• 1. OVERVIEW
• 2. THE MAIZE MARKET, STRATEGIC GRAIN RESERVE AND PRICE
• 3 FOOD PRODUCTION
  • 3.1 Background to agricultural production
  • 3.2 1998/99 Vuli maize production
• 4. EARLY PROSPECTS FOR 1999 MASIKA AND MUSUMI CROPS
  • 4.1 Masika crop in bimodal areas
  • 4.2 Musumi rains in unimodal areas
  • 4.3 Common issues
• 5. REGIONAL AGRICULTURAL SUMMARIES
  • 5.1 Mwanza
  • 5.2 Mara
  • 5.3 Arusha
  • 5.4 Kilimanjaro
  • 5.5 Morogoro
  • 5.6 Tanga
  • 5.7 Coast
  • 5.8 Mbeya
  • 5.9 Kigoma and Kagera
• 6. MAIZE SUPPLY/DEMAND BALANCE
  • 6.1 Emergency food aid assessment
• 7. MEDIUM/LONG TERM CONSIDERATIONS

1. OVERVIEW

Abnormal weather patterns, including serious floods in late 1997 and delayed rainfall and drought during the current Vuli season, have had a serious impact on domestic food production. Amidst reports of serious food shortages developing in the country and escalating prices of maize, the country’s main staple, the Government requested an FAO/WFP Crop and Food Supply Assessment Mission in early January 1999. The mission was to appraise the current situation, specifically with regards to maize and to update the findings of an earlier FAO/WFP assessment in August 1998. In accordance with this request a rapid appraisal mission was fielded to the country between 17-29 January. The specific objectives were to appraise prospects for current Vuli maize production, review the overall food supply situation, prepare a maize balance sheet for the remaining four months of the current marketing year and provide early indications of prospects for the next main (Musumi and Masika) crops. The assessment was based on field visits to main bimodal and uni-modal rainfall areas and on discussions with key Government Ministries and Departments and UN, bilateral, private sector agencies and NGOs involved in the food sector

The Mission found that with the exception of Kigoma and Kagera, rainfall in all other Vuli areas was significantly delayed and well below normal, which seriously affected land preparation and planting. The worst affected regions were the Coast and the lowlands of Arusha, Morogoro and Kilimanjaro where rainfall was less than 25 percent of normal. As a result of reduced rainfall, planted area and yields fell sharply and overall Vuli maize production will be significantly below normal. Rainfall over the next two months will also be critical for crop production and food supply prospects, as it will heavily influence the outcome of the main Musumi and Masika crops to be harvested from May. In the event that these crops fail, the food situation is likely to deteriorate significantly.

In addition, domestic maize supplies for the current marketing year were also reduced by higher than anticipated storage losses. These losses were due to a combination of ineffective pesticide use, generally poor storage facilities and high levels of grain borer infestation. On the basis of post harvest inspections of storage facilities, the level of loss is estimated to be much higher than was anticipated by the last FAO/WFP mission in August, which did not have the benefit of such inspections. Supplies may also have been reduced by increased unoffical cross border trade in maize to Zambia, in response to higher prices as a result of significantly reduced production last year. To some extent, however, this flow was countered by imports in the north from Kenya, where prices have remained low due to a good maize harvest and substantial maize imports last year. In addition to supply constraints, there was increased demand for maize this year in regions like Dodoma and Singida, where the principal sorghum and millet crops failed almost totally last year. In these regions the food situation is more precarious than other parts of the country as the harvest has failed for three successive years and the population has less access to alternative foods, such as bananas and tubers.

The sudden rise in maize prices in late October/early November over a relatively short period, suggests that in addition to developing shortages, which would have had a more gradual impact on price, there was a considerable degree of hoarding. The hoarding was principally due to (i) the failure of the rains during these months, which are essential for both short and long season preparations and (ii) considerable coverage in the national press warning of impending food problems. This view is supported by discussions the mission had with small and medium traders in surplus areas in the south, who were observed to have reasonable stocks available and had begun reducing prices in mid/late January as rainfall prospects improved.

Although the availability and price of maize have given cause for concern, there is satisfactory supply of foods, other than sorghum and millet, in the main producing areas at reasonable prices. This is due to the favourable production of rice, banana and tuber crops last main season as a result of heavy rainfall. Rice however, still remains beyond the reach of poorer sectors of the population who have limited purchasing power, whilst crops like cassava and banana are only available in main producing areas as they are not easily transported to food deficit areas over large distances.

Based on area and yield estimates and the pattern of rainfall this year, 1999 Vuli maize production is estimated at 228 000 tonnes, some 60 percent lower than last year and 40 percent below the long term average. Taking into account estimated stocks of 120 000 tonnes, available at the beginning of February, the domestic availability of maize for the remaining four months of the current marketing year is projected at 348 000 tonnes. Most of the stock is held by large traders and the Strategic Grain Reserve (SGR). Against this, utilisation needs, including allowances for food, seed, feed, waste and closing strategic stocks, are estimated at 909 000 tonnes. To meet these needs, therefore, the country has an estimated import requirement of approximately 561 000 tonnes, of which contracted commercial imports are estimated at 75 000 tonnes and food aid in the pipeline a further 3 000 tonnes. This leaves the country with an uncovered import requirement of 483 000 tonnes, of which 20 000 will be requested as further food assistance through WFP. In the meantime the Government has released 10 000 tonnes through the SGR in relief assistance. Although the aggregate deficit in maize remains sizeable, it is recognised that considerable substitution by other food crops will reduce overall requirements for maize. The commercial sector may also increase the volume of imports, in view of the lifting of the import duty on maize and providing domestic prices in relation to international prices remain high.

The shortfall in current Vuli production and successive poor harvests in parts of the country has significantly increased the numbers of people that are vulnerable to food shortages. Moreover, as excessive rainfall was beneficial for pastures last year, the animal population increased markedly, which has contributed to considerable worsening in the terms of trade between livestock and cereals this year. Pastoralists, therefore, constitute an important category of the population who have been made increasingly vulnerable to food shortages. At the time of the last FAO/WFP assessment, it was estimated that 300 000 people were in need of some form of emergency assistance. However, as this assumed a normal Vuli harvest, the fact that crop production has fallen sharply this season means that a larger number of people will need additional food assistance. The estimate of people needing food assistance, therefore, has been revised to approximately one million, for whom an additional 20 000 tonnes will be requested.

In view of the serious food problems that have emerged due to natural disasters in recent years, it is imperative that the Government implement appropriate strategies to cope with the effects of such disasters in future. In the short term such a strategy will require that adequate strategic grain reserves are maintained, monitored and utilised in meeting emergency food needs and in capping price rises if necessary. Longer term food security requires adequate investment in agriculture. Irrespective of the need for funding, however, the mission notes with some concern that there has effectively been dis-investment in essential support services like research and extension. This in turn has led to low morale and affected productivity in agriculture. The Special Programme for Food Security jointly formulated by the Government and FAO provides an extremely useful framework for improving people’s access to food in the long term, through appropriate and sustainable production enhancing technologies. The establishment of a special working group on food security with the aim of co-ordinating various bilateral and multi-lateral efforts in the food and agriculture sector is also seen as being an extremely important initiative.

2. THE MAIZE MARKET, STRATEGIC GRAIN RESERVE AND PRICE

The Tanzanian economy is heavily dependent on the performance of agriculture. Approximately 57 percent of GDP is from agriculture, with other sectors such as manufacturing contributing less than 10 percent and transport and commerce 22 percent combined. The main cash and export crops in order of importance are coffee, cotton, tea, sisal, tobacco, cashew nuts and cloves. In spite of its importance, government expenditure on agriculture in the last two decades has been declining, and there is growing concern that key agriculture support services such as research and extension are becoming increasingly ineffective. Such breakdown obviously has long term consequences for the country’s food security and medium/long term interventions are essential to redress the problem.

In general, maize marketing in Tanzania is characterised by a large number of small traders, operating from main producing areas and urban centres. Recent studies [ Wangwe and Associates (1997) and Bryceson (1993)] , indicate that domestic trade is mostly undertaken directly between traders and producers with a limited role played by middlemen. In the main surplus areas retailers take on the functions of intermediaries (inter-regional traders and wholesalers) as in normal years there is easy access to plentiful supplies. In areas where there is usually a deficit in maize, however, there is clearer distinction between producers, traders and retailers. The presence of bigger specialised wholesalers is largely confined to Dar-es-Salaam.

The main surplus regions for maize are Iringa, Mbeya, Rukwa and Ruvuma, which collectively supply most of the country. Rukwa mainly supplies the North-Western Region, which also receives maize from Moshi and Arusha. Most of the maize from Iringa and Mbeya is destined for Morogoro and Dar-es-salaam. Dodoma receives most of its maize from Iringa, Singida and Arusha, whilst the main source of supply for Lindi and Mtwara in the south, is Ruvuma.

Nationally, it is estimated [ MDB, Market Review of Maize and Rice (1996/1997)] that approximately 40 percent of maize produced is marketed, though the estimate is much higher (60 percent) in surplus producing regions like Iringa. Estimates of the volume of maize marketed in the period 1991/1992 to 1997/1998 are indicated in Table 1.

Table 1. Estimated Volume of Maize Marketed 1991/1992 to 1997/1998.

Source: FSD and MDB

In view of the extremely large distances and the agro-diversity of Tanzania, the flow of food from surplus to deficit regions, particularly to pockets of need, is of paramount importance to national food security. Irrespective of this need, however, the flow of food is heavily constrained by grossly inadequate transport systems and insufficient haulage capacity, vehicles and rail wagons. Most commercial transporters operate out of Dar-es-salaam, along main routes, and have little incentive to transport food grain to remote and distant areas, where roads are difficult, markets restricted and purchasing power generally low. In adverse years, therefore, irrespective of the national food picture, large numbers of people remain vulnerable to food shortages. Although this reinforces the necessity of the Government to develop a longer term food security policy based on maintaining an adequate Strategic Grain Reserves (SGR), in recent years the capacity of the SGR to intervene has been declining.

The Strategic Grain Reserve was established in the 1970s with a target of 150 000 tonnes, which was considered adequate to meet needs for three months (to allow imports to take place) in the event of an emergency. In an increasingly liberalised market, the reserve was also seen as an important control instrument to cap prices as the need has been this year. The capacity of the SGR to meet emergency needs, however, has generally been eroded in the 1990s, as stocks have been progressively depleted and maintained below target. In addition to low reserves, the capacity of the SGR to intervene in markets to cap prices is further limited by administrative and policy constraints which have resulted in significant lags between price increases and effective grain releases. This year for example the price of maize began rising appreciably in November, but purchase from the SGR remained extremely low as issue prices were set too high (see Tables 2 and 3 and Figure 1).

Table 2: Strategic Grain Reserve (SGR) Stocks 1991/92 – 1998-99 (tonnes)

Table 3. SGR Allotted and Actual Sales and Balance as of 21.1.99

Source USAID FEWS

3 FOOD PRODUCTION

3.1 Background to agricultural production

Tanzania has 63 agro-ecological zones based on variations in altitude, soil type and rainfall pattern. Such diversity supports a wide range of cash and food crops from some 100 000 sq.km under cultivation. Around 16 million cattle and 15 million sheep and goats are also maintained under pastoral systems from an estimated 600 000 sq.km of grazing and browse [ Ministry of Agriculture and Cooperation (1998), Basic Data, Agriculture and Livestock Sector.] .

Rainfed cropping systems can be classified into three broad categories, a short rains (Vuli) season, from September/October to January/February, a long rains (Masika) season from February/March to June/July, and a combination of the two (Musumi) from November to June. Of these, Vuli and Masika patterns occur together in northern regions. However, only in limited districts in ten of the regions do Vuli rains support the production of cereal crops. In the central and southern regions, uni-modal Musumi pattern predominates.

On average Vuli production has contributed approximately 17 percent to national cereal production over the past six years. However, Vuli cereal production is, in most areas, erratic and unpredictable. Only in Kagera, Kigoma and Mwanza regions and in a few districts of Kilimanjaro and Mara are the Vuli rains the main cereal producing rains. Elsewhere, Vuli crops are important to supplement food and seed supplies during the gap between masika crops.

Maize usually contributes 70 percent to the overall cereal harvest during the Vuli season, whilst sorghum contributes around 22 percent, paddy 6 percent and wheat and millet the remaining 2 percent. During the season, sorghum is produced almost exclusively in Mwanza, where a similar Vuli harvest of maize is usually obtained.

Other crops vary in their contribution to Vuli food supply from region to region according to agro-ecological zone. Bananas, cassava and sweet potatoes have equal importance to maize in Kagera and Mara. Cassava makes a similar contribution to maize in Coast and Mwanza, as do bananas in Kilimanjaro. Combinations of all three crops are of equal importance to maize in Kigoma, Tanga and Morogoro. In Arusha, other than maize, only wheat is produced in significant quantities equivalent to some 28 percent of the maize crop. In Mbeya, no other Vuli crops of any significance are grown.

Grazing and browse usually benefit significantly from the Vuli rains in all regions. Agro-pastoralist and pastoralist movements and husbandry practices, therefore, are timed to capitalize on the advantages the Vuli rains provide.

In addition to the three rainfed cropping patterns, irrigated maize and rice are also produced during the Vuli season. Traditional irrigation from runoff capture and rainfall harvesting supports Vuli crops in upland areas and on escarpments out of the rain shadow. There are also numerous small irrigation schemes fed from permanent rivers in the lowlands.

3.2 1998/99 Vuli maize production

Maize is the main easily transferable carbohydrate food source produced during the Vuli season, but the area planted and yield per unit area harvested are highly dependent on rainfall and are susceptible to large fluctuations from year to year.

3.2.1 Rainfall

With the exception of Kagera and Kigoma regions, Vuli rains were universally delayed for up to 2 months.

In lowland areas of Arusha, Kilimanjaro, Morogoro and Coast, the rains either failed completely or were less than 25 percent of normal, compared with 200-300 percent of normal in 1997/98.

In the upland areas and areas out of the rain-shadow in Kilimanjaro, delays and reductions in rainfall were less pronounced, with Vuli rains starting in November. But even in these comparatively rain-secure areas, subsequent rainfall patterns were irregular in quantity and distribution.

In Mwanza, Mara, Tanga and Mbeya similar delays and erratic patterns were experienced with rainfall estimated at around 50 percent of normal by January. In all areas a continuation of Vuli rains into February in sufficient quantities, either to allow grain fill for the Vuli crop or to sustain late-planted maize until the beginning of the masika rain, is essential.

3.2.2 Planted area

Cultivation began in September/October in Vuli areas in anticipation of normal rains. No constraints relating to the ability to cultivate were noted by the Mission. In September/October draught animals were in very good condition because of excellent pasture resources last year. In the hand-cultivated upland areas, labour was readily available and in the mechanized sector fuel supplies were satisfactory and access to land was not impeded by water logging of fields or by flooding of roads. Given that in most Vuli areas 95 percent or more of farmers rely on carryover seeds from the previous harvest, maize seeds were readily available and dry planting in October was reported in various parts of Tanga, Morogoro, Kilimanjaro and Arusha regions.

The subsequent delays in rain resulted in seed loss and where germination had occurred, sharply reduced rainfall until November/December made replanting necessary. Further, access to early maturing varieties of maize suitable for late planting particularly in Mwanza, Mara, Mbeya and Morogoro was restricted due to high prices and the effective lack of credit. Problems of access to inputs also applied equally to other Vuli regions.

In view of low rainfall, the area from which Vuli production in Arusha, Kilimanjaro, Morogoro and Coast can be expected is significantly below last year. However, in Kagera, Kigoma, Mwanza, Mara, Mbeya and Tanga, the area of maize planted is estimated to be similar to or only slightly lower than usual estimates. However, in the latter four regions not all areas are expected to be harvested due to crop loss early in the season.

Area estimates used by the Mission were obtained from Regional Crop Officers and cross-checked with District Crop and Extension Officers and district level subject matter specialists. Such estimates include the area of irrigated maize, which was also noted to be less than usual due to a combination of lower water resources and damage to canals caused by last year’s unprecedented flooding.

Overall maize area planted for Vuli season production is estimated at 311 017 hectares, which is 20 percent less than the average area harvested over the last six years and 28 percent less than last year.

3.2.3 Yields

Factors influencing maize yield are (i) rainfall particularly at germination, flowering, seed set and grain fill; (ii) seeding rate; (iii) seed variety; (iv) fertilizer use; (v) pest and diseases and (vi) weeds.

Except in Kagera and Kigoma, low rainfall reduced germination, caused permanent wilting damage, premature flowering and seriously affected seed-set and grain-fill in the surviving crops in medium altitude areas. Maize crops in the lowland areas, except those that were irrigated, therefore, have failed. Only upland crops and crops grown on slopes out of the rain-shadow have not suffered severely from water stress.

Sowing rates of the local landraces, which are used throughout, give average yields from 1.2 to 1.8 tonnes in good years. Intercropping with beans is practised in 30 percent of the maize areas in normal years. This year beans have suffered from moisture stress in all but the upland areas.

Fertilizer use, except for farmyard manure, is negligible in Vuli maize production. Fertilizers are, however, available from private traders at prices around 10 000 Tsh per 50 kg bag and include urea, sulphate of ammonia, triple super phosphate and compound (NPK). Any fertilizer that is used, is principally applied to coffee in the uplands and irrigated horticultural crops in lowland areas. Other chemical inputs including pesticides, fungicides and herbicides, though available from private stockists, are generally not used in maize production.

The absence of credit exacerbates the problems farmers have in purchasing and using agricultural inputs, especially in marginal rainfall areas.

Due to reduced rains this year, no significant pests or diseases outbreaks in maize were noted, although infestations of stalk borer, at levels within tolerance limits, were reported in Mwanza, Mara and Morogoro. Rats and wild animals caused local concern throughout the crop cycle in many areas. No migratory pests were noted.

Weed competition, though lighter than last year, still necessitated weeding twice after planting in upland areas. Labour was noted to be available in all areas at wages similar to last year.

Overall, maize yields estimated by district crop/extension officers and confirmed by Mission field observations were lower than normal at some 0.7 tonnes per hectare compared with 1.3 tonnes per hectare last year and an average of 1.2 tonnes per hectare in the last six years.

3.2.3 Production

Maize production by region, based on area expected to be harvested and anticipated yield are indicated in Table 1, which also shows average yield and production data for the past six years (1992 to 1997) and last year’s Ministry estimates. As no final post harvest assessments were made, the validity of such forecasts remains unknown. Area estimates are based on aggregated returns from village extension workers compiled at district level and forwarded to the regional offices for presentation. No verification is apparently conducted as cadastral records of farm areas do not exist.

Similarly, yield forecasts remain best guesses as crop cut surveys are restricted to pilot studies and most district offices lack basic equipment and facilities to undertake such exercises. With regard to yield, the regional estimates were adjusted by the mission following field observations, measurements and discussion with a wide range of key informants including NGOs, traders and farmers.

Consequently, and with due acknowledgement of the fragility of the data base, the Mission forecasts the 1998/99 national Vuli maize harvest to be 228 222 tonnes from 311 017 hectares. This is around 60 percent lower than last year and some 50 percent less that the average contribution to the annual domestic food supply from 1992 to 1997.

3.2.4 Other crops

Cereal crops other than maize were conspicuous by their absence. Only in Kagera and Kigoma were sorghum crops expected to be normal. Sorghum harvests in Mwanza and Mara were expected to be significantly reduced. Elsewhere no growing rainfed cereal crop other than maize was noted.

In irrigated areas, paddy was noted to be late and unlikely to contribute to food supplies in the next 3-4 months.

Normal sweet potato and bean crops are forecast for Kagera and Kigoma only. In the other Vuli regions, they have only been planted and survived in upland areas, so reductions in harvests similar to maize are anticipated.

By contrast, perennial crops such as bananas and cassava have clearly benefited from last year’s heavy rains with current high levels of production reflected in low local market prices throughout the areas where the crops are grown.

Similarly, sisal and coffee growth benefited although the 1998/99 coffee crop was reported to be lower than usual in areas where heavy masika rains damaged inflorescence.

Table 1: Vuli maize production by region

4. EARLY PROSPECTS FOR 1999 MASIKA AND MUSUMI CROPS

4.1 Masika crop in bimodal areas

Failure of lowland rains and delays in start up in the upland zones in bimodal areas, have reduced maize Vuli planting by some 27 percent below last year and some 20 percent below the average area planted from 1992 to 1997.

Rainfall in January has, however, encouraged farmers to plant in anticipation of an overlap between Vuli and Masika rains. Planting was noted to be underway across the lowland areas in Arusha and Kilimanjaro. In the former region farmers were seed priming (soaking maize seeds for two days) prior to planting to hasten and standarize germination date.

In the other regions, in the bimodal zone, Masika targets provided by the regional and district crop officers were optimistic. Success of the early masika planting observed will depend on the continuation of useful rainfall throughout February. If a break in February occurs, not only will growth be affected adversely but farmers may also have difficulty in securing adequate seeds for replanting when normal Masika rains begin in March. As the use of fertiliser and improved seed is low, yields similar to the recent 6-year average are anticipated, providing that the season is relatively pest and disease free.

4.2 Musumi rains in unimodal areas

Mission visits to central and southern regions, where the musumi crops are usually planted from November onwards, noted delayed onset of rains in most districts with concomitantly later planting dates. Precipitation in all regions was much lower than average in October, November and December.

The situation in Tabora was reasonably good in three districts out of five and requests were made by farmers for fertilizers to be made available on credit.

In Singida and Dodoma, there is deepening concern over food and seed shortages. If the rains which began late in mid-January are interrupted by dry spells, replanting will be difficult. Sorghum and millet seed were noted to be in short supply due to poor crops last year.

In Shinyanga, up to six weeks delays in rainfall had reduced areas planted and short cycle sorghum seeds were requested by farmers.

Planting has also been delayed in Lindi due to late rains. Complete failure of the Vuli harvest in the Coast region means that the areas will need to be replanted and seed stocks are expected to be low.

Elsewhere, the forecasts favour a normal Musumi crop, provided that the rains continue.

4.3 Common issues

Delays in planting will mean delays in harvesting, extending the gap between successive harvests, forcing small farmers to rely on remaining coping mechanisms.

Household stocks were noted to be low in all regions due in part to increased selling, as cereals, especially maize, were used as cash crops to meet the farmers’ immediate needs after the last harvest. Also, storage losses last year were assessed as being particularly high. Maize grain borers were universally cited as a pest of great significance, attacking the standing crop and the produce in store. Pesticides commonly used throughout the country were noted as having little effect and losses of up to 70 percent were reported. The threat of such losses was also the main reason why post-harvest selling of maize was so prevalent.

The low use of inputs on cereal crops places an artificial ceiling on production in masika and musumi areas with adequate rainfall. Presently, access to credit for the small farmer is non-existent, and cash payments for fertilizer, at 10 000 Tsh per 50 kg bag and 5 000 Tsh for 2 kg of improved seed, are beyond the means of most households. Consequently, farmers rely heavily on NGOs and other development agencies to supply seeds and other inputs at little or no cost. Further, given a spate of sales of "fake" or diluted chemicals, farmer confidence in private sector input supply appears to have been shaken and they look towards agencies/government to provide guarantees of quality. These two issues, i.e. credit and input supply, need to be addressed if future production targets are to be met.

5. REGIONAL AGRICULTURAL SUMMARIES

The situation in different bi-modal Vuli maize producing regions is summarized below. With the exception of Kagera and Kigoma, where data were obtained from the Tanzanian Food Security Office, the Mission visited key producing districts in each region.

5.1 Mwanza

Mission teams visited five out of Mwanza’s seven districts namely Mwanza, Missunguri, Sengerema, Magu and Kwimba. In each district a similar situation was noted. Rains were up to two months later than usual and irregular on arrival. As the Vuli harvest usually provides 94 percent of the annual maize crop such delays have serious consequences for maize availability in the next few months. Further, area planted to maize is estimated to be 30 percent below normal. Areas planted to pulses and cotton, the main cash crop in the region, are estimated to be 80 percent below normal. Cassava, of equal importance to maize as a staple in the region, is performing well having benefited from last year’s heavy rains. Sorghum is, however, also reduced in area and delayed planting is expected to reduce yield and availability.

Maize seed stocks are low due to high post-harvest losses of 20-30 percent. Opportunities to reseed with short cycle cereals are limited. Requests for assistance in the provision of suitable varieties of maize and sorghum were received by the Mission.

Market prices of maize during the Mission were between 16 000 and 18 000 Tsh per 100 kg, but a downward trend was reported to have begun due to imports from Kenya.

If the rains continue through February and merge with the masika rains, maize supply will be secured, after the harvest, providing, however, that the season remains relatively pest and disease free.

5.2 Mara

Mission teams visited four of the five districts in Mara, namely Terime, Bunda, Serengeti and Musoma Rural. Vuli rains began on time in Tarime and Serengeti but failed at maize flowering time in November and December. In Bunda and Musoma Rural, a late start was followed by irregular rains and crop performance is expected to be poor. Area under maize is similar to normal, however, yield is expected to be lower due to crop failure. It should be noted that other food crops, bananas, cassava and sweet potatoes each contribute similar quantities to maize annually. This year, bananas and cassava are expected to produce well, whereas the sweet potato crop is likely to be reduced.

Due the delay in rains, most farmers have opted not to plant cotton, the main cash crop. The range of alternative crops available to farmers will be restricted by seed availability and the ability of farmers to purchase given the lack of credit.

Maize prices are currently 15 000-18 000 Tsh per 100 kg bag, which compare unfavourably with the prices of livestock at 1 000 to 2 000 Tsh for adult goats.

Post-harvest losses last year were estimated at around 30 percent, mainly due to the ineffectiveness of pesticides against the grain borer. Storage conditions last year were also poor, particularly for the Vuli crop that was harvested and stored in humid conditions.

5.3 Arusha

Out of the region’s ten districts, the Mission teams visited six which produce Vuli crops, namely Monduli, Karatu, Mbulu, Kiteto, Simanjero and Arumeru. Rainfall patterns observed were very similar in that in lowland areas Vuli rains were virtually non-existent until mid-January. Pockets of rainfall in the uplands in November/December had encouraged planting of maize in Karatu but in general the only crops growing were perennials or annuals cropped under small-scale traditional irrigation systems. Area planted was some 70 percent below the average Vuli maize area for the last six years and wheat planting had been postponed during the last quarter of 1998.

Rains in January, however, encouraged widespread planting of maize and wheat in the expectation that in non-Vuli years the late Vuli and masika rains would merge. Farmer carryover seeds were used almost exclusively for this planting. Seed priming (soaking for two days prior to planting) of maize is being practised to hasten and standardize germination dates.

High post-harvest losses last year were said to have encouraged immediate post-harvest sales of maize that left farmers with low levels of household stocks. Market prices at 18 000-20 000 Tsh/100 kg for maize and between 2-3000 and 20-30 000 for a goats and cattle mean that the livestock/cereals terms of trade are extremely unfavourable for agro-pastoralists/pastoralists

Other crops , particularly bananas, are readily available in most markets at low prices due to high production following last year’s rains. Food security in the region does, however, depend on the continuation of rains through February until the masika rains begin.

5.4 Kilimanjaro

Mission teams visited five out of the six districts, namely Moshi, Hai, Rombo, Mwanga and Same. Conditions varied according to location. In the lowland areas the Vuli rains had failed completely. Consequently, there was no planting on the lowlands except in irrigated areas until the onset of rains in mid-January.

In medium altitude areas, perennial crops such as cassava, bananas and coffee were flourishing, however November/December planted maize and beans were observed to be suffering from stress caused by irregular and light rainfall.

In the uplands, the crops were well established and reasonable yields are expected.

On the eastern and north-eastern escarpments crops were also well established, benefiting from not being in the rain-shadow. Paddy rice crops had been planted using traditional irrigation methods, grazing and browse were in very good condition. As a result of the failure of lowland rains, the total area planted to maize is some 30 percent of previous years.

Last year, sales of maize in most districts were greater than usual due to a combination of factors including humid conditions, poor storage facilities, grain borer and apparently ineffective pesticides. Farmers who had sold 100kg of maize for as little as 2 000-3 000 Tsh post-harvest would now have to buy back similar maize at seven or eight times the price.

Other crops, such as bananas and cassava, were available at low prices in all the markets. Livestock condition was good, although viral diseases were noted to be of concern. Given the proximity to the Kenya border, livestock prices appear to have been sustained.

5.5 Morogoro

The Mission visited the two districts in Morogoro which normally provide a Vuli maize crop, namely Morogoro rural and Kibsa. Observations and discussions confirmed the late arrival of Vuli rains which delayed planting and reduced area planted by 60 percent. Rainfall remained light and patchy until January, so much lower maize yields than normal are expected.

Household stocks are low and have been ravaged by pests since the harvest. Other crops are, however, in good condition, particularly sugar cane, which has benefited considerably from last year’s rains.

Similarly forage/browse is still in good supply and livestock condition is good.

Maize market prices are high, from 17 500 to 24 000 Tsh per 100kg bag, representing a great change in prices over the past four months.

5.6 Tanga

Mission teams visited four districts in Tanga covering a cross section of upland and lowland areas, namely Pangani, Lushoto, Tanga Urban and Handeni. Despite late onset of the Vuli rains, area planted to maize was relatively close to recent average planting, mostly using farmer carryover seeds. The Vuli rains have, however, been light and irregular so significantly reduced yields are expected.

Other crops particularly cassava are important contributors to food supply. Production of cassava was noted as normal. Similarly important cash crops, viz. coconuts, cashew and sisal, were noted to be in very good condition having benefited greatly from last year’s heavy rains. Maize prices in the market during the Mission reflect reduced availability, exacerbated by post-harvest/storage losses reported at 15-20 percent.

5.7 Coast

Mission teams visiting the Coast region reported that no maize crops had been sown due to delayed rainfall. The region is not normally recognised as a Vuli maize planting area, the average area planted from 1992 to 1997 being only 3 500 hectares. Last year was exceptional with an estimated tenfold increase in area and reasonable yields noted. Other crops, particularly perennial cash crops, were noted as normal. The main problem of food supply in parts of the coast region has been the failure of successive harvests in recent years.

5.8 Mbeya

Mission visits to south west regions noted much reduced Vuli rains, which delayed planting of maize. Yields are expected to be lower than normal for the Vuli crop in the upland areas. However, most maize is produced from the main crop later in the year. For this, early maturing seed varieties were requested as soon as possible in anticipation of poor main rains. In general, planting has been delayed throughout the lowlands until rains to avoid loss of seed.

5.9 Kigoma and Kagera

Reports of near normal rains were confirmed by meteorological statistics and information from the Tanzanian Food Security Unit. In general, crops are likely to yield normally and have been free from migratory pests and diseases. No labour constraints that would affect weeding or harvesting were reported.

6. MAIZE SUPPLY/DEMAND BALANCE

In estimating the maize supply/demand balance sheet for the remaining four months (February-May) of the current marketing year, the Mission used the following estimates:

• A current population of 31.66 million people, based on a mid November 1998 population of 31.51 million estimated by the last FAO/WFP mission and a growth rate of 2.8 percent per annum.
  
  
• In accordance with norms used for food consumption in Tanzania, a per caput consumption of 74.5 kg maize per annum as part of an overall diet of various foods. This translates into 24.8 kg/caput for the four months covering February to May 1999.
  
  
• Other utilization include seed for replanting up to 200 000 hectares, feed (25 percent of 290 000 tonnes of concentrates used annually or 24 000 for the next four months) and post harvest losses equivalent to 15 000 tonnes.
  
  
• Maize stocks at the beginning of February are estimated at 120 000 tonnes, of which major traders are estimated to have 60 000 tonnes, the Strategic Grain Reserve (SGR) 35 000 tonnes, and smaller traders and farm households the remaining 25 000 tonnes.

Closing stocks for maize assume that the SGR needs to have significant build up of stocks to cover future strategic needs, especially in the next marketing year if the next main season crops are significantly reduced due to the failure of rains.

The national maize balance sheet for February to May 1999 is shown in Table 4.

Table 4: Tanzania – Maize Balance Sheet, February to May 1999 (000 tonnes)

1/ Based on 120 days.
2/ Includes 50 000 tonnes through traders and 25 000 through Government.

6.1 Emergency food aid assessment

As a result of serious drought in early 1997, WFP approved an emergency operation for assistance to approximately 1.5 million people, in 14 of the country’s 20 regions. Due to subsequent serious flooding, which on the one hand limited the distribution of emergency food aid, and on the other caused further destruction of livelihoods, this WFP emergency operation was extended in time after the 1997/98 harvest. The extension was limited to Dodoma and Singida regions, which were particularly affected by flooding. In these two regions food aid was being provided to about 600 000 people in January 1999, targeted through on-going vulnerability assessments and analysis of information by the WFP VAM office, working closely with FAO and non-government organisations involved in food security and emergency interventions, as well as with FEWS and local and central Government. Food resources available under the existing operation were only sufficient to allow the continuation of the existing level of assistance to these two regions until March 1999.

The number of people exposed to short-term food insecurity, and the number of areas affected, is now increasing significantly as a result of the cumulative impact of abnormal rains since 1997. The scarcity and high prices of cereals and the low price of livestock are further exacerbating problems faced by rural households in areas where other staples are in short supply and pastoralists who rely on livestock sales to meet food needs.

The Government carried out an assessment, in November 1998, which identified other regions containing pockets where deteriorating food supply and the high price of maize had eroded the food security of the most vulnerable households.

The Government responded to the situation by authorising, in November 1998, the release of 35 000 tonnes of maize for sale from the SGR, although release was slow due to the high price set initially. In January 1999, the Government approved the free distribution of 10 000 tonnes of maize from its own stocks held in the SGR. The Government took steps to import 25 000 tonnes of maize to help replenish the SGR, and at the same time requested another joint FAO/WFP assessment mission to update the earlier mission’s assessment of the situation.

The present Mission confirmed that there has been a significant increase in the number of people at risk of hunger and malnutrition. While food security continues to be particularly precarious in the Central Zone, many of the regions affected by the 1997/98 emergency were again experiencing food insecurity among the most at-risk groups of the population.

There is a serious risk of widespread severe deterioration in food availability should the Masika and/or Musumi rains fail during the current growing season. There was a 45 percent risk of failure of the main rains according to meteorological forecasts in January; there had been no rains in most areas in the first dekad of February.

While famine conditions, evidenced by weight loss, destitution, migration, epidemic mortality and morbidity etc. do not yet exist, in several regions all of the "trailing indicators" are in evidence: successive crop failures, very reduced food supplies, inflated staple prices and steeply declining livestock/cereals terms of trade.

For Singida, which has suffered greatest food insecurity, the mission also recommends an increase in the basic ration of maize, to ensure an adequate level of calorie availability. The households already identified by recent assessments to require emergency food aid through until March 1999, will require a further period of assistance to meet minimum nutritional requirements until the delayed next harvest - i.e. until May 1999.

While several of the areas suffering from the Vuli rains failure still had other staples such as banana or cassava, 11 other regions were identified, through assessments prior to and during the current mission, to have pockets where the failure of the Vuli rains has led to severe difficulties for the most at-risk households. In the affected districts, which have in many cases experienced three successive bad agricultural cycles, a maximum of five percent of the population has been targeted, which are those households whose traditional coping strategies have been eroded due the cumulative effects of successive disasters.

Areas currently receiving assistance

The areas already being assisted – Dodoma and Singida Regions – are in unimodal production areas. The two regions had been singled out in the mid-1998 FAO/WFP mission report as being severely food insecure due to drought and floods in 1997/98 which caused poor production, especially of sorghum and millet, in 1998. While 712,000 people in the Central Zone were targeted for emergency assistance in 1997/98, only less than half that number were reached due to devastation of infrastructure caused by severe floods.

In Singida a high proportion of the population was experiencing the cumulative effects of several consecutive poor seasons. Food stocks were low and prices exceptionally high for the little grain available, and livestock prices were exceptionally low. The latter problem was compounded by a regional quarantine due to Contagious Bovine Plural Pneumonia (CBPP). In addition, sweet potatoes, which are the main staple after maize, sorghum and millet, were becoming increasingly scarce. From November 1998, 88 villages had been receiving WFP assistance, and in January a further 52 villages were selected for assistance starting in February. The districts affected are Singida Urban, Singida Rural, Manyoni and Iramba. The assessments of needs were based on household food security surveys carried out by Save the Children (UK). Subsequent food basket monitoring and food usage surveys in 30 villages confirmed that almost all (more than 90 percent) of the food aid was consumed by the targeted households. Assuming a reasonable harvest in May/June, a continuation of food assistance until May was recommended by the mission.

A market intervention of 4 000 tonnes was also recommended for Singida by the mission.

Dodoma had a poor season in 1998. A survey by the Government and WFP before the mid-1998 joint FAO/WFP mission had estimated that the cereals harvest would cover only six months or less of consumption requirements. In fact a very high proportion of the maize production was sold early in the season and left the region. At the time of the mission the situation was beginning to stabilise because of Government market intervention, and the increase in WFP assistance, which had been extended recently from 107 villages to 327 villages. While the general situation was similar to Singida, the terms of trade between livestock and maize were slightly less unfavourable in Dodoma. A continuation of assistance until April was recommended for Dodoma, subject to continuing assessment of the situation.

Areas not assisted during the post 1997/98 harvest period

The failure of the short rains in many bimodal areas, and cumulative problems in several other unimodal areas has resulted in additional pockets of vulnerability. While certain areas suffering the consequences of abnormal rains still have other staples such as banana or cassava, there are areas where the latest failure of the rains has caused the most vulnerable households to exhaust all alternative livelihood and survival mechanisms to take them through to the next harvest. The districts and villages affected are to a large extent the same that had already been targeted under the 1997/98 WFP emergency operation, although the total number of individual households targeted at that time was higher than the number assessed to be in need of assistance in 1999.

In Shinyanga region the coping mechanisms of low-income families without livestock have been steadily eroded in Bariadi, Maswa and Meatu districts. In these eastern districts the delay in the rains had worsened the condition of the cattle, thereby affecting the terms of trade between cattle and grain. While some people were coping by engaging in casual labour, tending the fields and livestock of wealthier farmers, these options were limited and not available to all.

In Tabora region, which is a unimodal production area, the districts of Nzega and Igunda, bordering the affected areas in Shinyanga, were identified to have been worst affected by the decline in crop production in the 1997/98 season and subsequent floods and pest infestations. Food prices and the condition of livestock is these districts indicate that there is a serious food security problem.

In Mwanza and Mara the bulk of the food which is now being utilised was produced during the previous 1997/98 season. This mainly constitutes cassava and to a lesser extent maize and sorghum from Serengeti and Tarime districts. Under normal circumstances, harvests from the short rains would have been available to cover the food gap. However there has been a complete absence of Vuli rains in many areas, which in normal years should contribute up to 60 percent of the overall food balance in the region.

In Mwanza the last season was not very good for maize and cotton but very good for rice. However the rice crop was sold, mainly to neighbouring countries, early in the season and at very low prices. The region relies on the Vuli rains for over 90 percent of the annual maize and sorghum production. However this year a little rain in September did not continue and most farmers had insufficient seeds to replant. Prices in markets for maize had declined briefly due maize imports from Kenya, but these were now restricted due to the export ban from Kenya. Rice prices had risen, and household stocks were low or nil in several districts. People were consuming dried cassava and sweet potato chips, and some families were skipping meals. Livestock prices had fallen drastically from 50-80,000 shillings per cow in Aug-Oct to 15-25,000 shillings at the time of the mission. The most affected districts are Magu, Kwimbe, Missungwi. Although there was still an abundant supply of cassava in Sengerema and Geita districts, some divisions have been affected.

In Mara all districts have been affected at varying levels, with Musoma and Bunda being hardest hit.

In Arusha the bimodal areas were suffering due to very late rains. Vuli harvested crops were expected to be virtually nil. Replanting was taking place and if rains should continue prospects were judged to be reasonably good for the next harvest due in May. However the poor terms of trade between livestock and grains were creating severe problems especially for small-scale pastoralists. Pockets where the population was facing special difficulties in coping with the food situation were identified in the districts of Simanjaro, Kiteto, Ngorongoro, Mbulu, Babati, Arumeru, Monduli and Karatu.

In the Kilimanjaro region the western lowlands were very dry, with no planting. Limited pockets of need were identified in Mwanga, Rombo, Same and Moshi districts.

In Morogoro and Tanga regions the Vuli rains have failed and the prices of cereals are high. The Vuli crops account for about half the annual main food crops. Land preparation was in progress, giving opportunities for employment as a coping mechanism and the rains were sufficient to allow vegetable production. Pocket of critical food shorage identified in the districts of Kilosa, Morgoro Rural and Ulanga in Morogoro; and in the districts of Handeni, Korogwe and Muheza in Tanga.

In Lindi region three districts have been identified as most affected: Ruangwe, Liwale and Kilwa. A drought in 96/97, and flooding in 1998 affected these areas. In Liwali District where two divisions are affected by food shortages, household level food stocks are almost non-existent, and families were resorting to eating wild roots.

The Coast Region had similarly received rains only very late in January. Difficulties were cumulative and mounting. While in many districts established coping mechanisms have been successful, pockets of serious food insecurity were observed in Bagamoyo, Kbaha and Rufiji districts.

While the Southern Highlands were judged to have a generally good harvest from the 1998 season and planting was well under way at the time of the mission, pockets were reported where food problems exist. In Iringa Region the Pawaga division of Iringa Rural district was facing extreme food shortage and the poorest members of the villages were resorting to consumption of wild berries, which are only consumed in times of extreme stress; however, this chronically food-insecure area is already included in a WFP food-for-work project. In the event that the food-for-work project were to be delayed or considered to provide insufficient coverage, there would be a need for additional emergency assistance.

Food aid needs assessment and targeting

The number of people assessed to be in need of assistance is based on a rapid assessment carried out by the mission, and on information available from VAM, USAID/FEWS, SCF and on previous experience of WFP and its implementing partners in responding to earlier food crises.

The mission examined nutritional information available. Nutritional data is collected routinely in MCH Clinics in Health Canters and dispensaries throughout Tanzania. The common parameters measured are weight for age. Weight for age is not a useful index for screening or for assessment in nutritional emergencies. Because it ignores height it fails to distinguish between short children with an adequate body weight and tall thin children. In Tanzania, which has a high prevalence of stunting, such a measurement cannot give an indication of mortality risk of individual children or nutritional risk of the population that weight for height does. Weight for Height is also more sensitive, albeit very late indicator of food insecurity.

The nutritional data currently collected in Tanzania cannot be used to assess the severity of the current food supply situation nor can it be used for targeting, whereas the Household food Security Assessment tools developed by SCF/UK and other NGOs, in association with the WFP VAM unit, have proved to be very effective for both targeting and monitoring in the current situation and it is recommended that they continue to be used to monitor the situation.

Food distributions should only be undertaken through reliable organisations working alongside local government, so that market disruption and weakening of household coping strategies may be avoided. There is considerable recent experience in assessments and vulnerability analysis, coordinated by WFP’s VAM unit. Monitoring of the current situation will however call for additional human resources to ensure adequate control and supervision in the current situation.

The overall requirements in food assistance for the period February – May 1999 (120 days) is estimated to amount to about 50,000 metric tonnes of maize, including that provided by the Government of Tanzania. The existing ration of 400 grams of maize per person/day is adequate for most areas, except in the most critically affected areas where a 500 gram ration is recommended.

Taking into account in-country emergency food aid stocks under WFP emergency operation no. 5889 (Assistance to Drought Affected Persons in Tanzania), the need for additional external emergency food aid is estimated to be about 20,000 metric tonnes of maize.

The following table sets out the number of people assessed to be in need of emergency assistance:

Longer-term measures to reduce food insecurity and provide adequate safety-net provisions are essential to reduce the frequency and extent of emergency food aid. In this regard the Government has indicated that food security will be given the highest priority. WFP is also re-examining its longer-term strategies through a country programme preparation exercise, which will focus mainly on ways to contribute through food aid to the Government’s efforts in this area.

7. MEDIUM/LONG TERM CONSIDERATIONS

In addition to ensuring short term food needs to the most vulnerable sectors of society it is important that steps are taken by the Government, with assistance from the international community, to address the longer term issue of food security. This will require that adequate strategic stocks are built up to meet future emergency needs, especially in areas which are difficult to access. Effective early warning of impending difficulties will also assist in countering disasters. In this regard, the following may be considered.

a) Improvements in the collection and analysis of agricultural statistics is needed to assess production and assist in monitoring and planning. District agricultural offices lack basic equipment and training to undertake simple area surveys or yield assessments and little post harvest evaluation is undertaken.

b) In view of changing weather patterns appropriate efforts in adaptive/action research need to be undertaken to modify cropping systems. For instance, the wider application of seed priming, noted by the Mission in Keratu-Arusha, needs to be investigated. Other techniques such as cereal transplanting, tied-ridging and contour strip farming should be investigated with appropriate training packages for pilot programmes. Small scale irrigation techniques, presently being sponsored by FAO’s Special Programme for Food Security need to be expanded if shown to be cost effective and replicable.

c) Current low livestock prices highlight the vulnerability of pastoralists and agro-pastoralists in times of low cereal (maize) availability. Appropriate meat marketing strategies, based on product diversification, need to be introduced into the livestock industry. On a wider scale, to make Tanzanian meat products improved internationally, acceptable health and slaughtering standards need to be implemented.

d) The high level of post-harvest losses noted in maize last year, particularly through grain borer infestation, needs effective counter measures. In pesticide use, for example, a wide range of problems persist including ineffective pesticide application rates, adulteration of chemicals, which reduces the level of active ingredient. Appropriate quality control of chemicals marketed through the private sector should be strengthened, to re-establish farmer confidence in the products sold, whilst grain borers need to be tested for pesticide resistance.