. He therefore has a total of 90 trees and in 1995 was planning to plant 40 more ‘Original Ndijani’ from seed. Oranges are now his main cash crop, and are intercropped with coconuts. He does not know who organises the rotating marketing system and does not take part in it. He usually picks all his own fruit even if the prices are low. Sometimes he leases out the harvest to middlemen from the village. He reckons that 3 masusu of fruits per tree can be lost because of damage caused by bugs.

1973 Farmer B, Binguni. He originally lived and farmed on family land in Ndijani village, before purchasing a farm about a mile away in Binguni so that he had greater freedom to plan his own agricultural activities. He bought 13 acres of land for Tshs 750, and paid an additional Tshs 250 to obtain the relevant documentation. In 1973 he planted 100 orange trees on this land, planting a further 180 in 1975, 250 in 1985, and 300 more trees in 1990

300 trees. All of the 830 trees he has planted were the ‘Original Ndijani’ variety grown from seedlings, and he financed each stage of this expansion with his income from orange sales.

He stopped planting

oranges in 1990 because of the low price which was then being offered in Zanzibar market. He fear that continued planting by many other farmers will result in an oversupply of fruits and cause a drastic fall in prices, possibly in 3-5 years’ time. In response to this prediction he has begun to plant ‘Boribo’ mangoes and coconuts instead of oranges, noting that the markets for mangoes and copra provide much better future prospects. He has heard about the rotating marketing system which operates in Ndijani and Binguni villages, but notes that farmers in other areas are not interested in it: he does not participate in it himself or know who organises it. Instead he prefers to picks his oranges as soon as they are mature and thinks that there is no point in leaving them unpicked on the tree while waiting for higher prices on the market. Instead he thinks that orange prices should be set by a commission of farmers with advice from the government, and that price information should e broadcast on the radio. In his view a reasonable price would be Tshs. 1,000-1,500 per susu in the peak of the orange season and Tshs 3,000-3,500 per susu when oranges are in short supply (more or less the same as the current price). He says that about 2-3 mausu per tree can be lost from bugs, the rotting of fruit, and theft.

1975 Farmer C, Ndijani Uwandani. In the mid-1960s he and an Arab friend worked together as orange middlemen, buying the picked fruits from local farmers and selling them in Zanzibar town. In early 1975 they started to lease orange trees for the harvest season from their owners, picking the fruits themselves before trading them. Sometimes tree owners went back on their verbal contracts, claiming to have agreed to lease out fewer trees than they actually had (for example 10 trees instead of the 14 originally promised). Frustrated by this he decided to grow his own oranges, and later in 1975 bought 2 acres of land at a cost of Tshs 600 for this purpose. He started out with a very small number of trees, 10-20 in all, planted in

both masika and vuli. Now he has 120-150 mature trees. One tree can produce 10-12 masusu of oranges per season, one or two of which may be lost as a result of pests, the rotting of fruit, and theft. He says that the current rotating marketing system was introduced in Ndijani and Binguni villages. He follows the system and waits for his turn to sell, but does not do this to get a better price, just to conform. In his opinion the future of the orange market is uncertain, and he thinks that ‘Boribo Muyuni’ mangoes should now be grown on the uwanda instead of oranges.

He also

suggests that the government should establish a multipurpose processing plant to absorb the increasing supply of oranges and other fruits.

Farmer D, Binguni / Uwanja wa Ndege. He has access to 8 acres of family and private land. He began to grow oranges in 1975 by planting 300 trees, a mixture of ‘Original Ndijani’ and ‘Budded Ndijani’. Between 1985 and 1990 he planted 350 of the same, and another 500 between 1990 and 1994, making a grand total of 1,150 trees. He has expanded orange cultivation to increase the profits he makes from sale of the fruit. Although he knows about the rotating marketing system he prefers to sell to middlemen who come from town

and pay what he thinks is the

reasonable price of Tshs 10-15 per fruit. Because he sells to middlemen he is not worried about supply to the market. He picks the fruit himself as soon as they are mature because if they are left for longer on the tree will not grow well before the next season. Each tree produces around 10-15 masusu of oranges per season, while 3-4 masusu per tree are lost because of pests and diseases and theft.

Farmer E, Ndijani Mseweni / Congo. He started to grow orange trees in 1975 because by then oranges had become the most important economic crop in the village. Over a period of eight years he planted 70 trees every other year: in 1975 (a mixture of all three types), 1977 (50 ‘Original Ndijani’ and 20 ‘Budded Ndijani’), 1979 (all three varieties), 1981 (all ‘Original Ndijani’), and finally in 1983 (variety unspecified), making a total of 350 trees. He did not plant any more after 1983 because he had no more land on which to do so. In the valley areas he loses 4-5 masusu of oranges per

tree each season because of the damage caused by bugs: sometimes all the fruits on a tree are lost. In addition to this 1-2 masusu of fruit per tree may be lost from rotting.

Farmer F, Ndijani Mtakuja.

He turned to orange farming because his

salary from paid employment was insufficient. He bought 3 acres of land for Tshs 1,500 and planted 100 orange trees on it in 1975 (a mixture of varieties). In 1976 he planted 50 more, and between 1977 and 1980 150 trees, making a total of 300. About 60% of these were budded tres, the remaining 40% ‘Original Ndijani’. He stopped planting on this farm after 1980 because he had no more land on which to expand production. These oranges are intercropped with jackfruit, coconut, mangoes and miembe ya kizungu (Golden Apple, Spondias cythera) and cover about 70% of the total cultivated area. In 1989 he was able to start another orange farm in Dunga which now has about 200 trees, all of them ‘Original Ndijani’. In all he therefore owns a total of about 500 trees. He was one of the first participants in the rotating marketing system, in the northernmost of the three zones which comprise it. When the market is flooded with oranges he has to delay picking his fruits. As a result about 30% of them fall to the ground, and they are then collected by his children and sold by them at the roadside. The fruits which remain on the trees are harvested and sold in Zanzibar town at the end of the season. He reckons that 3-4 masusu of oranges per tree can be lost, and all of the fruit if the tree is invaded by bugs.

1977 Farmer G, Ndijani Mseweni / Congo. According to him there were no oranges (mature trees) in this area in 1975. He started to grow oranges in 1977, planting 50 trees in masika and another 75 orange trees, followed by 200 (all ‘Original Ndijani’) in 1990, and 100 trees in 1991. He now has a total of some 675 orange trees. He estimates that 80% of these are ‘Original Ndijani’, 15% ‘Budded Ndijani’, and the remaining 5% ‘Budded Binti Juma’ (if it is true that he planted 50 of these in 1980 this last figure should be 8% or higher). Only 20% of the total are budded because they

have always cost more than Ndijani seedlings, and are more difficult to obtain. He picks and sells all of his oranges (though he did not specify how). During masika and especially at the end of the rains he may lose up to 15 masusu of oranges because of bugs (mabungu), and the equivalent of one susu of fruit because of poor picking practices (when the picker in the tree throws fruit down to an assistant who does not always catch them). He has not experienced any problem with crop theft.

1978 Farmer H, Ndijani Mseweni / Congo. Like Farmer G, he says that there were no orange trees in this area in 1975. In 1978 he planted 35 ‘Original Ndijani’ seedlings which he obtained from the Forestry Department nursery in Ndijani. In 1980 he planted a further 20 trees. He did not plant any more after this because he had no more land on which he could do so. If he can raise the capital from orange sales or other business activities he will buy land on which to plant orange trees.

1982 Farmer J, Binguni. In 1982 he planted 100 orange seedlings, obtained from the government nursery at Machui, together with 40 coconuts. In 1984 he raised his own orange seedlings and planted 40-50 on his farm. Thereafter he expanded production as follows: 1986, 30-40 trees; 1988, 30 trees; 1990, 30 trees; (plus 5 tangerine trees) 1991, 30-45 trees; (plus 20 tangerine and 10 jackfruit trees) 1992, 42-50 trees; and 1993, 40-50 trees. He now has a total of about 370 orange trees and plans to plant 50 more in 1995. He increased the number of trees gradually so that it would be easier to weed and therefore manage the farm. In 1985 he borrowed 20 cattle for the first time to manure and improve the fertility of the soil. He has also planted some tangerine trees (5 in 1990 and 20 in 1991) and jackfruit trees (10 in 1991) to insure against a possible future decline in the price of oranges. He reckons that he can lose a total of 4-5 masusu of oranges if he delays picking and up to 20 masusu from theft.

1983 Farmer J, Binguni. In 1982 he bought a farm for Tshs 11,000 and the following year planted 150 ‘Original Ndijani’ seedlings on it. He added another 30 trees of the same variety between 1985 and 1990. Every year he has lost some trees, 10 in all, because of fire invading from a neighbouring farm. So he now has a total of 170 trees. He usually lets other lease his trees for harvest and never picks the oranges himself because he is too busy and does not like the trouble of taking them to market. One tree can produce 12-15 masusu of fruit per season, while 4-6 might be lost. He thinks that the government should find an overseas market for oranges, should buy them itself, or set prices in conjunction with farmers and traders. He also thinks that the rotating marketing system (which he does not participate in) should be extended to the whole of Unguja island.

1985 Farmer K, Ndijani Uwandani.

He moved into orange production after

hearing a speech by the President in mid-1984 in which he advised farmers to grow other cash crops (including oranges, ‘Boribo’ mangoes, and coconuts) because of the falling price of cloves. Soon after this he bought 5 acres of land for Tshs 2,000 and in 1985 he planted 200 budded trees from the government nursery at Machui. Between 1985 and 1990 he planted another 300, and by 1994 he had planted a further 350 trees, making a grand total of 850. He has expanded production by reinvesting his profits from orange sales. He says that the rotating marketing system was introduced by large farmers, but does not help small farmers like him because Ndijani is not the only area which supplies oranges to the Zanzibar market. Nonetheless he participates in the system, and does so to conform rather than to secure better prices through delayed harvesting and marketing.

He thinks that delaying is unhelpful because a lot of

oranges can therefore be lost while a farmer is waiting for his turn in the system. He estimates that one tree can produce 8-10 masusu of oranges on average and that 2-3 are lost because of pests and the delay in picking. However, he does not want to lease out his trees because his children

would then be unable to collect the fallen fruits to sell at the roadside. The people who lease trees will not allow anyone to collect fallen fruits. They usually pick the fruits as soon as they are mature (and have had less opportunity to fall) because they do not take part in the rotating marketing system and are not therefore obliged to delay picking. He thinks that it would help if he government found an export market for oranges.

1986 Farmer L, Binguni. Before he become an orange farmer he used to take his father’s oranges to the market in Zanzibar town. When he saw the large amount of cash which this trade earned he decided to grow oranges himself on 2 acres of spare land owned by his father. In 1986 he planted 15 ‘Original Ndijani’ seedlings. Thereafter (year not specified) he planted another 120 of the same, and 200 more in 1990, making a total of 335 trees. He plans to plant another 500 trees in 1995. He participates in the rotating marketing system and especially at the end of the season he leaves mature fruits on the tree so that he can get a higher price. He estimates that one tree produces an average of 6-7 masusu of oranges and that 2-3 masusu can be lost from various causes including the delay in picking. He thinks that overseas markets should be found to solve the problem of oversupply.

1987 Farmer M, Tunguu. In 1986 he bought 4 acres of uwanda land for Tshs 200 and moved there from his natal home because of the growth of the family and the pressure which this put on their land. In the following year he planted 16 orange trees, another 16 in 1990, and 45 trees in 1992. The total of 77 is a mixture of ‘Original Ndijani’ and ‘Budded Ndijani’.

He

stopped planting because of the high price of seedlings and his decision to concentrate upon the management of his existing trees. In 1995, however he plans to start to raise his own trees from seed.

1988 Farmer N, Binguni. He was a late entrant to orange growing because previously he lived and worked in Zanzibar town. He left his government job as a printing engineer because of the low salary and dim pension prospects, having decided that he could earn a better living from agriculture. He bought some land for Tshs 1,700 in 1980 and in 1988 planted 30 ‘Original Ndijani’ seedlings on it. In 1990 he planted more trees (he cannot remember how many), and plans to plant 200 in masika 1995. He prefers ‘Original Ndijani’ to the budded trees because he believes that the latter have a short life and do not produce as well. He knows about the rotating marketing system but does not participate in it. Instead he leases out his trees when the fruits are mature and does not become involved in the picking. He leases out to anyone who is willing, whether they are from the village or Zanzibar town. He thinks that the rotating marketing system was introduced by people with many trees primarily in order to protect their own interests, because they could not all send to market on the same day. His own position is that this system should be removed, a step which he believes would ensure free competition and improve the market for and price of oranges.

31. While these case studies comprise a somewhat loosely assembled sample, and could have been expanded with a lot more detail (had time allowed), collectively they provide a number of important insights into the development of orange production and marketing and form a good starting point for further analysis. In the sections which follow we will look at the key issues which they raise (and some which they do not) in greater depth.

The Availability of Land 32. Land availability has been a factor of critical importance in the development of oranges as a cash crop in Ndijani.

Before 1964

settlement and agricultural production were concentrated in the plantation area, where the landscape was dominated by tree crops (including cloves and coconuts) and multi-storey cropping in the home

gardens around human habitations.

Land ownership followed the

typical pre-Revolution pattern of large estates interspersed with the smaller farms of squatters and others who also worked on these estates.

The nearby uwanda land was barely utilised, except for

sporadic shifting cultivation of annual crops, and was generally considered to have little potential for further agricultural development. 33. The Revolution of 1964 ushered in a major transformation in land ownership and, ultimately, in land use. Many large landowners left the area, and their estates were nationalised, broken up, and redistributed among smallholders. The new opportunities which this transformation created encouraged an influx of settlers from outside Ndijani, many of whom moved in from the coral rag villages to the south and east of this area.

Although the pattern of land ownership had changed, the

existing system of land use by and large persisted through the second half of the 1960s. Farmers in the home garden area, Ndijani viamboni, continued to cultivate the traditional plantation crops as their main source of cash, along with scattered plots of annual food crops. The greatest change took place in the large rice valley of Cheju, which lies to the south of Ndijani. The nationalisation of land in the valley and the development of mechanised and irrigated rice production under government supervision can be assumed to have had a considerable impact upon the livelihood of farmers who were able to gain access to the seasonally-allocated plots.

Rice production in Cheju certainly

provided an added stimulus to immigration into the area, seasonal or otherwise, and affected all of the villages around it, some of them more distant than Ndijani. 34. By the early 1970s, however, population growth had created something of a land shortage in the plantation area. This shortage was most keenly felt by young farmers who were entering agricultural production for the first time. Although they had access to family land owned and managed, sometimes jointly, by their fathers and other family members, the fact that it was often in short supply and subject to the competing interests of different relatives gave them relatively little freedom to develop it as they wished. This is a problem, in part a

function of the prevailing system of land inheritance, which can be found to a greater or lesser degree throughout the plantation areas of Zanzibar, and it becomes particularly acute when land becomes a scarce resource relative to the number of family members who have rights in it and no other option but to exercise them by farming. In this situation it may be next to impossible for an individual and especially a junior family member to plant permanent tree crops.

This is both

because there may be little space in which to plant more trees and because planting them may produce friction with other joint owners who perceive this as an attempt to stake an individual claim to part of their collective property. 35. The solution to this problem which was widely adopted in Ndijani was to move onto the hitherto underutilised uwanda, though for many farmers this may initially have seemed less of a solution than the only available option.

Buying land in the plantation area was another

possibility, but this land was more expensive and the available plots few and relatively small.

Orange growers who remained in the

plantation area soon ran into difficulties as a result. The case studies show that a number of small to medium producers stopped planting in the early 1980s because they had no more land on which to do so. Even some large producers had problems in converting their orange profits into more land in the plantation area. One successful farmer in Binguni related how, in 1984, he had tried to secure another farm in Jendele, about two miles from his home. The chairman of Jendele CCM branch agreed to sell him a 3 acre plot for Tshs 9,000, and presented him with a written contract when he handed over this sum (which he raised by selling a herd of exotic Jersey cattle for Tshs 7,000 and adding a further Tshs 2,000 from his orange income). However, the branch committee subsequently refused to confirm this allocation because he was not a resident of Jendele. Eventually they paid him Tshs 6,000 to return the land, by which time he had planted 55 orange trees on it. 36. Therefore throughout the 1970s and 1980s many farmers, and especially young farmers, started new farms on the uwanda and built

their homes there, creating a series of new hamlets away from the older plantation settlements. As a result the uwanda was transformed from the older plantation settlements. As a result the uwanda was transformed from an area of shifting to permanent cultivation. It is not entirely clear how individual rights in the uwanda were established, or exactly when this took place, though it seems to have happened quite early in some parts of Ndijani where a land market had already developed by the mid-1970s. In some cases existing use rights were converted into more permanent ones simply by continuous cultivation, the planting of tree crops, and finally by settlement on the land. This process can still be seen in its early stages in other parts of the coral rag on Unguja, where a land market has begun to emerge prior to permanent settlement, driven by people who have staked early claims and then sold them off before developing the land further. In other cases the local party branches seemed to have played a role in allocating and even selling land, drawing their mandate from the government’s de jure ownership of the uwanda and the relative weakness of existing de facto claims. 37. Apart from the independence of decision-making which it afforded farmers, the advantage of owning land on the uwanda was that there was plenty of land on which to expand orange production. The case studies illustrate vividly how many farmers progressively planted more and more orange trees, using their profits from the crop to do so. Those with insufficient land and the capital to do so purchased additional farms. Prices in some areas seem to have remained quite low through to the mid-1980s, ranging between less than Tshs 100 and Tshs 500 per acre. More recently, however, there has been a dramatic jump in the cost of land, even allowing for the post-liberalisation devaluations of the Tanzanian shilling. In mid-1995 the going rate for an acre of uwanda land in Ndijani Mseweni was said to be between Tshs 100,000 and Tshs 150,000. Two main causes can be posited for the recent inflation of land values in this area: first, because there is no more uncultivated uwanda left; and second, because the land itself has been improved considerably by manuring and management practices.

38. The filling-up of the uwanda and the current prohibitive cost of land in areas where it has already filled up will make future expansion of orange production there increasingly difficult. Farmers have already begun to move onto poorer maweni (coral rag) soils, and it is possible that history will repeat itself if these can be improved. Nonetheless it seems unlikely that orange farming will be the starting point for this improvement in the way that it was on the uwanda, where orange farmers are currently in the process of diversifying into other cash crops.

Land Improvement 39. In many ways oranges were a fortuitous choice of crop for growing on the uwanda. They readily fostered the development of a mixed farming system which has resulted in an increasing improvement of soil fertility, while

the

orange

trees

themselves

have

contributed

to

this

improvement by providing shade and organic matter from their leaf-fall. The evolution of an efficient mixed farming system has been the key to the sustained development of agricultural production and the further diversification which is now taking place in Ndijani.

This system is

described in greater detail in ZCCFSP’s study of Ndijani (Mseweni) Farmer’s Research Group, Mixed Farming and Improving Fertility on the Uwanda of Central Unguja, and only the main points need to be summarised here. 40. The soils of Unguja, indeed Zanzibar as a whole, are for the most part poor in structure and deficient in minerals.

In common with many

tropical soils of non-volcanic origin they are only as good as the top layer of organic matter which vegetation provides them with.

The

balance is a fragile one, and in the absence of adequate vegetation cover (or artificial fertilisers) soil fertility can decline very rapidly, with disastrous consequences for agriculture. exception.

The uwanda soils are no

Although the uwand is sometimes characterised as

‘grassland’ and ‘low bushland’, it seems likely that it originally supported woodland or forest, and tthat its present appearance when uncultivated is the result of more than a thousand years of shifting

cultivation by Unguja’s indigenous farmers. The implications of this are that the uwanda can support tree crops, and in fact might do a lot better by them than by periodic firing and the cultivation of grains and other annuals. 41. However, orange trees cannot perform this role alone, certainly not in short period of time and when the land between them has to be kept clear of weeds or is cultivated with annual crops. Ndijani’s farmers have (almost literally) filled the gap with cattle and their manure. How this came about is unclear: presumably some cattle were kept on the uwanda when it was under shifting cultivation, but their value as fertilising agents only came to the fore when farmers started to corral them on their farms and graze them between their orange trees. Informants are agreed that manuring only became important after orange production had been established.

Thereafter cattle-keeping

became an integral part of the farming system, to the extent that farmers in Ndijani do not feel the conflict between cattle and crops which forces farmers elsewhere on the islands to limit their livestock holdings or keep them away from cropland and in the coral rag bush. 42. A high proportion of the orange farmers in Ndijani own cattle, having purchased them with income from their harvests.

Relatively few

owners, however, both manage and herd their own animals. Small herds are often given to others to manage and/or herd, while large herds may be split up and similarly parcelled out. These arrangements may involve different kinds of payment, often in kind (milk and calves). Farmers without cattle are frequently able to borrow them for manuring. Otherwise a herd can be hired for this purpose: in early 1995 the going rate in Ndijani Mseweni was between Tshs 10,000 and Tshs 20,000 for a 1-2 month hire, depending on the exact period and the number of cattle hired. Farmers interviewed said that ideally 25-30 cattle should be grazed on an acre of land before both the masika and vuli rains, though most farmers are unable to manure all of their land so intensively. Nevertheless they report that as a result of manuring soil fertility has increased on the uwanda. This is in marked contrast to farmer’s experience elsewhere in Zanzibar, where continuous cropping

without fertilisers (natural or artificial) has resulted in a noticeable decline in fertility and therefore yields in recent years. 43. Other fertility management practices employed on the uwanda include mixed cropping and crop rotations, and to a lesser extent fallowing. A variety of root crops and vegetables are grown between the orange trees, providing additional income outside of the orange seasons. Manuring is usually concentrated on the most responsive crops, because there are not enough cattle in Ndijani to manure every plot every year.

While it is difficult to assess the precise impact of

manuring on orange farming per se, it has certainly enabled farmers to develop an efficient system around it. Agricultural diversification has become an important part of this system, helping to support orange production as well as providing alternatives to it. If, as some farmers predict it will, the market price of oranges falls significantly, then many of them will be well positioned to exploit other opportunities.

The Availability of Labour 44. It is difficult to imagine that orange production could have expanded as it did and the uwanda developed as it has without an adequate supply and inputs of labour. The fact that this development was driven by local farmers, many of them young men, is significant here. ZCCFSP studies in different parts of Zanzibar suggest that, as a group, young men (especially those in their twenties and thirties) tend to be the most innovative in exploring new sources of income, both on-farm and offfarm. There are a number of reasons for this. The initial impetus derives from their need to secure economic independence and raise the cash for bridewealth so that they can marry.

Once they have

established their own households, then the need to provide for these in their critical phase of growth (as children are born and place increasing demands on household resources) further exercises the energy and ingenuity of the men who head them. While individual heads of mature households can be equally innovative and energetic in their pursuits, many of them are relatively conservative and content to consolidate the position that they have secured for themselves in their own youth.

45. This conservatism is most marked in the plantation areas of Zanzibar, where opportunities for agricultural development are limited by a number of factors, not least of which is the difficult of removing ‘permanent’ tree crops and replacing them with something else. As already mentioned above, the prevailing pattern of land tenure in these areas acts as a further constraint. In this situation one of the most attractive options for young men is to seek off-farm employment, and to move into town (whether Zanzibar town or elsewhere) to do so. However, in places like Ndijani which are on the edge of the plantation area and border underutilised coral rag (uwanda or maweni) land, the ready availability of this land provides an alternative to labour migration. The successful development of this land by young farmers in Ndijani appears to have largely stopped and even reversed the flow of male labour from farms to the town. Whether this will continue to be the case remains to be seen. It is possible that the current shortage of cheap

uwanda

land,

coupled

with

growing

opportunities

for

employment and enterprise in Zanzibar town, will result in many members of the next generation moving out of farming. 46. To date the growth of the orange industry and of a mixed farming system around it have created a wide range of local employment opportunities. In addition to providing agricultural labour of different kinds, an expanding network of backward and forward linkages has developed which embraces small-scale manufacture (for example basket-making, to provide the masusu which oranges are marketed in), the provision of services (including ox-cart hiring) and various trading enterprises (from shops through to the activities of middlemen engaged in the marketing of oranges and other farm produce).

The further

development of these off-farm linkages and the jobs which they create could to some extent counter the lure of urban employment for householders who have little land on which to produce cash crops, and no doubt this is already happening. 47. Unpaid family labour is only used to a limited extent in orange production, most often on the smaller plantations.

Farmers usually

plant their own seedlings (and in some cases seeds to be later

transplanted), while all of the members of a household, including women and children, may take part in picking the fruit. Sometimes friends and neighbours are also called to help in harvesting oranges, in which case they may be given some cash reward after the crop has been sold, though not a predetermined sum. As noted in the case studies, children are sometimes allowed to collect fallen fruits and sell them at the roadside. Otherwise many farmers take the main crop to market themselves, though they usually pay for them to be carried to and from the bus, in addition to the cost of transporting the oranges by bus and the farmer’s fare for the journey. 48. Paid labour has played a critical role in the development of orange production in Ndijani, and continues to do so. Most uwanda plots were cleared by hired labourers, who also do most of the weeding and harvesting on all but the smallest plantations. Most of this paid labour is and has been provided by Nyamwezi (and Sukuma) men and women, originally from western Tanzania. Although some local Swahili men will also perform these tasks for payment, most orange-growers prefer to employ labourers of mainland origin because they have a reputation for being more hard-working, work for lower rates than the locals, and will often accept delayed payments for their labour until after the crop has been sold. In May 1995 farmers in Ndijani Mseweni quoted the following rates for weeding orange farms: between Tshs 12,000 and Tshs 15,000 per acre for a pure stand of trees, and as little as Tshs 3,000 per acre for oranges intercropped with bananas, sweet potatoes and/or vegetables. Labour rates are generally highest during masika, when the demand for labour is greatest. Harvesting oranges was said to cost between Tshs 1,400 and Tshs 2,000 per acre for a single harvester, each acre of oranges requiring four to six harvesters (giving a total outlay in the range of Tshs 5,600 to Tshs 12,000 per acre). 49. Nyamwezi labourers have been an important feature of the rural economy in Ndijani and other parts of Unguja, especially in the plantation areas, since long before the 1964 Revolution. It is difficult to assess what proportion of the present population is permanently settled

on the island. The Nyamwezi in Ndijani are certainly highly mobile, with the exceptions of individuals who have bought land and/or married into the local community, and to some extent have adopted a Swahili identity.

They often arrive in Ndijani as couples, sometimes with

children, and are invited to settle on the land of local farmers, where they are allowed to build temporary houses and cultivate their own plots.

Technically they are squatters, though always with the

agreement of the landowner, who can expect to benefit from having favoured access to their labour (which they still pay for) and from the security which they provide. The majority of them leave after two or three years, ostensibly to take their earnings back to the mainland, though it is said that many of them simply move elsewhere in Unguja. The apparent reason for this pattern of mobility is their wish to avoid greater integration into the community, which would have the effect of reducing their ability to accumulate cash as they became enmeshed in local networks of social and economic obligation and all that this entails for resource-sharing. Farmers in Mseweni estimated that at any one time Nyamwezi comprise around 25% of the total population of the area. It is some measure of their importance to the local economy that orange-growers freely admitted that without their labour inputs it would not be possible to sustain the current farming system.

Social and Economic Differentiation 50. The intensification of agricultural production, and especially production for the market, is often associated with increasing social and economic differentiation, and in this respect Ndijani is no exception.

The

development of orange production has seen the emergence of a pattern of differentiation which is absent or only very weakly developed in most other rural communities in Zanzibar, though it can be predicted that sooner or later they will follow suit. This might be considered as a negative impact of cash crop development, particularly in view of the egalitarian ethos which has prevailed since the 1964 Revolution, though in many ways it would seem to be inevitable (much as

economic

liberalisation

has

fuelled

the

process

of

economic

stratification in urban Zanzibar). 51. At the risk of over-simplification, it is possible to identify three different strata which have emerged or are in the process of emerging in Ndijani.

These are the large landowners (and orange-producers),

small landowners, and the shifting pool of agricultural labourers. There is no sharp boundary between the first and second of these categories, which comprise interrelated members of the same ethnic group and others who have assimilated to it. They are, however, quite clearly distinct from the third category, and this distinction is reinforced by their separate ethnic origins.

In economic terms the key factors

distinguishing all three strata are their differential access to land and their different roles in the labour process.

Large and Small Farmers 52. Wealth differences (in terms of property ownership) in rural Zanzibar frequently correlate with the stag which a household has reached in its developmental cycle. These differences are also commonly negated by the social and economic integration of individual households within wider groupings of kin, as a result of which the individual ownership of resources may be more apparent than real in terms of the number of people who can claim access to them. To the extent that this is the case wealth differences may be neither permanent nor perpetuated, though at any particular time certain households and certain lineages may command more resources than others. In Ndijani, however, this pattern has been significantly altered, at least for the time being. 53. The uwanda was opened up by young male householders in response to their need for resources in a context in which these had become severely limited in the overpopulated plantation areas of Ndijani. The planting of oranges and the improvement of the uwanda, however, transformed it into a valuable resource which was individually owned from the very start.

For successful orange-growers and their

households this marked a break in the normal pattern of development, which would typically only have enabled them to accumulate resources

gradually as their households grew and matured. As we have seen, farmers who remained in the plantation areas have experienced considerable difficulty in expanding their farms and orange production. As uwanda land too has become scarce, and land prices have risen, differences in the area of land owned and planted with oranges and other cash crops have become critical, and can be expected to become an even more important factor in the future. 54. Farmers themselves frequently distinguish between large and small orange producers.

The largest individual holdings of land on the

uwanda are said to reach up to 20 acres or more in total (combining separate plots), while large orange farmers may possess 800-1,000 plus trees. One farmer in our sample had planted 1,150 trees on 8 acres (an average of 144 trees per acre); another 850 trees on 5 acres (an average of 170 trees per acre); and a third 830 trees on 13 acres (an average of only 64 trees per acre).

The latter farmer stopped

planting oranges in 1990 because of his fears for the future of the orange market. With the exception of one farmer with 675 trees, the rest of the orange-growers in our sample had fewer than 400 trees, and three or four of them less than 100. These smaller orange farmers typically owned 2-3 acres of land suitable for orange production, and a number of them had stopped planting oranges because they had insufficient land on which to expand production. It should be noted here that our sample was biased in favour of large-scale producers, and we can safely assume that they form a rather smaller proportion of the total population of farmers than these figures imply. In absence of data from a large and random sample of Ndijani farmers, however, it is difficult to be more specific on this point. 55. This picture is complicated by the fact that in recent years many orange-growers on the uwanda have diversified into other kinds of cash crop production as well as other types of economic activity. In Ndijanii Mseweni vegetables, root crops and irrigated rice provide farmers with alternative sources of income which in some cases have become more important than oranges. Livestock-keeping is another important source of income for some farmers. The wealthiest farmers

have also begun to invest their profits from farming in off-farm enterprises. These include the establishment of roadside shops and the construction of houses for rent in Zanzibar town. Ownership of orange trees is therefore not the only criterion by which the most successful class of farmers can be recognised. The total acreage of land owned by an individual, including uwanda and valley (reinfed and/or irrigated) land, provides a more reliable guide. 56. One of the most striking features of the stratum of wealthier farmers in Mseweni is their relative youth. Some of the richest farmers are aged in their late twenties and early thirties. Another way in which they have invested their agricultural wealth has been to marry two and in some cases three wives. This contrasts with the more common pattern in Unguja whereby men are usually only able to marry additional wives at a somewhat later age, when they have accumulated sufficient resources to do so.

In deed most of the fathers of Mseweni’s

successful young farmers still have only one wife. Polygny is a clear mark of their sons’ status. When asked why he and others like him had married more than one wife, one young farmer frankly admitted that it was purely for “pleasure”, brushing aside any suggestion that this practice might have an economic motive. Nevertheless, such motives can be reconstructed: additional wives may ease the management of widely separated land holdings (when, for example, these are in both plantation and uwanda areas), while polygyny doubtless also functions to create wider networks of potentially useful social and economic alliances. 57. On the other hand it is possible that these polygynous farmers may thereby be sowing the seeds of their own future decline, or at least that of their descendants.

Without more land to expand onto, one

consequence of this marital practice may be the accelerated fragmentation of their land holdings among large numbers of offspring. In the long term the gap between large and small farmers might therefore be narrowed, and the current process of differentiation reversed. It is easy, however, to imagine alternative scenarious: for the time being all of these remain purely speculative. One possibility is,

perhaps, worthy of further investigation at present. In Mseweni there was some suggestion that aside from individual differences in wealth, the benefits of cash crop development had spread through at least one local kinship group (ukoo) in its entirety, marking it off from other local groups. If this has indeed happened, then it might be interpreted as evidence

both

for

the

persistence

inter-household

levelling

mechanisms as well as for the emergence of an inter-lineage (and therefore more readily reproducible) pattern of differentiation.

Squatters and Agricultural Labourers 58. As noted above, the social and economic boundary between local landowners and the labourers of mainland origin who are allowed to squat on their land is much sharper, though not entirely impermeable. Although a landless class has yet to appear among the indigenous population of Ndijani, all but a handful of Nyamwezi (those who have bought land) do not possess land of their own in this area, but are obliged to borrow it from their hosts. The greater part of their income derives from the sale of their labour to local farmers, and very little from the sale of crops grown on their borrowed, which are cultivated primarily for subsistence purposes. In other respects, however, the Nyamwezi do not form a landless class of labourers in the normal sense.

On their own account temporary settlement in Ndijani (and

elsewhere in Unguja) comprises part of a strategy of accumulation which, at least in theory, is completed when they return with their earnings to their homes in western Tanzania.

What subsequently

happens to these earnings, and indeed how likely they are to make this journey, are subjects which remain to be clarified. It seems that a significant number of the Nyamwezi on Unguja are more or less permanently resident, in Zanzibar town as well as in the rural areas. 59. The Nyamwezi (and Sukuma who fall under the same label) have performed a similar role in the economy of Zanzibar since at least the early colonial period, adapting to the changing political and economic circumstances of the islands.

A detailed history would presumably

show significant shifts in their role as agricultural labourers, especially

following the abolition of slavery and, more recently, as a consequence of the nationalisation and redistribution of the larger estates.

The

development of the uwanda and cash crop production in Ndijani has provided a new set of opportunities in this particular area of Unguja, and the transient population of mainlanders there has increased accordingly.

We have no evidence to suggest that this has either

worsened or improved their social and economic position, which to some extent is the result of their own economic choices (hence the decision of most Nyamwezi to remain in Ndijani for no longer than two or three years, allegedly in order to escape further integration within the local community and the likely consequences of this for their ability to earn and accumulate cash). Our data on the Nyamwezi, however, are largely second-hand, and it is quite possible that more thorough research on their economic strategies will reveal factors which we have overlooked.

Gender Differentiation 60. The pattern of stratification which we have outlined above is crosscut by another, based on gender differences. Women play a very much secondary role in the farming system of Ndijani. While some local women do own and cultivate land of their own, the majority of them do not, but only provide labour inputs on farms which are owned, borrowed, and otherwise managed primarily by their husbands or other male relatives. These labour inputs tend to be limited to particular operations (especially weeding and harvesting) and/or focused upon particular crops (especially the cultivation of irrigated rice and the small-scale production of vegetables for domestic consumption). Women in resource-poor households which cannot afford or do not have the need to employ hired labourers typically perform more agricultural tasks than those in wealthier households, though most important farming decisions are still made by men. Male farmers in Ndijani Mseweni declared that women’s primary place is in the home, performing domestic chores, rearing children and undertaking smallscale income-generating activities (such as kofia-making). While in the

field we were presented with graphic evidence of this by farmers’ reluctance to involve their wives in interviews and their keeping them more or less hidden from our view (including that of a woman interviewer). 61. We did not come across any cases of women who had planted oranges, though an estimated 2% of all the orange trees in Ndijani are owned by women who had inherited them.

Oranges have been

developed as a cash crop and the uwanda opened up and therefore claimed by men. As might be expected, women play no part at all in the marketing of oranges (or indeed of other cash crops which are sold outside of Ndijani), though they do provide some of the labour in weeding and harvesting the crop. It will be interesting to see what happens over time as women gain more ownership rights in orange trees as part of their share in the inheritance of farms.

Unless

inheritance practices are modified, it seems likely that events will follow a similar course to that which can be traced for mango tree ownership in Muyuni. Over the past 100 years women’s ownership share has increased from as little as 5-10% (estimates vary) to around 30% of the total number of trees.

The trees owned by women, however, are

almost exclusively managed by their male relatives, who in many cases lease the trees out for harvest. Women’s share of the proceeds is therefore less than it might be, and their income from mangoes is often superseded by that from other sources over which they have more direct control, such as the collection and sale of firewood (for a more detailed account see ZCCFSP’s report on The Development of Mangoes as a Cash Crop in Muyuni, Unguja). 62. A comparison of the experiences of Ndijani and Muyuni indicates that in the initial phase of development (which in the case of Ndijani has already lasted for a quarter of a century) women’s rights of ownership in the new tree crop are negligible. Gradually these rights increase, though the management of the crop remains largely in the hands of men, and women’s direct access to the income from it continues to be limited as a result. On the other hand, women can and do benefit indirectly from the development of tree crops owned and/or managed

by their husbands and other male relatives. In Ndijani the wives of successful orange growers are spared from bearing the burden of agricultural labour, most of which is performed by their husbands and the labourers which they hire. From this point of view the ability to stay at home (and concentrate upon domestic tasks and other off-farm income-earning activities) is a mark of improved (and urbane) status, and widely considered to be so. A large proportion of the Nyamwezi labourers who work the fields in their place are, nonetheless, women and children. We have no information on how the development of cash crop production in Ndijani and elsewhere has affected their status. It is perhaps significant that Nyamwezi women who become the wives of local men are said to switch roles quite readily, as far as possible withdrawing from agricultural labour (on the family farm) and confining themselves to the home.

The Evolution of the Marketing System 63. Farmers employ a number of different strategies for selling their oranges. These include (1) the seasonal leasing out of trees and their harvests to middlemen, who do the picking themselves before taking the fruits to market; (2) the sale of picked fruits to middlemen; (3) the transport of picked fruits to the town market by farmers themselves, without the mediation of middlemen; and (4) an organised version of the latter, whereby farmers take turns to harvest their fruits and take them to the town market. This last strategy is designed to counter oversupply in the market during the peak harvesting season, and enable farmers to take advantage of the higher prices which oranges command at the end of the season. 64. These four main strategies differ in the extent to which orange growers themselves participate in the marketing process and the degree to which they can control this as a means of obtaining higher prices for their crop. The first of the strategies listed above gives farmers the least leverage and, unless they have negotiated shrewdly, the smallest return for their crop. Potential returns increase when farmers pick their own crop, and more so when the role of middlemen is eliminated. The

fourth strategy, participation in Ndijani’s rotating marketing system, offers, at least in theory, the highest possible returns. As the case studies indicate, farmers tend to adopt a single marketing strategy, though they may switch strategies over time.

To some extent

marketing strategies mirror the pattern of differentiation (large versus small farmers) outlined in the preceding section, and a correlation can be traced between the volume of an individual farmer’s production and the strategy adopted. At the same time, however, and especially in the case of participation in the rotating marketing system, there is some mismatch between farmers’ practice and what they feel to be their best interests. We will examine this mismatch, which emerges quite clearly in the case studies, in greater detail below.

Leasing Arrangements 65. The seasonal leasing of tree crops has a long history in the plantation areas of Zanzibar. It is often the preferred strategy of farmers who only have a modest number of trees or who are unable to play an active role in managing, harvesting and selling the crop, as is often the case with absentee owners and women owners. Given that there are very few absentee or women orange farmers in Ndijani, we should expect to find that this practice is most common among small-scale producers. Our farmer case studies support this conclusion: the three farmers who leased out their crops (one of them only in some seasons) all had less than 200 orange trees in total. An estimated 20% of all orange growers in Ndijani lease out their trees, the vast majority of whom presumably have such relatively small holdings. 66. Although leasing out generally gives the farmer the lowest returns, it makes economic sense for those who expect to derive relatively little income from oranges (for whom the proportional loss of income is not so significant)), and whose agricultural labour and cash resources are therefore best invested elsewhere. At the same time farmers who have entered into leasing agreements may subsequently wish that they had not done so, especially when harvests and/or prices are much better than they had expected. In this situation some farmers are not averse

to backing down on earlier verbal agreements, as one of our interviewees (Farmer C, who leased trees in the mid-1970s) discovered to his cost. One solution to this problem, at least for the middlemen, is to ensure that leasing agreements are written down and witnessed.

However, we have no more information on this point,

though we would expect that written contracts have become the norm, especially when large numbers of trees are involved (certainly more than the 14 which our informant gave as an example). 67. Those who lease trees also risk paying more than they would otherwise for a poor crop. Presumably, however, the rates they offer take account of this, and by the same token they can also gain from a bumper crop.

In order to maximise harvests leasing agreements

extend to fallen fruits as well as those on the tree, while the latter are picked and taken to market as soon as they are mature.

Lessees

employ their own pickers and transport the fruits to town themselves. Most of them are local farmers who may also act as ordinary middlemen, buying oranges which farmers have already picked.

Sale to Middlemen 68. Some farmers pick the fruits themselves (or hire people to do it for them) and then sell to local middlemen. This practice is less common than either the leasing out of trees (a strategy of minimal labour investment) or picking and taking them to the market (to gain maximum returns from a much greater labour investment). It seems likely that it was more common before and during the early stages of the expansion of orange production, when fewer farmers had the experience or the need to seek their own markets outside of Ndijani. Now, however, only an estimated 5% of orange farmers pick and then sell their oranges to middlemen. 69. Only one farmer in our sample said that he sold his crop in this way. Interestingly enough, this was the farmer with the largest declared number of orange tree (1,150). He harvests as soon as the fruits are mature because, he says, delayed picking affects tree growth in the following season. As a result he does not participate in the rotating

marketing system. Although he did not say so, it is possible that he prefers to sell to middlemen because of the amount of his own and time and labour which he would otherwise have to invest in taking such a large crop to town. He said that he is quite happy with the price offered by middlemen, in the range of Tshs 10-15 in 1994, suggesting that he gets a large enough income from the crop to be relatively unconcerned about any extra sum he might earn by taking it to market himself. 70. Middlemen sometimes select the fruit they buy and pay a premium for it, otherwise they buy all of the fruits which have been picked, at a lower price (one example was given of Tshs 12 per selected orange and Tshs 10 per orange of variable quality). None of the middlemen in Ndijani participate in the rotating marketing system, but derive their profits primarily from the difference between farm-gate and market prices during the peak season.

Sale Direct to the Market 71. An estimated 70% of Ndijani’s orange growers pick and take their own crop to the Zanzibar market. This includes a mixture of both small and large farmers.

They incur a number of costs in the process.

The

following list shows prices quoted in late 1994 (the main harvest season being June to September):

Hire of orange pickers:

Tshs 100 per susu

Produce permit (cheti):

Tshs 50 per consignment

Transport by ox-cart from farm to roadside:

Tshs 50 per susu

Transport by bus to Zanzibar town:

Tshs 70 per susu

Return bus fare for farmer:

Tshs 400 per trip

This list excludes the cost of the masusu themselves.

It might be

added that the cost of produce permits, which are obtained from the local sheha, is said to vary somewhat, and may be as high as Tshs 200.

72. At the same time (in 1994) the following range of prices were paid to farmers in the central market for their oranges:

Start of season:

Tshs 800-1,000 per susu

Peak season:

Tshs 1,000-1,500 per susu

Late season:

Tshs 1,500-3,000 per susu

The low prices obtained at the start of the season presumably reflect the fact that early-maturing oranges are generally of lower quality than the main crop. These and other figures we recorded in the course of this study do not provide a rigorous enough basis on which to analyse the costs of production in detail, and we will not attempt to provide such an analysis here. If they wish, however, readers can get some idea of the sums involved by relating the figures quoted in this section to the production data which appear in the individual farmer case studies.

73. Farmers who pick their own fruit but who do not participate in the rotating marketing system usually pick them as soon as they are mature. They do not therefore make any special attempt to sell at the end of the season when prices are highest. Why this should be will be discussed below, when we look at the rotating marketing system in more detail. Many farmers do, however, pack their fruits according to size, separating the larger from the smaller oranges to conform with the preferences of their town buyers.

Orange awareness of the urban

market and its requirements has certainly increased as orange production has expanded.

Marketing has also become easier as

transport links have improved between Ndijani and Zanzibar town. In the past few years, for example, buses have started running to and from some of the uwanda settlements. In part this improvement has been brought about by growth of the orange industry.

In part this

improvement has been brought about by growth of the orange industry. In turn it has increased farmers’ willingness to market their oranges themselves.

74. Although most of the oranges grown in Ndijani are taken to Zanzibar town market, some of them are still marketed locally. Farmers who pick their own fruits, including those who participate in the rotating marketing system, often make use of fallen fruits by allowing family members, especially children, to gather them and sell them direct to consumers at the roadside in Ndijani. An estimated 5% of farmers sell all of their crop in this way, all of them small-scale producers. This practice predates the development of oranges as an important cash crop, and was once much more common than it is now. We did not come across any cases of farmers bartering their orange crop for other produce or goods, which they are also said to have done in the past. Oranges are also consumed directly by farmers and their families, and are presumably of some nutritional importance in the local diet. Leasing arrangements are unpopular among some small farmers because they restrict domestic consumption and the sale of fallen fruits, despite their other advantages.

The Rotating Marketing System 75. A large number of orange growers who take their own produce to the urban market now do so according to an organised rota. This is said to have been introduced in recent years on the initiative of large-scale producers, including Maulid Mussa and Abdallah Yussuf of Binguni. Farmers participating in the rota are assigned to one of three zones, and each zone is allotted two consecutive days during which the farmers in it can take their oranges to market before the next zone has its turn. Farmers who have just had their turn must therefore wait four days before they can sell their produce again. Sometimes, however, if a particular zone has a large harvest waiting for market, then its two days can be extended, with the effect that farmers in other zones have to wait a little longer before they can sell their oranges. 76. The object of this rotating marketing system is to prevent oversupply during the peak season and therefore ensure that prices do not drop as much as they otherwise might at this time. It also has the effect of increasing the supply of oranges towards the end of the season and

encourages farmers to delay harvesting so that they can take advantage of the higher prices which they then command (given that the supply still does match that of the peak season, when many farmers and middlemen who do not participate in the system sell most of their oranges).

As noted in an earlier section, ‘Original Ndijani’

oranges grown from seed are particularly well suited to this system, because they can be left on the tree after they have matured much longer than other varieties without suffering the same degree of loss. Nonetheless, and as the case studies make clear, delayed picking does result in higher losses than would otherwise occur, and not all farmers are convinced that the higher prices they obtain at the end of the season will compensate them for this loss. 77. A number of interviewees voiced the opinion that the rotating marketing system works most to the advantage of large-scale producers (though, ironically, one of the farmers who said this and counted himself as a small farmer in fact had 850 trees). Our data make it clear, however, that both large and small-scale orange producers participate in the system.

Judging by their comments some of these farmers only

participate in order to conform, and not because they think it gives them any economic advantage.

It seems then that the rotating

marketing system has not captured the minds of all of its participants, much as it has failed to incorporate all of the farmers and middlemen who are content to operate outside of it. However, given the extent of participation in the system, especially by large farmers, it has clearly had a significant impact upon the market and thereby helped to sustain the expansion of orange production. 78. It is difficult to predict what will happen to this system in the future, and there are some suggestions that it is in the process of breaking down. The fact that a large number of farmers are engaged in orange production in Ndijani and have been for a relatively long period of time makes consensus particularly difficult to attain. This contrasts with the recent experience of egg-plant producers in Gamba in northern Unguja, where a rotating marketing system has developed for this new cash crop and is continuing to expand without any apparent dissent (for

further details of this case see ZCCFSP’s report on The Expansion of Cash Crop Production and Development of Marketing System in Northern Unguja) 79. The basic problem seems to be one of seasonal oversupply.

It is

difficult to determine whether this situation has arisen because a lot of farmers do not participate in the rotating marketing system or in spite of the fact that many of them do. Ndijani farmers point out that the urban market is also supplied with oranges from other areas of Unguja which are not subject to organised marketing systems, thereby further undermining the impact of their own system.

Farmers who are

dissatisfied with the Ndijani system suggest alternatives which would not necessarily be any better. One of these is a call for official price controls managed by the government and set following the advice of farmers. This and other farmers’ comments suggests that many of them have an incomplete understanding of the marketing strategy which lies behind the rotating marketing system, which is designed to achieve the same result without government intervention. The call for the development of markets outside of Zanzibar seems to be a more sensible proposal, though it remains to be seen whether Ndijani oranges can complete against other varieties in the region.

They

certainly have no obvious future on the international market. 80. Many Ndijani farmers have already indicated that they have no faith in the future market for their oranges by switching their attention to other cash crops.

These include a range of vegetable crops (which are

intercropped with oranges) and, most recently, cassava, which is now being grown on land in the Cheju valley formerly planted with rainfed rice.

As the case studies illustrate, farmers have also begun to

experiment with other tree crops on the uwanda, including ‘Boribo Muyuni’ mangoes. It may well be that the orange industry in Ndijani has reached a point beyond which it will be very difficult to develop further.

Whatever the case, orange production has had a wide-

reaching impact upon the farming system of Ndijani, and provides one model for the sustainable and dynamic agricultural development of the uwanda land of Unguja.

Conclusion: Lessons Learned from the Study

81. The following is a summary of the principal lessons which we think can be drawn from this study. 82. The development of oranges as a cash crop in Ndijani was achieved by local farmers with minimal inputs from the government agricultural research and extension services. Extensive efforts during the colonial period to improve orange germplasm had no direct impact upon this development, which was based upon an indigenous variety of orange which was already being grown. The two budded varieties of orange which farmers have adopted have been less important, and their introduction played no part in stimulating the much later expansion of cash crop production.

Nonetheless government nurseries have

provided an important source of planting material for farmers. 83. The most important factor in the expansion of production from the early 1970s onwards was the opening up of new land on the uwanda. Oranges were planted on the uwanda as a known cash crop with a growing market. Farmers invested the profits from their orange sales in improving the land further, and thereby evolved a mixed farming system in which cattle play a key role in increasing soil fertility. In turn this sustained the further expansion of orange production and, more recently, it has fostered diversification into other cash crops. Cash crop production is currently expanding throughout the uwanda and other coral rag areas of Zanzibar. Ndijani provides a model of how this can be done sustainably. While other tree crops can take the place of oranges, the presence of tree and livestock in a mixed farming sysem are essential in addition to the cultivation of short-term vegetable and root crops. 84. This expansion was driven by young men seeking better livelihoods than they could obtain by remaining in the overcrowded plantation areas, where their rights to land and economic independence were severely restricted. As population increases, this process is likely to repeat itself throughout Zanzibar, and to some extent is already

happening.

One implication of our analysis is that young male

household heads are potentially the most innovative of all groups in the farming community, and that they are most likely to innovate outside of the plantation areas.

Cash crop development programmes should

therefore pay particular attention to targeting this group as opposed to older men. 85. The development of orange production in Ndijani has fostered a pattern of economic differentiation in the form of a growing distinction between large and small farmers. It has also reinforced the sharp boundary between them and squatters and agricultural labourers of mainland origin. To a certain extent this seems to be inevitable, and mirrors a phenomenon which is recognised as widespread and by no means confined to Zanzibar. This suggests that cash crop development and the equitable distribution of its benefits may e irreconcilable goals, though measures can be taken to ensure that benefits are spread as widely as possible. There is still, however, a lot that we do not know about social and economic differentiation in Zanzibar and how this is developing and might develop in future. The economic position and strategies of labourers from the mainland are very poorly understood. In the short-term, therefore, further research is required to find out more. 86. Women have played relatively little direct role in the development of oranges as a cash crop in Ndijani. Only a few women own orange trees which they have inherited, though their numbers can be expected to increase over time. Otherwise most of the decisions about orange farms, while women of mainland origin form a large component of the hired labour force. Women’s access to resources relative to men’s has certainly declined, though the extent that some women benefit from their husbands’ incomes and have been able to withdraw from agricultural labour their status has improved. However, we know that gender relations are structured quite differently in different areas of rural Zanzibar, the most apparent contrasts being between the plantation areas and the villages on particular kind of crop, about the

impacts of cash crop development upon women’s social and economic status. 87. The marketing strategies adopted by men often reflect the importance of their holdings. An estimated 20% of mostly small-scale producers lease out their trees, while another 5% harvest and sell the fruits at the local roadside. Relatively few farmers (5%) pick and sell their crop to middlemen. The vast majority, an estimated 70% of farmers, harvest and market the crop in Zanzibar town themselves. A large number of these participate in an organised marketing rota designed to counter oversupply and enable farmers to obtain higher end of season prices. Many farmers, however, including some participants in the system, feel that it mainly serves the interests of large scale-producers, including those farmers who founded it.

Some of them still look to the

government to intervene and control prices or find alternative markets. It is not clear whether their concerns follow from the failure of the rotating marketing system to solve the problem of oversupply, or from their own failure to understand how it might work. It would be useful to examine this issue further, in order to understand better the pros and cons of marketing associations and how, if necessary, they might be assisted. 88. Otherwise, and partly as a result of their fears for the future of the orange market, farmers are already in the process of diversifying into other cash crops. The development of one successful cash crop, and a sustainable farming system around it, has evidently accelerated the pace of agricultural development. It is quite possible that this pattern of increasing diversification will, in time, become general throughout the islands.

References: Davies, F.S and Albrigo, L.G. International.

1994.

Citrus.

Wallingford: Cab

Findlay, A. J. 1933. ‘Experimental Work, Zanzibar’, in Annual Report of the Department of Agriculture for the Year 1933. Zanzibar. 12-13. (abstracted in Smith 1992b).

Freeman-Grenville, G. S. P. 1975. The East African Coast: Select Documents from First to the Earlier Nineteenth Century (second edition). London: Rex Collings.

John, R. et al.

1935.

‘Other Perennial Crops’, in Annual Report,

Kizimbani Reseach Station, 1934.

Zanzibar. 33-39.

(abstracted in

Smith 1992b).

Kirkham, V. H. et al.

1930.

Annual Report of the Department of

Agriculture for the Year 1929. Zanzibar. (abstracted in Smith 1992b).

Main, R. and Walsh, M. T. 1995. The Development of Mangoes as a Cash Crop in Muyuni, Unguja, Working Paper No. WP 95/draft, Zanzibar Cash Crops Farming Systems Project, Ministry of Agriculture, Livestock and Natural Resources, Zanzibar.

Muir, J. C. et al. 1937. Annual Report of the Department of Agriculture for the Year 1936. Zanzibar. (abstracted in Smith 1992b).

Purseglove, J. W.

1968.

Tropical Crops: Dicotyledons.

Harlow:

Longman Scientific and Technical.

Samson, J. A.

1986.

Tropical Fruits (Second Edition).

Harlow:

Longman Scientific and Technical.

Smith, P. D. (ed) 1992a. Agricultural Research and Development in Zanzibar: An Analysis of the Literature. Bangor: Centre for Arid Zone Studies, University of North Wales.

Smith, P. D. (ed) 1992b. A Bibiliography of Agricultural Research and Development in Zanzibar (Volumes I and II). Bangor: Centre for Arid Zone Studies, University of North Wales.

Walsh, M. T. (ed) 1995. Mixed Farming and Improving Fertility on the Uwanda of Central Unguja: Principal Findings of a Study of Ndijani Farmers’ Research Group, Working Paper No. WP 95/draft, Zanzibar Cash Crops Farming Systems Project, Ministry of Agriculture, Livestock and Natural Resources, Zanzibar.

Walsh, M.T. (ed) 1995. The Expansion of Cash Crop Production and Development of Marketing Systems in Northern Unguja: Principal Findings of a Study of Gamba Farmers’ Research Group, Working Paper No. WP 95/draft, Zanzibar Cash Crops Farming Systems Project, Ministry of Agriculture, Livestock and Natural Resources, Zanzibar.

Williams, R. O.

1947.

‘Plantation Crops’, in Annual Report of the

Department of Agriculture for the Year 1946.

Zanzibar.

21.

(abstracted in Smith 1992b).

Williams, R. O. 1949. The Useful and Ornamental Plants in Zanzibar and Pemba Zanzibar: Zanzibar Protectorate.