SPECIAL REPORT

FAO/WFP CROP AND FOOD SUPPLY ASSESSMENT

MISSION TO ETHIOPIA

28 January 2005

• Mission Highlights
• 1. OVERVIEW
• 2. SOCIO-ECONOMIC CONTEXT
• 3. FOOD PRODUCTION IN 2004
• 4. CROP PRODUCTION SITUATION BY REGION
• 5. CEREAL SUPPLY SITUATION

Mission Highlights

1. OVERVIEW

An FAO/WFP Crop and Food Supply Assessment Mission visited Ethiopia from 8 November to 8 December 2004 to estimate the main meher season cereal and pulse production; review the final estimates of the 2003 meher and 2004 secondary belg season harvests; forecast the 2005 belg season production; assess the overall food supply situation; and estimate cereal import requirements, including food aid needs, for the 2005 marketing year (January/December). Accompanied by experts from the Federal Ministry of Agriculture and Rural Development (MoARD) and an EU observer, the Mission, in six teams, visited fifty zones and special woredas (districts), over a 20-day period, in all the grain producing regions. Parallel to the crop assessment teams but spread over a longer period, 23 teams, led by the Government’s Disaster Prevention and Preparedness Commission (DPPC) and with members comprising WFP, bilateral donor agencies and NGO personnel, visited marginal localities and vulnerable zones and woredas to determine their current and prospective food security situation.

The assessment teams obtained planted area and yield data for all major food crops from woreda, zonal, and regional agricultural bureaux, which were cross-checked against information from farmers, traders, NGO and donor project staff and remote sensed data from early warning systems. Crop inspections, spot-check crop cutting, market surveys, livestock condition observations and transect recordings of crops and their conditions were conducted en route to audit the information received. Where necessary, yield forecasts were fine-tuned to take into consideration latest and broader information.

The overall agricultural performance of the meher 2004 season is better than the previous year due to improved yields from a greater cultivated area in the main production zones. Despite generally poor belg rains in 2004 and a staggered start to the meher season throughout the country, more favourable conditions later in the year culminated in rains that continued until October/ November in many places. This has allowed farmers to take advantage of a better input supply and a migratory pest free year.

Good rainfall in the central highlands and in the western pastoral areas increased available forage and water, improved livestock condition, decreased mortality rates of young stock and removed the necessity for unseasonable and early migration of herds and flocks. However, pastoral areas in Somali Region and the southern zones of SNNPR and Oromiya have not had similar benefits until the recent rains and are reported to be subject to water and forage shortages with associated difficulties that need to be monitored carefully. In the north-east, where no late rains are expected, premature migration of livestock herds out of Afar is anticipated along the Afar-Amhara and South Tigray borders.

Overall, the Mission puts total meher cereal and pulse production at 14.27 million tonnes, about 24 percent above the previous year’s MoARD post-harvest estimates and 21 percent above the average for the previous five years. With a predicted belg harvest of 250 000 tonnes in 2005, total domestic availability of cereal and pulses is estimated at 14.59 million tonnes. As a result, the country’s cereal import requirement in 2005 is estimated at about 117 000 tonnes.

The grain price movement observed in 2004 followed a typical seasonal pattern of a steady month-on-month decline in grain prices soon after the start of the main meher harvest (September-January), then increasing thereafter during the pre-harvest months (July-September). As is generally the case, grain prices are expected to decline when cereal flows into the market peak toward mid January through February 2005. Continued and increased local purchase of cereals for food assistance needs, which averaged around 200 000 tonnes in the last four years, would provide some firmness to the grain market. There should also be scope for a significant build-up of stocks at all levels.

For the first time in the history of food aid appeals for Ethiopia, there is a difference in approach in addressing the needs of chronically food-insecure people and acutely food-insecure people. Some 2.2 million acutely food-insecure people will need emergency food assistance in Ethiopia during the year. In addition, some 683 000 people in Somali Region and 250 000 people in Afar Region, who will eventually be covered under the safety-net programme, will require emergency food assistance for the first half of 2005. Total emergency food needs in 2005 are, therefore, estimated at 387 500 tonnes (see Tables 13 and 14). The Productive Safety Net Programme (PSNP), a new food security programme, is aimed at tackling longer-term food security needs and is expected to eventually reach 5 million chronically food-insecure people in 2005 with cash and food transfers.

Thus while the total number of people targeted for assistance has not drastically changed from that of the previous year, the number that now fall under the category of emergency food needs is significantly reduced. Safety Net transfers for the pastoralist populations in Afar and Somali Regions are scheduled to start later in 2005, and the emergency programme will cover all food needs in the two regions for the first half of 2005.

Results of nutrition surveys conducted by the Government and NGOs in worst affected woredas throughout the country in 2004 indicated a national average Global Acute Malnutrition (GAM) of 9.6 percent and Severe Acute Malnutrition (SAM) of 0.9 percent. Since March 2004, targeted supplementary food has been distributed to malnourished children and pregnant and lactating women in SNNPR under the joint WFP/UNICEF/Government of Ethiopia Enhanced Outreach Strategy (EOS) for Child Survival Interventions (CSI). The programme is being extended to several other regions, and in 2005 is expected to target 6.8 million children aged 6 to 59 months in more than 320 drought-affected woredas (emergency and Safety Net woredas) of the country. It is estimated that around 700 000 malnourished children and 300 000 pregnant and lactating mothers will require supplementary feeding in 2005 under the EOS, with requirements for this programme being 89 000 tonnes of fortified blended food and vegetable oil.

2. SOCIO-ECONOMIC CONTEXT

2.1 Macroeconomic situation¹

The Ethiopian economy is highly dependent on agriculture, which contributes to about 45 percent of GDP, followed by 43 percent from the service sector, and 12 percent from the industrial sector. Growth in GDP surged to 11.6 percent in 2003/04 after a negative growth of 3.8 percent in 2002/03 due to a severe drought. This recovery is mainly due to the rebound in agricultural production in 2003/04.

The momentum in the balance of payments surplus, which began in 2001/02, has continued in 2003/04 depicting a favourable development. Though lower than the US$ 278.1 million and US$ 302.4 million surplus recorded in 2001/02 and 2002/03, respectively, it registered a surplus of US$ 138.9 million. Therefore, compared to previous year, the surplus in the overall balance declined by 54.1 percent due to the huge trade deficit and the decline in official transfers despite the significant surge in net services and private transfers. Export earnings registered a 24.4 percent growth as its level reached US$ 600.7 million in 2003/04 while at the same time imports rose sharply by 39.4 percent in the same period to reach US$ 2 587.4 million.

In line with the observed 5.5 percent slowdown in official transfers and the widening of the trade balance deficit, the current account deficit worsened significantly from US$ 156.9 million in 2002/03 to US$ 495.8 million in 2003/04. The trade deficit has increased from US$ 1 373.7 million (20.7 percent of GDP) in 2002/03 to US$ 1 986.7 million (24.7 percent of GDP) in 2003/04. The increase in the trade deficit is mainly due to the significant surge in the international price of imported commodities (e.g. fuel and steel) and the ongoing capacity building activities of the country.

Ethiopia’s capacity to import goods and non-factor services has improved to reach 5.1 months compared to 4.5 and 3.4 months of import in fiscal years 2002/03 and 2001/02, respectively. This improvement in the level of international reserves is mainly due to the build-up of the net foreign asset position of the National Bank of Ethiopia and the secured US$ 81.5 million exceptional financing in the form of debt relief. While this latter is very encouraging, looking at the absolute figure, this exceptional financing in the form of debt relief could barely compensate for the hike in the world prices of fertilizers and fuel products which led to a drastic jump in their respective import bills by up to 97.9 percent and 8.1 percent, respectively.

As of end-June 2004, the external debt stock of Ethiopia has shown a 6.2 percent increase compared to the previous year and stood at US$ 7200.3 million. However, the external debt measured in terms of its ratio to GDP or to exports of goods and non-factor services registered a decline in 2003/04 owing to the enhanced HIPC relief assistance. Indeed, while the external debt to GDP ratio went down from 101.9 percent in 2002/03 to 89.7 percent in 2003/04, the external debt servicing as a percent of exports of good and non-factor services has continued to decline from 14 percent in 2002/03 to 12.3 percent in 2003/04.

Reflecting the Government commitment to enhance the competitiveness of the export sector, the official exchange rate of the Birr continued to slightly depreciate. Indeed, the weighted average exchange rate data revealed that the value of the Birr against the US Dollar has shown a 0.45 percent annual depreciation to stand at US$1=Birr 8.61 in 2003/04 and US$1=Birr 8.58 in 2002/03 and US$1=Birr 8.54 in 2001/02.

Table 1 - Ethiopia: Key economic indicators, 2001-2004

Ethiopia’s major agricultural export commodities are coffee, pulses, oil seeds and chat (Table 2). Other exports include sugar and molasses, leather and leather products, live animals, canned meat and frozen foods, fruits and vegetables, gold, etc. Export revenues in 2003/04 increased by 24.4 percent to US$ 600.7 million compared with earnings of US$ 482.7 million in 2002/03.

Table 2 - Ethiopia: Major commodity exports (2001-2004)

Coffee export earnings increased by 35.3 percent and stood at US$ 223.6 million in 2003/04 compared to the previous year, due mainly to the jump in the volume of coffee exported by 24.3 percent, registering a record high of 159.7 thousand tonnes and coupled with the mild recovery in world prices. The increase in the export earnings from pulses is due to the 10 percent pick-up increase in its volume exported in 2003/04 compared to the previous year.

The best-ever performance of oilseeds in 2003/04 is due to a rise in both the quantity exported (28 percent) and the commodity international price (39.2 percent). On the other hand, the improved performance of chat is solely accounted to the 79 percent hike in its unit price.

Total imports increased by 39.4 percent and 52.6 percent to US$ 2 587.4 million in 2003/04 from US$ 1 856.4 million in 2002/03 and US$ 1 695.7 million in 2001/02, respectively. The major import commodities in 2003/04 are semi-finished goods (US$ 435.3 million), fuel (US$ 310.6 million), capital goods (US$ 876.7 million), consumer goods (US$ 895.8 million) and miscellaneous goods (US$ 43.1 million).

While Official Development Assistance (ODA) trends for Ethiopia reflect a steady increase in magnitude from US$ 925 million in 2000 to about US$ 1 920 million in 2003, the humanitarian/relief component has consistently been the largest representing on average about the third of total ODA. In fact, in 2000 the humanitarian/relief aid consumed about half of total ODA and has been several multiples of that invested in other socio-economic sectors such as agriculture, forestry and fisheries, health and education in subsequent years. On average, about 33 percent of ODA funds are allocated annually for humanitarian assistance compared to 6 percent for agriculture, forestry and fisheries and 10 percent for transport infrastructure (Figure 1).

Figure 1 - ODA by selected sectors, average 2000-2004

2.2 Population

The population of Ethiopia for the mid-year 2005 is estimated at 73.044 million comprising 61.369 million rural (84 percent) and 11.675 million urban (16 percent), respectively. These estimates are based on the 1994 population and housing census of Ethiopia conducted by the Central Statistical Authority under the auspices of the Office of the Population and Housing Census Commission in 1994 and released in June 1998. The overall annual population growth rate is estimated at 2.78 percent.

Table 3 - Ethiopia: Total population size by region and by sex in 2004 (millions)

2.3 Agricultural sector

Agriculture in Ethiopia is the main economic activity, contributing to about 45 percent of GDP with some 80 percent of the population earning a living directly or indirectly from agricultural activities. The agricultural sector is nearly totally dependent on rainfall with only 2 percent of the total arable land being irrigated. The proportion of area under improved seeds is less than 3 percent and the proportion of area treated with pesticides is less than 10 percent. This, coupled with low fertilizer use, susceptibility to pest and disease outbreaks and extensive highland soil erosion, has meant high variability in year-to-year agricultural production, which is predominantly in the hands of peasants working smallholdings². This high variability in agricultural production increases food insecurity in the country.

The relatively low crop sub-sector productivity in Ethiopia is not only the result of the low adoption rate of the many yield enhancing technologies, but also the poor promotion and marketing schemes, Government policy in relation to tenure security and agricultural terms of trade. Indeed, volatility of prices of agricultural products has seriously constrained production and adversely affected farm income, particularly as prices collapse in periods of bumper harvest.

On the institutional side, the prevailing land tenure system in Ethiopia and the constraints over the issues of transferability of land rights do not encourage farmers to invest on land and water development and adopt expensive technological packages. This coupled with high population growth in the rural areas can only bring about disincentives to investment in the land and lead to inappropriate land management.

The livestock population in Ethiopia is estimated to be the largest in Africa and ninth in the world. However, this sub-sector is also characterized by a low-input system based on common grazing and the use of crop residues. According to CSA survey, almost 99 percent of the cattle, sheep and goat population in the country are indigenous. Basic production statistics regarding average birth rates and losses are not available. The current national production of milk from indigenous cattle is about 400 litres for a lactation period, whereas cross-breeds (which represent only 1 percent of the whole population) give over 3 000 litres during the same period.

2.3.1 Agricultural input credit

The Commercial Bank of Ethiopia (CBE) is the largest source of agricultural credit in the country. During the current cropping year (2004/05), CBE approved a total of 978 million Birr of agricultural input loans based on credit requests submitted by the regional governments - Oromiya, Amhara, SNNP, Tigray, and Addis Ababa. Table 4 presents the total agricultural input credit approved, disbursed, and overdue for the last five years.

Table 4 - Total agricultural input credit approved, disbursed and overdue (2000 to 2005)

The amount of agricultural credit approved by CBE for the cropping year 2004/05 is about 25 percent higher than 2003/04 and the highest for the last five cropping years. The credit repayment default rate has been kept under control and it is expected that it will decline in the coming year given the good harvest prospects. Furthermore, the regional governments in their capacity as guarantors of agricultural input loans are implementing measures to reschedule part of the past due loans. The interest rate on these loans is 7.5 percent shared between the CBE which receives 5.25 percent on the disbursed amounts and regional governments which receive 2.25 percent for loan disbursement, recovery, and administrative charges.

3. FOOD PRODUCTION IN 2004

3.1 General

In Ethiopia, of the 11 million hectares presently farmed to all crops, only some 190 000 ha are irrigated; consequently, production varies considerably from year to year depending on the quality and quantity of the annual rains. The crops grown are diverse following the complicated mosaic of agro-ecologies derived from soil types ranging from vertisols to sand and cropping altitudes ranging from more than 3 000m to less than 600m above sea level. The main cereal staples include wheat, barley, teff, finger millet, maize and sorghum grown in varying proportions according to soils, altitude, and the prevailing climatic and market conditions of the year. Other carbohydrate sources include the stem of enset or false-banana, cassava, potatoes and sweet potatoes, all of which are found in either the middle altitude or highland areas of the south/central regions of the country. Cash crops include oilseeds, spices, coffee, chat and eucalyptus, the tree crops being found as hedgerows, on-farm woodlots and in forests in the middle altitude and highland areas.

Common grasslands provide extensive pasture and browse for livestock in most regions, but are particularly important to livestock producers in the eastern regions of Afar and Somali, the southern zones of Bale, Borena and South Omo, and in the western lowlands that reach from Gambella to Tigray. National livestock production from such pastoral areas, is augmented by the settled agro-pastoralism of peasant farmers throughout the Central Plateau and the escarpments of the Rift Valley, using common grazing, browse and crop residues to produce sheep, goats and less frequently, dairy cow products for sale and home use. Livestock are further integrated into the farms through the universal use of animal traction for ploughing, secondary cultivation, threshing, and the transportation of goods and commodities.

3.2 Rainfall 2004

Rain in Ethiopia falls in two distinct seasons: (i) the belg, a minor season that usually begins in January-February and ends in April-May; and (ii) the meher or kiremt, the main rainy season, which starts in June-July and ends in September-October. In some ten zones in central and northern parts of the country, belg rains are regularly enough to support the generally opportunistic belg harvest which may, in a good year, account for 5 percent of national cereal production. Elsewhere, belg rains offer the opportunity for land preparation and improve pasture and browse after the dry season. The melding of belg and meher rains in the south-west zones often generates one long season without clear-cut breaks, which although good for perennial crops and the long-maturing stover cereal varieties, is less than ideal for the early maturing grains.

In 2004, except for good rain in April the belg season was poor with hardly any rain in May being recorded across the country, which is reflected in the poor belg harvest given in Section 3.8 below.

Regarding the meher season, the six Mission teams dispatched throughout the country to determine agriculture production and conditions, collected qualitative and quantitative meher season rainfall data from all zones and woredas visited. The combined returns confirm the National Meteorological Institute’s rainfall gauge data from 80 stations provided to the Mission that show that in most of the 53 zones and special woredas identified as Mission entry points, meher rains were considered to be as good or better than the previous year. In 40 Mission entry-points, the 2004 meher rains were considered to have been “normal,” that is to say they conformed to the expected pattern, they began on time, they were reasonably evenly distributed during the season and they either finished on-time or later than expected. Only in 13 zones/special woredas were less than satisfactory reports filed. These reports included later starts and an erratic distribution with breaks of 10 to 20 days noted and early finishes. Such sites are located mostly in the south, with the worst examples being in South Omo, Konso, Borena and Gamo Gofa and in the east/north-east including woredas in East Tigray, adjacent woredas in a triangle between Central and South Tigray and Weghamra (Amhara) and in the Afar zones.

Despite the variable nature of the rainfall inherent in the semi-arid areas of Ethiopia, which means that in any zone and in any year there are always communities, particularly in the lowlands, that will experience a less than satisfactory rainfall, the 2004 meher rains may be characterised as being reasonably well-received. Except in SNNPR, where a 60 000ha shift to short cycle crops is noted, the timely start sustained the area sown to long cycle crops that had shown a dramatic increase in 2003, following the reduction in maize and sorghum cropping in 2002; supported germination and vegetative development of all cereals and pulses and provided adequate moisture at flowering and grain fill. The continuation of the rains into October and November encouraged late, opportunistic planting of short cycle crops and supported their development, adding a further positive aspect to the season. Rains in December that fell during the Mission have, at the time of reporting, had only positive effects. In Central Tigray, harvesting campaigns conducted as a precaution against possible storm-related losses, secured all vulnerable crops. Elsewhere in the north, the harvests of short-cycle crops were either completed or well- advanced at the time of the Mission, or, as in the cases of late planted sorghum in North-West Tigray and very late planted barley and pulses throughout Awi zone, crop production will only benefit from the continued precipitation.

Regarding the effect of rainfall on pasture and browse, the poor belg rain generally and the absence of rain in May in particular, reduced the availability of an early-bite compared to the previous year in most areas and raised concern for the well-being of livestock. However, the main season rains have been widely distributed and prolonged and, even if erratic in nature in some places, have increased forage production and improved water supply in all areas except the north-east. As the November/December rains in the south and south-east are reported to be benefiting the herders and grazers previously thought to be at risk in Somali, Borena and South Omo, only in the north- east are early pastoralist movements anticipated.

3.3 Area planted

Mission estimates for area planted are derived from several sources. The main source is area data collected by Bureau of Agriculture (BOA) Development Agents (DA) based at Peasants’ Association (PA) level throughout the country. Such data are aggregated at woreda level, passed on to the agricultural desks at zonal level, where they are reviewed and transmitted to the Regional level offices.³ In addition, the Mission teams collect data on area and production from farming companies, investors, and any state farms that are operational in the zones visited. These data are all included in the regional estimates shown in Table 5 below.

It should be noted that, in 2004, all the Woreda Agricultural Bureaux and, consequently the Zonal Agricultural Desks in Amhara Region have used forecasts issued by the Central Statistics Authority as a basis for their meher crop estimates rather than continuing with their own data collected time-series. This parallel data set resulting from the 2001 Agricultural Sample Census contains markedly different area estimates, which are not considered by the other major grain producing regions to describe either their agricultural areas or the proportional distribution of crops grown. In 2003 all but three Amhara zones opted to retain their own methods of assessment. The 2004 decision to base all returns in Amhara on CSA data reduces the previous year’s comparator by 24 percent (716 000ha) with a concomitant effect on production data at regional level and area and production at national level. Presently, pilot studies are being conducted by CSA on the census data comparing, inter alia, the international unit value of various local land measures used by farmers and DAs to assess crop area. The EU observer to this Mission accompanying the team in Amhara undertook a pilot study to compare CSA/BOA land area estimates in six major production zones against remote-sensed data identifying potential agricultural areas in the wurch, kola, weina-dega and dega ecological zones, with a view to determine the closest fit. Initial findings from the two pilot studies point to a smaller than recognised value for the local land measures in one zone in Oromiya, and, by contrast, a closer match between BOA data and remote-sensed potential agricultural areas in the six zones in Amhara. Further studies of a similar nature are clearly required urgently to resolve the dichotomy.

For the time being, the Mission estimates are based on data collected from the BOA offices, which with the exception of Amhara, Addis Ababa and Afar regions where they are linked with CSA data sets, have been compiled by the system of aggregation the BOAs usually apply, as noted above.

Consequently, the Mission estimates that the national area planted to cereals and pulses during the 2004 meher season is 10.639 million hectares, which is 4 percent lower than the previous year’s Mission estimate of 11 055 million ha but is 4 percent greater than the BOAs’ adjusted post harvest estimates for 2003. Similarly derived estimates for 2004 belg season suggest that in 2004’s meher planting follows a 6 percent reduction in belg planting. The increased meher area is, therefore, explained by increased meher use of the belg area as well as Mission noted (i) expansion of commercial farms in the western zones, (ii) the resettler programme and (iii) an increased use of fallow land in Oromiya.

Closer examination of major cereal areas at national level reveals that maize and sorghum areas have been sustained at the levels achieved in 2003. However, these data mask return to lower maize and sorghum areas in SNNPR due to a 54 000 ha (13 percent) and 27 000 ha (21 percent) reductions, respectively. At national level, this drop is compensated by numerically similar area increases in maize and sorghum planting in Oromiya and Amhara. At the same time, teff, wheat and barley and the total pulse crops register a 76 000 ha increase in SNNPR, sustaining the level of land use. National area increases in teff (5 percent), wheat (7 percent), barley (5 percent) and pulses (11percent) confirm the anticipated, greater interest in late-planted, short-maturing grain crops in 2004, due to the shortage of rain in May. In the SNNPR where maize is grown, most farm families eat enset, therefore, despite fairly similar farm sizes to the peasant farms in Oromiya, the remarkably high level of carbohydrate production from well-established enset gardens at around a minimum of 130 tonnes fresh weight to the hectare, releases the maize crop for sale. Maize has also the advantage of an early, green- cob market that provides a lucrative income when early rains are favourable. As an effective cash crop, financial returns from maize influence planting decisions more strongly than where maize is needed mostly for subsistence. In such areas teff and selected pulses provide cash crop substitutes for maize, being agro-ecologically interchangeable and having a firmer recent price history; the data from SNNPR suggest that such a shift occurred in 2004.

Countrywide, given the sustained levels of cultivation achieved in 2003, there does not appear to have been any widespread constraints on ploughing capability. However, in the wetter, forested areas, where the risk and effects of trypanosomiasis are high and the small size of the farms precludes the effective use of the normal four- wheeled tractors but where timeliness of cultivation, sowing, and weeding is of paramount importance for the production of a satisfactory series of crops to achieve food security, the general lack of alternative power sources to oxen, points towards the need to consider the introduction of the diesel engine, two-wheeled, hand-tractor to improve the efficiency of husbandry practices and place less stress on the natural resources used to maintain the oxen.

Following the 2003 good harvest, seed supply was not a constraint on 2004’s meher planting. Given that some 680 000 tonnes of seed grains were estimated to have been used in 2004, most seed sown came from farmer carried-over stocks from the previous year, however, returns from the National Agricultural Input Suppliers’ Association show that in 2004 improved seed sales increased by 32 percent to 21 000 tonnes from the previous year’s final estimate of 15 700 tonnes.

3.4 Factors affecting yields

The national yield averages compare favourably with averages estimated over the past five years reflecting a similar or an improved performance of crops in almost all regions except for regional cereal yields in Tigray and Addis Ababa; and regional pulse yields in SNNPR. Presently, under the prevailing BOA system, woreda specialists assess yields at pre-harvest and post-harvest stages for all field crops. Such data are then transferred to the zonal or regional desks for review, analysis and onward passage. Because of the timing of the exercise, the Mission teams usually only receive the earlier yield assessments, which are then adjusted during the Mission with the assistance of the key informants, to take into account field observations, measurements and any changing conditions regarding the weather and late pest and disease challenges. All teams used the Pictorial Evaluation Tool (PET), developed by the Centre for Arid Zone Studies, University of Wales, Bangor, UK, in 2004 to good effect to add more consistency to the auditing approach adopted. Such assessments are subject to rigorous review when the Mission teams return to base. At this stage, assessed performance is reviewed in respect of seed type, timing of sowing, extent and timing of fertiliser use, the season’s pest and disease profile, the performance of similar crops in neighbouring localities, time-series data and compared with any other independent assessments available for the zone.

In 2004, 97 percent of the seeds used were local seeds carried over from the previous harvest. In the surplus areas, such seeds are mostly open-pollinated releases from government seed agencies that have stabilised over the last two decades and have acquired local identities reflecting their provenance (in Ofla, South Tigray the Global 2000 programme wheat releases, locally called sekokwa, are noted by the Mission to be producing 3.2 tonnes/ha under prevailing peasant conditions). In the more marginal areas, as well as such seeds, local landraces are also in evidence and are exchanged/sold between farm families as needed. In 2004, seed assistance programmes were virtually absent being restricted to a few emergency programmes in response to localised problems. Of the remaining 3 percent, amounting to the 21 000 t of improved seed sold, 5 000 t were maize seeds and 14 000 t were wheat seeds. It is anticipated, therefore, that 12 percent of the maize and 8 percent of the wheat sown in 2004 will have been newly purchased improved varieties, sustaining the levels of introduced varieties noted the previous year. Regarding maize, given the low seed rates, the propensity for farmers to carry over seeds from year to year and the fact that that most seeds released are open-pollinated varieties that will produce at the expected level for several generations, if the present levels of sales are sustained it is possible that the maize population, in the serious maize growing areas, will be completely upgraded within another four years or even sooner given the likelihood of farmer to farmer exchanges. Under reasonable environmental conditions and provided that fertilisers are used maize production should increase by 600 000 tonnes per year (200 000 ha @ 3 tonne per ha increment) through seed change alone. Coincidentally, in 2004 production is estimated by the Mission to have increased by 586 000 tonnes but this also includes better estimates of performance in zones where no improved seeds have been released.

A similar, if slower, progression is to be expected with the wheat crop. Sales of 14 000 tonnes of improved wheat seed connect to a possible production increment of 225 000 tonnes from 112 000 ha or 7 percent of the wheat area. In 2004 Mission estimates suggest a production increment of 568 000 tonnes again including better estimates of performance from all zones, with or without improved seed sales.

Given the favourable rainfall, no widespread replanting was necessary, and the continuation of the rainfall meant that where replanting did occur, the rains supported the growth and development of the replacement crop as well as the main crops in most areas, reinforcing the role of rainfall as the single most important determinant regarding crop performance in Ethiopia.

Reversing the trend noted by the Mission in the past three reports, fertiliser use as indicated by cash and credit sales, increased by around 19 percent to 323 000 tonnes and is higher than in any year since 1996 despite significant increases in prices of DAP (diammonium phosphate) and urea that have raised DAP prices to c.365 Birr per quintal (US$ 420/tonne) and urea prices to 280 birr per quintal (US$ 320/tonne). These current prices connect to FOB prices for DAP at around US$ 200 per tonne are now around twice the retail price of compound/nitrogenous fertilisers sold to farmers in the UK. Nevertheless, demand appears to have matched supply, except in Illubabor where supplies arrived late and use dropped by 11 percent; and in Tigray where unfavourable rains reduced use in the North West, Central and East zones. Notwithstanding the foregoing, fertiliser distribution to the regions was similar in proportion to the distribution in meher season the previous year. Tigray received the lowest share at 2.7 percent (5 percent in 2003); Amhara received 31 percent (29 percent); Oromiya received 46 percent (44 percent); SNNPR received 10 percent (8 percent) and the remainder being 10 percent (13 percent in 2003), was sold to farmers in the remaining regions and to various commercial enterprises. Zonal distribution patterns are more informative and indicate a return to recommended levels of use in the Oromiya wheat growing zones of Arsi, NW and SW Shewa where increases in use of 54 percent and 30 percent are noted; and in the maize and teff growing zones of East and West Gojam, where similar increases occurred.

Regarding pests and diseases, the only migratory pest scares noted by the Mission teams were outbreaks of migratory quelea birds that were controlled by aerial spraying in the four zones, Oromiya, North Shewa, Borena and East Hararghe and Konso special woreda, where they made an appearance.

Non-migratory pests are noted to have been of little significance in 2004, nevertheless, they included infestations of sorghum chafers in East Tigray, the eastern zones of Amhara bordering Afar Region; stalk borers, termites, boll-worms, and birds, the control of which through bird-scaring continues to place a great and underestimated demand on household labour, particularly in sorghum growing areas, if heavy losses are to be avoided.

Storage pests, especially weevils, are noted to be, as usual, as a cause for concern throughout the country but they are particularly important in the wetter south- western zones, where stored maize losses are noted to be as high as 40 percent in good rainfall years. In 2004 an increased use of storage chemicals was reported, however, no sales figures were available to confirm its improved availability. The increased production of maize (20 percent) and pulses (59 percent) suggests that annual grain storage losses will be higher in marketing year 2005 than during 2004 and have been proportionally calculated for use in the grain balance sheet to be 11.5 percent.

The adverse effects of crop diseases were also mild, with only yellow rust on wheat in Arsi and smut on sorghum in East Hararghe and Somali reported as causes for concern. The presence of sorghum smut was also identified by Mission teams in the fields of South Wollo and South Tigray, but was seen to be of very little concern to the farmers whose fields were infested. Local seed treatment is carried out using cows’ urine in some of the other localities visited in the same zones and in West Tigray, where the disease is also present. It is noted that in the sorghum growing areas of Sudan bordering the expanding mechanised farming units in the western lowlands of Tigray and Amhara, treatment of sorghum seed against smut with proprietary brand seed dressings is almost universally practised by the commercial farmers.

Weed competition was again fierce in 2004, as the good distribution of rain generally enhanced all plant growth. The Mission teams noted an increase in frequency of hand-weeding of most crops in all regions, and more reports of “shillshallo” or animal- powered thinning of maize and sorghum crops, followed by inter-row cultivation, were evident in 2004. There was also an apparent increase in the use of herbicides. The Mission notes the use of 2.4 D by farmers as diverse as investors in West Tigray, (where labour rates were noted above the previous year’s 30 Birr per day for workers harvesting sesame under piece work contracts), wheat farmers in Arsi and Bale, teff farmers in Jimma and mixed cereal farmers in East Gojam, the common elements linking their choice being the vigorous growth of weeds and a shortage of labour.

The combination of the positive factors noted above and the well-distributed rainfall described previously, explains the overall improvement in crop performance in all the main production regions except Tigray as manifested by the universal increase in yields compared to the previous year for cereal and pulses listed in Table 5 below. These are considered by the Mission to be due to (i) the direct effects of well-distributed rainfall on crop growth and development, (ii) timely pre-season cultivation and main season husbandry, (iii) increased investment in inputs due to improved availability of fertilisers, improved seeds and credit, (iv) better financial returns to cereal growers the previous year.

3.5 Other crops

Crops contributing to household food security vary from north to south and from east to west. In the north, oilseeds, particularly nuq and sesame, are important to both peasant farmers and commercial producers. In 2004 national oil seed area increased by 38 percent or 230 000 ha of which 135 000 ha is noted to be increased sesame area in Amhara and Tigray, the remaining increment coming from expanded crop area in West and East Wellega in Oromiya region.

Given the diverse nature and generally favourable conditions for plant growth of the southern half of the country, a greater range of other crops contribute to the household’s economy. In SNNPR and the southern zones of Oromiya, crops other than cereals and pulses occupy 12 percent and 32 percent respectively of the planted area compared to 3 percent and 7 percent in Amhara and Tigray. Of these the importance of enset, which provides the main carbohydrate staple for some 8+ million people and makes a substantial contribution to the diet of an additional 4 million people, is well understood. Data from southern zones are incomplete in 2004, however the Mission teams in the enset area noted no reasons to suppose that enset harvesting is not in balance with replanting, suggesting that the area noted the previous year will have been sustained. Enset condition is noted to be good with yields at normal levels. Annual roots and tubers, mostly in the same agro-ecological zones as enset, have also performed well during the meher season, with sweet potato and potato yields of 15t-30t/ha as recorded by Mission team members, during a separate study in 2003 being achieved, in 2004.

Coffee production in 2004 is expected to be similar to the previous year, according to the Coffee and Tea Authority specialists interviewed by Mission teams. Growing conditions during the year were good in all zones. Coffee berry disease remains a serious problem except where trees have been replaced by resistant varieties through GOE supported nursery programmes.

Production from the other industrial field crops such as tea, sugarcane and cotton and the performance of chat, a mild narcotic cash crop grown throughout the southern half of the country in small backyard plots, are reported to be similar to the previous year.

3.6 Livestock

Ethiopia has the largest livestock inventories in Africa, including more than 38 million cattle, 30 million small ruminants, <1 million camels and 4.5 million equines and 40 million chickens (CSA, 2004), with livestock ownership currently contributing to the livelihoods of an estimated 80 percent of the rural population. In the arid and semi-arid extensive grazing areas in the eastern, western and southern lowlands cattle, sheep, goats, and camels are managed in migratory pastoral production systems. In the highlands, livestock are kept under settled or transhumant systems utilising common pastures, many of which have a high clover content and crop residues. Such livestock includes some 9.3 million oxen providing draught power, for the mixed farming system that prevails.

Recovery from the 2002 drought, which reportedly reduced livestock numbers in some parts of the pastoralist areas, notably Afar, by nearly 50 percent, has been set back in the same areas in 2004 due to meagre rainfall and until the recent rains, conditions have been cause for concern in the southern pastoral areas of the country. Elsewhere, improving pasture and reasonable availability of drinking water has resulted in enhancing livestock body condition in the central highlands and western zones to the good body condition scores of 3-4 noticed the previous year with the notable exceptions of Meket, (North Wollo), Mehoni (South Tigray) and associated tabia bordering Afar Region

No unusual disease outbreaks have been noted except concerns about internal/external parasites and reports of endemic infectious diseases such as pasteurellosis, anthrax, blackleg, CBPP, and CCPP. Trypanosomiasis is routinely cited as a concern in western and southern lowlands, but the condition is not reflected in the observed body condition scores in 2004. In several zones, farmer and BOA sources indicated shortages of vaccines and other veterinary pharmaceuticals at local levels, where decentralisation has lead to difficulties in ensuring ready supplies of prophylactic and curative medicines.

Mission teams report no early migration of pastoralists due to water shortages and poor pastures inducing premature livestock movements. However, reports of imminent movement from Afar to the neighbouring zones of Oromiya, South Wollo and South Tigray reached the Mission on arrival in Addis Ababa. In southern Borena, Hararghe while good November/ December rains have been reported, pasture growth and browse remain fragile and heavily dependent on rainfall in the coming months.

In Afar, in South Omo and in two zones in Somali livestock prices are falling as pastoralists sell to reduce numbers to carry during the dry season on the limited forage. Elsewhere, throughout the country livestock prices are firm and rising, boosted by firm cereal prices in the central plateau and the west and the opening of meat processing units in the east.

Regarding the use of feed grains, information is scarce. On the one hand, the modern poultry industry producing eggs and broilers is served by private feed mills generating some 80 000 tonnes of poultry feed per annum to accommodate an estimated 1.5 million layers and 1 600 tonnes of broiler meat produced annually. About 70 percent of the components of the rations are estimated to be home-grown cereals. Feed grain use in the traditional backyard poultry industry, on the other hand, is far less easily assessed. Given that the backyard chicken population has recently been estimated by MOA at 56 million birds (7million households, 8 birds per household) and by the CSA at 38 million birds, assuming that every household feeds one menelik/wollo (0.7 kg) of home-produced cereals to the birds once a week, then the feed use is in the order of 254 000 tonnes per year. Mission observations suggest that both the grain ration and frequency of feeding are usually greater than assumed above. Therefore, the traditional and modern poultry industries may consume around 330 000 tonnes of cereals per year. In addition to chickens, cereal – based feeds are also given in limited quantities to working equines, draught oxen at ploughing time, fattening stock for the elite markets and 156 000 grade and pure bred dairy cows. Again, information on rations and frequency of feeding, outside the small modern sector is scanty, however, whereas it is understood that the bulk of the supplementary rations for large ruminants rations comes from household waste and cereal by-products, brans, mill-sweepings, brewers’ grains, and oil-seed cakes, at household level home-grown cereals are also fed directly to livestock as cut sheaves and as grain. Consequently, a further 70 000 tonnes per annum has been added by the Mission to animal feed use in the cereal balance sheet, to cover such eventualities this coming year when grains will be more freely available.

3.7 Cereal and pulse production forecast

The CFSAM teams’ visits coincided with all stages of the harvest from crop cutting to threshing depending on crop and location. The wide range of harvesting activities underway at the time of the Mission enabled a ready assessment of actual production per unit area to be observed by the teams. Where crops were still standing, samples were taken, threshed using local techniques, and weighed to crosscheck agricultural bureaux yield estimates and farmers’ predictions of production. Where harvesting was over, quantities of stored grains or cobs were matched against the areas from whence they came; and where threshing or combining of fields had been accomplished, information was obtained directly from the harvesting contractors, regarding the median yields in their areas of operation. In such ways, additional information was obtained to make adjustments to estimates and predictions received and to counter-balance glaring inconsistencies and false declarations or to supply figures for missing data. Transects driven by the Mission teams moving from location to location enabled observers to take detailed records of crop conditions, standardised using PET.⁴

Regional totals of area and production, prepared by the Mission, are presented in Table 5 by crop. They indicate a 2004 meher cereal harvest of 13 million tonnes from 9.23 million hectares. This is 3 percent higher than 2003’s post harvest estimate corrected for the switch to CSA data in Amhara and 10 percent higher than the CFSAM estimate. Pulses return at 1.26 million tonnes from 1.41 million hectares, a harvest that is 59 percent better than 2003’s corrected post harvest estimate and identical to the previous year’s CFSAM figure. Time series data for the past five years are provided in Table 6 for comparison purposes. They show that 2004’s production estimate for cereals and pulses is the highest that has been achieved to date. Its validity hinges on the accuracy of the area and yield estimates prepared by the BOA offices. The differing area estimates described earlier generate doubt in the minds of assessors, who may then opt for lower yield estimates to redress possible overestimated areas, therefore, the sooner the area differences can be practically identified and resolved at woreda or PA level the better. In 2004 the formalised approach to transect-based field recording and crop cutting by all Mission teams resulted in “spot- check” yield estimates that are significantly higher than the zonal averages, particularly in the drier zones, where many of the low average yield estimates of 2-3 quintals are thought to be highly unlikely and connect to the absence of training in crop assessment and the availability of suitable equipment viz. manuals, accurate balances and quadrats, at woreda level. This points to the need for a programme to address these issues as soon as possible, coupled with technical support from zonal and regional specialists to help BOA staff at local levels to resist local pressures to underestimate production to sustain the flow of food aid.

Table 5 - Ethiopia: Area ('000 ha), production ('000 tonnes) and yield (tonnes/ha) of cereals and pulses in 2004/05 Meher season