Summary

Ethiopia continues to suffer from food deficits and poverty despite a wealth of natural resources large tracts of arable land, diverse biological diversity, abundant water and a large pool of human resources. Nearly 85% of the human population lives in rural areas and is engaged in agriculture. Therefore, it is on agricultural front that efforts to alleviate poverty and achieve food security must be concentrated. It is in recognition of this fact that the Federal Democratic Republic of Ethiopia formulated Agricultural Development Led Industrialization (ADLI) during the early 1990s as its fundamental economic policy. One of the pillars of ADLI is Participatory Demonstration and Training Extension System (PADETES), which has been in practice since 1994/95. The major objectives of PADETES are to improve the income and the living standard of the rural population. This is possible by increasing agricultural productivity; becoming self-sufficient in food production; facilitating the establishment of farmers’ organizations; increasing and improving the production of agro-industrial and export crops so as to diversify export and substitute import; conserving and developing natural resources for sustainable utilization; and by encouraging the participation of rural women in development. The objective of this study was to evaluate whether PADETES has met its objectives particularly in terms of crop productivity and production as well as natural resources management by taking a large sample of households (4587) from all regions of Ethiopia, except Gambella.

Agricultural extension, as an educational and communication tool, makes a vital contribution to agricultural production and rural development. Similarly, the results of present study showed that there was increment in production and productivity whenever extension packages were used, confirming previous demonstration and adoption studies. Yet, there was a problem of sustainability and extending the scope of the extension package in Ethiopia. For instance, results showed that no more than 56% of the sampled households used any one modern inputs and improved crop production management. In certain cases, farmers were found to regress back to their own old practices and no sustainable use of modern agricultural inputs, such as fertilizers and improved seeds. It is recommended that agricultural extension services be more enhanced in terms of making more and more resource poor farmers participate in the extension package, delivering alternative packages, reducing the ratio of development agents to farmers, increasing input supply, increasing access to credit and improving input and output distribution and marketing. Moreover, the present study showed that the percentage of new entrants into the extension package was low, ranging from 4% in 1994/95 to 16% in 2002/03, suggesting low number of graduates as well. However, certain agricultural extension professionals feel that achieving 10% involvement of farmers in the extension system is sufficient, provided this is followed by effective inputs and outputs distribution and marketing. Input suppliers should become more responsive to the needs of farmers graduating from the extension package by making available the necessary quantity and quality of inputs such as improved seeds, fertilizers, pesticides, and farm implements, etc. at the right time and for affordable prices. Individual contacts between extension agents and farmers have to be expanded; implying the need for training and deploying increased number of extension agents. One of the chief impediments to the use of modern agricultural inputs is the poor distribution channels as well as poor marketing for surplus production. As a solution to this there is a strong need to enhance the process of upgrading of rural transport in association with appropriate market development, for both agricultural inputs and outputs.

Improved seed is one of the primary requirements for enhancing crop productivity and income generation. Seed security should also be first ensured for ensuring food security to farmers. However, the results of the present study, in agreement with several previous studies, revealed that only about 8% of the participating households used improved seeds and over one-half (54%) of the participating households used own saved seed, suggesting annual fresh seed purchase to be low; still 30% of the participating households were not replacing seeds within the acceptable period of 1-4 years. Further disturbing case is that only about 39% of the participating households used complete package (improved seed + fertilizer + improved cultural practices). Incomplete package is not as profitable as complete package. The levels of participation in high value crops and natural resources management were also found to be low. The amounts of inorganic fertilizers used for major crops were found to be about one quintal per hectare. Ensuring Ethiopian farmers access to quality improved seed timely and at a reasonable price can only be achieved if there is a vibrant seed supply system to multiply and distribute the seeds and if mechanisms of getting effective credit is established. It seems that formal seed system such as ESE and private seed companies could not significantly do the job in this regard. Thus, it is recommendable if farmers-based seed production and distribution mechanism is enhanced at least for open-pollinated crops; with cooperatives/unions playing a major role in distributing and marketing of improved seeds. Formal seed system is usually profit-oriented and likes to operate on principle of economy of scale and deals with a narrow range of crop species and varieties. However, given the agro-ecological and farming system diversity of the country no single variety will occupy a wide area of cultivation. Seed production, distribution and marketing for specific areas would seem to be possible only under farmers-based seed production and marketing system. Farmers-based seed production and dissemination should receive more priority than formal seed sector in view of the fact that no single variety will have wide adaptation for economically profitable seed production and distribution by the formal sector. The formal seed sector tends to focus on a few crops such as wheat and maize and on a few varieties for which there is a large market, giving less attention to other crops especially pulses, oilseeds and horticultural crops. Farmers-based seed production and distribution is also suitable for rural Ethiopia, where transportation and other infrastructures are poorly developed. As local demand for improved seeds of many varieties of crops takes off, farmers will gain new economic opportunities in small-scale seed production. Experiences in Ethiopia and other African countries have shown that, with adequate support, smallholder farmers are capable of producing high quality seed of improved varieties. Extension service is expected to develop training materials and offer training on principles and procedures of operating small-scale seed production. Such practice would generate income for the farmers and greatly promote the adoption of improved varieties on a wide scale. Wide adoption of improved varieties in turn ensures food security and increases productivity. The role of private seed producers should not be undermined, however. It is only because of the fact that Ethiopia’s dispersed and underdeveloped markets, poor infrastructure, and low farmer income present obvious challenges to the expansion of private seed markets profitably.

Cooperatives and unions would also play a major role in seed, fertilizer and pesticides distribution and marketing in the sense that they would have more number of marketing outlets in remote rural areas than the already known companies in this regard. This presumption of more number of distribution and marketing outlets for cooperatives would solve the lack of demand-supply determination in the use of agricultural inputs and the carry-over that the companies are currently facing. Carry-over, for both seeds and fertilizers, is because of poor distribution channels and inadequate demonstration and promotion of newly developed improved varieties, rather than due to lack of demand in actual sense.

Many agricultural research centers have been established to assist in development efforts by way of developing and disseminating research-based technologies. Yet, many agriculturally potential areas and moisture stress areas did not receive research services. Since agro-ecologies and farming systems are highly varying in Ethiopia, the type of crops grown, the technological input requirements and the major objectives also differ within Woredas, zones, regions and the country as a whole. Hence, the agricultural research system is required to extend its research activities into uncovered areas. All agricultural research centers should also have formal linkage with bureaus of agricultural and rural development of their mandate Woredas and farmers training centers.

Along with the supplies of improved seeds, fertilizers and other agricultural inputs, rural infrastructure conditions in Ethiopia are the major and most common constraints to development of agriculture, limiting the input-output marketing. Surplus production cannot easily be transported to grain deficient parts of the country. In most cases, it is difficult to obtain market information. This entails development of effective rural transport and road networks. Such development would let farmers in remote parts of the country to have access to agricultural inputs and easily market their surplus outputs at reasonable price.

In the present study, the participating households in the extension package have rated diseases and insect pests as major production constraints. As a solution to this, diversification of production system is imperative. The diversified production system should involve crop rotation, frequent replacement of varieties of the same crop and effective use of integrated disease, insect pests and weed management. Crop rotation, involving cereals, pulses, oilseeds and horticultural crops, must be practiced to reduce disease, insect and weed incidences as well as to improve soil fertility and to enhance productivity. The present study revealed that high fertilizer price is one of the constraints in using fertilizers and the sampled households were found to apply less than required without also taking into account annual nutrient removal. Practices such as crop rotation, application of green manure and organic fertilizers may compensate for nutrients removed by the previous crop. It is to be noted that organic wastes, manure and biological nitrogen fixation are among the principal sources of nitrogen, which is the most deficient in almost all soils but highly essential for plant growth and development. Conservation-oriented cropping systems such as agro-forestry should be promoted for not only land conservation but also for fodder for livestock and for fuel supply.

The use of BBM should substantially be increased to improve the drainage problem of vertisols. Vertisols are the most important soils for agriculture next to nitosols and cambisols in Ethiopia. However, the agricultural potential of vertisols cannot be fully utilized unless the drainage problem is improved by using such techniques of drainage using BBM. The widespread use of BBM means improved drainage of black soils, which in turn would enhance the practice of double or multiple cropping patterns taking the advantage of the long growing season available in highland black soils. The use of BBM in the lowland black soils, if coupled with irrigation, would contribute greatly to the increase of the area under irrigation in the country. Similarly, the use of Tier-ridger should substantially increase in moisture stress areas as means of soil and water conservation practice.

Results of the present study have the following implications for the research and extension systems of the country.

· Increase the level of awareness of farmers about existing technologies through demonstration, adaptive trials, field days, media, extension manuals and so forth. Further more, farmers research groups, farmers extension groups, farmers training centers, and research-extension farmers advisory council should play a decisive role in ensuring that research and extension activities of the country are based on real and immediate problems and solving local problems. The level and role of farmers’ participation in planning and implementation of research and extension programs should be enhanced so that they develop confidence and eventually adopt the emerging technology.

· Increase the amount of improved seeds produced for all categories of crops (all cereals, pulses, oilseeds, horticultural crops and industrial) rather than concentrating on few varieties of a limited number of crops, such as only maize and wheat.

· Develop location specific (based on soil type, moisture regime and inherent soil fertility status) fertilizer rates and types and promote wide scale adoption.

· Enhance sustainable cropping system (crop rotation, inter-cropping, double cropping, rely cropping) instead of mono-cropping to improve soil fertility and reduce the building up of diseases, insect pests and weeds.

· Improve linkages among research, extension, farmers and agro-processing industries. Here the role of each must be defined and formalized, at least for such crops like bread and durum wheats, malt barley, canning beans, and soybean.

· Reduce post harvest losses by improving storage structures and processing to add value.

· More number of technologies be developed and disseminated for farmers in the moisture prone areas (arid and semi-arid). This involves opening more number of research centers and testing sites in such areas. Technology development in such areas should focus on integrated technology development and transfer instead of going for piecemeal.

· Give more research and extension focus on high value crops such as pulses, oilseeds and horticultural crops, as the present focus is more on cereals.

1. Introduction

The Ethiopian economy is dominated by agriculture that accounts for 40 to 50% of the total GDP, employs about 85% of the population and provides close to 90% of the export earnings in which coffee alone accounts for 60% of the total value of agricultural output (RATES, 2003; Library of Congress, 2005). Agriculture also provides 70% of raw materials required for agro-based industries. Within agriculture, crop production, livestock and forestry as well as related activities contributes about 60-64%, 23-30% and 10-13% of agricultural GDP ), respectively, (SIDA, 2003; FAO, 2004. Crop production is essentially rainfed; with subsistence smallholders, accounting for about 95-97% of the agricultural output (FDRE, 2002) and hence agricultural GDP significantly varies with rainfall patterns (amount and distribution), registering a trend rate of growth of about 1.8% per annum during the period 1981/82 to 2002/03 (World Bank, 2004). In contrast, human population is growing at about 2.9% per year. Indeed, food insecurity in Ethiopia is largely the result of dependence on low-input, low-output rainfed agriculture and limited diversified livelihoods.

The agricultural system is, however, characterized by low technology, low productivity and natural disasters such as drought and irregular rainfall pattern (including timing, amount and/or distribution). As a result, the majority of the rural population is vulnerable to food insecurity and famine. Indeed, poverty is pervasive, deep and persistent. On the Human Development Index (HDI) of UNDP, Ethiopia ranks nearly at the bottom, 171^st of 174 countries. According to an FAO (2001) estimate, about 45-51% of the population is undernourished or suffered from malnutrition of vitamins such as vitamin A, minerals such as zinc, iron and iodine as well as from protein-energy malnutrition and over two million people are considered to be chronically food insecure (FAO/WFP, 2001). Domestic production is estimated at only to supply about 70% of the food requirement and each year about 4 to 6 million people need food assistance, despite the existence of potential resources (human, land, water, and biological and agro-ecological diversity) for food self-sufficiency and even surplus production in terms of land. To mention just the under exploitation of one of the resources, the potential gross irrigable area is estimated to be 3.7 million ha, of which only 3% is currently utilized (FDRE, 2002).

About 74 million hectares (66%) of the total land areas of 111.6 million hectares are arable. However, only 15-22% of these is under cultivation with annual and perennial crops and about 96% of the cultivated land is believed to be under smallholder farming (CRDA, 2000). Cereals are the most important food crops occupying about 76.6% of the total cropped area. Other crops grown include pulses, oil crops, coffee, fruit trees, and cotton.

Despite the importance of agriculture in the Ethiopian economy and despite Ethiopia is known to be one of the dozens of countries where several crops have been domesticated and cultivated from time immemorial, initiation of government organization to lead agricultural development came late with the establishment of the Ministry of Agriculture in 1908. Not only the establishment of the Ministry was late but also there was not organized agricultural education, research and extension up until the establishment of agricultural colleges at Ambo in 1947, Jimma in 1952 and at Alemaya in 1953. The extension service was only further consolidated and extended its scope in 1963 under the Ministry of Agriculture, while the Institute of Agricultural Research was established to lead nationwide agricultural research system in terms of technology generation and dissemination in 1966. All lines of evidence points to the fact that Ethiopian agriculture was virtually unaffected by any modernization as recently as 1950s (Bhagavan, 1989).

Another major pitfall in the development of Ethiopian agriculture was the less attention given to smallholder agriculture until the 1990s. During the Imperial regime emphasis was given either to industrial development or to the development of big commercial farms. Similarly, during the Derg regime more emphasis was given to big state farms and cooperative farms, which consumed about 95% of agricultural inputs (improved seeds, fertilizers, pesticides and farm implements) but contributed only 5% to the total production. On the other hand, the smallholder agriculture accounted for about 95% of the total area under crops and for more than 90% of the total agricultural output. It also produced 94% of all food crops and 98% of the coffee, whereas the remaining was generated from state and commercial farms (CRDA, 2000).

Over years agricultural production has been affected by fast increasing population, leading to reduced arable land surface per capita, as well as by environmental degradation (manifested by soil removal by sheet and gully erosion, nutrient depletion due to biomass burning, acidity, salinity, water-logging, deforestation, overgrazing, short fallow period, etc.), increasing top soil erosion, and desertification, impacting negatively on overall rainfall cycles resulting in decreasing soil fertility. Deforestation is caused by the cutting of trees to clear land for cultivation, fuel, and construction.

Poor infrastructure as well as market and marketing development feature prominently as major constraints impending agricultural development in Ethiopia. These factors have prompted the need to mobilize from a sector characterized by predominant subsistence farming (low input-low output agriculture) towards promoting the use of inputs (improved seeds, fertilizers, pesticides, farm implements and irrigation) as well as to intensification and diversification of farming systems so as to ensure food security and commercialization of Ethiopian agriculture (USAID, 2004).

Recognizing the key role of agriculture in the economic welfare of the country and its associated shortcomings, the Ethiopian government has been engaged in a highly ambitious effort of rural and agricultural development. To this effect, the much-proclaimed ADLI strategy was put in place to guide the economic development of the country since the early 1990s. Rural development led by agricultural growth is also one of the pillars of the Ethiopia’s SDPRP. With regard to agriculture, the SDPRP seeks gains in factor productivity, output diversification, and greater market orientation to help farmers move beyond subsistence farming to small-scale market-oriented agriculture. In other words, the ADLI strategy sets out agriculture as a primary stimulus to generate increased output, employment and income for the people, and as the springboard for the development of the other sectors of the economy. Industrial development is expected to absorb rural surplus labor, and expand domestic markets for agricultural commodities while the increase in agricultural production is expected to raise the country’s export earnings to meet domestic demands for food and raw materials and to expand markets for industrial commodities.

Unlike previous development policy and strategies, ADLI gives strong emphasis to raising the productivity of smallholder farmer to make agriculture a strong foundation for the final industrial development. The rural and agricultural sector is viewed as the focal point of development for several reasons (Addis Tribune, 1995): First, since 85% of the country’s population is located in the rural areas and engaged in agriculture, mobilizing the country’s resources for fast development requires working with the rural population to improve agricultural productivity; second, a broad-based development strategy which shares the benefits of development among is necessary to maintain peace and ongoing support for the development process; third, a focus on increasing rural and agricultural productivity is the key to finding a lasting solution to Ethiopia’s chronic famine problem. Other parts of the economy will also benefit from faster growth in the agricultural sector in that the urban population will have access to cheaper food; with the growth of agricultural production, the availability of raw materials for industry will also increase, stimulating off-farm employment; rising rural incomes will accelerate the demand for consumer items and agricultural implements, strengthening these industries. In other words, the ADLI strategy views agriculture as the engine of growth because of its potentially superior growth linkages, surplus generation, market creation, and provision of raw materials and foreign exchange (Haile et al., 1991; Demeke, 1996; MOFED, 2002). ADLI has set the goal of doubling per capita income over 15 years and to resolve the food security problem within five years. The policy instrument chosen to attain these goals included a new system of agricultural extension, Participatory Agricultural Demonstration Training Extension System (PADETES) in 1994/95. The major objectives of PADETES are to improve the incomes and standards of living for the rural population by increasing agricultural productivity; become self-sufficient in food production; facilitate the establishment of farmers’ organizations; increase and improve the production of industrial and export crops so as to diversify export and substitute import; conserve and develop natural resources; and to encourage the participation of rural women in development.

The PADETES attempts to merge the training and visit (T&V) system with the technology diffusion experience of the Sasakawa/Global 2000 (SG2000). The principle of PADETES is to demonstrate to farmers the benefits of a package of inputs, notably balanced and higher rates of fertilizer, improved seeds, pesticides and better cultural practices. Evidently, SG2000 has convincingly demonstrated productivity increment when farmers were provided with appropriate research messages, adequate extension assistance, and agricultural inputs such as improved seeds, fertilizers and pesticides, delivered on time at reasonable prices (SG2000, 1996). Results from SG2000 indicated that if technologies like chemical fertilizers, improved high yielding varieties, optimum plant population, timely weeding, pest control and harvesting are used, the yield can be easily doubled or even tripled as compared with traditional practices (Dercon, 1999; SG2000, 2000). Farmers who participated in the PADETES allocate ½ ha of land (in some cases ¼ or less) for demonstration and make a 25-50% down payment on the inputs used at the time of planting with the balance due after harvest. The farmers manage the plots under a close supervision and advice of extension agents. Both the SG200 and government led-PADETES have demonstrated that the yield levels of the crops (maize, wheat, tef, sorghum, etc) included in the package program to be 3-4 times higher, on average, than the conventional national averages, implying that the country has a potential to solve the chronic food shortage and insure food security.

The purpose of this paper is to assess whether PADETES has met its intended objectives - increasing crop production and productivity and thereby contributed to food security and poverty reduction.

2. Methods

A total of 4587 households (drawn from 9 regional states, 44 zones and 94 Woredas) were sampled in 2003/04 cropping season for the purpose of the study (Table 2.1 and Annex 1). Majority of the Woredas (88%) were drawn from Oromia, Amhara, SNNPR and Tigray in that order of importance, keeping with the size of households in respective regions.

Table 2.1. Sample size (number of households, zones and Woredas by region)

Region	Number of households	Number of Zones	Number of Woreda
Tigray	350	4	7
Afar	200	2	4
Amhara	994	8	20
Oromia	1891	13	39
Somali	100	1	2
Benshangul Gumuz	100	2	2
SNNPR	852	11	17
Dire Dawa	50	1	1
Harari	50	2	2
Total	4587	44	94

The 4587 households were sampled for 18 potential commodities in their localities (Table 2.2). In this respect, the list of potential commodities provided by the Market Development Strategy (‘Market Plan’) document of the Ministry of Agriculture (unpublished data) was used to identify the sample framework. Male-and female-headed households constituted 88% and 12% of the sample, respectively. More than three-quarter of the sampled households (83%) are found to be followers of Orthodox Christian and Muslim, while the remaining 17% was accounted for by other types of Christianity. The sampling also took into account agro-ecology, moisture regime (reliable and stress areas), length of growing period, market opportunity, and availability of other enabling institutions.

Table 2.2. Sampled households by major types of commodities and region

Commodity	Tigray	Afar	Amhara	Oromia	Somali	Benshangul Gumuz	SNNPR	Dire Dawa	Harari	Total
Agro-pastoral	0	100	0	0	0	0	0	0	0	100
Barley	0	0	51	153	0	0	48	0	0	252
Chickpea	50	0	51	50	0	0	0	0	0	151
Coffee	0	0	0	249	0	0	153	0	0	402
Dairy	100	0	100	243	0	0	151	0	0	594
Enset	0	0	0	50	0	0	97	0	0	147
Groundnut	0	0	0	51	0	0	0	0	50	101
Haricot- bean	0	0	49	96	0	0	53	0	0	198
Honeybee	0	0	99	50	0	0	49	0	0	198
Horse bean	0	0	51	54	0	0	0	0	0	105
Maize	0	0	200	140	0	0	151	0	0	491
Meat	0	100	195	150	50	0	0	50	0	545
Potato	0	0	0	50	0	0	50	0	0	100
Sesame	50	0	0	0	0	50	0	0	0	100
Sorghum	50	0	50	99	50	50	0	0	0	299
Tchat	0	0	0	49	0	0	0	0	1	50
Teff	50	0	100	251	0	0	0	0	0	401
Wheat	50	0	49	154	0	0	100	0	0	353
Total	350	200	995	1889	100	100	852	50	51	4587

3. Crop Production Extension Package

3.1 Participation in the crop extension package

Of the total sample of 4587 households sampled for this study, 4575 (99.7%) responded to the questionnaire whether they have participated or not in the extension package. Close to 56% of the respondent households were found to have ever participated in extension package programs and the remaining 44% did not participate (Table 3.1). Region-wise, only Tigray, Amhara, Oromia and Beneshangul Gumuz have managed to let more than 50% of their respective sampled households to have participated in the extension during the period 1994/95-2003/04. The percentage of participant households was about 62% in moisture reliable Woredas as compared to 46% in moisture stress Woredas. Apparently, there was an increment in the number of households involved in the package program nationwide when compared to involvement of 37% of the sampled households in 1998 (Ayalew, 1998). However, percentage of participating households was low in agro-pastoral Woredas, accounting only for 29%, suggesting that relatively less households have adopted the extension package in arid and semi-arid areas, where agro-pastorals are widely residing.

Table 3.1. Percentage of households ever participated in the extension package programs during the period 1994/95-2003/04

Region	Sample size	Yes	No	Total
Tigray	350	55.4	44.6	100.0
Afar	199	21.1	78.9	100.0
Amhara	992	63.8	36.2	100.0
Oromia	1885	63.6	36.4	100.0
Somali	99	15.2	84.8	100.0
Beneshangul Gumuz	100	60.0	40.0	100.0
SNNPRS	849	43.7	56.3	100.0
Dire Dawa	50	24.0	76.0	100.0
Harare	51	45.1	54.9	100.0
Total	4575	55.7	44.3	100.0

Despite the importance given to PADETES, the percentage of households annually entering into the extension package was found to be low, ranging from 4.3% in 1994/95 to 16.3% in 2002/03 and there was no clear trend during the period 1994/95-2003/04 both for the country as a whole and region-wise (Fig. 3.1a, b and Annex 2).

Fig. 3.1a. Nationwide annual percentage of households who have started participating in the extension package for the first time during 1994/95-2003/04.

Fig. 3.1b Region-wise annual percentage of households who have started participating in the extension package for the first time during 1994/95-2003/04.

The same trend was observed while analyzing the data for moisture reliable and stressed Woredas. It is to be noted that the rural based agricultural development strategy document issued in 2002 targeted that by the year 2005, all farming households, numbering close to 9 million households, would be accommodated in the program. The present study implied that it may be too ambitious to achieve such a target. Probably those who once started participating will not graduate and leave room for new entries presumably because continuing to remain in the extension system would ensure access to inputs and credit and the DAs also maintain such experienced farmers so as to reduce the load of training fresh and inexperienced ones. Another likely reason is that the extension package is always accommodating only resource-endowed farmers, leaving aside the less resourceful farmers. Meeting the participation criteria such as having a plot size of 1/4^th of a hectare and 25% down payment to be entitled for credit for inputs may also discourage households to participate in the extension package. An economic analysis of the Ministry of Agriculture/SG2000 experiments with improved cereal technology in Ethiopia showed that extension package participants are richer than the average farmers, with larger land holdings and better access to resources such as animal traction and labor (Howard, et al., 1998). To redress this limitation, future extension services should equally focus on both resource-rich and resource-poor farmers to avoid creation of rural wealth gap like the previous extension systems (Nichola, 1985; Bhagavan, 1989). Moreover, the resource poor farmers are more succumbed to risks than the resource rich farmers are, unless special support is given to them so that they cop with circumstances. This should in turn be coupled with increasing the number of DAs and conducting adaptation trials of the intended technologies to verify their suitability to particular localities, especially in more marginal areas.

On the other, some extension professionals argue that achieving about 10% involvement in the extension system is adequate on the ground that extension is expected to only communicate idea with the farming community and inputs demand emerging as a result of the extension program need to be addressed by designing effective ways of inputs delivery to farmers. This involves concerted efforts of higher level policy makers, effective and efficient input supply organizations, output marketing agencies and generally the full involvement of all stakeholders engaged in agriculture and rural development.

It is generally acknowledged that agricultural research efforts and results will be of no value unless the approved technologies are transferred and widely adopted by farmers. Results of the present study showed that only two-third (67%) of the sampled households participated in the different package programs. Of the participating households, 58% of the households participated in cereals (maize, wheat, tef, sorghum, barley and finger millet) production package, while those involved in high value crops such as pulses, coffee, vegetables, oil seeds and fruit crops production constituted low (Fig. 3.2 and Annex 3). Unfortunately involvements in natural resources management (soil and water conservation, agro-forestry,

Fig. 3.2. Extension packages in which the sampled households have been participated.

^a barley, maize, sorghum, tef, wheat, finger millet; ^b chickpea, faba bean, field pea, haricot bean, lentil and soybean ; ^C pepper, tomato, head cabbage, carrot, onion, shallot, garlic; ^d potato, sweet potato, enset; ^e Noug, linseed, gomenzer, groundnut, sesame; ^f banana, mango, avocado.

community forestry), use of farm implements to improve soil drainage such as Broad Bed Maker (BBM) and to improve soil and water conservation such as Tie-ridger, and post harvest were also very low. BBM is used to drain vertisols. If appropriate surface/subsurface drainage measures were implemented on vertisols, such as using BBM, potential food grain production would be increased substantially. Similarly, Tied-ridging is a technique of making embankments at regular spacing between ridges so as to prevent the run-off water and to create basins of water into micro-catchments, making more water available for crop production. Past research results and demonstrations have shown that effective management of soil water using tie-ridger to have appreciably improved land productivity in moisture deficient areas. For instance, experiments conducted in the semi-arid areas of Ethiopia (Melkassa, Meisso, Kobo and Mekele) using tie-ridges resulted in increased yield of maize, sorghum and low land pulses mainly because of the storage of moisture in the soil. Previous studies have also shown that farmers who were growing cereals have gained the most from the extension efforts and also the cultivation of vegetables seems to become an increasingly profitable activity most likely due to improved marketing of vegetables on the local market (Broeck and Dercon, 2001). Focusing largely on major crops such as cereals has also been previously criticized on the ground that most of them are not suitable for dry land areas (Abesha, et al., 2000). For such areas, cultivation of drought tolerant pulses having multiple uses such as cowpea, and planting of multipurpose (food, fodder, and fuel) tree species should be encouraged.

More than two-third (69%) of the participating households indicated that involvement in package program has been participatory and based on their ability and willingness (Table 3.2). Non-participatory means of including households in package program such as the one based on top-down, quota and administrative decisions accounted for 27% of the households. This finding is contrary to the conclusions reached by previous studies that the extension work in Ethiopia was non-participatory and coercive, giving little consideration to farmers’ experiences and knowledge (Abesha et al., 2000; Kassa, 2003, 2004). Participatory extension should be based on the farmers’ knowledge and farmers should be involved in the planning and implementation of extension programs so that the extension becomes more responsive to the needs of the different farmers backgrounds in the communities.

Table 3.2. Opinion of sampled households about extension package programs (%)

Region	Sample	Participatory	Non participatory	No suggestion	Total
Tigray	235	66.4	28.5	5.1	100.0
Afar	108	74.1	20.3	5.6	100.0
Amhara	528	74.5	22.3	3.2	100.0
Oromia	1014	67.3	28.3	4.4	100.0
Somali	45	53.3	28.9	17.8	100.0
Benshangul Gumuz	62	87.1	9.7	3.2	100.0
SNNPR	391	64.2	32.6	3.2	100.0
Dire Dawa	25	84.0	16.0	0	100.0
Harari	26	69.2	26.9	3.9	100.0
Total	2434	68.7	27.2	4.1	100.0

3.2 Extension information about package programs

In the present study, a greater proportion of the participating households (80%) were attracted to extension package upon awareness creation by extension agents/Bureau of Agriculture and woreda or peasant association officials. Observation-based (good result of extension package plots and increased production achieved by neighboring farmers) attracted only about 14% of the households. Need to get access to inputs (seeds and fertilizers) and to obtain loan, pressure from woreda and peasant association officials as well as listening to extension program on radio attracted only insignificant percentage of households (6%). A study aimed at studying the adoption of improved bread wheat varieties and inorganic fertilizer by small scale farmers in Yelmana Densa and Fata districts of northwestern Ethiopia also revealed that the major actors in the dissemination of information about wheat technology are extension agents and neighbors, while relatives, researchers, traders, producers and service cooperatives played limited roles (Zegeye, et al., 2001b).

The present study further revealed that the same above agents also served as major sources of information about extension packages in that more than three-quarter of the participating households (81%) got informed about extension package from awareness creation made by either extension or development agents and about 15% from their neighbors (Table 3.3). Field days, demonstrations, media and SG2000 played insignificant (4%) role in informing farmers about extension. Results from wheat and barley seed systems in Ethiopia and Syria (Bishaw, 2004) also revealed that although farmers use multiple sources of information to acquire knowledge on agricultural inputs such as improved seeds and fertilizers, the extension agents were the main source of information for new technologies generated through the package program. Similarly, assessment of the diffusion and adoption of agricultural technologies in Chilalo by Waktola (1980) indicated that farmers preferred direct contact by extension and other field workers to indirect contact through model farmer approach and the former approach was found to be more effective in ensuring adoption and diffusion of technologies.

Table 3.3. Sources of firsthand information about extension package for the sampled households (%)

Region	Sample	Extension agent	Neighboring farmers	Field days	Demonstration plots	Mass media	SG2000	Own experience	Total
Tigray	194	88.1	9.3	0.0	0.6	1.5	0.5	0.0	100.0
Afar	41	82.9	2.4	0.0	0.0	0.0	9.8	4.9	100.0
Amhara	635	84.4	12.9	0.0	1.6	0.3	0.0	0.8	100.0
Oromia	1192	77.1	18.0	0.3	1.3	2.6	0.2	0.5	100.0
Somali	14	71.4	28.6	0.0	0.0	0.0	0.0	0.0	100.0
B. Gumuz	56	91.1	5.4	0.0	1.8	1.7	0.0	0.0	100.0
SNNPR	369	85.4	10.6	0.0	2.4	0.5	0.0	1.1	100.0
Dire Dawa	12	83.3	16.7	0.0	0.0	0.0	0.0	0.0	100.0
Harari	23	34.8	52.2	8.7	0.0	4.3	0.0	0.0	100.0
Total	2536	81.0	14.7	0.2	1.5	1.6	0.3	0.7	100.0

3.3 Types of technologies disseminated and used

Generation of improved technology and its adoption by farmers are the key factors for increasing crop productivity and, therefore, improving the standard of living of farming communities. Several factors determine, directly or indirectly, the decisions of farmers to accept or otherwise the newly developing agricultural technologies/inputs. In this section, attempts are made to examine the inputs utilization patterns of the sampled households and assess factors associated with such adoption or non-adoption of the inputs being delivered.

3.3.1 Improved seeds

Despite the increasing supply of improved seeds in PADETES, 75% of the participating households were found to use local seeds (Table 3.4). Only about 8% responded to use improved seeds and another 15% used both local and improved seed and the use of hybrid maize is only 0.5% nationwide. Analyzing the data for potential crops in potential Woredas generally gave results of the same trend. However, the situation seemed to be better for maize in that in the maize dominated sampled Woredas, 49% of the households indicated that they use local seed most of the time, 13% used improved seed, 34% used both improved and local seed and 3% used hybrid seed. Although tef is grown on 32% of the cultivated land, seed production of this crop is not attractive to large commercial seed producers for several reasons, such as admixture. As a result, small-scale farmers are now being encouraged by ESE to grow seed of this crop on contract (Kugbei and Fikru, 1997).

Table 3.4. Kinds of seeds mostly used by the sampled households (%)

Region	Sample	Improved seed	Local seed	Both local & improved	Seed of unknown source	Hybrid maize	Total
Tigray	340	7.9	84.4	6.8	0.9	0.0	100.0
Afar	136	20.6	68.4	5.1	5.9	0.0	100.0
Amhara	978	9.0	73.9	16.7	0.1	0.3	100.0
Oromia	1849	9.0	70.5	20.2	0.2	0.1	100.0
Somali	97	3.1	91.8	1.0	0.0	4.1	100.0
B. Gumuz	98	9.2	84.7	6.1	0.0	0.0	100.0
SNNRPS	829	4.6	80.1	13.0	0.4	1.9	100.0
Dire Dawa	48	6.2	89.6	4.2	0.0	0.0	100.0
Harari	50	16.0	62.0	22.0	0.0	0.0	100.0
Total	4425	8.4	75.0	15.6	0.4	0.6	100.0

Previous studies also estimated the use of improved seeds at about 7% of the seed requirement (Amha, 1999), which is definitely low and the proportion of households in 1995/96 who bought improved seeds was 2.3% (MOLSA, 1997). The same studies pointed out that the use of improved seed is influenced by price, access to credit, fertilizer use, economic status of the household, chemical use, visits of extension agents and infrastructure development. Generally farmers use different sources of seed for planting such as own saved seed from previous year’s harvest, seed obtained from relatives, neighbors or other farmers as well as seed purchased through local markets or grain traders and seed purchased from the formal sector. This overall low rate of use of improved seeds corroborates previous estimate that 85% of the Ethiopian farmers are believed to depend upon seeds of local cultivars (Tafesse, 2002). The reasons for largely depending on own saved seed could be that the farmers do not have to pay cash for it nor travel to procure it; that the farmer may not have well organized supply system for the particular seed; and that the farmer knows that his cultivar is superior or does not have enough knowledge about the merits of improved seeds. Another reason could be that the formal sector more often provides seeds of a limited range of crops and varieties, which not always fulfill the needs of the farmers to diversify production.

The major source of seed was own saved seed (54%) (Table 3.5). The formal seed sources (ESE, agricultural research centers and colleges and Pioneer Hybrid Seed supplied, directly by respective institutions or by bureaus of agriculture) provided for about 19% of the households’ seed requirement. For about 25% of the households a purchased seed (from seed producing farmers, neighbors, relatives, market, etc.) was source of seed for planting. In agreement with these results, several previous studies have shown that farmers produce crops using farm-saved seed with limited adoption of improved varieties (Kahsay and Kugberi, 1998; Gebeyehu et al., 2001).

Table 3.5. Major sources of seeds for the sampled households (%)

Region	Sample	Own saved seed	Formal seed sources	Purchased seeds	Relief seeds	Community produced seeds	Total
Tigray	350	56.0	20.0	22.3	0.3	1.4	100.0
Afar	200	48.5	11.5	36.0	0.0	4.0	100.0
Amhara	995	59.2	17.5	21.4	0.4	1.5	100.0
Oromia	1889	52.1	20.1	25.7	0.1	2.0	100.0
Somali	100	53.0	20.0	25.0	0.0	2.0	100.0
B. Gumuz	100	43.0	25.0	32.0	0.0	0.0	100.0
SNNRPS	852	52.2	20.3	25.0	0.2	2.3	100.0
Dire Dawa	50	66.0	20.0	14.0	0.0	0.0	100.0
Harari	51	66.7	11.8	21.5	0.0	0.0	100.0
Total	4587	53.9	19.2	24.8	0.2	1.9	100.0

In the present study, for the improved seed the major source was found to be ESE, supplied through Bureau of Agriculture (17%) and direct supply by ESE accounted only for 1%, indicating that direct sales of commercial seeds to the smallholders by ESE is very minimal. Regional bureaus of agriculture are known to supply 95% of seeds produced by ESE and other formal sectors. A study made in 1998 estimated that the ESE provided 3.5% of national seed requirement, while the bulk of seed used nationwide was farm-saved from harvests of previous crops (Kahsay and Kugbei, 1998). Improved seed marketing and adoption study in Ethiopia in 1999 revealed that the capacity of the ESE to reach all areas at a reasonable price is very limited (Amha, 1999). It seems that seed sales performance by farmers based seed production and marketing schemes is better than that of ESE and Pioneer Hybrid Seed Corporation. This may be due to the flexibility these small seed distributors have in adapting prices to reduced demand to secure sale of their available stocks and they are also closely located with other farmers (Raymakers, 2002). In general, seed distribution and marketing is weak in the Ethiopian seed system because of inadequate means of determining demand-supply, inefficient distribution systems, and low price of grains during bumper harvests reducing farmers’ interest to purchase improved seeds the following year. The present study also revealed that other formal sources such as agricultural research centers and colleges as well as Pioneer Hybrid supplied very insignificant amount of improved seeds (1.2%). Despite the formulation of a national seed policy in 1992 to encourage private seed production and distribution, the contribution of the private sector to Ethiopia’s seed system is imperceptible. Empirical evidence has shown that small-scale seed enterprises by contracting farmers and farmer-based seed production and marketing scheme are more effective in supplying improved seeds of desirable quality than by public and private seed companies (NSIA, 1997). Such farmers are also found to be effective in producing seeds of crops, which are not commercially attractive to large seed companies. About 25% of the households were found to use seeds of unreliable source, such as purchased seeds from markets and neighbors or relatives as well as donations as emergency relief. Seeds from such sources were not only unreliable but also may completely fail to germinate and be poorly adapted to local conditions. Such source of seed would also be unable to restore the adapted crop varieties and often rather known to introduce new diseases, insect pests and noxious weeds.

To increase productivity of crops, farmers should have access to improved seeds of the right type, at the right time, at the right place, at a reasonable price and with right-sized seed packages. To this effect, the formal seed sector in Ethiopia started with the beginning of agricultural research in early 1950s with the establishment of higher learning agricultural colleges. Seeds developed by the research system were initially introduced to farmers by research centers and production of improved commercial seeds of major crops was institutionalized in 1979 with the establishment of the Ethiopian Seed Corporation, renamed Ethiopian Seed Enterprise in 1994 (Tafesse, 2002). It was only in 1990 that a private seed company started operation in Ethiopia (Gebeyehu, et al., 2001). However, the annual sales of these two formal seed companies are not only low but also limited only to a few crop varieties, primarily maize and wheat (Amha, 1999). Creating conducive policy environment in 1992 through the announcement of a National Seed Industry Policy to encourage the participation of private seed companies has not yet to bring about significant improvement in improved seed supply to farmers. Probably the major obstacle to the formal seed companies in supplying seeds to farmers is not the small quantity of seeds they produce annually but poor distribution outlets to reach out the inaccessible smallholders located in different parts of the country.

Seed marketing is the weakest link in the seed production/marketing chain in Ethiopia. As a solution to this problem, farmers’ based seed production and marketing scheme was initiated in 1996 with a financial assistance of IDA, IFAD and the government of Ethiopia. Farmers at selected locations were trained in improved seed production, cleaning and seed exchange system. Empirical evidence shows that this scheme was more effective in distributing improved seeds to farmers (Tafesse, 2002).

The frequency of replacing seeds of improved crop varieties is one of the important indicators of modern farming. The present study revealed that nearly 70% of the participating households replace their seeds within 1-4 years (Table 3.6). The remaining 30% households replaced their seeds after more than 4 years. A study made in 1998 by Afri-consultants indicated that about 57% of the households renewed their seed stocks within 1-4 years and 3% renewed between 4 to 5 years, and the rest did not have specific renewal patterns (Afri-Consultants, 1998). The general recommendation for seed replacement duration by seed technologists is to renew within 4 years for self-pollinated crops and every year for hybrid varieties.

Table 3.6. Frequency of seed replacement of released (both hybrids and open-pollinated) varieties for the sampled households (%)

Region	Sample	1-4 years	4-6 years	6-10 years	>10 years	Total
Tigray	162	73.5	8.6	0.6	17.3	100.0
Afar	95	18.9	3.2	0.0	77.9	100.0
Amhara	506	83.0	3.9	1.8	11.3	100.0
Oromia	1168	71.6	9.9	1.3	17.2	100.0
Somali	17	47.1	23.5	0.0	29.4	100.0
Benshangul Gumuz	48	75.0	10.4	4.2	10.4	100.0
SNNRPS	398	58.0	5.8	2.3	33.9	100.0
Dire Dawa	26	80.8	3.8	0.0	15.4	100.0
Harari	29	58.6	3.4	3.5	34.5	100.0
Total	2449	69.7	7.6	1.5	21.2	100.0

Frequent seed replacement should also involve frequent replacement of old variety by a newly released variety to take advantage of the better yielding and quality as well as better disease and insect pest tolerance of such varieties. In Ethiopia, it takes several years for newly released varieties to enter into production for several reasons such as inadequate seed production and limited demonstration and popularization.

Results of the present study showed that only 22% of the participating households get improved seeds timely for purchase, while more than one-half (58%) responded to get late and another 20% had no idea about the time of supply (data not shown).

Among those households who have used improved variety, 26% responded to have used the right type of crop variety, while about 17% did not get the right type of crop variety (Table 3.7).

Table 3.7. How often households receive improved seeds of the right variety (%)

Region	Sample	Mostly	Seldom	Not at all	No suggestion	Total
Tigray	169	42.6	18.9	4.8	33.7	100.0
Afar	120	15.0	6.7	4.1	74.2	100.0
Amhara	514	28.4	24.5	22.0	25.1	100.0
Oromia	1134	27.2	34.5	14.5	23.8	100.0
Somali	23	17.4	8.7	21.7	52.2	100.0
Benshangul Gumuz	49	14.3	44.9	26.5	14.3	100.0
SNNPR	374	17.4	28.3	24.3	29.9	99.9
Dire Dawa	25	44.0	48.0	8.0	0.0	100.0
Harari	31	35.5	22.6	12.9	29.0	100.0
Total	2439	26.4	28.9	16.6	28.1	100.0

While commenting on the quality of seed supplied, 70% of the participating households regarded the seed to be of medium to high quality, whereas 20% judged the seeds to be of poor quality, manifested as mixture or resulting in crop failure. Some 10% of the households could not judge the quality of seed they received (data not shown).

The source of information for the participating households’ about the seeds they have used included own prior knowledge/information and reading extension manuals about the importance of seeds of a particular variety (18%), demonstration and observation of field days (12%) and upon recommendation by neighbors, friends, relatives, NGOs and research centers (13%). More than one-half (57%) of the participating households received improved seeds of a particular variety upon recommendation by extension agents (Data not shown). Demonstrating crop cutting and weighing as well as organizing field days at different levels where prizes are given for farmers who achieved the highest yield are often known to attract farmers to adopt improved seed varieties (Zegeye et al., 2001a).

Out of the participating households, 39% planted improved seeds on 1-20% of their total arable land during the past agricultural year (2003/2004), and those who planted on 81-100% of their arable land numbered only 10% (Table 3.8). The results suggested that despite development and official release of many varieties of cereals, pulses, oilseeds, and horticultural crops, there is limited adoption of the developed improved varieties by farmers who still predominantly practice traditional agriculture mainly because of the absence of appropriate seed delivery and promotion systems.

Table 3.8. Percentage of total arable land area of the sampled households which was actually planted to improved seed during the 2003/04 cropping season

Region	Sample	1-20%	21-40%	41-60%	61-80%	81-100%	Cannot estimate	Total
Tigray	112	48.2	14.3	8.0	2.7	5.4	21.4	100.0
Afar	63	3.2	3.2	3.2	11.1	33.3	46.0	100.0
Amhara	413	33.7	12.8	21.3	14.3	13.1	4.8	100.0
Oromia	898	45.3	18.0	13.4	6.5	9.9	6.9	100.0
Somali	14	28.6	7.1	42.9	7.1	0.0	14.3	100.0
B. Gumuz	42	33.3	21.4	4.8	2.4	7.1	31.0	100.0
SNNPR	266	31.6	24.1	17.7	5.2	2.2	19.2	100.0
Dire Dawa	21	38.1	28.6	14.3	0.0	19.0	0.0	100.0
Harari	29	17.2	24.1	34.5	10.3	7.0	6.9	100.0
Total	1858	38.6	17.2	15.4	7.9	10.0	10.9	100.0

Slightly over fifty percent (53%) of the participating households indicated the presence of improved seed distribution center nearby, while 47% pointed out the absence of nearby improved seed distribution center (Table 3.9). For Dire Dawa, Benshangul Gumuz, Tigray, Amhara and Oromia more than 50% of the households were found to have nearby seed distribution center, whereas Afar, Somali, SNNPR and Harari seemed to have nearby seed distribution center for less than 50% of their farm households. About one-quarter of those households with far away located improved seed distribution center were found to travel 5 to 50 km to get improved seed. Previous studies have indicated the presence of acute shortage of infrastructure facilities where close to 75% of the farmers are more than half-day walks from all weather roads (Amha, 1999).

3.3.2 Inorganic fertilizers

Over one-third (36%) of the households responded that they did not use inorganic fertilizer during the 2003/04 crop season. Previous study also indicated that the proportion of households in 1995/96 who bought fertilizer was 28.5% (MOLSA, 1997), suggesting about 29% increase over the decade. For those households who used inorganic fertilizer, the major suppliers were Bureau of Agriculture (35%), AISCO (29%), other trade firms (19%), cooperatives/unions (13%), and NGOs supplied about 4% (Table 3.10).

Table 3.9. Percentage of sampled households with or without close by seed distribution center

Region	Sample	Yes	No	Total
Tigray	346	61.3	38.7	100.0
Afar	172	32.6	67.4	100.0
Amhara	992	55.9	44.1	100.0
Oromia	1868	55.6	44.4	100.0
Somali	99	6.1	93.9	100.0
Benshangul Gumuz	100	74	26	100.0
SNNPR	845	47.2	52.8	100.0
Dire Dawa	49	81.6	18.4	100.0
Harari	51	45.1	54.9	100.0
Total	4522	52.9	47.1	100.0

Table 3.10. Major agencies which supplied households with an inorganic fertilizer during the past year (2003/04) (%)

Region	Sample	AISCO	Bureau of Agric.	Cooperatives/ Unions	NGOs	Other trade firms	Total
Tigray	217	32.7	44.7	13.4	3.7	5.5	100.0
Afar	83	24.1	42.2	2.4	6.0	25.3	100.0
Amhara	600	31.5	29.0	19.0	2.8	17.7	100.0
Oromia	1284	28.7	38.9	12.0	3.7	16.7	100.0
Somali	46	47.8	34.8	6.6	0.0	10.8	100.0
B. Gumuz	69	17.4	55.1	13.1	4.3	10.1	100.0
SNNPR	525	26.5	22.3	9.1	5.1	37	100.0
Dire Dawa	25	16.0	64.0	12.0	8.0	0	100.0
Harari	29	13.8	55.2	31.0	0.0	0	100.0
Total	2878	28.8	35.1	12.9	3.8	19.4	100.0

Out of the sampled households who used inorganic fertilizer during 2003/04 cropping season, about 23% responded to have used inorganic fertilizer on less than or equal to 20% of their total arable land area. Another 26% of the participating households applied inorganic fertilizer on 81-100% of their total arable land, which is one and half times that of improved seed planters for the same range of land (Table 3.11). This suggested that the level of inorganic fertilizer use is more than that of improved seed. However, farmers cannot gain potential benefit or profitability from fertilizers unless they use with improved seed and improved cultural practices.

Table 3.11 Percentage of total arable land area actually fertilized with inorganic fertilizers during the 2003/2004 cropping season by the participating households

Region	Sample	1-20%	21-40%	41-60%	61-80%	81-100%	Cannot estimate	Total
Tigray	157	14.6	12.7	22.3	12.7	27.5	10.2	100.0
Afar	65	6.2	4.6	1.5	13.8	37.0	36.9	100.0
Amhara	610	19.5	9.0	18.7	18.2	29.4	5.2	100.0
Oromia	1284	24.7	10.8	16.9	12.0	30.1	5.5	100.0
Somali	16	18.8	12.4	31.3	12.4	18.8	6.3	100.0
B. Gumuz	48	29.2	16.7	10.4	2.1	14.6	27.1	100.0
SNNRP	466	25.8	15.9	18.7	9.9	7.1	22.6	100.0
Dire Dawa	9	0.0	11.1	22.3	22.2	44.4	0.0	100.0
Harari	32	21.9	15.6	18.8	15.6	28.1	0.0	100.0
Total	2687	22.6	11.4	17.6	13.0	25.6	9.8	100.0

The mean amount of fertilizer applied for major cereals during the 2003/04 cropping season in the four major cereal-producing regions is given in Table 3.12. The result showed that relatively more amount of fertilizer was applied for wheat and maize in all the regions. The amount applied for sorghum was less than one quintal per ha except in Oromia, but one cannot be sure why the farmers have reduced the rate for sorghum. Generally, it seems that about one quintal per hectare of inorganic fertilizer is applied for cereals in the regions. Similar to the case with improved seed distribution, close to one-half (46.5%) of the households indicated absence of fertilizer distribution center nearby, while slightly more than one-half (53.5%) pointed out the presence of nearby fertilizer distribution center (Table 3.13). About one-quarter of those households with far away located fertilizer distribution center were found to travel 5 to 50 km to get inorganic fertilizer.

Table 3.12. Mean inorganic fertilizer applied for major cereals during meher 2003/04 cropping season

Region	Crop	Sample	Inorganic fertilizer applied (Qt/ha)	Region	Crop	Sample	Inorganic fertilizer applied (Qt/ha)
Tigray	White tef	33	1.3	Oromia	White tef	443	1.2
	Sergegna tef	71	1		Sergegna tef	681	1.1
	Barley	33	1		Barley	393	1
	Wheat	37	1.7		Wheat	916	1.3
	Maize	20	1.9		Maize	444	1.4
	Sorghum	4	0.9		Sorghum	68	1.3
Mean			1.3	Mean			1.2
Amhara	White tef	378	1.4	SNNPR	White tef	69	0.7
	Sergegna tef	235	1.1		Sergegna tef	41	1.1
	Barley	118	1.3		Barley	105	1.6
	Wheat	357	1.7		Wheat	227	1.3
	Maize	337	1.8		Maize	228	1.2
	Sorghum	23	0.6		Sorghum	5	0.5
Mean			1.5	Mean			1.2

Table 3.13. Percentage of sample households who were with or without fertilizer distribution centers

Region	Sample	Yes	No	Total
Tigray	346	69.1	30.9	100.0
Afar	171	15.8	84.2	100.0
Amhara	989	57.8	42.2	100.0
Oromia	1863	56.2	43.8	100.0
Somali	99	6.1	93.9	100.0
Benshangul Gumuz	99	86.9	13.1	100.0
SNNPR	847	45.90	54.10	100.0
Dire Dawa	49	51.00	49.0	100.0
Harari	51	47.10	52.9	100.0
Total	4514	53.50	46.5	100.0

3.3.3. Organic fertilizers and other soil fertility ameliorating practices

Other major soil ameliorating practices used by the participating households included application of manure (animal dung, farmyard manure and compost) (57%), crop rotation (32%), and fallowing (8%). Intercropping with legumes, green manure and terracing as well as tie-riding played insignificant (about 3%) role in soil fertility improvement. The major soil types in Ethiopia, which include nitisols, cambisols and vertisols, are deficient in nitrogen and phosphorus, making the use of inorganic fertilizers imperative for increased productivity of crop production. Majority of farmers use inorganic fertilizers along with local seeds, leading to low yield responses, probably due to lack of awareness of the fact that combining fertilizer and improved seeds could further improve yield and improve profitability. It is encouraging to find that the use of crop rotation is considerable among the farmers given the fact that mono-cropping is one of the major reasons for soil fertility depletion and building up of diseases, insect pests and weeds. The negative consequences of mono-cropping have already been observed in Arsi-Bale highlands (where long years of continuous mono-cropping of wheat resulted in reduced yield) (Bale Agricultural Development Enterprise, 2003) and western Ethiopia (where mono-cropping of maize has led to reduced yield) (Bako Agricultural Research Center, 2005). The negative consequences are not only yield reduction but also building of weeds, diseases and soil fertility depletion.

The mean amount of organic fertilizers applied for major cereals during meher 2003/04 cropping season in four major cereal producing regions of Ethiopia is given in Table 3.14. Overall, the application seemed to be higher for the SNNPR but lower for Oromia, followed by Amhara regions. The latter two regions are known for their higher amount of inorganic fertilizer consumption in that order than the other regions in the country.

Table 3.14. Mean amount of organic fertilizer (both compost and manure) applied for major cereals during meher 2003/04 cropping season in four major cereal producing regions of Ethiopia

Region	Crop	Sample	Organic fertilizer applied (qt/ha)	Region	Crop	Sample	Organic fertilizer applied (qt/ha)
Tigray	White tef	6	45.1	Oromia	White tef	16	3.3
	Sergegna tef	11	16.8		Sergegna tef	10	3.8
	Barley	13	17		Barley	26	9.2
	Wheat	10	8.9		Wheat	30	6.1
	Maize	22	43.1		Maize	189	8
	Sorghum	7	19.6		Sorghum	99	9.9
Mean			26.8	Mean			8.1
Amhara	White tef	25	11.6	SNNPR	White tef	2	52.7
	Sergegna tef	26	8.6		Sergegna tef	3	64.3
	Barley	39	15.2		Barley	6	21.4
	Wheat	51	12.7		Wheat	17	36.1
	Maize	149	10.5		Maize	28	37.5
	Sorghum	18	10.5		Sorghum	1	30
Mean			11.4	Mean			37.2

Several studies on technology adoptions have shown that the probability of adopting fertilizer, improved seeds and other related technologies is positively influenced by the size of the farm holding, size of the family, number of oxen owned, ability to read and write, etc. For instance, study conducted on maize technology adoption in southwestern Ethiopia concluded that adopters of high yielding improved maize were resource endowed in that they operated, on an average, larger farm sizes, had more arable area of land and maintained more herd sizes, and had adequate number of draught animals for their farm operation as compared with non-adopters (Broeck and Dercon, 2001; Zegeye, et al., 2001). Such strategy may lead to increasing the gap between resource-rich and resource-poor farmers as often noted in previous extension systems such as the maximum and minimum packages of the 1970s and early 1980s.

3.3.4 Farm implements

The results of the present study showed that labor and draught animals are the primary power sources for most of the farm operations such as primary and secondary tillages, planting, weeding, harvesting, threshing/shelling and transport for the overwhelming majority of the sampled households (>90%). A very limited use of tractor and combine harvester was noted only in Oromia, reiterating the fact that the Ethiopian agriculture is less influenced by modern farm implements. Aggravating the situation, about 43% of the households did not have farm implements or tools that are essential for their farm operations. This predominant use of labor and draught animals in Ethiopia is comparable to the situation in USA and Europe before 1917 (Wayne, 1922). Similarly, the proportion of labor force engaged in agriculture around this period was 40% for France, 42% Italy, 59% for Portugal, and 65% for USA. By 1990, however, the proportion of labor force engaged in agriculture in respective countries was shrunken to 6%, 9%, 15% and 3% (Marena, 1996), following farm mechanization, which led to surplus production and increased labor productivity.

In Ethiopia, smallholder farmers are still predominantly using traditional equipment/and farm tools in that roughly one-third of the households (36%) used plow, i.e. maresha (traditional plow). Uses of BBM, maize Sheller, wheelbarrow, trailer, combine harvester, enset squeezer and honey refiner were found to be at low percentage. The present sampling was able to capture only 16 households (1 each for Benshangul Gumuz and Afar, 6 each for Amhara and Oromia and 2 for SNNPR) using BBM. The BBM has been proved to be useful equipment for draining the black vertisols for increased production and also for double cropping, which is otherwise poorly drained (waterlogged) for optimum crop production. Given the significant occurrence of vertisols in the Ethiopian highlands and the limited distribution and use of BBM at the moment, there is urgent need to expand the utilization of such useful farm equipment. Likewise, there is a need to increase the distribution and use of enset scraper and enset squeezer. Close to one-half of the households (48.4%) indicated that they need farm implements for efficient utilization of labor, 33% need to increase farm production and 16% need the implements to reduce post harvest crop losses. These purposes match with the fact that the mechanization of USA and European agriculture after the Second World War has lead to surplus production and consequently reduction of the labor force engaged in agriculture as result of the increased labor productivity (Glennon and Garraty, 1999).

3.3.5. Storage structures of agricultural products

In the present study, 64% of the sampled households were found to store their crop produce before sale, 11% sale immediately after harvest and 26% had no surplus for sale (Table 3.15). This result implied that about a quarter of the farmers are subsistent farmers, revealing the challenge of transforming Ethiopian agriculture to market-oriented agriculture.

Table 3.15. Percentage of sampled households who store their crop produce before sale

Region	Sample	Store for some time	Sale immediately after harvest	Had no surplus for sale	Total
Tigray	342	48.8	14.6	36.5	100.0
Afar	150	38.0	42.0	20.0	100.0
Amhara	992	76.8	6.1	17.0	100.0
Oromia	1868	63.8	9.3	26.9	100.0
Somali	98	59.2	12.2	28.6	100.0
B. Gumuz	100	54.0	13.0	33.0	100.0
SNNPRS	848	62.1	12.3	25.6	100.0
Diredawa	50	8.0	2.0	90.0	100.0
Harari	51	80.4	9.8	9.8	100.0
Total	4499	63.6	10.7	25.7	100.0

Regarding the type of storage structures being used to store produce before sale, 39% of the households store in bags/sacks, 44% in traditional granaries such as gotera, debegnet and guota, 9% in underground pits and the remaining 8% used structures like barrel and pots (Data not shown). Of the more than three-quarter (78%) of the sampled households who could estimate the storage capacity of their storage structure, about three-quarter of the households possessed storage structures with only up to 10 quintal. It can be concluded from this result that virtually all households are using traditional storage structures of limited capacity. The use of traditional storage structures leads to high post harvest crop loses and may force farmers to sell their excess produce early when market price is low due to high supply. Probably this lack of adequate storage facility and the need to pay loan early have forced farmers to sell their produce early.

The present study provided support to the above arguments in that over one-half of the households (57%) were aware of post harvest crop losses as one of their major problems. Generally, pests, poor on farm transport, poor storage facilities and wastage due to traditional food processing and preservation are the major causes of crop harvest losses in the country. The study further revealed that cereals were mostly susceptible to post harvest loss (88%), of which maize (31.5%), wheat (18%) and sorghum (17%) were found to be the most susceptible. Weevils (43%) and rodents (39%) were found to be the major agents for post harvest crop losses, while other agents such as grain mold accounted for another 18%. About 84% of the households estimated the loss to be 1-10% of their crop produce, 8% indicated that they lose 11-20% and another 8% estimated the amount of post harvest loss to be 21-50%. Generally, post harvest crop losses are estimated to be between 20-40% in Ethiopia.

Most households use chemicals (43%) and aeration (36%) to minimize post harvest crop losses (Table 3.16). Use of cat or rat poison (9.5%) and sell and repurchase (7%) were found to be other practices of considerable importance. About 5% of the household used cultural control methods such as polishing with salt, ash, and pepper as well as secure placement.

Table 3.16. Major practices used by the sampled households to avoid post-harvest crop losses (%)

Region	Sample	Aeration	Pesticides	Cat or rat poison	Cultural control	Sale out	Total
Tigray	301	33.6	48.2	11.3	2.0	5.0	100.0
Afar	125	44.8	36.0	7.2	4.8	7.2	100.0
Amhara	828	38.9	41.7	8.9	4.6	5.9	100.0
Oromia	1738	31.9	43.2	10.1	5.6	9.3	100.0
Somali	63	54.0	31.7	9.5	0.0	4.8	100.0
Benshangul Gumuz	100	39.0	43.0	11.0	6.0	1.0	100.0
SNNPR	752	38.4	42.6	8.1	4.8	6.1	100.0
Dire Dawa	50	38.0	38.0	8.0	4.0	12.0	100.0
Harari	51	23.5	43.1	15.7	11.8	5.9	100.0
Total	4008	35.6	42.6	9.5	4.9	7.3	100.0

3. 4 Trend of input use

The present study showed that only 39% of the participating households used complete package of crop production, i.e., improved seeds, fertilizer and improved cultural practices (Annex 3.3). The remaining households (61%) who were participating in the extension packages used incomplete package of crop production lacking one or more of the major components. This incomplete use of extension package often reduces the profitability of input use and reduces household’s confidence in the sustainable use of modern agricultural inputs. Farmers using fertilizer would achieve limited gain from the input, which absorbs nearly all of their cash expenditure, unless used together with other inputs. This in turn calls for an integrated approach to technology promotion and delivery services.

About 43% of the participating households increased and 35% maintained the use of inputs such as seed and fertilizer as they continue participating in extension package, whereas 22% decreased the use of such inputs. Some households responded to have abandoned certain package programs, primarily applying fertilizers and packages on crops. Most of the abandoned packages have involved fertilizer, probably because of the increasing fertilizer price and tight repayment of fertilizer credit right after harvest, when usually grain price is low. Households mentioned variable reasons as to why they have abandoned certain package programs. The main factors included difficulty of getting credit for inputs (27%), too low grain price to encourage surplus production (26%), low profitability/productivity of the packages per se (18%), fear of rain shortage (9%). Other reasons included low response to inorganic fertilizers, declining productivity, and land shortage (4.4%) (Annex 3.4).

3.5 Menu-based extension system

Over three-quarter of the households (78%) did not know menu-based or family-based extension system (Table 3.17). About 15% of the households responded to participated in menu-based extension system. Among those households who have participated in the menu-based extension system, 47% indicated the existence of difference between the regular and the menu-based extension systems, while the remaining 53% households said there was no difference or could not tell whether there is a difference or not (Data not shown). Close to two-third (66%) of the participating households participated in food crops (mainly cereals) production in the menu-based packages, indicating great similarity between the menu-based and the regular extension systems (Table 3.18). Involvement in high value crops was found to be low, despite the attention given to it. Similarly, involvement in natural resources management, such as agro-forestry, community forestry and soil and water conservation as well as in farm implements was minimal.

Table 3.17. Percentage of households who have participated in menu-based extension system

Region	Sample	Yes	No idea	Did not want to participate	Did not participate	Total
Tigray	347	42.9	40.6	13.3	3.2	100.0
Afar	194	4.1	86.6	0.5	8.8	100.0
Amhara	991	13.5	80.8	1.9	3.8	100.0
Oromia	1848	14.7	78.8	2.1	4.4	100.0
Somali	97	3.1	95.9	1.0	0.0	100.0
Benishangul Gumuz	98	5.1	54.0	3.1	37.8	100.0
SNNPRS	840	10.0	85.4	0.8	3.8	100.0
Dire Dawa	50	0.0	100.0	0.0	0.0	100.0
Harari	51	7.8	92.2	0.0	0.0	100.0
Total	4516	14.6	78.1	2.5	4.8	100.0

Table 3.18. Menu-based extension packages in which households have participated (%)

Region	Sample	Food crops	Cash crops	Natural resources management	Livestock	Total
Tigray	50	80.0	0.0	4.0	16.0	100.0
Amhara	172	66.9	0.6	4.7	27.8	100.0
Oromia	464	64.4	0.4	1.7	33.5	100.0
Somali	38	76.3	0.0	2.6	21.1	100.0
Benshangul Gumuz	50	24.0	0.0	0.0	76.0	100.0
SNNPR	255	71.4	1.2	2.3	25.1	100.0
Harari	19	100.0	0.0	0.0	0.0	100.0
Total	1048	66.4	0.6	2.4	30.6	100.0

The major positive aspects of menu-based extension packages included accommodation of the entire family (26%), saving government expenses (12%), more productivity (8%), inclusion of all packages in one system instead of separate packages (7%), more target oriented (6%), adequate support from expertise (5.5%) (Data not shown).

3.6 Yield and production assessment

Results of the present study revealed that roughly three-quarter (73%) of the participating households responded that participation in the extension system has increased the level of their production and another 13% could not observe clear trend. Those who declared decrease of and no change in production were only about 5% and 7%, respectively, and about 2% had no suggestion (Data not shown). In an attempt to qualitatively estimate the extent of increment in production, 46% of the households responded that their production has increased by up to one-fourth, about 50% reported production increment ranging from two to four folds (Table 3.19). Many lines of evidence indicate that the increase in crop production from 1995 to 2004 has been mostly due to area expansion and not due to increase in yield, despite the increasing use of fertilizers and promotion and adoption of the technological package. Generally, good rains encourage farmers to cultivate more area than they would otherwise. However, such area expansion is achieved largely through cultivation of hillsides with deep slopes, reducing or eliminating fallow land, and converting pasture, woodland and forest areas into farmland, constraining the sustainability of the natural resource base.

Table 3.19. Household perception of extent of increase of production since households started participating in the extension program (%)

Region	Sample	Increased <25%	Increased 25%	Doubled	Tripled	Quadrupled	Increased >4x	No suggestion	Total
Tigray	96	25.0	31.3	24.0	9.4	8.3	2.1	0.0	100.0
Afar	65	10.8	27.7	35.4	15.4	10.8	0.0	0.0	100.0
Amhara	248	11.3	34.7	40.7	6.0	3.6	2.0	1.6	100.0
Oromia	634	11.5	34.5	38.0	7.9	3.5	3.5	1.1	100.0
Somali	28	3.6	53.6	35.7	3.6	0.0	3.6	0.0	100.0
B. Gumuz	4	50.0	0.0	50.0	0.0	0.0	0.0	0.0	100.0
SNNPR	296	7.8	35.1	37.2	9.8	5.4	3.0	1.7	100.0
Dire Dawa	7	14.3	57.1	14.3	0.0	14.3	0.0	0.0	100.0
Harari	26	11.5	38.5	38.5	7.7	0.0	0.0	3.8	100.0
Total	1404	11.5	34.6	37.1	8.3	4.5	2.8	1.2	100.0

Table 3.20 presents quantitative estimation of the mean yield (qt/ha) of five major cereals in four regions of Ethiopia. Mean productivity was found to be higher for Oromia than other regions. Comparison of the overall average yield of these crops with package and traditional practices of previous years revealed that the average yield of each crop in the present study was greater than the corresponding yield in the traditional practices of the previous years, except for tef (Table 3.21). However, while comparing the yield of each crop in the present study with corresponding yield under the previous years package yield, there was a decrease for tef, maize and sorghum, suggesting lack of continued sustainability of increased productivity. It seems that the yield level was more or less maintained for barley and wheat. Comparison of adopters of the extension package and non-adopters for major cereals revealed that generally adopters harvest more yield than non-adopters, albeit low percentage difference except for maize and sorghum and the difference tends to be negative for white tef (Table 3.22).

Table 3.20. Mean productivity (Qt/ha) of major cereals during meher 2003/04 cropping season for households who participated in extension package in four major cereal producing regions of Ethiopia

Region	Crop	Sample	Yield (Qt/ha)	Region	Crop	Sample	Yield (Qt/ha)
Tigray	White tef	44	10	Oromia	White tef	215	7.8
	Sergegna tef	62	6.2		Sergegna tef	423	11.1
	Barley	49	7.9		Barley	278	27.6
	Wheat	50	23.5		Wheat	524	35.4
	Maize	35	11.7		Maize	660	29
	Sorghum	79	20.3		Sorghum	307	12.3
	Mean		13.8		Mean		23.1
Amhara	White tef	240	11.2	SNNPR	White tef	53	13.2
	Sergegna tef	266	9.6		Sergegna tef	64	8.8
	Barley	91	11.3		Barley	98	17.1
	Wheat	193	16.6		Wheat	81	20.8
	Maize	347	34.5		Maize	316	18.3
	Sorghum	185	12.8		Sorghum	44	16.7
	Mean		18		Mean		17

Table 3.21. Comparison of mean productivity (Qt/ha) of 5 major cereals of the present study with previous report^*

Crop	Overall average for the present study	Average for previous package practices (1995/96-2000/01)	Average previous traditional practices (1995/96-2000/01)
Tef	10.0	14.8	8.5
Barley	20.9	21.0	10.0
Wheat	29.0	28.6	11.8
Maize	27.5	46.3	15.8
Sorghum	13.8	26.8	10.9

^* Source: Data for the previous years was obtained from Quarterly Bulletin of the Ministry of Agriculture 7^th Year No. 10, Dec. 2001 (in Amharic).

Table 3.22. Mean productivity (Qt/ha) of major cereals between extension participants and non-participants
Crop	Sample		Yield (Qt/ha)		% Difference
Crop	Participant	Non-Participant	Participant	Non-Participant	% Difference
Barley	308	220	13.5	12.7	6.4
Maize	954	506	22.2	17.5	26.3
Sorghum	377	362	12.1	9.9	22.2
White tef	392	168	9.8	9.9	-1.1
Sergegna tef	507	315	8.4	7.8	7.9
Wheat	555	286	18.5	16.8	10.3

Similarly, comparison of adopters of the extension package and non-adopters for using inorganic fertilizers on major cereals showed that adopters harvest more yield than non-adopters, still except for white tef, where the difference was found to be negative (Table 3.23). Still similar trend was observed while using improved seeds (Table 3.24). In general, comparison of adopters and non-adopters recorded more difference while using or not using inorganic fertilizers than while using or not using improved seeds, implying the much deteriorated soil fertility of the arable land and the need to increase the rate of application of fertilizer per hectare. From among the cereals, maize and sorghum tended to show more positive yield differences between adopters and non-adopters, suggesting that these crops are more responsive to the use of modern inputs. There is no plausible reason why the yield difference for white tef was negative. In general, red seeded tef variety is higher yielder than white seeded one.

Table 3.23. Mean productivity (Qt/ha) of major cereals when using inorganic fertilizers

Crop	Sample Size		Yield (Qt/ha)		% Difference
Crop	Participant	Non-Participant	Participant	Non-Participant	% Difference
Barley	237	293	16.2	10.7	51.2
Maize	727	736	25.4	15.8	60.8
Sorghum	106	635	13.2	10.7	23.8
White tef	394	166	9.4	10.7	-12.3
Sergegna tef	465	357	8.6	7.7	12.4
Wheat	578	264	19.7	13.9	42.2

Table 3.24. Mean productivity (Qt/ha) of major cereals when using improved seeds

Crop	Sample Size		Yield (Qt/ha)		% Difference
Crop	Participant	Non-Participant	Participant	Non-Participant	% Difference
Barley	40	460	16.8	13.0	29.5
Maize	558	861	24.5	18.2	34.6
Sorghum	112	585	13.3	10.6	25.9
White tef	43	480	9.3	9.6	-2.7
Sergegna tef	41	739	9.4	8.2	14.0
Wheat	210	606	19.8	17.3	14.3

Technology evaluation by the participating households revealed that productivity increment was medium to very high when such technologies like improved seeds, fertilizers, herbicides, insecticides, proper and timely seedbed preparation, BBM, tie-ridge, terraces, etc. were used (Annex 3.5a-d). The results showed the productivity increases that could be achieved when farmers were provided with appropriate technologies as well as information.

While commenting on the sustainability of the increased crop production, 44% of the participating households responded that the increment in production remained on the first increase and another 17% declared that the production is improving from time to time (Table 3.25). Still another 36% of the households pointed out the production has come down to its previous state. It is interesting to note that the percentage of households who reported decrease of production is negligible, suggesting the positive aspect of participating in the extension system.

Table 3.25. Sustainability of increased productivity after once participated in extension package programs (%)

Region	Sample	Decreased	Remained same	Increased	Fluctuated	Total
Tigray	141	45.4	48.2	5.7	0.7	100.0
Afar	36	30.6	58.3	11.1	0.0	100.0
Amhara	570	31.6	50.2	17.2	1.1	100.0
Oromia	1045	33.2	43.2	21.0	2.7	100.0
Somali	12	41.7	58.3	0.0	0.0	100.0
B. Gumuz	47	46.8	36.2	12.8	4.3	100.0
SNNPR	284	50.7	35.2	9.9	4.2	100.0
Dire Dawa	10	30.0	30.0	30.0	10.0	100.0
Harari	19	57.9	26.3	0.0	15.8	100.0
Total	2164	36.3	44.3	16.9	2.4	100.0

Two-third of the participating households (67%) have continued using the extension package once they had involved in the package and some 7% abandoned some of the components of the package, whereas 26% did not continue using the package (Table 3.26).

Table 3.26. Whether households continue using extension package once they have involved in the package (%)

Region	Sample	Maintained	Abandoned	Abandoned some of the components of the package	Total
Tigray	175	78.9	19.4	1.7	100.0
Afar	32	56.3	43.8	0.0	100.0
Amhara	602	75.4	19.4	5.1	100.0
Oromia	1114	64.4	25.3	10.3	100.0
Somali	14	92.9	7.1	0.0	100.0
Benshangul Gumuz	54	48.1	48.1	3.7	100.0
SNNPR	344	59.6	36.6	3.8	100.0
Dire Dawa	10	50.0	30.0	20.0	100.0
Harari	23	34.8	43.5	21.7	100.0
Total	2368	66.9	25.9	7.2	100.0

3.7. Challenges faced in crop production extension package

3.7.1. Improved seeds

The problems associated with the use of improved seeds included susceptibility to insect pests, diseases, weeds and drought (21%), poor adaptation to local conditions, low quality involving poor grain quality, poor germination and mixture (24%) and low yield (8%) (Table 3.27). Lack of systematic provisions and clear ownership for maintenance of variety both during initial breeder seed multiplication and later on during repeated multiplication of the same variety for restocking may lead to deterioration of seed quality (EARO, 2000). Shortages of improved varieties for high value crops and for moisture stress areas are often noted as one of limitations of PADETES (Devereux, 2000; Fantahun and Minuye, 2000). Several studies have also indicated the need for better disease and insect tolerant varieties of crops in order to enhance the adoption of improved varieties by farmers ((Zegeye, et al. 2001b; SG2000, 2002). Poor adaptation includes that the varieties lack one or more of the traits most desired by farmers such as cooking quality, taste, etc. Thus increasing farmers’ access to their preferred varieties would result in a faster rate of diffusion through farmer-to-farmer seed exchange (Mulatu and Belete, 2001; Mulatu and Zelleke, 2001). Utilization of quality seed for planting is also the most essential factor for increased crop production and income of farmers.

Table 3.27. Major problems associated with the use of improved seeds (%)

Region	Sample	Susceptibility to diseases & insect pests	Low yield	Poor adaptation & quality	Marketing & distribution	No problem	Total
Tigray	192	30.7	5.2	15.6	1.6	46.9	100.0
Afar	113	29.2	5.3	15.1	10.6	39.8	100.0
Amhara	463	18.4	9.1	23.3	4.5	44.7	100.0
Oromia	995	19.7	7.9	23.5	3.0	45.9	100.0
Somali	41	9.8	7.3	29.3	2.4	51.2	100.0
B. Gumuz	45	15.6	0.0	26.7	13.3	44.4	100.0
SNNPR	390	23.1	9.0	31.3	3.0	33.6	100.0
Dire Dawa	9	11.1	0.0	33.3	0.0	55.6	100.0
Harari	16	31.2	0.0	37.5	0.0	31.3	100.0
Total	2264	21.2	7.8	24.0	3.7	43.3	100.0

3.7.1. Fertilizer

The key problems in fertilizer use included high price (39%), late supply 33%), tight credit repayment schedule (16%) and long transportation distance from the distribution center (6%) (Table 3.28). Close to three-quarter (74%) of the participating households responded to have encountered problems in fertilizer market (Table 3.29).

Table 3.28. Major problems households encountered in purchasing fertilizer (%)

Region	Sample	Late supply	High price	Tight credit repayment schedule	Far located distribution centers	Low quality	Others	Total
Tigray	265	29.7	49.3	14.8	2.4	1.8	2.1	100.0
Afar	67	28.6	49.0	18.8	2.6	0.5	0.5	100.0
Amhara	798	28.9	45.7	12.5	9.2	1.8	1.9	100.0
Oromia	1542	25.8	49.7	13.7	6.4	2.2	2.2	100.0
Somali	25	30.2	42.9	12.7	7.9	1.6	4.8	100.0
Benshangul Gumuz	80	12.0	70.0	10.0	4.0	0.0	4.0	100.0
SNNPR	586	23.3	54.1	13.2	4.7	2.5	2.0	100.0
Dire Dawa	26	32.7	38.8	16.3	6.1	6.1	0.0	100.0
Harari	40	31.4	39.2	19.6	2.0	3.9	3.9	100.0
Total	3429	26.3	49.7	13.7	6.1	2.1	2.1	100.0

Table 3.29. Percentage of households’ who faced problems in fertilizer marketing

Region	Sample	Yes	No	Total
Tigray	265	37.4	62.6	100.0
Afar	67	61.2	38.8	100.0
Amhara	798	65.7	34.3	100.0
Oromia	1542	83.9	16.1	100.0
Somali	25	68.0	32.0	100.0
Benshangul Gumuz	80	52.5	47.5	100.0
SNNPR	586	79.4	20.6	100.0
Dire Dawa	26	38.5	61.5	100.0
Harari	40	82.5	17.5	100.0
Total	3429	73.6	26.4	100.0

It is noteworthy to indicate that various sources have shown that despite the increasing level of fertilizer import and consumption, Ethiopia is still one of the lowest users of fertilizer in the world (World Bank, 1995). Moreover, Ferede et al. (2000) argued that despite the relatively strengthened extension services over the last decade, farmers continue to apply less than the recommended rates. The current combined rate of application of about 31 kg fertilizer per ha is very low compared to the recommended combined optimal rate of 150-200 kg per ha of DAP and Urea, suggesting that the increase in quantity of fertilizer consumed over the past few years has been due mainly to area expansion and not intensified application. The recommended application rates are also considered to be unreliable; they are blanket, without any scientific basis and are not area specific (SG2000, 2002) and also do not take into account annual soil fertility depletion, soil type and rainfall pattern. The implication is that the actual rates could be much lower than the real need (Zegeye et al. 2001b). The Ethiopian government established 17 regional soil laboratories through the World Bank loan before ten years to develop location specific fertilizer rate recommendations, though the outputs of such laboratories are so far limited.

Fertilizer in Ethiopia is often applied by broadcasting, leading to considerable losses with low efficiency. Such application method is applicable only for small cereals like tef, wheat and barley. It is advisable to use pocket or band application for crops such as maize, sorghum, and potato, requiring row planting of these crops. Moreover, calculations of the profitability of fertilizer use by small farmers have shown that the return to investment in fertilizer, albeit positive, is not commensurate with the risk or it is low, suggesting that the barriers to increased fertilizer use are knowledge gap, risk aversion and marketing inefficiencies, which can be overcome through strengthening extension services, improving transport infrastructure and marketing.

With regard to the provision of farm implements, the present study revealed that the main constraints were high purchase price and maintenance cost (including poor quality leading to high unaffordable initial price and high maintenance cost) as indicated by 90% of the households and unavailability of the implements on the market (10%). Waktola (1980) also reported that technical and economic constraints were hindering wide adoption of farm implements developed by CADU/ARDU and recommended the need for establishing and operating agricultural tools/implements repair and maintenance workshops and to promote low-cost appropriate technology in order to raise the productivity and income levels in rural areas.

4. Extension package in natural resources management

In the present study majority of the households (84%) used owned land for cultivation. The proportion of households renting out their land for fixed payments (0.2%) and shared cropped out (1.4%) was insignificant. Similarly, the percentage of households rented in plots for fixed payments (4%) and shared-in crop (4%) was low. Unfortunately, the percentage of households putting land for fallow, man-made woodlot, natural forest, virgin land and communal grazing land was also low. It is noteworthy that the country has suffered severe deforestation with forest cover dropping from an estimated 30% at the beginning of the last century to under 4% today. The results suggest the opportunity of employment creation for landless farmers is gloomy by renting land in or out in the rural areas.

4.1 Technology dissemination

4.1 .1 Soil and water conservation

Land degradation, as manifested by low soil fertility, is the major problem facing Ethiopian agriculture. In the present study about 39% of the households applied either organic or inorganic fertilizers or both to enhance soil fertility, about 23% used crop rotation, and about 20% used various forms of natural resources management such as agro-forestry and terracing (Annex 4.1). Mixed cropping, tree planting, agro-forestry, ridges and micro-basin development played very low role in mitigating land degradation/soil erosion. In Ethiopia, cow dung and crop residues are often burned because of lack of fuel-wood, leading to gradual depletion of soil nutrients. An increasing amount of land is being withdrawn from cultivation due to land degradation, and the demand for additional land for cultivation is being met by extending cultivation to areas previously designated as pasture and forestland (Alemu, 2002).

The limited practices of tree planting and agro-forestry are unfortunate, given the fact that 39% of the sampled households were witnessing that land degradation is a problem for their effective crop production as well as their environment (Table 4.1). The present study further strengthened the issue that 58% of the households identified soil erosion as one of the major causes of land degradation. Another 24% of the households pointed out over cultivation and 8% soil erosion by wind as the major cause of land degradation/soil erosion. To a limited extent, overgrazing (3%), invasive weeds (3%) and flood damage (4%) were also cited as causes of land degradation/soil erosion. Annual soil loss in Ethiopia is estimated between 1.5 and 3 billion tonnes, of which 50% occurs in croplands where soil loss may be as high as 296 tons/year on a steep slope (Ministry of Agriculture, 1987).

Table 4.1. Percentage of sampled households encountered land degradation problem for their effective farming

Region	Sample	Yes	No	Total
Tigray	347	26.5	72.0	100.0
Afar	190	15.8	56.8	100.0
Amhara	988	39.7	59.8	100.0
Oromia	1863	46.2	53.4	100.0
Somali	99	43.4	47.5	100.0
Benshangul Gumuz	100	63.0	36.0	100.0
SNNPR	847	24.6	73.1	100.0
Dire Dawa	49	79.6	18.4	100.0
Harari	51	58.8	37.3	100.0
Total	4534	38.8	59.0	100.0

In an attempt to rehabilitate the degraded land so as to restore soil fertility, 66% of the households have used some land management practices in the 2003/04 cropping season on their farms. The commonly applied land management practices (data not shown) included, crop rotation and intercropping (28%), application of cow dung (manure) (22%), application of inorganic fertilizer (21%) and terracing (16%). Still the practices of tree planting and agro-forestry were limited, amounting to only about 4%. Other practices such as ridges, contour plowing, etc. accounted for 9%. Fortunately, majority of the households (92%) continued using whatever land management practices they have started in the past, whereas the remaining (8%) have abandoned (Data not shown). Major practices abandoned (Data not shown) included crop rotation (42%), applying inorganic fertilizer and manure (29%), fallowing (13%, as well as terracing (9%). The remaining 7% included agro-forestry, mixed cropping, and ridges. Major reasons for abandoning some of the practices (data not shown) included expensiveness of fertilizers (27%) and intensive labor requirements (25%) of most of the natural resources conservation practices such as terracing and ridging.

4.1.2 Irrigation and water harvesting

For majority of the households (97%), the main water source for belg crop production was found to be rain-fed and those using irrigation (river, water harvesting and dam) accounted for only 3%. Similarly, for majority of the households (98%), the main water source for meher crop production was found to be rain-fed and those using irrigation (river, water harvesting and dam) accounted for only 2%. Food insecurity is caused, in part, by shortages and the unreliability of seasonal rains. Water harvesting, when available for use in times of scarcity, will greatly contribute to food security. However, despite the proclaimed water harvesting in the media, only 32% of the interviewed households have engaged in water harvesting schemes during the period 2002/03-2003/04 (Table 4.2).

Table 4.2. Percentage of households participated in water harvesting during the period 2002/03-2003/04

Region	Sample	Participated	Non-participated	Total
Tigray	349	45.3	54.7	100.0
Afar	192	9.9	90.1	100.0
Amhara	989	26.6	73.4	100.0
Oromia	1861	36.5	63.5	100.0
Somali	97	10.3	89.7	100.0
Benshangul Gumuz	99	6.1	93.9	100.0
SNNPR	848	31.0	69.0	100.0
Dire Dawa	49	55.1	44.9	100.0
Harari	51	49.0	51.0	100.0
Total	4535	32.0	68.0	100.0

Probably believing in the value of water harvesting, 53% of the households have a plan to construct some sort of water harvesting structure, whereas still 47% of the households do not have such a plan in the future (Table 4.3).

Those households who did not engage in constructing water harvesting structures cited the following reasons: some did not want any water harvesting (28%), some lacked water catchments (21%), some argued that water harvesting is not profitable (13%), some lacked the required skill (11%), some speculated problems related to rapid accumulation of silt (7%), some lacked the required materials (6%) and some fear wastage of water due to evapo-transpiration (6%). The remaining 8% pointed out lack of suitable land digging hole for water harvesting, lack of capacity due to old age, fear of disease development such as malaria, etc. More than one-half of the households (57%) who harvested water responded that their structures somehow held water, whereas for 43% of the households the structures could not hold water (Table 4.4). Earlier evaluation reports of water harvesting in Tigray, Amhara and Oromia showed the problem of design, shortage of construction materials, leakage, siltation, and lack of participatory approach (Oromia Bureau of Agriculture, 2003; UNDP/OCHA, 2003). The same evaluation reports indicated that water harvesting is beneficial wherever appropriately applied.

Table 4.3.Whether households have any plan to construct any one type of water harvesting structures in the future (%)

Region	Sample	Yes	No	Total
Tigray	172	38.4	61.6	100.0
Afar	90	64.4	35.6	100.0
Amhara	657	51.4	48.6	100.0
Oromia	1,083	55.6	44.4	100.0
Somali	76	89.5	10.5	100.0
Benshangul Gumuz	85	57.6	42.4	100.0
SNNPR	506	46.8	53.2	100.0
Dire Dawa	20	65.0	35.0	100.0
Harari	24	66.7	33.3	100.0
Total	2,713	53.3	46.7	100.0

Table 4.4. Percentage of households for which the structure held water

Region	Sample	Yes	No	Total
Tigray	153	74.5	25.5	100.0
Afar	16	56.2	43.8	100.0
Amhara	259	70.3	29.7	100.0
Oromia	662	49.5	50.5	100.0
Somali	9	100.0	0.0	100.0
Benshangul Gumuz	6	50.0	50.0	100.0
SNNPR	257	52.9	47.1	100.0
Dire Dawa	27	40.7	59.3	100.0
Harari	24	54.2	45.8	100.0
Total	1413	57.0	43.0	100.0

Close to two-third of the households (62%) who harvested water reported to have used the water, whereas 38% of the households did not use the water (Table 4.5). About two-third (64%) of those households who used the harvested water used for irrigating garden crops such as vegetables, while 36% used for other purposes (Table 4.6).

Results of the present showed that more than one-half of the households (54.5%) have faced problem during the construction of the water structures or the use of the water, whereas 45.5% reported facing no problem (Table 4.7).

Table 4.5. Percentage of households that used the harvested and stored water for some purposes

Region	Sample size	Yes	No	Total
Tigray	84	73.7	26.3	100.0
Afar	9	66.7	33.3	100.0
Amhara	180	49.4	50.6	100.0
Oromia	324	61.7	38.3	100.0
Somali	9	100.0	0.0	100.0
Benshangul Gumuz	3	33.3	66.7	100.0
SNNPR	136	58.8	41.2	100.0
Dire Dawa	11	100.0	0.0	100.0
Harari	13	100.0	0.0	100.0
Total	799	61.7	38.3	100.0

Table 4.6. Purposes for which the harvested water was used (%)
Region	Sample	For planting	For other purpose	Total
Tigray	60	75.0	25.0	100.0
Afar	4	75.0	25.0	100.0
Amhara	68	58.8	41.2	100.0
Oromia	136	58.1	41.9	100.0
Somali	6	50.0	50.0	100.0
Benshangul Gumuz	1	100.0	0.0	100.0
SNNPR	57	70.2	29.8	100.0
Dire Dawa	10	50.0	50.0	100.0
Harari	12	100.0	0.0	100.0
Total	354	64.4	35.6	100.0

The major problems encountered during the construction of the water harvesting structure included failure of water holding structures to be full being built on area without water catchments (58%), shortage of the required materials (24%) and inadequate skill and associated problems (18%) (Table 4.8). In responding whether the households have engaged in other small irrigation activities like river diversion, use of spring or well water, only 19.5% responded positively, whereas a large proportion of the households indicated that they did not engage in such endeavors (Table 4.9).

Table 4.7. Percentage of households who faced problem during construction of water harvesting structures

Region	Sample	Yes	No	Total
Tigray	146	45.9	54.1	100.0
Afar	13	38.5	61.5	100.0
Amhara	245	29.8	70.2	100.0
Oromia	641	73.9	26.1	100.0
Somali	7	71.4	28.6	100.0
Benshangul Gumuz	5	0.0	100.0	100.0
SNNPR	240	27.5	72.5	100.0
Dire Dawa	25	100.0	0.0	100.0
Harari	25	76.0	24.0	100.0
Total	1347	54.5	45.5	100.0

Table 4.8. Major problems encountered by the households during the construction of the water structure or use of the water (%)

Region	Sample	Poor water holding	Shortage of the required materials	Skill & others	Total
Tigray	50	60.0	22.0	18.0	100.0
Amhara	172	62.8	21.5	15.7	100.0
Oromia	483	58.2	24.0	17.8	100.0
Somali	38	44.7	23.7	31.6	100.0
Benshangul Gumuz	50	50.0	40.0	10.0	100.0
SNNPRs	280	58.2	21.4	20.4	100.0
Harari	20	60.0	25.0	15.0	100.0
Total	1093	58.2	23.6	18.2	100.0

Table 4.9. Percentage of households who have been engaged in other small irrigation activities like river diversion, use of spring or well water
Region	Sample size	Yes	No	Total
Tigray	339	34.5	65.5	100.0
Afar	168	46.4	53.6	100.0
Amhara	954	25.9	74.1	100.0
Oromia	1768	16.1	83.9	100.0
Somali	80	1.2	98.8	100.0
Benshangul Gumuz	96	28.1	71.9	100.0
SNNPR	820	9.0	91.0	100.0
Dire Dawa	49	26.5	73.5	100.0
Harari	46	2.2	97.8	100.0
Total	4320	19.5	80.5	100.0

The most preferred irrigation types included small irrigation (36%), river diversion (32%) and water harvesting (29%), whereas development of wells was least preferred (Table 4.10).

Table 4.10. Major types of irrigation used by the sampled households (%)

Region	Sample	Water harvesting	Small irrigation	River diversion	Wells development	Total
Tigray	50	20.0	42.0	36.0	2.0	100.0
Amhara	154	22.1	48.1	25.3	2.6	100.0
Oromia	314	35.7	30.6	31.8	0.6	100.0
Somali	8	0.0	0.0	100.0	0.0	100.0
SNNPR	198	28.3	33.8	35.4	0.0	100.0
Harari	20	30.0	50.0	20.0	0.0	100.0
Total	744	29.3	36.0	32.1	0.9	100.0

5. Conclusions and Recommendations

One of the major policy agenda of Ethiopia is to achieve food security while reducing environmental degradation and the depletion of the natural resource bases. During the past one-decade or so, the government has given top priority for agriculture and rural development so as to increase smallholder farm production and productivity. For this purpose the PADETES has been pursued since 1995. Agricultural extension, as an educational and communication tool, makes a vital contribution to agricultural production and rural development.

Similarly, the results of present study showed that there is increment in production, whenever extension packages are used, confirming previous demonstration and adoption studies. Yet, there is a problem of sustainability and extending the scope of the extension package in Ethiopia. Results show no more than 56% of the sampled households used any one modern inputs (improved seeds, fertilizers, etc) and improved crop production practices. In certain cases, farmers were noted to regress back to their own old practices and no sustainable use of modern agricultural inputs such as fertilizers and improved seeds. It is recommended that agricultural extension services be more enhanced in terms of making more and more resource poor farmers to participate in the extension package, delivering alternative packages, reducing development agents to farmers ratio, increasing input supply, increase access to credit and improving input and output distribution and marketing. Moreover, the present study showed that the percentage of new entrants into the extension package is low, ranging from 4% in 1994/95 to 16% in 2002/03, suggesting low number of graduates as well. However, certain agricultural extension professionals feel that achieving 10% involvement of farmers in extension is quite sufficient, provided this is followed by effective inputs supply and outputs marketing. Input suppliers should become more responsive to the needs of farmers graduating from the extension package by making available the necessary quantity and quality of inputs such as improved seeds, fertilizers, pesticides, etc. at the right time and for affordable prices. Individual contacts between extension agents and farmers have to be expanded, implying the need for training and deploying more number of extension agents.

One of the chief impediments to the use of modern agricultural inputs is the poor distribution channels as well as poor market for surplus production. As a solution to this there is a strong need to enhance the process of upgrading of rural transport in association with appropriate market development both for agricultural inputs and outputs.

Improved seed is one of the primary requirements for enhancing crop productivity and increasing income generation. Seed security should also come before ensuring food security to farmers. However, the results of the present study, in agreement with several previous studies, revealed that only about 8% of the participating households used improved seeds and over one-half (54%) of the participating households used own saved seed, suggesting annual fresh seed purchase to be low; still 30% the participating households were not replacing seeds within the acceptable period of 1-4 years. Further disturbing case is that only about 39% of the participating households used complete package (improved seed + fertilizer + improved cultural practices). Incomplete package is not as profitable as complete package. The levels of participation in high value crops and natural resources management were also found to be low in the present study. The amounts of inorganic fertilizer used for major crops were found to be about one quintal per hectare. Ensuring Ethiopian farmers access to quality improved seed timely and at a reasonable price can only be achieved if there is a vibrant seed supply system to multiply and distribute the seeds and if mechanisms of getting effective credit is established. It seems that the formal seed system through public and private seed sectors could not significantly do the job in this regard. It is recommendable if farmers based-seed production and distribution mechanism is enhanced; with cooperatives/unions playing a major role in distributing and marketing of improved seeds that have specific adaptation and not handled by seed enterprises. Formal seed system is usually profit-oriented and would like to operate on principle of economy of scale and deal with a narrow range of crop varieties. However, given that the agro-ecologies and farming system diversity of the country no single variety will occupy a wide area of cultivation. Seed production, distribution and marketing for specific areas would seem to be possible only under farmers based-seed production and marketing system. Farmers based seed production and dissemination should receive more priority than formal seed sector in view of the fact that no single variety will have wide adaptation for economically profitable seed production and distribution by the formal sector. The formal seed sector tends to focus on a few crops such as wheat and maize and on a few varieties for which there is a large market, giving less attention to other crops, especially pulses, oilseeds and horticultural crops. Farmers based seed production and distribution is also suitable for Ethiopian rural condition where transportation and other rural infrastructures are poorly developed. As local demand for seeds of many improved varieties of crops takes off, farmers will gain new economic opportunities in small-scale seed production. Past experiences in Ethiopia and other African countries such as Zimbabwe, Mali, and Senegal have shown that, with adequate support, smallholder farmers are capable of producing high quality seed of improved varieties. Extension service is expected to develop training materials and offer training on principles and procedures of operating small-scale seed production. Such practice would generate income for the farmers and greatly promote the adoption of improved varieties. Wide adoption of improved varieties in turn ensures food security. The role of private seed producers should not be undermined. However, Ethiopia’s dispersed and underdeveloped markets, poor infrastructure, and low farmer income present obvious challenges to the expansion of private seed markets profitably.

Cooperatives and unions would also play a major role in seeds, fertilizers and pesticides distribution and marketing in the sense that they will have more number of marketing outlets in remote rural areas than the already known companies in this regard. This presumption of more number of distribution and marketing outlets for cooperatives would solve the lack of demand-supply determination in the use of agricultural inputs and the carry-over that the companies are currently facing. Carry-over, for both seeds and fertilizers, is because of poor distribution channels and inadequate demonstration and promotion, rather than due to lack of demand in real sense.

Many agricultural research centers have been established to assist in development efforts by way of developing and disseminating research-based technologies. Yet, there are many agriculturally potential areas and moisture stress areas that did not receive adequate research services. Since agro-ecologies and farming systems are highly variable in Ethiopia, the type of crops grown, the technological input requirements and the major objectives also differ within Woredas, zones, regions and the country as a whole. Hence, the agricultural research system is required to extend its research activities into uncovered areas. All agricultural research centers should also have formal linkage with bureaus of agricultural and rural development of their mandate Woredas and farmers training centers.

In the present study, diseases and insect pests have been rated as major production constraints by farmers. As a solution, diversification of production system is imperative. The diversified production system should involve crop rotation, frequent replacement of varieties of the same crop and effective use of integrated disease, insect pests and weed management. Crop rotation, involving cereals, pulses, oilseeds and horticultural crops, must be practiced to reduce disease, insect and weed incidences, to improve soil fertility and to enhance productivity. Crop rotation is part of a sustainable cropping system, which is characterized by efficient use of nutrients through recycling of nutrients in crop residues, minimizing leaching losses, promoting biological nitrogen fixation and low levels of fertilizer inputs. The present study revealed that high fertilizer price is one of the constraints in using fertilizers and they also apply less than required without also taking into account annual nutrient removal. Practices such as crop rotation, application of green manure and organic fertilizers may compensate for nutrients removed by the previous crop. It is to be noted that organic wastes, manure and biological nitrogen fixation are among the principal sources of nitrogen, which is the most deficient in almost all soils but highly essential for plant growth and development. Conservation-oriented cropping systems such as agro-forestry should be promoted for not only land conservation but also for fodder for livestock and fuel supply. Soil-test based fertilizer recommendation is exceedingly required through strengthening the capacity of the 17 regional soil laboratories.

The use of BBM should substantially be increased as it improves the drainage problem of vertisols. Vertisols are the most important soils for agriculture next to nitosols and cambisols in Ethiopia. However, the agricultural potential of vertisols cannot be fully utilized unless the drainage problem is improved by using such techniques of drainage using BBM. The widespread use of BBM means improved drainage of black soils, which in turn would enhance the practice of double or multiple cropping patterns taking the advantage of the long growing season available in highland black soils. The use of BBM in the lowland black soils, if coupled with irrigation, would contribute greatly to the increase of the area under irrigation in the country. Similarly, the use of Tier-ridger should substantially increase in moisture stress areas as means of soil and water conservation practice.

Developing location or agro-ecology specific technologies and effective utilization of such technologies by farmers should also accord due attention. This is because the results from the present study showed that one of the limitations in adoption of technologies is the poor adaptation of the technologies to the specific environmental settings of the farmers. The poor adaptation is manifested biologically, socially and economically. Post harvest losses should also be reduced by adapting simple storage and processing technologies. Extending extension services more and more into the lowlands (moisture stressed arid and semi arid areas) by of making available improved drought tolerant crops, increased access to credit, increased small scale irrigation schemes and enhanced natural resources conservation and utilization so as to ensuring food security.

The extent of irrigation was found to be 2% in meher and 3% in belg, indicating that Ethiopia has very low irrigated agriculture. Among the smallholders agriculture, uncertainty of rainfall is one of the major causes for crop failures. Small irrigation projects and effective water harvesting and its utilization should, therefore, take top priority to ensure sustainable extension services in moisture stress areas. It is estimated that Ethiopia’s mean annual rainfall reaches about 1090 mm. However, 70% of the total arable land receives annual rainfall of less than 750 mm, while about 110 billion cubic meters of rainwater annually is lost through surface runoff. Harvesting and utilizing such water should be enhanced by participating farmers from beginning to end. Use of harvested rainwater could supplement natural rainfall and makes possible the production of garden vegetables, which in turn could improve the nutrition and generates income of the family. Problems associated with water harvesting such as flaws in the design of the structures, wrong site selection, leakage and evaporation, lack of effective planning, non-participatory and quota systems should be solved positively to help farmers develop confidence in the practice of water harvesting and utilization. Possibility of using drip irrigation with water harvesting should also be explored in order to use the harvested water efficiently.

It is generally believed that sufficient technology and information have been generated by the national research system during the past 50 years to make a substantial impact on productivity in highland potential areas at least. However, the smallholder farmers have not benefited much from such technology and information to the required extent because of weak linkage between research and extension. Thus it is strongly recommended to strengthen the linkage by improving and making responsible research extension advisory council and expanding farmers' research group, farmers' extension group and forming a formal linkage with farmers training centers for demonstration of emerging research-based technologies. Such linkages should make concerted effort towards realizing the potential of improved technology and information that are being accumulating at the research centers.

On the whole, in Ethiopia, agriculture is highly influenced by external factors like weather, which, coupled with low input, poses a significant challenge in realizing the government’s Agricultural Development Led Industrialization (ADLI) strategy designed to bring about rapid growth and transformation in socioeconomic life of the people. After analyzing GDP and agriculture growth rates in Ethiopia for the period 1981 to 1997, Devereux (2000) concluded that the entire Ethiopian economy is dependent on low productivity of rainfed agriculture and rainfall is the single most important determinant of Ethiopia’s economic success or failure from year to year. Annual report of the National Bank of Ethiopia for the Ethiopian Fiscal Year 2002/03 also revealed that the economic performance is dependent on agriculture, which in turn shows upward or downward growth depending on rainfall. Many lines of evidence indicate that Ethiopian agriculture has derived more of its growth from expansion of arable land and good patterns of rainfall than from increased use of modern technological inputs that would lead to intensification of agriculture.

The present study and review of past studies by Degefe and Nega (2000), implied that for longer term food security in Ethiopia there should be promotion of wide scale adoption of research-based technologies and fertilizers to increase crop productivity, which should be accompanied by improved infrastructure (rural roads, small and medium irrigation schemes, rural electrification) and marketing of inputs and outputs (including market sites, storage facilities). However, given the inherent vulnerability of the agricultural sector, due attention should be given to significantly increasing growth in other sectors such as agro-industry, income diversification by way of off-farm activities, community-based tourism, and other services.

6. Implications for Research and Extension Systems

The following implications are drawn for policy makers as well as research and extension systems of Ethiopia based on the results of the present studies and review of related literature.

Increase the level of awareness of the farmers about existing technologies through demonstration, adaptive trials, field days, media, extension manuals and so forth. Despite the availability of a large number of recently released varieties having better yield and resistance than the older varieties, seed multiplication agencies and farmers are using older varieties. Farmers research groups, farmers extension groups, farmers training centers, and research-extension farmers advisory council should play a decisive role in ensuring that research and extension activities of the country are based on real farmers problems and solving local problems. The level and role of farmers’ participation in planning and implementation of research and extension programs should be enhanced so that they develop confidence and eventually adopt the emerging technology. The agricultural TVET should continue in producing more and more number of competent and skilled DAs.
Increase the amount of improved seeds (including seedlings and cuttings used for propagation) produced for all categories of crops (all cereals, pulses, oilseeds, horticultural crops and industrial) rather than concentrating on few varieties of a limited number of crops. However, the capacities of the existing seed multiplying agencies (ESE, Pioneer Hybrid International, research centers, NGOs and private farms) are limited. Similarly, the distribution coverage of such organizations is narrow. Organizing seed producing cooperatives for different crops in potential agro-ecologies is suggested as solution. The research system should make available basic seeds of high quality and extension should give proper services for such cooperatives.
Enhance the development of country’s biotechnology capacity so that such techniques like tissue culture could be used for rapid and disease free multiplication of enset, hybrid coffee, fruit and vegetable crops
Develop location specific (based on soil type, moisture regime and inherent soil fertility status) fertilizer rates and types and promote wide scale adoption.
Enhance sustainable cropping system (crop rotation, inter-cropping, double cropping, rely cropping) instead of mono-cropping so as to improve soil fertility and reduce the building up of diseases, insect pests and weeds.
Improve greatly input and output marketing system by improving the capacity of cooperatives and improving rural infrastructure (roads, storage structures). Farmers should obtain the right type of inputs in sufficient amount and timely.
Improve research-extension-farmers-agro-processing industries. Here the role of each must be defined and formalized, at least for such crops like bread and durum wheats, malt barley, canning beans and soybean.
Reduce post harvest losses by improving storage structures and processing so as to add value.
Develop more number of technologies and disseminate for farmers in the moisture prone areas (arid and semi-arid). This involves opening more number of research and testing sites in such areas. Technology development in such areas should focus on integrated technology development and transfer instead of going for piecemeal.
Give more research and extension focus on high value crops such as pulses, oilseeds and horticultural crops, as the present focus is more on cereals.
Develop more number of irrigation schemes (small, medium and large) to greatly supplement and complement the dominant rainfed agriculture.
Expand GIS support to research and extension system to map locations of similar growing conditions in terms of soil types, climatic factors and socio-economic factors.

References

Abesha, D., Waktola, A. and Aune, J.B. (2000). Agricultural Extension in the dry lands of Ethiopia, report No. 9. Addis Ababa, Ethiopia. Mimeographed, Pp. 39.

Addis Tribune June2-July 15, 1995

Afri-Consultants (1998). Seed marketing study, Vol. I and Vol. II, a consultancy study for NSIA, Addis Ababa, Ethiopia.

Alemu, Z.G. (2002). How successful the agricultural development led industrialization strategy (ADLI) will be by leaving the existing landholding system intact? Economic Focus 4(4): 16-22. Jan-Feb. Bulletin of the Ethiopian Economic Association.

Amha, W. (1999). Improved seed marketing and adoption in Ethiopia. Ethiopian Journal of Agricultural Economics Vol.3 (1): 41-87.

Ayalew, T. (1998). Food security in Ethiopia, paper presented to at the CTA conference: Food insecurity in ACP countries: from policy and program interventions to vulnerable group. Leuven, Belgium: CTA 26-30 October 1998.

Bako Agricultural Research Center (2005). Annual Report for 1997 E.C. fiscal year.

Bale Agricultural Development Enterprise (2003). Proceedings of the Agronomy Workshop, March 20-21, 2000. Addis Ababa, Ethiopia.

Bhagavan, M.R. (1989). Ethiopia: development of scientific and technological research and SAREC’s support 1979-1988. SAREC, Sweden, Stockholm.

Bishaw, Z.J. (2004). Wheat and barley seed systems in Ethiopia and Syria. Summary, WU dissertation No. 3575. Pp. 3.

Broeck, K.V.D. and Dercon, S. (2001). Agricultural production changes in Ethiopia. Mimeographed, Pp. 22.

Broeck, K.V.D., and Dercon, S. (2001). Agricultural production changes in Ethiopia. Addis Ababa, Ethiopia.

CRDA (2000). Agriculture and natural resources panel, proceedings of a workshop on NGOs and sectoral development, CRDA, Addis Ababa, Ethiopia. Mimeographed, Pp. 20.

Degefe, B. and Nega, B. (2000). Annual report on the Ethiopian Economy (1999/2000). Ethiopian Economic Association, Addis Ababa, Ethiopia.

Demeke, M. (1996). Constraints to efficient and sustainable use of fertilizer in Ethiopia. In: Sustainable intensification of agriculture in Ethiopia, Proceedings of the 2^nd conference of the Agricultural Economics Society of Ethiopia, 3-4 October 1996. Pp. 242-266.

Dercon, S. (1999). Ethiopian poverty assessment study. Addis Ababa, Ethiopia.

Devereux, S. (2000). Food insecurity in Ethiopia, a discussion paper for DFID. Mimeographed, Pp. 1-16.

EARO (Ethiopian Agricultural research Organization) (2000). Seed Research and Multiplication strategy. Addis Ababa, Ethiopia.

Fantahun, A. and Minuye, S. (2000). Constraints to agricultural input loan repayment. Economic Focus 3(2): 25-27.

FAO (2004). The Ethiopian agricultural situation and a framework for Ethiopia-FAO collaboration.

FAO/WFP (2001). Crop and food supply assessment mission to Ethiopia. Addis Ababa, Ethiopia for 2001.

FDRE (2002). Agricultural sector review, agricultural and rural development department ONAR.1.

Ferede, S., Verkijl, H., Tanner, D.G., and Gebre, T. (2000). Optimizing fertilizer use in Ethiopia: Results of a baseline survey of fertilizer adoption in Arsi Zone. SG2000, Addis Ababa, Ethiopia.

Gebeyehu, G., Dabi, G. and Shaka, G. (2001). Focus on seed programs: The Ethiopian seed. Mimeographed, Pp. 22.

Glennon, J. and Garraty, J.A. (1999). The 20^th century, an illustrated history of our lives and times.

Guinand, Y.F. (2002). Livelihood disruption in cash crop and surplus producing areas, consequences of persistent low cereal market prices in Ethiopia. UN-EUE situation analysis, July, Addis Ababa, Ethiopia

Halie, A., Negatu, W., and Retu, B. (1991). Technology transfer for wheat production in Ethiopia. In: Wheat Research in Ethiopia: A historical Perspective. IAR/CIMMYT, Addis Ababa, Ethiopia.

Howard, J., Demeke, M., Kelly, V., Maredia, M., and Stepanek, J. (1998). Can the momentum be sustained? An economic analysis of the Ministry of Agriculture/Sasakawa Global 2000’s experiment with improved cereal technology in Ethiopia. Department of Agricultural Economics, Addis Ababa University. Addis Ababa, Ethiopia. Mimeographed, Pp. 36.

Kahsay, M. and Kugbei, S. (1998). Alternative Seed supply Systems in Ethiopia. Ethiopian Seed Enterprise, Addis Ababa, Ethiopia. Mimeographed, Pp. 4.

Kassa, B. (2003). Agricultural extension in Ethiopia. The case of participatory demonstrations and training extension system. J. Social Development in Africa 18(1):49-83.

Kassa, B. and Abebaw, D. (2004). Challenges facing agricultural extension agents: A case study from southwestern Ethiopia. African Development Review 16(1): 139-168.

Kugbei, S. and Fikru, A (1997). The injera initiative. Caravan 6:22-25. ICARDA, Aleppo, Syria.

Lemessa, D. (2001). Filed assessment report: Maize price crisis badly hits farmers in western Ethiopia. UN-EUE situation analysis, June, Addis Ababa, Ethiopia.

Library of Congress-Federal Research Division (2005). Country Profile: Ethiopia. Pp.10.

Marena, G.C. (1996). Agricultural development in Sub-Saharan Africa: An engineer paper presented to at a regional workshop on agricultural mechanization in Africa, September 30-October 1, 1996. Arusha, Tanzania.

Ministry of Agriculture (1987). Soil conservation research project progress report, Addis Ababa, Ethiopia.

MOFED (2002). Ethiopia: Sustainable development and poverty reduction program. Addis Ababa, Ethiopia.

Mohammed, I and Terfa, T. (2001). Maize technologies: Experience of the Ministry of Agriculture, Proceedings of the 2^nd National Maize Workshop of Ethiopia, 12-16 Nov., 2001.

MOLSA (Ministry of Labor and Social Affairs) (1997). Agricultural wage employment and rural non-farm income in Ethiopia, survey results.

Mulatu, E. and Belete, K. (2001). Participatory selection in lowland sorghum in eastern Ethiopia: Impact on adoption and genetic diversity.

Mulatu, E. and Zelleke, H. (2001). Farmers’ highland maize (Zea may L.) selection criteria: Implications for maize breeding for the Hararghe highlands of eastern Ethiopia.

Nichola, T. (1985). Agricultural research and extension in Ethiopia: the state of the art, IDR research report No.22. , Institute of Development Research, Addis Ababa University. Addis Ababa, Ethiopia.

NSIA (1997). Implementation plan for farmer-based seed multiplication and marketing scheme. NSIA, Addis Ababa, Ethiopia.

Oromia Bureau of Agriculture (2003). Evaluation of water harvesting and usage in Oromia, assessment report. Addis Ababa, Ethiopia.

RATES (2003). Maize market assessment and baseline study for Ethiopia, Center for Regional Agricultural Trade Expansion Support, Nairobi, Kenya.

Raymakers, B. (2002). Review and consequences of reduction in agriculture in put sales in 2002, a situation analysis. UN-EUE, Addis Ababa, Ethiopia.

SG2000 (1996). SG2000 Agricultural project in Ethiopia, 1996, annual report for 1995 crop seasons. Addis Ababa, Ethiopia.

SG2000 (2002). SG2000/Ethiopia Project, activities and outputs: An assessment 1993-2001. Addis Ababa, Ethiopia.

SIDA (2003). Ethiopia: Country analysis. Addis Ababa, Ethiopia.

Tafesse, K. (2002). Towards seed industry development in Ethiopia, National seed Industry agency of Ethiopia. Mimeographed, Pp. 6.

UNDP/OCHA (2003). Ponds filled with challenges: water harvesting-experiences in Amhara and Tigray, assessment mission. Addis Ababa, Ethiopia.

USAID (2004). Integrated Strategic Plan for Ethiopia: Fiscal Year 2004-2008. Mimeographed, Pp. 38.

Waktola, A. (1980). Assessment of the diffusion and adoption of agricultural technologies in Chilalo. Eth. J. Agric. Sc. 2(1):51-68.

Wayne, D. (1922). The efficient use of animal power. Agricultural engineering 21(4): 23-27.

World Bank (1995). Ethiopia: National fertilizer project, staff appraisal report.

World Bank (2004). Ethiopian Rural Capacity Building Project, report No.: AB606

Zegeye, T., Tadesse, B. and Tesfaye, S. (2001a). Determinants of high yielding maize technology adoption: Empirical evidence from southwestern Oromiya, Research Report No. 38, EARO, Addis Ababa, Ethiopia.

Zegeye, T., Taye, G., Tanner, D., Verkuijl, H., Agidie, A., and Mwangi, W. (2001b). Adoption of improved bread wheat varieties and inorganic fertilizer by small scale farmers in Yelmana Densa and Fata districts of northwestern Ethiopia.

Annexes

Annex 2.1. Frequency of sampled households by region zone and woreda

Region	Zone	Woreda	Frequency	Percent
Afar	Zone 1 (Afar)	Asaita	54	1.18
	Zone 1 (Afar)	Chifra	46	1.00
	Zone 3 (Afar)	Amibarra	50	1.09
	Zone 3 (Afar)	Awash Fentale	50	1.09
Subtotal			200	4.36
Amhara	Awi	Dangla	49	1.07
	E.Gojam	Baso Liben	49	1.07
		Dejen	51	1.11
		Enemay	49	1.07
	N. Shewa (Amhara)	Debre Brehan Zuria	50	1.09
		Insarona Wayu	45	0.98
		Gonder Zuria	51	1.11
	N.Gonder	Wegera	49	1.07
	N.Wello	Habru	50	1.09
		Kobbo	49	1.07
		Meket	51	1.11
	S.Wello	Ambassel	51	1.11
		Dessie Zuria	50	1.09
		Legambo	51	1.11
		Were Illu	49	1.07
	W.Gojam	Achefer	50	1.09
		Bahir Dar	50	1.09
		Bure Wonberma	50	1.09
		Merawi	50	1.09
	Wag himera	Ziquala	50	1.09
Subtotal			994	21.67
Benshangul Gumuz	Assosa	Assosa	50	1.09
Benshangul Gumuz	Metekel	Pawe	50	1.09
Subtotal			100	2.18
Dire Dawa	DireDawa	Diredawa-Zuria	50	1.09
Subtotal			50	1.09
Harari	Harari	Harari	48	1.05
Harari	Harari	Harari	2	0.04
Subtotal			50	1.09

Annex 2.1. Continued.

Region	Zone	Wereda	Frequency	Percent
Oromia	Arsi	Gedeb	53	1.16
		Hitosa	48	1.05
		Kofele	49	1.07
		Limu Bilbilio	48	1.05
		Tiyo	51	1.11
	Bale	Adaba	50	1.09
	Bale	Sinanana Dinsho	50	1.09
	Borena	Moyale	51	1.11
	E.Hararge	Alemaya	49	1.07
		Babile	51	1.11
		Deder	50	1.09
		Fedis	49	1.07
	E.Shewa	Ada'a Lome	50	1.09
		Akaki	51	1.11
		Dugda Bora	48	1.05
		Gimbichu	50	1.09
		Shashamene	51	1.11
	E.wellega	Horo	50	1.09
	Guji	Liben	50	1.09
	Illubabor	Metu	49	1.07
	Jimma	Dedo	50	1.09
		Goma	50	1.09
		Kersa	50	1.09
	N.Shewa (Oromia)	Abichuna Gne'a	49	1.07
		Degem	45	0.98
		Gerar Jarso	55	1.20
		Wuchalena Jedu	54	1.18
	W. Shewa	Ada Berga	51	1.11
		Alem Gena	50	1.09
		Ambo	49	1.07
		Bacho	50	1.09
		Bako Tibe	39	0.85
		Welisona Goro	50	1.09
		Welmera	50	1.09
	W.Hararge	Alemaya	1	0.02
		Kuni	50	1.09
		Mieso	50	1.09
	W.Welega	Begi	50	1.09
	W.Welega	Gimbi	50	1.09
Subtotal			1891	41.23

Annex 2.1. Continued.

Region	Zone	Wereda	Frequency	Percent
SNNPRS	Alaba	Alaba	50	1.09
	Amaro	Amaro	50	1.09
	Bench Maji	Meinit	49	1.07
	Gedio	Bule	47	1.02
		Wenago	53	1.16
		Yirga Chefe	50	1.09
	Gurage	Gumer	48	1.05
	Hadiya	Badawoch	50	1.09
		Misha	50	1.09
		Soro	50	1.09
	Keficho Shekicho	Chena	51	1.11
	Kembata Alaba T	Angacha	51	1.11
	Sidama	Aleta Wondo	50	1.09
	Sidama	AWASSA (Incl. Wendogenet)	50	1.09
	Silte	Silti	50	1.09
	wollaita	Boloso Sorie	53	1.16
	wollaita	Sodo Zuria	50	1.09
Subtotal			852	18.57
Somali	Jijiga	Jijiga	50	1.09
Somali	Jijiga	Kebre Beya	50	1.09
Subtotal			100	2.18
Tigray	C.Tigray	Adwa	50	1.09
	E.Tigray	Awlalo	50	1.09
	E.Tigray	Hawzien	50	1.09
	S.Tigray	Alamata	50	1.09
		Enda Mekhoni	50	1.09
		Raya Azebo	50	1.09
	W.Tigray	Kafta Humera	50	1.09
Subtotal			350	7.63
Total			4587	100.00

Ethiopian Economic Association / Ethiopian Economic Policy Research Institute

Assessment of Crop Production and Natural Resources Management Extension Packages in Ethiopia

(1994/95-2003/04)

Issue Paper Series

No. 01/2006

December 2006

Addis Ababa

Table contents

List of Tables

List of figures

List of Annexes

Abbreviations and Acronyms