Adoption of Improved Maize Technologies and Inorganic Fertilizer In Northwestern

By Tesfaye Zegeye ail'd Alemu Haileye

Research Report No. 40

Ethiopian Agricultural Research Organization Adoption of Improved Maize Technologies and Inorganic Fertilizer In Northwestern Ethiopia

By Tesfaye Zegeye and Alemu Haileye

Research Report No. 40

§ Ethiopian Agricultural Research organization (EARD), 2QDI

E-mail: [email protected] Fax: 251-1-461294 Tel - 25M -4G 2633 P.O. Box: 2D03 Addis Abeba. Ethiopia

Copycditing. Matebu Tadesse and Abebe Kirub Design: Abebe Kirub Contents

Acronyms 1 Acknowledgments 2 Summary' 3

Introduction 5 Background and rationale * 5 Objectives of the study 6

Description of the Study Area 7 Amhara National Regional State 7 Northwestern Amhara 8

Maize Production, Technology Development and Transfer 10 Production 10 Technolog}'Development 12 Review of Extension Systems 14 Seed Production 16

M ethodology 19 Sampling procedure 19 Data Collection 19 Analytical procedure 19

Demographic, Socio-economic and Institutional Characteristics 23 Demographic characteristics 23 Socio-economic Characteristics 24 Institutional Characteristics 28

Maize Production Management 32 Land Preparation and planting 32 W eeds 33 Pest and Disease Management 34 Post-harvcst Management 36 Rate of Adoption of Improved Maize and Chemical Fertilizer 38 Adoption Rate of Improved Maize 38 Adoption Rate of Chemical Fertilizer 39

Factors Affecting Adoption of Maize Technologies 42 Logit Analysis of Improved Maize 42 Logit Analysis on the I -se of Chemical Fertilizer 44

Conclusion and Recommendations 48

References 5 1 Adoption of improved maize and chemical fertilizer 1

Acronyms

CDE: Center for Development ;uid Environment CSA: Central Statistical Authority IAR: Institute of Agricultural Research EARO: Ethiopian Agricultural Research Organization MED AC: Ministry of Economic Development and Cooperation AUA: Alemaya University of Agriculture ESE: Ethiopian Seed Enterprise MOA: Ministry of Agriculture CADU: Chilalo Agricultural Development Unit WADU: Wolayita Agricultural Development BOA: Bureau of Agriculture IMV: Improved maize MOA: Ministry of Agriculture TLU: Tropical Livestock Unit CIMMYT: International Center for Maize and Wheal Research BOPED: Bureau of Planning and Economic Development MRS: Maize Research Strategy EPID: Extension Project Implementation Department MPP1: Minimum Package Program One MPP2: Minimum Package Program Two PADETS: Participatory Agricultural Demonstration and Training System T and V: Training and Visit SG-2000: Sakakawa Global-2000 PAS: Peasant Associations SPSS: Statistical Packages for Social Sciences 2 Tesfaye and Alemu

Acknowledgments

We are very grateful to the Ministry ot Economic Development for accepting and recognizing the importance of the project and the USAID for the financial assistance which led to the implementation and finalization of this study.

W e are also very grateful to the north Conder, west and east Gojam farmers who have spared their precious time to respond positively to the lengthy questionnaire- without which this document could have not been written. We would also like to thank the enumerators (mostly technical assistants) drawn from Adet Research Center who have tolerated the hardship of moving from one place to die odier and for filling in the lengthy questionnaire.

We are also very much indebted to the Ethiopian Agricultural Research Organization (EARO), and Adet research center for their assistance and unreservecT support. The Regional Bureau of agriculture of Amhara, Zonal and Woreda Agriculture Departments deserve special thanks for providing background inf ormation and for providing staff whenever there was a need.

W e would also like to thank Mr. Demeke Nigussie for extracting the two maps from CD_ROM entitled Soil Conservation Research Program Database File System. Adoption of improved maize and chemical fertilizer 3

Summary

The overall objective of the study is to investigate and document adoption levels and to specifically determine the factors that affect the adoption process of improved maize and draw implications for research, extension and policy. The districts selected for the study were Dera from South Gonder, Bure Wonbera, Achfer and Bahir Dai' Zuria from West Gojam and Machakel from East Gojam. The selection of the sample farmers involved a two-stage sampling procedure. The sample peasant associations (PAs) were selected randomly using random sampling procedure. Following the selection of the peasant associations, the sample fanners were tiien selected from sampling frame obtained from development centers and/or peasant association offices of die respective PAs using random sampling procedure. A total of 369 fanners were selected and included in the study.

Generally as far as the demographic characteristics of fanners are concerned, only level of education and family size were found to significandy influence die adoption decision of improved maize and chemical fertilizer. The odier demographic characteristics like age, fanning experience are not strong enough to influence farmers’ decision behavior in adopting improved maize.

The size of livestock units was hypothesized to affect the adoption decision of improved technologies, for they are good source of cash for purchase of farm inputs. In light of this, a significant difference was found in tropical livestock units between adopters and non-adopters (t=-2.054; p<0.05). Adopters have average livestock units of 4.76 while non-adopters have 3.87. A significant difference was found in number of oxen owned by adopters and non-adopters (t =-2.244; p<0.05). The average number of oxen owned by adopters is 2.07 while that owned by non-adopters is 1.80. The study also revealed that diere is a systematic association between adoption of improved maize and access to draft power (x*= 11.351; p<0.01).

The chi-square analysis shows diat access to extension message (tiirough eidier of such sources as field day, fonnal agricultural training and listening to agricultural programs on radio) is systematically associated with adoption of improved maize (x'=50.965; p<0.01). Thus it can be concluded diat farmers who have got information dirough either of die sources are more likely to adopt maize technologies than those farmers who did not have access.

In diis study, it w^as hypodiesized diat increasing access to credit can enhance the adoption of new technologies. In light of tliis, die chi-square analysis shows diat diere is a systematic association between adoption of improved maize and access to credit (X =137.635; p<0.01). The result shows diat fanners who have access to credit tend to adopt improved maize and chemical fertilizer than diose farmers who have no access to credit.

The adoption rate of improved maize has increased from less dian 1% in 1976 to 43 % in 1998. The adoption rate has increased noticeably over die last five years ever since die national extension package program was started. In 1995 when die program was started, it rose to 16 % and reached 43% in 1998. Though different varieties w'ere introduced in die study area, die varieties adopted by 58% of adopters are A- 4 Tesfaye and Alemu

511 and BH-660. The preferred varieties as reported by 37%, 29% and 21% of adopters are BH-660, A-511, and BH-.540 respectively.

Among tile 14 factors considered in the logit model, only six factors had significant and positive influence on adoption of improved maize at less than or equal to 5% probability level. These were applying chemit ;d fertilizer, access to credit, access to extension information, distance to development center, distance to market center and family size. Of these factors, contrary to expectation distance to development and market centers are positively related to the probability of' adoption of improved maize.

The probability of adopting improved maize among farmers with the average values of the continuous variables included in the model is about 5%. With the application of chemical fertilizer the probability of adopting improved maize increased to 49%. With access to credit and extension information, die probability that a farmer would adopt an improved maize would increase to 38% and 36%, respectively. Finally rise in family size, for instance from seven to eleven has increased the probability of adopting improved maize from 7.4% to 17.5%.

Among the fourteen factors postulated to influence the adoption decision of chemical fertilizer use, half of them have been found to be significant. These are* access to credit, visit to demonstration plots, attendance to formal agricultural training, access to extension information, distance to market center, family size and use of improved maize.

The probability of adopting chemical fertilizer among farmers with average values of continuous variables included in the model is about 26%. With access to credit the probability of adopting chemical fertilizer increased to 50%. The use of improved maize increased the probability that a fanner would adopt chemical fertilizer to 82%. Adoption of improved maize and chemical fertilizer s

Introduction

Background and rationale

he Amhara National Regional State (ANRS) is @ne of the constituent states of the Federal Democratic Republic of Ethiopia, anfi is located in the northwestern T)art of the country (Map 1). The region is located between 8°45’ and 13°45’ North atitude and 35°46’ and 40°25’ East longitude. The boundaries of the ANRS adjoin Tigray in the north, Oromiya in die south, Afar in the east^ Benishangul Gumuz in the south-west, and Sudan in the north-west. ANRS is divided into 11 administrative zones, including the capital city of the region, . The odier 10 administrative zones are East Gojam, West Gojam, Awi, North Gonder, South Gonder, Wag Himra, Nordi Welo, $(outh Welo, Semien Shewa and Oromiya (BOPED, 1999). 'Hie region consists of 101 districts and 5,300 rural and urban associations (UNECA, 1996). rrhe total area of the region covers 170,752 km2. The topography of the region is divided mainly into plains, mountains, valleys and undulating lands. The high and :nid altitude areas - about 65 percent of die total area - are characterized by a chain of mountains and a central plateau. The lowland part, constituting 33 percent of die otal area, covers the western and eastern parts of die region; diese are mainly plains and constitute large river drainage basins. Available data indicate diat, of die total area of die region, 27.3 percent is under cultivation, 30 percent is under grazing and browsing, 14.7 percent is covered by forest, bush and herbs, and 18.9 percent is currendy not used for productive purposes. The remaining 9.1 percent represents settlement sites, swampy areas, and lakes (UNECA, 1996).

The population of die region was estimated to be 15 million in 1998/99. Of diese, 90.3 i>ercent live in rural areas. The mean population density of die region is 91 persons/km' and ranges between 39 in Wag Himra to 151 persons/km' in West Gojam (BOPED, 1999). Persons below 25 years of age account for more than 65% of die total population. A large proportion of die population in die ANRS is dependent upon crop and livestock fanning. Cropping systems are predominandy rain fed. Due to population pressure and poor land husbandry practices, die level of land degradation and environmental depletion is worsening over time.

’ The region has fertile farmland and water resources suitable for crop production and ivestock husbandly. High potential areas include die western lowlands and die densely populated surplus-producing areas of Gojam and Gonder (UNECA, 1996). Farmers produce a combination of cereals, pulses and oilseeds. Cereals account lor lie largest percentage of cultivated area (84.3%) and total production (85%). This study was undertaken in Yelmana Densa District of West Gojam Zone and Farta District of Soudi Gonder Zone. 6 Tesfaye and Alemu

Objectives of the Study

The overall objective of the study is to investigate and document adoption levels and to specifically determine the factors that affect the adoption process of improved maize and draw implications for research, extension and policy. The specific objectives are to: • investigate the rate and pattern of adoption of improved maize and fertilizer, • examine the characteristics of adopting and non-adopting farmers; • identify demographic, socio-economic and institutional factors diat affect the adoption of improved maize technologies; and • draw implications for research, extension and policy. Adoption of improved maize and chemical fertilizer 7

D e s c r ip t io n o f t h e s t u d y a r e a

Amhara National Regional State

he Amhara National Regional State (ANRS) is located in north western Ethiopia and is divided into 11 administrative zones, including the capital city of T(he region Bahir Dar. The other ten administrative zones are East Gojam, West Gojam, Agew Awi, North Gonder, South Gonder, Wag Himra, North Welo, South Welo, North Shewa and Oromiya (Maps land 2). The region consists of 101 ivoredas (Woredas) and 5,300 rural and urban Keble Associations.

Topographically the Region is divided into plains, mountains, valleys and undulated lands. The mid and high land comprises the largest part of the southern, central, northern and astern part of the Region. These areas are mountainous where severeal big rivers commence from them. The highland is also characterized by chains of mountains and plateau. The lowland covers mainly the western and eastern parts (33%). These areas are largely plain and constitute big river drainage basins, and are 1 lighly suitable for wide mechanized farming and irrigated agriculture (BOA, 1999). rFhe Region has three traditional agro-climatic zones, namely, the “Dega ”(2,300 m), “Woina dega ” (between 1,500 and 2,300 mj and “Kola” (below 1500 m).The Dega constitute 25%, Woina Dega (44%), and Kola (31%) of the total area of the region. rrhe annual mean temperature for most part of the region is between 15°c-21°c. A relatively high temperature (27°c) is observed at some valleys and arid marginal lands. The region receives die highest percentage of the total rainfall in die country (80%). rlTie highest rainfall occurs during die summer season, which starts in mid June and ends early September. The soudiem and central part of die region receives high amount of rainfall compared to north western and north eastern parts.

The major soil types are red (30%), black (32%), brown (25%) and gray (1 l%)(BOA, 1 )99). Red and brown soil types are die main features of mid and low altitude areas.

Out of the total areas of die Region, 27.3% is under cultivation, 30% is grazing and browsing land, 2.1 % covered by forests, 12.6% covered by bush and herbs, and 18.9% cuiTendy not used for production. The remaining 9.1% represent settlement si es, swampy areas, lakes, etc (BOA, 1999).

The human population is about 15 million. Of diese, 90% are rural, while 10% are urban. More than 65% of die total population is under 25 years of age. The population density is about 87 persons per square kilo meters. About 90% of die population is dependant on crop and livestock for living (BOA, 1999). Cropping is predominandy rainfed. Because of population (both human and animals) pressure and poor land husbandry, land degradation and environmental depletion are getting worse over time. The Region has plenty of fertile farmland, 8 Tesfaye and Alemu and water resources for crop production and livestock husbandry. The high potential areas are the western midlands and the densely populated fertile surplus producing areas of Gojam and Gonder (UNECA,1996). ITiese high potential areas are also known for major maize production. Large and private investors are located in West Gojam. Farmers produce a combination of cereals, pulses and oil seeds and cereals.

Northwestern Amhara

Northwestern Amhara constitutes W urchi (0.6%), Dega (11.6%), Woina Dega (45.2%) and KoJa( 42.5%). Particularly, the Dega and Woiiia Dega climates are most suitable for crop production. The Region s fertile lands and water resources are found in Gojam and Gonder. Kola is dotninated by Cambisols. The dominant soil types in the Woina Dega zone include clay, red and black soils. Crops such as tef, wheat, maize, pulses, and oilseeds ;ire important. W urc/z/climate is not suitable for crop production. Ras Dashen (4620m), the highest peak in the country, Guna (4236m), and Choke (4184m) are the peak mountains that are located in the Wurchi climate.

The mean annual rainfall ranges from 70mm to 20mm. About 47 to 75% of the annual rainfall is received between June and August depending on the locality. The annual mean temperature of northwestern Amhara ranges fromlO°C to 30°C. The lowest mean minimum temperatures are from October to February, whereas die highest mean temperatures are observed between March and May.

There are three seasons of crop production based on rainfall patterns and growing periods. These ;ire the main rain season (June to August), residual season (September to November) and the Belg or short rain season (March to May). The main crop production season is the most important one where more than 95% of annual crops are produced. Residual moisture crop production season is limited to the highlands, where moisture is sufficient.

The human population of northwestern Amhara is about 9.1 Million, i.e., 60% of the population of ANRS. O f this, the rural people accounts for about 90% (BOPFJD, 1999). Majority of the population in NW E live in Dega and Woina dega areas. Nearly all rural inhabitants are Orthodox Christians. The proportion of followers of other religions is very small, and mainly live in urban areas. The inhabitants, except few places in Agew Awi Zone, speak . Map 2 shows the specific study areas in northwestern Ethiopia. Adoption of improved maize and chemical fertilizer 9

Mip 1 loc»:ion of At Amhtrt Region in Ethiopia

Source MOA »nd CDS, 1999

Map 2: The Amhara National Region Zonal Adniinstrative Divisions

THE STUDY AREAS

1 Bahir Dar 2 A chcfc 3 Bunc Wemben •* Mac hake! 5 Dera

Source MOA and CDE. 1999 10 Tesfaye and Alemu

M a iz e p r o d u c t i o n , t e c h n o l o g y DEVELOPMENT AND TRANSFER

Production

;iize was introduced to Ethiopia in the 16" or 17u' century (Kebede et al, 1993). MMaize has a wide range of adaptation in Ethiopia. The bulk of the production of maize comes from Oromiya, Amhara and Southern Nations, Nationalities and Peoples Regional States in a descending order (I ARO, 2000). Maize is grown under short and long season rain-fed cultivation. In Ethiopia, maize is mainly produced for consumption for both human and livestock feed. The green leaves and stalks are used to feed domestic animals, for fuel and construction material.

Mean annual maizo output amounted to nearly 1.45 million tons between 1980/81 and 1996/97 and increased to 2.3 million tons, (denoting die largest share (32.6%) among cereals). Maize output increased by nearly 2% per year between 1981 and 1997. In the last two years, maize output had reached a record level of 2.5 million tons. Yield also increased by 0.5% per year, but it was not statistically significant. Area under maize rapidly expanded alter the reform, increasing from 808,900 hectares in 1993 to over 1.45 million hectares in 1998 (Table 1).

According to Kebede, et al (1993), the major constraints limiting maize production in Ethiopia are identified as: shortage or excess rainfall; pests such as stock borers, common diseases such as rust, blight, streak virus and downy mildew; weeds such as striga {Strign hennontJiicn, S. aspcra, S. asiatica), Orobanche and continuous use of land without proper soil and water conservation.

A \ RS is the second maize producing Region of the country following Oromiya. It accounts for about 20.5% and 19% of the national maize area and production, respectively. Maize is third in area coverage and second in production next to tef and sorghum (CSA, 1999). Though farmers could not d ace the introduction of maize to northwestern Amhara, some elderly farmers in Achel'er and Burie weredas reported that maize was introduced during tiie worst periods of food shortage (Kifu ken = famine year), in 1898. Al that time it was believed that a farmer went to Welega (western Ethiopia) in search of food and brought back maize seed. The great famine period which occurred at that lime has enabled maize and potato lo be widely adopted . Northwestern part of Amhara constitutes about 82% and 89% of the loud maize area and production respectively (t SA, l‘)'»S).

In the past ten years, maize area expansion .uid productivity, as measured by yield per hectare, have increased by 6% and 8%, respectively. The main reasons for this trend are environmental suitability, the growing demand for food self-sufficiency, increased awareness among farmers about the lood value of maize, wide range of Adoption of improved maize and chemical fertilizer II uses and introduction of high yielding varieties, etc.

Table 1 Maize area under cultivation ( ‘000 ha) and production ('000 tons) Cereals Total Maize Year Area Prod Area Prod Area Prod %of %of %of %of total total cereal cereal area prod area prod 1989 4848.3 5685.9 5576.8 6320.0 1021.1 1688.7 >14.6 26.7 21.0 29.7 1990 4915.5 6088.8 5705.9 6755.9 1277.8 2055.6 22.4 30.4 26.0 33.8 1991 4295.2 5577.9 5154.5 67621.7 1121.4 13479.2 21.8 19.9 26.1 24.2 1992 4263.3 49290.5 5114.2 65905.0 986.5 15106.2 19.3 27.0 23.1 30.6 1993 3954.1 51487.7 4856.8 57949.6 808.9 13915.5 16.7 24.0 20.5 27.0 1994 5287.4 51052.6 7157.6 57001.4 1207.7 13378.8 16.9 23.5 22.8 26.2 1995 6448.6 58484.9 7680.7 67428.1 1418.3 13637.4 18.5 20.0 22.0 23.3 1996 7670.5 92654.0 9026.4 102876.7 1851.1 31053.9 20.5 30.0 24.1 33.5 1997 6688.6 86293.3 8011.1 91152.1 1316.9 25320.0 16.4 27.8 19.7 29.3 1998 6312.7 71974.4 8186.9 80662.2 1448.9 23443.0 17.7 29.1 23.0 32.6 Prod = Production Source: CSA ( 1990,92.94.95.97,98.99)

Maize varieties and hybrids were introduced to northwestern Amhara when cooperative farms were expanded during die Derg (a military government that Riled die country staring from 1975 to 1991). Particularly varieties such as A-511 and Alemaya Composites were introduced widi improved management. Since die extension services, provision of seeds, credit facilities and training were limited only to producer cooperatives the dissemination of improved maize to small-scale maize producers was very low. When Producer Cooperative dissolved in 1991, members shared the commonly-owned resources including different crop varieties. This has assisted in die diffusion of improved varieties of maize and management practices such as row planting, cultivation and fertilizer application. Among the technologies developed by IAR fertilizer, row planting and improved varieties were adopted in sequential order. Maize is a highly versatile crop used for many purpose and enabled farm households to be food self-sufficient.

The major maize production constraints in northwestern Amhara could be categorized into biological, socio-economic and institutional. The biological constraints include: • pests such as termite, stalk borer, rodents and weevils; ® wild animals like fox and pigs; • diseases such as leaf blight and rust; • infestation of weeds; and ® poor soil fertility.

The socio-economic problems were: 12 Tesfaye and Alemu

• low market price; • poor absorption capacity of local markets; • limited food habit and food preparation methods; • high cost of inputs; • shortage of improved seed; • poor quality seed supply; and • poor infrastructure.

The major institutional problem were availability of credit and high down payment, which discourage farmers from using credit.

Technology Development

Research on maize was initiated at Jima College of Agriculture in 1952 and the College of Agriculture at Aleinaya in 19.53. These two teaching institutions were the planners of research in Ethiopia by undertaking simple experiments on maize and oilier crops. IAJR was established in 1966 with the mandate to formulate a national policy for agricultural research and implement die policy through coordinated programs of applied research. In die same year, Bako Research Center was established in die western part of the country based on agreement between the Ethiopian government and die Federal Republic of Germany. During tiiis time, various maize germplasm were introduced to die country. The research on maize focused on screening varieties, cultural practices such as seed rate, time of planting, spacing, fertilizer type and rate, etc.

Starting form 1986 maize research was nationally coordinated from Bako Research Center. Maize research was carried out based on a team approach composed of researchers with specialization in breeding, agronomy, pathology, entomology, weed science, soil science, agricultural economics, ;md research- extension. The major objective ot the maize research was to develop high yielding varieties or hybrids along with dieir improved management and protection technologies for different agro-ecologies. During die last 40 years different open pollinated and hybrid maize varieties were developed for different agro-ecologies of die country as indicated in (Table 2). Agronomic and crop protection recommendations were also developed for both huge and small scale farmers (Tables 3 and I).

In ANRS, there are three research centers, which were established in different agro­ ecologies. Adet Agricultural Research Center was established in 1986. The main objective of die center is to improve the living standard of smallholder farmers in ljorUiwestem Amhara through research. Since its establishment, ii has generated a number of improved agricultural technologies, including crop varieties, agronomic practices, and crop protection practices. Adet Research Center has 10 testing sites each having 2.5 ha representing Administrative Zones rather dian agro-ecologies. Adoption of improved maize and chemical fertilizer 13

Maize research at Adet Research Center was started in 1995 in collaboration with the national maize research program of EARO. Research on maize focused on: • variety development and adaptation; ® pest and disease control; • management practices; and • technolog}' transfer.

So far, no improved maize technology has been generated at Adet Research Center. Currendv, farmers are using different improved maize imported by Pioneer Seed Company and developed by Bako Research Center.

Table 2. Maize varieties/hybrids developed by research and their areas of adaptation

Variety Altitude (m) Rainfall Days to Year Experimental On-farm (mm) maturity released yield (q) (q)

Open-pollinated varieties A -5 ir* 500-1800 800-1200 150 1970 50-60 30-40 UCB* 1700-2000 1000-2000 163 1975 50-70 40-45 Alemaya* 1600-2200 1000-1200 163 1975 50-70 40-45 Composite Katumani 1550 600-1000 105 60-70 40-45 ACV-3 1550 600-1000 110 1996 35-50 25-30 ACV-6 1550 600-1000 110 1996 35-50 25-30 Abo-Bako* 500-1000 1000-1200 150 1986 50-70 35-45 Kuleni* 1700-2200 1000-1200 150 1995 60-70 40-45 Gutto* 1000-1700 800-1200 130 1988 30-50 25-30 Hybrids

BH-140* 1000-1800 1000-1200 140 1988 80-90 50-60 BH-660* 1600-2200 1000-1500 165 1993 90-120 60-80 BH-540* 1600-2000 1000-1200 145 1995 80-100 50-65 BH-530* 1000-1300 1000-1500 137 1997 80-90 50-60 Beletech* 1500-2000 800-1200 160 1990 50-70 40-45 Source: EARO. 1999, Mosisa, 2001 * = Under production

Bako Research Center is the center of excellence for maize research and is responsible for introduction and distribution of germplasm and other breeding materials for maize variety development and adaptation. 14 Tesfaye and Alemu

Table 3. Agronomic recommendations for maize

Practices Recommendation Land preparation 2-3 times plowing with maresha (local plough) Planting depth 5-7 cm, planted in rows Spacing 80 cm X 50 cm, two plants/hill for full season varieties Fertilizer rate 100 (46 N/P2O5) kg/ha for open-pollinated varieties Weed management hand weeding: twice hand weeding at 25-30 days and 55-60 days supplemented with slashing herbicides: premagram- 2kg/ha Note: Herbicide use should not obviate the need for supplementary hand weeding Precursor crop noug followed by haricot bean Crops suitable for inter-cropping haricot bean, sweet potato as well as forage legumes and relay cropping

Review of Extension Systems

Impacts from research investments could only be assessed through ch uiges in farm productivity. Hus envisages the use of research-generated technologies. A strong and efficient national agricultural extension service that stimulates the adoption of recommended scientific farming techniques and ideas is thus a prerequisite for the successful technology diffusion.

Agricultural extension in Ethiopia had its beginning in the early 1950s with the establishment of the Alemaya College of Agriculture. In the early 1960s the extension function of the college was transferred to die Ministry of Agriculture, which followed die conventional approach to implement die extension service. W hen peasant agriculture gained more attention during the diird five year development plan of 1968-1973, comprehensive* agricultural projects like Chilalo (CADU) and Wolaita Agricultural Development I nit (WADIJ) were initiated (Tenassie, 1985). These projects, besides agricultural extension proper, included development of infrastructural services like roads and water. They also served as models to be expanded to odier areas later. The comprehensive approach of extension also gradually phased out because the miming cost had been expensive to duplicate to odier areas. Nevertheless, the program left a consistently positive effect and major gains in extension knowledge in the project areas. The high financial demand of the comprehensive packages led to the initiation of die minimum package projects in the 1970s under the Extension and Project Implementation Department (EPID). The minimum package extension approach comprised limited extension components like inputs, credit and extension advice. It had wider area coverage though limited to ten kilometers of either side of all weadier roads. This project continued to operate in two phases Minimum Package Program (MPP1 and MPP2) until 1985 when die training and visit system was introduced. Adoption of improved maize and chemical fertilizer 15

Table 4. Protection recommendations for maize

Pests and diseases Recommendations Stalk borer Early planting after on-set of rain Horizontal placement of the maize stalk in the sun for 4-6 weeks in the field Storage pests Drying of the grain to the optimum moisture level(%) Use of insecticides such as Primiphosmethyl dust Diseases Use of resistant/tolerant varieties Management practices -Proper tillage -Crop rotation -Timely weeding -Plant at the optimum sowing date -Optimum planting density -Balanced fertilization -Crop sanitation -Removal of crop residues -Removal of subtle -Removal of alternate hosts -Seed dressing with chemicals Management of vectors that would transmit viral diseases like streak virus Use fungicides to control foliar diseases when Justifiable

Despite various extension efforts of the past, die performance of agriculture in die country has not been improving. The major problem in die technology diffusion process was to make die products of technologies physically available to farmers, mainly improved varieties, fertilizers and crop protection products. Some agricultural technologies like improved seeds were not'produced in sufficient quantities. There are also constraints from die farmers side. Only a few fanners would have cash to purchase inputs. Credit for input purchase existed, but involved administratively cumbersome procedures often repelling farmers.

The Sasakawa Global (SG-2000) project initiated in 1993 has proved diat technologies generated by die national agricultural research systems, if properly utilized, could double and even uiple yields of major cereals grown in die country. The secret behind the SG-2000 technology transfer program was simply filling the major gaps that had existed during die various extension systems of die past. These, among odier things, include access to improved technologies and odier inputs and make diem physically available through the provision of credit. Intensive practical U-aining of extension staff down from different levels and improving mobility of extension workers through provision of vehicles, motorcycles and bicycles have greatly facilitated die success of SG-2000. The odier sUengdi of SG-2000 is die big effort it is making to bring stronger linkages between research extension and input distributors, which is a key issue for a successf ul transfer of agricultural technology.

The experience of SG-2000 and die Training and Visit (T and V) system have gready contributed to die formulation of an extension strategy known as die 16 Tesfaye and Alemu

Participatory Demonstration and Training Extension System (PADETES). It is a syndiesis of' the SG-2000 approach, which uses large plot, usually one-fourth to half of a hectare, to demonstrate improved farming practices. Training.is given bodi to the extension staff and farmers. The regular visits to demonstration plots provide ample opportunity to discuss widi farmers about problems encountered in the process. Important packages of recommendations of the strategy include : • improved seed varieties; • seedbed preparation; • optimum seeding rate; • mediods of fertilizer application; • fertilizer type and rate; and • use of pesticides

The recommended production packages tor maize comprise • tillage; • forming furrow's; • applying half die fertilizer; and • dien row' planting.

Fanners acquire improved variety and treat it \vidi pesticide to limit damage from pest. At die second weeding, application of second dose of fertilizer near the base of the plants and then passes down die row widi a plow' burring the side dressed fertilizer into the soil and simultaneously destroying weeds.

Seed Production

The Ethiopian Seed Enterprise (ESE) is an autonomous commercial public enterprise. At present the ESE is die country’s principal producer and supplier of improved seed. It has five stationary seed processing plants at Asella, Nekemt, Kofelle, Awasa and Bahir Dar and four mobile seed cleaners. 'Hie enterprise owns diree basic seed farms found at Gonde/Etaya, Sliallo and Kunzila. These farms are located in agro-ecologies suitable for seed production. In addition to die seeds produced on diese farms ESE procures seed from contract growers by inspecting die performance of the crop from planting to harvesting (ESE 1099). Until the change in government in 1991, out of die total seed produced by die enterprise about 80% was sold direcdv to state farms. The share of NGOs and die Agricultural Input Supply Corporation which distributed improved seed w'as about 2()%. After die institutionalization of the new extension package program die ESE distributed about 80% of its seed to small-scale fanners dirough die different agricultural bureaus of die regions. Adoption of improved maize and chemical fertilizer 17

A large quantity of maize seeds has been produced in Ethiopia from 1980 to 1997 (Fig 1). It is evident that quantity of hybrid maize seed production has increased from 4799 in 1995/96 to 42526 quintals in 1998/99 and it is further expected to increase significantly in die near future as maize hybrids possibly possess considerably higher yield potential than composites. Composite and hybrid maize seeds are processed and graded at the processing plants of the ELSE. Seeds of maize are treated and packed into 4, 7, 12.5 and 25 kg bags according to the needs of client sand sold to fanners through the MOA. The prices of seed with which hybrid and composite varieties are sold in 1998/99 were 547and 222 bin/q.

Figure 1. Certified and commercial seed sale (q)

300.000

180.000

160,000

140.000

120.000 •

100,000

80,000

60,000

40.000

20.000 ill <£> (v^ ofV o5i) (A cO'' CV^ o!^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ Year

Source: Ethiopian Seed Enterprise, 1999 18 Source: Ethiopian Seed Enterprise (1^99) Enterprise Seed Ethiopian Source:

B irr/Q t 97 1997/ 8 /9 7 9 9 1 7 /9 6 9 9 1 iue : eln Prkt I Btr 19/7 n 1997/98 and 1994/97 l Q trr/ B In i d e e S f to k r P Selling 2: Figure Tesfaye and Alemu *r e* Y S W h c a t t a c h W S ■ B ir ley ir B ■ l mize e iiz m ile u p m o C ■ ghum m u h rg o S □ e r u m rid b y H B f c T Q Adoption of improved maize and chemical fertilizer 19

M e t h o d o l o g y

Sampling Procedure

he study was conducted in ELast Gojam, West Gojam and South Gonder Zones. These Zones were selected on the basis of maize production area, number o f Tgrowers, potential for maize production, accessibility and representation of the farming systems. They were selected in collaboration with relevant extension experts of the Amhara Regional Bureau of Agriculture.

Once die Zones were selected, die same procedure and selection criteria were used to select die study Woredas. The relevant extension experts at the zonal agricultural department were involved in die selection of die study Woredas. l he Woredas selected for die study were Dera from South Gonder, Bure Wonbera, Bahir Dar Zuria and from West Gojam and Machakel from East Gojam (Map 2).

The selection of die sample farmers involved a two-stage sampling procedure. The sample peasant associations (PAs) w^ere selected randomly using random sampling procedure. In the course of selection of the sample PAs precaution was taken not to select inaccessible and non-maize growing PAs. Sample fanners were then selected from sampling frame obtained from development centers and/or peasant association offices of the respective PAs using random sampling procedure. A total of 3(59 farmers were selected and included h\. die study.

Data Collection

Data were collected from primary and secondary sources. The secondary source of ^ infonnation included published and unpublished information about agricultural production in particular and the study area in general. These information were collected from regional, zonal and Woreda level offices of agriculture and planning and knowledgeable individuals. The primary data were collected from sampled fanners using structured questionnaire. Before starting the actual data collection, the questionnaire was pre-tested to modify some inelevant questions. Experienced enumerators from Adet Research Center were involved in administering the questionnaire. The enumerators were given training on the content of the questionnaire, methods of data collection and on how to approach fanners. The data collection was done between March and April 1999.

Analytical Procedure

Adoption studies often attempt to analyze and understand die observed adoption patterns (CIMMYT, 1993). The examination of farmers’ opinions and observations and statistical comparison of adoption measures with characteristics of the farmers’ n regarding a new technology are o: ;x» The purpose of this analysis identify the most important factors ^adoption.

The lata collected were analyzed using SWSS 8. The different analytical techii ({ues applied were t-tests, chi sqyarev test, lation analysis, and logistic regref:{ion models. Frequency and meilns were ijput&d for dikerent variables, TJie lest was run to see if there is any statistically s| ■cant difference in continuous variai> es of farm characteristics of farmers who adopted improved maize and thos< rvho have not done so. The chi-square test run to se6 if there was any syste ratic association between adoption an^ sonij 'atom characteristics. Out of the two logistic and probit) related multi-fatf - ; - that 'are used for adoption studi :i a logistic adoption model fitted toBeterd| ^-factors affecting adoption of imprp yed maize and chemical fertilizers.

The ogit model is used to estimate the prOpabilit ^adoption of improved maize that lakes either of the two values of Y-l; forjadopt tndY - 0 for non-adoption of im p lied maize and chemical fertilizer. The [on|I form of the model is presi i ted as follows:

M'X) P r ^ r = 1) = l.+ e w>jr> Wlide p *X is defined as:

P’X- 3t+piXi+ptX»+pjXa+...+PiXi+7i

W hiie, P- is the constant. j$, - i ^l,2,...n are e ^coefficient of the exogenous van es to be estimated. X is a vector of e&pl r# variables; y» is the error term with iero mean and constant variance. Fatmet eciston to adopt or reject new techii*ologies was influenced by the combines effe number of factors related to fam dr'’s objectives and constraints such as^farm'B *: socio-economic circumstances (age &nd formal education etc); farmers’ resburce dbwments as measured by (size o f f ; inily labor, farm size and oxen ownefsl # institutional support systems avaife Me to farmers (credit, extension and availabi of inputs), (CIMMYT, 1993). A n l nber of variables were hypothesized !to ini irice the adoption of improved maiie and inorganic fertilizer. These variables in< ieif '.'£•■

Le\ el of education Levil of education was assumed to increase farmei ability to obtain process and use info ination relevant to the adoption ojf imgrovt and fertilizer. Education is theri ore expected to increase the probability of a&^fotion of improved maize. r Farmers’ experience Exppience of fanner could affect confidence, h more experience, farmer can *

1 . J L k Adoption of improved maize and chemical fertilizer 21 avert risks to adopt improved maize technologies, so this variable can positively or negatively affect farmers’ decision.

Household size Household size was expected to affect the probability of adopting improved maize and chemical fertilizer. Large households will‘be able to provide the labor that might be required by improved maize.

Use of hired labor Use of hired labor was hypothesized to be positively related to the adoption of improved maize.

Access to credit Access to credit can relax the financial constraints of fanners, and in some cases, access to credit is tied to a particular technology package. Access to credit was expected increase the probability of adopting improved maize and fertilizer.

Agricultural extension service Agricultural extension services provided by the Bureau of Agriculture of the Amhara National Regional State were the major sources of information in the study area. Contact with extension agents (development agents) was assumed to increase farmers’ likelihood of adopting improved maize and fertilizer technologies.

Tropical Livestock Unit (TLU) The number of livestock owned by farmers was hypothesized to be positively related to the adoption of improved maize and fertilizer. Tropical livestock unit is an index where livestock numbers are aggregated using the following weighing factors: cow~0.8, goat=0.4, and sheep=0.4.

Distance from nearest development center The further an extension office from farmers’ homes, the lesser access of farmers to information will be. This was expected to be inversely related to adoption.

Attending field days, visiting demonstrations, and attending formal agricultural training Farmers who had attended field days and formal agricultural training, and visited demonstration fields were expected to have positive attitudes to adopting improved maize and chemical fertilizer.

Farm size Increasing the production and productivity of maize depends on cropping intensity of improved maize and chemical fertilizer. Therefore, farm size was hypothesized to be inversely related to adoption of improved maize and chemical fertilizer. Iff i ! f m SH IP 11*

2 2 TesfayeiandAl T

Adequacy of draft power 1 adequacy of draft power was hypothesised to j| positively related to the probability f adoption of improved maize.

stance from market center tfstance from market center was h: ltd “be negatively related to the robability of adoption o f improved t households living near market < enters tend to use improved maize,; for have easy access to dispose of their production.

J se of chemical fertilizer lie of chemical fertilizer was hypothesised to positively related to the probability 1 adoption of improved maize in tfyat improve are alleged to perform better 4ilh chemical fertilizer. T Adoption of improved maize and chemical fertilizer 2 1

D e m o g r a p h ic , s o c io - e c o n o m ic a n d INSTITUTIONAL CHARACTERISTICS

Demographic Characteristics

able 6 shows the demographic characteristics of maize farmers in the study area. Level of education was one of the demographic characteristics hypothesized to Tinfluence the adoption decision of improved maize technologies. This was based on the assumption that exposure to education increases farmer’s ability to obtain, process, and use information relevant to the adoption of improved technologies. The level of education tended to be systematically associated with die adoption of improved maize (%*= 7.244; *P< 0.05). Besides, a significant difference was found in level of education of adopters and non-adopters (t=2.757; P =0.05). The average level of education of adopters is 1.72 while that of non-adopters was 0.84 years. Both non-adopters (55 %) and adopters (62%) have attended primary school. About 44% of non-adopters and 31%> of adopters were illiterate.

As expected, a significant difference was found in family size of adopters and non­ adopters of improved maize (t=2.280; P<0.05). The average family size of non­ adopters was 5.35 while that of adopters is 6.01 . Almost all farmers in the study area were Orthodox Christians.

It was hypothesized that farmer’s age can increase or decrease the probability of adopting improved maize. However, no significant difference was found in age of adopters and non-adopters. The average age of both adopters and non-adopters was about 40 years.

Although the experience of farmers was assumed to influence the decision behavior of adopting improved maize and chemical fertilizer, no significant difference was found in years of experience of adopters and non-adopters. The average farm experience of both non-adopters and adopters was about 17 years.

As for the demographic characteristics of farmers, only level of education and family size were found to significantly influence the adoption decision of improved maize and chemical fertilizer. The other demographic characteristics like age and farming experience were not strong enough to influence farmers’ decision behavior in adopting improved maize. 24 Tesfaye and Alemu

Table 6: Demographic characteristics of maize farmers in the study area, 1998

Characteristics Non-adopters Adopters t-statistic Mean SD Mean SD Age of head of household 40.8 13.8 40.12 11.23 0.466NS (years) Level of education (grades) 0.84 1.51 1.72 2.69 2.757** Total farming experience (ye^rs) 24.97 13.51 24.39 11.78 -0.375NS Own farming experience(years) 16.96 13.19 17.35 11.22 0.260NS Family size 5.35 1.92 6.01 2.34 2.280** Characteristics Non-adopters Adopters ^statistic N % N % Religion Orthodox 76 100 289 99.7 NA Level of education Illiterate 34 44.2 89 30.7 7.244** Primary 42 54.5 181 62.4 Secondary 1 1.3 20 6.9 Significant at 10 %: * * = Significant at 5 %: and • • • = Significant at 1 %. NS = Nonsignificant NA = Not appropriate

Socio-economic Characteristics

Labor Table 7 shows the socio-ecoiiomic characteristics of maize fanners in the study woredas. Socio-economic characteristics like size of labor force, land size, and tropical livestock units were postulated to iulluence the decision of farmers in adopting improved maize, and chemical fertilizer.

Having a job off-farm was one of the characteristics assumed to influence the adoption decision of improved maize widi the assumption that farmers with off-farm jobs may earn a certain sum of money to invest in the purchase of improved maize. However, no association was found. Besides, no significant difference was found in die amount of income earned from off-farm jobs between adopters and non­ adopters. The majority of farmers (93% of non-adopters and 88% of adopters) did not have off-fann jobs. Only 7% of non-adopters and 12% of adopters were found to have off-farm jobs. The off-farm jobs include guarding, carpenter, petty trade, weaving and daily labor. Of these, petty trade is the most important off-fann job.

It was also hypothesized that if there is any one in the family working off-farm, die household could receive a certain sum of money to invest in the purchase ot improved maize. A systematic association w'a.s found between existence of a family member working off-farm jobs and adoption of improved maize (%*=3.2/J7; P <0.1). However, no systematic association was found between having a family member working off-farm and using die income earned for purchase of farm inputs and adoption of improved maize technologies. No significant difference in the amount of off-fann income between adopters and non-adopters. Adoption of improved maize and chemical fertilizer 25

The size of labor force in the household was assumed to bring about variation in decision behavior of farmers in adopting new technologies as larger households be able to provide additional labor for improved maize or application of fertilizer. However, no significant difference was found in size of labor force between adopters and non-adopters. The mean size of labor was 3.57 for non-adopters and 3.90 for adopters of improved maize (Table 7).

A systematic association was found between hiring labor and adoption of improved maize (x*=7.009;P<0.05). However, no significant difference was found in labor force between farmers who hire seasonal labor and those who don’t do so. The majority of the fanners (non-adopters 90% and 75% adopters) did not hire seasonal labor. Seasonal labor is hired for all farm operation from plowing through threshing. Of these, weeding and harvesting are the most important operations, which involve hiring of seasonal labor. A similar result was also found with hiring of permanent labor in that adopters tended to hire permanent labor than, non-adopters 8.118; p<0.05). Besides, a significant difference was found between those farmers who hire permanent labor and those who did not. About 31% and 15% of adopters and non­ adopters respectively hired permanent labor during the study year. The majority of farmers (55% non-adopters and 85% adopters) use community labor for maize production. A systematic association was found between adoption of improved maize and using community labor (x*=4.439; P<0.05). A significant difference in labor force between households who used community labor and those who did not use. The most important operations which used community labor were weeding and harvesting. The type of community labor used by the majority of both adopters and non-adopters was Wonfel.

Table 7: Socio-economic characteristics of maize farmers

Characteristics Non-adopters Ado liters X^statistic N% N % Off-farm jobs Yes 5 6.6 35 12.1 1.865NS No 71 93.4 255 87.9 Does any one in the Yes 6 9.5 10 4.0 3.257* family do off-farm jobs No 57 90.5 243 96.0 Use off-farm income for Yes 9 26.5 51 38.9 1.811NS purchase of farm inputs No 25 73.5 80 61.1 Mire seasonal labor Yes 8 10.5 66 14.7 7.009" No 68 89.5 208 75.3 Use community labor for Yes 56 74.7 238 85.0 4.439** maize farm operations No 19 25.3 42 15.0 Maize tenure -Own 56 83.6 220 84.3 NA Share cropping 4 6.0 32 12.3 Hire permanent labor Yes 11 14.7 89 31.2 8.118“ No 64 85.3 196 68.8 Significant at 10%: ~ Significant at 5 %; and NS = Nonsignificant at less than I %o. NA - Not appropriate j trig .* i f

armland Ifhe study revealed that there is significant ence in farm size between adopters improved maize and that of non-adopt : In the study area farmers make Afferent arrangements Co get access to'faddi't i t land. Such arrangements include e asing in land or a system of sharecropping, te proportion of farmers using these rrangements vary depending on the t|pe ot rops grown. Sharecropping is more important than teasing in land.

fi ible 8; Mean values of farm resources of maize f$T eitudy area

Characteristics Non-g Copters t-statistic Mean m jSD ib o r force 3.57 I1.638NS ;arm size (Ha) ______1.57 ?.394** \jinuai income earned from off- 634.40 D.596NS rm jo b s ______tynual income earned by family ttembers from off-farm jobs - 887.50 ik Significant at 5 %; and NS ~ Non-sigmfwdjit ;

ivestock ; j .u livestock is an important component jc if the ning systems in the ANRS. In the tudy area, bovines, equines, small n| min Jbo^try, ai$l bees are important :pmponents of the livestock systent> Tfr e tfijke livestock units was hypothesized to tifecl the adoption; decision of improve d tecjj as th$y are good sources of iish for purchasing tann iilputs. InjliglWof l ^ighpicant difference was found in ippical livestock units between adopters 1 ‘liori-adopfets (t=-2.054; P

'ible 9. Size Adoption of improved maize and chemical fertilizer______2 7

A significant difference was found in number of oxen owned by adopters and non­ adopters (t =2.244; P<0.05). The average number of oxen owned by adopters was 2.07; while that owned by non-adopters was 1.80. A systematic association between adoption of improved maize and means of access to draft power (%3= 11.351; P<0.05). The proportion of farmers using their own oxen are higher among adopters than non-adopters (Table 10). Adopters of improved maize (73%) and non-adopters (54%) reported that the available draft power was adequate for maize production. The chi-square analysis shows that adoption of improved variety/ hybrid was systematically associated with the adequacy of draft power for maize production (X*“9.694;P<0.01). Farmers with adequate draft power appeared to have adopted improved maize than those with no adequate draft power. This implies that unavailability of adequate draft power is a constraining factor for adoption of improved maize and chemical fertilizer

Farmers with inadequate draft power indicated that they obtain additional draft power through borrowing, using a local arrangement called mekenajo and sometimes renting. About 75% of non-adopters and 83% of adopters reported that they use mekenajo.

Farm implement Only few adopters reported to have ever used improved farm implements. The reasons for not using improved farm implements are unawareness and unavailability of the implements. About 41% of non-adopters and 40% of adopters mentioned unawareness to be the reason for not using improved farm implements. A good proportion of farmers (29% of non-adopters and 34% of adopters) mentioned unavailability as another important reason for not using improved farm implements. This implies that a concerted effort has to be made by concerned institutions to develop and distribute the required farm implements in order to facilitate the adoption process of improved farm implements in the study area. I institutional Characters

Extension Farmers in the study area could have ai tension messages through field days cr visiting a demonstration plot, fo <£iltural training, and listening to agricultural programs on radio. It thesteed that access to extension i iformation through one of these ch lild increase fanners’ likelihood of a dopting improved maize and chemica ^%In light of this, access to extension r messages through one of such sources i^ys, formal agricultural training and stening to agricultural programs on it^ma0cally associated with adoption of improved maize (%S=5Q.965; P

Adoption of improved maize is systematically associated with having access to extension messages through field days or demonstration plot (x*= 41.221;P<0.01). This result indicates that those farmers who have attended field day are more likely to adopt improved maize as compared to farmers who have not attended. About 48% of adopters reported to have visited a demonstration plot or attended a Held day while only 8% of non-adopters reported to have done so.

Table 11: Institutional characteristics of maize farmers

Characteristics Non-Adopters Adojrters ^•statistic N%N % Access to extension information Yes 10 12.7 168 57.9 50.965*** No 69 87.3 122 42.1 Ever attended a field day/visited Yes 6 7.8 138 48.1 41.221*** maize demonstration plots No 71 92.8 149 51.9 Ever attended a formal Yes 4 5.2 39 13.6 4.140** agricultural training No 73 94.8 247 86.7 Have a radio Yes 3 4.9 57 22.8 10.318*** No . 59 95.1 193 77.2 Listen to any agricultural Yes 2 9.6 49 53.8 13.515*** programs on radio No 19 90.4 42 46.2 Visited by an extension agent in Yes 13 25.0 181 74.5 46.587*** 1998 No 39 75.0 62 25.5 Get credit for maize production Yes 26 34.2 270 93.4 137.635*** No 50 65.8 19 6.6 Ever disqualified from getting Yes 16 24.2 57 20.5 0.446NS credit No 50 75.8 221 79.5 Have credit problem Yes 35 ' 50.0 112 39.9 2.368NS No 35 50.0 169 60.1

Used credit for maize production Yes 43 59.7 258 90.5 41.239*** in 1998 Member to an organization Yes 55 77.5 227 82.8 1.094NS No 16 22.51 47 17.2 * * = Significant at 5 %; and * * * = Significant at / %. NS ~ Nonsignificant at less than 1 %

Besides, the study had revealed that attending a formal training in agricultun facilitates the adoption of improved maize. This was confirmed by die result that adoption of improved maize was systematically associated with attending a formal training (x*= 4.140; P<0.05). About 14% of adopters have got formal training as compared to 5% of non-adopters (Table 11).

Visit by extension personnel was also found to be systematically associated with adoption of improved maize (%i= 46.587; P<0.01). About 75% of adopters and 25% of non-adopters were visited by development agents in 1998. During die same year, the number of extension contacts made with adopters was significandy greater than 30 temu that of non-adopters (t=-7.702;'j The average number of visits paid to adopters were 2.85 and thatofnoj 0.34 (Table 12).

Adoption of improved varieties/• *1 ; >eared to be realized by those farmers who own a radio x= 10.3l8;B<0.j( poportion of farmers owning radio were higher among adopters than non>j |bOut 23 %cof adopters and only 5 % of non-adopters Own a radio Fahne ined to agricultural programs are more likely to adopt improved maiie i, x*~ 13.515;P<0.01). About 54% of the adopters and only 10% of nion^ Ifften to agricultural programs on radio. Commenting on the content O f the agricultural radio programs broadcasted by the Ethiopian |jc Service, about 51% of adopters and 11% of non-adopters indicated Factory. Broadcasting time was identified as the most important probletn*

Distance from development: centers vfiR ^assumed to influence the decision of farmers in adopting improved irapei jiW f&rmers living near development centers were more likely to be visited more brajeyelopment agents, thus, getting access to information and inputs of iniproyed tecBnolo^es thaii those households far away. However, no significant difference was fotnd between adopters and non-adopters.

able 12. Extension contacts at different tini teinth^ud^aregin1998.

Time of visit and other extensi0j| ‘ Ndn-afll iteirs Ado Dters t-statlstic variables '1 VSD Mean SD v

Before plowing i l 1 .0 0 'jl j p o z 1.41 ■ 0.80 -1.155N3: During plowing tf] W i 1.50 ! 1.03 4.908NS; At planting • ;j 1.20*1 :®45|>S 1.19 T 0.48 0.041 NS : During weeding 1 i.O O ll 'I q o ii 169 ^ 0.87 -2.105NS At harvesting ■! 1 .0 0 ,ffl i i o t r 1.18 ' 0.48 *0.846NS: During slack period : 'U iM 1.78 \ 2.20 -1.438NS Total number of extension contact J 0.34 j | Q.95 2.85 2.85 -7.702*** during the year J J Distance to travel to development center 1 m m 28.95 26.59 -0.056NS (minutes) 1 1 f ' 0 Distance to travel to market center ; 1 81.92 59.32 1.057NS (minutes) ) •* = Significant at i %. Ns - Nonsignificant at lessfy " t Credit There was a systematic associatio|i betvra iX adoption o! improved maize and access to credit (x“ 137.635;P<0.01), indicating! lajt farmers who have access to credit tend to adopt improved maize and c|emicajj fd|tili^er than those farmers who did not have access to credit. About 669o of n® adopters and 7% of adopters reported to have no access to credit. (Table l|). Adoption of improved maize and chemical fertilizer 31

Farmers in the study area were constrained by shortage of cash to purchase farm inputs. About 50% of non-adopters and 40% of adopters reported that diey have credit problems. The issue of credit assumed various natures. However, about 46% of non-adopters and 40% of adopters identified high interest rate as the most important credit problem. Unavailability of credit facilities was identified as the second most important problem by 20% of non-adopters and 14% of adopters of improved maize.

Fanners who did not have access to credit financed their maize production through sale of crop and livestock products. About 58% of non-adopters reported that they financed their maize production by income obtained from side of crops. About 89% of non-adopters reported that they take credit to purchase fertilizer while 92% of adopters took credit to purchase both improved maize and fertilizer.

About 62% of non-adopters and 70% of adopters reported that BOA is the major source of credit f or farm inputs.

Farmers reported that they fail to obtain credit because of lack of cash to effect the down payment, which is 25% of the total sum of the loan. About 24% of non- adopters and 21% of adopters reported to have been disqualified from getting credit because of their failure to pay down payment. Most farmers, 60% and 91% of non­ adopters and adopters of the sample farmers used credit in 1998. The use of credit for maize production in 1998 was systematically associated with adoption of improved maize (x‘=41.239;P<0.01) (Table 12). About 79% of non-adopters and almost all adopters bought both DAP and urea using tiie credit obtained from the BOA. 32 T&faye q jiAlem

Ma iz e p r o d u c t io n M|NA||EfMiNT

Land Preparation and Plani

rmers prepare land for' ms^ize pk< uction using a pair of oxen, •A.F proportion ni of fanners (76% of fion?adii tsrs iind 90% of adopters) were found to use own pair of oxen to cultivate their |fields. Fanners plow their farms up to four times between Defcember and■6 Marc! . reported by 68% of adopters and 05% of non-adopters. The second plowing is rie in these months as reported by 72% non-adopters and 68% adopters. Tjie plowing is done between March -June as reported by 93% of non-adopters and 8< of adopters. The fourth plowing is done between March and May as repoilfcd by| non-adopters and 91% adopters and 76% of non-adopters and 74% ofaii e fourth plowing from April to May.

Planting of maize is usually done $*-to June as 91% of non-adopters and 88% of adopters reported diattthey $iay. The majority of both adopters and non-adopters are planting in the I second week of the respective planting month. The reasons for planting articular rimes as mentioned by 49% of non-adopters and 43% of adopter ty of adequate moisture. It wasjeamt that 86% of adopters use row pi ile ;51% o f non-adopters use broadcast planting. The chi-square analysis, t planting method was systematically associated with adoption of imprc (^ai7:762;P<0.0l). Row method of planting improved maize was st 76 by few fanners. However, more farmers started to use row p&ntii n$w extension package program was launched in 1995. ‘The rsite of adoj|tion: planting Was 4% in 1994 and reached 29% in 1998 (Table l6).

The majority of adopters (67%) use' 50 ( lacing between plants, which is similar to the recommended spacing between pi. A. good proportion of adopters (26%) also use 30 cm spacing between pkn (.spacing between rows practiced by the majority of adopters (76%) is $0cms, is siinilar to the one recommended by research and extension (Table 13).|TI jrce of information regarding row method of planting are mainly extension* anc iftoring farmers. 42% non-adopters and 86% adopters got the inform^UiOn; from [C^A followed, by neighboring farmers as reported by 39% of non-adopters and 7 if adopters. 'Fhe reasons (as indicated by 28% of non-adopters and adopters|; for OW method of planting were easiness in harvesting, fertilizer application and jsi seed requirement. The majority of adopters (57%) reported to use 2 sdeds jfj hole; while non-adopters reported to use only one seed. Adoption of improved? iaize Was systematically associated with number of seeds planted per holef (x^ t P<0.01). Adopters appeared to use two seeds per hole which is rfeco&me: 1 ;by research and extension while noil- adopters tend to use only onfe. Tiie jtfi |Iso showed a significant difference in number of seeds planted per hole (jr 3.7 p

■ -fei '-S' Adoption of improved maize and chemical fertilizer 33

Table 13. Spacing and amount of fertilizer used in the study area

Characteristics Non-adopters Adopters t-statistic Mean SD Mean SD Spacing between plants (cm) 42:29 10.83 44.17 10.31 0.848NS Spacing between rows (cm) 64.38 21.33 71.75 15.26 2.176** Number of seeds planted per hole 1.22 0.42 1.57 0.50 3.795*** Amount of DAP used when started (kg) 48.27 21.06 51.84 34.19 0.799NS using chemical fertilizer Amount of urea used when started (kg) 43.0 32.58 42.88 23.66 0.011NS using chemical fertilizer Amount of DAP applied in 1998 (kg) 44.36 29.53 53.60 48.59 1.171 NS Amount of urea applied in 1998 (kg) 26.92 12.59 41.19 22.16 -3.228“ * Frequency of weeding 2.33 0.88 2.13 0.85 1.737* Time after planting to start first weeding 3.66 1.53 3.97 1.85 -1.327NS (weeks) Time after first weeding to start second 3.59 1.84 3.73 2.04 -0.498NS weeding (weeks) * * * = Significant at 1%; ** = Significant at 5%; * = Significant at 10%; and NS = N onsignificant at less than 10%.

W eeds

Weeds are important biological factors constraining die production of maize in the study area. Almost ail farmers reported to practice weed control methods. About 40 % of non-adopters and 15% of adopters used only hand weeding. The larger proportion of both groups do hand weeding along with cultivation and hoeing. The chi-square analysis indicated a systematic association between adoption of improved maize and method of weed control (Table 10).

The frequency of hand weeding ranges from 1 to 4. 50% non-adopters and 47 % adopters weed their maize field twice, which is similar with research recommendation. Where as, 24% of adopters and 27% of non-adopters weed their maize field three times. The first weeding is done 2-4 weeks after planting. 36% of non-adopters and 31% of adopters weed their maize field four weeks after planting, which is in conformity with the recommended date of weeding. The second weeding is done by both the majority of adopters and non-adopters 2-4 weeks after first weeding. 35% of non-adopters and 31% of adopters of improved maize carried out the second weeding four weeks after first weeding.

In addition to hand weeding oxen-cultivation locally called shilshallo (hoeing) was also practiced. The majority of adopters (98%) and non-adopters (72%) use shilshallo for removing weeds from die field. The chi-square analysis shows systematic association between practicing cultivation and adoption of improved maize (xJ=59.559; p<0.0l) (Table 10). The result shows that farmers who practice cultivation appeal' to have adopted improved maize tiian diose who did not practice. 34 Tesfayeartf Alemu

This may probably be related to fee amjfflgibility of adequate draft power. The majority of both adopters (81%) and non*a||>p|ers (90%) practice hoeing before first hand weeding. Answering to the qug$ o|i as to whether they know the recommended weeding frequency aijid tim$ jfrweeding, 40% of adopters and 4% of non-adopters responded that they kifow tfjj ecommendation of researchers and that of the Bureau of Agriculture. The cl^-sqUd analysis showed that adopters are more likely to be knowledgeable about research dlextension recommendation than non- adopters (x‘=33.763; p<0*01). *

The source of information regarding freqri&icy and time of* weeding is reported to be BOA by 39% of the non«adopter| and of adopters. Almost all farmers in the study area have never used herbicide^ to cojffn?l weeds in their maize fields.

Pest and Disease Mana

Pests are important biological factors cau reduction in maize yield. Farmers in the study area reported that almosjtjall es are susceptible to pests. However, according to 28% of adopters, me !prietyf| -511 was the most susceptible one.- To reduce the loss, about 63% of* ad« 1 35% of non-adopters practice both cultural and chemical control nfethic tipn was npted between adoption of improved maize and practice of p« # 1 8 ,9 6 2 ; p<0.0lj (Table 10).

In the study area, there were bot a£fe (rodents, porcupine, fox, etc) and invertebrate pests (such as stalk boi , attnywoniii, etc). Of the invertebrate pests, cutworm, stalk borer, termi evils are the most important pests affecdng maize production. AccordiS ^lajprity of the farmers, the severity of these pests is medium to high J Ta jp f the vertebrate pest, the severity of damage of wild animals is consiOerc in to high by 56% of non-adop.ters and 61% of adopters. The damage caus< ) by rodents is considered fiiinorby both non-adopters (68%) and adopt (7 Adoption of improved maize and chemical fertilizer 35

Table 14: Severity and rank of pests affecting maize production in the study area.

Characteristic Non-adopters Adopters N% N % Severity of damage of stalk borer - Very severe 6 10.0 27 11.7 - Severe 34 56.7 106 46.1 - Minor 20 33.3 97 42.2 Rank of stalk borer First 12 20.3 58 25.2 Second 27 45.8 97 42.2 Third 17 28.8 60 26.1 Severity of damage of cut worm -Very severe 5 15.2 57 41.3 - Severe 22 66.7 44 31.9 - Minor 6 18.2 37 26.8 Rank of cut worm -First 14 42.4 68 48.9 - Second 11 33.3 34 24.5 - Third -- 23 16.5 Severity of damage of rodents - Very severe -- 5 5.6 - Severe 9 32.2 22 24.4 - Minor 19 67.9 63 70.0 Severeity of damage of wild animals - Very severe 10 24.3 33 18.3 - Severe 13 31.7 76 42.2 - Minor 18 43.9 71 39.5 Severeity of damage of weevils - /ery severe 6 46.2 33 67.3 - Severe 6 46.2 15 30.6 - Minor 1 7.6 1 2.1 Rank of weevils - First 10 71.4 33 67.3 - Second 2 14.3 12 24.5 - Third 1 7.2 4 8.2 Severeity of damage of termites -Very severe 12 75.0 47 77.0 • Severe 3 18.8 12 19.7 - Minor 1 6.2 2 3.3 Rank of termites ' • First 13 81.3 53 81.5 Second 2 12.5 9 13.8 • Third 1 6.2 1 1.5 36 1 Post-harvest Management

Harvesting of maize was: found to be done jp tweep October and December with the majority of both adopters (5196) anfl not outers (45%) harvesting in November followed by December (Table 15). Erior |j larvesting farmers select seeds for next ' V ' rs season. Diverse criteria such as size £>f cob] inability of grains, straightness of seed rows, and seed color are considered fn theS lotion of seeds. O f these, size of cobs, maturity of grains and seed color wire thf nbst important criteria. Seed selection was done mainly in the field and at|home| was usually done immediately after harvest as reported by 95% of adopters and^fe% of non^adopters. Seed selection and storing of maize grain was done indh^duaUjflr^pintly by husband and wife ais well as children.

Farmers who adopted improved maii pcle or use their own seeds while, Non-adopters maintain seeds from R zest for next season while adopters, Lte who are using mainly hybrid seed! from external sources (x*~4.243; P<0.05) such as extension service. Tt ty of non-adopters (70%) reported tfiat seed is not readily available while 613 ptfers get the seed on time for planning, A systematic association betweeflladoptiori iproved maize and readily availability of seeds (x*= 16.793; P<0.01). | ■ f The analysis also disclosed that ad pcared to buy improved maize seeds regularly than non-adopters. This tmte the fact that adopters using mainly hybrid seeds should buy roved maize hybrid varieties; wj|le non-adopters or farmers using com; varieties could recycle the seed with or without minimum loss in yield, <$f adopters reported to buy improved maize every year while 13% buy on@e ui ej two. years. The latter could be those farmers who either use composite oifrecycl hybrid m&ize variety that they have purchased and produced the previous ction season. The major reasons mentioned by 48% of adopters aind 5]6% n opters for not buying seeds regularly is unavailability and unaffordable pricje, resp ely.

The most important methods of seeil sto V e storing on crips with out shelling and putting in gottera after shelling as re by 75% of non-adopters and 81% adopters. A large proportion of adopters t) reported to treat maize grain with chemical during storage; while about(62% oj lon-adopters reported not to do so. A systematic association between chetrucal lent of maize grain and adoption of improved maize (x!,=45.969;P<0.01 )i| Farmj who have adopted improved maize were more likely to treat the product with1 ethical than non-adopters. About 46% of non-adopters reported that since j the t|tt df maize they produce was very small there is no need for using chej|ricalsi f adopters as one of the reasons for not treating maize grain during stor h^regard to the proportion of maize Adoption of improved maize and chemical fertilizer 37 consumed or sold as green, the majority of non-adopters (43%) consumed one-third while adopters (40%) consume little.

Table 15. Post harvest management and utilization of maize products in the study area

Characteristic Adopters Non-adopters N %N % Month of harvesting . October 13 17.1 35 12.2 ■* November 34 44.7 146 50.7 - December 28 36.8 97 33.7 Treat maize seeds with chemical -Yes 29 37.7 224 77.8 -N o 48 62.3 64^ 22.2 Reason for not treating maize with chemicals - No need 13 46.4 17 54,8 - No money to buy chemicals \ 8 28.6 1 12.9 - Chemicals are unavailable ' v 2 7.1 5 16:1' - Others Method of seed storage - Shell and put in storage 19 26.0 74 26.9 • Store it on crips with out shelling 44 60.3 154 56.0 - Others Criteria considered in the selection of maize seeds - Big cobs 16 21.3 41 14.9 - Big cobs + mature grain 21 28.0 81 29.5 • Big cobs + seed color 12 16.0 38 13.8 Big cobs + mature grains + seed color 17 17.3 54 19.6 is seed readily available - Yes 16 30.2 135 61.4 -N o 37 69.8 85 38.6 Use own maize seeds -Yes 74 97.4 260 90.0 4.243** -N o 2 2.6 29 10.0 Purchase seeds regularly - Yes 7 13.0 114 49.4 23.721“ * -N o 47 87.0 117 50.6 Reasons for not purchasing improved seeds - Cannot afford to buy 10 55.6 14 29.2 - Unavailability 3 16.7 23 47.9 Are maize products used for income generation -Yes 6 11.3 31 14.1 0.280NS -N o 47 88.7 189 85.9 *** = Significant at 1% ** - Significant at 5%; NS = Nonsignificant at less than 10 P f f f mm * r* > p

tV? l 18 T 1 Li 61 Ra t e o f ADOPTioN o l lilPl ■ CHEMICAL FERTIUZEl ; ■:% . ,r :' -i he common procedure usedl f ^ Karate of- adoption was a Ipj^^ curve, which capliires the histori* fejjti^n^er a giVSn tim ^'TO? Togisric curve is ijisft * prpportioh of farmers who had adopted i^rove# maize and crai |r Jrer a give perip4, The/< bsi i Lssumplion fbUowe^fifo ^n lc p c ^ if HBjs 3 fc w e was *fchat adoption in

993). Mathematically the Jogistic curVi fei wessed fey the following fomiula^ r - K - : " '.(I.'* " I ■ H .(\

Adoption Rate of;lifiprev< M $■ The adoption rate of improved i s^from Iplfr than 1% in 19*76 to" <>3% in 1998. The, adoption rate; jce^ly ovcr the last five year$ aS'dtf tie commencement of/rijiti(|nal ..program. In 1995, when -t$i program was, ^ j l j l in; 19^8 (Fig 2). The growjft lr ?ite of adoption of fcaL,, b:^ the logistic model was abo||t /8%. Though diJRferen^variytifesi i4er^n|r| iii the study area, the varieti|is a dopted by 58% of adapters a$ •$30. Th^* preferred varieties asc r sported by 37%, $9% ahd 2l%1of 5doj>] ^re £H-€60> A-511, and B tt-5 # :spectively. pisf^jShglh# ejneirttioned improved maize vyere; • high yielding abili#: f v • tolerance to lodging; , v; • better germinatiotii and • quality and Higher stover yi^ld. 1 'it

The study revealed that abput 80% of i le s| xers in t998 have adopted one or more of improved mafc^introdticei }area finable 15). Only 2Q% Pf :he farmers reported to haVe never ia$ic

S. ■■ Year Started Planting Improved Maize Variety

Figure 2. Adoption rate of improved maize in the study area

Adopters reported that they obtained improved maize seeds from different sources such as extension, neighbors, relatives, and service cooperatives. About 82%

Adoption Rate of Chemical Fertilizer

In the study area soil fertility is a constraining factor for maize production. Almost all adopters and 95% of non-adopters revealed that their maize field require fertilizer (Table 16). It was identified that, 96% of adopters and 55% of non-adopters applied fertilizer on maize during the 1998 production period. The use of chemical fertilizer is systematically associated with adoption of improved maize (x*= 85.263; p <0.01) indicating that farmers capable to apply chemical fertilizer were more likely to adopt improved maize than diose who were unable to do so.

The history of chemical fertilizer use dates back to 1960. The growth rate of adoption of chemical fertilizer was found to be 77% (fig 3). The study also revealed that use of credit has resulted in facilitating the adoption of chemical fertilizer. A systematic association between access to credit and adoption of improved maize (x =69.204; p<0.01).

Currendy 38% of non-adopters and 61% of adopters are using basal method of application for DAP and top dressing for urea. During the survey year it was reported that only 3% of adopters and 31% of non-adopters used broadcasting method of fertilizer application, which is not promoted by research and/ or extension. 40 Tesfaye and Alerm

Table16. Adoption of chemical fertilizer by maize farmers

Characteristics NcMi­ Adopters X^statistic ado sters N % N % Ever applied chemical Yes 25 34.2 251 87.8 93.704*** fertilizer No 48 65.8 35 12.2 Method of application as Broadcast 53 85.5 200 72.2 NA started using chemical Basal 4 6.2 17 6.1 fertilizer Broadcast + Basal 1 1.5 20 7:2 Know fertilizer Yes 10 19.6 224 85.5 98.665*** recommendation by No 41 80.4 38 14.5 research/extension Get chemical fertilizer Yes 48 71.0 191 66.6 0.645NS on time No 20 29.0 96 33.4 Used chemical fertilizer Yes 41 55.4 275 95.5 85.263*** in 1998 No 33 44.6 13 4.5 Types of fertilizer used DAP 15 36.6 18 6.9 34.335*** in 1998 DAP + Urea 26 63.4 255 93.1 Method of application Broadcast 10 31.2 9 3.4 43.350*** used in 1998 Basal 7 21.9 62 23.3 Top dressing 2 6.3 12 4.5 Basal + Top dress 12 37.5 161 60.5 Ever discontinued using Yes 14 20.9 40 14.1 1.932NS chemical fertilizer No 53 79.1 244 85.9 • = Significant at 10 %; • * = Significant at 5 %: and • * * = Significant at I %.NS=Not significant NA= Not applicable.

8(5% and only 20% of adopters and non-adopters knew the recommended fertilizer rate by research and extension. The analysis done to see if there is any systematic association between being knowledgeable about the recommendation and adoption of improved maize showed that fanners who have adopted improved maize are iriore likely to know die recommended fertilizer rate than non-adopters (X=98.665;P<0.01). Adoption of improved maize and chemical fertilizer 41

Y ear

Figure 3: Adoption rate of chemical fertilizer in the study area

Sixty-seven percent of adopters and 71% of non-adopters reported that they get fertilizer on time. The chi-square analysis does not show systematic association between adoption and obtaining fertilizer on time. In addition, it was reported by 79% of non-adopters and 86% of adopters that they have never discontinued using chemical fertilizer once they started using it. About 96% of adopters reported to have used chemical fertilizer in 1998. The use of chemical fertilizer was systematically associated widi adoption of improve maize variety (x*=85.263; p<0.01). 93% oi adopters reported that they use both DAP and urea. 37% of non-adopters and 7% of adopters used only DAP. The adoption of improved maize is systematically associated with the types of chemical fertilizer used. Adopters appeared to use both ; while non-adopters use only DAP (xi=34.335; p<0.01). The major source ol fertilizer as reported by 63% of non-adopters and 72% of adopters were the ministry of agriculture and Ambasel. 42 Tesfaye and Alemu f a c t o r s A f f e c t in g A d o p tio n o f MAIZE TECHNOLOGIES

Logit Analysis for Improved Maize

ogistic model estimates of factors affecting adoption of improved maize in the Lstudy area were computed (Table 17). The model indicates that 90% of the total variations in the model was explained by the logistic model, which was very good for cross sectional data. Figures for correctly predicted adopters were 96% and for non­ adopters it was 68%. The chi-square shows that the parameters included in the model were significantly different from zero at less than 1% probability level.

Among the 14 factors considered in the logit model, only six had significant and positive influence on adoption of improved maize at less than or equal to 5% probability leve». These were applying chemical fertilizer, access to credit, access to extension information, distance from development center, distance from market center and family size. Of these factors, contrary to expectation distance from development center and that of market center were positively related to the probability of adoption of improved maize.

The level of education was found to have no significant influence on the adoption decision of farmers for improved maize. This might be because information about improved maize was already available to all farmers through the new extension package and regular extension programs. Tliis result agrees with the finding of Chilot, et al (1996), that level of education of farmers has no impact on the adoption decision of modem varieties of wheat varieties in Addis Alem Areas of Ethiopia. However, Getahun, et al (2000) and William et al (1997) found out that the odds in favor of adopting improved maize increased by a factor of 17.7 and 4.26 for literate and illiterate fanners, respectively.

With the application of chemical fertilizer, the probability of adopting improved maize increased by a factor ot 16.6 indicating that farmers applying chemical fertilizer are more likely to adopt improved maize than those who were not using it.

Access to credit was also found to influence the probability of adopting improved maize by a factor of 10.6. Getahun, et al (2000) and also confirm that credit has a statistically significant impact on farmers’ choice to adopt improved maize. Adoption of improved maize and chemical fertilizer 43

Table 17. Parameter estimates of a logistic model for factors affecting adoption of improved maize

Explanatory variables Parameter Wald statistic Exp (B) Mean values estimates Use hired labor 0.9989 2.4168 2.7154 Adequacy of draft power 0.3282 0.5252 1.3884 Use chemical fertilizer 2.8098 36.1625 16.6073 Access to credit 2.3580 *** 31.2472 10.5993 Attend a field day or visit a 0.7136 0.4196 2.0414 demonstration plot Attend a formal agricultural training -0.8297 0.5686 0.4362 Access to extension messages 2.2824 ** 4.1089 9.8001 Distance to development centers 0.0206 * 3.6140 1.0209 30.0 (minutes) Distance to market centers (minutes) 0.0104*** 6.8161 1.0105 80,0 Level of education 0.0476 0.1982 1.0488 1.53 Farming experience (years) 0.0083 0.2631 1.0083 17.0 Family size 0.2450 ** 5.2322 1.2777 5.68 Tropical livestock units (TLU) -0.0297 0.1226 0.9707 4.57 Total farm size (ha) -0.1454 0.6521 0.8647 1.63 Constant -5.5305 1.8295 Model X2 189.480 *** Overall cases correctly predicted 90.24 % ' Correctly predicted adopters 96.21 % Correctly predicted non-adopters 68.35 % Sample size 369 Note: * = Significant at 10 %; * * = Significant at 5 %; and * * * = Significant at I %. NS = Nonsignificant at less than 1%.

A similar result was also found with regard to access to extension information. The probability of adopting improved maize increased by a factor of 9.8 % among farmers who had access to extension information. This agrees with the finding of Kaliba, et al ( 998) and Chilot et al (1996) that extension contact has a positive and significant influence on the proportion of land allocated to improved maize in central Tanzania, and adoption of improved wheat varieties in Addis Alem areas of Ethiopia, respectively.

Household size was also found to have significantly influenced the probability of adoption of improved maize. Households with above average family size are more likely to increase the adoption of improved maize by a factor of 1.28 than those households with less than average family size.

The regression coefficients and the model were used to calculate predicted probabilities of adoption of improved maize by keeping the continuous variables constant at their mean levels and the dummy variables at zero. The predicted probabilities showed the likely effects of changes in the significant variables. The changes in the probability of adopting improved maize as a result of changes in using chemical fertilizer, getting access to credit and extension services; distance to development center and market center; total number of persons in the household were significant. 44 Tesfaye and Ale mu

The probability of adopting improved maize among farmers with average values of continuous variables included in the model was about 5%. With access to credit the probability of adopting improved maize increased to 38%. The use of chemical fertilizer as complementary input increased the probability that a farmer would adopt chemical fertilizer to 49% (Table 18).

Table 18. Impact of significant factors on predicted probabilities of the use of improved maize among sample fanners

Factor Changes in probabilities (%) Application of chemical fertilizer No 5.0 Yes 49.0 Access to credit -No 5.0 -Yes 38.0 Access to extension information -No 5.0 -Yes 36.0 Family size - Average 5.0 -Seven 7.4 -Nine 11.5 - Eleven 17.5

Logit Analysis of Chemical Fertilizer

The logistic model estimate of factors affecting adoption of chemical fertilizer use in the study indicates that 85% of the total variation in the model wasexplained by the logisuc model (Table 19). Figures for correcdy predicted adopters and non-adopters were 94% and 58%, respectively. The parameters included in the model are significandy different from zero at p

Among the 14 factors postulated to influence the adoption decision of chemical fertilizer use, six were found significant. These are: • access to credit; • visit of a demonstration plot; • attendance of formal agricultural training; • access to extension information; • distance to market center, and • family size and use of improved maize.

Of these factors, cot . i p - to anticipation, attendance of field day and access to extension information .verc negatively and significantly related to the adoption Adoption of improved maize and chemical fertilizer 45

decision of chemical fertilizer use. However, other studies have found out that an increase in the intensity of extension service have positive and significant effects on the probability of fertilizer use (Kaliba, et al (1998)

In this study, farm size, diough positive, was not found to have a significant influence on the adoption decision of chemical f ertilizer. Mulugeta (1994) also confirmed that it has a positive but insignificant effect on fertilizer adoption of wheat in the south eastern highlands of Ethiopia. Getahun, etal, (2000) also found out that it has positive and significant impact on the adoption decision of chemical f ertilizer.

Access to credit and use of improved maize are the most important factors found to positively and significantly influence die adoption decision of chemical fertilizer. Access to credit has increased die probability of adopting chemical fertilizer by a factor of 2.8.5, indicating that farmers who have access to credit are more likely to use chemical fertilizer compared to those who have no access to credit.

The same result was also obtained widi die use of improved maize, i.e., die probability of adopting chemical fertilizer increased by factor of 13.50 among farmers who used improved maize. The result also revealed diat farmers who use improved maize are more likely to use chemical fertilizer.

In this study, though not significant, distance from development center was negatively related to die probability of adoption of improved maize. Chilot, et al (1996) also validated diat distance of extension center from residence of respondents was negatively and significandy related to die intensity of fertilizer use in Addis Alem areas of Ediiopia.

As anticipated, distance from market center was found to be negatively and significandy related to die probability of adoption of chemical fertilizer (by factor of 0.99). This means diat die probability of adoption of chemical fertilizer use increases among households living in adjacency to market center than those households far way. This was probably because households living nearby market center have easy access to purchase inputs and dispose of outputs thus leaning toward adopting improved technologies.

The probability of adopting chemical fertilizer was enhanced among households with smaller family size than among larger households. This could be probably because larger households lack cash to buy chemical fertilizer. However, Mulugeta (1994) disagree with diis finding. He found out diat die number of persons in the household have both positive and significant effect (5% level) on die adoption of chemical fertilizer in wheat in die southeastern highlands of Ediiopia.

The regression coefficients and die model were used to calculate predicted probabilities of adoption of chemical fertilizer for change in the significant explanatory variables. Probabilities were calculated keeping die continuous variables 46 Tesfaye and Alemu constant at their mean levels and die dummy variables at zero. The predicted probabilities showed the likely effects of changes in the significant variables.

Table 19. Parameter estimates of a logistic model for factors affecting adoption of chemical fertilizer among sample farmers, in the study area

Explanatory variables Parameter Wald exp (b) Mean values estimates statistic Use hired labor 0.1298 0.0942 1.1387 Adequacy of draft power -0.0928 0.0578 0.9^14 Access to credit 1.0464*** 5.9984 2.8475 Attend a field day or visit a demonstration -1.4130*** 6.0380 0.2434 plot Attend a formal agricultural training 1.3862*** 6.6211 3.9995 Access to extension messages -1.1269* 3.1753 0.3240 Distance to development centers -0.0091 1.86773 0.9909 30.0 (minutes) Distance to market centers (minutes) -0.0141*** 20.7138 0.9860 80.0 Level of education 0.0321 0.1253 1.0326 1.53 Farming experience (years) 0.0004 0.0008 1.0004 17.0 Family size -0.1690** 4.1210 0.8445 5.68 Tropical livestock units 0.0455 0.4944 1.0466 4.57 Total farm size (ha) 0.0831 0.2343 1.0867 1.63 Use improved variety 2.6035*** 35.5281 13.5115 Constant 0.8993 0.2028 Model X2 140.278*'* Overall cases correctly predicted 85.09 % Correctly predicted adopters 94.22 % Correctly predicted non-adopters 57.61 % Sample size 369

Note: * = Significant at 10%; ** = Significant at 5 %. and • • * = Significant at I %

The probability of adopting chemical fertilizer among fanners with average values of continuous variables included in the model is about 26%. With access to credit the probability of adopting chemical fertilizer increase to 50%. The use of improved maize increased die probability that a fanner would adopt chemical fertilizer to 82% (Table 20). Adoption of improved maize and chemical fertilizer 47

Table 20. Impact of significant factors on the predicted probabilities of chemical fertilizer use among sample farmers

Factor Changes in probabilities (%) Access to credit -No 26.0 -Yes 50.0 Attend formal agricultural training -No 26.0 -Yes 58.0 Use improved maize -No 26.0 -Yes 82.0 48 Tesfaye and Alemu

C o n c l u s io n a n d recommendations

he study revealed that access to extension information is systematically associated with adoption of improved maize. Thus, it could be concluded that Tfanners who have got information through different means are more likely to adopt maize technology than those farmers who did not have access. Hence, the existing extension package program need to be strengthened more to increase the flow of information to fanners. More demonstration sites both for improved technologies and fertilizer application should be organized in order to create awareness am .g fanners about the new technologies. The yontact between extension agents and lanners must be strengthened more by reducing the ratio between farmers and development agents. The program should provide more transport facilities to development agents to increase their capacity to travel within their mandate area. In addition, frequent training must be organized for development agents and supervisors about existing and newly developed improved technologies and new methods of agricultural practices. Tliis is expected to develop the confidence of the agents to transmit appropriate and useful information to farmers

Improved maize (bodi hybrids- and composites) are responsive to fertilizer and fanners do obtain economic yields with fertilizer. But the rate of growth of fertilizer adoption growth rate is high particularly alter the implementation of PADETS. However, use of fertilizer was constrained by high price of fertilizer and farmers lack of knowledge about how to use it. An efficient marketing system for input and output will benefit fanners by paying higher prices for maize and reducing the cost of fertilizer.

Farmers who have access to credit tend to adopt improved maize and chemical fertilizer more than those fanners who did not have access to credit. Shortage of rural credit appeared to be a key factor limiting use of purchased inputs and investment. The most important credit problems in the study area were identified as unavailability of loan from formal and informal sources, high interest rates, and unfavorable loan repayment terms. It has been noted that with rising input prices fanners’ need for credit has become cnicial.

Absence of credit for marketing and other functions also may constrain rural development, farmers income and food security. The rural credit organization in the region presently call for agricultural input loans to be paid back at harvest time, forcing farmers to sell when prices are low, rather than storing and selling when prices are higher and food is in short supply. There does not appear to be any regional or federal government policy preventing such institutions for extending loans for a longer period (for at least one year than six months). Increasing the loan repayment period to one year could help to facilitate storage and orderly marketing. The availability of credit services for marketing, storage and processing facility Adoption of improved maize and chemical fertilizer 49 development could further promote die development of competitive output marketing system.

Incentives available tiirough market for surplus output and availability of cash through market sales play major role for adopting technologies. This indicates diat government policies for die expansion of output market networks and provision of inducement for die spread of output marketing services can promote die use of technologies by households.

Because of poor bargaining power of small scale producers, high underdeveloped market information which tends to harm more the smallholder in times of market crises, poor rural roads, concomitant high transport costs that denies small farmers access to major markets, lack of storage facilities that force fanners to sell dieir surplus widiin 2-3 month? after harvest helps merchants to easily establish a monopolistic role in times of good harvest widi detrimental effect 011 smallholder production. To insulate maize farmers from market forces and to provide incentives for investment 011 maize production through die adoption of maize technologies, it is necessary' to introduce and implement a system of floor prices supported by appropriate market intervention mechanisms; while at die same time leaving die market free to operate under competitive conditions.

Limited degree of competition in fertilizer marketing was also observed in die study area. This could limit the use of fertilizer. The participation of traders or retailers in marketing fertilizer and improved seed is very limited because of uncertainty and unfamiliarity of agents in marketing inputs, diough transport costs, storage and die need for credit to finance working capital are also important constraints. Therefore , efforts must be made to promote the participation of small traders and retailers in input distribution.

Limited infrastructure development is also another factor limiting the production and distribution of maize. This obviously limits farmers capability to participate in market transactions, and explains the predominance of subsistence maize farming in the region. Widiout improvements in market access, impacts from improved technologies, improved availability of inputs, credit, and price liberalization may be trivial for maize farmers in remote areas, because of high transportation and marketing costs.

The study also has revealed dial the mean farm size of sample farmers included in die study was 1.57 ha for adopters of improved maize and 1.46 ha for non- adopters. Under the present Ethiopian constitution, all land is die property of die state, and it cannot be sold or mortgaged. The right of peasants of free access to land is guaranteed. Although die constitution has resolved some issues, it does not address some important issues. Given die scarcity of land, it is not clear how much peasants’ right of free access to land can be assured in practice, and what effect diis may have on tenure security of diose currendy possessing land. Nor is it clear how much land 50 Tesfaye and Alemu peasants are entitled to. These issues have been left to regional governments to resolve, and there has been important differences across the regions. In Amhara , a general land distribution was completed some times back and no policy has been established regarding future distributions. Given these circumstances, tenure insecurity may be more of a constraint to land improving investment and technology adoption in ANRS.

The present land policy (federal or regional) could: • constrain effort to reduce land fragmentation; • limit fanners’ ability to obtain sufficient income from farming; • limit incentives to invest in land improvements; • constrain fanners’ ability to take advantage of better economic opportunities outside of farming or in other locations; and • inhibit land mortgaging, thus, reduce farmers’ collateral and access to collateral-based credit.

Therefore, in order to promote agricultural production and productivity in particular and rural development in general, there is a need to study in detail the existing land issue in the region and improve the situation based on the findings. Adoption of improved maize and chemical fertilizer 51

References

Amemiya, T.1981. Qualitative Response Models: A Survey. Journal of Economic Literature 19: 1483-1536. BOPED (Bureau of Planning and Economic Development). 1999- Adas of Amhara National Region. Bahir Dar. Bureau Planning and Economic Development. 1999- Atlas of the Amhara National Region, Bahir Dar. CDE (Center for Development and Environment) and MOA, 1999. Soil Conservation Research Program. Data Base System, ETHIO-GIS. Vol. 2. Chilot Yirga, B.I. Shapiro, and Mulat Demeke. 1996. Factors Influencing Adoption of New Technologies in Wolmera and Addis Alem Areas of Ediiopia. Ethiopian journal of Agricultural Economics. Vol. I. pp .63-84, Agricultural Economics Society of Ethiopia. Addis Ababa, Ethiopia. CIMMYT Economics Program. 1993- The Adoption of Agricultural Technologies: a Guide to Survey Design, Mexico DF: CIMMYT CSA (Central Statistical Authority). 1990. Statistical Abstract. CSA, Addis Ababa. CSA. 1992. Statistical Abstract, CSA, Addis Ababa. CSA. 1994. Population and Housing Census of Ethiopia, Result at National Level. Volume 1, Statistical Report, CSA, Addis Ababa, June 1998 CSA. 1995. Statistical Abstract, CSA, Addis Ababa ,April 1996. CSA. 1997. Statistical Abstract, CSA, Addis Ababa , March 1998 CSA. 1998. Statistical Abstract, CSA, Addis Ababa, February 1999. CSA. 1999. Report on Area and Production of Major Crops, Statistical Bulletin CSA, Addis Ababa. March 1999. EARO. 1999- Maize Research Strategy. Addis Ababa, Ethiopia EARO. 2000. Crop Research Strategy. EARO, Addis Abeba, Ethiopia Epoug. J.1996. Linkage between research an technology users. Some issuers from Africa. 1SNAR Briefing paper No.30, The Hague, The Netherlands. ESE(Ethiopian Seed Enterprise). 1999- 1998/99 Business Plan. September 1998. ESE. 1999. Seed System Development Project Status Report As of June 1999- Feder, G., R.E. Just and D. Zilberman. 1985. Adoption of Agricultural Innovations in Developing Countries: A Survey. Economic Development and Cultural Change 33:255-298. Getahun Degu, Wilfreed Mwangi, Hugo Verkuijl, and Abdushukur Wondimu, 2000. An Asessment of the Adoption of Seed and Fertilizer Package, and the Role of Credit in Small Holder Maize Production in Sidama and North Omo Zones of Southern Ethiopia. Mexico, D. F.: International Maize and Wheat Improvement Center (CIMMYT) and Ethiopian Agricultural Research Organization (EARO). Kaliba, A.R.M.,H.Verkujil, W. Mwangi, A.J.T. Mwilawa, P. Anandajayasekeram and A.J.Moshi.1998. Adoption of maize production technologies in central Tanzania. Mexico, D.F.: International maize and wheat improvement center (CIMMYT), the United Republic of Tanzania, and the Southern Africa Center for Cooperation in Agricultural Research (SACCAR). Kebede Mulatu, Gezahegn Bogale, Benti Tolessa, Mossisa Worku, Yigzaw Desalegn and Asefa Afeta. 1993- Maize Production Trends and Research in Ethiopia. In: Benti Tolessa and Joel. K Ranson (eds). Proceeding of the First National Maize Workshop of Ethiopia, 5-7 May 1992,Addis Ababa, Ethiopia. LAR/CIMMYT, Addis Ababa. 52 Tesfaye and Alemu

Ministry of Economic Development and Planning (MEDAC). 1999-Survey of the Ethiopian Economy. Review of Post Reform development (1992/93-1997/98). MEDAC, Addis Abeba, Ethiopia. Mosisa Worku, Jemal Abdushikur, Leta Tulu, Haji Tuna, Legesse Wolde, Kasa Yilma, Wonde Abera, Aschalew Gutu, Sewagegne Tariku, Teshle Asefa, Tamirat Birhanu, Yoseph beyene, and Habtamu Zeleke. 2001. Improved germplasm development for the mid and low altitude Sub-humid agro-ecologies. Paper presented at the Second National maize Workshop of Ethiopia. 12 -15, November 2001. Addis Abeba, Ethiopia Mulugeta Mekuria. 1994. An economic analysis of small holder wheat produ'tion and technology adoption in the southeastern highlands of Ethiopia. Pn.D. thesis. Department of agricultural Economics, Michigan state university, USA NSLA( National Seed Industry Agency.2000. The seed sector basic data. Issue No 2. Addis Abeba, Ethiopia. September 1999.Addis Ababa. Tenassie Nichola 1985.Agricultural Research and extension in Ethiopia .the state of the Art IDR. Report No. 22 Addis Ababa UNECA (United Nations Economic Commission for Africa). 1996. Sustainable Agriculture and Environmental Rehabilitation Program (SAERP). Statistical Master Book on Sectoral Conditions and Activities in the Amhara Regional State. Vol. 1. July 1996. UNECA, Addis Ababa. William Ntege-Nanyeenya, Man.' Mugisa-Mutetikka, Mwangi and Hugo Verkuijl. 1997. An assessment of factors affecting adoption of maize production technologies in Iganga, Uganda, NARO/IIMMYT.