WFP/UNDP-EUE/DPPC: Joint 2002 mid-season assessment to Gambella region, Ethiopia

Attachments

• Download Report (PDF | 28.56 KB)

Assessment Mission: 25 August - 1 September 2002
By Kerren Hedlund, WFP, Abraham Sewonet, UN-Emergencies Unit for Ethiopia & Habtamu Beyene, DPPC

1. Introduction and background

The team from DPPC, WFP and UNEUE conducted a rapid mid-season 2002 assessment to Gambella region from 25 August to 1 September 2002. Visited weredas include Gambella in zone 1, Gog and Abobo in zone 2, Dimma in zone 4. Due to security problem, the team was unable to visit zone 3. For the weredas mentioned above, the Meher share ranges between 80-90% with two harvest (June, September), and 10-20% recessional agriculture (January/February).

Because it assists in defining target groups but also because it reveals the complexity of targeting due to risk of conflict, it is important to note several facts:

• Anyuak are largely crop dependent. The majority are found along the riverbanks including Itang from which they have recently been displaced. They are the majority in Gambella, Abobo, Dimma and Gog. They practice shifting cultivation. The incumbent government in Gambella is largely Anyuak. The population of the Region is 27% Anyuak (1994 Census).
  
  
• Nuer are largely livestock dependent. They are found in the majority in Akobo and Jikaw and parts of Itang. As Akobo and Jikaw are largely inundated in the wet season, they migrate to the uplands with their livestock during that period then return to the riverbank once it has receeded to cultivate. The population of the Region is 40% Nuer.
  
  
• The Mezenghir (6%) the majority live in Godere and in pocket areas of Gambella and Abobo. They live in settlements, practice some cultivation but are mostly dependent on beekeeping.
  
  
• Other minority clans include the Opo and Komo.
  
  
• Highlanders from Amhara (e.g. Wollo), Tigray, SNNPR (e.g. Wolayita) live in settlements in Abobo, Gambella and Itang and practice sedentary agriculture. Many highlanders reside in town and are business men and women (total 24% of region, however 49% of urban population).
• There are 3 refugee camps in Gambella Region in Gambella (Bonga), Gog (Pugnido) and Dimma (Dimma). They are largely Sudanese Anyuak (29% in Pugnido), Nuer (57% in Pugnido and 79% in Dimma), Dinka, Shuluk, Uduk and others. Given their ethnic affiliation, they often serve as a coping mechanism for Ethiopian tribes of the same clan. There is a steady "trickle" into the camps during the hungry season (January - April) (Table 2)

2. Mission Results

2.1 Rainfall

Normal rainfall in Gambella starts in May and ends in late September (Akobo Baro Basin Study, 1999). Total rainfall is 720 to 1350mm in areas between 500 and 1000 m.a.s.l. during this period. This is approximately 80-90% of total annual rainfall (900 to 1500mm).

In 2000, the total rainfall for the period May to August was 560mm; in 2001 405 mm and to date (as of 25 August) in 2002 320 mm. It must also be noted that the rainfall gauge in Gambella town can only be said to represent that area however it may provide indication for parts of the region.

As early as May, Regional and Woreda officials (Gambella Region Early Warning Bulletin) reported below normal rains in the west (Jikaw, Akobo and western Jor). Again in June, the same report was made.

However in the eastern part of the region (Abobo, Gog, Itang, Gambella and Godere) rainfall was reported to be normal in May and again in June with a note that amount and distribution was more intensive in Abobo.

In July, throughout the Region with the exception of Godere, rainfall was reported to be below normal.

In general as of August, it was reported that that centre south Zone Two (Abobo, Gog and Jor) with the exception of western Jor was better relative to centre north Zone One (Itang and Gambella). For western Zone Three (Jikaw and Akobo) no data was available given insecurity and lack of surveillance, however it was assumed that the west was more affected.

Many areas reported a dry spell in July and rains resuming only in August (Gog and Gambella Officials). In Dimma and Jog, the dry spell had started in June and has lasted until September. Officials noted that normally dry spells can occur for up to one month, but that this three to four month dry spell was abnormal.

River levels of the Akobo, Gillo and Baro Rivers (in order of severity) which depend on rains in the highlands were substantially below normal for this time of year. However if rains continue in the highlands the river levels should assume their normal levels in September and October.

Secondary sources of data confirm both ground reports and rainfall gauge data. Satellite imagery denotes much below average rain in June and July (more than 75% below average) in the west, that is parts of Akobo, Jor and Dimma. And much below average rain throughout the region in August (25-50% below average). NMSA data reported in northern Gambella normal to above normal rainfall in July. While southern Gambella reported normal to below normal for the same period.

At the same time, when discussing with elderly farmers, they consistently reported that this years rains were as poor as they had seen in their lifetimes (maybe with the exception of 1984/85 when there was reported literally no rain) and they were expecting substantial reductions in production (25-50%).

2.2 Crop Production

The Region is not cereal self-sufficient and in fact in a good year (2001) less than 12,000 Ha were planted and less than 70kg per person per year were produced. In 2000, a year that the Region received no food assistance, production was only 44kg per person. As recently as 1998 and 1999 total production was reported to be 3095 Mt and 3775 Mt (17kg/ppyr and 21kg/ppyr). This implies a regular and heavy dependence on the purchase of cereals from the highlands or on other sources of food such as wild foods, fishing, livestock products, etc. However it is interesting to note the steady increase between 1998 and 2001 in agricultural production.

Total area planted in 2000 for woredas with historical data (Itang, Abobo, Gog, Jikaw and Godere) was 7743 Ha with 8,912 Mt of production (maize and sorghum only). Average yields for maize were 10 Qt/ha and sorghum 6 Qt/ha. As this was a year that the ENA teams reported a 43% reduction in production (50% reduction in Jor due to displacement from conflict and 50% reduction in Gambella, Itang and Jikaw due to pest infestation), 2000 is here considered a "bad year".

In 2001 for the same woredas, there was nearly twice as much area planted and production, i.e. 11,835 Ha were planted resulting in 14,835 Mt or an average yield for maize of 12 Qt/ha and sorghum 9 Qt/ha. As this is the highest production on record, 2001 is here considered a "good year".

The yields for maize and sorghum even in a good year are below the national average for these crops, this may be due to the lack of plough oxen used to prepare fields. This is further elaborated below in recessional agriculture.

In 2002, Regional BoA officials were in the field collecting data and were unable to provide it to the teams before leaving. However, data was provided by DoA in Gog and demonstrated no change in the area planted. In 2001 (a good year) 1750 Ha were planted with maize and 430 Ha were planted with sorghum (2180 Ha). In 2002, 1676 Ha were planted with maize and 459 Ha were planted with sorghum (2135 Ha). While it is a slight shift in crop type, the total area covered compares to 2001.

In Gambella, if drought occurs, it is unlikely farmers will make a shift to short cycle crops. At best, they shift to sorghum, a drought resistant crop often replanting after maize fails. For early consumption, Gambella farmers also produce a local variety of maize "Utella" that matures in 3 months. At this time (August), farmers are consuming this indigenous variety.

BoA reported that farmers are planting increased amount of sorghum, but given the low percentage of total crop (around 12% Annex 1), a slight increase in area planted will have little impact on total production.

While the 2000 ENA team reported a reduction of 43% in maize production in 2000/2001 (nonetheless not recommending food assistance given the Region’s capacity to cope with a maize production shortfall), the rainfall was substantially higher in 2000 than in 2001. In fact with less rainfall in 2001, production was 65% higher than in 2000. This may be due to the fact that too much rain caused flooding and damage of crops still in the fields in 2000.

Satellite imagery denotes area with below average vegetative cover (NDVI) for June, July and August, substantially below average in the vicinity of Jor in July however recovering in August, and substantially below average in the vicinity of Gog and Dimma in August.

Given the team’s ground observations it can be confirmed that maize was in general in two states:

1) Either maize had been planted in May/April and had undergone moisture stress during emergence stage and if it survived, was at ripening stage however stunted and expecting below average yields. There were a few fields, e.g. Bonga, Gambella woreda and Gedu, Dimma woreda where sorghum was at ripening stage and in good condition.

2) OR maize/sorghum had been planted late and was in good condition however late in phenological stage compared to normal that is at flowering or seed setting stage. In the centre north (Zone One) maize was 70% at flowering stage and 30% at seed setting. Sorghum was roughly 90% at growth stage. In the centre south (Zone Two) maize was 40% at flowering, 40% at seed setting and 20% at ripening.

However it is also true that areas were not uniformly affected and within 50 m of a poor field was an "okay" or even good field. BoA officials, NGOs and even farmers attributed this to soil type e.g. sandy vs. loamy soils or tillage practices; machine, ox-plough or "chala" (spear and hoe).

Recessional agriculture is also very important to Gambella as perennial rivers normally overflow in June, July and August and recede in September or October. In 2001, recessional farming is 20% of total area planted. Total area in 2001 was 11,835 Ha, of this 2,360 Ha (1,895 Ha maize and 464 Ha sorghum) was recessional agriculture. At the moment the river levels are substantially below normal as rains in the highlands were also late. Normally the river overflows in June or July in Gambella, and then break it’s banks in Itang and Jikaw later. In Gambella the river only started to break its banks in August. Reports from NGOs and Churches indicated that in Gambella many dry season villages are still inhabited when normally they would have been vacated with rising waters and floods, including Jikaw where the team was unable to visit due to security reasons (Nuer Development Association).

2.3 Livestock Production

Livestock is a primary source of income for the Nuer in Akobo and Jikaw and a secondary source of income for Anyuak, particularly in Jor.

Livestock condition, pasture and water availability at present are good. While some disease outbreaks have been reported (FMD, CCPP, CBPP and pasteurollosis), the numbers are not of epidemic proportions.

In 2001, in Jikaw, 124 cattle and 300 goats died of FMD and CCPP respectively. In 2000, again in Jikaw, 138 cattle and 644 shoats died. In 2002, 130 and 395 cattle and goats have died. Of note, FMD appears to have "emerged" affecting 1,100 animals and killing 234 in Jikaw (per BoA), however livestock experts with ACORD, an NGO working in Jikaw, note that FMD is actually endemic. Before outbreaks occurred every 3 to 4 years but since 2001 they have occurred every year. FMD normally affects milk production and does not normally cause death of animals.

The BoA with the assistance of ACORD is conducting a vaccination campaign that has already concluded in Abobo, Gog, Gambella, Itang and is ongoing in Godere.

2.4 Market Conditions

Cereal Prices

Similar to other parts of the country, Gambella Region benefited from very low prices in 2001 due to relative surpluses in the highlands in 2000 and 2001. With the exception of Abobo, prices are increasing but are at 2000 levels, i.e. returning to their normal state.

Prices of green maize indicate relative food availability in markets; Gambella 4 cobs/1 ETB, Gog 3 cobs/1 ETB, near Godere (Woina-dega ; Gechu) 7 cobs/1 ETB, Dimma 5 cobs/1 ETB. It should be noted that local production in Dimma is negligible and is not even assessed by the BoA. The people don't market their own production and maize available in the market arrives from Mezan (Bench). Another source of cereals in the market in Dimma is the refugee camp where refugees regularly exchange a part of their wheat ration for maize (thus encouraging a regular flow of maize from the highlands to the lowlands). The price of wheat is slightly increased in the last three months, however it is still lower than in 2000 (104 bir/qt July 2002 vs 140 bir/qt July 2000).

Livestock Prices

Livestock prices are reported to be low due to lack of supply caused by constraints in movements of livestock to markets due to the conflict (particularly in Itang). However, the data provided are inconclusive or incomplete. Furthermore, if supply is low normally prices would rise therefore the present market situation is unclear. The supply of livestock to the market normally increases until it peaks in January (Acord).

2.5 Income Sources and Other Coping Mechanisms

Wage labour opportunities include the State Cotton Farm in Abobo ; increasing demand for wage labour following years where area planted declined due to lack of wage labour. According to wage labourers it is a natural increase in demand and not due to food shortage. Wage labour rates (rural) are 7 to 8 bir per day.

The majority in Dimma engage to some degree in gold mining. According to UNDP study (1999) less than 50% of total woreda income is derived from agriculture and the remaining from gold mining, trade, and other sources. However due to lack of water necessary for "washing" the gold, supply to markets has decreased and subsequently the price of gold has doubled (2001 45 bir/gr vs 2002 75 bir/gr). Woreda officials reported that in school attendance decreased in June due to increased numbers of children going to the mines and attributed this to food shortage. It appears this income source is "expandable" however it may be at the cost of school attendance.

Godere appears to be less affected not only from a crop production perspective as rainfall was relatively better but also because the people of Godere (Mezenghir) rely on honey production.

Fishing is conducted along major rivers in Jikaw, Itang and Gambella along Baro River, Gog and Jor along Gillo River, Jor and Dimma along the Akobo River. While when water levels in the river are high, prohibiting fishing with a spear. Regardless, in areas such as Jor, it is reported that when the river overflows it’s banks it fills ponds that are more easily fished during the wet season. In fact, at present, residents of Jor are bringing fish to sell in Pugnido markets. In Dimma, Woreda Officials reported that people were still fishing as the river levels are low but that once river levels rise it will be impossible.

Communities also engage in making a local beer "araki" from maize that in Dimma they buy from the market at 180bir/qt. They produce 25 liters that they sell for 10bir/lt thus a gross income of 250 bir or a 70bir profit (almost enough for a 50kg bag for consumption).

Wild foods are a common part of a normal diet given the rich landscape including the forests. Wild foods include leaves (awake and anado), roots (moda, key and ajegno), tree products such as tamarind and wodo which produces a rich oil. Only one farming family on Abobo reported eating wild foods but indicated this is a normal part of their diet. Abnormal consumption of wildfoods is not seen to be widespread at the moment but deserves close monitoring.

Communities of the same tribe have a tendency to share what is available. Indicative of this, refugee camp personnel indicated that it is normal that a steady trickle of relatives from the surrounding area will come to the camps to share food rations during the hungry season. At the moment, there are no unusual influxes.

2.6 Human Health

No reports of abnormal human health (epidemics) or malnutrition. The Sisters of Charity (Missionaries of Mercy, Gambella) report that in normal years during the wet season they receive and treat/feed 15 children per week who are moderately or severely malnourished. Beginning in December, this number goes up to 50 per week. However it must be noted that the children are not weighed and measured so only look malnourished and often malnutrition is secondary to malaria, which accounts for more than 80% of total admissions in the Sister’s Health Centre. Therefore while these figures cannot be used to indicate numbers of malnourished children, constant contact with the Sister’s who may notice a change in trend either in early onset (October/November) or increasing numbers, would indicate cause for further and immediate investigation.

In the UNICEF Situation Analysis, it was reported that in HIV/AIDS incidence has increased from 12.3% in 1993 to 19% in 1997 among women visiting antenatal clinics in Gambella town. The rising incidence of HIV/AIDS is also reported a concern to Churches and local NGOs. This is slightly higher than the national average (for urban areas 16%) and should be kept in consideration in food security analysis over the coming years.

3. Conflict

Conflict is increasing in Gambella since early 2002. Present conflict is largely contained between the Anyuak and Nuer in Itang. The underlying cause of the conflict is reportedly struggles for the arable land along the riverbank that the Anyuak normally farm. Several distributions (see below) of food assistance have been made to victims of displacement including recently in Itang to both Nuer and Anyuak, shelter, medicine and other non-food items by ICRC.

As the conflict is over resource management, it is clear that any emergency distribution whether for food, shelter or medicine, has to be targeted in a transparent manner or the distribution itself will be the cause of further conflict. Discussions with the DPPB indicate they are willing to facilitate a stakeholders consultation which will include government, local NGOs including Anyuak and Nuer, international NGOs, churches, UN agencies and others to ensure targeting and distribution modalities do not enflame the conflict.

It is also recommended that if food distribution is necessary in the future that ICRC who has a presence in the area and who is practiced in distribution of humanitarian assistance in conflict situations be consulted.

Late in 2001 and early 2002 there was also conflict between Anyuak and Mezenghir in Gog and Abobo (end of 2001) and between Mezenghir and Sheko in Yeki (SNNPR) and extended to Godere (March 2002).

Further discussion on the cause and effects of conflict in the Region will be elaborated in a subsequent EUE publication.

4. Relief Assistance in 2001

For the drought operation in 2002, only one distribution in July out of a recommended 4-6 distributions from March to August occurred, with the exception of Jor where three distributions occurred. While distribution was substantially less than recommended, there were no signs of food security stress during the period (DPPB Early Warning Bulletins February, March, April, May and June).

Analysis of previous estimated needs and amounts delivered without subsequent signs of serious food crisis leads one to question whether or not estimated needs have been overestimated over the last 4 years (with the exception of 2001 wherein no needs were identified). And while it cannot be said that there was not asset depletion or other negative coping mechanism that resulted from a lack of distribution, further analysis of this problem is deserved (see Scenarios).

On the other hand due to ethnic conflict in the region, in Feburary and March 90Mt (2x45Mt) were delivered to Abobo (conflict between Anuak and Mezenghir), 20Mt (2x10Mt) to Gog (conflict between Anuak and Mezenghir) and 34Mt (2x17Mt) to Godere (conflict between Mezenghir and Sheko). In April, an additional 36Mt were delivered to Gog (conflict between Anuak and Mezenghir). In June an additional 75Mt delivered to Akobo (conflict between Anyuak and Nuer).

The donor resources used were Swedish and Liberian and while they might not disagree that their resources were used for this purpose, their donations were given to the GoE (bilateral) for the Drought Appeal in early 2002.

4.1 Assistance Requirements for 2002 (September to December)

When prioritising the woredas as either 1) facing starvation, 2) malnutrition, 3) loss of productive assets, migration, other negative coping mechanisms or 4) close monitoring, none of the woredas are facing starvation in the next 3 months. Worse affected Peasant Associations in Abobo, Gog and Gambella may be at risk of malnutrition (less than 5% of population) while the rest of these woredas are at risk of negative coping mechanism. Worse affected are Dimma and Jor with a higher percent of population (more than 5% of population) at risk of malnutrition if the rains do not resume.

While it is unlikely that Gambella Region will require assistance before November, depending on the cessation of the rains (see below), it is moderately likely that the Region will need some assistance at the end of 2002 or beginning of 2003. Areas likely to require assistance are Dimma and Jor and pocket areas in Gog, Gambella and Abobo.

The team did not visit Akobo, Itang and Jikaw due to security constraints and therefore cannot rate the woredas on the same scale, however if the situation deteriorates it is likely to be due to conflict. Food assistance may be necessary for affected populations.

5. Scenarios for 2003

The interpretation of rainfall data, crop production, annual estimated needs and actual distributions is complex.

While in 2000, the rainfall was high, crop production was low, needs for 2001 were estimated to be nil and in fact in general the food security situation was stable possibly with the exception of Itang and Gambella. There were reports of malnutrition due to food scarcity in Itang and Gambella, and presumably a Disaster Area Assessment was done that resulted in two months food assistance to 7,500 beneficiaries in Itang and in 5,400 Gambella, however the results of the Assessment were unavailable in DPPB Gambella. There were floods in 2001 (Jor, Jikaw, and Dimma) that necessitated emergency food assistance but this was beyond the capacity of the 2000 ENA team to predict in December 2000.

In 2001, rainfall 25% less than the year before, production was an all time record high, cereal prices a record low, the ENA teams nonetheless recommended assistance for over 56,000 beneficiaries. At the same time, less than 40% of the assistance recommended was received and no reports of food insecurity were made by Woreda and Regional officials.

In fact, in Gambella Region, analysis of beneficiary data from 1978 reveals that between 1978 and 1991 there was little to no assistance recommended. Only in 1979, 1980, 1984, and 1986 on average 2,500 beneficiaries received assistance. While after 1992, each year Gambella has received assistance (with the exception of 2000) and on average 36,000 beneficiaries. It happens that in 1991/2, Gambella experienced an influx of refugees from Sudan however none of the assistance received since then has mentioned the refugees as a cause for food assistance to the general population.

However, for the sake of the exercise, let us assume that:

Best Case Scenario: Rains last until Mid October, rivers overflow and recessional agriculture takes place with normal area planted in October (Gambella, Itang, Abobo) or November (Jikaw), second harvest is obtained in January and February. Zero (0) beneficiaries.

Mid Case Scenario: Rains last until late September, recessional agriculture takes place with less than normal area planted in early October, less than average second harvest is obtained in January. In some areas, Woreda Officials reported that recessional agriculture can compensate for losses during the main Meher season (Gog). Twenty-five thousand (25,000) beneficiaries.

Worse Case Scenario: Rains last until early September, recessional agriculture takes place with much less than normal area planted in late September, very small second harvest is obtained in December. If rains receed early and rivers do not overflow their banks there will also be a concern for pasture in the dry season between January and June. At the same time if rains continue, and the river receeds as late as late October, recessional agriculture is at risk of low yields due to likelihood of high temperatures in January during flowering period. Fifty-thousand (50,000) beneficiaries.

Indicators to Monitor include:

6. Conclusion and Recommendation

At present there are no signs of severe food insecurity in the Region. Green maize is available for consumption. With river levels still low, fishing is still an option. Prices of staple cereals have not risen above normal levels for the season. Furthermore, as the people of Gambella are extraordinarily resourceful and the team received complaints of food aid dependency that reduce this initiative, it is not recommended that the present assistance be extended.

With regard to assistance in the near future (as soon as December), several factors need to be closely monitored and considered by the pre-harvest Meher Assessment teams:

• With regard to main season production, while production may be reduced from 2001 (a good year), it remains to be seen how it compares to 2000, which reported a reduction of 43% in crop production, but nonetheless survived without food assistance
  
  
• With regard to recessional agricultural production, river levels may not increase or overflow their banks or river levels may rise and then receed late (after October) thus affecting production
  
  
• With regard to conflict, if incidences increase and people are displaced

If the situation is seen to deteriorate, as a safety net for vulnerable groups, supplementary food assistance can be provided through clinics preferably supported by NGOs who can assist in growth monitoring and identifying malnourished children. These include ICRC in Itang (Ilea), Africare in Abobo and Gog and Missionaries of Mercy in Gambella.

It is therefore recommended that close monitoring of the cessation of rains (September or October) be conducted, in addition to close monitoring of the situation in Jor and Dimma as rains have been remarkably less in these woredas, and that the next assessment team visit Gambella in late October or early November to assess not only the final outcome of the main season but also the potential for recessional agriculture.

Finally of note, that the efforts of BoA, Woreda Officials, NGOs and others to encourage ox-plough farming should be supported. As evidenced in the increasing production in the Region (3095 Mt (1998), 3775 Mt (1999), 8912Mt (2000) and 14,835 Mt (2001) it appears to be having some success.

DISCLAIMER

The designations employed and the presentation of material in this document do not imply the expression of any opinion whatsoever of the UN concerning the legal status of any country, territory, city or area of its authorities, or concerning the delimitation of its frontiers or boundaries.

UN-EUE
PO Box 60252
Addis Ababa
Ethiopia
Tel.: (251) (1) 51-37-25
Fax: (251) (1) 51-12-92
E-mail: un-eue@un.org
www.telecom.net.et/~undp-eue/

Annexes

Abbreviations

Glossary

Glossary of important meteorological and seasonal terms used for Ethiopian highland areas

Meteorological Drought Defined

Drought is a period of insufficient water initiated by reduced precipitation. The impacts of drought on crops and society are critical but not easily quantified. The result is that "drought" does not have a universal definition. "Meteorological drought" is defined as a sustained period of deficient precipitation with a low frequency of occurrence. While crops may be damaged by lack of precipitation and high temperatures in just a few days, such short periods are not considered to be meteorological droughts. A three-month period is defined by the American Meteorological Society to be the shortest period that can be defined as a drought. (Source: The American Meteorological Society)

Ethiopia's 'Keremt' or 'Meher' Rains Defined

Since Ethiopia and Eritrea are in the tropics, physical conditions and variations in altitude have resulted in a great diversity of climate, soil, and vegetation. Rainfall is seasonal, varying in amount, space, and time. There is a long and heavy summer rain, normally called the big rain or keremt, which falls from June-September. It is followed by the baga hot, dry period from October through February (see below for definition). In some areas there are short and moderate spring rains in March and April known as the little rains or belg. These rainy periods correspond to Ethiopia's primary and secondary agricultural seasons, known as the meher and belg. (Source: FEWS)

Ethiopia's 'Belg' Rains Defined

In spring, a strong cyclonic centre develops over Ethiopia and Sudan. Winds from the Gulf of Aden and the Indian Ocean highs are drawn towards this centre and blow across central and southern Ethiopia. These moist, easterly and southeasterly winds produce the main rain in southeastern Ethiopia and the little spring rains to the east central part of the north-western highlands. The little rains of the highlands are known as belg rains, referring to the second most important sowing season of the region. (Source: FEWS)

Ethiopia's 'Baga' Season Defined

Since Ethiopia is in the tropics, physical conditions and variations in altitude have resulted in a great diversity of climate, soil, and vegetation. Rainfall is seasonal, varying in amount, space, and time. There is a long and heavy summer rain, normally called the big rain or keremt, which falls from June-September. It is followed by the baga hot, dry period from October through February. In some areas there are short and moderate spring rains in March and April known as the little rains or belg. These rainy periods correspond to Ethiopia's primary and secondary agricultural seasons, known as the meher and belg. (Source: FEWS)

(pdf* format)

Potential for Shea oil production in Gambella Region: a concept note to donors, NGOs and local Government

The author of this concept note is an independent consultant who has been working in rural development and natural resource management in East Africa since 1989 (and in Ethiopia for the last six years). Following a recent (but very brief) visit to Gambella Region, it became apparent that there may be a potential for developing of a valuable natural resource that could have a range of positive contributions to local livelihoods. This concept note is has therefore been produced to raise awareness among the different organisations who may have an interest to pursue the idea further.

Introduction to the shea tree and its produce

The shea tree, Vitellaria paradoxa (previously known as Butyrospermum paradoxum and B. parkii), a deciduous hardwood indigenous to Africa, grows wild in a narrow sub-Saharan belt of savanna wood land from Senegal in the west to Ethiopia in the east. For centuries the natural oils extracted from the kernel of its edible fruits have been valued by the local people as a high quality food and skin ointment with unique healing and moisturising properties. Because the golden oil solidifies at room temperature, the colonial explorers and botanists who first documented it in the 18th and 19th centuries referred to it as shea "butter".

The nuts are collected from the wild trees for processing by women in the villages and in many communities the shea tree is protected and left standing in areas cleared for cultivation. However, traditional processing of the kernels by hand is a laborious and complex process; it can take a single woman over 10 hours of hard labour to produce just a single litre of the oil. As such, the bulk of the shea resource remains largely unexploited in the countries in which it grows. Initially in west Africa, and more recently in Uganda and Sudan, several promising initiatives have been undertaken by NGOs to introduce appropriate technology that allows rural women to significantly improve the oil extraction process, significantly increasing both quantity and quality of oil produced and reducing the workload. These projects have helped not only to improve dietary intake but have made a considerable impact on generating household income.

The shea tree grows wild in several areas of Gambella Region where it is also valued by the indigenous rural communities as a source of a high quality oil, both for consumption and skin treatment. However, very little is known about the tree and its local use, either among the Government departments or NGOs operating there (it is not even mentioned in Azene Bekele’s "Useful Trees and Shrubs for Ethiopia" which is usually so reliable. That said, although not part of any official programme, the Regional Forestry expert (Ato Haile Mariam) in Gambella BoA has been conducting his own informal study of the tree, having attended an awareness raising workshop on shea butter in Senegal in 2001. There is even confusion as to the local name, but it seems that the tree and its oil are known as "wodo" is the Anuak language. Local informants also recognised that the oil is sold in markets (albeit in small quantities), for a as much as 15-20 Birr a litre.

The potential for improving local livelihood food security, empowering women and promoting sustainable natural resource management

Support for the appropriate development of this natural resource could make a significant impact to strengthening local livelihoods, by improving production for local production and income generation. Since collecting of the fruit and processing of the oil is traditionally a woman’s responsibility, the right sort of intervention could be directly linked to the economic and social empowerment of women. Further more, there is a clear link to sustainable natural resource management. The shea tree itself grows slowly and does not produce a full harvest of oil bearing fruit until it is some 15-20 years old (it lives for 200-300 years), and while natural regeneration can be prolific, is not readily grown from nurseries. Its potential derives more from the sustainable management of the wild resource which can grow in densities of 20 to 200 fruit-bearing trees per hectare. By adding value to the tree therefore, local communities can become motivated to protect it and the ecological conditions on which it depends.

Considerable progress has been achieved over the last 10 years in developing the appropriate technologies and methodologies for assisting rural women to benefit from the development of shea trees and oil. A simple and very low maintenance hand press has been designed and tested which can be manufactured locally (for around $500 on a commercial basis) and easily operated by women.

One such press can allow up to 40 women to produce in the region of 4,000 litres of high quality oil per year. Initial training and support is needed (not just in technical issues, but also for organisational management and marketing) but thereafter such rural press groups can sustain their own enterprises autonomously. Where appropriate, mechanical grinding units can be introduced, which can increase annual production from one group up to 40,000 litres (about 40 mt). Furthermore, local groups can easily manufacture high quality soap from sub-standard oil with simple equipment and materials (assuming a supply of caustic soda is available). Many such groups can be established and sustained in an area well provided with shea trees.

Already there are examples in Sudan of international aid agencies switching from using imported relief oil to purchase of locally produced shea oil (and soap). This avoids the negative impact of suppressing local production while allowing a considerable stimulus to local economic growth and development, helping once aid-dependent communities to initiate locally sustainable processes of self-help.

Ideas on how to proceed

Any agencies who would be interested to explore further the potential for developing the shea resource in Gambella would need to first undertake a simple feasibility study. This would involve:

• mapping out the main areas where the tree is grown
  
  
• consultation with local communities to understand existing indigenous knowledge, practices and ideas
  
  
• a brief assessment of local demand and market opportunities
  
  
• chemical analysis of the traditionally produced oil to ascertain its nutritional value

If such assessments were to generate encouraging information, a small scale pilot could be supported to introduce the technologies and methodologies and see in practice the potential of the intervention in the Gambella context. The author of this concept note, who has hands-on experience in shea development, could either advise directly or help identify other resource persons who could assist in feasibility studies and pilots. It would be advisable for any interested agencies to start communicating with each other so that a coordinated and collaborative approach could be established from the outset. The author would be prepared to give an initial briefing to programmers from all agencies, if someone was to take the initiative to organise such a meeting in which all interested parties were invited to attend. However, because of existing commitments, this would probably have to wait until November.

Prepared by:

Justin Corbett, independent consultant
tel: 1-514570; or e-mail: Kerren.Hedlund@wfp.org

Date: 5.9.2002

Circulated to:

Accord
Afri-Care
Catholic Mission
GTZ
ZOA
Forestry Department, BoA Gambella
FAO, UNICEF, UNHCR & WFP