Regional Wheat Research for Development

Edited by Tadesse Dessalegn

East African Agricultural Productivity Project Wheat Regional Center of Excellence

vhlw? fiflc? rcrc Ethiopian Institute of Agricultural Research Regional Wheat Research for Development

Proceedings of the 3ld Regional Wheat Progress Review September 2014, Adama, Ethiop

Copyeditor: Abebe Kirub TABLE OF CONTENTS Acronyms and Abbreviations 2 Words from ASARECA 3 Genotype by Environment Interaction and Yield Stability in Bread Wheat Genotypes in East Africa 6 Evaluation of Spring Bread Wheat Advanced Lines across Different Environments of Tanzania 15 24 24Responses of Bread Wheat Genotypes to Fertilizer and Seed Rates Arsi Zone, Ethiopia 24Responses Responses of Bread Wheat Varieties to N and P Moist and Humid Midhighland Vertisols of Arsi Zone, Ethiopia 35 Monitoring of Major Foliar Diseases ofWheat in Ethiopia, Kenya, Tanzania and Uganda 46 Evaluation of Herbicides in Wheat 55 Developing and Introducing Pre-Harvest Implements 61 Developing and Fabricating Small-scale Wheat Thresher 69 Disseminating Improved Wheat Technology through Pre-Extension Demonstration in Ethiopia 72 Determinants of Fanners’ seed demand for improved wheat varieties in Ethiopia: A Double Hurdle Model Approach 84 Enhancing Adoption of Improved Wheat Technologies, Innovations and Management through Dissemination, Up-Scaling and Knowledge Management 94 Wheat Production Efficiency in Major Producing Areas of Ethiopia 103 Developing manually-operated single row precision Wheat-Cum-Fertilizer planter 112 Distribution, Physiologic Races and Reaction of Wheat Cultivars to Virulent Races of Leaf Rust in Southeastern Zone of Tigray, Ethiopia 125 Durum Wheat Research and Achievements 140 Yield and Yield Stability of Bread Wheat Genotypes 149 in Lowland Irrigated Areas 149 Evaluation of Bread Wheat Genotypes for Yield and Yield Components in Irrigated Lowland Areas 155 Closing Remarks 159 List of participants 161

1 Acronyms and Abbreviations

AFrll African Innovation Innovative AMMI Additive main effect and multiplicative interaction APPRC Ambo Plant Protection Research Center ARARI Amhara Region Agricultural Research Institute ( ASARECA Association for Strengthening Agricultural Research in East and Central Africa ASV AMMI stability values BBM Broad Bed Maker CADDP Comprehensive Africa Agriculture Development Program CIMMYT Centro Intcmacional dc Mcjoramiento de Mai/ y Trigo (International Maize and Wheat Improvement Center) DRRW Durable Rust Resistance in Wheat EAAPP Eastern Africa Agricultural Productivity Project EIAR Ethiopian Institute of Agricultural Research Eol Expression of Interest ESMP Environmental and Social Management Plan FARA Forum for Agricultural Research in Africa FIG Farmer innovation grant system, FREG Farmer Research Extension Group GEI genotype by environmental interaction GTP Growth and Transformation Plan ICARDA International Center for Agricultural Research in the Dry Areas IPCA Interaction Principal Component Axis IPMP Integrated Pest Management Plan KARLO Kenya Agricultural Research and Livestock Organization MARC Mekele Agricultural Research Center MMT Million Metric Tons MRR Marginal Rate of Return NEPAD New Partnership for Africa Development NIR Near Infrared Reflectance NRI Natural Resources Institute RCoE Regional Center of Excellence SNNP Southern Nation Nationalities People SSA Sub-Saharan Africa T&D Training and Dissemination USAID United Stales Agency for International Development WRCoE Wheat Regional Center of Excellence

2 Words from ASARECA Vincent Akulumuka1 and Apophia Muhimbura2 1 EAAPP Program Manager; ASARECA-Uganda 2 Environmental and Social Safeguard Specialist; ASARECA-Uganda

What is cxpectcd from the reviews: Reviews of the regional projects implemented under the RCoEs jointly by researchers from the 4 EAAPP implcmeting countries is an annual event which track the implementation progress of the past one year. The wheal review is the last commodity reviewed in 2014. Others reviewed included project under dairy RCoE, which took place in May 13-15, 2014, followed by projects under rice RCoE in July 16-18, 2014 and projects under cassava RCoE in July 28-30. 2014. It should be noted that these reviews arc the last carried out in phase I of EAAPP. The main expectations from the 2014 wheat review included:

° Monitor regional integration in terms of seeing how regional projects arc clearly demonstrating regional dimensions such as addressing regional priorities, demonstration of RCoE leadership in ensuring the agreed designs are followed, communication among research team is enhanced, fostering information sharing, and regional projects have clear focus on the “end products” such as new varieties, innovative technology pathways, clear strategy for publishing results into journals and other media; • Quality of presentation in terms of visibility, clarity, content, andengaging the audience; and • Reflection and way forward

The Roles of RCoEs Experience gathered from the past reviews show that the roles and responsibilities of the RCoE arc somewhat forgotten and hence a need for reminders is apparent. In that respect, the agreed roles and responsibilities of the Wheat RCoE developed in November 2012 during the Mid Term Review Implementation Support Mission was shared again to the scientists and development specialists presen in this review. It was agreed for the Wheat RCoE to:

o Lead in the development of wheat technologies; o Coordinate the implementation of all pro jects' activities; o Provide training to scientists and other stakeholders; o Enhance access of proven technologies by other countries; o Ensure sharing of information and knowledg across countries; and o Strengthen linkages with national, regional and international institutions

Planning, implementing and reviewing projects It was felt important to also revert to the agreed steps to follow when planning, implementing, reviwe ng and concluding regional projects. This is thought important because EAAPP phase I is coming to a close and transition to phase II is yet uncertain. Thus, the steps agreed by wheat stakeholders in 2012 were:

• Identifying regional priority areas with participation of partner countries; • Developing concept notes on identified priority areas; 8 Approving regional projects using the agreed procedures under the RCoE; • Implementation of the projects; and • Reporting in the reviews and other fora

Features of regional projects Alongside the planning and review process, also agreed on were the key features qualifying the regional projects as:

• they should be implemented in more than one EAAPP implementing country;

3 • they should demonstrate the potential for improving regional intergration; • they should have potential for spill-overs; • they should have potential for building new strategic partnerships and in addition they should have potential to positively impact on economic growth, social welfare, natural resource management, and capacity building

Rotes sind responsibilities of Principle Investigators under Wheat RCoE Principle Investigators are key in spearheading the implementation of the regional projects so as to ensure implemcnters are on track. In that respect, it was fell necessary to remind them of their roles and responsibilities and more so as the project comes close to the end of Phase I. The roles are:

• Coordinating project planning and implementation; • Monitoring implemetation to achieve set objectives and results and promote(?) the findings to wider stakeholders; • Developing practical linkages amongst project implementation teams; • Enhancing information and knowledge sharing; • Organizing meetings and training sessions; and • Preparing technical reports

End of EAAPP Phase I Evaluation As EAAPP comes closer to the end of phase I, it is a pre-requisite to carry out an independent end of phase I e\aluation. The workshop participants were informed that a consulting firm formed by the Natural Resources Institute (NRI) of the University of Greenwich (UoG) based in the United Kingdom had won the award for carrying out the evaluation. NRI/UoG had teamed up with the African Innovation Iniative (Afrll) based in Kampala Uganda. The consulting firm had its first inception meeting in Naivasha during the 8lh EAAPP mission wrap-up and had one on one discussion with Kenya, Ethiopia and Tanzania teams. The focus of the evaluation will include the following:

• Undertaking the economic analysis and assessing the implemnted projects paying attention to the number of direct and indercct beneficiaries reached, and quantities of benefit accrued • Assessing EAAPP performance in meeting its project Development Objectives (PDO) as stipulated in its results framework. Key Performance Indicators at PDO level are:

o rate of change in regional specialization and collaboration in agricultural research; o rate of increase in information and knowledge transfer across national boundaries; o rate of change in adoption of new technologies; o rate of change in increase of land areas with seed of improved cultivars; o increase in productivity at farm level over control technology; and o level of stakeholders satisfaction with technologies and innovation • Generating key lessons learned that could be used to inform the design of EAAPP Phase 2.

Transition of EAAPP from Phase I to Phase II Workshop participants were informed that the High Level Leaders (Permanent Secretaries, Principal Secretaries, State Minister) of the four countries implementing EAAPP are highly satisfied with the implcmenttion progress the project is achieving. The leaders registered this statement during the 3rd EAAPP Regional Steering Committee meeting held in Bahar Dar Ethiopia in February 2014. In that meeting, iJie leaders agreed to submit the Expression of Interest (Eol) to the World Bank for phase II. Kenya and Tanzania have submitted, Ethiopia's Eol is with the Ministry of Finance for approval and submission to the Bank and Uganda is at advanced stage of submission. No official communication has come from the World Bank concerning Phase II. Meanwhile, during the conclusion of the 8th EAAPP mission by the World Bank and ASARECA, it was recommended that countries should apply for a No Cost Extension (NCE) to meet the following:

4 • Bridging the gap between Phases I and II; • Providing opportunity for the implementing countries to document achievements, success stories and lessons learned; and • Developing a strategy for continuity of the initiated activities in case Phase II docs not materialize

It was noted that there was significant progress in compliying with ESS with the;

9 National Action plans submitted for Kenya, Ethiopia and Uganda; • EIA for civil works carried out across the region; • ESS instruments development improved but still lacking especially for Ethiopia; and • The compliance rating had improved from 0% at the beginning of 2013 and stood as follows;

Regional RCcE Civil works projects Compliance rating Comment Wheat 55 42 48% - Not Compliant All regional projects screened but no ESMP or IPMP submitted. EIA done though approval certificates not seen. Half yearly report had minimal ESS content Rice 70 59 64% - Somewhat All regional projects screened, not all ESMP and IPMP Compliant submitted. EIA done. Half yearly report had minimal ESS content Dairy 75 69 72% - Compliant All regional projects screened and all ESMP / IPMP submitted though one was not yet complete. EIA for civil works done. Needed to improve ESS content in the half yearly report to align it with the national action plan Cassava 68 61 64% - Somewhat All regional projects screened, one ESMP not yet Compliant submitted. EIA for civil works done though approval certificates not seen. Half yearly report had minimal ESS content

There was need to improve on:

• Reporting on the status of implementation of the national action plans; • Development and implementation of the ESS instruments - most especially for Ethiopia; and • Implementation of recommendations for EIA for civil works

5 T heme 1 Gerrnplasm Enhancement, Variety Development, and Breeder Seed Production Genotype by Environment Interaction and Yield Stability in Bread Wheat Genotypes in East Africa

Zerihun T1‘, Dawit A1, Habtemariam Z1, Njau P2, Mohammed A3 and Muluken B4 1 Ethiopian Institute of Agricultural Research (EIAR); P.O. Box 2003, Addis Ababa. Ethiopia 2Kenya Agricultural and Livestock Research Organization (KALRO), Private Bag (20107) Njoro, Kenya 3Oromia Agricultural Research Institute (OARI), Ethiopia 4Amhara Regional Agricultural Research Institute (ARARI), Ethiopia

Abstract The existing heterogeneous agroecology in eastern Africa emphasizes the importance of a robust and eftlcien variety testing scheme to identify stable and widely adapted varieties in the region. Twenty-six elite bread wheat genotypes were evaluated to estimate the magnitude of genotype by environmental interact:on (GEI) and yield stability under different environments. The varieties were tested at nine environments in 2013, i.e., eight in Ethiopia and one in Kenya. The experiment was laidout in alpha lattice design with three replications. Additive main effect and multiplicative interaction (AMMI) and AMMI stability values (ASV) were computed. The main effects of environments, genotypes and GEI were highly significant (p<0.01) for grain yield and accounted for 47%, 11.7%, and 25.4% of the total variations, respectively. The highest environmental and GEI effect indicated major influence of environment on genotype performance and inconsistent performance of genotypes across tested environments. The mean grain yield of genotypes across environments was 4.08 t ha'1. Genotypes ‘G8\ ‘G18’, ;uid ‘G30’ were the best performing varieties across environments with mean grain yields of 5.41, 5.14, ard 4.94 t h a 1, respectively. The stability statistics. AMMI. showed that the first four IPCA were highly significant (p<0.01) and explained 91% of the total GEI. Njoro from Kenya and Meraro from Ethiopia showed higher discriminating power of genotypes. In addition, AMMI biplot confirmed that Njoro was different from other test locations, which implied that testing of genotypes at Njoro is very essential to develop stable varieties across the region. Based on IPCA and ASV, ‘G15’, ‘G26’ and ‘G3’, in that order, were stable genotypes across locations. Despite the high contribution of environment to the total variations of the genotypes performance across locations, some high yielding and widely adapted genotypes were identified and confirmed the importance of testing genotypes across the region to investigate best performing and stable varieties for future use.

Introduction

Raising level of poverty, food insecurity, and high rate of unemployment especially among the youth are the three major challenges facing the Eastern Africa sub-region. To alleviate these challenges, the region has to articulatc its development agenda covering all key sectors of the economy. Agriculture is one of ihc sectors considered the most critical economic pillar throughout the region contributing over 45% of the regional GDPs and directly employing over 75% of the population (Negassa et a I. 2012). Wheat is one of the main agricultural commodities in the world. Today, it is growing on more land area than any other commercial crops and it provides 19% of our total available calories (FAOSTAT 2014). Similarly, it is one of the most important cereal crops cultivated in a wide range of agro-ecologies in Eastern Africa i.e., Ethiopia, Kenya. Tanzania, and Uganda. The wheat economies of these countries are characterized by a growing gap in wheat supply amidst rising wheat prices and price volatility. Economic and demographic changes have led to rapid growth in wheat demand and dependency on wheat imports, which have aggravated the vulnerability of nations in the region to political instability. Despite enormous economic and dietary values of the crop in the region, the average yield has remained extremely low. This has been attributed to multifaccted biotic and abiotic factors. The region is mainly known by the epidemic development of wheat rusts and there exists high

6 rale of evolution of new races of the pathogens because of which most of the newly released cultivars succumb to such new pathotypes within a short period leading to the historical boom and boost cycle in wheat production. Poor grain quality, lack of varieties for specific growing conditions (drought, heat, rrigation and acid soils), poor crop management practices were found bottlenecks for increased wheal productivity in East Africa. Moreover, despite huge potentials for irrigated wheat production, little has been done to develop varieties and suitable crop management practices. As a result, the region is a net importer of wheat.

The changing environmental conditions of ihc region, evolution of new races of the rust pathogens, the expansion of wheat to new agroecologies coupled with inadequate wheal varieties available for the different environments necessitate a rigorous and continuous effort to develop improved wheat varieties in the region. Therefore, this research was conducted to evaluate and identify high yielding bread wheat genotypes across location and to assess the nature and magnitude of genotype by environment interaction across Ethiopia and Kenya.

Materials and Methods

A total of 26 elite and four check (Hidasse, Danda'a, Kakaba, Hawi) wheat genotypes from CIMMYT and ICARDA origin w'ere evaluated for grain yield performance across nine different environments of Ethiopia (8) and Kenya (1) with in altitude range of 2120 - 2990 m.a.s.l. The locations were different in altitude, mean annual rainfall, and soil types. These locations represent the major wheat growing agro-ccologies ranging from mid to extreme high altitude. The details of the experimental locations and materials are listed in Tables 1 and 2.

Table Description of the test locations

Location Location Altitude Annual rainfall Agro-ecology code (m) (mm) Kulurrsa KUL 2200 820 Optimum area Bekoji BKJ 2780 1020 Optimum & high RF area Meraro MER 2990 1196 Extreme highland & high RF area Asasa ASS 2340 620 Terminal drought prone area Ginch GNC 2200 1140 Water logging area Sinana SNN 2400 950 Optimum area Adet ADT 2240 869 Optimum area Geregera GGR 2650 1105 Terminal drought prone area Njoro NJR 2120 - Optimum area

The genotypes were planted in alpha lattice (5 x 6) with three replications in all experimental sites luring 2013 main cropping season. Each plot had six rows of 2.5 m length with spaced 0.2 m. Planting dates of each location was on the onset of the main rainy season. Fertilizer and other agronomic practices also carried out as per the recommendation of each location. Grain yield data was recorded on plot basis and convert to ton ha'1 for analysis. The additive main effects and multiplicative interaction (AMMI) model combines the standard analysis of variance (ANOVA) with principal component analysis fits both additive and multiplicative effects for GEI (Gauch and Zobel 1997). It determines the relative magnitude of sums of squares attributable to G, E, and GEI. A biplot was constructed in the dimension of first two princ pal components, using a singular-value decomposition procedure (Yan et al. 2000). The genotypes were represented on the biplots as the points derived from their scores and the environments as the vectors from the biplot origin to their points. The AMMI model (Zobel el al. 1988) is: k-H+bj+Ej+J ^-„a inYjn + eij n=l Where; Yij is the yield o f the ith genotype in the j'h environment; u is the grand mean; Gi and Ej are the genotype and environment deviations from the grand mean, respectively: An is the eigen value of the PCA axis n; ain and yin are the genotype and environment principal component scores for axis n. respectively; N is the number oj principal components retained in the model and eij is the error term.

AMMI model does not make provision for a specific stability measure to be determined, such a measure is essential in order to quantify and rank genotypes according their yield stability. Since the 1PCA 1 (Interaction Principal Component Axis) score contributes more to GEI sum of squares, it has to be weighted by the proportional difference between IPCA 1 and IPCA 2 scores to compensate for the relative contribution of IPCA-1 and IPCA-2 in to the total GEI sum of squares called AMMI stability values (ASV). The following measure was proposed by Purchase (1997):

fPCA\ sum of squares (IPCAl score) ASV = + [IPCA2 score)1 IPCA2 sum of squares

Table 2 Description of genotypes used for the study

Code Designation Pedigree G1 Kakaba KIRITATI//SERI/RAYON G2 Darda’a KIRITATI//2*PBW65/2*SERI.1 B G3 ETRW 6434 PRL/2*PASTOR//PBW343*2/KUKUNA G4 ETEIW 6435 SNB//CMH79A.955/3*CN079/3/ATTILA/4/CHEN/AEGIL0PS SQUARROSA (TAUS)//BCN/3/2*KAUZ G5 ETEIW 6468 CROC_1/AE.SQUARROSA (205)//BORL95/3/2*MILAN/4/TIMBA G6 ETBW 6138 SNI/TRAP#1/3/KAUZ*2/TRAP//KAUZ/4/PARUS/PASTOR G7 ETEIW 6139 SOKOLL/KENNEDY//LANG G8 ETEIW 6161 CROC 1/AE.SQUARROSA (213)//PGO/10/ATTILA*2/9///FNAJ/3/BZA/4/TRM/5/6/SERI/7/VEE#10/8/OPATA G9 ETEIW 6184 MTRWA92.161/PRIN IA/5/SERI*3//RL6010/4*YR/3/PAST OR/4/BAV92 G10 ETBW 6133 BERKUT/3/ALTAR 84/AE.SQUARROSA (219)//SERI G11 ETEIW 6189 BERKUT/EXCALIBUR G12 ETEiW 6320 GIRWILL-13/2*PASTOR-2 G13 ETEIW 6216 HAMAM-4/ANGI-2//PASTOR-2 G14 ETEIW 6224 HUBARA-16/2*SOMAM A-3 G15 ETEIW 6230 HUBARA-16/2*SOMAMA-3 G16 ETBW 6231 HUBARA-16/2*SOMAMA-3 G17 ETEIW 6202 IZAZ-2/TEVEE-2 G18 ETBW 6236 HAIEL-1*2/ABREG-4 G19 ETBW 6266 BOW #1/FENGKANG 15//NESMA*2/261-9/3/DUCULA G20 ETBW 6460 KEA/TAN/4/TSH/3/KAL7BB//TQFN/5/PAVON/6/SW89.3064/7/SOKOLL G21 ETBW 6204 PAST OR-2/BOCRO-2 G22 ETBW 6226 HUBARA-16/24SOMAM A-3 G23 ETBW 6473 LERKE/5/KAUZ/3/MYNA/VUL//BUC/FLK/4/MILAN/6/RABE/LAJ3302 G24 ETBW 6497 D67.2/P66.270//AE.SQUARROSA (320)/3/CUNNINGHAM G25 ETBW 6220 HUBARA-16/2*SOMAMA-3 G26 ETEiW 6271 BOW #1/FENGKANG 15//NESMA*2/261-9/3/DUCULA G27 ETBW 6536 UP2338/3/HE1/3*CN079//2*SERI/5/STAR//PVN/STAR/3/WH 542/4/M ILAN/KAUZ G28 ETBW 6657 FRET2/KUKUNA//FRET2/3/YANAC/4/FRET2/KIRITATI G29 Hawi CHIL/PRL; CM92803-91Y-OH-OSY-5M-ORES-OSY G30 Hidasse YANAC/3/PRL/SARA//TSI/VEE#5/4 ETBW - Ethiopian Bread Wheat

8 Result and Discussion

Mean grain yield performance of genotypes: The mean grain yield of the genotypes across nine locations were 4.08 ton ha'1 with grain yield range of 3.14 ton ha’1 for G21 to 5.41 ton ha'1 for G8 (Table 3). This shows that there was broad diversity in genetic yield potential between tested genotypes. The observed environmental mean grain yield ranged from 2.35 to 6.46 ton ha'1 from Geregera and Bekoji, respectively. Bekoji, Meraro, and Adet were the highest yielding environments with mean values of 6.46, 4.94 and 4.29 ton ha'1, respectively, whereas Geregera (2.35 ton ha'1), Ginchi (3.44 ton ha'1) and Njoro (3.52 ton ha'1) were the lowest yielding environments (Fig 1). All genotypes showed more than mean grain yield at Bekoji, except G21 (4.06 ton ha'1). In contrary, all genotypes showed grain yields below grand mean at Geregera and Ginchi, except G30 at Ginchi. Across location G8 (5.41 ton ha' '), G18 (5.14 ton ha'1) and G30 (4.94 ton ha'1) showed higher mean grain yields. G21 (3.14 ton ha’1), G22 (3.20 ton ha’1) and G29 (3.31 ton ha’1) were lower yielding genotypes across tested locations (Table 3). Figure 1 revealed simple observation of the interaction between genotypes and locations on grain yield performance. The graph shows a clear yield potential difference between environments like Bekoji and Geregera. Moreover, genotypes showed different performance in different environment showing high influence of environmental effect over genotypes performance and leading to interaction effect.

9 Table 3 Mean grain yield (ton ha1) of each location, IPCA 1 and 2 scores, ASV and rank of 30 genotypes tested at nine locations in 2013

Geno KUL ADT MER BKJ GGR ASS NJR SNN GNC Mean Rank IPCA-1 IPCA-2 ASV ASV Rank G1 3.86 4.53 5.18 6.69 2.58 4.52 3.75 4.02 3.67 4.31 9 -0.60 -1.24 2.56 5 Q2 3.48 4.14 ■1.70 6.30 2.20 4.13 3.37 3.63 3.28 3.92 18 -1.13 0.05 4.23 G3 3.89 4.56 5.21 6.72 2.61 4.55 3.78 4.05 3.70 4.34 8 0.46 -1.00 2.00 3 G4 3.97 4.64 5.29 6.80 2.70 4.63 3.86 4.13 3.78 4.42 7 1.26 0.54 4.77 15 G5 4.09 4.75 5.41 6.92 2.81 4.75 3.98 4.25 3.90 4.54 5 1.65 0.99 6,26 17 G6 3.35 4.01 4.66 6.18 2.07 4.00 3.24 3.51 3.16 3.80 19 -3.14 1.36 11.85 28 G7 3.23 3.90 4.55 6.06 1.95 3.89 3.12 3.39 3.04 3.68 23 -1.74 0.25 6.53 19 G8 4.96 5.62 6.27 7.79 3.68 5.61 4.85 5.12 4.77 5.41 1 1.66 1.94 6.54 20 G9 3.86 4.52 5.17 6.68 2.58 4.51 3.75 4.01 3.66 4.30 10 0.90 0.21 3.39 8 G10 3.52 4.18 4.83 6.35 2.24 4.17 3.41 3.68 3.33 3.97 16 0.93 0.24 3.49 9 G11 3.99 4.66 5.31 6.82 2.72 4.65 3.89 4.15 3.80 4.44 6 -3.70 1.49 13.96 30 G12 3.31 3.97 4.62 6.14 2.03 3.96 3.20 3.47 3.12 3.76 20 1.16 -0.65 4.39 12 G13 3.15 3.82 4.47 5.98 1.88 3.81 3.05 3.31 2.96 3.60 25 -1.20 0.27 4.50 13 G14 3.13 3.80 4.45 5.96 1.86 3.79 3.03 3.29 2.94 3.58 26 -1.02 -0.90 3.94 10 G15 3.51 4.17 4.82 6.34 2.23 4.16 3.40 3.67 3.32 3.96 17 0.07 0.72 0.77 1 G16 3.19 3.85 4.50 6.01 1.91 3.84 3.08 3.34 2.99 3.63 24 -0.16 -2.06 2.14 4 G17 3.31 3.97 4.62 6.14 2.03 3.96 3.20 3.47 3.11 3.76 21 2.49 -2.02 9.56 25 G18 4.69 5.36 6.01 7.52 3.41 5.35 4.58 4.85 4.50 5.14 2 3.00 1.02 11.31 26 G19 3.04 3.70 4.35 5.87 1.76 3.69 2.93 3.20 2.85 3.49 27 -1.29 -1.51 5.05 16 G20 3.85 4.51 5.16 6.67 2.57 4.50 3.74 4.00 3.65 4.29 11 1.86 -0.95 7.03 21 G21 2.69 3.35 4.01 5.52 1.41 3.35 2.58 2.85 2.50 3.14 30 0.68 -1.82 3.15 6 G22 2.75 3.42 4.07 5.58 1.47 3.41 2.64 2.91 2.56 3.20 29 -1.71 -0.77 6.45 18 G23 3.84 4.50 5.15 6.67 2.56 4.49 3.73 4.00 3.65 4.29 12 2.09 1.09 7.91 24 G24 3.79 4.45 5.10 6.62 2.51 4.44 3.68 3.95 3.60 4.24 14 -3.48 0.30 13.07 29 G25 3.27 3.94 4.59 6.10 2.00 3.93 3.17 3.43 3.08 3.72 22 -0.61 2.21 3.18 7 G26 3.58 4.25 4.90 6.41 2.30 4.24 3.47 3.74 3.39 4.03 15 -0.18 -1.86 1.98 2 G27 3.83 4.50 5.15 6.66 2.55 4.49 3.72 3.99 3.64 4.28 13 1.97 -0.28 7.40 22 G28 4.31 4.97 5.62 7.14 3.03 4.96 4.20 4.47 4.12 4.76 4 0.88 3.33 4.70 14 G29 2.88 3.54 4.19 5.71 1.60 3.53 2.77 3.04 2.69 3.33 28 -3.14 -0.95 11.83 27 G30 4.49 5.15 5.80 7.32 3.21 5.14 4.38 4.65 4.29 4.94 3 2.02 -0.03 7.59 23 Mean 36.27 4.29 4.94 6.46 2.35 4.28 3.52 3.79 3.44 4.08 CV (%) 12.19 11.36 23.22 15,43 20,04 10,49 18.66 21.34 12.36 LSD (5%) 6.07 6.69 27.12 13.68 6.46 6.16 13.84 11.09 5.83 IPCA 1 -1.36 -1.00 9.35 -0.04 -1.46 -1.08 -1.33 -1.77 -1.31 IPCA 2 2.14 -2.51 0.10 -0.73 -2.15 1.04 5.28 -0.64 -2.52 BEKOJI MERARO ASASSA NJORO KULUMSA SINANA ADET GINCHI GEREGERA

'r0 > 0 '-T r0 Q o o c\jo o o co ro r- n racDOxrco^iocnoo DOJ>tOOOCNJr-COC\JCMO(OfOfO fONO)I3fO(DCONrtCO CO CD w w N(N<(NJCNCNJC'jT-r0C\JCNjT-OJr--O(NTj- B J in ^ T -^ ^ T fi-T r(0 05 cm r- :0 CO -j- CO CO CO CO CO CO CO CO CO CD CO C CO CO CO CO CO CO CO CO CO CO CO ’O CO CO I I S 5 5 § § 5 § 5 § 5 § <3 § 5 5 § § § 5 S 5 5 § § = § § 20 CO CO CO 00 CO CD CO CO CO CD CO CO CO CO CDCOCOCOCDCDCDCDCO CO CO ~~ h— i I— h - I— 1 I— I— 1— I— h~ I— I— h- I— 1— 1— I— I— 1— h— 1— 1— I— |— -LILLI LUUJLUUJLULULULULULULLJ LULU LULULUlUlULULUlULU LULU Fig 1 Genotypes against their yield potential (Q ha-1) across nine locations in 2013

AMMI model analysis: AMMI model analyses of variance showed highly significant (P<0.01) differences for environments (E), genotypes (G) and GEI (Table 4). Mean performance of the genotypes for grain yield differed from location to location. This is obvious phenomena like Ethiopia having diverse agro-ecologies. The significant difference observed among the genotypes also revealed the presence of genetic variability among the tested genotypes (Table 4). The presence of significant GEI shows inconsistencies in the performances of wheat genotypes across environments i.e., the relative performances of the genotypes were significantly affected by the varying environmental conditions. Single genotype was consistently neither superior nor inferior across locations indicating the need to develop cultivars that are specifically adapted and the need to identify widely adapted and stable genotypes across environments. But, this was not the objective of the study.

From the total treatment sum of square, 47.0% was attributed to environmental cffccts and the rest to genotypic effects (11.7%) and GEI (25.4%). Hcnce, environments were diverse with large differences among environmental means causing most of the variation in grain yield. Results from AMMI analysis also showed that GEI component of variation was partitioned into eight possible IPCA (Table 4). The first four IPCA were highly significant (P < 0.01) explaining 91% of the total GEI sum of square or the variation which were taken as adequate dimensions for this data set. However, the prediction assessment indicated that AMMI model with only two IPCA was the best predictive model (Yan ei al. 2000). This model (AMMI1 and AMMI2) had explained 74% of GEI sum of squares. Therefore, the interactions of the 30 genotypes with nine environments were best predicted by the first two principal components.

11 Table 4 AMMI analysis for grain yield of 30 genotypes tested at nine locations in 2013

Soiree of df SS MS Sum of square Explained (%) Variation Total G xE G xE variation explained Cumulative Environments 8 94450.7 11806.4*** 47.00 Reps within enfironment 18 2715.5 150.9 1.40 Genotype 29 23599.4 813.8*** 11.70 GxE. 232 51134.7 220.4*** 25.40 IPCA 1 36 30142.2 837.3*** 58.95 58 IPCA 2 34 8034.1 236.3*** 15.71 74 IPCA 3 32 5282.1 165.1*** 10.33 84 IPCA 4 30 3382.3 112.70** 6.61 91 IPCA 5 28 1773.6 63.30^ 3.47 95 IPCA residual 72 2520.4 102.1 4.93 100 Error 522 29084.4 55.7 14.47 Grand mean = 40.76 R2 = 0.85 CV (%) = 18.3 **= significant at P < 0.01 and ns= nonsignificant: IPCA - Interaction principal component axis

Based on the biplot analysis, environments and genotypes showed high variability for both main effects and interaction effects (IPCA1) for grain yield (Fig 2). AMMI1 classified genotypes, environments into two broad groups based on their IPC'Al scores and the genotypes and environments interacted positively within a group, and generally, the genotypes adapted for those environments (Yan 2002). Accordingly, Meraro and Bekoji had positive FPCA1 scores and positively interacted with genotypes having positive IPCA1 scores (Table 1) whereas other locations having negative IPCA1 score and positively interacted with genotypes had a negative IPCA1 scorcs (Fig 2).

MER

7.2

G18 G17 G23 G30

G12 G10 G9* G28 G21

93 BKJ -o*6- GlS *et>e - - *G25 * G1

• GGR ASS G22 G7 SNN

G29 G6 G24 * G11

23 31.4 39.8 48.2 56.6 65 Grain Yield Means (Q/ha) Fig 2 AMMI1 biplot for grain yield of 30 genotypes evaluated across nine locations in 2013

The biplo also showed Meraro had more than average grain yield with high positive IPCA1 score indicating that the location interacted highly with the performance of the genotypes followed by Njoro with below average grain yield. Geregera was the least favorable environment for all genotypes with

12 negative IPCA1 score (Fig 2). Based on AMMI1 biplot genotypes G15, G16, and G26 were found the most stable genotypes with below average grain yield. G11 and G24 were the most unstable.

AMMI2 biplot presents the spatial pattern of the first two IPCA interaction effects (IPCA1 and 2) and helps in visual interpretation of the GEI patterns and identify genotypes or locations that exhibit low, medium or high levels of interaction effects (Yan 2002). The biplot showed that environments and genotypes were distributed highly in all quadrants (Fig 3). Meraro and Njoro were the most discriminating environments among the genotypes indicated by longer vectors projected from the origin i.e., these locations gave good information on the performance of the genotypes. In addition, it confirmed the importance of testing genotypes across these two countries (Ethiopia and Kenya) to investigate high yielding and stable varieties across the region. Asasa and Sinana identified as the least interactive environments with the tested genotypes as indicated by the shortest vectors from the origin zero (Fig 3). Genotypes near the origin are non-sensitive to environmental interactive forccs and these distant from the origin are sensitive and have large interactions (Samonte et al. 2005). Accord ngly, genotypes G2, G13, and G9 were less sensitive to environmental interactive forccs; and hence, these genotypes were considered as stable genotypes. Genotypes G28, G25 and G8 were highly influenced by the interactive force of environment and sensitive to environmental changes i.e., considered as unstable genotypes due to their long projections from the origin (Fig 3).

AMMI stability value (ASV): ASV is the distance from the coordinate point to the origin in a two dimensional plot of IPCA 1 scores against IPCA2 scores in the AMMI model (Purchase ct al. 2000). According to Farshadfar (2008), as the IPCA1 scorc contributes more to the GEI sum of square, a weighted value is needed. Since AMMI model does not make provision for a quantitative stability measure, ASV is essential in order to quantify and rank genotypes according to their yield stability (Purchase, 1997). In ASV method, genotypes with least ASV arc the most stable than genotypes with higher ASV (Purchase et al. 2000). Accordingly, genotypes with small ASV values were G15, G26 and G3 were found stable in the studied genotypes. On the contrary, they had low grain yield performance across locations. The most unstable genotypes, according to the ASV approach, were Gil, G'M, and G6 with high ASV values. However, these genotypes had above average grain yield exccpt G6. References

FAOSTAT 2014. Statistical Database of the Food and Agriculture of the United Nations. Farshadfar E. 2008. Incorporation of AMMI stability value and grain yield in a single non-parametric index (GSI) in bread wheat. Pakistan Journal of biological sciences. 11(14): 1791-1796. Gauch HG and Zobel RW. 1997. Identifying inega-environinents and targeting genotypes. Crop Science 37(2): 311-326. Negassa A. Koo J, Sonder K. Shiferaw B, Smale M, Braun HJ, Hodson DP, Gbegbelegbe S, Guo Z, Wood S, Payne T, Abeyo B. 2012. The potential for wheal production in Sub-Saharan Africa: Analysis of biophysical suitability and economic profitability. CIMMYT, Addis Ababa, Ethiopia Purchase JL. 1997. Parametric analysis to describe Genotype x Environment interaction and yield stability in winter wheat. Ph.D. Thesis, Department of Agronomy, Faculty of Agriculture. University of the Free State, Bloemfontein, South Africa. Purchase JL, Hatting H and Denventer CV. 2000. Genotype by environments interaction of wheat in South Africa: stability analysis of yield performance. South Africa Journal of plant science 17:101-107. Samonte SOPB, Wilson LT, McClung AM and Medley JC. 2005. Targeting Cultivars on to Rice Growing Environments Using AMMI and SREG GGE Biplot Analyses. Crop Science 45:2414-2424. Yan W. 2002. Singular-value partition for biplot analysis of multienvironment trial data. Agronomy Journal 94:990-996. Yan W, Hunt LA, Sheng Q and Szlavnics Z. 2000. Cultivar evaluation and mega environment investigation based on the GGE biplot. Crop Science 40: 597-605. Zobel RW, Wright MJ and Gauch Jr HG, 1988. Statistical analysis of a yield trial. Agronomy Journal 80: 388- 393.

14 Evaluation of Spring Bread Wheat Advanced Lines across Different Environments of Tanzania

Rose Mongi1, Athony Elanga \ Emmanuel Kadogholo2 Ibrahim Mamuya3, Salome Munisi3 The Uyole Agricultural Research Institute 1 Kifyulilo Agricultural research Institute2 Selian Agricultural Research Institute 3; Tanzania

Abstract The performances bread wheat varieties depend on the influence of genotype, environment, and their interactions (GEI). 1 he objective of this research was to determine the contribution of genotype, environment and GEI to variation in grain yield potential in advanced bread wheat lines of Tanzania, evaiuatc yield performances and select lines better suited to the needs of farmers in the country. Eighteen genotypes were grown using randomized complete block design with four replications. Six environments were used over a period of two years, 2012, and 2013. Agronomic management was done according to the recommendations for wheat production in Tanzania. Data were collected on agronomic traits including yield and yield components. Additive Main effect and Multiplicative Interaction model (AMMI) and Genotype and Genotype by environmen (GGE) biplot graphic analysis was made to assess grain yield interactions and identify best performers across environment. Low incidences of septoria tritici blotch were observed and two variants races of Ug99 were detected. Varieties UW91012, UW20302 and UW20311 yielded highest and are recommended for on-farm trials under high rainfall environments. These are in addition to UW20006, UW20010, and UW20101, which were tested under farmer’s field conditions.

Introduction

In Tanzania, the national wheat yield average is 1 ton ha'1 and the annual production is 100,000 tons, while importation is 643,000 tons, six times more than the amount produced in the country. Among the major factors contributing to yield reduction in wheat, is inadequacy of improved cultivars that have the ability to withstand biotic and abiotic stresses such as moisture stress and foliar diseases, respectively. Moistures stress particularly in the northern part of the country (Kilimanjaro and Arusha is attributed mainly by crratic rainfall patterns that are linked to the global climate change (Chang a et al. 2010). Consequently, in the Southern highlands zone, where unimodal type of rainfall is experienced, wheat is grown as a second crop after potatoes and beans. In ease of harvest delays of the previous crop, the wheat crop falls into severe moisture stress attributed by late plantings resulting into low yields.

Foliar diseases particularly Septoria tritici blotch and the three rusts (stem, leaf and stripe or yellow rust) have been serious biotic constraints to wheal production as the diseases interfere with the photosynthetic ability of the infected plants by reducing radiation capture and radiation use efficiency (Corretcro et al. 2011). It is known that the East African region is the host spot for cpidcmics of the wheat rusts with high possibilities of new races. The presence of barberry plants that are alternative hosts for stem rust, increases chances of having new races as the plant facilitate completion of sexual life cycle of the pathogen that involves exchange of genetic material during mciosis. An example is the new racc of stem rust, Ug 99, that was first reported in Uganda and thereafter in Kenya that causcd yield losses of 71% in experimental plots (CIMMYT 2007). In this regard, the success of the wheat crop in mproving livelihoods of the people in East Africa depends largely on its ability to yield well across environments.

To join efforts in increasing wheat productivity, regional wheat nurseries have been assembled, that contained different genotypes with different levels of tolerance/resistance to biotic and abiotic stresses that exists in the region, including the stem rust caused by Ug99. However, wheat-growing

15 environments in the region arc diverse requiring thorough understanding of the influence of genotypes (G), environment (E) and genotype by environment interaction (GEI) on the performances of the advanced lines. However, the effects of G and E are not additive due to the interaction between them making it difficult to select superior genotypes/lines without multi-location evaluations. This is because genotypes grown in a certain environments may perform well but when the same is grown in different environments performs poorly (Allard and Bradshow 1964; Hill 1975; Dclacy et al. 1996)

To identify and select best genotypes for a wide cultivation, several statistical models and procedures have been developed and exploited for studying GEI effects, stability, and their relationship in yield performances (Finlay and Wilkinson 1963). Combined Analysis of Variance (ANOVA) and Additive Main effect and Multiplicative interaction (AMMI) model, including the AMMI biplot graphic analysis have been used to draw' conclusion regarding phenotypic stability, genotypic behavior of the cultivar and the degree of genetic divergence between cultivars. Consequently, genotype and genotype oy environment (GGE) biplot compliments the AMMI biplot by stratifying environments into mega environments that optimize performance in such mega-environment and it has been used in evaluating performances of wheat and maize varieties across locations to identify those which arc well adapted (Yan el al. 2007; Yan and Tinker 2006). This study was designed to evaluate:

• the influence of G, E and GEI on eighteen wheat advanced lines selected from national and regional wheat breeding programs. • identify stable lines under different wheat growing conditions • evaluate the yield performances of each genotype and examine the relationship among test environments

Materials and Methods

The expe iments were conducted for two years under rain fed conditions at three locations (Uyole, Milundikwa and Igeri) that represented wheat growing environments of Tanzania. Eighteen advanced lines of wheat from national and regional wheat nurseries were used as planting materials. A randomized complete block design (RCBD) with four replications was used in each site. Plot size was 1.2 x 3m with rows spaced at 0.2m apart and seeded at a rate of 120 Kg ha’1. All plots w'ere fertilized with 75N and 20 P kg h a 1. When the plants were at 4 - 6 leaf stage, copper was applied in a form of copper sulphate at a rate of 2 kg ha 1 to supply cu as a micronutrient that plays a role during seed formation. Weeds were controlled by 2-4D amine followed by hand pulling where necessary. Data were collectcd on stand density, days to heading, days to maturity, plant height foliar diseases in particular' Septoria tritici blotch and stem rust. Considering low infections of the rust diseases that occurred during the seasons, infected stem samples were collected and sent abroad for race identification through DNA sequencing. Additional data were collected on bulk weight, 1000 kernel weight and grain yield.

Data were subjected to statistical analysis using Genstai 14th edition software where combined analysis of variance (ANOVA) was done. Additive main effect and Multiplicative Interaction (AMMI) model was used to determine the magnitude of the main effect and interactions. The GEI was partitioned into two principal components axis. Stable lines across locations were identified by analyzing the contributions of the variations into total sums of squares. AMMI stability values (ASV) were calculated for the purpose of ranking genotypes using the formula below proposed by Purchase. 1997;

IPCAl sum o f squares ASV = (IPCA square 1) + [IPCA2 Score]2 IIPCA2 sum of squares Where: IPCA1 - interaction principal component analysis axis 1; IPCA2 - interaction principal component analysis axis 2. AMMI and GGE biplots graphic analysis were used to draw conclusions on genotype performances and identify mega environments and cultivars that optimize performance in such mega environments.

Results and Discussion

The combined analysis o f variance indicated that there is a significant variations between genotypes (P< 0.05) in days to heading, days to maturity, plant height, septoria leaf blotch, kernel and bulk weight (Table 1 and 2). This indicated difference in responses of the advanced lines to environmental conditions of the localities. Days to heading ranged from 60 - 69 while days to maturity were between 107 and 113 suggesting all the genotypes were early to medium maturing, traits highly preferred by farmers. Consequently, all the genotypes were semi dwarfs with plant height ranging from 77 - 97 cm as measured from the soil surface to the top of the spike. Considering foliar diseases, septoria leaf blotch ranged from 2.2 - 5.0 scores at the scale of 0 - 9 indicating most of the advanced lines were tolerant/resistance to the disease with the exception of the local check Juhudi. In general, the range of mean performances o f the advanced lines indicated the presence o f sufficient genetic variability for the traits studied.

17 Table 1 Mean of agronomic traits of tested genotypes in 2012/2013 wheat cropping seasons

Genotypes Stand density Days to Days to Plant height Septoria leaf 1000 kernel Bulk weight Grain yield (0 - 9) heading maturity (cm) blotch (0 -9 ) weight (Kg hM) (kg ha-1) nnnrs olrA 0.0 ft/i04 lUr OO.IQO i 0.09 Q HH. U 01.mi C y O 003 UW 20104 6.6 65 108 86.4 2.8 39.8 80.7 3391 UW 20302 6.5 69 113 85.1 2.5 40.3 78.5 3808 UW 91006 6.7 64 106 79.2 3.4 40.4 81.9 3862 UW 90032 6.4 63 107 86.3 3.2 43.1 80.3 3594 UW 86085 6.2 67 110 80.0 2.2 37.6 81.5 2850 UW 90039 6.7 62 107 82.2 4.5 43.1 80.3 3617 UW 20311 6.5 60 106 83.6 2.9 42.9 82.4 3724 ETBW5927 6.5 68 113 76.4 3.0 41.5 80.0 2894 ETBW5981 6.6 69 107 80.3 3.0 45.8 80.4 3140 JUHUDI 6.6 60 101 74.1 5.0 40.5 81.4 2949 UW 20009 6.8 62 107 83.8 4.2 43.3 80.5 3681 ETBW6198 6.5 68 112 74.8 3.0 39.7 80.4 3291 UW 91012 6.5 65 109 80.4 3.2 42.3 81.8 3762 SELIAN 87 6.6 64 109 81.5 3.4 38.9 82.5 3384 UW 20219 6.5 68 112 76.8 3.3 43.5 80.1 3452 NJOMBE 7 6.7 69 111 97.1 2.4 41.5 80.3 3109 UW 91020 6.4 64 109 78.6 3.9 43.4 80.2 3701 Mean 6.6 64.8 109.1 81.7 3.6 39.8 80.8 3448 LSD 1.2 4.5 4.7 8.3 0.86 7.8 11.9 679.9 CV (%) 13.2 5 6.03 7.3 32.9 4.1 10.7 18.9 P (5%) ns ** ** ** ** ** * ** Table 2 Mean square for yield and other agronomic traits of 18 genotypes

Source of df Stand density Days to heading Days to maturity Plant height Septoria leaf blotch Kernel Bulk Grain variation weight weight yield Replication 3 6.1 0.75 6.36 83.1 5.3 16.7 165.5 2.3 Genotype (G) 17 1.45 440.6** 461.9** 1500.5** 29.5** 123.0** 126.5* 1.67** Environ (E) 5 90.44** 3484.6** 29749.2** 5163.2** 105.3** 410.3** 2577.3** 5.92** GxE 85 0.98 27.5** 60.3** 64.2** 3.4* 23.5** 71.1 5.35** Residual 28 - 10.5** 18.8** 35.9** 2.7 9.14** 73.8 2.36 Error 306 Total 481

19 Low infections of stem rust were also observed in few lines with disease severity ranging from 5S - 30 S at a sea e of 0 - 100% and severity scale that identify resistant genotypes (R), moderate resistant (MR), moderate susceptible (MS) and susceptible (S). Out of 38 samples collected in the country, one was a non Ug 99 race while 27 samples were two variants of Ug 99 (TTKSK). Grain yield ranged from 2850-3869 kg ha'1 with a significant GEI, suggesting that some of the advanced lines were not stable when evaluated across locations. The contributions of variations of grain yield to the total sums of squares were 8.2%, 49.1 %, and 21.1 % for genotypes, environment and GEI, respectively (Table 3). Environments contributed most to the total sums of squares indicating that they were quite diverse. This might be due to differences in soils and rainfall patterns.

Table 3 ANOVA table for AMMI model

Source df SS MS F Explained % Total 481 576629197 1337887 - - Genotypes 17 47428817 2789930 9.46** 8.2 Environments 5 282947046 5.7E+07 29.78** 49.1 Interactions 85 121781493 1432723 4.86** 21.1 IPCA1 21 72235683 3439794 11.66** 12.5 1PCA2 19 20075196 1056589 3.58** 5.7 Residuals 28 16238487 579946 1.97** 2.8 Error 306 90272646 295009 - -

Mean grain yield results of the AMMI analysis with IPCA1 and IPCA2 for the advanced lines of wheat is presented in Table 4. Some of the genotypes had higher IPCA scores, regardless of being positive or negative, indicating that, they were more adapted to specific environments. Among them are UW 20302, UW 20219, and ETBW6198. On the other hand, line ETBW5981 and UW 86085 had IPCA scores close to zero indicating that they were more stable lines. However, ihe highest yielding genotype was SIFA, followed by UW 20302, UW 91006 To identify environments having the same cultivar as superior in yielding ability (Mega environments) GGE biplot was used involving the first two components of multivariate analysis that accounted for most of the data variance. The GGE biplot allowed visualization of the interaction pattern in a graph through a polygon (Yan and Kang 2003). The polygon was formed by connecting the vertex genotypes (advanced lines) with straight lines and the rest of the genotypes were placed within the polygon. The vertex genotypes were SIFA, UW 20219, UW 20311, ETBW6198, NJOMBE 7, JUHUDl, UW 86085 and ETBW5927. SIFA and UW20311 were among the best genotypes in terms of yield in some environment while UW20219, ETBW6198 were moderate yielders. The poorest genotypes were Juhudi, ETBW5927 and UW86085 as they were far from the origin of the biplot (Fig 1). According to the graph, UW90039 and UW91012 are specifically adapted and will perform better if grown in mega environments of Uyole, Igeri and Milundikwa. Since the polygon was divided into sectors, the environments that were within the same sector share the same high yielding genotype and environments in different sector has different high yielding genotype. Hence, the genotypes fell into four sections and the environment into three sections. The first section contained environment 1 (Uyole) and 6 (Milundikwa) with line UW 90039 as the best for these environments while the second section contained environment 2 (Igeri) with genotype UW 91012 as the best yielder. The vertex lines ETBW5927, UW 86085, NJOMBE 7, ETBW6198 were not the top yielding lines in any of the six environments.

20 Table 4 Genotype yield means, IPCAs and ASV scores

Genotype Gm IPCA1 IPCA2 ASV SIFA 3869 1.77049 -0.93735 4.3 UW 20104 3391 -0.56399 -1.54112 2.4 UW 20302 3808 -3.17961 5.9764 6.5 UW 91006 3862 -1.85293 1.0734 4.4 UW 90032 3594 1.01389 -1.138 3 UW 86085 2850 1.35295 1.0135 1.2 UW 90039 3617 1.2283 - 1.77753 4.1 UW 20311 3724 0.94687 -1.68522 3.4 ETBW5927 2894 -1.59883 0.93343 3.7 ETBW5981 3140 0.59938 -1.88536 1.3 JUHUDI 2949 0.6.1091 1.40001 2.5 UW 20009 3681 -0.09538 3.57381 3.7 ETBW6198 3291 1.72323 1.73152 5 UW 91012 3762 12.4939 0.08092 2.3 SELIAN 87 3384 9.72177 -1.38339 3.2 UW 20219 3452 -3.48711 -6.46804 7.1 NJOMBE 7 3109 1.41323 1.50814 4.1 UW 91020 3701 -1.19001 6.713 2.8 Gm = genotype means; ASV = AMMI stability' values

-3-2-101234 PC1 - -1 0 .7 0 % Fig 1 Polygon view of GEI for the advanced wheat lines over six environments

21 Comparison biplct

PC1 - 40.70%

Genotype scores Environment scores AEC

AEC - the average - environment coordination Fig 2 Ranking of advanced lines of wheat relative to an ideal genotype

The advanced lines were compared with an ideal genotype created within the comparison graph and expected to have highest yield mean performance and be absolute stable (Yan and Kang 2003). The ideal genotype is located towards the centre of the concentric circles and pointed with an arrow in Fig 2. Any advanced line close to the ideal genotype was considered more favorable. In this regard, UW90039, UW20311 and UW91012 were considered as the best in ascending order.

Conclusion

All the advanced lines evaluated were relatively resistance/tolerant to septoria tritici blotch. However, the presence of two variants races of Ug99 suggested more efforts in fighting the devastating pathogen. Consequently, the results confirmed genetic variability between the advanced lines and that the test environments were quite diverse probably due to differences in soil types and rainfall patterns. The highly significant differences were observed in all of the agronomic traits evaluated with the exception of stand density. Yield performances variation of genotypes were highly influenced by environments and by GEI.. The presence of GEI allowed genotypes to be ranked by AMMI stability values to identify stable genotypes. The GGE biplot graphic analysis complemented the ASV in defining mega-environments and allowed visualization of the interaction. Based on the overall stability analysis, UW91012, UW20302 and UW20311 yielded highest and are recommended for on- farm trials under high rainfall environments. These arc in addition to UW20006, UW20010 and UW20101 which are tested recently under farmer's field conditions.

References

Allard RW and Bradshaw AD. 1964. Implications of genotype - environmental interaction in applied plant breeding Crop. Sci. 5: 503 - 506. Chang’a LB Yanda PZ and Ngana J. 20I0. Indigenous knowledge in seasonal rainfall prediction in Tanzania. A case of the South-West highlands of Tanzania. ./. Geography and Regional Planning 3(4): 66-72.

2 2 Corretero R. Bancal MO and Miralless DJ. 2011. Effcct of leaf rust (Puccinia trichina) on photosynthesis and related process ofleaves in wheat crops grown at two different contrasting sites and with different nitrogen levels. European J. of Agronomy 64 (1): 134- 141. CI V1MYT. 2007. Sounding the Alarm on Global stem rust. An assessment of Ug99 in Kenya and Ethiopia and potential for impact in neighboring regions and beyond. CIMMYT circular, 29 May, 2007. FAOSTAT 2012. FAO food crop production data. 2012. FAOSTAT 2013. FAO Food crop production data. 2013. Hill J. 1975. Genotype environment interaction. A challenge for plant breeding. J. Agric. Sci. 85:477-493. Delacy IH, Basford KE, Cooper M. Bull JK. 1996. Analysis of multi-environment trials-an historical perspective. Plant adaptation and crop improvement. Eds. Cooper M and Hammer GL. CAB International. Finlay KW and Wilkinson GN. 1963. The analysis of adaptation in a plant breeding program. Aust J. Agric Res 14:742-754. Yan W, Kang MS, Ma B, Woods S and Cornelius PL. 2007. GGE biplot vs AMMI analysis of genolype-by- environment data. Crop Sci. 47: 643-655. Yan W and Tinker NA. 2006. Biplot analysis of multi-environment ti ial data: Principle and applications. Can J. Plani Sci. 86:623-645. Purchase JL. 1997. Parametric stability to describe G x E interaction and yield stability in winter wheat. South Africa. Yan W and Kang MS. 2003. GGE biplot analysis. CRC Press. New York.

23 T heme 2 Development of Integrated Crop, Soil, and Water Management Practices Responses of Bread Wheat Genotypes to Fertilizer and Seed Rates Arsi Zone, Ethiopia

Dawit Habte1, Kassu Tadesse2, Alemayehu Asefa3, Amare Tadesse4, Bahiru Tilahun5 1,2.4.5 Ethiopian Institute of Agricultural Research, Kulumsa Agricultural Research Center, Wheat Regional Center of Excellence, P.O. Box 489, Asella, Ethiopia Eastern Africa Agricultural Productivity Project (EAAPP), Ethiopian Institute of Agricultural Research, Addis Ababa, Ethiopia

Abstracts Bread wheat varietal responses to fertilizers in grain yield, N uptake. N use efficiency, and apparent N recovery have been reported in various studies. Varieties also differ in their tillering potential, seed size, and older agronomic characteristics, which directly or indirectly influence the optimum plant population densities that can maximize expressions of their genetic potential for grain and biomass yields. Most agronomic trials in Ethiopia should consider wider range of interactions and information on varietal differences in the yield and yield component responses to management and environment is necessary. A trial was conducted in 2013 with the objective of determining the responses of different bread wheat cultivgxs to different combinations of seed and fertilizer rates under different environments of Arsi zone. Experimental design used was a split-split plot replicated three times. Three recently released bread wheat varieties (ETBW5795, ETBW5483, and HAR3116) as main plot factor were combined in factorial arrangement with three fertilizer rates (recommended NP or RNP. 1.5RNP, and 2RNP), and four seed rates (100, 150, 200 and 250 kg ha'1 for Vertisols areas and 75, 125, 175, 225 kg ha'! for the non- Vertisols). Data on agronomic responses, grain quality, and N recovery was collected and the nitrogen recovery efficiency (NRE) and agronomic efficiency (AE) calculated. The main effects of fertilizer, seed rates, and varieties on the grain and biomass yields were significant (p<0.05). ETBW5795 (5.95 t ha"1) out yielded across locations followed by ETBW5483 (5.2 t ha'1) and HAR3116 (4.5 t h a 1). Location effects, were also highly significant (P<0.01). The average responses of the three varieties at Kulumsa (Tiyo district) and Kofele was higher by 1.34 and 1.1 t ha 1 as compared to Digelu-Tijo sites. Varietal responses across locations showed that ETBW5795 and ETBW5483 out-yielded HAR3116 with mean extra yields of 1.42 and 0.66 t ha'1, respectively. Introduction

Increased usage of fertilizers, particularly of N, has been recommended as a primary means of increasing bread wheat grain yields in Ethiopia (Amanuel el al. 1991; Tanner el al. 1993; Asefa el al. 1997; Shambel el al. 1999; Minale el al. 1999; Taye et al. 2002; Minale el al. 2004). However, the recommendations have limited considerations of varietal, locations, and seed rates interactions with fertilizer rates. Crop yield responses are influenced by both genetic and environmental factors and the interaction between them (Tisdale el al. 2002). Differences in tillering potential, seed size, and other agronomic characteristics exist among bread wheat genotypes released in Ethiopia. Variations in the environmental and management factors and their interactions control the degree of expressions of the genetic potential of different varieties.

Generation of information on the response of varieties to seed rates and nutritional levels and environments of Ethiopia is important. Therefore, trials were conducted to determine the effect of seed and fertilizer rates on yield, yield components,nutrient use, and agronomic efficiency of bread wheat varieties under different environments of Arsi Zone.

24 Methodology

Location description and soils The experiment was conducted during the 2013 main cropping season on farmers' fields and on research stations of Arsi Zone in the districts of Digelu-Tijo, Kofele, and Tiyo. The study areas generally lie between latitudes of 7°00’ and 8°00’and longitudes of 38°45’ and 39°30\ The soils vary from Haplic Luvisols in Tiyo district to Eutric Vertisols and Humic Nitosols in Digelu-Tijo and Kofele districts, respectively. The agroecologies of the study areas vary from Tepid to cool moist mid highlands to tepid to cool humid mid-highlands (Ethio-Italian Development Cooperation, 2002)

Arsi GIS data (Ethio-Italian Development Coop. 2002) and Arcview 3.2 GIS software (ESRI, 1999)

Treatments and experimental designs Three recently released bread wheat varieties (ETBW5795, ETBW5483, and HAR3116), main plot factor were combined in factorial arrangement with three fertilizer rates (recommended NP or RNP, 1.5RNP, and 2RNP), and four seed rales (100, 150, 200 and 250 kg ha'1 for Vertisols areas and 75, 125, 175, 225 kg ha'1 for the non-Vertisols). Fertilizer rates and seed rates were used as the sub-plot, and sub-sub plot factors, respectively. In another experiment sel up two varieties (ETBW5795 and HARSI 16) were combined with three fertilizer and four seed rates. The three variety trials were conducted on research sub-stations located at the three districts and the second trial was conducted on four sites under farmers’ conditions, making seven sites. The number of varieties was reduced from three in the first to two in the second trial in order to reduce the field sizes by one third due to the difficulties of obtaining the required field size under farmers’ fields. All P20 5 and half of the N was applied at planting and the remaining N before booling stage. The gross and net plot size were 5 m x 4 m and 3 m x 3 m, respectively.

Data collection and analysis Soil sampling was done using standard sampling procedure (Carter and Gregorich, 2008). Each composite soil sample was be subjected for physico-chemical analysis (soil texture, Bulk density, soil pH, organic carbon (OC), total N available P). Plant samples for analysis of total N were collected at 25 one growth stage from two sites at Digelu-Tijo district from trials conducted on farmers’ fields. The sampling time was two weeks after top dressing, and 14-16 young leave samples from the upper part of plants were collected to make one composite assay per plot. Agronomic data on grain yield and yield components such as seedling density, number of tillers per plant, spike length, number of kernels per spike, thousand kernel weight (TKW), plant height, grain and biomass yields were collected at the recommended time. Information on disease and pest incidences and lodging was also collected. Harvesting was done by hand using sickles. Hundred culm weight (lOOcw) were collected from four to five points within a plot and slashed from close to the ground surface and the dry matter yield of above-ground biomass was determined. Grain yield was determined from 9m2 net plot by hand threshing of the harvested samples. Yield adjustments were made based on 12.5% moisture content. Above ground biomass yields were determined based on data of hundred culm weight and the harvest index (HI) calculated as the ratio of grain yield to above ground biomass yield expressed as a percentage. TKW was determined by weighing 1000 grains and adjusted under 11% moisture content. The number of grains/spike was determined by hand counting of the grains from 5 spike samples and averaging them. Straw N contents were determined by micro-Kjeldahl analysis at KARC soil laboratory from the oven-dried bulk samples. Grain Protein, starch, wet gluten, and zeleny values were determined using Near Infrared Reflectance (N1R) at Amhara Region Agricultural Research Institute (ARARI), Bahirdar, Ethiopia.

Grain N values were calculated by multiplying grain yields by the respective N content. Apparent N recovery (AR) of the grain for each treatment N was calculated as: (GNU of treatment - GNU from the control treatment)/fertilizer N applied. The GNU values were calculated from N treatments averaged over the ranges of P2O5 levels and replications making the degree of freedom 30. Agronomic efficiency (AE) of fertilizer N was calculated as: (grain yield of treatment -grain yield of control)/fertilizer N applied. Again, the main effects of N were considered. Efficiency values calculated based on known procedures (Novoa and Loomis, 1981; Cassman and Dobermann, 2002; Fageria and Baligar, 2003, Doyle and Holford, 1993).

To determine the effect of location and varietal differences in yields, the linear regression model from SPSS 20 software using Digelu-Tijo and HAR3116 as reference location and variety, respectively. Yield and yield component data were analyzed using SAS 9.0 statistical software. The DMRT test (P<0.05) was used to assess differences among treatment means. For graphical analysis of yield and yield, components Origion 8 GUI and SPSS 20.0 softwares (Origin Lab Coop., 1991-2007; IBM, 1989-2011) were used.

Results and Discussion

Grain and biomass yield responses The grain yields and yield components of varieties (ETBW5795, ETBW5483 and HAR3116) under different fertilizer and seed rate combinations are presented in tables 1, 2, and 3. The average grain (AGY) and biological yields (ABY) of ETBW5795 varieties at Kofele are significantly affected by the treatments with mean grain and biomass yields of 5.65 and 16, respectively. In contrast to the control treatment, i.e., recommended NP and 125 kg ha'1 seed rate grain yield of 5.4 t ha'1, the highest and lowest AGY obtained were 6.71 and 4.8 t ha 1 with treatments of 1.5RNP/225 kg ha'1 and RNP/75 kg ha’1 fertilizer and seed rate, respectively. Similarly, the highest and lowest ABY obtained were 20.1 and 12.6 t ha' 1 with corresponding treatments of 1.5RNP/225 kg ha seed rate and RNP/75 seed rate kg ha'1, respectively, against the control treatment yield of 13.6 t ha . The grain and biomass yields of FIAR3116 were 4.95 and 15.6 t ha*1, significantly. The treatment difference of the AGY of ETBW5795 at Digelu-Tijo was highly significant, but NS (p< 0.05) for the ABY. The treatment effects of same variety on both parameters at Kulumsa (Tiyo) were highly significant contrary to the non-significant effects on the yields of the other two varieties. The yields of HAR3116 were significantly affected at Kofele, but not at the other two locations, while ETBW5483 was highly significantly affected at Digelu-Tijo. Table 1 Main and interaction effects of SR, FR, variety, and location on the yield and yield components of three varieties

Source Main TRTs PH SPM AGY ABY HI NSPS TKW HLW Variety ETBW5795 106 541 5958 16084 37.6 46.9 37.7 74.5 HAR3116 116.5 493 4541 14114 32.5 46.9 33.1 76.1 ETBW5483 104.5 573 5227 14436 36.4 45.5 32.6 76.8 k ic k k k k k-kk ****** NS "kick *** Mean 109 536 5242 14878 36 46 34 76 FR FR1 107 527 5089 14205 36.0 45.1 34.9 76.1 FR2 110 536 5242 14870 35.7 46.8 34.4 75.6 FR3 110 543 5397 15591 34.9 47.5 34.0 75.6 NS NS *** k k k NS *** k Mean 109 535 5243 14889 36 46 34 76 SR SR1 107 509 4981 13645 36.8 47.5 34.9 76.0 SR2 108 529 5293 14675 36.3 46.1 34.8 76.1 SR3 108 551 5357 15231 35.4 46.0 34.2 75.6 SR4 112 553 5326 15973 33.6 46.3 33.9 75.4 NS k k k *** k k k *** NS k ic k •kk Mean 109 536 5239 14881 36 46 34 76 LOC ****** k k k XT* NS *** ****** Rep NS * ***** NS *** NS NS FR*LCCA NS NS *** NS NS NS k ic k k k FR*SR NS NS NS NS NS NS NS NS Variety*FR NS NS k k k ** NS NS NS NS SR*LGCA NS * k k k ** NS NS X* NS Variety'LOC NS k k k *** ■kick k k k k k k *** Variety*SR NS NS NS NS k k k NS NS NS Mean 109 536 5242 14884 35.5 46.5 34.5 75.8 CV (%) 11.5 10.8 12.8 17 11.2 15.1 5.4 1.8

27 Table 2 Two way table of economic yields (FR*SR) per variety (t ha 1) arranged across locations

Disrict SR ETBW5795 Mean HAR3116 Mean ETBW5483 Mean (kg ha-1) RNP 1.5RNP 2RNP GY RNP 1.5RNP 2RNP GY RNP 1.5RNP 2RNP GY Kulumsa 75 58.9 70.43 75.29 68.21 40.33 43.88 43.42 42.54 54.13 59.04 58.62 57.26 (Tiyo) 125 70.29 74.24 73.46 72.66 42.71 43.28 44.64 43.54 52.45 57.6 65.38 58.48 175 72 74.13 72.73 72.95 52.12 42.17 41.43 45.24 56.25 61.54 56.83 58.21 225 67.73 72.3 76.76 72.26 42.32 42.17 44.51 43 58.66 54.25 56.81 56.57 Mean 67.23 72.78 74.56 44.37 42.88 43.5 55.37 58.11 59.41 Kofele 75 47.72 51.06 44.6 47.79 53.54 44.68 30.27 42.83 54.8 49.33 57.67 53.93 125 53.57 50.97 65.95 56.83 59.17 46.01 50.41 51.86 60.36 60.39 55.93 58.89 175 59.24 58.63 57.52 58.46 58 40.69 52.9 50.53 62.7 59.61 64.35 62.22 225 57.31 67.17 64.56 63.01 59.84 51.84 47.04 52.91 65.15 64.97 66.4 65.51 Mean 54.46 56.96 58.16 57.64 45.81 45.16 60.75 58.58 61.09 Digelu-Tijo 75 50.59 50.2 57.3 52.7 38.83 39.1 50.44 42.79 32.2 39.37 40.23 37.27 125 47.53 50.56 61.23 53.11 38.1 40.6 44.27 40.99 28.2 42.07 41.55 37.27 175 39.04 55.85 57.44 50.78 45.01 51.02 39.84 45.29 34.22 38.3 43 38.51 225 39.77 48.46 47.37 45.2 37.69 41.89 47.9 42.49 33.93 36.91 43.19 38.01 Mean 44.23 51.27 55.84 39.91 43.15 45.61 32.14 39.16 41.99

Table 3 Differential responses of varieties to the main and interaction effects of SR, FR, and Location in yield (t ha 1) and yield components (V1 - ETBW5795; V2 - HAR3116; V3- ETBW5483)

Source AGY ABY HI SPM PH TKW NSPS HLW V1 V2 V3 V1 V2 V3 V1 V2 V3 V1 V2 V3 V1 V2 V3 V1 V2 V 3 V1 V2 V3 V1 V2 V3 FR *** NS ** **★ NS *j** NS NS NS NS NS NS NS ** NS * NS NS NS NS ** NS ** Rep NS * V* NS # NS NS NS NS NS NS NS *** NS NS NS * NS NS NS NS * NS LOC *** *** **+ *** *** #** t** NS *** *irk *** *** * *** *** *★* *** *** ** *+# *** *** H* SR NS NS NS *** NS * *** NS NS * NS * NS NS * * NS NS NS NS NS ** NS FR*REP NS NS # NS NS ** NS NS NS NS NS NS NS NS NS NS NS NS ** NS NS NS NS FR*LOC NS *★* ♦ NS ** NS NS NS NS NS NS NS ** NS NS * NS NS NS NS NS ** * NS FR*SR NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS SR*REP NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS SR*LOC *** NS * NS * NS NS NS NS NS NS NS NS NS NS ** NS ♦ NS NS NS NS NS NS Mean 5.96 4.54 5.22 16.1 14.1 14.4 37.6 32.5 36.4 541 493 573 106 116 105 37.7 33.1 32.6 46.9 46.9 45.5 74.5 76.8 76.1 CV (%) 11.2 15.7 10.1 17.2 17.8 14.2 10.8 9.5 11.8 13.3 8.9 9.8 3.6 17.5 2.8 5.4 4.8 4.6 11.2 13.6 13.1 1.7 2.1 1.6

28 Location and varietal effects Location cffccts on varietal responses arc illustrated in Figures 2, 3, 4. ETBW5795 was the highest yielding at Kulumsa (7.67 t ha'1) with yield increments of 964 and 1553 kg ha'1 as compared to Kofele and Sagurc area (Digelu-Tijo district), respectively. HAR3116 performed better at Kofele district with the highest economical yields of 5.9 t ha' 1 and yield increments of 705 and 815 as compared to the cl feet at Kulumsa and Digelu Tijo (Sagurc), respectively. ETBW5483 outperformed at Kofele and Kulumsa (Tiyo) with the highest AGY of 6.54 and 6.51 t ha'1, respectively. The differences in the varietal responses at each location are illustrated in figures 5, 6, and 7. Both ETBW5795 (6.7 t ha'1) and ETBW5483 (6.5 t ha'1) outperformed HAR3116 (6.0 t ha'1) at Kofele. In the other two locations, the highest yield was recorded by ETBW5795 with 6.1 (Digelu-Tijo) and 7.67 t ha'1 (Kulumsa). Using SPSS20 linear regression model and setting Digelu-Tijo district (Sagure area) and HAR3116 variety as the dumy variables, the effcct of location on the varietal responses was analyzed. The average responses of the three varieties at Kulumsa (Tiyo district) and Kofele was higher than Digelu-Tijo. ETBW ^795 and ETBW5483 out yielded HAR3116 across locations with mean extra yields of 1420 and 656 kg ha’1, respectively.

Fig 2 Grain yield Response curve of ETBW5795 4.2.1 Fig 3 Grain yield response curve of under 3 locations: Kofele, Sagure, Kulumsa HAR3116 at Kofele (blue color), Sagure (green), and

29 Fig 4 Grain yield response curve of ETBW5483 at 4.2.3 Fig 5 Grain yield response curves of ETBW 5795 Kofele (bl ue), Sagure (green), and Kulumsa (ye llo w ) (red), HAR3116 (green), and ETBW5483(blue) at Kofele

Fig 6 Average grain yield response of ETBW5795 Fig 7 Grain yield response of ETBW5795 (red), (red), HAR3116 (green), and ETBW5483 (violet) at Sagure HAR (green), and ETBW5483 (violet) at Kulumsa

SR and FR effects on AGY and ABY One-Way analysis of variance for seed rale (SR), fertilizer rated (FR), and variety are presented in table 1. Differential grain yield responses of varieties was analyzed to SR and FR at different locations i.e., FR*SR arranged per variety and location wise (Table 2); shows the main effects of treatment factors on selected crop response parameters The main effects of both SR and FR were significant and both have positively affected AGY and ABY. SPM and NSPS were also positively affected by SR and FR increments. At Kulumsa (Tiyo district), ETBW5795 responded positively to either increased SR (up to 175 kg ha'1) under RNP or to increased FR. The same variety at Kofele responded positively to decreased SR at the RNP, but to increased SR of up to 175 kg ha 1 at higher fertilizer rates. When the FR was at 2RNP. the recommended SR (125 kg ha l) gave high yields of 6.6 t ha'1. Wlicn the FR was 1.5RNP, a SR of 225 kg ha'1 gave 6.7 t ha against 5.4 t ha'1 yield obtained from recommended SR and FR. At Digelu-Tijo, u responded to increased fertilizer rates. HAR3116 responded best at the recommended lowest fertilizer rate and recommended SR (125 kg ha'1) al Kofele; il responded better at the lowest FR and increased SR (175 kg h a1) at Kulumsa, and to increased FR at Digelu-Tijo, with yield increments of 941 and 1231 kg ha :, respectively, than the recommended rates. ETBW5483 responded positively to increased SR at the RNP at Kulumsa; although, the highest yield (6.5 t ha ) obtained was at increased FR and recommended SR (2RNP/125 kg ha'1 seed) compared to 5.4 t ha ! of the recommended fertilizer and seed rate. At Kofele, the same variety responded more positively to increased SR than FR. At Digelu-Tijo, it generally responded positively to both increased FR and SR.

Similar trial was conducted under two farmers’ fields at Digelu-Tijo and on-station at Tiyo (Kulumsa) and Kofele districts using two varieties (HAR3116 and ETBW5795). The results at Digelu-Tijo district showed that Hidase variety (ETBW5795) responded positively to increased FR, and also to increased SR up to 175 kg/ha. An average yield increment of 2374 kg ha ! was obtained at 138-92 N- P20 ? kg ha'1' seed rate of 175 kg ha'1 than responses al RNP and recommended SR. At kulumsa, the response of ETBW5795 to increased FR was positive, but with no response to increased SR above the recommended. The yield increment was 805 kg ha at 138-92 N-P20^ kg ha' and seed rale of 125 kg ha'1. Location effects on the responses of HAR3116 to increased FR were highly significant. At Digelu-Tijo, positive response with yield increments of 1404 kg ha at 138-92 N-P2O5 kg ha 1 was abserved as compared to RNP. At Kofele, the response to increased fertilizer rate above the RNP was negative.

30 Seed and fertilizer rate effects on grain quality The effect of seed rate and fertilizer rate on grain quality parameters and leaf N absorption was studied using samples from two on-farm trials, which contained two varieties (HAR3116 and ETBW5795). Both varieties were significantly (p< 0.0001) affccted by increased fertilizer rates (Table 4). The protein content of varieties increased from 9.8 and 10.9 at FR1 to 11.0 and 12.52 at FR3. The wet gluten contents increased from 23.98 to 27.99 for HAR3116 and 24.0 to 28.71 for d ETBW5795. Similarly, the zeleny values increased with increased fertilizer rates. Varietal variations were very high i.e., ETBW5795 responded better than HAR3116 across treatments. Generally, the results grain quality analysis agrees with Genene et al. (2003).

Table 4 ANOVA result of treatment effects on grain qualities presented variety wise

HAR3116 ETBW5795 Source of Protein Starch Wet Gluten Zeleny Protein Starch Wet Gluten Zeleny Variation (%) (%) (%) (ml) (%) (%) (%) (ml) SR DMRT NS NS NS NS NS NS NS NS FR FR1 9.8 66.5 24.0 27.4 10.9 65.7 24.0 31.0 FR2 10.4 68.4 25.8 30.2 11.8 65.1 27.5 34.3 FR3 11.0 676 28.0 32.0 12.5 64.6 28.7 37.6 DMRT kk NS kkk kkk kkk kkk kkk *★* Mean 10.4 67.5 25.9 29.8 11.7 65.1 26.7 34.2 CV (%) 4.7 9.2 5.4 5.8 5.6 1.3 12.0 8.5

Leaf absorption of N was also significantly (p< 0.05) affccted by fertilizer rates with the response of FIAR3116 higher than ETBW5795. The leaf N content increased from 3.12 to 3.50 for HAR3116 and from 2.62 to 3.05 for ETBW5795. The effect of seed rate was not significant on leaf absorption of N. However, the effect of seed rate on the fertilizer recovery and agronomic efficiency would be different.

Fertilizer N recovery and agronomic efficiency The higher response of ETBW5795 in grain qualities to increased fertilizer rate was discusscd in previous section. Nevertheless, the statistical analysis did not show significant effect of seed rate on grain quality for both varieties in contrary the effcct of seed rate on the NRE and AE of the two test varieties. The NRE and AE of ETBW5795 increased dramatically al every level of increase in seed rate o 'up lo 175 kg ha'1 (Figure 8 and 9). The NRE of HAR3116 increased from near 0 % to mean 15.1% when the fertilizer rate increased from 46-46 N-P:05 kg ha'1 to 92-46 N- P20? kg ha'1. The change in seed rate of FIAR3116 al every level of fertilizer rate did not bring any positive effcct on the efficiencies. On the other hand, the effect of the treatments on the NRE and AE of ETBW5795 was different. The NRE constantly increased from 6.4% lo 44.4% al every level of seed rate when the fertilizer rate is increased from 46-46 to 92-46 N-P2O5 kg ha'1. Both seed and fertilizer rates affected the AE of the higher yielding variety (ETBW5795). AE increased from 2.6 to 18.7 kg of grain/kg of N. The very high AE and NRE of ETWB5795 was a consequence of its high yield response to both seed and fertilizer rate increased applications. The rate of increase in the efficiencies declined above the 92-46 fertilizer treatments. The results agree with Genene et al. (2003).

31 TRT (Kg/ha) TRT (Kg/ha) Fig 8 Varietal differences in the effect of fertilizer 4.2.4 Fig 9 Varietal differences in effect of seed rates and seed rates treatment on N recovery efficiency and fertilizer rates on agronomic efficiency

Bread wheat cultivars varied in their yield and yield component responses to the treatments. Site and locations also differentially affected the capacity of the three varieties to express their yield potential. ETBW5795 (Hidase) gave the highest yields al all locations. It responded positively to both increased fertilizer and to seed rates of up to 175 kg ha1. The second highest yielder was ETBW5483 (Shorima).

Conclusions and Recommendations

The response of HAR3116 (Degelu) varied from location to location where its response was the highest at the recommended seed and fertilizer rates at Kofele. Increased seed and fertilizer rates above the recommended decreased its responses. At Sagure (Digelu-Tijo district), Degelu responded positively to both increased fertilizer rates and to seed rates of up to 175 175 kg h a 1. At kulumsa, the highest response was obtained at the recommended fertilizer rate (RNP) but at increased seed rate (175 kg ha 1). Their responses to increased seed rate remain constant or unchanging above the recommended fertilizer rate.

Shorima variety responded positively to both increased seed and fertilizer rates at Digelu-Tijo. At Kofele, increased yield response was obtained at increased seed rate indicating that the recommended fertilizer rate was sufficient. At Kulumsa. increased fertilizer rates increased its yield with seed rates up to 175 kg ha'1. But at the highest fertilizer rates (2RNP), it gave the highest grain yield (6.54 t ha ) at the recommended seed rate.

The variations in the responses of the tested cultivars across locations could be due to the differences in the soil fertility status. The soils of Kofele include humic Nitosols; it is fertile and docs not require much fertilization. Increasing seed rates from blanket recommendation of 125 kg ha 1 for row planting lo 175 kg ha 1 can provide additional yield advantages of up to 0.5 t ha'1 or more. Yet. other factors should be considered for investigation to further improve wheat productivity in the area.

Degelu-Tip area was generally lower yielding though highly responsive lo increased fertilizer rates and seed rales of up to 175 kg ha 1 (or 200 kg ha 1 as it is a Vcriisols area). Generally, the lower yield obtained could be attributed to the seedbed preparation method. The furrow spaces used for drainage in the ridge and furrow system of seedbed preparation method, commonly used in the areas, can rcduce the land use efficicncy of the soils. During heavy rains of August, the furrow widths become wider killing plants at the edge of furrows and consequently reducing yields. Therefore, an alternative seedbed preparation method, like the broad bed and furrow for traditional oxen plough and the camber bed for mechanized farming introduced to increase the productivity of w heat and their yield potentials in Degelii-Tijo.

32 The differential effect of seed and fertilizer treatments on grain qualities, NRE, and AE of ETBW5795 and HAR3116 is also very useful information. Both varieties responded positively in grain protein, wet gluten contents, and zeleny values to increased fertilizer rates with higher records from ETBW5795. On the other hand, the NRE and AE of ETBW5795 dramatically increased with increases in both seed and fertilizer rates, while NRE and AE of HAR3116 were affected by fertilizer rates increments only.

The very high NRE and AE of ETBW5795 was a consequence of its high yield potential and high response to increased fertilizer rates and population densities. The additional attribute of such cultivars could provide additional merits as a choice for wider dissemination under different favorable environments, from agronomists’ point of viewr. Its resistance to diseases (rust and septoria) under some environments should be cautioned before wider circulation. The results obtained under seven sites in 2013 alone cannot be used to generalize recommendations at respective locations as fertilizer management is complicated by variations in cropping systems and management, microclimate, and landscape configurations. This trial will be repeated in 2014 cropping season. Finally, results shall exclusively apply to subsistence farmers whose farming system is based on traditional tillage and planting method.

References

Amanuel Gorfu, Asefa Taa, Tanner DG and Mwangi W. 1991. On farm research to Derive Fertilizer Recommendations for Small-Scale Bread Wheat Production: Methodological Issues and Technical Results. Research Report No 14. IAR, Addis Ababa, Ethiopia. 37pp. Asefa Taa, Kefyalew Ginna and Shambel Maru. 1997. On-farm N and P fertilizer trial in Bread wheat on vertisols in South eastern Ethiopia. Agronomy and Crop Physiology progress report. Kulumsa Agricultural Research Center. EIAR. Ethiopia Carter MR and Gregorich EG. 2008. Soil Sampling and Methods of Analysis. Canadian Society of Soil Sciences. Tailor and Francis Group. LLC.6000 Brocken Sound Parkway. NW. Suite 300 Cassman KG, Doberman A and DT Walters. 2002. Agro-ecosystems, nitrogen use efficiency, and nitrogen management. Ambio 31:32-140. Doyle DF and Holford ICR. 1993. The uptake of nitrogen by wheat, its agronomic efficiency and their relationship to soil and fertilizer nitrogen. Aust. J. Agric. Res. 44:1245-1258. Environmental Systems Research Institute (ESRI). 1991-1999. Arc View GIS 3.2. Software Program. USA. Etliio-Italian Developemtn Cooperation. 2002. Atlas of Arsi zone. Arsi-Bale Rural Development Project. Arsi zone Planning and Economic Developemnt Office. Regional State of Oromia.GIS sub unit. Fager a NK and Baligar VC. 2005. Enhancing Nitrogen use Efficiency in Crop Plants. Advances in Agronomy, Volume 88 (97-185). Elsevier Inc. Fager a NK and Baligar VC. (2003). Methodology for evaluation of lowland rice genotypes for nitrogen use efficiency. J. Plant Nutr. 26:1315-1333. Genene Gezu. 2003. Yield and Quality Response of Bread Wheat (Triticum aestivum L.) Varieties to Rate and Time of Nitrogen Fertilizer Application in Kulumsa, Southeastern Ethiopia. MSc Thesis. Alemaya University, Alemaya. IBM Corporation. 1989-2011. SPSS Statistical Software. Version 20. Minale Liben, Alemayehu Assefa, Tanner, D.G. and Tilahun Tadesse. 1999. The response of bread wheat to N and P applications under improved drainage on Bichena vertisols in northwestern Ethiopia, pp. 298-308. In: The Tenth Regional Wheat Workshop for Eastern, Central and Southern Africa. Addis Ababa, Ethiopia: C1MMYT. Minale Liben, Alemayehu Assefa, Tilahun Tadesse and Abreham Mariye. 2004. Response of Bread Wheat to Nitrogen and Phosphorous Fertilizers at Different Agroecologies of Northwestern Ethiopia. In: Proceedings of the 12th Regional Wheat Workshop for Eastern, Central and Southern Africa. Nakuni, Kenya. Novoa, R and Loomis RS. 1981. Nitrogen and plant production. Plant Soil 58, 177-204. Revised edition. Mexico, D.F. Origin Lab Coorporation. 1991-2007. Origin 8 Graphical User Interface (GIU). Northampton. MA 01060 USA SAS Institute. 2002. The Statistical Analysis Software System for Windows. Version 9.00. TS Level 00M0. SAS Institute Inc., Cary NC. USA

33 Shambel Maru, kefyalew Girma, Workiye Tilahun, Anianuel Gorfu and Mekonnen Kasaye. 1999. On-farm N and P fertilizer trial in Bread wheat on vertisols in South eastern Ethiopia. Agronomy and Crop Physiology progress report. Kulumsa Agricultural Research Center. EIAR.Ethiopia Tanner DG, Amanuel Gorfu and Asefa Taa. 1993. Fertilizer effects on sustainability in the wheat-based small holder farming systems of southeastern Ethiopia. Field Crops Research 33:235-248. Taye Bekeb, Yesuf Assen, Sahlemedhin Sertsu, Amanuel Gorfu, Mohammed Hassena, Tanner DG, Tesfaye Tessemraa and Takele G. 2002. Optimizing Fertilizer Use in Kthiopia: Correlation of Soil Analysis with Fertilizer Response in Hetosa Woreda, Arsi zone. Addis Ababa, Sasakawa-Global 2000. Tilahun Geleto, Tanner DG, Tekalign Mamo and Getinet Gebeyehu. 1996. Response of rainfed bread and durum wheat to source, level and timing of nitrogen fertilizer on two Ethiopian Vertisols: II. N uptake, recovery and efficiency. Fertilizer Research 44:195-204. Tisdale LS, Nelson WL1, Beaton JD, Havlin JL. 2002 Soil Fertility and Fertilizers. 5ed. Prentice Hall. India. New De ihi. Responses of Bread Wheat Varieties to H and P Moist and Humid Midhighland Vertisols of Arsi Zone, Ethiopia

Dawit Habte1, Kassu Tadesse2, Wubengeda Admasu3, Tadesse Desalegn3’’ Asrat Mekonen4 12 3 4 Ethiopian Institute of Agricultural Research, Kulumsa Agricultural Research Center, Wheat Regional Center of Excellence, P.O. Box 489, Asella, Ethiopia 3'Eastern Africa Agricultural Productivity Project (EAAPP), Kulumsa Agricultural Research Center, Wheat Regional Center of Excellence, P.O. Box 489, Asella, Ethiopia

Abstract Field experiment was conducted in 2012 and 2013 in ihree districts of Arsi zone, Ethiopia, to evaluate the response of bread wheat cultivar “Danda’a” under highland vertisols. The treatments consist of 20 factorial combinations of five N rates from urea and four P2O3 rates from TSP. The treatments were laid oul in a complete factorial arrangement using RCBD replicated three times. All yield and yield components, plant N absorption, and grain quality data were collected and analyzed and differences among treatment means were assessed. Grain and biomass yields prediction models were developed and graphical analysis of yield and yield components was made. Economic profitability of the various treatment options were discovered through partial budget analysis. Grain and biological yields across different treatment effects were significant (P<0.00()l). Number of spikes m'2, number of seeds per spikes, and plant height were also significantly affected. The main effects of fertilizer N on grain yield, biomass yield, protein content, wet gluten content, and zeleny values were also highly significant. Leaf absorption of N was increased with increased rates up to 92 kg ha-1 N. The N recovery efficiency at 46 and 92 kg N ha-1 was 20.9% and 29.4% and the agronomic efficiency was 10.8 and 13.3 kg grain/kg N applied, respectively. Above 92 kg ha- 1 N, the increase in both N recovery efficiency and agronomic efficiency declined or fell reaching 31.4% and 12.6 kg grains/kg N. Hence, 92-46, 138-69, and 115-46 kg ha- N-P2O5 recommendations were made based on agronomic data, economic analysis, capacity of fanners to invest, the complexity in management history of different farms, the need to progressively meet the high national yield goals, and environmental considerations.

Introduction

The relatively slow growth in mean national yield for bread wheat (Triticum ctestivum L.) from 1.46 t ha‘l in 2004/2005 (CSA 2005; Kate and Leigh 2010) to 2.01 t ha'1 (CSA 2011) is due lo several constraining factors, such as poor crop management that include the prevalence of poor weed control, exacerbated by the limited availability of herbicides in the market and its improper use when available, depleted soil fertility and a low level of fertilizer usage, particularly of N fertilizer are among the most important factors (Tanner et al. 1993; Payne et al. 1996). Nitrogen and phosphorus deficiency is often encountered in wheat growing areas of Ethiopia, in which the severity of the problems predominate the frequently waterlogged soils- highland Vertisols (Tekalign et al. 1988; Syers et al. 2001, Asgelil et al. 2001). The very high potential of vertisols for wheat production have been w ell recognized by highland farmers sincc the introduction of vertisols technologies in the 1990s such as BBM and ridge and furrow seed bed preparation methods. Nonetheless, it has been unden. tilized mainly due to the very low input use of fertilizers and poor pest management strategies. Development of site and crop specific fertilizer recommendations has shown modest progress possib y due to the limited resources. The old bulk recommendations continued to be practiced in many areas. The importance of specific fertilizer recommendation has gained the attention since the 1990s. In this respect, the progress indicated that increased applications and rates of N and P increased grain yields with a very strong and significant linear response (Amanucl et al. 1991; Tanner et al. 1993; Asefa at al. 1997; Shambel et al. 1999; Minale et al. 1999, 2004; Amsal et al. 2000; Taye et al. 2002; Tcklu et al. 1996, 2000). Further, such efforts should continue considering the variability of soils, climate, and cropping systems. The demand for site-specific fertilizer recommendations has been increasing from time to time. Development of agroecology based or site-specific N and P fertilizer recommendations and increased implementations of recommended practices is one of the

35 primary means of increasing wheat yields in Ethiopia. Therefore, a fertilizer trial was conducted in 2012 and 2013 with the objective of developing economic optimum fertilizer recommendation for bread wheat productions in three highland Vertisols dominated districts of Arsi zone, Ethiopia.

Materials and Methods

The experiment was conducted on farmers' fields in poorly drained heavy dark clayey soils (pellic Vertisols) of Digelu-Tijo, Arsi Robe, and Tiyo districts. Arsi Robe, Digelu-Tijo and Tiyo are located from 8.4N to 8.6N and 40.IE to 40.4E, from 8.01N to 8.15N and 039.15Eto 039.3E and from 7.77N to 8.03N and 38.94 to39.31E, respectively, all in degree decimal. The altitudes of the locations vary from 2200-2500 masl. The long-term average annual rainfall for Arsi Robe is 1040 mm, above 840 mm for Tiyo and similar for Digelu-Tijo (800 and 1000 mm). Tepid to cool moist mid-highlands and Tepid to cool humid mid-highlands are the agroecological classification for the study areas (Ethio- Italian Development Cooperation 2002). Even though the long term average annual rainfall for Arsi Robe is higher than the other location, its distributions arc uneven.

The exper ment was conducted to evaluate the response of recently released bread wheat variety Danda’a (KIRITATI/2*PBW65/2*SERI.IB; Damphe) to treatments consisting of 20 factorial combinations of five N rates (0, 46, 92, 138, 184 kg ha'1) from urea and four P2O5 rates, i.e. 0, 46, 92,138 kg ha'1 from triple super phosphate. Complete factorial arrangement using RCBD replicated three times. The gross plot size of the trial was 4 m x 5 m (20 nr) and net plot size of 3 m x 3 m (9 m2). The fields were prepared using the traditional oxen-plow system of the ridge and furrow with 0.7 m wide inter-furrow spacing. Seeds were broadcasted on the plots and then the ridge and furrows were prepared with well experienced fanners to keep the inter-furrow spacing of 0.7 m using a small ridge and furrow maker commonly called BBM. All P fertilizer and half of the N fertilizer treatments were applied at planting and the remaining N was top-dressed at booting stage. Existing recommendation of seed rate (150 kg ha 1) and onetime herbicide "pyroxyslam" sparyed for weed control.

Data was recorded on grain yield and yield components i.e., seedling density, number of tillers per plant, spike length, kernels per spike, thousand kernel weight (TKW), plant height, grain and biomass yield. Information on disease and pest incidences was also collected. Plant samples and grain samples were collected from two experimental sites at Digelu-Tijo district at the start of heading and after harvest, respectively.

Plant height and number of productive spikes m 2 were determined for each treatment before harvest. Hundred culm weight (100 cw) were collected from four to five points within a plot and slashed from elose to the ground surface and the dry matter yield of above-ground biomass determined. Grain yield (12.5% moisture) was determined from 9m: net plot. Above ground biomass yields were determined Harvest index (HI) calculated as the ratio of grain yield to above ground biomass yield expressed as a percentage. The number of grains spike'1 was determined by hand counting of the number of grains of 5 spikes. Straw N contents were determined by micro-Kjeldahl analysis of straw sub samples (Bremner and Mulvaney 1982) at Kulumsa Agricultural Research Center soil laboratory from the oven-dried bulk samples. Grain Protein, starch, wet gluten, and zeleny values were determined using Near Infrired Reflectance (NIR) at Amhara Region Agricultural Research Institute (ARARI), Bahirdar, Ethiopia. Grain N values were calculated by multiplying grain yields by the respective N content. Apparent N recovery (AR) of the grain for each treatment N was calculated as: (GNU of treatment - GNU from the control treatment) / fertilizer N applied. The GNU values were calculated from N treatments averaged over the ranges of P2O5 levels and replications. Agronomic efficicncy (AE) of fertilizer N was calculated as: (grain yield of treatment - grain yield of control)/Fertilizer N applied. Again, the main effects of N were considered. Efficiency values were calculated based on known procedures (Novoa and Loomis 1981; Cassman and Dobermann 2002; Fageria and Baligar 2003; Doyle and Holford 1993).

36 Partial budget analysis (CIMMYT 1988) was used to evaluate the economic profitability of the various treatment options and determine the economic optimum rate. Prediction of yield response under alternative fertilizer treatments was generated using the regression model. All variable costs including land preparation, planting, weed control, and harvesting costs are estimated based on the actual field prices at the lime of planting and immediately after harvest and averaged over locations. The yield data used for economic analysis was 2013 data due to better management conditions. The costs of P2O5 and N fertilizer were estimated based on the cost of DAP and Urea, respectively. The Urea atcs were adjusted based on the contribution ofN from each treatment level of DAP to N source levels. Dominance analysis (CIMMYT 1998) was applied to screen treatments with higher variable costs, but lower net benefits; and dominated treatments eliminated from further considerations in Marginal analysis. The minimum acccptable rate of return was taken as 100%; and treatments with lower minimum rates of return were also removed from further analysis. Finally, sensitivity analysis was conducted on the selected best treatments lo evaluate the effect of variability in input prices over time and space on the strength of acceptability of recommended practices under all recommendation domains.

All c op parameters data were subjected to analysis of variance using SAS 9.0 statistical software (SAS 2002). Data were analyzed for trials combined across site and seasons. The DMRT test (P<0.05) was used lo assess differences among treatment means. SPSS 20.0 statistical software was used to analyze the correlations between yield and yield components and the treatments, and for developing prediction models for grain and biomass yields. For graphical analysis of yield and yield components Origin 8 GUI and SPSS 20.0 softwares (Origin Lab Coop. 1991-2007; SPSS 1989-2011) were used.

Results and Discussion

Grain and biomass yield responses to N and P205 rates Grain yields and yield components under different fertilizer rates arc presented in tables 1 and 2. Main effecis of each fertilizer rates on the yield and yield component responses across locations and years are summarized. The main effects of N and P20 5 on grain yield (GY) and biomass yield (BY) arc also illustrated in figures 1-4. The GY and BY at Arsi Robe and Digelu-Tijo districts in 2012 showed that main effects of N and P arc significantly different (p<0.0001) with mean grain and biological yields of 2861 and 6940 kg ha' 1 as compared lo the control 1606 and 3910 kg h a 1, respectively. The highest GY' and BY at Robe were 4229 and 11145 kg ha''and at Sagure and Tiyo 4658 and 9745 kg ha'1, respectively.

Main effects of N and P2O5 were significantly different (p<0.0001) with mean grain and biological yields of 4134 and 9831 kg ha'1, respectively, at Digelu-Tijo and Tiyo districts of Arsi zone in 2013. In contrast, to the control treatment results of 2589 GY and 6554 BY kg ha'1, the highest records were 5431 GY and 13299 BY kg ha'1. The highest mean number of spikes per meter (SPM) and number of seeds per spikes (NSPS) at all sites of the two districts in 2013 were 512 and 43.5, as compared to the controls results of 367 and 36.6, respectively. SPM, NSPS, and plant height (PH) were significantly affected by increased rates of N at all locations in the both seasons.

37 Table 1 Effects of N and P2O5 application on selected agronomic parameters grown on highland ertisols of Arsi zone in 2012 and 2013

2012 summary results of Robe and Digelu-Tijo districts 2012 and 2013 summary results of Digelu-Tijo and Tiyo districts PH NSPS SPM GY BY HI HLW PH NSPS SPM GY BY HI HLW (cm) (ka ha-i) (ka ha-1) (cm) (ka h * 1) (kg ha-1) N (kg ha-1) 0 84.4 41.5 241 1926 5106 39.5 76.3 75.5 38.4 360 2385 6176 39.7 75.0 46 89.6 47.0 252 2554 6067 40.2 75.5 82.8 40.9 376 2852 7160 40.2 75.4 92 101.4 50.8 266 2960 7103 40.1 75.4 87.8 41.6 423 3370 8327 41.0 74.5 138 98.4 55.3 268 3368 7935 40.0 74.7 92.6 43.4 430 3916 9381 42.2 74.9 184 98.0 54.8 263 3516 8685 39.0 74.6 93.8 45.2 461 4271 10365 41.6 74.4 ** *** * *** *** * *** *** DMRT NSNS NS NS P2O5 (kg h a j 0 88.4 47.1 247 2287 5603 39.9 75.6 82.2 40.6 386 2788 6846 40.6 75.1 46 93.1 49.2 249 3004 7018 40.5 75.07 88.1 42.1 416 3511 8551 41.5 75.0 92 94.6 49.9 260 2973 7303 39.9 75.4 88.3 41.9 432 3653 9087 40.5 74.2 138 101.2 53.0 274 3205 7876 39.1 75.12 88.4 43.3 417 3637 8952 41.5 74.7 *** *** *#* *t* *** *** DMRT NS NS NS NS NS NS N‘ P NS NS NSNS NS NS NSNS NS NSNS NS NS NS *** * * REP NSNS NS NSNS NS NS NS NS NS NS N*REP NS NSNS NSNSNS NSNS NSNS NS NS NS NS P2O5 * REP NSNS NS NS NS NS NSNS NSNS NS NS NS NS Control 81.3 40.4 231 1606 3910 40.7 76.4 70.3 36.5 337 1969 5266 38.47 75.02 Mean 94.4 49.8 258 2861 6940 39.8 75.3 86.7 41.96 413 3390 8341 40.98 74.8 CM (%) 19.3 14.2 15.8 20.4 23.2 15.9 4.5 8.96 24.6 28.9 32.9 30.82 14.6 3.94 *, **, and *** shows significance at p<0.01. p<0.001 and p<0.0001, respectively; NS - nonsignificant

38 I

Table 2 Effects of N and P2O5 application on selected agronomic parameters grown on the highland Vertisols of Arsi zone in 2013

PH NSPS SPM GY BY HI (cm) (Kg ha-1) (Kg ha1) N (Kg ha-1) 0 77.4 39.2 408 3025 7450 40.6 46 84.2 39.3 439 3769 9249 40.5 92 86.8 41.0 465 3973 9245 43.8 138 91.2 40.4 497 4755 11069 43.0 184 93.4 43.5 501 5111 12170 42.6 DMRT *** NS *** hic k *** NS P2O5 (Kg ha-i) 0 83.7 40.6 445 3653 8658 42.1 46 87.5 41.3 463 4206 9993 42.2 92 88.6 41.1 472 4370 10428 41.9 138 87.3 40.3 474 4312 10264 42.9 DMRT NSNSNS ** NS TRT *** NS 4 ** *** *** NS N*P NSNSNSNSNSNS REP NSNSNSNSNSNS N*REP NSNSNSNS NS NS P2Os*REP NSNSNSNSNSNS Control 72.8 36.6 367 2589 6554 39.4 Mean 86.8 40.8 463 4134 9831 42.3 CV (%) 7.7 18.0 12.2 22.9 23.3 13.2

Generally, increased rates of P2 O5 significantly (p<0.05) affected SPM and PH while NSPS did not show any significant difference in across years. Nonetheless, effects of P2 0 5 on NSPS were significant at some individual sites. The interaction effects of N and P205 are non-significant for most parameters at all locations. N than P. significantly controlled the grain yield component parameters such as SPM, NSPS, and TKW. But, P was crucial to increase the yield responses within the ranges o f 46-92 kg ha' 1 P2 0 5 (Figures 1 and 2). Generally, the yield increments in 2012, 2013, and combined across years vary from 111-146% and 103-142% for the highest treatments of GY and BY, respectively, as contrasted to the control treatment.

The relationship between GY and different rates of N and P can be expressed using the following second-degree polynomial equation, with R2 =0.97 for GY. The predicted average grain yield is expressed by:

PGY = c + aN + bP2Os - dN2 -e(P2Os )2 + f(N * P20 5) Where: PGY is predicted grain yield, c is a constant with a value o f 2667 kg ha'1; a, b, d, e, and f are coefficients with values o f 12.84, 12.7, 0.015, 0.07, and 0.019, respectively. The parameters PGY, N, P2O5 are all in kg ha'1.

Grain yield responses of the test variety to N and P2 0 5 combinations, not included in treatments, were predicted using the equations developed for GY and the values applied during economic analysis. The results obtained so far are in agreement with the works done on bread and durum wheat by Amanuel et al. (1991), Asmare et al. (1995), Teklu tesfaye (1996), Asefa et al. (1997), Shambel et al. (1999), Minale et al. (1999, 2004), Tanner and Payne (2001), and Teklu et al. (2000).

39 ■

N levels as P2O5 levels (colored curves) change from 0-138 kg ha-1 levels ( colored curves) change from 0-138 kg ha '1

Fig 3 Relationsiip between GY responses to P2O5 levels as N levels Fig 4 Relationship between BY responses to P2O5 levels as N (colored curves) change from 0-184 kg ha-’ levels (colored curves) change from 0-184 kg ha-1

Effect of N and P205 on grain quality One hundred twenty grain samples obtained from two sites of Digelu and Tijo District were subjected to grain quality analysis using NIR at Amhara Region Agricultural Research Institute (ARARI), Bahirdar, Ethiopia. Increased rates of N had significant effect on grain protein and wet gluten contents, and zeleny values. Responses of the variety in grain quality parameters measured from samples of two sites in 2013 are shown in Figures 5, 6 and7. The average protein contents obtained from highest lo lowest were 13.4, 12.5, 12.0, 11.0, 10.8% from applications of 184, 138, 92, 46, 0 kg ha 1 N, respectively. The wet gluten contents from highest to lowest: 30.7, 29.6, 27.52, 23.7, and 23.6% were obtained from applications of 184, 138, 92, 46, and 0 kg h a1, respectively. Similarly, N had very high significant effcct on the Zeleny values, but no significant effect on starch content. Generally, N increased the grain protein content, wet gluten content and zeleny values by 25%, 34% and 44%, respectively, at the highest rate of N. The result was clearly indicative that N was the main determining factor for improving the grain quality of bread wheat. The effect of P nutrition on the grain quality was insignificant (Fig 6). Protein and starch contents were not significantly affcctcd by increasing rates of P, but Zeleny values negatively affected. The wet gluten content slightly raised at the 92 kg ha 1 P 2 O 5 rates.

40 Fig 5 Relationship of grain quality parameters with N nutrition across wo locations nutrition across two locations

Fig 7 Relationship between GY and grain N with changes in N and P treatment combinations

Another interesting relationship was between grain N uptake and GY (Fig 7) which showed similarity between the two curves. At the point where the P2Os values fall to 0 kg ha'1, both grain N uptake and GY fall together, indicating the complementary effects of P20 5 for grain yield and N uptake.

Leaf analysis Hundred twenty-leaf samples were collected from two sites of farmers’ fields for analysis of N absorption by the plant. Absorption of N was significantly (p>0.001) affected by increasing N rates. The mean absorption was 2.6%. The highest rates of absoiption was 2.9% from plots that received 138 kg ha 1 N followed by 92 kg ha'1 N rate (2.8%) and 184 kg ha'1 N rate (2.58 %). The lowest absorption recorded was from plot without N application. Contrary to the effcct of the highest rates of N on the grain protein or N contents, leaf absorption was intermediate to the highest rates. This may be due to, possibly, higher associated losses. The effect of different rates of P on the leaf absorption was insignificant; and no interaction cffect of N and P was observed.

Fertilizer N recovery and agronomic efficiency Fertilizer N uptake and agronomic efficiency was calculatcd based on data from quality analysis. The N recovery efficiency (NRE) steadily increased with N rates up to 92 kg ha'1, remained constant between 92-138 kg ha‘! N, and slowly rose after that. Similarly, the agronomic efficiency (AE) of N increased up to 92 kg ha'1 N, and then declined. The NRE was 20.9% and 29.4% and AE was 10.8 and 13.3 kg grain/kg N applied at 46 and 92 kg N ha'1, respectively. Between 92-138 kg N ha'1, the mean NRE and AE was 29.3%. The highest NRE was 31.4 % at the highest N. On the other hand, the AE at the rates of 138-184 was 12.6 kg grain kg ha'1 of N applied. Generally, the highest response values

41 were obtained at 92 kg N ha ! (Fig 8). The results obtained so far are also in agreement with the works done on bread and durum wheat (Tilahun et al. 1996; Bemnet et al. 2006).

axis) and AE (right y-axis)

Economic analysis Every shift in investment from the lower to higher selected treatments resulted to more than 100% return (Fid 9). The marginal rate of return (MRR) generally varied from 1.14 lo 4.19. The highest MRR was obtained from applications of 92-46 kg ha 1 N-P20>. Further increases in fertilizer use, particularly of nutrient N, still had positive yield rewards. The values in the vertical axis show the return in birr for every 1 birr invested on fertilizers. The highest MRR (4.19) was obtained from application of 2.61 units of fertilizer, which is equal to 160 kg ha 1 Urea + 100 kg ha 1 DAP. The MRR generally declines with increased applications of Fertilizer N (Fig 10).

o • » 57 a ci icn s h h »n cqrt

Fig 9 Relationship between total variable costs and net benefits as Fig 10 Graph showing the relationship between MRR and per treatments selected based on dominance analysis Units of fertilizer applied

Sensitivity analysis was made based on data used in MRR analysis and with treatment results above 100% minimum rate of return, except for the control. The if-analysis was done with the assumption of an average of 30% rises in all variable costs within 3 years time, keeping the prices of the produce constant. The analysis showed that the recommended rates still hold positive benefit cost ratios.

42 Conclusions and Recommendations

Using below optimal fertilizer rate is one of the major bottlenecks for improving wheat yields in almost all highland Vertisols areas of Ethiopia. Farmers also consider as the most important input to sustain or improve their production. Flowever, use of the appropriate quantity of N and P should be the major consideration to increase productivity, reduce poverty and food insecurity.

Three forms of recommendations are made considering the various factors affecting decision making i.e., fertilizer recommendations should be flexible. If farmers cannot afford to cover the costs of the highest recommended fertilizer rates, they normally perceive that any other lower rates cannot work. Consequently, they resort back to their already adapted practices. Therefore, due consideration w'as given to the various factors that need to be considered for pract cal applications and acceptance by various stakeholders. Based on farmers preference of fertilizer use rate and their tendency to gradually adopt higher rates the 92-46 (N- P2O5) kg ha'1, which is equivalent to 160 kg ha'1 Urea + 100 kg ha"1 DAP is recommended. This rate w as with the highest marginal rate of return (MRR). It would be better to advice poor fanners to start with this lowest level recommendation. With the increasing benefits they experience they can progressively develop to higher levels. Based on the need to attain the long term high yield goals set by planners, the 138-69 kg ha'1 N- P2O5, which is equivalent to 240 kg ha'1 Urea + 150 kg ha'1 DAP is recommended. However, due to the buildup of P nutrient in the soil through continued use, farmers with history of good P fertilizer use should be advised to use 100 instead of 150 kg ha"1 DAP. For some lead farmers and for progressive use by poor fanners, an intermediate recommendation of 115-46 kg ha'1 N-P2O5, which is equivalent to 210 kg ha'1 Urea + 100 kg ha'1 DAP is given.

Under Ethiopian conditions, however, inflation affects both input and output prices, because of which the total revenues could proportionally increase with increasing benefits. What is more important is Ethiopian government usually takes some measures to control price rises in fertilizers than in wheat. This fact clearly indicates that the validity of the recommendations made can have little effect over the course of time on the recommendation made across the recommendation domains. Additional benefits that could be obtained from the production process were straw' yields, soil quality improvements, and grain quality improvements. The addit onal economic advantages that could be obtained from such additional benefits wrere not considered in the economic analysis due to the difficulties posed in the estimation of the market values of straw, residues, and the lack of grain quality standards to set premium prices. The future, therefore, holds promise to farmers to generate additional income from grains produce with higher improved grain qualities resulting from applications of the recommended practices. Straw yields and soil quality improvements are additional factors that can increase the benefits-cost ratios of the recommended practices.

References

43 Asefa Taa, Kefyalew Ginna and Shambel Maru. 1997. On-farm N and P fertilizer trial in Bread wheat on vertisols in South eastern Ethiopia. Agronomy and Crop Physiology progress report. Kulumsa Agricultural Research Center. EIAR. Ethiopia. Asefa Taa, Tanner DG, Kefyalew Ginna and Amanuel Gorfu. 1997. Grain yield of wheat as affected by cropping sequence and fertilizer application in southeastern Ethiopia. African Crop Science Journal 5: 147- 159. Asgelil Dibfcbe, Taye Bekele and Tekaligm Maino. 2001 Nutrient management in highland Vertisols of Ethiopia In: Paulos Dubale, Asgcleil Dibabc, Asfaw /eleke, Gezahegn Ayle and Abebe Kirub, eds. Advance; in Vertisols management in the Ethiopian highlands. Proceedings of the international symposium on Vertisols management, 28 Nov-1 Dec, 2000. Debre Zcit. Ethiopia, pp. 81-98. Asmare Yallow, Tanner DG, Regassa Enserinu and Alemu Haile. 1995. On-Farm Evaluation of Alternative Bread Wheat Technologies in NW Ethiopia. African Crop Science Journal 3-4: 443-449. Bemnct Gashawbeza, Solomon Assefa, Amelia Yaekob, Alemayehu Zemede. Jemanesh Kifetew and Bekele Mekuria. 2006. Effect of Nitrogen Fertilizer Levels and Varieties on Gluten Content and Some Rheological Characteristics of Durum Wheat Flour. Proceedings of the 12th Regional Wheat Workshop for Eastern, Central mid Southern Africa. Nakuru. Kenya, 22-26 November 2004. Bremner JM and Mulvaney CS. 1982. Total nitrogen. In: Page AL, Miller RH, and Kenney DR, eds. Method of Soil Amilysis, Part 2, Agronomy Monograph No 9, American Society of Agronomy, Madison, WI, 595- 624. Cassman KG, Doberman A and Walters DT. 2002. Agroecosystems, nitrogen use efficiency, and nitrogen management. Ambio 31:132-140. CIMMYT. 1988. From Agronomic Data to Farmer Recommendations: An Economics Training Manual. Completely revised edition. Mexico, D.F. CSA. 2005 Report on Area and Production of Major Crops (Private Peasant Holdings, Meher Season). Statistical Bulletin. Volume I. May. 2005,Addis Ababa. CSA. 2011 Report on Area and Production of Major Crops (Private Peasant Holdings, Meher Season). Statistical Bulletin. Volume I. April 2011, Addis Ababa. Doyle AD aad Holford ICR. 1993. The uptake of nitrogen by wheat, its agronomic efficiency and their relationship to soil and fertilizer nitrogen. Aust. J. Agric. Res. 44: 1245-1258. Ethio-Italiari Developemtn Cooperation. 2002. Atlas of Arsi zone. Arsi-Bale Rural Development Project. Arsi zone Planning and Economic Developcmnt Office Regional State of Oromia. GIS sub unit. Fageria NK and Baligar VC. 2003. Methodology for evaluation of lowland rice genotypes for nitrogen use efficiency. J. Plant Nutr. 26:1315 -1333. Kate Schne dcr and Leigh Anderson. 2010. Yield Gap and Productivity Potential in Ethiopian Agriculture: Staple Grains and Pulses. Evans School Policy Analysis and Research (EPAR). EPAR Brief No.98. University of Washington. Minale Liben, Alemayahu Assefa, Tanner DG and Tilahun Tadesse. 1999. The response of bread wheat to N and P application under improved drainage on Bichena Vertisols in north-western Ethiopia. In: The Tenth Regional Wheat Workshop for Eastern. Central and Southern Africa. Addis Ababa, Ethiopia: CIMMYT. pp. 298-308. Minale Liben. Alemayehu Assefa, Tilahun Tadesse and Abreham Mariye. 2004. Response of Bread Wheat to Nitrogen and Phosphorous Fertilizers at Different Agroecologies of Northwestern Ethiopia. In: Proceedings of the 12th Regional Wheat Workshop for Eastern, Central and Southern Africa. Nakuru, Kenya. Novoa R and Loomis RS. 1981. Nitrogen and plant production. Plant Soil 58:177-204. Origin Lab Corporation. 1991-2007. Origin 8 Graphical User Interface (GILJ). Northampton. MA 01060 USA Payne TS, Tanner DG and Abdalla OS. 1996. Current issues in wheat research and production in Eastern, Central and Southern Africa: changes and challenges. In: Tanner DG, Payne TS and Abdalla OS, eds. The Ninth Regional Wheat Workshop for Eastern. Central and Southern Africa. Addis Ababa. Ethiopia: CIMMYT^ pp. 1-27 SAS Institute. 2002. The Statistical Analysis Software System for Windows. Version 9.00. TS Level 00M0. SAS Insiitute Inc., Cary NC. USA. Shambel Maru, Kefyalew Girma, Workiye Tilahun. Amanuel Gorfu and Mekonnen Kasaye. 1999. On-farm N and P fertilizer trial in Bread wheat on vertisols in South eastern Ethiopia. Agronomy and Crop Physiology progress report. Kulumsa Agricultural Research Center. EIAR. Ethiopia SPSS Statistical Software. 1989-2011. IBM Corporation. Version 20. Syers JK, Craswell ET and Nyamudeza P. 2001. Research Needs and Opportunities for Farming Vertisols Sustainably. In: Syers JK. Penning FW and Nyamudeza P, eds. 2001. The Sustainable Management of Vertisolsl CABI Publishing. New York. USA. Tanner, D.G., Amanuel Gorfu and Asefa Taa. 1993. Fertilizer effects on sustainability in the wheat-based small-holder farming systems of southeastern Ethiopia. Field Crops Res. 33:235-48.

44 Tanner DG and Payne TS. 2001. CIMMYT wheat research and capacity building in eastern, central, and southern Africa. In: Research Highlights of the CIMMYT Wheat Program 1990-2000 (Online). http://www.cimmyt.org Taye Bekele, Yesuf Assen, Sahlemedhin Sertsu, Amanuel Gorfu and Mohammed Hassena, Tanner DG and Tesfaye Tessemma. 2002. Optimizing Fertilizer Use in Ethiopia. Correlation of Soil Analysis with Fertilizer Response in Hetosa Were da, Arsi Zone. EARO/SG-2000 Collaborative Research. Tekalign Mamo, Haque I and Kamara GS. 1988. Phosphorus Status of Some Ethiopian Highland Vertisols. In Proceeding of Management of Vertisols in the Sub-Saharan Africa. International Livestock Center for Africa. Addis Ababa. Ethiopia. Teklu Erkossa, Tekalign Mamo, Selamyihun Kidane and Mesfin Abebe. 2000. Response of some durum wheat landraces to nitrogen application on Ethiopian Vertisols. In: The Eleventh Regional Wheat Workshop for Eastern, Central and Southern Africa. Addis Ababa, Ethiopia Teklu Tesfaye. 1996. On-farm verification of an improved durum wheat production package in Ada and Akaki areas of Ethiopia. In: The Ninth Regional Wheat Workshop for Eastern, Central and Southern Africa. Addis Ababa, Ethiopia: CIMMYT. Tilahun Geleto, Tanner DG, Tekalign Mamo and Getinet Gebeyehu. 1996. Response of rainfed bread and durum wheat to source, level and timing of nitrogen fertilizer on two Ethiopian Vertisols: II. N uptake, recovery and efficiency. Kluwer Academic Publishers. Netherlands.

45 T heme 3 Surveillance, Monitoring and Developing Management Options for Monitoring of Major Foliar Diseases of Wheat in Ethiopia, Kenya, Tanzania and Uganda

Worku Denbel1, Wanyera R2, Rose M3 and Wasukira A4 Kukumsa Agricultural Research Center, Wheat Regional Center of Excellence, Kulumsa, Ethiopia 2Kenya Agricultural and Livestock Research Institute (KALRO), Njoro, P.O. Private Bag Njoro 20107 3Uyole Agricultural research Instituted Sealian Agricultural Research Institute2, Tanzania ginyanya Zonal Agricultural Research and Development Institute, P. 0. Box 1356, Mbale, Uganda

Abstract The rust diseases of wheal are major wheat production bottleneck in many countries of east Africa due to the favourable environmental conditions and susceptible wheat cultivars that provides optimum condition for the pathogen to evolve continuously. Wheat cultivars are withdrawn from production after a few years of their deployment as a result of the breakdown of the deployed rust resistance genes. A three year wheat diseases survey was conducted in the major wheat producing regions of the four countries and the result indicated that the rust diseases of wheat are the major wheat diseases in the region. However, the incidence, severity and distribution of the three rusts vary from country to country, from variety to variety and from season to season. In countries such as Ethiopia, where bi- modal rainfall distribution prevails the rusts are more important in the main season than in the off season (belg) an

Introduction

East African countries arc known ‘hot-spots’ for the evolution of new rust races and this has been evidenced in the recent evolution and discovery of race Ug99 (TTKSK) in Uganda in 1999 (Pretorius et al. 2000). Similarly, the >>9-virulent Puccinia striiformis race first evolved in Eastern Africa and then migrated to different parts of the world within 10 years and caused severe epidemics in its migration path (Singh et al. 2004). The stem rust race Ug99 (TTKSK) is widespread in the eastern part of Africa on commercial wheat fields and endangers wheat production in the region (Singh et al. 2006). In Kenya, a yield loss ranging from 5.6-66.3% has been recorded due to this race (Macharia and Wanyera 2012). Currently most of the wheat cultivars grown in East African countries are completely susceptible to Ug99 (Worku et al. 2013; Njau et al. 2009).

In Ethiopia, a yield loss of 71% has been recorded due to yellow rust on susceptible bread wheat variety Wabe in Bale (Derejc Hailu and Chemeda Fininsa 2009). In 2010 cropping season, major epidemics of yellow rust resulted in the susceptibility of the two mega bread wheat cultivars viz., Kubsa and Galama due to the breakdown of the Yr27 gene deployed in both cultivars.

Monitoring diseases and races is of great importance for sustainable wheat production in the region. There is no detailed information on all the major diseases of wheat in East Africa region. Disease survey is basic to design problem solving research projects and implementing effective diseases 46 control strategies. In view of this, carrying out wheat diseases surveys annually is crucial to assist the wheat breeding program in the region. Therefore, the major objective this paper is to describe the regional importance and geographic distribution of major wheal diseases in cast Africa and recommend possible strategics lo alleviate the problem.

Materials and Methods

Wheat diseases surveys were conducted in ihe wheat producing areas of the four EAAPP countries at least once a season. Under Ethiopian condition, the survey was repeated in the short (belg) rainy season in Arsi and Bale Zones as wheat crop is grown twice a year. The assessments were done at even. 3 - I Okm interval following the pre-dctcrmined routes. GPS machine was used lo know the grid references of ihe locations. Incidcncc and severity of wheal rusts were scored for each data point and the prevalence was calculated in percentages. The severity of wheat msts was recorded based on the modi led Cobb's scale (Peterson et al. 1948). The septoria diseases of wheat were recorded using the doub c-digit (00-99) scale (Eyal et al. 1987). The growth stage of the crop was recorded based on the Zadoks Scale (Zadoks el al. 1974). The collcctcd data were compiled and analyzed al national and regional levels.

Results and Discussion

Diseases situations, 2011

Ethiopi a: 781 wheat fields w'cre covered in the survey in the four major wheat growing regional stales including Tigray, Amhara, Oromia and Southern Nations, Nationalities and Peoples (SNNPs). Yellow (stripe) rust of wheat was recorded in 506 (64.8%) of the wheat fields; stem rust of wheat w;as recorded in 173 (22.2%) of the wheat fields; leaf rust infection was recorded in 141 (18.1%) of the wheat fields and Septoria tritici blotch was encountered in 350 (44.8%) of the wheat fields covered in the survey. During the main (meher) season of 2011, yellow rust, septoria leaf blotch, stem rust and leaf rust were important diseases both in terms of distribution and disease intensity (incidence and severity) in order of their importance in the wheat growing regions of Ethiopia (Table 1). In the off season (belg) of 2011, the survey was conducted in Bale zone of Oromia. The major diseases in the off-season were Septoria tritici blotch, stem rust and leaf rust. Yellow rust was recorded in only one field at Gassera district. Other diseases of wheat such as take-all, eycspot, root-rot. head-scab, tan-spot and ioose-smut were recorded in some parts of wheat producing areas of the country with various distribution and intensity.

The variety Kubsa (HAR1685) was dominantly grown throughout the surveyed regions in 2011. It occupied 32.2% of the cultivated varieties during the survey year. Wheat varieties Digalu and Galama occupied 15.6% and 12.1% of the land in the surveyed fields in the country. More number of wheat varieties was recorded in Oromia region followed by Amhara Region. Tigray and SNNPs had the least number of varieties.

Table 1 Distribution of major wheat diseases in 2011 in Ethiopia

Infected fields Surveyed Region Season Septoria leaf fields Yellow rust Stem rust Leaf rust blotch Tigray Main (meher) 52 33 4 6 29 SNNPs Main (meher) 14 7 2 6 3 Anrnara Main (meher) 250 193 9 28 48 Orcmiya Main (meher) and off 465 273 158 101 270 season (Belg) Total 781 506 173 141 350

47 Kenya: Surveys were conducted in farmers’ fields in years 2011, 2012 and 2013 to determine the prevalence and distribution of wheat diseases in the key wheat growing (South Rift, Mt. Kenya region, North and South Rift) regions of Kenya. In 2011, 327 farms were sampled. Stem rust was detected in 209 (63.9%), yellow rust in thirty-seven (11.3% ) and leaf (brown) rust) forty-nine (14.9% ) of the farms with disease severity ranging from trace (TR) to 100S, TR to 60S and TR to 40S for stem rust, yellow rust and leaf rust respectively. Stem rust infection ranged from irace to 100S with minimum infection in North Rift (48.3%), and a maximum infection in Central Rift (70.9%), South Rift (68.8%), and Mt. Kenya (68.0%). Yellow rust infection ranged from trace to 60S with minimum infection in South Rift (6.4%), Mt. Kenya (10.4%), North Rift (10.9%), and maximum infection in Central Rift (17.7%). Leaf (brown) rust infection ranged from trace to 40S with minimum infection in North Rift (10.9%), Mt. Kenya (13.4%) and maximum infection in Central and South Rift (17.7%). Two hundred and seventy-five (84.1%) of the farms were sprayed to reduce/ suppress stem rust and fifty-one (15.6%) of the farms were not sprayed. Two hundred and sixty-one (94.9%) and sixty-five (23.6%) of the sprayed farms had low (TR-15S) high (20S and above) infection respectively. Other diseases of importance were also recorded. The incidence of Septoria diseases was highest in North Rift (27.4%) followed by Central Rift (16.1%). Fusarium head Scab was highest in North Rift (12.3%).

Tanzania: Leaf and stem rusts occurred in all of the four districts of the northern zone and were recorded in 61% of the fields visited. Stem and leaf rusts were highly prevalent in Monduli and Hanang districts (Table 2). Likewise stripe rust was more prevalent in Hanang and Hai districts compared to the other two districts. Karatu district has a high prevalence of powdery mildew and take all. Disease incidence varied among the districts and among the farms. Even fields in the same district had different incidences of wheat foliar diseases. Twelve percent of the fields had disease incidence of more than 50%. Hanag and Monduli districts showed higher incidences of leaf and stripe rust compared to Hai (Table 2). All the diseases recorded varied among the four districts surveyed (Table 2). Leaf rust severity scores were highest in Moduli district with a range of 10-80 %. Hai district had lower scores for all of the six wheat foliar diseases recorded with the exception of septoria leaf blotch and stripe rust.

48 Table 2 Prevalence, incidence and severity of wheat diseases (%) in 2010/2011 wheat cropping season in Tanzania

District Altitude (m) Parameter Septoria Stem rust Leaf rust Stripe rust Powdery mildew Take-all Hanang 1500- 1850 Prevalence 0 91 75 58 8 25 Incidence 0 20-80 10-60 25-85 0 -3 0 0-20 Severity 0 5 -4 0 5 -3 0 5 -50 0 -1 0 30-50 Hai 1300 -1700 Prevalence 60 60 0 40 0 20 Incidence 20-70 0-15 0 0- 15 0 0-10 Severity 30-50 1 -1 0 0 0-50 0 0-10 Monduli 1400-1950 Prevalence 14 57 43 29 14 0 Incidence 10-45 20 -30 40-70 25-75 0-15 0 Severity 0-40 5-10 10-80 5-40 0-3 0 Karatu 1500-1750 Prevalence 28 14 57 14 57 57 Incidence 25-55 15-20 45-75 10-15 5-90 10-35 Severity 0-50 0-20 5-70 0-50 0-9 0-40

49 Diseases situation in Ethiopia, Kenya and Tanzania in 2012

Ethiopia: In the 2012 cropping season, a total ol 1171 wheat fields were covered in the survey in Ethiopia. Septoria tritici blotch, yellow rust, leaf rust and stem rust were recorded in 604 (56.1%) 368 (31.4%), 342 (29.2%) and 295 (25.2%) fields, respectively. The incidence, severity and range of the recorded diseases is presented (Tables 3 and 4).

Kenya: 278 farms were sampled in 2012. Stem rust was detected in 109 (39.2%), yellow rust in eleven (3.9%) and leaf (brown) rust in thirteen (4.6%) of the farms with disease severity ranging from trace to 80S, 5S to 70S and 5S to SOS respectively. Septoria diseases were highest in South, Central and North Rifts (46.5%, 42.8% and 41.8%), respectively, while Barley yellow dwarf virus (BYDV) was highest (16.4%) in Central Rift. Fusarium species were highest in Central Rift (8.9%).

Diseases situation in Ethiopia, Kenya and Tanzania in 2013 Ethiopia Septoria tritici blotch, stem rust, yellow rust and leaf rust occurred in 589 (61%), 291 (30.1%), 229 (23.7%), and 178 (18.4%) wheat fields covered in the survey in 2013. The incidence, severity and the range and mean of the corresponding incidence and severity values is given in Tables 5 and 6.

Kenya: 333 farms were sampled in 2013. Stem rust was detected in 165 (49.6%), yellow rust in twenty-five (7.5%) and leaf rust in sixteen (4.7%) with disease severity ranging from trace to 100S, 50S and 50S respectively. Septoria diseases infections were highest in Mt. Kenya region (45.3%) and North Rift (28.2%). Fusarium species were highest in Central Rift (6.2%).

Table 3 Summary of rust diseases of wheat on regional basis in 2012 cropping season in Ethiopia

Region Total Yellow rust Leaf rust Stem rust field FI Inci %) Sev %) FI Inci %) Sev (%) FI Inci (%) Sev %) R M R M R MRM R M R M Tigray 147 90 0-100 39.9 0-100 18.6 68 0-100 36.4 0-100 13 8 0-100 2.5 0-60 1.3 Amhara 407 147 0-100 10.9 0-75 5.1 55 0-100 4.8 0-40 0.1 19 0-100 1.8 0.40 0.8 SNNPR 69 20 0-60 6.3 0-40 6 35 0-100 11.5 0-60 5.6 31 0-50 2.7 0-30 2 Oromiya 548 111 0-100 0.8 0-60 0.3 184 0-100 1.6 0-70 0.8 237 0-100 2.1 0-60 1.5 Total 1171 368 342 295 Fl-Field infected. R-Range, M-Mean, Inci-lncidence, Sev-Severity

Table 4 Regonal summary of septoria leaf blotch in 2012, Ethiopia

Region Total Septoria leaf blotch fields Incidence Severity Index FI (%) (00-99) RMMRM Tigray 147 88 0-100 49.6 42 0-0.52 0.2 Amhara 407 306 0-100 21.9 42 0-0.6 0.14 Oromiya 523 210 0-100 5.2 11 0.1

50 Table 5 Distribution and intensities of wheat rusts in Tigray. Amhara, and Oromia regions during 2013/14 main cropping season

Region Total Yellow rust Leaf rust Stem rust field FI Incidence (%) Severity (%) FI Incidence (%) Severity (%) FI Incidence (%) Severity (%) R M RM RM RM RM R M Tigray 137 72 0-100 40.94 0-100 20.1 32 0-100 22.6 0-100 5.9 16 0-100 7.8 0-100 5.1 Amhara 353 77 0-100 3.3 0-60 1.3 33 0-100 1.6 0-30 0.8 10 0-100 0.4 0-60 0.11 Oromia 476 80 0-100 4.48 0-75 2.02 113 0-100 8.85 0-50 3.92 265 0-100 14.62 0-100 7.43

Total 966 229 -- -- 178 ---- 291 -- - 4.2.6 FI-Field infected, R= Range, M Mean

Table 6 Distribution and intensities of Septoria leaf blotch in Tigray, Amhara and Oromia regions during 2013/14

Region Total Septoria Leaf Blotch Fields inspected Field infected Incidence (%) Severity index Range Mean Range Mean Tigray 137 52 0-100 27.6 0-1 0.13 Amhara 353 254 0-100 31.4 0-0.69 0.07 Oromia 476 283 0-100 57.73 0-0.89 Total 966 589 --- -

51 Tanzania: 43 farmer fields and one research station were surveyed in the southern and northern highlands of Tanzania during 2013 season. All three wheat rusts were observed during the survey. Stem rust was observed in 23 (53%), stripe rust in 16 (37%), and leaf rust in 15 (35%) of the farmer fields. The three wheat rusts were observed at the Uyole research station and stem rust predominated at this site.

Uganda: 35 fields were surveyed in 2013. Stem rust was the major disease; recorded in 23 out of the 35 fields surveyed (66%) and was widespread in both the east and the south-west. 52% of the fields in southwest were infected with leaf rust with low observation in the east. Yellow rust was recorded in 70% of the fields surveyed in the southwest.

Discussion

Wheat rusts were common in the wheat fields and stem rust was widespread in all die surveyed regions in Kenya. In all the three years where the survey was conducted, stem rust of wheat was the number one wheat production bottleneck in major wheat producing regions of Kenya in terms of both frequency of occurrence and its distribution in the field. The second important disease in both 2011 and 2012 season was leaf rust of wheat and this disease took the third position in 2013 season. Unlike in 2011 and 2012 seasons, yellow rust of wheat was the second important diseases of wheat in 2013 season in Kenya. The three-year disease data clearly shows that there is always favourable environmental condition for the development of stem rust epidemics in Kenya. This in turn may be associated with the cultivation of susceptible bread wheat cultivars in the country. Commonly grown varieties include KS Mwamba, NJRBW2, Kwale, Robin, and Eagle-10. In Kenya, evolution of new stem rust races with various virulence combinations is a recent phenomenon and the deployment of variety KS Mwaba that carries Sr24 gene has increased the frequency of race TTKST that is virulent to Sr24 gene (Ismail et al. 2012). For instance, variants of stem rust carrying virulence for Sr24 and Sr36 genes were detected in wheat fields in Kenya (Jin et al. 2008; Jin et al. 2009). Data on Septoria diseases revealed that the diseases are on die increase in Kenya. The susceptibility of the varieties, favourable climatic conditions, and additional costs of f jngicides qualify the diseases as damaging with strong impact on wheat production.

In Ethiopia, yellow rust of wheat was the number one foliar disease in the wheat growing major agro-ecologies of the country in 2011 season. This may be attributed to the presence of sufficient amount of yellow rust inocula that were produced from the 2010 major yellow rust epidemics in the country (Worku 2014) and the significant wheat area covered by the super susceptible varieties such as Kubsa and Galama that took 32.2% and 12.1% of the wheat area in the same year. Septoria tritici blotch was the second most important diseases under Ethiopian condition. In Ethiopia, cereal mono-cropping is the common practice especially in the major wheat belts of Arsi and Bale where wheat is sown after wheat or wheat is sown after barley and this situation provides optimum condition for the build-up of the stubble- born pathogen, Septoria tritici blotch. Stem rust of wheat was in the third place both in terms of frequency of occurrence and distribution in wheat growing districts. This reduction in stem rust distribution may be associated with in the slight increase in the area covered with Ug99 resistant bread wheat cultivars that were released and distributed to farmers especially in Oromia Region. The two Ug99 resistant bread wheat cultivars including Danda’a and Kakaba each covered 1.2% of the wheat area in the 2011 season and this definitely decreased the amount of stem rust inoculum for the 2012 season. Leaf rust of wheat was the fourth important disease of wheat and this disease is particularly important in the major durum wheat producing regions of Ethiopia. In 2012 season, septoria was the number one disease

52 and it was distributed in all wheat producing regions of Ethiopia at epidemic proportion. This is also attributed to the prevailing favorable environmental condition to the pathogen and the continuous cultivation of susceptible varieties of wheat in the country. Yellow rust of wheat was the second important disease in the season. In both 2010 and 2011 seasons in Ethiopia, yellow rust was the major wheat disease and its distribution started to decline because of the increase in the area allocated to the resistant bread wheat cultivars such as Danda’a, Kakaba, and Digalu. Though Digalu was susceptible to stem rust epidemics in Bale zone in 2013 season it was the most important variety in the immediate utilization and deployment after the 20 0 nationwide yellow rust epidemics. In 2013 season, as that of 2012 season, septoria was the number one disease in Ethiopia, which was recorded, in the major wheat producing regions. Stem rust, yellow rust and leaf rusts were also prevalent in their order of importance and distribution.

In Tanzania, in general the weather for the 2010/2011 cropping season in which the first survey was done, did not favour disease development in some areas due to severe droughts that caused 100% crop loss. Diseases of economic important in Tanzania include the three nists, Septoria tritici blotch, and powdery mildew especially in the northern part of the country.

As far as fungicide application is concerned, majority of the farmers sprayed the fungicides to reduce suppress disease infections, particularly the rusts, but some sprayed farms were noted to have high disease infections under Kenyan condition. These are farms that either had been sprayed late or the timing/fungicide concentrations were not right. This illustrates that majority of the fanners do not know how to spray fungicides timely.

Conclusion and Recommendation

The three wheat rusts are major wheat production bottleneck in the four countries. Septoria leaf blotch is also becoming more important nowadays and all the national wheat breeding programmes must incorporate resistance breeding for Septoria tritici blotch. In addition to resistance breeding, cultural practices such as crop rotation technologies must be promoted to fanners in the region for the management of septoria. Besides, continued monitoring of disease virulence throughout the region is crucial to detect shifts in the padiogen population as early as possible and therefore to effect an appropriate breeding strategy.

References

Dereje Hailu and Chemeda Fininsa. 2009. Relationship between stripe rust (Puccinia striiformis) and common wheat (Triticum aestivum) yield loss in the highlands of Bale, southeastern Ethiopia. Archives of PI ytopathology and Plant Protection 42(6): 508-523. Eyal Z, Scharen AL. Prescott JM and Van Ginkel M. 1987. The Septoria disease of Wheat: concepts and Methods of disease management. CIMMYT, Mexico, D.F. Ismail SG, Kinyua AM, Kibe AM and Wagara IN. 2012. Wheat stem rust severity and physiological raccs in North Rift region of Kenya. Asian Journal of Plant Pathology 6(2):25-32. Jin Y, Szabo LJ, Pretorius ZA, Singh RP, Ward R and Fetch TJ. 2008. Detection of virulence to resistance gene S/24 within raceTTKS of Puccinia graminis f.sp. tritici. Plant Diseases 92:923-926. Jin Y. Szabo T.J, Rouse MN, Fetch TJr, Pretorius ZA, Wanyera R, Njau P, 2009. Detection of virulence to resistance gene Sr36 within the TTKS race lineage of Puccinia graminis f.sp. tritici. Plant Diseases 93:367-370. Macharia JK and Wanyera R. 2012. Effect of stem rust race Ug99 on grain yield and yield components of wheat cultivars in Kenya. Journal of Agricultural Science and Technology 2:423-431.

53 Njau PN, Wanyera R, Macharia GK, Macharia J, Singh RP and Keller B. 2009. Resistance in Kenyan bread wheat to recent eastern African isolate of stem rust, Puccinia graminis f.sp. tritici, Ug99. Journal of Plant Breeding and Crop Science 2:022-027. Peterson RF, Campbell AB and Hannah AE. 1948. A diagrammatic scale for estimating rust intensity of leaves and stems of cereals. Canadian Journal o f Research 60:496-500. Pretorius ZA, Singh RP, Wagoire WW and Payne TS. 2000. Detection of virulence to wheat stem rust resistance gene Sr31 in Puccinia graminis f. sp. tritici in Uganda. Plant Diseases 84:203. Singh RP, Hodson DP, Jin Y, Julio J, Kinyua MG, Wanyera R, Njau P and Ward RW. 2006. Current status, likely migration and strategies to mitigate the threat to wheat production from race Ug99 (TTKS) of stem rust pathogen. CAB Reviews: Perspectives in Agriculture, Veterinary Science, Nutrition and Natural Resources. 1, No. 054. Singh RP William MM, Huerta-Espino J and Rosewarne G 2004. Wheat Rust in Asia: Meeting the Challenges with Old and New Technologies. New directions for a diverse planet: Proceedings of the 4,h International Crop Science Congress, Brisbane, Australia. 26Sep-lOct.2004. (Online): www.cropscience.org.au. Worku Denbel. 2014. Epidemics Puccinia striiformis f. sp. tritici in Arsi and west Arsi zones of Ethiopia in 2010 and identification of effective resistance genes. Journal of NaUiral Sciences Research 4(7): 33-39. Worku Denbel, Ayele Badebo and Tameru Alemu. 2013. 1,valuation of Ethiopian commercial wheat cultivars for resistance to stem rust race UG99. International Journal of Agronomy and Plant Production. Vol., 4 (1): 15-24. Zadoks JC’, Chang TT and Konzak CF. 1974. A decimal code for the growth stages of cereals. Weed Research 14:415-421.

54 Evaluation of Herbicides in Wheat

Hussien Sareta1 and Wogayehu Worku1 Ethiopian Institute of Agricultural Research, P.O.Box 2003, Addis Ababa, Ethiopia

Abstract Field evaluation trial of post-emergence herbicides for the management of annual grasses and broadleaf weeds in wheat was conducted at Kulumsa Agricultural Research Center main station. Bekoji and Lole farmers’ fields during 2011/12 and 2012/13 main cropping seasons and to be incorporated in an integrated weed management program. "Mesosulfron methyl + Idosulfuron methyl sodium and Pyroxsulam herbicides at a rale of 1 lit ha'1 and 0.45 lit ha1, respectively, two hand weeding and untreated weedy check treatments were used in the study. Most annual grass weeds like Snowdenia polvstachya, Avena fatua, Bromus pectinatus, Phalaris paradoxa, Setaria pumila and most broad leaf weeds like Polygonum nepalense, Gizotia scabra, Ga/insoga pcirvijlora, and Gallium spurium were 85- 100% controlled by Mesosulfron methyl + Idosulfuron methyl sodium and Pyroxsulam herbicides. Yield wise, both Mesosulfron methyl + Idosulfuron methyl sodium (liquid) 1 lit ha'1 a.i, Pyroxsulam (liquid) 0.45 lit ha"1 a.i and two hand weedings (30-35 and 55-60 DAE) outperformed in yield than weedy check by 37.7%, 29.3% and 21%, respectively. Mesosulfron methyl + Idosulfuron methyl sodium herbicide has a yield advantage over Pyroxsulam, two hand weeding and untreated weedy check by 12%, 21.3% and 37.7%, respectively. Maximum grain yield (5184 kg h a 1), crop biomass (12808 kg ha '), TKW (48.6 g) and Hectoliter weight (74.2) was obtained at Mesosulfron methyl + Idosulfuron methyl sodium treated plots. Pyroxsulam gave next highest better traits expression after Mesosulfron methyl + Idosulfuron methyl sodium. Generally, Mesosulfron methyl + Idosulfuron methyl sodium herbicide application at a rate of 1 lit ha'1 is better for the effective control of annual grasses and broad leaf weeds in wheat and can be used as one of the components in an integrated weed management program in wheat.

Introduction

Up to 25% could be lost clue to weeds (Akobundu 1987). In Ethiopia, a yield loss of above 36.3° o was recorded in wheat due to uncontrolled weed growth (Rezene 2005). Similarly, in a competition study of Avena abyss mica, Lo/ium temulentum L., Snowdenia polvstachya and Phalaris paradoxa L. with bread wheat, a yield loss of 48-86% were recorded by the maximum weed density of 320 weed seedlings rrf (Taye et al. 1996). In Durum wheat, Convolvulus arvensis and Cyperus spp poses significant yield loss. Besides, considerable yield loss has been recorded in irrigated wheat due to Sorghum antndnaceae, Cyperus esculentus, Cyperus rotundus, Portulaca oleraceae, Corchorus olitorius, and Sorghum arundinaceae around 60% at Werer research center (Kassahun 1987).

Bromus pectinatus and Snowdenia polvstachya are such weed species that recently became prominent in the affected cropping systems due to a weed population shift attributed primarily to continuous cereal cropping and frequent use of selective herbicides against previously common grass weeds such as Avena fatua (Tanner and Giref 1991; Amanuel et al. 1992: Rezene and Yohannes 2003). Therefore, the study was designed with the objective to evaluate the different herbicides for the control of annual grasses and broadleaf weeds in whea to be incorporated in an integrated weed management program to increase the productivity of wheat.

Materials and Methods

The trials were conducted at Kulumsa Agricultural Research center main station, Bekoji, Lole (Ego) during the main cropping season of 2011/12 and 2012/13. Kulumsa is situated in the

55 main wheat belt of Ethiopia at an altitude of 2200 m in the north periphery of Asella town, which is about 168 km southeast of Addis Ababa. It is found at 8°01,10,,N and 39<,09'H "E and receives an average of 832 mm. The mean minimum and maximum temperature is 10°C and 23°C. Bekoji is found at 7°32'37"N and 39°15'21" E with an altitude of 2780 m and receives an average of 1066 mm rainfall and the mean minimum and maximum temperature is 9.6°c and 24°c. Dominant soils in these areas are Luvisol and Nitosol. Bekoji and Lole (Ego) are 61 km and 39 km away from Asella, respectively. The treatments were Pyroxsulam, Mesosulfron methyl + Idosulfuron methyl sodium both applied as post- emergence at a rate of 0.45 and 1 lit h a 1, respectively, two hand weeding and weedy check left as control. Post-emergence herbicides were applied at 30-35 days after emergence (DAE) and hand weeding at 30-35 and 55-60 (DAE). All the necessary agronomic practices were applied equally for all treatments. Danda'a wheat variety was used for the trial at different locations at a seeding rate of 150 kg ha'1, row planted and 100 kg ha'1 DAP and 50 kg ha'1 Urea fertilizers were applied at sowing. The design was RCBD with three replications in a plot size of 4m by 5m. The necessary agronomic data such as plant height, number of tillers, spike length, weed count before, two and four weeks after herbicide application on 1.0 nr quadrates, general weed control score (1-5 scale), fresh and dry weed biomass, crop biomass, grain yield, Thousand Kernel Weight (TKW), hectoliter weight (HLW) were collected. Data were subjected to statistical analysis software using Proc GLM procedure in SAS (1994). Comparisons among treatments with significant differences were computed based on LSD test. Linear correlation was used to determine the association between grain yield and yield components. .

Results and Discussion

Efficacy of herbicides Efficacy esult over locations indicated that all the treatments except untreated weedy check were effective for the control of the target annual grass weeds like Snowdenia polystachya, Avena fatua, Bromus pectinatus, Phalaris paraduxa, Seiaria pumila and broad leaf weeds like Gallium spurium, Gizolia scabra, Galinsoga parvijlora and Polygonum nepalense. Effectiveness of the treatments in controlling S. polystachya by Mesosulfron methyl + Idosulfuron methyl sodium (liquid) 1 lit ha'! a.i, Pyroxsulam 0.45 lit ha'1 a.i, and two hand weeding (30-35 and 55-60 DAE) was 100%, 85% and 100%, respectively (Table 1). A.fatua and P. paradoxa weeds were controlled at efficacy rate of 100% by Mesosulfron methyl + Idosulfuron methyl sodium 1 1 lit ha"1 a.i, Pyroxsulam 0.45 lit ha'! a.i, and two hand weedings (30-35 and 55-60 DAE). Whereas, Bromus pectinatus was controlled at 85%, 100% and 100% efficacy level, respectively. Rezene et al. (2007) reported that Propoxycarbozone- sodium (Attribut 70WG) was effective against Bromus pectinatus and gave satisfactory suppression of Snowdenia polystachya constantantly across all experimental locations. On the other hand, Shambel et al. (2000) reported that the herbicidal chemical sulfosulforol and ethiozin exhibited significant potential to control problematic grass weeds including Brome grass in ihe wheat growing areas of Ethiopia. Similarly, both herbicides and the fanner’s practice controlled Gallium spurium, Gizotia scabra, Galinsoga parvijlora and Polygonum nepalense at 89-100% efficacy level (Table 1). The negative values in the efficacy of the applied herbicides were resulted from the increasingly late emergence of the weeds after herbicide applications.

Hence, Mesosulfron methyl + Idosulfuron methyl sodium is best recommended in areas where Snowdenia polystachya, Avena fatua, Pha laris paradoxa, Setaria pumila and broad leaf weeds like Polygonum nepalense, Galinsoga parvijlora, Gizotia scabra, Gallium

56 spurium are dominant weed problems. For areas where Bromus pectinatus, Avena fatua, Phalaris paradoxa, Setaria pumila, Loium temulentum and broad leaf weeds like Polygonum nepalense, Galinsoga parviflora, Gizotia scahra are dominant problems, it is better to use Pyroxsulam.

Yield and yiefid components Grain yield showed significant (P<0.001) difference due to Mesosulfron methyl + Idosulfuron methyl sodium, Pyroxsulam and 2 hand weedings. The highest grain yield (5184 kg ha'1) was recorded in Mesosulfron methyl + Idosulfuron methyl sodium application followed by Pyroxsulam (4567 kg ha'1) and two hand weedings (4079 kg ha'1). However, the lowest grain yield (3228 kg ha'1) was recorded in weedy check treatment (Table 2). The combined analysis over locations indicated that no significant effect of the treatments on plant height, spike length, TKW and HLW but there was significant differences (P<0.05) among treatments with respect to weed biomass, crop biomass and grain yield at compared to the unweeded treatment (Table 2). Both Mesosulfron methyl + Idosulfuron methyl sodium 1 lit ha 1 a.i, Pyroxsulam 0.45 lit ha'1 a.i and two hand weedings (30-35 and 55-60 DAE) outperformed in yield than weedy check by 37.7%, 29.3% and 21%, respectively. Similarly, Mesosulfron methyl + Idosulfuron methyl sodium 1 lit ha'1 a.i has a yield advantage over Pyroxsulam 0.45 lit ha’1 a.i, two hand weeding and the weedy check by 12%, 21% and 37.7 %, respectively (Table 3).

57 Table 1 Efficacy of Mesosulfron methyl + Idosulfuron methyl sodium as compared to Pyroxsulam major grass weeds two weeks after application at two locations in Arsi and West Arsi Zones

Locations Species Mesosulfron methyl + Pyroxsulam (liquid) 0.45 lit ha-1 a.i Two hand Untreated weedy Idosulfuron methyl sodium weeding check (liquid) 1 lit ha-1 a.i •* CD < AA Efficacy BA AA Efficacy BA AA Efficacy BA AA Efficacy (%) (%) (%) (%) Bekoji Snowdenia polystachya 80 0 100 40 6 85 120 0 100 160 160 0 Avena fatua 68 0 100 56 0 100 42 0 100 0 0 0 Bromus pectinatus 3400 510 85 2180 0 100 1740 0 100 4200 4300 -2.3 Phalaris paradoxa 25 0 100 260 0 100 100 0 100 300 340 -11.7 Gallium aparine 58 0 100 43 0 100 5 0 100 3 4 -25 Polygonum nepalense 117 0 100 55 0 100 50 4 92 45 47 -4.2 Gizotia scabra 17 0 100 23 0 100 15 0 100 18 18 0 Galinsoga parviflora 0 0 - 0 0 - 68 3 95 56 56 0 Lole Snowdenia polystachya 1260 0 100 860 130 85 1140 0 100 1420 1460 -2.7 Avena fatua 48 0 100 32 0 100 72 0 100 0 0 0 Bromus pectinatus 1720 256 85 1820 0 100 1220 0 100 2080 2140 -2.8 Phalaris paradoxa 17 0 100 21 0 100 30 0 100 0 0 0 Gallium aparine 94 5 99 102 0 100 19 2 89 5 6 -16.6 Polygonum nepalense 55 0 100 30 0 100 62 3 95 50 54 -7.4 Gizotia scabra 5 1 90 9 0 100 11 0 100 15 15 0 Galinsoga parviflora 16 1 99.9 28 0 100 23 0 100 68 68 0 * BA- Before Application; A A-After Application

58 Table 2 Mean grain yield after Pyroxsulam and Mesosulfron methyl + Idosulfuron methyl sodium at two locations in Arsi and West Arsi Zones

Treatment PH SL No. of TKW HLW CBM GY Weed (cm) (cm) tillers (g) (kg ha-1) (kg ha-1) biomass plant-1 (kg ha-1) Pyroxsulam 96 7.4 3.4 47.85 73.7 10750 4567b 709 Two hand weeding (30- 98.5 7.7 3.25 47.6 73.6 8792 4079c 684 35 and 55-60 DAE) Weedy check 101 7.0 2.85 46.8 72.9 7467 3228d 1492 Mean 98.4 7.5 3.35 47.7 73.6 9954 4265 LSD NSNS NSNS NS 7763 381 CV (%) 5.6 14.8 16.4 1.1 0.03 37.2 3.7 GY- Grain yield, SL-Spike length. NS/S-Number of tillers per plant, PH-Plant heighI TKW-Thousand kernel weight, HLW- Hector liter weight. CBM- Crop Biomass yield, EC-Efficacy, ns- statistically nonsignificant

able 3 Mean grain yield (kg ha-1) after Pyroxsulam and Mesosulfron methyl + Idosulfuron methyl sodium at two locations

Treatments Locations Mean Advantage of Yield advantage of Bekoji Lole KARC three treatments Atlantis 37.5 OD over (Station) over the check the others treatments (%) (%) Pyroxsulam 5867 4085 3750 4567 29.3 12 Mesosulfron methyl + 6633 4435 4485 5184 37.7 Idosu furon methyl sodium Two hand weeding (30- 4667 3285 4285 4079 21 21.3 35 and 55-60 DAE) Untreated weedy check 3400 2615 3670 3228 37.7 KAR' - Kulumsa Agricultural Research Center

The highest value of general weed control score (1-5 scale), individual weed score (1-5 scale); weed biomass (kg ha'1) was recorded in the untreated weedy check (Table 2 and 4). The lowest dry weed biomass (317 kg ha'1) was recorded in Mesosulfron methyl + Idosulfuron methyl sodium applied herbicide followed by two hand weeding (684 kg ha'1) and Pyroxsulam (709 kg ha’1) while the highest (1492 kg ha'1) was recorded in untreated weedy check. Dry weed biomass showed significant difference (P<0.05) due to Atlantis 37.5 OD. Pyroxsulam and two hand weeding (Table 2).

Table 4 Visual assessment score on general control of annual grasses and broad leaf weeds 15 days after application and at maturity at two locations

Treatments Locations Mean 15 days after application Bekoji Lole(Ego) Pyroxsulam 1 1.5 1.75 Mesosulfron methyl +ldosulfuron methyl sodium 1 1 1 Two • and weeding (30-35 and 55-60 DAE) 1 1 1 Untreated weedy check 5 4 4.5 At maturity Stage Pyroxsulam 1 4.2.7 2 4.2.8 1.5 Mesosulfron methyl -Hdosulfuron methyl sodium 4.2.9 1 4.2.10 1.5 4.2.11 1.75 Two hand weeding (30-35 and 55-60 DAE) 4.2.12 1.5 4.2.13 1 4.2.14 1.75 Untreated weedy Check 4.2.15 5 4.2.16 4.5 4.2.17 4.25 Key: 1-Complete eradication; 2-effective destruction; 3-proper reduction in growth and population; 4- reduced growth and population and 5-healthy

59 Conclusion and Recommendation

Mesosulfron methyl +Idosulfuron methyl sodium is best recommended in areas where Snowdenia polystachya, Avena fatua, Phalaris paradoxa, Setaria pumila and broad leaf weeds like Polygonum nepalense, Galinsoga parviflora, Gizotia scabra, Gallium spurium are dominant weed problems. For areas where Bromus pectinatus, Avena fatua, Phalaris paradoxa, Setaria pumila, Loium temulentum and broad leaf weeds like Polygonum nepalense, Galinsoga parviflora, Gizotia scabra are dominant weed problems are major weeds species, it is better to use Pyroxsulam. Therefore, it can be recommended that Mesosulfron methyl +Idosulfuron mediyl sodium at a rate of 1 lit ha’1 is better for the effective control of annual grasses and broad leaf weeds in wheat than Pyroxsulam and can be recommended as one of the components in an integrated weed management program in wheat.

References

Akobundu (O. 1987. Weed Science in the tropics, Principles and practices. John Wiley and Sons, Ltd. New York. Amanuel Corfu, Tanner DG and Assefa Taa. 1992. On-farm evaluation of pre-and post emergence grass herbicides on bread wheat in Arsi Region of Ethiopia. In: Tanner DG and Wilfred Mwangi, eds. Proceedings of the Seventh Regional Wheat workshop for I astern. Central and South Africa, Addis Ababa, Ethiopia: CIMMYT. pp. 330-337. Kassahun /ewdie and Tanner DG. 1998. Pre- and Post- Emergence Herbicides for irrigated wheat in Ethiopia. The Tenth Regional Wheat Workshop for Eastern. Central and Southern Africa. September 14-18, 1998, University of Stelenbosch, South Africa, pp. 309-315. Rezene Fessehaie and Yohannes L. 2003. Control of Snowdenia polystachya in large scale wheal production: Herbicide Resistance in context. Proceedings of the Agronomy Workshop, 20-21 March 2000, Melkassa, Ethiopia Bale Agricultural Development Enterprise (BAD1 ), Addis Ababa, pp. 79-88. Rezene Fessehaie. 2005. Weed Science Research and Extension in Ethiopia: Challenges and Responses. Key note address. Ethiopian Weed Science Society 7th Annual Conference. 24 -25 November 2005. EARO, Addis Ababa, Ethiopia. Rezene Fessehaie, Natenael Wassie and Kedija Demsiss. 2007. Effect of Propoxycarbozone-sodium and mesosulliiron-methyl for annual grass weed control in wheal. Ethiopian Journal o f weed management 1: 53- 61. SAS. 1994. SAS Institute Inc. 1994. Shambel M, Kefyalew G and Tanner DG 2000. Evaluation of herbicides for the control of brome grass in wheat in Southeastern Ethiopia. In: CIMMYT. The Eleventh Regional Wheat Workshop for Eastern, Central and South Africa, Addis Ababa, Ethiopia: CIMMYT Tanner DG and Giref Sahle. 1991 Weed control research conducted in Ethiopia. Pp.235-276. In: Hailu Gebremariam, Tanner DG and Mengistu Hulluka, eds. Wheat Research in Ethiopia: A Historic Perspective. Addis Ababa: IAR/CIMMYT. Taye T, Tanner DG and Mengistu H. 1996. Grass weeds competition with bread wheat in Ethiopia: I. Effects on selected crop and weed vegetative parameters and yield components. African Crop Science Journal 4: 399- 409.

60 T heme 4 Development of Small implements and I Machinery

Developing and Introducing Pre-Harvest Implements

Friew Kelemu1, Nasirembe VJW2, Mubarake Mohammed1, Abiy Solomon1 and Tilahun Teka1 1Melkassa Agricultural Research Center, Ethiopian Institute of Agricultural Research (EIAR) 2Kenya Agricultural Research and Livestock Organization (KARLO), Kenya

Abstract Ploughing has been a back breaking experience especially in wheal growing regions where ihe soil is heavy, high weed infestation is exhibited and where land is ploughed up to five times using the local plough. Seed is also broadcasted, which requires higher seed rate and also makes difficult subsequent operations like weeding. Poor seed viability were also observed due to non-protective measures in storage. It was ascertained with the survey work, where farmers indicated ploughing, seed dressing and planting were the priority areas to be addressed in the pre-harvest category. Accordingly, the already developed plough was batch produced and given out to 106 farmers in Robe and Sagure districts, Ethiopia. A six row planter was also developed, tested and introduced to the project site and a seed dresser was introduced to Kenyan farmers. Two field days were conducted on farmers’ field that used the plough and the planter. A number of people participated at the field days and favorable responses were received from farmers. Accordingly, manufacturers from the vicinity and different parts of Ethiopia were trained in the manufacturing skill of the implements so that these groups will handle the multiplication works hereafter. Introduction

Timeliness is very important and crucial in crop production lo get good yield and quality product. Land jreparation, planting, weeding and the other consecutive operations should be handled in time to get reasonable yield and quality product acceptablc along the value chain. To attain these, a proper package of mechanization technology comprising of tillage through crop establishment and including post harvest handling is necessary. In most wheat production areas, , the land is ploughed up to 5-6 times before planting using the traditional plough. This could be attributed to the narrow width, shallow depth, and non soil turning or inefficient trash burial nature of the implement. Ploughing takes long time, which inhibits the fanner from using the whole growing period, thus resulting in reduced yield especially in areas where late onset and early session of rainfall is exhibited. In most places, seed is also broadcasted which results in higher seed rate and makes weed management laborious. Use of untreated seed even if planted in time or planted in row, seed germination could be low, resulting in low crop stand and productivity. These shortcomings have resulted in poor quality of work timeliness problem and low yield at the end of the season. To overcome these problems an efficient animal drawn moldboard plough, a six row wheat planter and seed dresser have been developed by the Ethiopian Institute of Agricultural Research (EIAR), Mechanization Research program and KARLO, respectively. Hence, the objective of this paper is to introduce the improved pre-harvest technologies for increased wheat production.

Methodology

The work comprised of survey, prioritization of intervention areas, design and development, testing, demonstration, provide training for fanners and manufacturers.

61 Survey A survey was conducted in Robe and Tiyo districts in June 2011. Six kebeles and twenty farmers were selected with the recommendation of the two district agricultural office. From Degelu Tiyo District; Delu Bora, Ashebka Welkite, and Lolle Abosera Kebles were selected, while Habe Daengize, Gina Gedemessa and Habe kebles were selected from Robe district. Hundred twenty farmers were interviewed from the two districts. The questionnaire focused on capturing the crop enterprise, land holding, draft power condition, crop production techniques, and constraints ranging from land preparation, crop establishment, harvesting, and post harvest handling including the priority areas the project needs to address. Collected data was computed (SPSS 20) and priorities were set for appropriate interventions.

Plough The plough has been in operation for some time. It was fine-tuned at the project sites in association with farmers. It was tested with fanners and farmers were given hands on training on the operation of the implement with proper guidance of the oxen. Sixty and 42 farmers were selected from Sagure and Robe districts, respectively, in association with the district bureau of agriculture. The selected farmers were given practical field training on the use of the plough for three days at their respective districts. Each trainee was given a plough as well. Seven manufacturers from Assela, Arisi Robe, Sire, Eteya, and Huruta were trained on the manufacturing of the plough at Melkassa ARC workshop for five days. Farmers were advised to use the plough for their operation. A field day was organized on fanner’s fields, which used the plough to grow wheat. There were more than 100 participants. Farmers explained about the performance of the equipment and the participants were able to observe the merits of the technology practically.

Planter As there was no animal drawn wheat planter, the work started with the actual design. The focus was the seed-metering device, where the crop physical characteristic and recommended seed rate were the starting pcint. The seed geometry (length, depth read as major diameter and minor diameter), and 100 grain weight were determined. The seed rate of 150 kg ha ! and 20 cm distance between rows were taken as basis for the design based on the agronomic recommendation. These gave the seed metering flute depth and width, the ratio between the ground wheel and the seed-metering sprocket. The hopper was constructed from 1 mm sheet metaL with six openings at the base to make it a six-row planter. It has six furrow openers with an integrated covering device on each. The handle is made from a 0.5- inch water pipe with a provision for controlling the animal’s movement. The planter was calibrated in the workshop and on unploughed land. Adjustment was made on the degree of the opening, which regulates the seed from the hopper to the metering device. The equipment was tested on station and on farmer’s fields. The test was done on a plot size of 10 x 40 m, using the implements test procedure at the Agricultural Implements Research and Improvement center Data on speed, crop emergence, seed rate, and time of operation were collected. A comparative test was also conducted with the traditional practice at one farmer’s field.

Seed Dresser A design concept was developed from which working sketches were drawn leading to a bill of quantities (Fig 1). Materials were acquired from local hardware outlets, cut and shaped from local machine shops for assembly. Prototypes of a electrically and gasoline operated small grain seed dresser (KARI Seed Dresser) have been fabricated and tested for efficiencies and were under observation at KARI Njoro, Mwea, Kibos and Framers’ farm.

62 Figure 11 Schematic fabrication flow

Two models, gasoline and motor powered dressers, were tested. To determine the capacitywheat grains were weighed in batches of 5 kg. The drum door was slid open and loaded and switched on with an additional 5 kg batch at a time in each model until the machine was unable to rotate the loaded drum in each case. This was done without adding the liquid insecticide; it was rotated by switching on the engine/ motor. In each case when the drum could not rotate, offloading in steps of 1 Kg was done until the respective drums cold just rotate. Overhanging load was the calculated as follows;

To calculate overhung load, gear drive manufacturers use the formula:

Overhung Load = 126,000 x HP x FC x LF x PD x RPM

Where: HP - Horsepower; FC - Load connection factor; Lf- Load location factor; PD- Pitch diameter o f the sprocket, sheave or gear; RPM - Revolutions per minute o f the shaft

The load connection factor, or FC, describes the type of sheave, sprocket, or pinion mounted on the shaft. A flat-belt has relatively high tension in order to transmit the load by friction. A V-belt has moderate tension in order to seat the V-belt and to transmit the load by friction. A timing-belt has some tension and a chain has very little tension since the teeth transmit the load. A pinion or gear has a separating force related to the pressure angle of the tooth fonn. Therefore, each connection is given a different factor to account for this additional load. Flat-belt = 2.50, Pinion or gear = 1.25, V-belt = 1.50. Timing-belt = 1.30, Sprocket = 1.00.

Hence, 30% o f load was removed to acquire factor 1.00 in V-belt as in sprocket. The operating load capacity was calculated from:

Load factor for V-belt =1.3; Determined load capacity = 70 kg Load (gasoline) less overhanging load = 70 - (0.30 *70) kg = 49 kg recommend 50 Kg Load (Electric) = 45 - (0.30 x 45) Kg = 31.5 Kg recommend 30 Kg

63 Gasoline (rpm) — .Electric(rpm)

Figure 2 Selecting capacities for specific dresser

Uniformity of pesticide coverage To determine the uniformity of seed treatment on the electric seed dresser, two series of tests were carried out. In the first series, 30 Kg of samples of rice were treated with 700 ml of water containing 1% wt/vol brilliant sulphaflavine dye. Individual seeds were washed in 10 ml of distilled water and the washings were then analyzed with a fluorometcr to determine relative amounts of dye picked up by each seed. In the second series, five 30 Kg samples of the seed were treated with a mixture of 350 ml of Vitrax RS fiowable (Uniroyal limited) seed dressing and 350 ml of dye solution. Ten seeds from each sample were washed individually in 5 ml aliquots of distilled water and the washings were analyzed is before. The procedure was repeated for the gasoline powered dresser. The coefficient of variation was calculated for the individual seeds within cach sample.

Results and Discussion

Survey The result of the survey is indicated Table 2. According to the survey, plowing was the first priority area to be addressed followed by threshing and planting. Thus, the project decided to embark in that order of priority.

Table 2 Survey result indicating priority areas of intervention

Variables Percent case Quantity Land allotted for wheat 77 < 1 hectare Power Sourc* 90 A pair of oxen Start time/ Firlish time 80/83 Feb-March/June-July Number of pkDughings 83 3-5 Major weed t}ipe 94 Grass weed No of weedin 3 71 3 times Priority 43.4,18.4, 18.4 Plowing, threshing, planting

Plough Components of the plough include, beam, handle, moldboard and share (Figure 4). The handle and the beam are made of wood, the main ground engaging parts, the moldboard is made of 3 mm sheel metal, while the share is made from heat treated 6 mm sheet flat iron. The other specifications are specified in the Table 3.

64 MM

Fig 4 Components and parts of the plough specifications aid working Features

Table 3 Specification and working features of the plough

Overall dimension (mm) Weight (kg) Working Length Width Height With attachments Without depth (mm) attachments 660 310 243 16 5 125 Field capacity (m2/hr) Draft Unit draft (kg/cm2) Labor requirement (kg) 434 106 0.33 1 Source: Improved small scale Agricultural Equipment (Ratalogue 2009)

Demonstration Farmers were given hands on training al the testing sites and field days were conducted to crcate awareness among the main stakeholders (Fig 5).

Fig 5 Field day on the moldboard plough

Training of manufacturers As an exit strategy, the program has trained manufacturers so that farmers will have access for ihe plough in their locality. Accordingly seven people from Assela, Arisi Robe, Sire, Eteya and Hurula were trained on the manufacturing of the plought at Melkassa workshop for five days from March 4 to March 8, 2013. They were able to manufacture ploughs and carts and were successful in their project. The results of the plough tests are as shown in Table 4.

65 Table 4 Tes results of planter in June 2013

hejme Area Time Grain Application Fertilizer application Field capacity (.m2) {min) weight rate rate {kg ha-1) (hr h a 1) {kg h a 1) Kulumsa (pfanter) 400 11 7.5 187.5 100 4.57 Manually 400 58 6.2 156 100 24 Farmer 1 (f’lanter-moh) 400 13 7.8 195 100 5.4 Farmer 2 (3lanter-tad) 320 8 6 150 100 3.33 Melkassa / 3C (planter) 1000 23 18 180 100 3.83

Training of farmers and field evaluation At the beginning, lack of training of animals, poor degree of soil pulverization and some manufacturing problems were observed, which had some bearings on the performance of the machine. The prob cms encountered in 2013 season were corrected in 2014. Farmers in the project sites of Sagure and Robe were selected and training was given on the operation of the planters (Fig 6 and 7). The farmers also trained their oxen. The planter was tested at Melkassa, Kulumsa, Sagure, and Robe sites in 2014 crop season. Good crop emergence and stands were observed.

66 F ig 7 i raining of farmers on the use of the planter

Training of manufacturers Training was given for manufacturing of planter and cassava processor for 16 manufacturers who came from Oromya, Amhara, Tigray, and SNNP Regions. Each group manufactured one planter and a cassava processor. These groups are expected to produce the respective implements at least to their respective Regions.

Seed dresser The two dressers are electric power and gasoline operated (Fig 8). The machines main features are portability, easily and locally assembled, affordability, timely usage, requiring basic skills to operate, locai serviceability and gender friendliness (Table 5). The dressing efficiency was found to be over 95% and the lowest was 85% in offloading. The seed dresser coated 35 Kg of seed with insecticide in 30 seconds with an efficiency of over 95%.

able 5 Special features for the seed dresser

Characteristic G a s o lin e E le c tric D rum s iz e 0.21 n r 3 0.21 n r 3 Prime mover rating, Hp 5 .5 Hp 2 Hp Mean drum speed, rpm 41 3 7 Coating uniformity coefficient 0 .9 8 0 .9 6 "im e to acquire desired coating, sec 40 30 Offloading time, minutes 3 5 Rate of output, Kg h r1 5 4 5 .5 4 9 0 .9 Labor requirement, md 2 2 Cost of the assem bly, $ 56 8 5 2 4

67 Fig 8 Gasoline seed dresser in farmers grain store

The two models carrying capacity rating were determined as 30 kg for the electric dresser and 50 kg for the gasoline one. The amounts of chemical used per kilogram were the same in each case according to their manufacturers’ recommendation. The coefficients of dye on individual seeds compare favorably in the two types of seed dressers, for which coefficients of variation approaching 100%. The difference between the two coefficients of variation in this study is likely due to the effects of the difference in prime mover capacity, which led to a differential in ability to take overhanging load. Though increased motor horsepower will lead to increased cost of production and power requirement, the ability to take overhanging load may disappear. Time for attaining those respective coefficients was found to be 30 and 40 seconds for electric and gasoline dressers, respectively.

Conclusion and Recommendations

Training was given to the local Ministry of Agriculture staff, manufacturers, and farmers in Ethiopia. Thus, periodic awareness creation and experience sharing and the use of the distributed implements should be monitored. The status of job orders for local plough manufacturing from the trained manufacturers should be assessed. Methods should be devised with local fmancing institutes on the means of getting credit for both the manufacturers and fanners to avail the technology to the wider community. The planter has shown good performance and farmers are giving favorable response. From the planter tests conducted so far, it needs well-leveled field with less heavy clouds and training of the animals as well. The output of the two seed dresser models were 515 Kg hr'1 and 720 Kg h r1 for the electric and gasoline driven, respectively. The dressing treatment method indicated provides relatively uniform application of liquid seed dressings to batch quantities of seeds with less man-day requirement. Work should be undertaken to standardize pesticide application into the drum. Operators should be trained before using the machine. More work is required to establish energy requirement and germination trends of seed treated by the dresser.

References

Agricultural Mechanization Research Process. 2009. Improved small scale Agricultural Equipment. SPSS Statistical Software. 2011. IBM Corporation. Version 20

68 Developing and Fabricating Small-scale Wheat Thresher Nasirembe. W.W. Kenya Agricultural Research and Livestock Organization, P.O. Box 100, Molo Correspondence: [email protected]

Abstract To increase efficiency in agricultural production among small-scale farmers, mechanization was found lo be the main driving tool. When all the mechanization aspects were ranked through a surv ey, harvesting and seed processing were found to be the most crucial areas of need. Hence, the decision was made lo fabricate the small-scale wheat thresher and seed dresser machines. A prototype of a manually operated small grain thresher (KARLO wheat thresher) has been fabricated and tested for efficiencies and were under observation at Njoro, Mwca, and Kibos in Kenya. The threshing efficiency was up to 90% and the lowest was 85% in cleaning. Aggressive mechanization can contribute the overall output in food production and food security.

Background

To increase efficiency in agricultural production among small-scale farmers, mechanization was found to be the main driving tool (Wheat Baseline Report, 2012; unpublished). Isolated farmers have practiced mechanization in agriculture since the advent of agriculture in an unstructured manner (NAMS 1992). Almost all agricultural activities can be mechanized at different levels and magnitudes along the production value chain (FAO 2008). Critical areas of mechanization include seedbed preparation, sowing, pest control, harvesting, and post harvest handling (Singh 2000). When all the mechanization aspects were ranked through a survey, harvesting, and seed processing were found to be the most crucial areas of need (Wheat Baseline Report, 2012; unpublished) and necessary to fabricate the small scale wheat thresher and seed dresser machines. Another survey was carried out to take an inventory of available machines in use and determine their appropriateness (Nasirembe and Ooro 2012; unpublished). The outcome of the survey exhibited how small scale fanners incurred huge losses in terms of grain lose when they threshed wheat by using sticks (Hassane/ al. 1993). Again, the process is not cost effective. Attention has therefore turned to mechanization of threshing and seed dressing to reducc losses following the dwindling labor force as the youth opt to undertake white color obs, increase in demand for food, and the high cost of imported machinery. Imports are not appropriate to local conditions and may become unnecessarily expensive when the imported tools and equipments breakdown for luck of spares or obsolescence. While the rest of the developing world has movec from traditional agricultural production methods during the last 50 years, Kenya and the region has remained a sole importer of agricultural machinery. The objective of the study was therefore to innovatively design, fabricate, and test thresher for small grains in order to contribute to the food productivity.

Methodology

Understanding the crop physiology at maturity and forces involved while detaching the grain from the car formed the basis of original design while creating room for improvement processes. Sketches of the prototype and working diagrams or chronology of events were drawn from which a bill of quantiiies was derived (Fig 1). The materials were acquired and using the working drawings, materials were cut and machined to desired shapes. The parts were assembled and prototype tested. Data was collected on power requirements, threshing drum speed as a ratio of pedaling speed, winnowing fan speed, cleaning efficiency and rate of output. There are other issues that could be captur:d such as gender inclusiveness of the design, aesthetics, and consumer acceptability. Some of the daia were continually collected.

69 Fig 12 Schematic fabrication flow chart

Results and Discussion

The prob em areas were identified through a social economic survey that captured information from farmers aad through crop researchers. A design concept was developed from which working sketches were drawn leading to a bill of quantities. Materials were acquired from local hardware outlets, cut and shape d from local machine shops and assembled. A prototype of a manually operated small grain thresher (KARLO wheat thresher) has been fabricated and tested for efficiencies and were under observation at KARLO Njoro, Mwca and Kibos. The machines main features were portability, easily assembled, affordability by small-scale fanners, timely usage, requiring basic skills to operate, local serviceability and gender friendliness. The following are the major testing parameters that were included: dimension, weight, performance (field capacity, efficiency, losses) depending upon the machine for particular operation and crop, and durability. The threshing efficiency was up to 90% and the lowest was 85% in cleaning (Table 1).

Table 1 Requirements and results of the tests of the machine

Component Rating Power requirements 0.15KW (0.2 hp) Threshing drum speed 500 rpm Winnowing fan speed 6000 rpm Average output 110 kg h r1 Cleaning efficiency 85% Cracking 4.2.18 <2%

Special features for the thresher It is manually operated by two people, one to pedal and another to feed the harvest for threshing operation The pedal has the speed of 60 rpm. The machine is easily transportable on a small vehicle and can be easily assembled for use.

The power requirement of 0-15 KW is the amount of energy required to propel the thresher due to inertia developed by a drum cum flywheel effect. This amount of energy is equivalent to one used to cycle a bicycle on flat land. The cost as compared to the conventional thresher is only 1/10lh. The machine allows a farmer to thresh a crop at will as the crop can be preserved in a store when sure that it will be threshed. The head loses are reduced by timely sickle harvesting, drum loses are low at <2%

70 and no cracking was noted and the thresher reduces field losses from shuttering and sprouting. The machine can be serviced locally is an added advantage to its appropriateness.

Conclusion

Aggressive mechanization can improve the overall output in food production. Hence, wheat seed threshing prototype machine was designed, constructed, and tested. Seed moisture content and machine speed significantly affected the performance of the machine. Cleaning efficiency o f the machine increased with the increase in moisture content and speed within the range of moisture under consideration. The machine still requires improvement on cleaning mechanisms. Suggested improvements are gender bias, drudgery, move from prototype to product, fit wheels, and make it multi-crop.

References

FAO. 2008. Crop production. The United Nations. Hassan RM, Mwangi W and Karanja D. 1993. Wheat supply in Kenya: Production technologies, sources of inefficieneyand potential for productivity growth. Economics Working Paper No. 93-02. Mexico D.F., CIMMYT. Nasirembe, W.W and Ooro PA. 2012. Back to office report. NAMS (National Agricultural Mechanization Strategy). 1992. Formulation Report. Singh G. and Chandra H. 2000. Analytical approach to growth dynamics of agricultural inputs and their effect in increasing productivity in Madhya Pradesh. Agricultural Situation in India. March. 2000. pp 723-731.

71 T heme 5 Socio-economics; Value Chain Analysis and Development; Pre-extension Demonstration and Validation of Improved Technologies

Disseminating Improved Wheat Technology through Pre-Extension Demonstration in Ethiopia Bedada Begna1', Tesfaye Sollomon1, Mesay Yami1, Tadesse Dessalegn2, Workiye Tilahun1 Ethiopian Institute of Agricultural Research (EIAR), Ethiopia 2Wheat Regional Center of Excellence (WRCoE), Kulumsa, Ethiopia

Abstract Demonstrating, validating and popularizing the newly released rust tolerant wheat varieties to smallholder farmers is vital in facilitating rapid adoption of improved wheat technologies. In popularising newly released and existing rust tolerant wheat varieties, 118.4 tons of seed that covered 847 ha have been dispatched to 1,637 farmers since 2011 in major wheat producing areas of Ethiopia. To laci itate the extension effort of the technologies, around 3,800 (81% fanners) and 2,000 (88% farmers i individuals attended trainings and field days, respectively, organized on wheat production since 2011. Through demonstration and validation, nist tolerant wheat varieties have been transferred within short period by reducing the prevailing long chain of the formal system. Kakaba and Danda'a varieties were disseminated throughout wheat growing areas of Ethiopia within 3 years of their release. As a result, the income of farmers has significantly increased. Some farmers have begun investing on modern farming implements such as tractor and other household assets. Improvements in wheat production and productivity have been recorded in Ethiopia. However, increased population has concealed the impact, urbanization subsequently, increased demand for consumption of wheat products like bread, pasta, and macaroni. Hence, to sustainably increase wheat production, continuous supply of better yielding and rust tolerant varieties is crucial.

Introduction

Different studies indicated that wheat consumption is increasing at faster rate in Ethiopia as well as in Africa. According to Asfaw et al. (2013), the gradually rising incomes and urbanization have contributed to the shift in dietary preferences from other grains to wheat and rice. Similarly, Nicole et al. (2012) results suggested that rising incomcs, growing populations, women’s participation in the labor force increasing at a faster rate than men's, and wheat food aid are the key drivers for rising wheal consumption. In consideration of wheat importance for food security in Africa, Kamruzzaman and Mohammad (2008) recommended the need of increasing area under wheat production to bridge the gap between wheat demand and supply in the region. Despite enormous economic and dietary values of the crop in Ethiopia, the average yield has remained low due to multifaceted biotic and abiotic factors including insufficient and erratic rainfall, poor agronomic practices, diseases and insect pests and low technology utilization. Given wheat is among the most important staple food crops grown in Ethiopia, wheat research program under Ethiopian Institute of Agricultural Research (EIAR) has exerted considerable efforts in generation of high yielding rust tolerant wheat varieties. To justify investment in technology generation, improved technologies must reach users. However, more wheat varieties have been generated; their transfer to the smallholders for utilization takes longer time. For example, in the 2004-2009 study, only 3% of the land under wheat cultivation was planted with improved varieties of wheat seeds. As citcd in Abebe and Lijalem (2011), 96.5% is covered by local seed and 3.5% is by improved seeds of the total annual arable land coverage by major food crops in

72 Ethiopia. In order to ensure the dissemination of improved wheat varieties to farmers in shorter period, capacitating the farmers’ skill, knowledge, and acceptance is vital for sustainable diffusion and adoption of the technologies.

A sustained increase in agricultural production and productivity largely depends on the continuous development of new and improved varieties of crops accompanied with effective utilization of other inputs and required agronomic managements. In considering to food self-sufficiency in East African countries, East Africa Agricultural Productivity Project (EAAPP) has been established to support sustainable agricultural production and productivity in the region (Fekadu et al. 2012). In line to this, with the support of EAAPP, wheat technologies have been disseminated to the farmers in the country since 2010. To enhance rust tolerant bread and durum wheat varieties adoption, the varieties were demonstrated and evaluated on the farmers’ field. Hence, this manuscript is designed to describe the approaches used in technology dissemination, achievements attained, and impacts brought because of pre-extension demonstration and validation activity in Ethiopia.

Improved Wheat Technologies

Over the years, quite a number of wheat technologies have been developed and promoted for different agro-ecological zones of the country by the national wheat research program, regional agricultural research institutions and certain higher learning institutions. However, the ongoing efforts to improve access to these technologies are low in terms of area coverage and the number of beneficiaries (Bedada et al. 2014). In 2010, there was extended rainfall all over the Ethiopia throughout the year with prolonged precipitations and cool temperatures, which created suitable weather for stripe rust (Puccinia striiformis). The existence of susceptible wheat varieties, especially Kubsa and Galema, which were highly adopted, exacerbated the problem. Consequently, the resulting damage was severe and wheat production reduced by 8.03 % (GAIN 2012). The situation was a challenge for Ethiopia that strives to ensure food security by doubling agricultural production and productivity at the end of the Growth and Transformation Plan (GTP, 2010/11- 2014/15).

The circumstance has tremendously alerted wheat-breeding strategies towards broadening the genetic diversity by generating varieties of high yielding and disease resistance with acceptable quality traits. Developing stress tolerant wheat varieties for irrigated farming system is also an additional goal as well. Since 2010, as a remedy to the devastating danger of yellow rust, the national wheat research of EIAR has released nine bread wheat and two durum wheat varieties of which one bread wheat was released for irrigated areas (Table 1). Such effort of and proper dissemination of improved varieties has also helped to a steady increase in production and productivity i.e., 3.4 million tons of grain produced, productivity increased from 1.7 t ha'l to 2,1 t h a 1, and expansion of cultivated land (CSA 2013). However, the noticed improvement of wheat production and the growth has been a fraction of what is possible if more focused extension efforts are in place.

73 Table 1 Rust tolerant bread and durum wheat varieties available under production in Ethiopia

Plant Productivity Y e a r of D ays to height Rainfall Altitude (t ha-1) Resistance to Variety r e le a se maturity (cm) (mm) (m) at research field vellow rust Bread wheat

Huluka 2012 133 104 >600 2200-2800 3.9-7.1 Resistant G a'am bo 2011 91 102 Imgated 750 3 . 5 - 5 . 7 Moderately resistant Hogina 2011 135 95 >700 2200-2800 4.3-6.9 Resistant M ekelle-01 2011 84-90 7 7 .9 300-500 1980-2500 30-35 Resistant to stem rust M ekelle-02 2011 84-90 76.5 300-500 1980-2500 30-35 Moderately resistance to stem rust Shorim a 2011 126 102 >500 1800-2400 4.5-6.3 R esistan t T se h a y 2011 138 76 >900 2600-3100 3.8 R esistan t D an d a’a 2010 9 0 -113 110-145 >600 2000-2600 3.5-5.5 Resistant Galil 2010 1 13 73 >550 1800-2400 3.5-52 - K akaba 2010 85-100 90-120 500- 800 1500-2200 3.3- 5.2 Resistant Digelu 2005 9 5 -110 100-120 >600 2000-2600 3-5 Moderately resistant KBG -01 2001 80-90 80-100 >600 2000-2400 6.6 R esistan t Hawi 2000 90-100 105-125 >500 1800-2200 1.5-4 R esistant Madda-Waiabu 2000 95-110 100-125 >600 2300-2800 3.54.5 Resistant Tuse 1997 75-110 125-130 >600 2000-2500 4-6.5 Resistant Pavon-76 1982 90-100 120-135 >500 750-2500 3-6 Resistant E T -13 1981 105-120 127-149 >600 2200-2900 3-6 Moderately resistant

K 6295-4A 1980 100-115 128-131 >600 1900-2400 3-6 Moderately resistant Durum wheat Toltu 2010 125-13 5 78 500 2300-2600 4.4-6.0 Moderately resistant Mangudo 2004 110-124 88 700-1000 1880-2700 43-50 Mukiye 2004 106-116 88 700-1000 1880-2700 40-56 Denbi 2001 106-122 84 800-1200 1800-2650 40-56 Hitisa 2001 105-121 85 800-1200 1800-2650 40-60 Yerer 1994 110-130 84 800-1200 1800-2400 30-50 Ude 1994 105-115 80 800-1200 1800-2400 30-50 Ginchi 1992 120-140 106 800-1200 2000-2300 30-50 Source: Variety registry book (2010 and 2011)

To enhance the improved technology access to the farmers, newly generated and the existing best-bet wheat technologies were disseminated. Technology deployment started with inventory of best-bet wheat technologies. The inventory identified 32 bread and 33 durum wheat varieties that were released in the country between 1980 and 2012 and still in production (Table 1). M oreover, eight integrated disease and crop management, one integrated weed management, and two farm implements were identified (Fikadu et al. 2012). Due to the weak extension system prevailed in the country access to improved agricultural technologies is limited. To improve the prevailing dire situation. EAAPP came to the forefront in 2010 to stride for availing wheat technologies among the small holding fanners of .Ethiopia.

Deploying the technologies In addition to technology generation and inventory, the research extension system is engaged in evaluation and validation of wheal technologies among selected farm households only in limited areas under the supervision of the researchers to popularize the promising technologies. The wheat technology dissemination was conducted in a collaborative approach among researchers, extensionists, and farmers. The pre-extension demonstration activities were accompanied by capacity building of development agents (DAs), MoA subject matter specialists (SMSs) and fanners. The pre- extension demonstration has offered trainings and prepared showcases lo strengthen the capacity of farmers in technology up-takc and utilization. The dissemination strategy has included full technology packages for higher productivity and production. Multidisciplinary team of socio-economists, extensionists, breeders, pathologists, agronomists and weed scientists was organized to accomplish the various duties of the technology deployment activities. Close linkage was crcated among farmers, extension and research system for better technology dissemination and utilization. The team from

74 different disciplines delivered the information, and created awareness about the technologies with required trainings and advices with the support of EAAPP.

Achievements Delivering wheat technologies: The pre-extension demonstration and validation of wheat technologies commenced in 2011/12 in 11 districts in Arsi and West Arsi Zones. During this time, 9.9 t of three bread wheat varieties namely Kakaba, Danda’a and Digalu were dispatched among 247 beneficiary’ farmers (Table2). Since the pre-extension demonstration and validation, was accompanied with all necessary trainings and advisory services, some participant farmers had obtained up to 7.6, 7.2 and 6.4 t ha'1 from Kakaba, Danda’a, and Digalu varieties, respectively, in Digalu-tijo and Arsi- robe districts (Bedada et al. 2014). With success stories in 2011/12, the activities were expanded to Oromia, Amhara, Tigray, and SNNPP regions, which has the potential for wheat production. Twelve bread wheat varieties (BW) and 3 durum wheat varieties (DW) were demonstrated and similarly 10 BW and 5 DW varieties were used in pre-scaling activities in 2012/13. About 52 t of improved seeds was dispatched and covered 426 ha of land in 2012/13 (Tables 3 and 4). Similarly, 66.4 t seed of bread and durum wheat varieties were distributed among 56 districts in the four regions in 2013/14 and a total of 118.4 t seeds of improved varieties had been accessed to 1,637 households of which 184 were female headed households (Tables 5 and 6).

Table 2 Quantity of seed and wheat varieties used in pre-extension demonstration in different districts of Arsi zone in 2011/12

Wheat varieties distributed in tons (t) Districts Kakaba Danda’a Digalu Total Digalu tijo 0.6 0.64 0.12 1.36 Arsi rcbe 0.68 0.52 0.12 1.32 Hetosa 0.62 0.3 0.12 1.04 Limu bilbilo 0.6 0.6 Tiyo 1.9 0.38 0.08 2.36 Munesa 0.28 0.84 0.04 1.16 Honkclowabe 0.2 0.1 0.3 Guna 0.2 0.2 0.2 0.6 Gedebassasa 0.4 0.4 Lodehetosa 0.26 0.1 0.36 Sire 0.4 0.4 Total 5.54 3.68 0.68 9.9 Area coverage (ha) 34.625 23.00 4.25 61.785 Total number of farmers who took bread 247 whea- seed for pre-scaling up

Table 3 Beneficiary farmers in pre-extension demonstration in four regions of Ethiopia (Oromia, Amahara, Tigray and SNNPP), 2011/12-2013/14

Y e a r Beneficiaries Varieties (No) Seed No of beneficiary farmers (No) (No) dispatched A rea Zone District Demo Pre-scaling(t) Demo Pre-scaling co vered (ha) M | F M | F 2011/ 2 2 11 3 BW - 9.9 247 6 1.7 8 5 2012/ 3 19 54 12 BW+ 3 DW 1 0 B W + 5 D W 52 212 12 17 4 70 426 2013/ 4 22 56 14 BW+ 6 DW 11 BW+6DW 66.4 331 38 7 3 7 64 421 Total 41 110 118 .4 543 50 910 134 847

75 Table 4 High yielding bread and durum wheat varieties used for demonstrations and pre-scaling up in 2012/13 season

Areal V arieties No. of farmers S e e d co v e rag e disiriDuiea P re-scannq D em o (ha) R egio n s Z o n es Districts Pre-scaling Demonstration (ton) MFM F Oromia Arsi; lllu-ababora; Kellem- 38 5 bread wheat varieties ('Kakaba' 2 7 .7 13 50 40 150 9 306 6 bread wheat varieties (Shorima: Wellega; North Shewa, west and 'Danda'a'; ‘Digelu’;’Meraro’ and Huluka: golcho; Hidase: Jeferson; Shewa and Gurage zone ‘Alidoro’) and 5durum varieties Galil) and 3 durum varieties (Ude; Asasa; Quamy; Ginchi and (Mangudo: Mukiye and Hitossa) Y e r e r )

Tigray South, Eastern and South 6 3 bread wheat varieties (Digelu, 1.0 7 24 15 5 bread wheat varieties (Kakaba; 6.1 eastern rain fed and Kakaba and Mekelle-03) Danda'a; Mekelle-01; Mekelle-02, and moisture-stress areas of M ekelle-03 Tigray

North Shewa; East and 8 4 bread wheat varieties 9.73 8 7 5 4 7 3 38 Am hara West Gojam; North Gondar; (Menze;Tay: Gassay and 5 bread wheat (Menzie; Shorima; South and North-Western Dinknesh) Tsehay;Tay and Digalu) W ollo

SNNPR Kamash and 2 'D an d a'a' 4 7 9 10 - - 22.25 - B am basi

T on go 1 D anda'a' - 9.45 201 15 -- 54 Total 19 54 10 bread and 5 durum wheat 12 bread and 3 durum wheat 52 17 4 1 70 212 12 426 varieties varieties Code: Demo-Demonstrations; M= Male; F= Female (Mesayet al. 2013)

76 Table 5 Demonstrations and pre-scaling up of bread wheat activity in 2013/14 crop seasons

Varieties Seed No. of farmers Areal (q) Pre-scaling Demo coverage Region Zones Districts Pre-scaling Demonstration M F M F (ha) Oromia Arsi and West Arsi Digeu and tijo, Arsi robe, Aseko, Merti, Jeju, 'Kekeba, Dendea, Ogolcho and Hidasse 84 167 18 45 5 52.5 zones Ghole, Diksis, Sude, Shirka, Assasa, Tiyo, Shorima and Sire, Tena, Hetosa, Ziway-Dugda, Limu and Huluka Bilbilo, Munesa, Enkolo-Wabe, Guna, Dodota, Arsi-Negele and Huruta Bale Dodola, Agarfa and Ginnir Tusie Danda’a, Kakaba, Digalu, 5.3 165 12 1.77 M/Walabu Huluka, Hidase and Shorima East and Klaim Seyou and Chora Digeluand 4.8 114 6 - ■ 30 Wellega Dendea Amhara North wollo Dawunt and Meket Kekeba, Digelu, 56.5 136 15 -- 37.57 Minzie, Dinkinesh South-Wollo Woreilu and Jama Dinkinesh - 10.5 46 3 - - 8 North Shewa Siyadebirnawayu Menze 31.5 55 1 - - 18 North-western Y/Densa, Dembecha.Mecha, Enarje Tay and Gassay 40.5 87 7 - - 27 Enawga and Gozamin Tigray South eastern zone S/samre - Mekelle 3 and Mekelle 4 1.2 -- 18 2 0.8 Eastern and South Hawzen and H/wojerat - Mekelle 1, Mekelle 2, 3 - - 41 9 2 eastern zone picafior Total 41 11 varieties 14 varieties 233 605 50 269 28 175 Code: Demo—Demonstrations; M— Male; F= Female

TJ Table 6 List of high yielding Durum wheat varieties used for demonstrations and pre-scaling up in 2013/14 season

Varieties Seed No. of farmers Areal distributed Pre-scaling Demo Coverage c c \H/In) MIVI 1 M r (ha)V'al Oromia Bale zone 9 6 Durum Wheat varieties 3 Durum wheat varieties 409 119 11 10 2 233 Eastshoa; Ada, Lume, (Ude Mangudo, Mukiye, (Mangudo, Mukiye and Ginbichu, Adulala (Ziqala) Denbi, Assass and Hitosa) Yerer Northshoa; Aleletu Tigray Southern zone of Tigray 1 3 Durum wheat varieties 1 5 1 1 (Mangudo, Mukiye and Yerer Amhara North Shewa; East Gojjam; 5 1 Durum wheat variety (Ude) 5 Durum wheat varieties 12 11 1 23 3 8 South and North-Western (Mangudo, Mukiye, Yerer, Wollo Hitosa, and Ginchi SNNPR Guragezone: (Mesken, 4 4 Durum Wheat varieties 6 Durum Wheat varieties 2 2 24 4 4 Mereko, Sodo) (Mangudo, Mukiye, Denbi, (Mangudo, Mukiye, Denbi, Hadiyazone; (Lemo) and Hitosa Assass, Ude and Hitosa ) Total 10 15 6 durum wheat varieties 6 durum wheat varieties 431 132 14 62 10 246 Code: Demo Demonstrations; M = Male; F Female

78 Capacity building through information delivery: To better meet the challenge of sustaining and increasing agricultural productivity, increased access to quality education and training in agriculture at the fac litator and farming community levels must be exploited to continually update the knowledge and sk 11s of extension workers and farmers. Producers’ knowledge and skill of wheat production plays a prominent role. Organizing trainings to farmers, DAs, Agricultural experts or SMSs and others, on field advisory service during supervisions of the demonstrations were the major methods used to upgrade the capacity of producers and other stakeholders. Accordingly, 604, 1553 and 1679 individuals (farmers, DAS and SMSs, researchers and others) attended the training organized on wheat production in 2011/12, 2012/13 and 2013/14 production seasons, respectively, in Oromia, Amhara, Tigray and SNNPP (Table 7). The skill and knowledge gap filling training was organized by multidisciplinary team of researchers from the respective agriculture research centers in the regions. The thematic areas of the training were inputs required for wheat production and their utilization, seed production and its system, agronomic and disease management, mechanizations and crop diversification, and chemicals (fertilizers, herbicides, pesticides and fungicides) utilization and precautions while its use.

Table 7 Wheat production training attendee in 2011/12 to 2013/14

Farmers DAs and Experts Researchers Others Year M FMFM FMF Total

2011/12 417 84 46 7 29 - 21 - 604 2012/13 1152 185 114 24 52 5 21 0 1553 2013/14 1078 195 270 43 27 1 52 13 1679 Total 2647 464 430 74 108 6 94 13 3836

Popularizing wheat technologies: Sharing agricultural information and improving awareness of farmers towards improved agricultural technologies and gathering the feed backs and consicering it into agricultural researches and developments enhances the acccplances of farmers to the technologies and leads to effectiveness of research outputs. In order to create showcase for non- beneficiary fanners and improve rust tolerant wheat varieties adoption, different field days were organ zed in many EAAPP districts of Ethiopia (in three districts of Arsi zone in 2011/12 (Bedada et al. 2014), 19 EAAPP districts in 2012/13 (Mcsay et al. 2013) and nine EAAPP districts in 2013/14. The pre-extension demonstration and validation activity conducted among the four regions built greater awareness about improved rust tolerant wheat varieties and increased the interests of smallholders surrounding the activity. Organization of field days were used as means of bringing stakeholders together to discuss status of the ongoing activity and to gather the opinions, ideas and perceptions of the stakeholders to be considered in future research activities. Different stakeholders including DAs, SMSs, researchers, higher officials, and farmers attended the field days during promotion of the rust tolerant wheat varieties. A total of 4951 participants (241 in 2011/12, 2687 in 2012/13 and 2023 in 2013/14) attended wheat field days (Table 8). Relatively a significant proportion of female participants also attended the wheat field days. From the total wheat field day attendee about 12.3% (608) of them were female participants in those years (Table 8).

Outputs and impacts: However, the outputs and impacts of pre-extension demonstration and research efforts are varying in different parts of Ethiopia; different sourccs witnessed the significant impacts and outputs of the activity. The existing rapidly increasing demand for wheat consumption, which attributed lo different factors like rising incomes, and growing populations outweighs wheat production which further hides the impacts and outputs of research and extension efforts. Whatsoever the case, the pre-extension demonstration and validation activity popularized rust tolerant wheat varieties, improved the access of their seed among the producers and increased incomes of significant wheai farming communities in Ethiopia.

79 Table 8. Wheat pre-extension demonstration and validation field day participants (2011/12 to 2013/14)

Field day participants Year Farmers DAs and Agri. Others or higher Total experts officials M F M | F MF 2011/12 157 69 15 241 (7.5%) 2012/13 1566 261 574 105 163 8 2687 (14.3%) 2013/14 1596 188 113 21 98 7 2023 (10.7%) Total participants 4951 (12.3%) Note: percent in parenthesis represent proportions o f female attendee

Rust tolerant wheat varieties: Majority of the rust tolerant wheat varieties used in pre extension demonstration and validation program has been widely popularised. The pre-extension demonstration and validation program of EAAPP has contributed for popularizing the rust tolerant wheat varieties directly through training 3,836 individuals and involving 4,951 individuals in field days (Tables 7 and 8). Stakeholders indicated some important traits for some varieties that could be considered and be focused in the breeding program. In general, farmers observed that improved wheat varieties are more productive, relatively easy to thresh except “Danda’a'’ than local varieties and have high potential in disease resistance (rusts). However, the feedback from farmers varied among different locations in the country, which is discussed as follows.

Although majority of the wheat varieties that has been released since 2010 were considered as rust resistant, farmers in Arsi-robe district have indicated that “Danda’a' is susceptible to yellow rust although il is preferred among farmers in Bale zone of Oromia region due to its disease resistance and late maturity trait. Further, farmers in highlands of Arsi and Bale indicated that “Kakaba” is early maturing type and recommended it for areas with short rainy season (lowland and middle altitude areas of th; country). Farmers’ Field observation and assessment in Bale zone indicated that “Digalu” is hit adversely by stem rust in 2013/14 production season and the situation has become the concern of majorly “Digalu” producing farmers in Arsi highlands and other areas with similar agro-ecologies in the country. High shattering, low bread quality due to low elasticity, and less market demand when compared to “Kubsa” were the drawbacks complained againest “Digalu" by the smallholders and processors Similarly, “Danda’a” is relatively difficult for manual threshing and cool weather condition accompanied by wind affect the crops at maturity (grain filling) as pointed by farmers in Digalu-tijo district in Arsi zone (Table 9). Although farmers in Arsi zone are interested in “Shorima”, they stated that the variety lacks uniformity in m a tu r ity (different m a tu r ity level w it h in a single plant).

Table 9 Somo traits pointed by fanners as problems for some varieties

Variety Drawbacks of the variety as perceived by the farmers Digalu High shattering problem, low bread quality when compared to "Kubsa\ low market demand due to flour factories limited demand Danda’a Affected by cool weather during maturity, and difficult for threshing manually Shorima Lacks uniformity (different maturity level within a single plant)

Enhancing access to seeds of rust tolerant wheat varieties: However, Ethiopian seed system is composed of both the formal and the informal system (sometimes called local or fanners’ seed system). The formal system is the original source of improved seeds for the informal system, the informal seed system (self-saved seed or fanner-to-farmer seed exchange) accounts for 90% of the seed used by smallholder farmers (Belay 2004). The share of formal seed system is less than 10% (FAO-CDMDP 2010). Currently, the share of formal seed system includes 10-20%, while the informal system covers 80-90%.

Given the critical role that improved wheat varieties play in increasing wheat production, a key question is how to facilitate the development of a seed system that is capable of availing quality seed

80 of improved varieties at appropriate time and in a cost-effective way. The EAAPP support on pre- extension demonstration and validation has contributed for easily accessing the rust tolerant varieties in a short cut through delivering the prc-basic seeds to the farmers. The initially delivered seed was revolving which shortens the long chain it took through formal system. Farmers who have initially received the varieties in the area were advised to use the first product only for seed for themselves and local farmers in their system through exchange, sell, gift, and other seed exchange mechanisms (see case 1 in the following section). Generally, the impact of pre-extension demonstration and validation together with emphasized agricultural extension was high for reaching of new and improved wheat varieties (“Kakaba ’ and “Danda ’a”) within two to three years of their release almost in all parts of the country.

Case 1 Activities and achievements obtained among participant farmers

Hajo Mergo lives in Kechma Murqicha kebele in Digalu-tijo district in Arsi zone and the life of her family depends on agriculture. She cultivated different crops like wheat, maize, potato, faba bean and barley. Among these crops she allocates wider plots for wheat relatively. In the past years, she was growing local wheat varieties that were less productive and susceptible to diseases. In the recent two years she started growing improved wheat varieties that given by KARC with the help EAAPP.

In 2011/12 cropping season, she has 40 kg of “Danda'a” together with training with other farmers on wheat production from KARC in Sagure, where the district’s BoA is located. Then she planted it on 0.25ha of plot by appliying all the lessons from training she produced 20quintals of “Danda’a ” from the plot. She used certain of the produce for her own seed, she gave some to her neighbors and relatives through exchange with local varieties and crops for home consumption, and she sold major part to other farmers for seed with attractive prices ranging from 1000-1400 Ethiopian birr. Because of the benefits and experiences obtained in 2011, she expanded the plots to 2ha and additionally she has 40kg of Kakaba, 40kg of Shorima and 1.75kg of Jaferson in 2012/13 cropping season. Adde Hajo said that the frequent supervision and advice from KARC researchers initiated her to look at her field every morning and evening to manage it on time as required. By the support, she got from EAAPP, adde Hajo told that she benefited economically and she improved her skill and knowledge regarding to wheat production. The farmers of the area have also learnt more through observing her fields and practice and many ______farmers of the surrounding have attracted by her practice and changes.______Source: Bedada el al. 2014

Increase of households’ income: Despite difficult to conclude that merely the pre- extension demonstration and validation work done by EAAPP funding increased households’ income, it is also hard to deny its contribution in improving wheat production and productivity among wheat producing community in Ethiopia. The combined supports of EAAPP and other international research organization and projects (ICARDA, CIMMYT, DRRW, and USAID) have played a prominent role of wheat research, development, and extension in Ethiopia since reccnt years. The increase of wheat production and productivity among farmers in Ethiopia is noticed. However, the impact has been swallowed by increased population, urbanization, and increasing wheat demand for consumption of processed wheat products such as bread, pasta and macaroni with relatively low price and easier in processing.

8 1 The impact of pre-extension demonstration is significantly observed on farmers who initially involved in the activity. The intervention has benefited the participant and the surrounding farmers in two ways. The first one is satisfactory supply of quality seed (pre-basic) for the household and seed contribution to the community. The second aspcct is higher and sustainable income for the family from sale pf quality seed at a substantially higher price compared to grain price (case 1). For instance, farmers ill Digal-Tijo district sold a quintal of “Danada’a” variety on average at 1,200 birr while wheat grain price remained at about 700birr per quintal in June 2012. Moreover, the community in target areas benefited from easy access of wheat seed through purchase, seed exchange, and gifts as mentioned in case 1. Due to improvements in productivity and production, some farmers in wheat belt areas of the country has bought tractor (individually or in-group) to transform the traditional wheat farming system towards mechanization in recent two years. Furthermore, majority of farmers in Arsi and Bale zones have started wheat production through full packages by incurring all production costs that could directly related to rising incomes of wheat farming community in the country.

Conclusions and Recommendations

The pre-extension demonstration and validation attributed significantly for the dissemination and popularizing rust tolerant wheat varieties in target areas. Prominent yields were recorded by those farmers who adopted the technologies with its packages, which indicated the potential productivity of the varieties at farmers' field and management. The collaborative work of research extension and extension under MoA contributed for the observed wheat production improvement in the country; however, the growth has been a fraction of what could be possible with more focused extension efforts in place. Due to the developments of new rust types, all the previously released rust tolerant wheat varieties are exposed to new rust races that have adversely affected “Digalu”, which is high yielding in many palces in the country. The variety became susceptible to new stem rust race as of 2013/14 production season in Bale Zone. Such situation of short life of varieties due to rusts is a concern to sustained wheat production, to realize wheat self-sufficiency and rcduce grain importation. The pre-extension demonstration and validation had popularized and identified the farmers’ interests and views that need consideration in breeding program. In addition to improved varieties, farmers highly emphasized the constraint of row planters, unavailability of improved wheat seed at required time and quantity, unavailability of chemicals (herbicides and fungicides for rust and septoria) and development of different grass weeds due to continuous mono-cropping and unexpectedly changing rainfall pattern and distribution. Based on the results and impacts noticed during the intervention, the following points are the issues that need to be addressed in the future to enhance wheal production and productivity in Ethiopia as well among other East African countries.

In addition to accessing improved rust tolerant wheat varieties, initiating and fostering the farmers towards utilization o f full wheal production packages that attribute to enhance wheat productivity; Farmers well recognized the advantages of row planting and simple row planting technology has been remained the farmers’ main issue. Hence, availing row planting technology needs concern in the future; Strong linkage among research, extension, and farmers is indispensible for participatory effective technology generation, evaluation and rapid up-scaling of the best technologies fulfilling the farmers’ need and interest; Because of the frequent evolution of pathogens (leaf, stem and yellow rusts), ensuring the availability of fungicides for the farming community needs attention as integrated option. Necessary trainings have to be delivered to wheat producers for effective utilization of the chemicals (when to apply, how to apply, and all safety measures during application); Effective delivery of technical advices and support is required from the respective agricultural experts such as DAs and SM Ss to farmers to improve wheat productivity through their own management; and W idening and strengthening the pre-extension demonstration and validation is supreme to enhance wheat productivity as well production in the region in the future.

82 References

Abebe Atilaw and Lijalem Korbu. 2011. Recent Development in Seed Systems of Ethiopia. Debre Zeil Research Center, Ethiopian Institute of Agricultural Research, Addis Ababa, Ethiopia. Asfaw Negassa, Jawoo Koo, Kai Sonder, Bekele Shiferaw, Melinda Smale, Hans Joachim Braun, Dave Hodson, Sika Gbegbelegbe, Zhe Guo, Stanley Wood, Thomas Payne and Bekele Abeyo. 2012. The Potential for Wheat Production in Sub-Saharan Africa: Analysis of Biophysical Suitability and Economic Profitability, September 2012, CIM M YT, Addis Ababa, Ethiopia. Bedada Begna, Messay Yarni, Werkiye Tilahun, Firdissa Eticha, Bcdada Girma and Fiqadu Fufa 2014. Participatory pre-extension and demonstration of newly released rust tolerant and high yielding wheat var eties: Experience of EAAPP in Arsi Zone, Ethiopia. In: Proceedings of the Eleventh international conference on Ethiopian Economy, Ethiopian Economics Association (EEA), June 2014. Volume (II): pp 16 .--18 2 . Belay Kasa, and Manig W. 2004. Access to Rural Land in Eastern Ethiopia: Mismatch betweenPolicy and R e a lity . Journal o f Agriculture and Rural Development in the Tropics and Subtropics 105 (2): 123-138. CSA (Central Statistical Agency). 2013. The federal democratic republic of Ethiopia, Central Statistical Agency (CSA) Agricultural Sample Survey 2012/13 (2005 EC). Volume I report on area and production of major crops. May, 2013. FAO-Crop Diversification and Marketing Development Project. 2010. Seed Value Chain Analysis as a means for Sustainable Seed System: a case of farmers based seed production and marketing in Arsi Zone, Oromia Region. Assela, Ethiopia. Fekadu Fufa, Bedada Begna, Mekonnen M, Workiye Tilahun, Mesay Yami, Firdissa Eticha, Karta Kaske, Solomon Galalcha, Bedada Girma. 2013. Exploring existing wheat technologies and speeding-up transfer to farmers: experience of EAAPP-wheat commodity. Wheat for food security in Africa: Science and policy dialogue about the future of wheat in Africa 1:28. GAIN (Global Agricultural Information Network). 2012. Ethiopia grain and feed annual report, Addis Ababa. http://gain. tas.usda.gov/ReccntGAINPublications/GrainandFecdAnnual AddisAhaba Ethiopia 5-24- 20 i 3.pdf. Accessed 23 May 2013. Kamruzzaman M and Mohammad HI. 2008. Technical efficiency of wheat growers in some selected sites of Dinajpur district o f Bangladesh. Bang. J. Agri. Res. 33(3): 363-373. Mesa> Yami, Bedada Begna, Bedada Girma, Terefe Fita, Tesfaye Sollomon, Werkeye Tilahun, Firdisa Eticha, A\ele Badebo, Alemayehu Assefa, Dawit Habite, Werku Denbel, ZerihunTadesse, Solomon Gelelcha, Firew Kasa, and Desta Gebre. 2013. Enhancing Adoption of Rust Tolerant Wheat Varieties: Expriance of EAAPP in Ethiopia. Ethiopian Institute of Agricultural Research, Wheat Regional Center of Excellence (W RCoE), EAAPP, 2013. Nicole MM, Jayne TS, and Bekele Shiferaw. 2012. Wheat Consumption in Sub-Saharan Africa: Trends, Drivers, and Policy Implications. M SU International Development, W orking Paper 127.

83 Determinants of Farmers’ seed demand for improved wheat varieties in Ethiopia: A Double Hurdle Model Approach

Tesfaye Solomon', Bedada Begna1, and Mesay Yami1 1Ethiopian Institute of Agricultural Research (EIAR), Ethiopia vCorresponding author: [email protected]

Abstract The study using the double hurdle model empirically identified the most important farm households socio-demographic characteristics that are affecting wheat seed demand and investigated their effects on wheat seed demand. Region and farm size were significant in explaining both the decision to participate in purchasing wheat seed and the level of wheat seed purchase. Improved wheat variety use in the last five years has been found to have a negative relationship on farm households’ decision to purchase wheat seed but it was not important on the quantity of wheat seed purchase. Farm households who only read and write have a less likelihood to participate in purchasing of wheat seed. In addition, economic factors such as income and livestock ownership were among the significant determinants of wheat seed purchase demand. Calculated non-farm income elasticities, for those who purchased wheat seed, indicated that fanners’ wheat seed demand sensitive to changes in non-farm income.

Introduction

Availability of quality seed of improved varieties al required amount and affordable prices has been a milestone of developments recorded in wheal production. Assured supply of breeder’s and pre-basic seed is crucial from public breeders to engage private companies to produce seed of public cultivars. In a regional seed market, companies may need breeders and pre-basic seed to produce and sell seed in several countries. Failure to use appropriate seed, while investing sufficiently on other inputs and management practices, usually yields against expectations. This can be observed in the improved seed coverage and national wheat productivity in Ethiopia. During 2009/2010, only 2.25% was sown with seeds from the formal sources of 1.68 million hectare of land covered with wheat (CSA 2010) indicating that the vast majority of seeds used by small farmers in the country is obtained from the farmers’ seed system. Moreover, Dawit Alemu and Spielman (2006) had summarized that only 20 % of the demand for improved seed was covered in 2005 demonstrating that nearly 3000 tons of improved wheat seed is required to satisfy the present demand. Experience has shown that the predicted demand for wheat seed usually does not conform to the demand at planting times. When farmers revise their expectations of rainfall, prices and other factors, they incline to shift their interests. This frequently causes significant coordination problems for seed suppliers. It was well evidenced by Ethiopian Agricultural Research Institute (EIAR) national scaling up initiative and seed sales reported by seed suppliers. Hence, clearly defining demand dynamism for wheat seed is crucial. The current research was carried out with the objective of presenting the important farm household socio-demographic characteristics that arc affecting wheal seed demand and investigate their effects on wheat seed demand.

Methodology

Multi-stage purposive random sampling procedures were followed from higher to lower administrative levels, with farmers being the sampling units. The survey was carried out in three regional states i.e., Amhara, Oromia, and SNNPR in Ethiopia. A four-stage sampling procedure was adopted involving the selection of zones, districts, peasant associations and wheat farmers. Purposive selection of administrative zones, districts, and peasant associations was carried out based on area of wheat coverage. Ultimately, 763 farmers were interviewed using a structured questionnaire.

84 Data analysis DcscriDtive statistics was used to describe the socioeconomics and demographic characteristics of the sample households. Means, percentage, frequency, and graphs were analyzed using SPSS computer program and significance lest was conducted using t-test, and Chi-square. To analyze the demand of improved wheat varieties using farm household survey data, the Tobit and a more flexible parameterization to the Tobit model (the double hurdle (D-H) model) were considered. The tobit (TOBIN, 1958) model specification is defined as: yL= y ’ if y/ > 0 (1) y L = 0 Otherwise

The latent function y* that defines household participation decision and amount of purchased improved wheat varieties is given by: yj* = Xj' J3 + £j , where £j~ N( ,u i, a 2) and i = 1,..., n

The latent is defined variable y* as a variable that may or may not be directly observable and is the corresponding actual observed the purchase of an improved wheat variety measured in terms of proportion of wheat area allocated to improved wheat variety.

X( is a set of individual characteristics that explain both participation and the purchase o f improved wheat variety', and (3 is vector o f Tobit maximum likelihood estimates, pi the independently and normally distributed error term assumed lo be normal with mean zero and constant variance a. The value o f yifor all non-users equals zero (Alene 2000). £ (- is assumed to be a homoskedastic, normally distributed error term.

Equation (1) states that the observed purchase of an improved wheat variety becomes positive continuous values if only positive purchase of improved wheat varieties is desired, but zero otherwise. This show^s the observed 0’s on y t can mean either a “true” 0 (i.e., due to the individual’s deliberate choice) or censored 0 (i.e., those caused by survey design) (Wodajo, unspecified).

The "obit model is estimated using maximum likelihood methods. The log-iikelihood function verif\ ing equality of the coefficients in the participation equation to those in the purchase equation is:

LnLT = Zyi, y. j [ln(2n:) + Ina2 + (r,^ f )2] + ZyU In 1-0, [ ^ ] (2)

X P When i denotes the standard norma! distribution function evaluated at - L- and the summation indexes refer to the limit and the non limit observations. The first termon the right hand side o f the equation (2) is the contribution of the non limit observations to the log-likelihood function, while the remaining terms represent the contribution of the limit observations (Reynolds 1990).

The D-H model is a parametric generalization of the Tobit model, in which two separate stochastic processes determine the participation decision to purchase and the amount of purchased of technology (Hailemariam 2006). The first equation in the D-H model relates to the decision to participate in purchase (y) can be expressed as follows: y£- = 1 if y ' > 0 and 0 if y* < 0 (3) y ’l = x{a + (Participation equation)

Where: y ’ is latent participation in purchasing o f wheat seed variable that takes the value o f 1 if a household purchased improved wheat variety and 0 otherwise, x is a vector of household characteristics and a is a vector o f parameters.

Equation (3) is a probit model that examines the probability that the ilh farmer would make a participation decision to purchase improved wheat varieties.

85 >ndThe second>ndThe hurdle, which closely resembles the Tobit model is expressed as: t, = t- > 0 and ify^* > 0 (4) tj = 0 O th e rw ise t‘ = Z,/? 4- u t (Purchased amount of wheat seed equation)

Where: t, is the observed response on how much Kilogram of wheat seed purchased, Z is a vector of the household characteristics and /? is a vector of parameters (Mignouna 2011). £, and u t are error terms

£,~/V(0,1) and u t ~ N (0 ,c j2).

Following (Cragg 1971) model, the study assumes independence between the two error terms. The log-likelihood function for the D-H model is as:

L n U = £< In [<*>(Z'/?) ; 0 ( + E o 'n fl -

Where

The double hurdle model of equation (3) (i.e, the first hurdle) is a probit model that examines the probability that the i* farmer would make a decision to purchase improved wheat varieties. Equation (4) (i.e, the second hurdle) is a truncated regression model that examines the amount of purchased improved wheat varieties (Bhunbaneswar 2008). Therefore, the log-likelihood of the D-H model is the sum of the log-likelihood from a probit model and the truncated regression model (Adam 2012). Whether a tobit or a double hurdle model is more appropriate can be determined by separately running the tobit and the double hurdle models and then conducting a likelihood ratio test that compares ihe tobit with the sum of the log likelihood functions of the probit and truncated regression models (Greene 1993 cited in (Gebremedhin 2003).

LR = -2[LotLT - (LogLp + LogLTR)] ~ x\ (6)

Where: LogLT= log-likelihood for the Tobit model, LogLp= log-likelihood for the Probit model, LogLTR= log- likelihood for the Tobit model and k is the number of independent variables in the equations (Teklewold 2006)

Results and Discussion

Descriptive statistics The t-test and chi-square comparison of means of selected variables by participation in purchasing of improved wheat varieties for the surveyed households is presented (Table 1). Some of these characteristics were the explanatory variables of the estimated models are presented. The average amount of improved wheat seed purchases for the farm households who participated in purchasing was 130.84 kg (Table 1). Among all farm households surveyed, 80.6 % purchased improved wheat seed during 2012/13 cropping season. The farm size was about 2.2 ha for wheat seed purchasers. There was significant (P < 0.01) mean difference between the average farm size, and between wheat seed purchasers and non-wheat seed purchasers. Generally, the result depicted that the wheat purchaser "armer categories were distinguishable in terms of their wheat seed prices, non-farm income, hcusehold head, educational levels and adoption of wheat varieties. The two groups of smallholders suggested that farmers who participated in purchasing of wheat seed and not participated differed significantly in some proxies of socio-economic characteristics.

86 I

Table 5 Descriptive statistics variables used in estimations

Variables Unit Wheat seed Non-wheat t-stat purchasers purchasers „ (Chi-square) (615) (148) Dependent variable Wheat seed purchased Kg 130.84 0 Participation in wheat seed purchase Yes=1 No=0 1(80.6%) 0(19.4%) Independentvariables Family size Count 7.4780 7.0851 1.228 Age Years 44.84 45.91 -0.952 Farm size Ha 2.2 1.9 189* Livestock ownership TLU 8.0056 7.2731 1.49 Wheat seed price Birr 854.77 995.14 -4.55*** Expected wheat grain price Birr 707.0354 681.0228 0.833 Non-farm income Birr 16522.728 10967.152 2.4** Gender Yes=1 No=0 0.75 0.16 5.8** Read and write Yes=1 No=0 0.26 0.093 11.773*** Primary Yes=1 No=0 0.28 0.062 0.33 Secondary Yes=1 No=0 0.114 0.009 9.662*** High school Yes=1 No=Q 0.11 0.01 'J *** College/university Yes=1 No=0 0.012 0.001 0.562 Fertilizer adoption Yes=1 No=0 0.096 0.012 4.17**

Crop production patterns: Mixed farming characterizes the farming system of the study areas. The major crops grown in the study area are wheat, pulse, oil, tef, and maize. From the total sample respondents, the average cultivated farm size for wheat, pulse and oil crops was 1.34 ha, 0.54 ha, and 0.51 ha, respectively. Wheat crop is the major crop grown with average farm size of 1.25 ha. Oromia has the largest wheat farm size of 1.94 ha (Table 2). The ANOVA test shows that the average farm size for wheat, pulse, oil, teff and maize crops significantly differed among the three regions.

Cropping calendar farmers’ wheat seed demand and procurement Wheat cropping starts from Mid-June to Mid-July in major areas of Oromia region but the cropping starts mainly from Early-July to Late-July in Amhara region (Fig 1). While for SNNPR region, the wheat cropping calendar is mainly from Early-July to Mid-July. Monthly farmers’ wheat seed demand varied and farmers participated in purchasing wheat seeds differently in different months (Fig 2). The majority of farmers need wheat seed from January to July. However, they fulfill their wheat seed demand only in June and July.

Farmers’ subjective preferences on the characteristics of the technologies affected their adoption decisions. The surveyed farmers used improved wheat varieties as per their preference criteria (Fig 3). The major farmers’ preference criteria on the characteristics of wheat varieties were adaptability, disease resistance, yield, nutritional value and feed/straw quality. Fanners widely used Hawi, Danda’a, Digelu, Kakaba, Kubsa, Madawalabu, Pavon, Shina, Simba and Tuse wheat varieties.

i Table 6 Cropping pattern by region

Crop Region N Mean SD F Maize Oromia 25 0.3736 0.24986 12.299*** Amhara 141 0.6073 0.47706 SNNP 46 0.246 0.44806 Total 212 0.5013 0.47389 Oil Oromia 8 0.7188 0.52504 4.748** crops Amhara 18 0.4167 0.19174 SNNP 0 Total 26 0.5096 0.34986 Pulse Oromia 113 0.5926 0.48506 4.616**. Amhara 33 0.5385 0.37792 SNNP 21 0.2767 0.15651 Total 167 0.5422 0.44708 Te=f Oromia 105 0.5638 0.35187 12.936*** Amhara 131 0.5095 0.27847 SNNP 21 0.196 0.12598 Total 257 0.5061 0.31651 Wheat Oromia 294 1.9355 1.61044 102.745*** Amhara 171 0.7632 0.54957 SNNP 143 0.4167 0.22606 Total 608 1.2486 1.34404 Barley Oromia 128 0.6083 0.54793 17.170*** Amhara 48 0.2896 0.2691 SNNP 37 0.1928 0.09526 Total 213 0.4643 0.47931

LMcJum I Mdjdx I Earij aujusl I Lat* aug-ust my | March MidJun* EsJip j-Jjr Late julyr MuJaojusJ Feburary Cropping cil«nd«rfor B. Wheat Wheat seed monthly demand

Fig 1 Cropping calendar for bread wheat Fig 2 wheat seed monthly demand

88 Preference 0 Adapt ablity 10- 3 Disease resistance DEarly maturing f i Feed/straw value □ Lack of other seed □ Market demand Nutritional value S Yield a>05 Cr s oCJ

Wheat varieties

Fig 3 Wheat variety preferences criteria Econometrics model Model specification: To identify the model that best identifies the determinants of purchasing decision and volume of purchase of improved wheat varieties, a model specification test was conducted. Therefore, the D-H model was tested against the Tobit alternative using a likelihood-ratio test, ' he result for the model specification test is presented in Table. The LR Result rejected the null hypothesis that the Tobit model is appropriate and indicated that the estimated D-H model is preferred. The test statistic for log likelihood was 423.87 that exceed the critical chi-square value of 30.144 al 19 degrees of freedom and at a less than one percent level of significance in favor of the D- H model. This showed that the existence of two separate decision-making stages during the adoption process. This result provides an empirical result of farmers’ independent decisions making regarding the purchasing and volume of purchase of improved wheat in the study area.

Table 7 Test statistics of double hurdle and Tobit models

Test Statistics Probit Truncated regression Tobit regression Chi2() 156.38*** 101.27*** 2.79***

Log-L -226.29 -381.76 -819.97 Number of observation (N) 665 434 478 LR-statistics 423. 37*** 2(19)= 30.144 AIC (-_og-L+k)/N Source: model output. **, *** significant at 5% and 1%, respectively

Determinants of wheat seed demand: To identify the determinants of the decision to purchase wheat seed, a probit model (the first hurdle) was estimated (Table 4). The results revealed that the variables improved wheat variety use in the last five years. Location variables, read and write education level and farm size found significant in influencing the purchasing decision of wheat seed. The log likelihood for the fitted model was -226.29 and the y2 value of 156.4 indicated that all parameters were jointly significant at 1%. Improved wheat variety use in the last five years has been found to have a negative relationship with the decision lo purchase wheat seed implying that farmers who used improved wheat varieties in the last five years are unlikely lo purchase wheat seed than olhc fanners. Once farmers purchased a given wheal variety, they recycled it for more than al least one year. Looking at the marginal effects, farmers who have experiences in improved wheat varieties in the last five years, the probability of purchasing wheal seed was less by 0.32 compared to those who did not use improved wheat varieties in the last five years. Oromia farm households arc less likely to purchase improved seed in 2012/13 cropping season relative to SNNP region farm households. The likelihood of Oromia’s farm households (Oromia dummy) decreased by 0.15 relative

89 to the SNNP region farm households. For Amhara region farm households, the probability of purchasing improved wheat seed is lower by 0.20 relative to SNNP region farm households. Consequently, farm households from SNNP region were the most likely to purchase improved wheat seed controlling for other socioeconomics and demographic factors. Regarding the level of education, farmers who can read and write read had a negative effect on the probability of wheat seed purchase. According to the marginal effects, the probability of purchasing wheat seed decreased by 0.17 for farmers who read and write relative to farmers w ho cannot read and write. The effect of farm size was positive and significant suggesting that the larger farm size the farmer is the more likely the farmer is willing to purchase wheat seed. The probability of wheat seed purchasing increased by 0.041 as farm households farm size increased by one unit.

Table 4 Parameters and Estimated Marginal Effects of purchasing decision of wheat seed

Variable Double hurdle method Probit Marginal effect Coef. Z Coef. t Age -0.033061 -0.79 -0.00646 -0.8 Agesqur 0.0003807 0.88 7.44E-05 0.9 Extension Access 0.2337736 1.41 0.049709 1.3 Pulse rotation dummy -0.0425382 -0.29 -0.00828 -0.29 Gender 0.4589621 1.55 0.110339 1.3 Married dummy -0.2071268 -0.59 -0.03617 -0.67 Never maTied dummy -0.0799772 -0.09 -0.01636 -0.09 Family size -0.0088972 -0.37 -0.00174 -0.37 Dependercy ratio 0.0567892 1.1 0.01109 1.1 TLU -0.0062772 -0.33 -0.00123 -0.33 Improved A/heat use in last five years dummy -1.663392 -10.71*** -0.32484 -9.63*** Oromia dummy -0.7121715 -3.08*** -0.15359 -2.9*** Amhara dummy -0.8463951 -3.64*** -0.20128 -3.22*** Read and write dummy -0.768872 -2.13** -0.1683 -1.91* Primary dummy -0.516417 -1.43 -0.11114 -1.29 Secondary dummy 0.126352 0.3 0.023303 0.32 High school dummy -0.057078 -0.14 -0.01144 -0.13 Farm size 0.2086207 2.34** 0.040741 2.45** Fertilizer adoption 0.2111215 0.89 0.037235 0.99 _cons 4.008955 3.61*** --

Determinants of the amount of purchase of improved wheat seed: The determinants of the amount of purchase of improved wheat variety seed was estimated using the second double hurdle (Truncated regression) model. The empirical result from table of truncated regression model indicated that livestock ownership, regional variables, farm size and off-farm income had a significant effect on the quantity of improved wheat seed purchased. Once the decision to purchase improved wheat seed has been made, a 3% increase in the quantity of wheat seed purchase for every unit increase in the livestock ownership for a given farm household (Table 5; Table 6). The effect of regional variable on the amount of purchase of wheat seed was positive and significant for the two regional dummy variables. Among those fanners who purchased wheat seed, Oromia region farmers purchased 4% more amount of wheat seed relative to SNNP region farmers. Similarly, Amhara region farmers purchased 1% more amount of wheat seed. Thus, the results suggested that SNNP region farmers purchased less volume of wheat seed compared to framers in Oromia and Amhara regions. The plausible explanation was Oromia and Amhara regions had easy access to improved wheat seed and wheat seed market. Because, more wheat research centers and seed enterprises were found in the two regions.

90 The runcated model revealed that die amount of improved wheat seed purchased was positively (p< 0.01) affected by farm size i.e., empirical evidence of a positive impact of farm size on amount of wheat seed purchased. Farmers w'ho had one more unit of farm size purchased 5% more volume of wheat seed among farm households that buy wheat seed. The role of non-farm income had a positive and significant effect on the volume of wheat seed purchase. Higher non-farm income w'as associated with a higher volume of wheat seed purchase. Once the farmers made the decision to purchase wheat seed, ceteris paribus, a 1% increase in non-farm income will result in a 6% increase in the quantity of wheat seed purchase on average.

Table 5 Parameters and estimated marginal effects of purchasing decision of wheat seed

Truncated Variable Coef. t Coef. t Age -0.0239581 -1.42 -0.23454 -1.42 Agesqur 0.0002478 1.46 0.116201 1.46 Extension Access 0.0685047 0.7 0.0126 0.7 Pulse rotation dummy -0.0620692 -0.94 -0.00749 -0.94 Gender 0.0167851 0.12 0.003374 0.12 Married dummy -0.0120226 -0.51 -0.01866 -0.59 Never married dummy 0.011755 1.32 -0.00044 -1.26 Family size -0.0910006 -0.59 0.019843 1.31 Dependency ratio -0.441859 -1.26 -0.00396 -0.51 Livestock ownership 0.0165237 2.27** 0.028844 2.27** Improved wheat use in last five years -0.1172979 -1.21 -0.028 -1.21 dummy Oromia dummy 0.4587644 4.66*** 0.043343 4.67*** Amhara dummy 0.1546228 1.6* 0.008858 1.61* Read and write dummy -0.0675369 -0.33 -0.00479 -0.33 Primary dummy 0.1077046 0.52 0.008119 0.52 Secondary dummy 0.1117247 0.52 0.00365 0.52 High school dummy 0.0542945 0.25 0.001664 0.25 Farm size 0.1016566 2.79*** 0.050796 2.79*** Whea- seed price2012/13 -0.0000982 -0.53 -0.01839 -0.53 Expected Wheat grain price2013 -0.0000492 -0.53 -0.00766 -0.53 Fertilizer adoption -0.0533823 -0.56 -0.00126 -0.56 None 'arm income 0.0314148 2.14** 0.060304 2.14*** _cons 4.2.19 4.43641 4.2.20 7.76***

91 Table 6 Estimates of the double hurdle and Tobit models

Variable Double hurdle method Tobit Probit Truncated Coef. Z Coef. t Coef. t Age -0.033061 -0.79 -0 0239581 -1.42 -0.0450863 -1.31 Agesqur 0.0003807 0.88 0.0002478 1.46 0.0004691 1.34 Extension Access 0.2337736 1.41 0.0685047 0.7 0.2049885 1.06 Pulse rotation dummy -0.0425382 -0.29 -0.0620692 -0.94 -0.4563281 -3.05*** Gender 0.4589621 1.55 0.0167851 0.12 0.3347915 0.84 Married durimy -0.2071268 -0.59 -0 0120226 -0.51 -0.5007805 -1.23 Never married dummy -0.0799772 -0.09 0.011755 1.32 0.2854563 0.47 Family size -0.0088972 -0.37 -0.0910006 -0.59 0.0039387 0.21 Dependency ratio 0.0567892 1.1 -0.441859 -1.26 0.0392949 0.86 TLU -0.0062772 -0.33 0.0165237 2.27** 0.0271382 1.87* Improved wheat use in last -1.663392 -10.71*** -0.1172979 -1.21 -1.101621 -3.84*** five years djmmy Oromia dummy -0.7121715 -3.08*** 0.4587644 4.66*** 0.0467014 0.21 Amhara dummy -0.8463951 -3.64*** 0.1546228 1.6* -0.1453219 -0.66 Read and write dummy -0.768872 -2.13** -0.0675369 -0.33 -0.6087359 -2.08** Primary dummy -0.516417 -1.43 0.1077046 0.52 -0.2634484 -0.9 Secondary dummy 0.126352 0.3 0.1117247 0.52 0.0268408 0.09 High school dummy -0.057078 -0.14 0.0542945 0.25 -0.3867785 -1.21 Farm size 0.2086207 2.34** 0.1016566 2.79*** 0.199899 3.33***

Wheat seed price2012/13 - -0.0000982 -0.53 -0.0017545 -4.04*** Expected Wheat grain price - -0.0000492 -0.53 0.0001319 0.7 2013 Fertilizer adoption 0.2111215 0.89 -0.0533823 -0.56 -0.0770988 -0.41 None farm income 0.0314148 2.14** 0.0616994 1.67* _cons 4.008955 3.61*** 4.43641 7.76*** 7.1514 6.28*** Source: model output: *, “ and *** significant at 10%, 5% and 1%, respectively

Conclusions

The main objective was to identify the determinants of farmers demand for improved wheat varieties. Wheat fami household characteristics that contribute to the farmers’ wheat seed demand were identified in Oromia, Amhara and SNNP regions in Ethiopia. The double hurdle model empirically distinguished possible separate decisions on wheat seed marketing participation decision and quantity of wheat seed purchase decision. Region and farm si/.e were significant in explaining both the decision to participate in purchasing wheat seed and the level of wheat seed purchase. Improved wheat variety use in the last five years had a negative relationship with farm households’ decision to purchase wheat seed but was not important on the quantity of wheat seed purchased. Farm households who only read and write, out of the education levels identified, had a less likelihood to participate in purchasing wheat seed,. In addition, economic factors such as income and livestock ownership were among the significant determinants of wheat seed purchase demand. Calculated non-farm income elasticities, for those who purchased wheat seed, indicated that farmers’ wheat seed demand was sensitive to changes in non-farm income. Furthermore, study should be conducted using panel data in order to assess farmers’ seed demand variation over time.

References

Adam B, Belay K, Belaineh l and Tesfaye L. 2012. Determinants of Crop Commercial Orientation of Smallholder farm Households in Drought-prone areas of the Central Rift Valley of Ethiopia. Ethiopian Institute of Agricultural Research Alene D, Poonyth R and Flassan M. 2000. Determinants of Adoption and Intensity of use of Improved Maize Varieties in the Central Highlands of Ethiopia: A 1 obit Analysis. Agrekon 39:4. 92 Berhanu G and Swinton S. 2003. Investment in soil conservation in northern Ethiopia: the role ofland tenure security and public programs. Agricultural Economics 29:69-84. Bhunbaneswar D, Hugh B and Ross C. 2008. Determinants of Forestry Investment and Extent of Forestry Expansion By Smallholders in New Zealand. Review of Applied Economics 4 (l-2):65-76. Carroll J, Siobhan M and Carol N. 2005. An Econometric Analysis of Charitable Donations in the Republic of Ireland. The Economic and Social Review 36(3):229-249. C’ragg J. 1971. Some Statistical Models for Limited Dependent Variables with Application to the Demand for Durable Goods. Econometrica 39(5):829-844. CSA. 2010. Agricultural Sample Survey. FDRE Central Statistical Agency (CSA), Addis Ababa, Ethiopia. Dawit Alemu and David J. Spielman. 2006. The Ethiopian Seed System: Regulations, Institutions and Stakeholders. In: ESSP Policy Conference 2006 “Bridging, Balancing, and Scaling up: Advancing the Rural Growth Agenda in Ethiopia” 6-8 June 2006, Addis Ababa, Ethiopia Hailemariam T, Legesse D., A Y. and N D. 2006. Determinants of adoption of technology: a double-hudle approach. Livestock Research for Rural Development 18(3):???. Mignouna BD, Manyong MV, Mutabazi SD and Senkondo ME. 2011. Determinants of adopting imazapyr- resistant maize for Striga control in Western Kenya: A double hurdle Approach. Journal of Development and Agricultural Economics 3(11): 572-580. Reynolds A. 1990. Analyzing Fresh Vegetable Consumption From Household Survey. Southern Journal Of Agricultural Economics. ??? Teklewold, Dadi, Yami and Dana IT. 2006. Determinants of adoption of technology: a double-hudle approach. Livestock Research for Rural Development 18(3)???. TOBIN J. 1958. Estimation of relationships for limited dependent variables. Econometrica 26:26-36. Wodajo TB. (unspecified). A Double-Hurdle Model of Computer and Internet Use In American Households. Department of Economics, Western Michigan University.

93 T heme 6 Technology Dissemination, Up scaling and Knowledge Management

Enhancing Adoption of Improved Wheat Technologies, Innovations and Management through Dissemination, Up-Scaling and Knowledge Management

Mekonnen Mekuria1, Justa Katunzi2, Catherine KinyanjuP and Charles Aben4 1 Ministry of Agriculture, EAAPP, Ethiopia. Tanzania, 3Kenya, 4Uganda

Introduction

To promote improved technologies of wheat generated in East African Agricultural Productivity Project (EAAPP) countries to smallholder farmers/agro-pastoralists, the training and dissemination sub component of the project will use different strategics and mechanisms. Best bet technologies will be identified based on different Agro-ecological /.ones (AEZs) and extension package will be formulated. The dissemination mechanism should be diverse, proven, and acceptable by stakeholders. Valuable and acceptable results will be up scaled within EAAPP countries. Thus, the project will create partnerships with a range of institutions, both public and private to facilitate the dissemination, adoption and up scaling of wheat technologies.

The objectives of this component are the following:

• To d sseminate proven wheat technologies, information and management practices to farmers in EAAPP counties; • To improve productivity and increase the area under improved wheat varieties; • To enhance capacity of wheat growers to commercialize wheat production, agro processing and value addition; • To increase availability of improved wheat seed to wheat growers; and • To promote the agricultural machines and implements to actors along the wheat value chain • To improve knowledge, information management and sharing system

The expected outputs and outcomes will be:

• Increased number of farmers adopting improved wheat technologies and management practices; and regional technology uptake pathways; • Increased productivity, area under improved wheat, availability and access to seed of improved wheat varieties, and adoption of new handling and processing methods; • Increased stakeholder satisfaction with the technologies, innovations and uptake pathways; • Technologies disseminated in more than one EAAPP country increased • Increased capacity of wheat growers to commercialize wheat production and agro processing

94 Materials and Methods

The methods of Training and Dissemination (T&D) vary from country to country depending on the institutional arrangement and extension methods used. The approach that will be used to disseminate the technologies will be entirely based on participatory, client and market oriented with due emphasis to value chains. Some of the dissemination methods include preparation of manuals, guidelines, field days, exhibitions, leaflets, and posters and practical demonstration of technologies management, practices, and enhancing farmer innovations. Such activities will be implemented by intensive training of targeted communities. Moreover, demonstration plots on smallholder farmers’ fields and Farmers’ Training Centers (FTCs) will be used in order to achieve the strategy. In targeted areas, different working groups will be organized which includc men, women, and youth and linked with research, extension, and market. An exchange of experience on effective extension delivery methods will be conducted among countries as part of knowledge sharing in this subcomponent. Technologies accepted by farmers will also be multiplied for further scaling up activities by communities and other actors.

Imp ementation As the dissemination pathways differ among countries, different methods are applied to disseminate proven technologies. However, some initial activities like inventory of TIMPs of each country of wheal from varieties to prc and post harvest handling technologies were done by all participating count"ies. The efforts made by each country will be discussed here after i.e., for Ethiopia, Kenya, and Tanzania.

Ethiopia Dissenination activities were started by selecting effective approaches for effective implementation. Demonstration of TIMPs in farmer fields in the form of Fanner Research Extension Groups (FREGs), starting demonstrations on farmer training centers (FTCs), supporting farmer innovations by establishing farmer innovation grant system (FIGs), establishing linkage platforms of actors along the value chain and seed multiplication were the main dissemination pathways. Demonstrations of different varieties and management practices was conducted by establishing demonstrations at 704 FTCs and at farmer fields by establishing 4270 FREGs in selected 41 districts (Table 1).

Table ' FREGS and F I Cs established

Year Establishment of FREGs Establishment of FlC demonstrations Target Achievement Target Achievement 2010-11 (2003) 108 177 32 2011-12(2004) 984 832 177 141 2012-13(2005) 1050 1005 177 177 2013-14 (2006) 1050 1313 177 177 2014-15(2007) 1050 1012 177 177 Total 4270 704

Physical capacity building: Strengthening the district agricultural offices and FTCs were implemented to smooth the progress of demonstration and consequently 550 FTCs are equipped with furniture, demonstration materials, and pedal cycles. Transportation and officc equipment like motor bikes, computers, printers, and photocopiers were procured and distributed to project district agricultural offices lo improve the efficiency of the extension delivery system.

Human resource capacity building: Different types of short term and awareness creation trainings were given on different technologies of wheat to capacitate subject matter specialists, development agents, and farmers (Table 2). Trainings were conducted by researchers, seed experts, and experts from regional bureaus of agriculture.

95 Table 2 Number of trainees

Number of trainees Trainees 4.2.21 Total 4.2.22 Male Female Farmers 4.2.23 20,747 4.2.24 16,790 4.2.25 3,957 DAs and SMSs 4.2.26 4,709 4.2.27 3,833 4.2.28 876

Experience sharing visits: As experience sharing events will play a great role in popularizing proven and good performing technologies and also serve as a feedback methods for researchers. Performing these events will add up to the scaling up of proven technologies. Therefore, each year experience-sharing events were conducted at country regional level and district levels (Table 3).

Table 3 Number of participants in experience sharing events of different wheat technologies

Year Male Female Total 2010-11 (2003) 2011-12 (2004) 696 127 823 2012-13 (2005) 4,506 846 5,352 2013-14 (2006) 9,436 3,220 12,655 Total 14,638 4,193 18,830

The Wheat Regional Center of Excellence (WRCoE) and project management unit (PMU) hosted Tanzanian wheat extension and research specialists (1 male and 2 females) for visiting and experience sharing on the status and operation of research, T&D and seed multiplication activities. As a follow up of the visit made by experts, 17 Tanzanian farmers visited the WRCoE and surrounding wheat farmers and North Shewa zone of Amhara region. The delegates visited the structural setup of MoA, effectiveness of FREGs and FTC demonstrations for technology dissemination activities. Similarly, five delegates from Kenya also paid a visit to WRCoE and wheat growing parts of the country with similar objectives.

The reflections of visitors arc presented as follows:

• The nethod of operation is established up to the grass root level (village level) is interesting, strong and working effectively; • Development agents at village level are aware of the project and know what is expected from them in implementing activities of the project; • The research, extension and farmer linkage is well established and strong; • The extension delivery system using FREGs and other methods is proven to be effective; and • From the FTCs management - learned that three development agents trained in crop, livestock and natural resource management are assigned in each village (kebele) are working closely with local leaders who can consult farmers on the specified fields.

The visitors also gave their observation and comments

• Women participation in visited areas is minimum and should be considered; • In some areas late disbursement of funds is reported and needs improvement; • It is better to introduce improved machineries like combine harvesters to improve manual harvesting by farmers. These materials are available in Tanzania and can exchange those materials based on which ever mechanism effective; and • Insufficient staffing at district and village level is observ ed and needs to be considered.

96 Due to the effect of such interventions, it was possible to directly involve 228,107 farmers by FTC demonstrations, FREG demonstrations, FIGs, community based seed multiplication and value chain grain production activities. Considerable number of farmers has been also benefited by being engaged in experience sharing events and farmer-to-farmer seed exchanges. Considerable efforts have been made to facilitate linkage among research, extension, and farmers by creation of joint planning and reviewing forums, production of training materials and training of extension staff, delivery of seeds and other inputs. Creating linkage among value chain actors was done by creating platforms, analyzing the value chain and organizing gap filling value chain agri-business groups. Because of this effort, seed producers, wheat processing, value addition and marketing cooperatives were established. Since improving availability of improved seed is basic for dissemination and scaling up, efforts have been made to multiply certified seeds at fanner fields as seed business through the support of regional seed enteiprises and regional bureaus of agriculture. The regional seed enterprises serv ed as sources of basic seeds and potential buyers while regional bureaus of agriculture take care of the quality assurance and linking producers with markets. Therefore, more than 463 tons of improved varieties were distributed to farmers and 926.9 tons of certified seed was produced from 219.1 tons of basic seeds so far (Table 4).The produced Cl seeds were procured by regional seed enterprises, cooperatives, district agriculture offices, bureaus of agriculture or other farmers. Revolving seed system was also implemented at district level and framers were expected to return the amount they received for EAAPP interventions. This helps the district agricultural offices to lend the Cl seeds to other leedy farmers for demonstration purposes. Through the interventions made to broadly disseminate different technologies by different approaches, the average productivity of wheat increased from 1.5 t ha"1 (from base line survey) to 4.41 t ha'1 in project districts.

Table 4 Seed distributed and yield increments (2011-2014)

Year Seed distributed for Productivity different activities (t ha-1) 2010-11 (2003) 26.67 Starting year 2011-12 (2004) 26.43 3.97 2012-13 (2005) 166.03 3.36 2013-14 (2006) 244.7 4.41 Total 463.8

Strengthening the wheat seed sector: facilitate exchange of wheat seed among EAAP0 countries, capacity of seed production should be strengthened, a strong seed quality control facility and seed exchange policy should be in place. Therefore, EAAPP has planned different measures that can strengthen the regional seed system.

Strengthening seed quality assurance: As an effort to strengthen the seed quality control, it was planned to construct two seed laboratories at two regional states (Benshangul Gumuz and Gambella). It will support the effectiveness of seed quality control or the inspection system upon accomplishing of the civil work and equipping with laboratory.

To strengthen the existing regional laboratories, budget was transfencd to Amhara, Tigray, Oromia and SNNP regions. The procurement of laboratory materials for different regions is another EAAPP support for the future wheat R4D.

Strengthening quality seed production: EAAPP supported regional seed enterprises (Amhara, Oromia, SNNP, Tigray) and Ethiopian Seed Enterprise by procurement and planting of 3 seed c caning machines, and construction of shelter to previously procured seed cleaning machines (SNNPR and Sinana ARC) to strengthen quality seed production. Construction of four 100 ton capacity seed stores at four regional states (Amhara, Tigray, SNNP and Oromia) was planned. The procurement of seed cleaning machines is completed and the machines are transported to their destinations. The construction of shelter for seed cleaning machine at SNNPR is completed. Co funding by other regional seed enterprises (RSEs) and bureaus of agriculture (BoAs) will made to place the gleaning machines at the appropriate shelter before the closing date of the project.

Construction of Seed Stores: The construction of seed stores at Hosaena-SNNP, Dodola- Oromia, Chagni-Amhara and Mekele-Tigray will be finalized and handed over to the respective RSEs. Construction of lavatories at each seed store will be managed by regional states.

Kenya Wheat is the second most important cereal crop after maize in Kenya in terms of production and consumption and contributes significantly to the country's food security. Kenya produced 441,750 tons of wheat in 2012 while the national requirement was one million tons and the deficit of 56% was met through importation. Demand for wheat products is increasing by 4% annually and this provides a big opportunity for Kenyan farmers to invest in wheat production. Large scale farmers produce 80% of the wheat and the rest is produced by small scale farmers. Challenges faced by small scale wheat farmers includes emerging pests e.g., UG 99 and Russian aphid, inadequate access to certified seed, inadequate extension services, weak rescarch-extension linkages, and inadequate knowledge and information among small scale farmers. EAAPP is promoting wheat production in Kenya by improving access to certified seeds of improved varieties, strengthening research extension linkages, building extension and farmers’ capacity. The project focussed its efforts on small holder farmers who cons1.itute 80% of wheat farmers. Five wheat project areas and twelve fanner groups were selected in consulation with stakeholders. The project adopted a three pronged approach targctting increased production of certified seeds, promorting new wheal varieties and improving skills, knowledge and information. The project engaged liccnccd seed merchants through Memorandum of Understandings (MOUs) who were involved in contracting and training the groups on seed production. The groups were also trainned on business development and preparation of five year business plans. Through enhanced research extension collaboration, dissemination and training materials were developed, and joint training, field days and seed inspection activities were undertaken. The project interventions have resulted in production of 290 tons of certified seed through small holder seed producing groups. The new varieties namely ‘Robin’ and VEagle l O’ have been popularised and adopted by both large scale and small scale farmers resulting in 80% replacement of old varieties. The total area under new varieties is now 30,000 ha while wheat productivity has increased from 3.2 to 4.5 t ha ' in project areas. Demand for seed of the two varieties has also increased and the project has expanded hecatrage under seed production from 64 ha (2012) lo 278 ha (2014). Farmers groups have obtained information and improved skills in seed production, business development and have earned a total of USD 167,000 from certified seed production.

Tanzania Training and Dissemination sub-component need to have a regional focus and use participatory strategies and mechanisms to train extensionists, farmers in the latest innovations and to scale up application of technologies. It creates partnership with a range of institutions, both public and private to facilitate the dissemination and adoption of av ailable technologies. As part of the planning process, the project developed an annual training plan of short courses in “best bet” technologies for service providers, farmer organizations, and other stakeholders. The effectiveness of technology transfer was facilitated by linkages between research, extension, and farmers through participatory extension methodologies and communication support. Regional training and dissemination activities focused on: (i) ensuring increased availability and access of inlbnnaiion and improved technologies; (ii) strengthening the capacity of the agricultural advisory service providers; and farmers (iii) strengthening linkages between research, extension and end users; and (iv) establishing a regional platform for exchange/share of knowledge and experiences in scaling up agricultural innovations.

Activities implemented: Inventory of wheat technologies, development of wheat training manuals, capacity building/training of farmers and extension staff, study tours, exchange visit and establishment of demonstration plots were made. Other activities include monitoring, supervision and documentation of success stories.

98 Inventory of wheat technologies: The objective of inventory' activity was to identify the available wheat technologies developed and recommended in Tanzania so that they can be disseminated and used by wheat farmers and other stakeholders in the wheat value chain. Other objective was to identify capacity gaps between farmer and extension staff and help in the development of training program and dissemination strategics. Technologies identified included improved varieties, agronomic packages, and farm implements. Improved wheat varieties developed and currently recommended arc Juhudi, Chiriku, Lumbesa, Riziki cl and Riziki c2. Agronomic packages available in research institutions and other development partners were fertilizer types and rates of application. Fertilizers commonly used are TSP. DAP with the rate of 20-25 kg P ha'1 and 50 kg N ha'1, respectively, and foliar application of 1-2 kg blue coppcr ha'1. Other technologies were farm implement related technologies i.e., threshing machine, use of Magoyc ripper and seed planters.

Threshing machine: The machine has the capacity of threshing 8-10 bags/day (8 hrs). It is engine driven and fuel consumption is 5L/8 hours. It can be transported lo the wheat field by animal cart. T lis machine is suitable for smallholder farmers especially in the Southern highlands where farmers are not accessible to combine harvesters.

Magoye ripper: It is animal driven implement manufactured by local companies including SEAZ IN Mbcya and Nandra engineering in Moshi. It is a single tine implement used for opening furrows for planting wheat on either ploughed or un-ploughed fields

Planter (direct seeder): It is animal or tractor drawn implement manufactured locally by Elmi Farm implement in Hanang District Manyara Region. It is like a chisel plough with a seed box mounted on top. It also has a metering device for controlling seeds that are directed to the soil by pipes. 1 has effective width of 3 feet and 6 feet for animal and tractor drawn, respectively. Animal drawn planter has a capacity of 5 acres/day.

Use of green manure to improve soil fertility in wheat production: The use of green manure is an alternative means of enriching soil fertility, which is important to be sought in order to increase wheat production in general. Leguminous plant species as an alternative source of nitrogen in wheat production is currently used in Mbeya district and proved to increase wheat yields significantly. Five t ha'1 of Sesbania sesbania or Leucaena divosfolia were identified as alternative sources of Nitrogen fertilizers in wheat production.

Integrated Disease Management (IDM): Septoria leaf blotch, stem rust, leaf rust and yeilow/stripc rust are major wheat diseases. Septoria alone can cause up lo 50% of grain loss. Rusts can cause yield loss of up to 90% in different parts of Tanzania. Due to economic importance of these diseases to wheat industry, the research has developed integrated disease management of which early planting and single application of fungicide at heading stage was recommended.

Foliar fertilizers in wheat production: Poly feed is recommended as foliar fertilizer in wheat production at a rate of 300 g/20 liters of w^atcr equivalent to 1kg /acre. Two applications were recommended whereas first application (starter) was done at 3-5 leaf stage and second application (finisher) w'as applied at heading stage. Poly feed contains NPK 19:19:19 + important micronutrients.

Developing wheat training manual: Two training manual in Swahili version one on Good Agricultural Practices (GAPs) and sccond on Developing Agribusiness skills for wheat farmers were developed to be used by extension staff, fanners and other wheat stakeholders. About 100 copics of GAP training manual were printed distributed and the process for printing the second manual is underway. These manuals were developed in collaboration of extension staff, researchers, training officers, Small Scale Industry Development Organization (SIDO), mechanization specialist and NGOs. The manual were displayed for sharing during Nanc Nanc Agricultural show held in

99 Dodoma Tanzania in 2013, also were available during Agricultural show in Kenya held in September, 2014.

Capaci ty building to farmers and extension staff: Training of Trainers (ToT) on improved wheat technologies was conducted. A total of 115 farmers i.e 24 Female (F) and 91 Male (M) and 47 Extension staff (F 5, M 42) were trained on wheat husbandry, agribusiness skills, and safe use of chemicals, group dynamics and record keeping. Participants for training came from 31 villages in 9 Districts of Siha (Namwai, Engarcnairobi, Matadi v illages), Karatu (Kilimatembo, Rotya, Kambi ya simba. Kitete villages), Hanang (Gitting, Katesh, Nangwa,Endaswald), Mbeya (Ikhoho, Ihango, Ilomba, Haporoto, Muungano, Galijembe), Njombe (Makoga ,Igosi, Wangama villages), Sumbawanga Municipal (Mponda. Mawenzusi, Ulinji).Nkasi (Milundikwa, Kipande,) Sumbawanga (Mata, Jaigwani, Msandamuungano) Njombe Town Council (Utalingolo, Kisilo, Kilenzi).

Establishment of wheat farmer processor groups: 56 (F 12 and M 44) fanners and 23 (F 2 and M 21) extension staff from Siha, Karatu, Hanang, Njombe , Mbeya, Sumbawanga M, Nkasi and Sumbawanga districts participated in sensitization training on processor group formation held from 16-19 September 2013. The training organized to equip participants with agribusiness skills that involves profit margin analysis, fann budget, handling and managing of internal and external wheat markets, obtaining business licenses and use of mobile phones in obtaining agriculture information especially market information. Also, the training aimed at linking farmers/processors with other service providers such as NMB, SIDO, RUDI and MVIWATA that can assist with provision of loans, mobilizing farmer groups, processing and trading of wheat and wheat products.

Study tfisit in Ethiopia: A team of 8 extension staff one researcher and 11 farmers (3 F and 8 M) visited WRCoE in Ethiopia. The visit involved various areas including FREGs, farm research trials, processors, FTC and research centcrsStations.

Lesson learned: The knowledge /innovations and experience gained from WRCoE that could be practiced by Tanzanian wheat farmers and other stakeholders include the establishment and management of FREG and FTC, GAP practices. They appreciated the use of fertilizer in wheat fanns and the production of wheat seeds by farmers and farmer groups.

Establishing demonstration plots Thirteen wheat demonstration plots were established in 9 Districts of Nkasi (Kipande), S/wanga Munic (Ulinji and Mponda), S/wanga (Msandamuungano), Mbeya (Ikhoho), Njombe Tc (Kisilo), Njombe/Wanging’ombe (Igosi), Siha (Namwai and Engarcnairobi), Karatu (Kambi ya Simba , Rhotya and Kitete) and Hanang (Gitting). Three hundred six farmers participated to learn GAPs, the performance was encouraging. The demonstration involved the applications of skills in GAPs including local and those gained from WRCoE (tilling the land more than 2 times, fertilizer application, row planting and weed control).

Farmers’ field day Farmers field days were conducted from 2-9 June 2014 in Southern highlands consisting regions of Mbeya (Ikhoho village), Sumbawanga (Msanda Muungano village), Sumbawanga Municipal (Mpandi and Ulinji village), Nkasi (Kipande village), Njombe (Igosi village). Total number of farmers participated were 404 (149 F, 255 M) and 18 extension staffs (4 F and 14 M). The field day was canied out to enable farmers to see the performance of wheat when the recommended technologies are applied.

Exchange visit The exchange visit was conducted from 15-20 June 2014 for 40 farmers and 15 extension staffs from Southern Highlands Zone (Mbeya. Njombe, S/wanga and Nkasi districts) to Northern Zone in Hanang District. Participants had an opportunity to visit Fanners Innovation workshop. District Demonstration

100 Plots, Basutu Wheat Farm (private farm), farmer’s fields, and demonstration plots at Gitting village. The ob ective was to learn and share experience on wheat production. In addition, participants assessed what happened in wheat farming after having capacitated through training, study tour, and establishment of wheat demonstration plots.

Supervision and follow-up Technical backstopping was conducted to wheat farmers and extension staff who participated in various trainings and study tour to find out what happened on wheat field after equipping farmers and extension with various knowledge and skills on production and agribusiness. Through supervision and follow-up, is success stories were documented and the new emerging challenges on wheat sub-sectors were understood that the plan could address the existing challenges.

Up scaling of wheat technologies The best information and available technologies especially improved wheat varieties, farm implements, GAPs (fertilizers, insecticide, fungicides and herbicides) and processing techniques were disseminated to wheat stakeholders. The approaches were through establishment of demonstration plots, use of print media, use of radio programme eg Kilimo Bora cha Ngcino) was aired through TBC Taifa and Radio Maria, use of ICT including emails, mobile phones, farmers field days and agricultural shows or exhibitions.

A dissemination strategy/guideline lo speed up-scaling process was developed (In Swahili version) and was guided by five clearly defined strategic statements that were:

0 Increase the dissemination/utilization of technologies/interventions with high returns; • Promotion of technologies that reduces labour burden and gender sensitive; • Promote market linkages and value addition; • Increase dissemination pathways; and • Coordinate service providers both public and private particularly the technologies going to the farmers

Regional sharing The Rice RCoE’s outputs including inventories of technologies, communications strategy, and two training manuals are available at the RRCoE website and were available during Agricultural Show in Kenya.

The following achievements were recorded during the project period: a) Training of stakeholders: 539 wheat stakeholders including farmers were trained against project plan of 650 (Tabic 5).

Table 5 Number of trainees

Training Type No of participants No of participants Percentage planned for trained (%) training Short term training for extension staff and ToT 90 47 52.22 Short term training for farmers and ToT 160 115 71.88 Farmers study tour 30 16 53.33 Exchange visit 120 55 45.83 Demo plots 250 306 122.4 Tota 650 539 b) Adoption of GAP: The trained extension officers and farmers practiced the use of improved wheat technologies including land tilling more than 2 times, use of improve seeds, application of fertilizer and weed control techniques. Through those practices, the productivity of those farmers

101 increase from 0.9 t ha 1 to 3.5 t ha 1 by using variety Juhudi; it is an example of a young fanner Seva Sikabenga of Ikoho Village Mbeya District. c) Production of certified seed and QDS by farmers d) Farmer to farmer extension approach

Trained farmers were able to train other farmers eg Sikabenga was training a group of 30 young fellow farmers and she was used as a Model farmer to sensitize youth to engage themselves in agriculture.

Constraints or challenges • Many farmers especially those who haven't received training still are using poor technologies such is growing wheat without fertilizer, poor land preparation, use of low yielding varieties • Wheat is considered to be as men crop and there is a low participation of women in wheat production • Use of hand-threshing tools particularly by small fanners

Way forward

The plan for T&D is up scaling of the initiatives brought in by EAAPP in a sustainable manner through:

• Continuing with establishment of deinonstrations/FFS of recommended technologies/practices in wheat producing areas and in agricultural show grounds (Themi Arusha) and Uyole Mbeya; • Capacity building of farmers and other stakeholders within the wheat value chain; • Work closely with Zonal Information Extension Liaison Unit and other stakeholders for wider up- scaling of wheat technologies using various pathways within EAAPP countries; • Establish innovation platforms particularly engaging youth; • Sensitize stakeholders in collaboration with gender unit to understand the importance of women participation in wheat farming and have gender sensitive plans that will benefit all; and • Promote use of motorized threshers and small harvesting machines through mobilizing and strengthening farmer groups to access loans

102 Wheat Production Efficiency in Major Producing Areas of Ethiopia

Tolesa Alemu Ethiopian Institute of Agricultural Research, Kulumsa Agricultural Research Center, P.o.box 489, Asella, Ethiopia. I Abstract The study was carried out in Arsi Zone of Ethiopia to measure the level of production efficiency and identify sources of inefficiencies in wheat production in selected wheat producing districts. Cross- sectional data were collected from 381 randomly selected farm households in 2012/13 cropping season. A Cobb-Douglas Stochastic Frontier Production Function was employed. The average technical efficiency estimates for lowland, midland, and highland districts were 57, 82 and 78 %, respectively. The efficiency estimates showed that there was a potential to increase wheat yield in all agro-ecologies given the current state of technology and input levels. Wheat output elasticities associated with land, labor, chemical fertilizers and other inputs (seed and pesticides) were positive and significant in the lowlands whereas only land and chemical fertilizers were significant in mid and highland areas. Age of household head, livestock holding size, practice of crop rotation, access to credit and improved seed, and household size were significant factors that affected technical efficiency in wheat production. The findings imply that agricultural extension policy should consider agro-ecology and socio-economic contexts of farmers and access to inputs should be integral of the extension system.

Introduction

Wheat is one of the major food crops in Ethiopia. It is the second important cereal crop with annual production of about 3.43 million tons cultivated on area of 1.63 million hectares (CSA 2013). Based on CSA data of 2013, wheat occupied about 17% of the total cereal area with average national yield of 2.1 i t ha'1. This is the lowest yield compared to the world average of 4 t ha'1 (FAO 2009). The demand for wheat has been increasing due lo growing population, urbanization and the expansion of food processing industries in the country. It has become a very important staple in recent decades throughout the country, in part due to the massive food aid shipments of w'heat (Guush et al. 2011). The low yield has made the country unable lo meet the high demand and the country remains net impor.er despite its good potential for wheat production (Rashid 2010). Increasing yield and meeting the high demand has become the focus of government’s agricultural policy and extension activities. The major challenges facing agriculture are low' productivity, low use of improved farm inputs, and dependency on traditional farming and rainfall. As a result, low food crops productivity and food insecurity are prevalent. Given that wheat is a prominent food crop, it is of critical importance to identi y ways to enhance its productivity growth. Increasing productivity requires efficient utilization of farm inputs and adoption of improved agricultural technologies (Dorosh and Rashid 2013). There 'ore, if the country is to feed the rapidly growing population and meet the high demand, it needs to increase the production and productivity of wheat through efficient production and adoption of improved farm inputs in potential producing areas. One of the major wdieat producing areas in Ethiopia is Arsi Zone, which has different agro-ecologics with different wheat producing potentials. Wheat is one of the major food and cash crops for the farmers in the zone. Even though the Zone has high potential in w'heat production, low productivity of 2.40 t ha'1 is a major concern (CSA 2013). Improving efficiency in production is one of the factors for productivity enhancement (Coelli et al. 2005). Several studies on production efficiencies of selected crops and adoption of agricultural technologies have been conducted and most indicated inefficiency of farmers in wheat production (Kaleab and Brehanu 2011; Mcsay et al. 2013).

However, documented empirical study was limited on comparative analysis of wheat production efficiency among agro-ecologies and their sources of inefficiencies. Therefore, the main objectives of this s udy were to evaluate production efficiency and identify sources of efficiency in major wheat producing districts that represent the major wheat producing agro-ecological zones of the country.

103 Research Methodology

Study area Arsi Zone is found in the central part of the Oromia National Regional State of Ethiopia. The zone lies between 7°08’ 58” N to 8° 49’ 00” N latitude and 38° 41’ 55” E to 40° 43’ 56” E longitude. The area is divided into five agro-climatic zones mainly due to variation in altitude. It is dominantly characterized by moderately cool (40%) followed by cool (34%) annual temperatures. The mean annual temperature is between 20-25°c in the lowlands and 10-15°c in the central highlands. The zone receives a monthly mean rainfall of 85 mm and an annual mean rainfall of 1020 mm. The rainfall is well distributed both in amount and in seasons. These characteristics makes the zone good potential for production of various agricultural crops. Wheat is a major crop and it accounts for 42% of the total cereal area cultivated, with total output of 5.12 million quintals from 0.21 million hectares of cultivated land (CSA 2013). These characteristics make the zone the first potential area for wheat production in the country.

Sampling Three stages probability-sampling procedures was used for sample selection. In the first stage, major wheat producing highland, midland, and lowland districts were listed. The criteria for inclusion in the list included high potential for wheat production both at regional and national perspectives, availability of research and extension intervention programs embracing wheat producers, and distribution of newly released improved wheat varieties. From separate lists of highland, midland, and lowland districts, one district was randomly selected from each agro-ecology (highland, midland, and lowland). The selected districts were, namely, Lemu-Bilbilo from the highland, Hetosa from the midland, and Dodota from the lowland districts. In the second stage of the probability sampling, major wheat growing lower administrative divisions (kebeles) were listed. Taking into account the resources available, two kebeles were selected from each district with simple random sampling. In the final stage, wheat households were listed for each selected kebele. Sample households were selected by simple random sampling. The sample size was determined based on the formula given by Krejice and Morgan (1970), and allocation of sample size to each district and kebele was made proportionate to the size of wheat farm households’ population of each district and kebele. Accordingly, from a total of randomly selected 381 samples, 165 households were selected from the highland district (Lemu- Bilbilo), 133 households were selected from midland district (Hetosa) and 83 households were selected from the lowland (Dodota) district.

Data collection Data were collected from both primary and secondary sources. Cross-sectional data was collected from the survey of randomly selected sample farmers. For the primary data collection, specifically designed and pre-tested questionnaire based on the objective of the study, and trained data enumerators was used. Both quantitative and qualitative information were collected. The data collection included households’ demographic and socioeconomic characteristics (family sizes, age and sex structures, education), land holding (agricultural, grazing, wheat land), farm inputs utilization (seeds, fertilizers, herbicides and fungicides, labor utilization, credit, extension services), farm outputs, input and output prices, agronomic practices including crop rotation, wheat row planting and its inputs end output, etc. Secondary information on mean rainfall amounts and temperature were also collected. The survey was carried out in the months of May and June 2013.

Analytical method Stochastic Frontier Cobb-Douglas Production Function was employed to analyze efficiency and identify sources of inefficiencies in wheat production. STATA computer software version 11 was used for analysis. Stochastic Frontier Model was introduced by Aigner et al. (1977) and Meeusen and Van den Broeck (1977); and for n sample farms, it can be written as:

Yi=nxl-,m+s (1)

104 Where Y\ is wheat output o f the j" household's farm; i = (1, 2, 3,...... n) are sample household farms, X,j is the i'k input used by the f h household and [3 is a vector o f unknown parameters and s is composed o f error term which can be written as:

e = V; - Ui, ( 2 ) where v, is a symmetric random error which represents random variations, or random shocks in (he production oj the i household, outside the control of the farmer assumed independently and identically distributed as N(0. a 2). The error term m, is a one-sided non-negative variable which measures technical inefficiency of the i,h household. the extent to which observed output falls short of the potential output for a given technology and input levels. The method helps to decompose deviation of the actual observed wheat output from the estimated frontier nto random variations and inefficiency. Hence,

Ui = Zj 6 + w ; (3)

Where, Z. is a vector o f variables that explain inefficiency of i ,h household; 8 is a vector of unknown coefficients that are to be estimated in the model; and w, > —Zfi to ensure that u, > 0 (Battese and Coelli, 1995).

The technical efficiency of production of the ith farm in the data set, given the level of inputs, is defined by the conditional expectation evaluated at the maximum likelihood estimates of the parameters in the model, where the expected maximum value of Y is conditional on u =0. The measure of technical efficiency (TE) must have a value between zero and one. Following from equations (1) and (3), technical efficiency will be estimated as:

T E i = E (K i |u{,A 'i) / E (V 'i|u l- = 0 , ^ ) = e x p (_Ui) = e x p ( - Z £- 5 - ivt) (4)

Given the specifications of the stochastic frontier model expressed in equations (6), the stochastic frontier output (potential output) for the zth farm is the observed output divided by the technical efficiency, and TE; is given by:

Y,= £ = E(xt T l>=exp(Xi8+V|) (5)

The parametric specification of frontier in the Cobb-Douglas form for one output and n inputs is given by: to y i = Po + Z ”= 1 Pi lnXi + vt - Ui (6)

Where, y, is wheat output o f ilh household; x, represents vector o f farm inputs used as listed in Table 1; /?0 is intercept: is vector o f production function parameters to be estimated.

To iden .ify the factors in Eq. 6, a linear model was simultaneously estimated with the coefficients of farm inputs variables. Given the level of technical inefficiency derived from equation (5) and the above specified X vector inefficiency explanatory variables (Table 1), the coefficients of inefficicncy variables were simultaneously estimated along with the coefficients of input variables. The linear form of ihe equation is:

(In) efficiency = PX + 8 (7)

105 Table 1 Descriptions of variables used in efficiency analysis

Variables Descriptions In output (V) Natural logarithm of wheat yield in kg Inputs In area Natural logarithm of cultivated wheat land (ha) In labor Natural logarithm of labor (man-days*) In fert Natural logarithm of chemical fertilizers used in kg In other Natural logarithm of quantity of seed and pesticides in kg Inefficiency variables Age Age of household head in years Education Educational level of household head in number of grades completed Household size Household size in adult equivalent Livestock holding Livestock size of household in tropical livestock unit (TLU) Experience Fanning experience of household head in years Crops types Number of different types of crops cultivated Access tc seed Access to improved seed (1 if yes, 0 otherwise) Row planting Household practice of planting wheat in row (1 if yes, 0 otherwise) Credit Access to credit service (1 if yes, 0 otherwise) Crop rotation Household practice of crop rotation (1 if yes. 0 otherwise) Income Household annual income from off-farm activities in thousand birr *A man-day is equivalent to 8 working hours in the study area

Results and Discussion

Descriptive results Age aflects farm households’ experiences in farming processes. Older people usually have accumulated knowledge that helps them to improve their agricultural production. Hence, age can positively relate lo agricultural productivity. Older people sometimes are resistant to changes and unwilling to accept and test innovations and age can negatively relate to agricultural production efficiency. Younger famers, in some cases, are more active, better educated, have more access to information, and can adopt improved farm technologies. The average age, educational level completed and farming experience of sample household heads were 46.62, 4.8 and 25.7 years, respectively (Table 2). The average age of household head was the highest in midland district (48.8 years) followed by highland district (46.52 years). The mean age was significantly different (p<0.01) among districts; and average farming experience of household head was also significantly different (pO.Ol'i among the districts. Similarly, average educational status of household heads in terms of grade completed was 5.41, 4.56 and 4.69 for lowland, midland and highland districts, respectively, and the mean levels were not significantly different. However, Bartlett’s test for equal variance for education was significant (p<0.05) implying the variance of education was significantly different among agro-ecologies. Educated and experienced farmers are better able to process information and search for appropriate technologies to alleviate their production constraints. The assumption is that education enables the farmer to perceive, interpret, and respond to new information much faster than an illiterate farmer.

Table 2. Descriptive statistics of age, education and farming experience of household head

Household head Lowland Midland Highland Total SD CO CO characteristics c II n = 133 n = 165 N = 381 Age 43.34 48.80 46.52 46.62 11.27 Educational level 5.41 4.56 4.69 4.80 3.68 Farming experience 22.47 21 A l 25.58 25.70 11.09

Farmers are generally better able to assess the relevance of new farming practices with increased farming experiences, which often comes from their interactions with their neighbors and the outside

106 people. Because of their experience, they also tend to be better placed to acquire the needed skills to use the farm technologies compared with younger ones which result in a fanning experience usually improves efficiency in production. The average persons per household were about seven. Male and female average family sizes were 3.44 and 2.96, respectively (Table 3). The average family sizes for lowlanci. midland, and highland districts were 6.63, 6.08 and 6.53, respectively. Households in lowland district have largest household size while households in midland district have the smallest household size. However, there was non-significance difference among agro-ecologies.

Table 3 Household size of the study area

Sex of household Lowland Midland Highland Total SD Male 3.43 3.35 3.52 3.44 1.68 Female 3.18 2.73 3.03 2.96 1.49 Total 6.63 6.08 6.53 6.40 2.43

The household size in adult equivalent ranges from 0.5 to 11.5 with an average of 4.16 for the whole study areas (Table 4). The maximum and minimum household sizes in adult equivalent were 11.5 and 0.5, respectively. The average household sizes in adult equivalents were 4.18, 4.23, and 4.09 for lowland, midland, and highland districts, respectively. The test for mean difference of household size in adult equivalent shows non-significance difference among the agro-ecologies.

Table 4 Average household size in adult equivalent for the study area

District/Agro-ecology N Mean SD. Minimum Maximum Lowland 83 4.18 1.85 0.75 10.75 Midland 133 4.23 1.73 1.00 9.25 Highland 165 4.09 1.71 0.50 11.50 Total 381 4.16 1.74 0.50 11.50

The ma or agricultural crops produced in the study districts are wheat, barley, faba bean, field peas, tef, maize, and potato in 2012/13 cropping season. Some sampled households also planted onion. The average agricultural land allotted to wheat was 1.1 ha (Table 5). Barley (malt and food barley) is the second dominant crop with average area of 0.69 ha. However, malt barley alone accounted for average area of 0.8 ha. Other crops like tef, faba bean, and field peas were also essential crops with average planted land of 0.5, 0.3, and 0.3 ha, respectively. Wheat was generally the major crop with average planted land of about 1.6 ha in both lowland and midland districts. However, in highland district, malt and food barley were the major crops followed by wheat in tenns of land planted with these crops.

Table 5 A /erage area cultivated (ha) area and yield of major crops (t ha-1)

Crops Agro-ecology Lowland Midland Highland Total Area Yield Area Yield Area Yield Area Yield Wheat 1.6 1.56 1.6 3.09 0.5 2.48 1.1 2.49 Malt barley 0.5 1.2 0.0 0.0 0.8 2.92 0.8 2.91 Food barley 0.6 1.91 0.3 2.2 0.7 3.38 0.6 2.75 Faba bean 0.1 0.8 0.3 2.21 0.4 2.2 0.3 2.2 Field pea 0.2 1.02 0.2 1.46 0.4 1.91 0.3 1.65 Tef 0.6 0.87 0.2 0.9 0.3 1.2 0.5 0.8 Maize 0.3 1.12 0.2 1.91 0.5 4.0 0.2 1.7 Potato 0.1 2.4 0.2 11.55 0.3 9.66 0.2 10.54 Cabbage 0.0 0.0 0.2 15.22 0.1 18.1 0.1 16.66 Onion 0.2 5.47 0.3 8.81 0.0 0.0 0.2 7.44 Source: Computation from own data

107 Average yield was the highest for both malt barley (2.9 t ha 1) and food barley (2.75 t ha ‘). Faba bean and field pea’s yield per hectare were 2.2 and 1.65 tons, respectively. Relatively, the least yield was reported for tef (0.88 t ha'1). Farmers in highland district mainly planted malt barley. Food barley was planted in the three districts and it had average yield of about 1.9, 2.2, and 3.4 t ha 1 in lowland, midland, and highland agro-ecologies, respectively. The yield of barley was highest (3.38 t ha ’) in highland. However, yield of wheat was highest (3.1 t ha'1) in midland district. Generally, the yield of almost ail crops was relatively better in midland and highland than in lowland district of Dodota.

Econometric estimation results Production efficiency is one of the methods of farm performance measurement. It can be used for making comparisons among farms over time or across geographical regions or agro-ecologies. Technical efficiency analysis was used to measure farm households’ wheat production efficiency and make comparisons among the selected agro-ecological orientations of farm households. It measured the physical relationship between output and inputs used in wheat production at the given level of technology. A Cobb-Douglas Stochastic Frontier Model represented wheat production of sample famers. 3ecause, a series of preliminary likelihood ratio tests revealed that Cobb-Douglas stochastic frontier model best fit the data given a more flexible translog frontier model. The likelihood ratio test was used to identify the functional form of production function, which properly fit the data.

The test result revealed that Cobb-Douglas Stochastic Frontier model best fit the data compared to the more flexible translog frontier model. The existence of inefficiency factor was also tested by Wald test. The null hypothesis was that no systematic inefficiency in the distribution. However, the test result showed significant chi-squared statistic for the study areas (wald chi' = 52.1, prob > chi" = 0.0000) implying rejection of the null hypothesis. This means that there was systematic inefficiency in the distribution. Based on this, the parameters of Cobb-Douglas stochastic production model was associated with the maximum likelihood estimates. The coefficient of land, labor, fertilizer and other inputs of stochastic frontier model of Cobb-Douglas production function is shown in table 6.

The signs of all the slope coefficients of the production function were positive and significant. This implied that all inputs (land, labor, fertilizers, seed and pesticides) have turned out to be significant in determining wheat output i.e., wheat output was responsive to inputs utilization. The coefficients associated with the inputs measure the elasticity of output with respect to the respective inputs. The elasticity of chemical fertilizer was the highest (0.229) for the study area, indicating that there was relatively more proportionate change in output per hectare due to proportionate change in amount of fertilizer used . The sum of elasticities of the four inputs like land, labor, and fertilizers is 0.616 i.e., less than one which indicated wheat production function exhibited decreasing returns to scale. It means that proportionate increase in all inputs results in a less than proportionate increase in wheat output. The maximum likelihood estimate of the model for each agro-ecology showed that wheat output elasticities associated with land, labor, chemical fertilizers and other inputs such as seed and pesticides were positive and significant in lowland district. The elasticity of output due to other inputs was the highest (0.247) followed by elasticity of output due to labor (0.211) in lowland district. In midland district, elasticities of output due to land and fertilizers were positive and significant, with the highest being elasticity of output due to chemical fertilizers (0.163). Similarly, the elasticities associated with land and fertilizers were significant in highland district with negative elasticity of output due to land. This might be due to more suitability of the highland agro-ecology to barley production, and more land allocation to barley production was observed in the highland district. There was less cultivated wheat area in the highland (Table 5). It might have led to negative proportionate change of output due to proportionate change in the land area planted to wheat.

108 Table 6 Maximum likelihood estimates of Cobb-Douglas stochastic frontier model

Maximum Likelihood Estimates (Standard error) Variable Lowland Midland Highland Total (n = 83) (n= 133) (n = 165) N=381 Constant 4.316*** 7.045*** 6.955*** 5.688*** (0.726) (0.358) (.489) (0.418) In (land) 0.192*** 0.086** -0.586* 0.124*** (0.065) (0.037) (0.033) (0.026) In (labor) 0.211** 0.059 -0.011 0.132** (0.086) (0.046) (0.041) (0.041) In (fertilizers) 0.182** 0.163*** 0.163** 0.229*** (0.084) (0.046) (0.076) (0.048) n (other inputs) 0.247*** 0.024 0.040 0.131*** (0.069) (0.028) (0.041) (0.032) Wald 02 statistic 25.78*** 19.27*** 8.89* 52.01** Log-1 kelihood -6.195 37.3 15.269 -87.99 ***p< 0.01; **p< 0.05; *P< 0.10, and figures in parentheses are standard errors

The mean technical efficiency was 75% for the whole study area with minimum and maximum technical efficiency of about 24 and 94%, respectively. The mean technical efficiency estimates for lowland, midland, and highland districts were 57, 82, and 78%, respectively. Given the current state of tec inology and input levels, there was a scope of increasing wheat output by up to 25% on average. However, the scope of wheat output increment in lowland, midland, and highland districts were about 43, 18, and 22%, respectively. The technical efficiency ranges from 24.4 to 88.6% in the lowland, 51.6 to 94.4% in the midland, and 34.5 to 94.3% in the highland agro-ecologies. There was significant technical efficiency difference in wheat production among and between districts. The results are more or less consistent with what have been found by Arega and Zeller (2005), Arega (2006), Kaleab and Brehanu (2011) and Mesay et al. (2013) in their smallholder or commercial farms production efficiency analyses. This study indicated that improving technical efficiency in wheat production could improve productivity of wheat in all agro-ecologies.

The mean technical efficiency difference among agro-ecologies was tested using one way ANOVA. The response variable was technical efficiency scores and the factor variable was district or agro- ecology. Technical efficiency difference among districts was significant (P<0.001). Similarly, Bartlett’s test of equal variance gave significant chi-square statistic of 19.29 district or agro-ecology. Technical efficiency difference among districts was significant (P<0.001) implying rejection of the assuirption that variances are homogeneous. That is, the variances of the technical efficiency of agro ecologies were statistically different. Multiple comparison test of mean technical efficiency estimate was conducted between a pair of agro-ecologies using Bonferroni normalization. The test result gave signif cant difference (P<0.01) of mean technical efficiency estimate between any paired two agro ecologies.

Various households’ socioeconomic factors can affect the efficiency of wheat production in Ethiopia. These factors include age and educational level of household head, total livestock holding size, household size, and access to improved seed, and adoption of crop rotation and planting of different types of crops, row planting and income generation from other sources than crop cultivation. The existence of inefficiency factor was tested by Wald test. The null hypothesis for the test was "no systematic inefficiency in the distribution". However, the test result showed significant chi-square statistic, depicting rejection of the null hypothesis of no inefficiency. There was inefficiency factor in the distribution, and the famers were inefficient in wheat production.

Age of household head, livestock holding size, practice of crop rotation and access to credit (p<0.05), row planting (p<0.01), household size and access to improved wheat seed (p<0.10) were determinant factors of households’ wheat production efficiency. Age, livestock holding size, practice of crop

109 rotation and row planting were negatively related to inefficiency of wheat production. The test statistic gave significant chi-square at P < 0.01 level. The presence of inefficiency in production among agro-ecologies was significant (P <0.01) i.e., for lowland and midland districts (p<0.01) and for highland districts (p<0.1). The sources of inefficiencies were attributed to some significant socioeconomic variables (Table 7). For instance, household size, practice of crop rotation and row- planting and access to credit were found to be significant factors in affecting the inefficiency of wheat production in highland districts. Similarly, access to improved seed, farming experience of farm household head, and row planting were significant factors influencing the efficiency of wheat production in midland districts. However, age of household head was significant inefficiency factor in the lowland district. The difference in sources of inefficiency factor in lowland would be due the difference in the agro-ecological setting of the district; and production efficiency might be more related to the biophysical nature of the area than to the socioeconomic characteristics of households.

Table 7 Maximum likelihood estimates of inefficiency variables for each district

Lowland Midland Highland Inefficiency variables Coef St.err Coef St.err Coef St.err Age -0.222** 0.106 -0.034 0.039 0.003 0.041 Education 0.007 0.009 0.093 0.066 -0.018 0.067 Household size 0.022 0.047 -0.211 0.147 0.353** 0.162 Livestock holding -0.023 0.038 -0.024 0.064 -0.098 0.064 Crop numbers 0.030 0.038 -0.490 0.129 0.144 0.209 Seed 0.793 1.845 1.851*** 0.572 -0.430 0.674 Rotation -1.060 1.227 -0.584 0.483 -1.283*** 0.490 Income 0.069 0.119 0.028 0.029 -0.023 0.035 Experience 0.132 0.113 0.109*** 0.042 0.000 0.042 Row plantinc -0.411 1.020 -0.993** 0.432 -1.962*** 0.652 Credit -1.305 1.275 -0.119 0.435 1.090** 0.449 Constant 1.783 3.070 -4.633 1.536 -2.502* 1.469 sigma_v 0.205 0.025 0.134 0.016 0.169 0.019 WaldO2 24.85*** 18.70*** 8.39* Prob > D2 0.0001 0.0009 0.07 Log likelihood 1.580 37.42 18.25 *p < 0.10, **p < 0.05, ***p < 0.01

Age and educational level of household head, farming experience of household head, household access to improved seed and credit service, household and livestock holding sizes are the most identified sources of production inefficiencies by several studies (Arega and Zeller 2005; Arega 2006; Kaleab and Brehanu 2011 and Mesay et al. 2013). Moreover, this study indicates that crop rotation and row planting are the sources of technical inefficiency were found to be significant in determining the level of technical efficiency.

Conclusion and Recommendations

Infficiencies in wheat production needs attention as it provides significant source of enhancement in wheat output. The sources of inefficiencies were mainly related to farming experiences and skills, availability of labor and livestock for farm operations and other inputs, access to improved seed and credit as well as adoption of row planting and crop rotation farming practices. Though production efficiency was relatively higher in the midland agro-ecology, there was disparity of production efficiency among agro-ecologies and within agro-ecologies.

Improving production efficiency, adoption of wheat row planting and crop rotation can enhance wheat yield in the study areas. Increased use of farm inputs especially chemical fertilizers, improved seed and pesticides improves wheat yield in all agro-ecologies. Agro-ecological orientation and socioeconomic characteristics of households need to be considered in improving efficiency of

110 production. Households’ socioeconomic characteristics such as education, fanning experience, livestock holding, acccss lo improved seed, agricultural extension services and pesticides use should be considered to increase wheal yield.

References

Aigncr DJ, Lovell CAK and Schmidt P. 1977. Formulation and estimation of stochastic frontier production funct on models. Journal of Econometrics 6: 21-37. Arega D and Rashid HM. 2006. The efficiency of traditional and hybrid maize production in eastern Ethiopia: An extended efficiency decomposition approach. Journal o f African Economies 15(1 ):91-116. Arega Alene and Zeller M. 2005. Technology adoption and efficiency in multiple crops production in eastern Ethiopia: A comparison of parametric and non-parametric distance functions. International Institute of Trop cal Agriculture.??? Battese GE and Coelli TJ. 1995. A model for technical inefficiency effects in a stochastic frontier production function for pane data. Empirical Economics 20:325-332. CSA (Central Statistical Agency). 2013. Report on area and production of major crops, private peasant holdings, meher season, Addis Ababa. Coelli TJ, Prasada Rao DS, O’Donnell CJ and Battese GE. 2005. An introduction to efficiency and productivity analysis 2“li ed. Springer, New York. NY. Dorosh P and Rashid S. 2013. "Food and Agriculture in Ethiopia: Progress and Policy Challenges". University of Pennsylvania Press. Philadelphia, USA. FAO (food and Agriculture Organization of the United Nations). 2009. How to feed the world in 2050.(www.fao.org/fileadmin/templates/wsfs/does/expertpapcr/Howto Feed the World in 2050) Accessed November 12, 2012. Guush B, Zelekawork P. Kibrom T, and Seneshaw T. 2011. Food grain consumption and calorie intake patterns in Ethiopia. Ethiopia strategy support program II (ESSP II) ESSP II working paper No. 23, International Food Policy Research Institute. Kaleab Kebede and Berhanu Adenew. 2011. Analysis of technical efficiency: lessons and implications for wheat producing commercial farms in Ethiopia. Journal of Economics and Sustainable Development: 8(2). (http://www.iiste.org). Accessed on November 05, 2012. Krejice RV and Morgan DW. 1970. Determing sample size for research activities. Educational and Psychological Measurement 30:607-610. Meeusen W and van den Broeck J. 1977. Efficient Estimation from Cobb-Douglas Production Functions with Composed Error. International Economic Review 18:435-444. Mesay Yami, Tesfaye Solomon, Bedada Begna, Fekadu Fufa, Tolesa Alemu and Dawit Alemu, 2013. Source of technical inefficiency of smallholder wheat farmers in selected watwcrloggcd areas of Ethiopia: A translog production function approach. African Journal of Agricultural Research 8(29): 3930 - 3940. Rashid S. 2010. Staple Food Prices in Ethiopia. A paper prepared for the COMESA policy seminar on “V iriation in staple food prices: Causes, consequence, and policy options”, Maputo, Mozambique, 25-26 January 2010, under the African Agricultural Marketing Project (AAMP).

I ll Developing manually-operated single row precision Wheat-Cum-Fertilizer planter Ashenafi Tariku1 department of Agricultural Mechanization, Melkassa Research Center, Melkassa, Ethiopia

Abstract Manual wheat broadcasting is the oldest and common method that is widely practiced by our fanners for sowing wheat. This method of sowing crops has been giving poor yield because of the inconvenience for weeding and pest control, difficult for cultivation, irregular placement of seeds (overlaps and skips), poor control over depth of seed placement and difficulty for harvesting. The objective of this study was to apply the precision metering on wheat seeding to overcome seed damage, seed loss, and non-uniform distribution. A prototype of the precision metering device for wheat was developed. The performance of the device, including quality of feed index (QFT), multiple index (MULI), miss index (MISI), seed and fertilizer rate was investigated both under laboratory and field conditions. An accepted average QFI value of (81%) with MULI and MISI of 9.5% has been attained. Both the seed and fertilizer rate was within an acceptable range with seed breakage of less than 5%. Three different types of seeders were evaluated for depth uniformity, width, field capacity, average tumin j, time and plant emergence. The planter types were the developed precision planter, cultural method of planting, and Melkassa planter. The precision planter performed well under all parameters except the field capacity. Plant emergence ratios were evaluated by mean emergence time, emergence rate indexes, and emergence percentage. The best seed emergence ratio was obtained with the precision planter having a mean emergence time (MET) value of 2.2 days and percentage of emergence (PE) of 73.9% The study demonstrated that wheat could be seeded within an acceptable precision range by precision metering device.

Introduction

The production technique of weheat is traditional, which is laborious and productivity is low. Its productivity in general is effected not only by direct inputs like improved seeds, fertilizer and agronomic practices, but also by physical science based inputs like improved agricultural implements, which create favorable environment for the crop to flourish and create unfavorable environment for disease and pests. Besides, other implements, crop establishing technologies like planters which deposit the seed to the proper depth and avoid soil crusting enhance crop germination and timely emergence of the crop, which will make the crop escape the critical period to be attacked by pests and disease. Under current farmers’ practices, the wheat seed is simply planted on the ploughed land or by using the traditional methods, which includes broadcasting manually, opening furrows by a country plough and dropping seeds by hand and finally the seeds are covered by using a "maresha" (Teklu 2010).

In manual planting (broadcasting), seeds sown per hill are more than the prescribed amount. This results to over population and consequently reduce yield due to insect build-up and nutrients and sunlight competition. One factor affecting the quality of the crop in mechanized crop production systems is the planting operation. The more precise the planting operation, the better the quality of crop harvested. Precision planting reduces seed scattering and excessive use of seeds due to uniform distribution of seeds and by preventing seed from bouncing in the furrow, which facilitates drill calibration on the basis of the number of seeds to be placed along a unit length of the row. Uniform germination and growth of plants makes the subsequent operations, such as weeding and harvesting easy with lsss costs. Therefore, these problems need attention and one way of tackling this problem is using different kinds of planter. The wheat planters available in the market are imported (needs limited foreign currency), designed to operate in large farms (land holding of most farmers is about half a hectare), expensive (most farmers income is low) and not suited to local conditions. Therefore, the use of big wheat planter under Ethiopia conditions might not economically feasible for most of the farmers with fragmented farms. The adoption of row planting is likely to be slowed by its complex indicated challenges. Hence, the present study was undertaken to develop a precision-type wheat-

112 cum-ferlilizer planter, simple that can be fabricated at local machine shops, light in weight, that can be manipulated by women, usable for other cereals, and is adaptable to local farm size. It will be fabricated using indigenous materials and off thc-shelf standard components. The precision metering device was designed specially to meet the requirements of sowing wheat crop within the recommended seed rate to increase the productivity and decrease the production cost of wheat cultivation.

Materials and Methods

Physical properties of wheat seeds Fifty grains of wheat kernels were sampled for measuring the three principal dimensions of a grain: length (L), width (W) and thickness (T). The dimensions were used to compute the physical property attributes of the seed like geometric mean diameter, sphericity and volume that arc helpful in designing the planter. The results of all physical seed properties are given below (Table 1) required for the design of the planter.

Table 1 Dhysical characteristics of wheat kernels, variety Hiller

SEof 95% Physica properties Min Max Mean mean confidence limit Length (L) (mm) 4.1 6.68 5.86 0.0604 5.86 ±0.12 Width (W) (mm) 2.15 3.29 2.79 0.0365 2.79 ± 0.07 Thickness (T) (mm) 1.95 2.97 2.39 0.0315 2.39 ± 0.06 Sphercity (%) 0.52 0.67 0.58 4.410E-03 0.58 ± 0.009 1000 seed mass (g) 30 40 32 0.1333 6.57 ±0.26 Geometric mean diameter mm 2.76 3.85 3.39 0.0350 3.39 ± 0.07 Volume (cm3) 0.022 0.055 0.038 8.184E-04 0.038 ± 0.0069

Product design and development

Power developed by the operator: According to Campbell et al. (1990), the power of useful work done by human being is given by:

HP = 0.35 - 0.092 logt

Where t = Time in minutes

Assuming the manually operated planter works for 3-4 continuous hours, the power developed by the operator would be 0.1- 0.13 hp. By taking the operating speed of the machine be 0.8 m s'1, the force developed by an average human worker is 10.5 kgf.

Size of planter: The size of the planter can be calculated as:

Z = D/d

Where Z = number o f furrow openers in the planter; D = draft o f the planter; d = draft o f each row, kgf (is 12- 15 kgf for shallow depth)

Therefore, taking the maximum draft, Z is found to be 1 (single row seed-cum fertilizer planter) with a working width of 20 cm.

113 Design of seed and fertilizer hoppers: (a) seed hopper Normally a trapezoidal shape is recommended for free flow of seed and fertilizer in hopper bottoms. But due to limitation on the fabrication process, we arc forced to make some modification on the box by making it rectangular at the upper part and a bit inclined (> angle of repose of the wheal seed) at the bottom rest. Volume of seed hopper is given by:

V b 4 l . l V s

Where Vb = volume o f seed hopper, cm*; Vs = volume o f seed, cmJ

For light weight and easy operation, let 2 kg seed is filled in the hopper at a time. Now, it is possible to use the above equations to calculate the volume of the box

v l . i «2ooofl 1%7 8cm 3 = 2000 cm3 1.118 g / c n i 3 bulk density of wheat is 1.118 g /cn r.

Fig 1 Seed hopper

The hopper is assumed to be square in size at lop and the dimensions have been given by trial and error method so that its volume can be calculated to be 2171.25 cm'. Sincc, the volume of designed seed hopper is 2171.25 cm3, which is higher than the theoretical volume (2000 cm’) then, the designed dimensions of hopper are correct.

Fertill2:er hopper: For easy construction, balanced operation of the seed-cum-fertilizer and symmetry in size, the same size, and shape of fertilizer box was selected as that of the seed box.

Design of seed and fertilizer metering mechanism Determination of seed and fertilizer plates diameter: (a) Seed plate - Since the planter was designed to be precise, one seed should be in one hole. The diameter of the ground wheel was 0.4m and the speed ratio between the two sprockets is 1:3. Therefore, the diameter of the seed plate was calculated as follows:

d„, = —Wd = — 40 = 13.33 cm p R 3

Hence, v/d =diameter of the ground wheel; dp = Diameter of the seed plate and R = Speed ratio between sprockets, 1:3

114 Taking the recommended wheat seed required for a hectare of land, 1.43 cm intra row spacing was calculated; then the plate was designed to have 133.3 mm diameter with 29 equidisrant cylindrical holes (Fig 2). The plate was made to have a depth of 27 mm in order to accommodate kernels. The diameter of the seed hole was 1.6 mm based on < 50% size of the geometric mean diameter for wheat seed. Depending upon the small diameter of seed hole along with the influence of gravity, the singulation could be achieved.

Fig 2 The designed seed plates

(b) Fertilizer plate: The diameter of the fertilizer plate will be the same, 13.33cm, for balanced operation and with different number and size of holes. The recommended DAP requirement for a hectare of land is 100 Kg, which means 1.67 g m '1. The average weight of 10 fertilizer grains found to be 0.3: gm and the hole having 6 mm depth and 8 mm diameter can contain 14 fertilizer grains. Therefore, the number of holes on the plate will be 2 and they will be made on the periphery of the plate ISO" apart (Fig 3).

Fig 3 The designed fertilizer plates

Power transmission system design Since he power transmitted in the planter is very low, for power transmission a 40 cm diameter wheel fitted with sprocket of 48 teeth wras selected. Another sprocket of 16 teeth was used for seed and fertilizer metering shaft to achieve transmission ratio of 1:3.

Chain force: The total load on the driving side of the chain was calculated by using the methods from Sharira and Mukesh (2010) and was found to be 199.73 N. A chain number of 25 had a tensile strength of 3470 N (ANSI TEST).

Chain Length (Lc): The length of the chain is given by:

("jLjya)2 Lp = 2CP + (Nt + N2)/2 + c p Lc=Lp* p Stepl. Cp = y = 6 7 . 7 Lp = 167.81

115 Where Lp = Chain length in pitches; Cp - center lo center distance b/n sprockets; Nj = number of teeth on the driven sprocket, 16; N 2 = number o f teeth on the driving sprocket, 48; P = chain pitch, 6.35 mm

Step 2. The chain length, Lc

We should take the pitches in whole number, 168 pitches Lc = 168 * 6.35 = 1066.8 mm; Cc = center to center distance between sprockets

Step 3. The corrected center to center distance of the chain sprocket was calculated to be 430 mm (Fig 4).

Shaft design Main shaft: it is the shaft where both the seed and fertilizer plates are attached to; here the analysis was done to select the appropriate shaft diameter and bearing sizes that overcome the load applied to it. After calculating all the forces, on both the X and Y axis of the shaft, the maximum bending moment was determined to be 17.47 N.m and the diameter of the shaft was calculated to be 17.4 mm. Therefore, from standard table (Mechanical design, Shigly 9th edition), single row 03 serious (medium sizes) bearings was selected having bore size of 17 mm and outside diameter of 40 mm.

Wheel shaft design: The same procedure like that of the main shaft has been followed and resulted with a diameter of 1.84 cm.

Design of handle A standard light weight M.S. 30 mm outside diameter conduit pipe was used for handle of the tool carrier. Length of handle was calculated based on average standing elbow height of the operators (Fig 5)-

Average standing elbow height of operator is about 1 m = 100 cm; Distance of wheel center from the operator in operating condition = 120 cm

Hence, angle of inclination (Oh) of the handle with the horizontal:

0h = tan‘(0.667) = 33.7°

Now, it is possible to calculate the length of the pipe (Lh) needed.

Lh = 80/0.5545 = 144.27 cm or 145 cm (for sake of simplicity)

Therefore, in order to accommodate most of the operators, a 27.5 mm outer diameter conduit pipe of 2*145 = 290 cm long was used for handle whose operating height can be adjusted from 95 cm to 105 cm from the ground. Fig 5 Handle of the precision planter

Construction of the planter: The materials used in constructing the planter were those that can easily be found locally in Ethiopian cities and towns. The design was as simple as possible; it was produced with available present day intermediate technology available at Melkassa ARC. The planter has one ground engaging wheels ($=40 cm) which produces the necessary force to drive the seed and fertilizer plates through chain-sprocket drive. The total weight of the planter at full loads of the seed and fertilizer hoppers was made to be 33 kg in order to create enough traction between the wheels and the ground (Fig 6a, 6b, 6c and 7).

Fig 6a Top view of the planter 4.2.29 Fig 6b Side view of the 4.2.30 Fig 6c 3D plot of the planter planter

Fig 7 Dictures of the designed planter

117 Experimental design Laboratory procedures: The germination rale of ihe seeds was tested with 10 randomly selected seeds after being placed in a Petri-dish and awaited for a week until most of the seeds start germinating. The seeds which had passed through the seed metering mechanism were also subjected for germination tests so as to know if there were internal damage which may be caused by the metering plate (change in its velocity) and the weights of the seeds in the hopper. The seeding and fertilizer aoplication rates of the planter were measured at three levels of the hoppers loads; quarter, half and full loads, so as to know if there were significant variations on rating and seed damage (external). The test was replicated there times on 25 m length ground (with negligible gradient and moderate moisture contcnt) by pulling the planter manually with average walking speed of 0.83 m s'1. Weight balance (0.1 g accuracy) and stop-watch were the instruments used for measurement.

A mat of saw dust has been used as a test ground so as the planter drops the seed; it will act as a coition preventing rolling or bouncing. The mean spacing and the standard deviation of the seed spacing arc: useful but do not completely characterize the distribution of plant spacing for single seed planters. For this reason, different kind parameter was discussed below as the distance between plants within a row is influenced by a number of factors including multiple seeds dropped at the same time, failure of a seed to be dropped, failure of seeds to emerge, and the variability around the drop point (Kachman and Smith, 1995).

Theoretical spacing is the targeted distance between plants, assuming no skips, no multiples, and no variability in seed drop. The existing stand was assessed relative to what the plant to plant spacing should be. Four measures of plant-spacing accuracy are discussed below.

• Multiples index (D (doubles, triples, etc.)) is a percent of spacings that are less than or equal half to the theoretical spacing;, • Miss index (M (skips)) is the percentage of spacings greater than 1.5 times the theoretical spacing. These skips could be due to the failure of the planter to drop a seed or the failure of a seed to produce a seedling. Smaller values of M indicate better performance than larger values; • Quality of feed index (A) is the percentage of spacings that are more than half but no more than 1.5 times the theoretical spacings. This is a measure of how close the spacings are to the theoretical spacing. It is another way to look at information in the other two indices since: 100 - (D + M) = QFI. Larger values of QF1 indicate better performance than smaller values; and • Precision (C) is a measure of the variability in plant spacing after removing the variability due to skips and multiples. Precision is similar to a coefficient o f variation lor the spacings that are classified as singles. It is not affected by outliers, multiples or skips. A practical upper limit is 29%. Smaller values o f C indicate better performance than larger values.

Field te»t procedures: The field test was conducted to determine the performance of planter under the actual field conditions at Melkassa Agricultural Research Center, Central Rift Valley of Ethiopia. The physical properties the soil were determined (Table 2) and three tratments were viz. Cultural method of planting plow (CP), Melkassa planter (MP) and developed precision planter (AP) were evaluated (Table 3). Actual field evaluation has been carried out by using a standard test procedure using RCBD design with three replications. The actual field tests of the machine were conducted on 7 x 10 m plot.

Table 2 Soil physical properties for the 0 to 0.1 m depth range

Physical property Value Bulk density (Mg nrr3) 1.2 Porosity (%) 53.6 Moisture contsnt (% d.b.) 11.46 Textural class Sandy loam

118 Table 3 The treatments compared

Treatments Tillage frequency Method of seed sowing Cultu'al method of planting plow (CP) Three times Manual row planting Melkassa planter (MP) Three times The machine itself The ceveloped precision planter (AP) Three times The machine itself

Seed Emergence: Number of seedlings emerged each day were counted for 15 days starting from [he first day of emergence. Mean emergence time (MET), emergence rate indexes (ERI), and percentage of emergence (PE) were determined using the following equations (Karayel and Ozmerzi, 2002);

MET = ( Nj T1 + N2 T2 + ...NnTn) / ( N, + N2 + ...Nn) ER = S te/M E T PE = 100 * Ste/n where, Nj...n = number o f seedlings emerging since the time o f previous count; T]...„ = number of days after sowing; S!e = number o f total emerged seedlings per meter; N = number o f seeds sown per met

Results and Discussion

Performance test of the planter Parameters including the seeding rate, percentage of broken of both the seed and the fertilizer were determined. The level of fill has no significant effect in the seeding rate, which was probably due to the size of the hopper as it was loo small (2 kg of seed at full load) to notice any sizable difference (Tabic 4). The full hopper delivered an average of 6.18 kg hr'1 which was closer to the targeted 6.72 kg hr than 1/4 and 1/2 filled seed hoppers which delivered an average of 6.03 and 5.82 kg hr'1, respectively (Table 5). The interest was in getting which level of seed fill in the hopper gave a better result, since there were different parameters to measure seed uniformity. By checking the percentage of the broken once again, the level of feed has no significant effect on the percentage of breakage to both seed and fertilizer. It was found that all had percentage of the broken seed not exceeding 5%. There was no significant difference in the fertilizer rate between treatments. The full and 1/4 filled hopper delivered an average of 5.48 kg hr'1 which is closer to the targeted 5.01 kg hr'1 than 1/2 filled fertilizer hoppers which delivered an average of 5.92 kg hr'1. Bucket type garlic planter had less value of 9.98 to 15.76% than the new vertical plate planter which had the value of 11.19 to 39.45%.

Table 4 Analysis of variance [P values) of seed, fertilizer rating and percentage of breakage values

WF seeding fertilizer Average WOF seeding rate Average percentage rate rate percentage of Source (kg h r 1) of the broken seed (kg h r1) (kg h r1) broken fertilizer Level of hopper 0.1333 0.1968 0.6937 0.4443 0.1415

Table 5 Means comparisons of seed, fertilizer rating and percentage of breakage values

WF seeding Fertilizer Hopper level of WOF seeding Average percentage rate rate Average percentage fill rate (kg h r1) of broken seed ( kg hr*1) ( kg h r1) of broken fertilizer full 6.18a 2.58^ 6.133 5 .48a 2 .3 0 a 1/2 full 5.82a 2.36a 6.03a 5 .92a 1.06a 1/4 full 6.03^ 1.31a 6 .03a 5 .48a 1.77a Means within each column followed by the same letter are not significantly different (p<0.05); WF, WOF are tests wiih and without fertilizer, respectively.

119 Laboratory seed spacing and uniformity test The seed spacing in the laboratory had been conducted on the saw dust was measured manually from 30-meter length; every 10 meter represents one replication. The data obtained were statistically analyzed to determine the performance of the planter (Statistix 8 Version 2.0). The statistical analysis revealed that the highest QFI (80.96%) was observed whereas the corresponding values of MULI and MISI at these levels were 9.52 % .The highest MULI was (14.3%), while the lowest MULI (9.52%) . MISI has a uniform value of 9.52% with a CV of 0 (Table 6).

• C lies within the range of 12.98 -15.54 and is well under the practical upper limit of 29%. Smaller values of C indicate better performance than larger values; and • QFI is very much closer to the target 82.3%, which is a requirement for precision seeding.

The precision metering device for wheat could achieve a uniform distribution and seed rates estimated at 56 KPM at 20 cm row spacing and within a reasonable percentages of QFI.

Table 6 Stati stical data of laboratory test on seed spacing and uniformity

Trial no Max Min Mean (mm) CV (%) SD MULI (%) 14.30 9.52 11.43 22.90 2.62 MISI (%) 9.52 9.52 9.52 0.00 0.00 QFI (%) 80.96 76.18 79.05 3.35 2.62 Precision {2) (%) 15.54 12.98 14.28 6.61 0.94 MULI is the multiple index, MISI is the miss index, QTFI is the quality of feed index, and PREC is the precision

Field seed spacing and uniformity tests The highest QFI was (73.4%) whereas the corresponding values of MULI and MISI at these levels were 13.3% (Table 7). The highest MULI was (20%), while the lowest MULI was (10%) For MISI. the highest value 20% was observed, while the lowest value was 13.3%. The precision was in the accepted range with a value of 13.54 -20.3, but the QFI was below the minimum requirement of 82.3%. One of the reasons was that the figures were calculated based on results obtained from laboratory test. A few tests used performance measures involving distance between seeds sown into soil (Panning, 1997).

Table 7 Statistical data of field testing on seed spacing and uniformity

Trial no Max Min Mean (mm) CV (%) SD MULI (%) 20.00 10.00 15.34 24.84 3.81 MISI (%) 20.00 13.3 15.32 19.59 3.00 QFI (%) 73.40 63.30 69.34 5.31 3.68 Precision C) (%) 20.30 13.54 17.67 17.14 3.03

The other reason may probably was due to the vibration of the planter as it travels along the unfavorable condition of the farm field which was ploughed by maresha attached mould board plough. The specification of planters used for the three treatments of wheat seeding in the field experiments is given in Table 8.

120 Table 8 Specification of planters used for wheat seeding

Planter type Specification Precision Cultural Melkassa

Dimens on width x length x height 0.6 x 1.2 x 1 - 1.2 x 1.3 x 1 Numbe' of rows 1 1 6 Spacing between rows (m) 20 20 20 Seed metering device Seed cell plate manual Furrow opener type Disc coulter Traditional Disc coulter maresha

Field performance test Seed spacing uniformity, depth uniformity, and seed emergence ratios were evaluated considering all the parameters. Planter effects on evaluation parameters of field capacity, furrow width, sowing depth and the Av.turning time per row were statistically important (p<0.05). The best field capacity (12.76 hr ha ') was obtained with the melkassa planter, while the field capacity for the other two planters was similar to one another (Table 9). The melkassa planter has a working width of 1.2m and six row planters, which gave it the capacity to cover a larger area at a time. The largest sowing width was obtained with the cultural method of planting, while the mean sowing width for the other two planters was similar to one another. A possible reason was that the cultural method of planting uses a marcsha to open the soil and a "digir" to widen the furrow, which was greater than the thickness of the rolling disc fu t o w openers. The largest sowing depth was obtained with the cultural method of planting, while the mean sowing depth for the other two planters was similar to one another. For the 6 cm target value of sowing depth, the precision planter produced the best mean sowing depth closely followed by the Melkassa planter. The furrow opener on these planters differed from the cultural one (uses a maresha), as these planters uses rolling discs which are good for a shallow depth.

The un form sowing depth was obtained with the precision planter. The standard deviation and CV of the depth of the precision planter were 2.8 mm and 4.94%, respectively. The actual planting depths for the precision and melkassa planters were 2.8%, 14 less than the target depth, respectively. The planting depth for the cultural method was 44.5% greater than the target seeding depth (Table 9). The precision planter achieved the least mean average turning time per row, while the mean average turning times per row for the other two planters were similar to one another. Hence, the precision planter was operated by a human being and had a small working width as to the others uses animal and larger working width (the case of the melkassa planter) which will takc-up more time to turn.

121 Table 9 Analysis of variance (P values) and means comparisons of field capacity, furrow width and depth and fertilizer, average turning time per row values

Arlalysis of Var iance Mean comparisons

Qrsiirr>r» 1 CIaU * • ja r c — k' ~ n : — i .. k ...... r ...... r ______:___ i u a ______. . . _____ r IC1MU7I [ UI IUW VVIUll 1 ruiiuw ucpui nveiaye lum m y capacity width depth time per row (kg h r1) (cm) (cm) time per row(s)

Planter 0 . 0 0 0 2 0.0016 0.0013 0 . 0 0 0 0 Cultural 3 4 . 3 8 a 1 6 . 3 3 ^ 8.67a 12.74t>

Melkassa 1 2 . 7 6 b 8 . 5 b 5 . 1 7 b 1 4 . 9 7 3

Precision 3 5 . 4 3 a 6 . 6 6 ^ 5 . 8 3 b 5 . 8 3 ^ Means within each column followed by the same letter are not significantly different (p < 0.05)

122 Seed emergence test Number of seedlings emerged each day on each plot for the three planting methods were counted and averaged for 15 days starting from the first day of emergence. All planting methods have a relatively close result but the precision planter used a smaller amount of seed (135 kg ha !) showing it had the best percentage of emergence compared to the other treatments (Fig 8).

’cultural planting

Melkassa planter

■precission planter

Days from first emergence Fig 8 Seedling emerged vs days from first emergence

Seedling emergence was conccntratcd between 4 - 6 days after first seedling emergence, which resulted in a better and uniform stand. All the planters were in the acceptable range of achieving a uniforn stand. Soil moisture availability had an cffcct on stand uniformity.

The wheat emergence data were results from single year and three replications (Table 10). Planter effects on evaluation parameters for emergence: mean cmcrgcncc time (MET) and percentage of emergence (PE) were statistically important (p< 0.05). The rate at which the crop germinates and emerges, as measured by the Emergence Rate Index (ERI), was not influenced by planter type indicat ng how fast the crop was emerging. It should be a function of planting conditions and secd-soil contact, which as there was no significant difference between the three treatments.

Table 10 Analysis of variance (P values) and means comparisons of seed emergence parameters, MET, ERI and PE values

Analysis of Variance Mean Comparisons MET ERI Source MET ERI PE Planter type (d) (seedlings d-1 nr1) PE (%) Planter 0.0546 0.1169 0.0011 Cultural 2.513 22.193 66.94b Melkassa 2.53a 22.45a 60.56= Precision 2.16^ 21.293 73.87a MET is the mean emergence time; ERI is the emergence rate indexes; PE is the percentage of emergence; Means within each column followed by the same letter are not significantly different (p< 0.05).

The highest value of MET was obtained with the precision planter, while the MET value for the other two planters were similar to one another. The greatest total percentage of emergence was 73.87% for the precession planter while the smallest cmcrgcncc was 60.56% for the Melkassa planter. The reason for the variation could be due to the difference in sowing depth, which has an effect on the availability of moisture. The larger the sowing depth, with in the acceptable range, the higher the moisture content.

123 Conclusion

The type of planter used for planting wheat seed was found to be important in affecting the mean seed spacing, mean emergence time, and the percentage of emergence (p< 0.05). The analysis of factors affecting seed spacing uniformity indicated that the best seed spacing uniformity results were obtained in the laboratory test (81%). The results were close to the minimum requirement (82%). The best depth uniformity was obtained with a precision planter followed by Melkassa planter and the cultural method of planting. The precision planter, the cultural and the Melkassa planter in descending order achieved the best seed emergence ratios. The non-spherical seeds were convenient to be metered with the device; Seed damage was observed to be <5% and was smaller for the precision planter.

References

Sharma DN and Mukesh S. 2010. Farm Machinery Design. Principles and Problems. JAIN BROTHERS, New Delhi. FAO Plant Production and Protection Series No. 30. Kachnia l SD and Smith JA. 1995. Alternative measures of accuracy in plant spacing for planters using single seed metering. Transactions of the ASAE 38(2): 379-387. Panning JW. 1997. Seed spacing performance general puq?ose and speciality type sugarbeet planters. MSc Thesis, University of Nebraska, Lincoln. NE. Teklu Tesfaye. 2010. An overview of tef and durum wheat production in Ethiopia (IAR).

124 Distribution, Physiologic Races and Reaction of Wheat Cultivars to Virulent Races of Leaf Rust in Southeastern Zone of Tigray, Ethiopia

Tesfay Gebrekirstos \ Getaneh Woldeab2 and T. Salvaraj3 1 Tigray Agricultural Research Center Mekele 2 Am bo Agricultural research Center, Ambo 3 Haramaya University, Haramaya

Abstract This study was carried out to determine the distribution and intensity of leaf nist, identify physiologic races and evaluate the seedling reaction of commonly grown wheat varieties to virulent races. Survey was conducted to compute the prevalence and intensity of the disease; race identification through inoculation of leaf rust populations, isolations, multiplication of mono pustules of the pathogen and determination of races by inoculating on leaf rust differential hosts; and evaluating ten wheat varieties to virulent and dominant races (TKTT, THTT and PUTT) at seedling stage. A total of 108 farmers’ wheat fields and experimental plots were assessed in five districts of which 95(88%) of the fields were affected with leaf rust. The overall mean incidence and severity of the disease were 48.4 and 18.2%, respectively. The highest intensity was recorded in Wukro wheat fields with incidence of 63.2% and severity of 37.3%. The lowest incidence and severity of the disease were recorded in D/Temben district with mean values of 7.4 and 4.1%, respectively. Forty mono pustules resulted in the identification of 22 races. Races PIITT and PHRT were predominant with frequencies of 20 and 15%, respectively, followed by THTT and FHRT with a frequency of 10% each. The remaining 18 races were confined to specific locations and detected once with a frequency of 2.5% each. The broadest virulence spectrum was recorded from TKTT race, making all Lr genes ineffective except Lr9. About 81% of the Lr genes were ineffective to more than 55% of P.triticina isolates. High virulence was observed on Lr3, LrlO, LrB and Lrl8 with frequencies of 90, 95 97.5 and 100%, in that order. However, Lr genes 9, 24 and 2a were effective to 100%, 95% and 82 5% of the tested isolates, respectively. The variety evaluation revealed that MekelIe-3, Mekelle-4, Picaflor, Dashen and local showed susceptible reaction to TKTT, THTT and PHTT races. Mekelle-1 and Mckelle-2 were susceptible to races TK’fT and THTT, but resistant to PHTT. Digalu was only susceptible to rKTT race. Unlike bread wheat varieties, the durum varieties, Ude and Dembi were resistant to these races. Most bread wheat varieties did not have adequate resistance: hence, gene pyramiding of Lr9, Lr24 anti Lr2a has paramount importance as the additive effects of several genes offer the variety a wider base for leaf rust resistance.

Introduction

Whca is one of the major cereal crops grown in the highlands of Ethiopia. The area under wheat production is estimated to be about 1.5 million hectare and ranks third after maize (Zea mays L.) and teff (.Eragrostis tef) (CSA 2009). In Tigray region, wheat has been selected as one of the target crops in the strategic goal of attaining regional food self-sufficiency. Wheat covers over 0.1 million hectare with total production of 1.93 million quintals annually (CSA 2011). Southeastern zone of Tigray is one of the major wheat growing areas and recognized as wheat belt in the region. Although, the area cultivated under wheat has been increased in the last few' years, the production and productivity of the crop in Ethiopia in general and Tigray region in particular is still very low. The national average yield is estimated to 1.7 t ha'1 (CSA 2009), although yields > 5 t ha'1 in the highlands arc possible from well managed fields grown with discase-frce varieties (Hailu 1991). Wheat Leaf rust, also known as brown rust is one of the most important foliar diseases of wheat (Triticum Spp.) in Southeastern Zone of Tigray. Yield loss due to leaf rust reached 75% in susceptible wTieat varieties at hot spot areas (Mengistu el al. 1991). Use of host resistance is the most economical and environmentally friendly method of controlling wheat leaf rust (Afzal et al. 2009). However, host resistance may not always be readiK available for use against leaf rust, and it requires regular surveying, race identification and continuous search for new sources of resistant genes in the cultivated and wild forms of wheat

125 (Kuraparthy et al. 2007). In Ethiopia, many studies on distribution, race analysis and identification of source of resistant genes against wheat leaf rust were carried out in different times. However, these works was covered mainly Central, West showa and Southeastern parts of Ethiopia. Other parts of the country like Tigray region in general and southeastern zone in particular, do not have information on the distribution, physiologic races and response of wheat varieties to leaf rust populations. Hence, this thesis was initiated to determine the distribution and intensity of wheat leaf rust, Physiologic races and to evaluate the reaction of wheat varieties to virulent races of wheat leaf rust in Southeastern zone of Tigray. 1

Materials and Methods

Survey of leaf rust in major wheat growing areas of South-Eastern Zone of Tigray: Field survey of leaf rust was carried out in the major wheat growing areas of southeastern zone of Tigray in 2013 growing seison. Private farms in five districts (Wukro, Hintalo wejirat, Saharti samre, Degua Temben and Enderta) and experimental plots of Mekelle Agricultural Research Center were included in the assessment. Wheat fields along the main and accessible roadsides were inspected at 5-10 kilometer intervals. A total of 108 wheat fields were surveyed at critical growth stage of the crop (flag leaf stage) during which leaf rust reached its maximum severity level (Seek 1985). Leaf rust assessment was made along the two diagonals (in ‘’X” fashion) of the field at five points using 0.5m2 quadrant. In each fie d, plants within the quadrant were counted as diseased/infected and healthy/' non-infected and the incidence, severity, and prevalence of wheat leaf rust were calculated as follows.

• Incidence (%) = No of diseased plants x 100 • Total plants assessed • Severity (%) = Leaf area infected • Prevalence (%) = No of affected fields x 100 • Total fields assessed

The prevalence, incidence, and severity data were analyzed by using descriptive statistical analysis (means) over districts, localities, varieties, altitudes and crop growth stages. Collection of wheat leaf nist samples: Leaf rust infected samples (five samples per locality) were collected from wheat fields and experimental plots of MARC. Infected leaves were cut from the mother plant using scissors and placed in an envelope. Samples collected in an envelope were labeled with all necessary information including name of the region, zone, district, locality, variety, GPS data and date of collection. Samples were kept in refrigerator until the surveys in all districts were finalized. Then after, samples were preserved in icebox and transported to APPRC laboratory for race analysis and variety evaluation studies. The samples were kept in the refrigerator at 4°c until used for the virulence analysis.

Isolation and Multiplication of P. triticina inoculums: The inoculum was increased and maintained on universally rust susceptible variety “Morocco' w'hich does not carry any known leaf rust resistant gene (Roelfs 1982). Six seedlings of “Morocco" were raised in suitable 8 cm diameter clay pots containing steam-sterilized soil, sand and manure in a ratio of 2:1:1 mixture, respectively. Seven day- old seedlings or when the primary leaves were fully expanded and the second leaves beginning to grow, the leaves were rubbed gently with clean (dis-infected with 97% of alcohol) moistened (with distilled v/ater) fingers to remove the waxy layer from the surfaces of the leaves. Green house inoculations were done using the methods and procedures developed by Stakman et al. (1962). Spores from the leaf rust infected samples were collected w ith scalpels and transferred on to a watch glass which contain distilled water to make spore suspension, and then it was rubbed on seedlings of Morocco with clean moistened fingers. Plants were then moistened with fine droplets of distilled water produced with an atomizer and incubated in a dark dew chamber for 24 hours at 18-20°C and 90% relative humidity. Then, the seedlings were transferred from the dew chamber to glass compartments where conditions were regulated at 12 hours photoperiod, at temperature of 18-25°C and 60-70% of relative humidity.

126 Multiplication of isolates After seven days of inoculation, when the flecks/chlorosis were clearly visible, leaves containing single flecks were selected from the base of the leaves and the remaining leaves within the pots were removed using scissors. Only 2-3 leaves, which contains mono pustule, were covered separately with cellophane bags (145 x 235 mm) and tied up at the base with a rubber band to avoid cross contamination (Fetch and Dunsmore 2004). After 12-14 days of inoculation, when the mono pustule was well developed, each mono pustule collected using power operated vacuum aspirator and stored separately in gelatin capsule. A suspension, prepared by mixing mono pustule urediospores with lightweight mineral oil (SolTrol 130), was inoculated on seven day-old seedlings of ‘Morocco' for multiplication purpose.

After inoculation, seedlings were placed in dew chamber for 24 hours at 18-22°C and with RH of 90%. Then after, seedlings were transferred to growth chamber where conditions were regulated at 12 hours pholoperiod, 18-25°C and RH of 60-70% following the procedures mentioned earlier. After 12- 14 days of inoculation, the spores from each mono pustule/ isolate were collccted in separate test tubes and stored at 4°C until they were inoculatcd on the differential hosts. This procedure was repeated until sufficient amount of spores are produced to inoculate the set of wheat leaf rust different al host (Table 1). By doing this a total of 40 isolates/mono pustules were developed from 40 wheat leaf rust samples.

Inoculation of wheat leaf rust isolates to differential hosts: Six seeds of the sixteen wheat leaf rust differentials with known resistance genes (Lrl, Lr2a, Lr2c, Lr3, Lr9, Lrl6, Lr24, Lr3ka, Lrl 1, Lrl7, Lr30, LrB, LrlO, Lrl4a and L rl8) and susceptible variety Morocco were grown in 3 cm diameter pots separate y in greenhouse. The susceptible variety Morocco (without Lr gene) was used lo ascertain the viability of spores inoculatcd to the differential hosts The single pustule derived spores (approximately 3-5 mg of spores per ml of liquid suspension) was suspended in distilled water and sprayed onto seven-day-old seedlings using atomizers. After Inoculation, plants were moistened with Fine droplets of distilled water produced with an atomizer and placed in dew chamber for 24 hours at 18-22°C and RH of 90%. Up on removal from the dew chamber, plants were placed separately in the growth chamber to avoid contamination. Greenhouse temperatures and day light were maintained between 18-25°C and 12 hours respectively.

Leaf rust Assessment on differential hosts: Twelve days after inoculation, the infection types were scored for each isolate using the 0-4 scoring scale of Long and Kolmer (1989). Infection types were grouped in to two, where, Low (Resistance) = incompatibility (infection types: 0 to 2) and High (Susceptible) = compatibility (infection type: 3 to 4).

Designation of P. triticina races: Race designations were assigned as described by Long and Kolmer (1989). Race designation was done by grouping the sixteen differential hosts into four sets in the following order: (i) Lrl, Lr2a, Lr2c, Lr3; (ii) Lr9, Lrl6, Lr24, Lr26; (iii) Lr3ka, Lrl 1, Lrl7, Lr30 and (iv) LrB, LrlO, Lrl4a, Lrl8 (Table 3). Each isolate was assigned a four-letter race code based on its reaction on the differential hosts (Long and Kolmer 1989). For instance, low infection type (L) on the four hosts in a set is assigned with the letter 'B1, while high infection type (H) on the four hosts is assigned with a letter'T' (Table 1).

Response of wheat cultivars to leaf rust races at seedling stage in green house: The isolates of prevalent and virulent leaf rust races identified from the southeastern zone of Tigray were multiplied on the universally susceptible variety Morocco following the procedures mentioned in section 2.4 and urediospores collected in separate test tubes to inoculate wheat cultivars. Ten wheat varieties (Mekelle-1, Mekelle-2, Mekelle-3, Mekelle-4, Picaflor, Dashen, Digalu, Ude, Dembi and local cultivar) were evaluated against the virulent race TKTT and dominant leaf rust races THTT and PHTT.

127 Table 1 Nomenclature of Puccinia triticina races on 16 differential hosts in ordered sets of four

Host set Infection type (ITs) produced on differential Lr lines Host set 1 1 2a 2c 3 Pt code Host set 2 9 16 24 26 Host set 3 3ka 11 17 30 Host set 4 B 10 14a 18 B LL LL C L LLH D L L L F L LH G LH LL H LH LH JLH L K LHH L H LL L MH LLH NH L L PH LH Q H HLL RHH LH S HHH L TH H H H Source: Long and Kolmer 1989

Results and Discussions

Survey of wheat leaf rust in South Eeastern Zone of Tigray: Whenever disease assessments are made, growth stage of the plants is essential for meaningful comparisons between varieties, locations and years (Stubbs et al. 1986). In view of this, 2.8% of the wheat crop was at tillering to booting, 50% at heading to flowering and 47.2% at milk to dough growth stages. Leaf rust was observed in 2(66.7%), 45(83.3%) and 47(92.2%) of the 3, 54 and 51 wheat fields inspected at the stages of tillering to booting, heading to flowering and milk to dough growth stages of the crop respectively. Leaf rust was found more important at heading to flowering and followed by milk to dough growth stages of the crop. During heading to flowering growth stages, mean incidence of 67.2% and severity of 26.9% were recorded. The disease also caused high infection at milk to dough growth stages, with mean incidence and severity of 44.2% and 12.7%, respectively.

Conversely, the lowest incidence and severity of leaf rust was recorded at tillering to booting growth stages of the crop with mean values 21.2% and 6.7%. respectively. This might be resulted from the fact that, the flag leaf, which is the critical growth stage of the crop during which leaf rust, reached its maximum severity level (Seek 1985), was not fully develop at tillering to booting growth stages. On top of this, damage is minimal during tillering due to the onset of colder temperatures that are likely to eliminate or reduce reproduction and spread of rusts.

Distribution and intensity of wheat leaf rust across districts The disease was more prevalent at H/wejirat and Wukro districts with prevalence of 100% and 96.7% respectively, while districts of Enderta and S/samre showed similar distribution of wheat leaf rust, 85% each. In contrast, the lowest prevalence (45.5%) ofleaf rust was registered at D/ Temben district. As a whole, wheat leaf rust was observed on 88% of the 108 wheat fields inspected in southeastern zone of Tigray (Table 2).This indicated that, the pathogen was widely distributed across the districts of the study area.

128 Different mean incidences and severities were rccordcd in all districts. The highest mean incidence of the disease was noted in Wukro district with range of 0-100% and a mean value of 63.2% and followed by S/samre, with range of 0-100% and mean incidence of 60%. Districts of Endcrta and H/wejirat also showed a considerable level of leaf rust incidence with mean values of 57.9% and 53.3%, respectively (Tabic 2). However, the lowest incidence of the disease (7.4%) was registered in D/Temben district.

Table 2 Prevalence and intensity of wheat leaf rust across districts in 2013 main cropping season

Number of Number of Incidence (%) Severity (%) Altitude range fields fields Prevalence Districts (masl) inspected infected (%) Range Mean Range Mean H/wejirat 1966-2198 27 27 100 5-100 53.3 1-75 14.1 Wukro 1927-2399 30 29 96.7 0-100 63.2 0-95 37.3 S/Samrs 1961-2339 20 17 85 0-100 60 0-85 26.1 Endertc 1912-2292 20 17 85 0-100 57.9 0-25 9.5 D/Temten 2431-2654 11 5 45.5 0-24 7.4 0-10 4.1 Total /Mean 1912-2654 108 95 88 0-100 48.4 0-95 18.2

Likewise, leaf rust severity showed similar trend as that of incidence in Wukro, S/samrc and D/Temben districts. The highest severity was recorded in Wukro district with a range of 0-95% and mean value of 37.3%. This was followed by S/samre district, with range of 0-85% and mean severity of 26.1%. The highest severity of 37.3% was recorded where the highest incidence of 63.2% was registered at Wukro district. In a similar trend, the lowest mean severity of 4.1% was recorded where the lowest mean incidence of 7.4% was registered at D/Tembcn. Generally, leaf rust was more important at Wukro, S/samre, H/wcjirat, and Enderta districts. The high level of leaf rust intensity in these districts might be the cultivation of local cultivar or “Shahan” which is susceptible to wheat iusts in general and leaf rust in particular. This cultivar covered about 53.3, 55.6, 57.9, and 65% of the wheat area in W'ukro, H/wejirat, Enderta, and S/samre districts respectively (BoARD 2005). Moreover, the cultivation of wheat in some areas of these districts during offseason could have played a significant role in rehabilitating the population of wheat leaf rust. Hence, these varieties could act as a green bridge to carryover the disease from offseason to main season. Earlier study also indicated that the presence of two overlapping seasons for growing wheat (meher and belg seasons) helps in the buildup of inoculum in one season and transferred to the other season as a source of primary inocuium, facilitating the availability of inoculum year after year in the country (Serbessa 2003).

However, the low prevalence and intensity of leaf rust in D/Temben district might be resulted from low temperature occurred during the season. In this district, the temperature was less than 10°C (ENMA 2013) which is below the optimum level of temperature. This low temperature likely eliminates or reduces the reproduction and spread of the rust. Earlier studies also confirmed that, at 10°C, infection developed very slowly and restricted in size (Dyck and Johnson 1983).

Distribution and intensity of wheat leaf rust across altitude ranges Of 108 wheat fields inspected, approximately 84% of the fields were found with altitude range of 1801-2300 masl, while the remaining 16% were found between 2300-2654 masl. The highest prevalence of the disease was recorded at altitude range between 1801-2300 masl. Out of 91 wheat fields inspected in this altitude range, the disease was found in 84 wheat fields. Similarly, the highest incidence and severity of wheat leaf rust were recorded at this altitude range of 1801-2300 masl with range and mean values of 5-100% and 59.9% and Trace-95% and 21.4% respectively.

However, the distribution and intensity of wheat leaf rust reduced at higher altitudes. Relatively low prevalence (64.7%) of leaf rust was recorded at altitude range of 2300-2652 masl. In the same way, the range and mean values of incidcnce and severity were reduced to 1-100% and 29% and Trace- 15% and 5.6%, respectively. The overall mean incidence and severity of wheat leaf rust for the midland (1801-2300 masl) and highland (2300-2654 masl) of the study area reached 44.5 and 13.5%,

129 respectively (Table 3). In general, though, the disease was more important in midland areas, it was also distiibuted in the highland of the study area with considerable amount of intensity and prevalence. This might be associated with the fact that, wheat leaf rust occurs wherever wheat is grown and it is the most widely distributed of all cereal rusts (Knott 1989). Moreover, the adaptability of leaf rust to different climates play a significant role for the widely distribution of the pathogen (Roelfs et al. 1992). This result is also similar with the findings of Dagnatchew (1967). He reported that, leaf ust is endemic at different altitudes of wheat growing regions of Ethiopia.

Table 3 Prevalence and intensity of leaf rust in different agro ecologies of Southeastern Zone of Tigray in 2013 cropping season.

Number of Incidence (%) Severity (%) Altitude Number of Infected Prevalence (ml) Inspected field fields (%) Range Mean Range Mean 1801-230 0 91 84 92.3 5-100 59.9 Trace- 95 21.4 2300 -2654 17 11 64.7 1-100 29 Trace-15 5.6 Total/Mean 108 95 78.5 1-100 44.5 Trcea-95 13.5

Prevalence and intensity of wheat leaf rust by wheat type and variety: During the survey, the prevalence and intensity of leaf rust varied between durum and bread wheat varieties (Table 4). Though, durum wheat varieties, Ude and Dembi showed prevalence of 100%, the intensity of the disease was lower on these varieties compared lo Dashen. Digalu, Picaflor, Mekelle-3, Mekell-1 and local cultivar (Shahan). The range incidence and severity of leaf rust in durum varieties varied between (>-10% each respectively. The lowest mean incidence (2.5%) and severity (2.5%) of leaf rust was recorded on variety Dembi and followed by Ude, with mean incidence of 5% and severity of 3.8% across the three fields. The low intensity in ihese varieties was resulted from their resistant response recorded in all wheat fields of the study area. This finding is in agreement with previous report that stated that most of the commercial durum wheat cultivars exhibited stable resistance to wheat rusts across seasons in hot spot areas of Ethiopia and they could be exploited in wheat breeding programs (Efrem et al. 1995).

In contrast, the highest intensity of leaf rust was recorded in bread wheat varieties at different levels of incidences and severities. The prevalence and intensities of leaf rust in Mekelle varieties were lower as compared to Dashen, Digalu, Picaflor, and Local cultivar. The prevalence of leaf rust in these varieties ranged between 50 -100% of the fields cultivated. The leaf rust incidences and severities in these varieties varied between 0-19.1% and 0-10%, respectively. The lowest incidence and severity of the disease was recorded in Mekelle-4 with mean values of 2.5% each, respectively (Table 4).

Table 4 Prevalence and intensity of leaf rust on varieties grown in Southeastern Zone of Tigray in 2013 cropping season

Number Number of Incidence (%) Severity (%) Altitude Prevalence Varieties of fields fields (m) (%) Range mean Range Mean Inspected infected Mekelle-1 1980-2142 6 4 66.7 5-19.1 5.9 1-5 5 R, MR Mekelle -2 1970-2155 4 2 50 0-15 7.5 0-5 2.5R, MR Mekelle- 3 1975-2165 3 2 66.7 0-15 7.5 0-10 5R,MR Mekelle- ^ 2012-2178 2 2 100 0-5 2.5 0-5 2.5R,MR Picaflor 2021-2420 7 6 86 0-65.5 32.3 0-50 12MR.MS Dashin 1994-2595 18 17 94.4 0-100 43.3 0-65 13.7MR.MS Digalu 1961-2006 2 2 100 54-75 64.5 5-25 15MR.MS Shahan (local) 1912-2654 61 55 90.2 0-100 72.9 0-95 27 S Ude* 2000-2626 3 3 100 0-10 5 0-10 3.8R Dembi* 1973-2614 2 2 100 0-5 2.5 0-5 2.5R

130 High in.ensily of wheat leaf rust was found on commercial bread wheal varieties of Dashcn, Digalu, Picaflor and Local cultivar as compared with the other varieties. The intensity of the disease in these varieties varied between 0-100% in incidence and 12-27% in severity (Table 4). Varieties, Dashen, Digalu and Picaflor demonstrated moderately susceptible to moderately resistant response to leaf rust across locations. However, the local cultivar consistently showed susceptible response to the disease in all the study areas. As a result, the highest mean incidcnce of 72.9% and severity of 27% were recorded on this cultivar. The disease was prevalent on 90.2% of the fields cultivated with local cultivar. The long period cultivation and increase in susceptibility from time to time by leaf rust population probably makes the local cultivar highly infected. In addition, the wide cultivation of this cultivar in the study area also played a significant role for its susceptibility, as leaf rust is probably more damaging when large areas are sown to single, genetically homogeneous or closely related cultivars (Ahmad et al. 2010). This idea is in line with the reports of Mamluk et al. (2000) who stated that, majority of the Ethiopian farmers grow landrace cultivars that are susceptible to the disease; even though a large number of improved cultivars of wheat have been released. Generally, most of the varieties demonstrated different response across localities, among varieties and even within the same variety This variation might be the result of differences in host growth stage. Susceptibility and resistance are often highly correlated with host growth stage even with races specific resistance. A host may be subjected to a heavy inoculum density with favorable infection period at critical growth stage, while other host may not be confronted with similar circumstances when it is at the critical stage (Roelfs 1982). Environmental difference across districts and variation in aggressiveness among population of wheat leaf rust can also result different responses even with the same varieties.

Identification of P. triticina races in Southeastern Zone of Tigray: Using the international system of nomenclature for P.triticina (Long and Kolmer 1989), 22 raccs were identified from 40 mono pustules or isolates based on their reactions on 16 differential hosts.

Distribution and diversity of P. triticina races across districts: As far as race distribution was concerned, though most of the races were confined to specific districts, some had wider spatial distributions. Four raccs (FHRT, PHRT, PHTT and THTT) were predominant, representing 55% of the isolates analyzed. Raccs PHTT and PHRT were the most predominant with frequencies of 20 and 15% respectively, followed by THTT and FHRT with a frequency of 10% each. These races were isolated from three or four districts of the study area (Table 5), which indicated that they were widespread throughout southeastern zone of Tigray. PHTT was detected eight times in the population of wheat leaf rust collected from Wukro, H/wejirat and Enderta districts while, PHRT was isolated six times from S/samre, D/ Temben, Wukro and Enderta populations of wheat leaf rust. On top of this, PHRT was identified as the most distributed race and adapted to wide agro ecologies of the study area. Races, THTT and FHRT also isolated four times each from districts of Wukro, S/samrc and H/wejirat and Wukro, Enderta and H/wejirat respectively. The predominance of races of P. triticina in these districts provides evidence of clonal lineages and short distance migration of this pathogen within the study area. On the other hand, approximately 82% of the races including the most virulent race KTT, were confined to specific locations and detected only once with a frequency of 2.5% each Table 5). The distribution and diversity of P.triticina races indicated that, genetic similarity among isolates of within and between districts of the study area was existed. The three adjacent districts (Wukro, H/Wejirat and Enderta) had two similar races, FHRT and PHTT out of eight, seven and seven raccs detected, in that order. Likewise, Wukro and S/samre districts had two races in common, PHRT and THTT out of eight each respectively. This genetic similarity between P.triticina isolates of these districts is in line with the findings of McVey et al. (2004), who reported similar level gcnetic similarity between P.triticina populations collected from Egypt in 1998 to 2000 and from southern and ccntral plains of United States in the same period. The within districts comparison also indicated that, similarities among isolates of Wukro, H/wejirat, S/samre and D/Temben were observed. Out of the 11 isolates collected in wukro district, 36.4% of the isolates showed gcnetic simil arity and resulted in race PHTT. In H/wejirat district, 20 and 30% of the isolates were resulted in FHRT and PHTT races respectively. This indicated that, only 50% of the leaf rust isolates showed genetic diversity in this district. Isolates of S/samre also showed gcnetic similarity as a result; races

131 THTT and PHRT were detected two times each from 10 isolates of leaf rust. In the same way, PHRT was identified from two isolates collected from D/ Temben District.

Their geographic proximity, absence of barriers and cultivation of similar bread wheat cultivars among Wukro, H/Wejirat and S/samre districts might have played significant role for race similarity. On top of this, these races might be more fit or easily adapted with the environment of these districts. D/Temben district on the other hand is geographically isolated by mountains from other places. Thus, the possibility of migration of urediospores of wheat leaf rust to and from this district is much restricted and low diversity among P.triticina population is expected in this District.

Table 5 Distribution of Puccinia triticina races across districts of Southeastern Zone of Tigray in 2013 cropping season

Districts Isolates Races Wukro Enderta S/samre D/Temben H/wejirat Frequency (%) BBBT -- 1 -- 1 2.5 BBQR -- 1 -- 1 2.5 CBBT - 1 --- 1 2.5 FGRT -- 1 -- 1 2.5

FGTT -- 1 -- 1 2.5

FHRT 1 1 -- 2 4 10

FHTT 1 ---- 1 2.5

LBBM --- 1 1 2.5

LBDC - 1 -- 1 2.5

MBBR 1 --- 1 2.5

MCST -- - 1 1 2.5 MGJT 1 --- 1 2.5

MHTT --- 1 1 2.5

PCRR •• --- 1 1 2.5

PGRT 1 ---- 1 2.5

PHRT 1 1 2 2 - 6 15

PHTT 4 1 -- 3 8 20

PJTT -- 1 -- 1 2.5

RCJT - 1 --- 1 2.5 RHTT i ---- 1 2.5

THTT 1 - 2 - 1 4 10 TKTT 1 ---- 1 2.5 Total 11 7 10 2 10 40 100

In contrast, the ‘within district’ comparison had also indicated that, isolates collected from Enderta showed genetic diversity among the populations of wheat leaf rust. The seven isolates collected from this district yielded seven races (RCJT, PHTT, CBBT, FHRT. MBBR, MGJT, and PHRT) (Table 5). The high level of race diversity in this district might be resulted from the windy nature of this area. Hence, the movement of P.triticina urediospores via wind from their sources to or from this area is a common phenomenon in rusts in general and leaf rust in particular. This circumstance might be resulted, more heterogeneity in the population of wheat leaf rust and finally the chance of detecting different races in this district become increased.

Virulence spectrum of P. triticina races The number of differential lines determined Virulence spectrum that the isolate showed virulence. In this case, an isolate having virulence on more leaf rust resistance genes was considered to have wider spectrum compared to those isolates with virulence to relatively lower number of differential lines (Sewalem et al. 2008). In view of this, approximately 73% of the races had virulence spectra raging from 9 to 15 Lr genes. The widest virulence spectrum was recorded from TKTT race making 15 Lr genes ineffective (Table 6). However, this race was not widely distributed; it seems to be important in that it attacks all the members of the differential hosts except Lr9. In addition, this race has a potential to cause heavy infection on many bread wheat varieties grown in areas where this race was discovered. Similarly, races THTT was also the second most virulent racc making 14 Lr genes susceptible.

The virulence spectrum of P.triticina indicated that, some races showed the same virulence spectrum on the Lr genes. For instance, three races (RHTT, PHTT and PJTT), (FHTT, MHTT and PHRT) and (FGTT FHRT and PGRT) were virulent equally to 13 (81.3%), 12 (75%) and 11 (68.8%) of Lr genes respectively. Likewise, races FGRT, MCST, PCRR and RCJT had the same virulent spectrum, each produced virulence on 10 or 62.5% of Lr genes. Race MGJT was virulent on 9 or 56.3% of the Lr genes tested. This indicated that, unless wheat varieties have combined Lr genes through pyramiding, the mentioned races above have a potential to cause heavy infection during what production in the region n general and Southeastern zone in particular.

In contrast, the remaining six races (BBBT, BBQR, CBBT, LBBM, LBDC, and MBBR) or 27% of the races had narrow virulence spectra ranging from 3 to 5 Lr genes (Table 6). The “L” group races, LBBM and LBDC were the least virulent, producing compatible reaction only on three Lr genes (Lrl, LrB and Lrl8) and (Lrl, Lrl7 and Lrl8), respectively. Races BBBT, BBQR, CBBT, and MBBR were also the least virulent, producing susceptible reactions on four, five, five and five leaf rust resistant genes n that order (Table 6). Approximately 55% of the races identified in Southeastern zone of Tigray varied from one another by single gene changes. For instance, races FGTT and FHTT were similar to FGRT and FHRT with additional virulence each to Lrl7, respectively. In the same way, races PHRT, PHTT, THTT, and TKTT were similar to PGRT, PHRT, RHTT, and THTT with additional virulence to Lr26, Lrl7, Lr2c, and Lr24, respectively (Table 6). This slight difference in virulence between these races of leaf rust may resulted from the continuous evolution of leaf rust through one or more of the mechanisms of variation (mutation, migration, recombination and selection pressure on race specific resistance). This idea is in line with the report of Green (1975) who stated that, single step changes in virulence were resulted from the main proccss of evolutionary change in wheat leaf rust Populations.

The present study indicated that, the identified races of P.triticina did not show similarities with the previously identified races in Ethiopia. This could be due to variation over location and time, as raccs are prevalent in specific season and region depend on the type of wheat cultivars grown (Singh 1991), and to some extent on the predominant environmental conditions, especially temperature (Roelfs et al. 1992). Generally, the virulence spectrum of the pathogen in this study area confirmed the presence of wider range of virulence among the population of wheat leaf rust races. This might be linked with the fact that, the large population size of leaf rust leads to greater probability of mutants and more diversity of virulence/avirulcncc combination existed in the crop (Schafer and Roelfs 1985).

Virulence frequency of P. triticina isolates to Lr genes: Virulence frequency of P. triticina indicated that, majority of the resistance genes were found ineffective by most of the isolates tested in this study. Approximately, 81% of the Lr genes were ineffective to more than 55% of the isolates. High virulence (>72.5%) has been exhibited on Lr genes Lrl, Lr2c, Lr3, Lrl6, LrSka, Lrl 1, Lr30, LrB, LrlO, Lr26, and Lrl4a. There was 100% frequency of virulence for leaf rust resistant genes L rl8. The Lr 17 has an intermediate virulence frequency of 55%, while the remaining genes, Lr9, Lr24 and Lr2a were found to have between 0-17.5% of virulence frequencies (Table 7). Some Lr genes such as, Lr2c and Lr26, LrlO and Lr30 and Lrl4a and L rll had the same virule frequency of 72.5%, 77.5% and 87.5% respectively. The Lrl8 displayed consistently high infection type to all isolates of P.triticina collected from southeastern zone of Tigray. All the identified raccs including the least virulent races, LBBN and LBDC were virulent on this gene and showed susceptible reaction just like the universally susceptible variety “Morocco”. This showed that absolute compatibility of Lrl8 to all races of P.triticina was demonstrated. This information may provide a clue either Lrl 8 is effective and expressed at adult plant or completely lost from its differential line, RL6009. However, further study on the effectiveness of Lii8 against wheat leaf rust population is required to consolidate this conclusion. Different authors have reported similar results on the ineffectiveness of the Lrl8. For instance, Torabi et al. (2001) reported that, the host with Lrl 8 appeared to be ineffective to all isolates

133 at seedling in the green house, but it showed considerable resistance at adult plant. Singh (1991) also reported ihat, virulence status in the pathogen for this gene could not be determined. Similarly, there was also 97.5% frequency of virulence for LrB. This gene was effective only to the least virulent race, LBDC isolated from the local cultivar in S/samre district. McIntosh et al. (1995) also reported that, LrB was ineffective to leaf rust isolates in most geographic areas of Australia. The ineffectiveness of the genes Lrll and Lrl7 at seedling stage were expected as they were reported to be adult plant resistant genes (Kolmer 2003).

Table 6 Virulence /avirulence spectrum of P. triticina races collected from Southeastern Zone of Tigray in 2013

Virulence Races Virulence (Ineffective Lr genes) AVirulence (effective Lr genes) factor BBBT LrB, 1 0,14a, 18 Lrl, 2a, 2c, 3, 9, 16, 24, 26, 3ka, 11, 4 17,30 BBQR Lr3ka, 11, B, 10,18 Lr1,2a,2c, 3,9,16,24,26,17, 30,14a 5 CBBT Lr3, B, 10,14a, 18 Lr1, 2a, 2c, 9, 16, 24, 26, 3ka, 11, 17, 5 30 FGRT Lr 2c, 3,16, 3ka, 11, 30, B, 1 0,14a, 18 Lr1, 2a, 9, 24, 26,17, 10 FGTT Lr2c, 3,16, 3ka, 11,17, 30, B, 10,14a, 18 Lr1, 2a, 9, 24, 26 11 FHRT Lr2c, 3,16, 26, 3ka, 11, 30, B, 10,14a, 18 Lrl, 2a, 9, 24,17 11 FHTT Lr2c, 3,16, 26, 3ka, 11,17, 30, B, 10,14a, 18 Lr1,2a, 9,24 12 LBBM Lr1,B,18 Lr2a , 2c, 3, 9, 16, 24, 26, 3ka, 11,17, 3 30,10 ,14a LBDC Lr1,17,18 Lr2a, 2c, 3, 9, 16, 24, 26, 3ka, 11, 30, 3 B, 10,14a MBBR Lr1, 3, B, 10,18 Lr2a, 2c, 9, 16,24, 26, 3ka, 11, 17, 30, 5 14a MCST Lr1, 3, 26,3ka, 11,17, B,10,14a,18 Lr2a, 2c, 9,16,24,30 10 MGJT Lr1,3,16,11,17, B,10,14a,18 Lr2a, 2c, 9, 24, 26, 3ka, 30 9 MHTT Lr1, 3,16, 26, 3ka, 11,17, 30, B, 10,14a, 18 Lr2a, 2c, 9, 24 12 PCRR Lr1,2c,3,26,3ka,11,30,B,10,18 Lr2a, 9 ,1 6 ,2 4 ,17,14a 10 PGRT Lr1,2c, 3,16, 3ka, 11, 30, B, 1 0,14a, 18 Lr2a, 9, 24,26,17 11 PHRT Lr1, 2c, 3,16, 26, 3ka, 11, 30, B, 10,14a, 18 Lr2a, 9, 24,17 12 PHTT Lr1, 2c, 3,16, 26, 3ka, 11,17, 30, B, 10,14a, 18 Lr2a, 9, 24 13 PJTT Lr1, 2c, 3,16, 24, 3ka, 11,17, 30, B, 1 0,14a. 18 Lr2a, 9,26 13 RCJT Lr1, 2a, 3, 26,11,17, B, 10,14a, 18 Lr2c, 9,16, 24, 3ka, 30 10 RHTT Lr1, 2a, 3, 16, 26, 3ka, 11 17, 30, B, 10. 14a, 18 Lr2c, 9, 24 13 THTT Lr1,2a, 2c,3,16,26,3ka, 11,17,30,B,10,14a,18 Lr9, 24 14 TKTT Lr1,2a, 2c.3,16,24,26,3ka,11,17,30,B,10,14a,18 Lr9 15

Table 7 Virulence frequency of P. triticina isolates on 16 Lr genes in 2013 cropping season

Lr Number of Virulence Lr Number of Virulence gene /irulent isolates Frequency (%) gene virulent isolates frequency (%) Lr1 30 75 Lr3ka 33 82.5 Lr2a 7 17.5 Lr11 35 87.5 Lr2c 29 72.5 Lr17 22 55 Lr3 36 90 Lr30 31 77.5 Lr9 0 0 LrB 39 97.5 Lr16 31 77.5 Lr10 38 95 Lr24 2 5 Lr14a 35 87.5 Lr26 29 72.5 Lr18 40 100

Moreover, the ineffectiveness of Lrl, Lr2c, Lr3 and LrlO might be due to these genes have been used in wheat cultivation for many years (Long 1986), during which virulence to these genes become

134 common and most races identified in recent years are virulent to these genes. Likewise, virulence for l.r26, LrJ6, Lr30, Lr3ka, and Lrl4a was very common by most isolates of leaf rust with virulence frequencies of 72.5, 77.5, 77.5, 82.5, and 87.5% respectively. This virulence could be result from the fact that, leaf rust differential lines have single and specific resistant genes, and race specific resistant genes have been proven vulnerable to selection and increase virulent raccs in rust population (Kilpairick 1975). On the other hand, Lr9, Lr24, and Lr2a were effective to most of wheat leaf rust populations (Table 12). The leaf rust resistant gene, Lr9 derived from Aegiolops umbcllulata, demonstrated an incompatible host-pathogen interaction to all isolates of leaf rust. This implied that, no virulence was observed on Lr9 (virulcnce frequency=0%) in all the districts of collection. In Ethiopia, this gene was also effective to wheat leaf rust isolates collected in 2004 from Ethiopia and Germany (Sewalem et al. 2008). This finding is also in a good agreement with the previous studies that stated that, no virulence to Lr9 was (Hussain et al. 1980). Similarly, Lr24 was found to confer resistance to 95% of the tested leaf rust isolates. This gene was ineffective only by two virulent races, TKTT and PJTT identified from Wukro and S/Samrc isolates respectively.

Kolmer (2003) also stated that, Lr24 is generally effective in many wheat-producing areas of the world. Virulence on Lr2a was also rare and found to be effective to 82.5% of leaf rust isolates. Antench (2011) also reported that, Lr2a was exhibited effectiveness lo wheat leaf rust races. Hence, these genes can be used as potential sources during wheat breeding programs for resistance to wheat leaf rust.

Response of wheat varieties to leaf rust races at seedling stage in greenhouse: In this study, seedlings of 10 commonly grown wheat varieties and the universally susceptible check, Morocco were screened against the most virulent and dominant races of leaf rust (TKTT, THTT and PHTT). The reaction of wheat varieties to leaf rust races in the green house revealed that none of the varieties were immune (no sign of infection to the naked eye) while the infection type varied from 1 (small urcdia surrounded by nec'otic area) to 4 (large uredia without chlorosis) (Table 8). Among the tested wheat varieties, eight (Mekelle-1, Mekelle-2, Mekell-3, Mekelle-4, Picaflor, Digalu, Dashen and local cultivar), seven (Mekelle-1, Mekelle-2, Mckelle-3, Mckelle-4, Picaflor, Dashen and local cultivar) and five (Mekelle- 3. Mekelle-4, Picaflor, Dashen and local cultivar) of them produced susceptible reaction to TKTT, THTT and PHTT races, respectively. Five bread wheat varieties namely, Mekelle-3, Mekelle-4, Picaflor, Dashin and local cultivar were susceptible to the three races. Varieties, Mekelle-1 and Mekelle-2 were susceptible to TKTT and THTT races, but resistant to PHTT. Digalu was only susceptible to TKTT but showed resistance to THTT and PHTT races. The data showed that, the more the vin lent race, the more susceptibility on many bread wheat varieties was recorded and vice versa.

Most varieties showed more resistance under the natural infection in the field than at seedling stage in the green house. All the Mekelle varieties, Mekclle-1, Mckelle-2, Mckelle-3, and Mckell-4 showed moderately resistance to resistance under the natural infection in the field, while they showed susceptibility to at least two of the virulent races of leaf rust at seedling stage in the greenhouse. Similarly, Picaflor, Dashin, and Digalu showed moderately susceptible to moderately resistance reaction under the natural infection, while they showed susceptible reaction to at least one of the races of leaf rust at their seedling stage. This variation might be resulted due to, these varieties may have genes responsible for adult plant resistance at field condition, but poorly expressed at their seedlings in the greenhouse. The presence of conducive environment in the greenhouse provided optimum development to leaf rust as compared to the field where environmental conditions might not regulated based on the requirement of the pathogen. Hence, this circumstance also contributed for the suscept ible response of these varieties at their seedling stage in the green house.

Moreover, varieties in the greenhouse were evaluated through the inoculation of the most virulent and dominant raccs of leaf rust and resulted susceptible reaction on the seedlings of the above mentioned varieties. However, varieties in the field have a chance to be infectcd with weak population of leaf rust. In effect, low infection type or resistance lo leaf rust could be observed in these varieties under field condition. The local cultivar however, showed susceptible reaction under natural infection and greenhouse inoculations for leaf rust population. This cultivar had high infection types similar to that

135 of “morocco”. Therefore, this cultivar can be assumed as having no effective gene (s) at its seedling and adult plant growth stages. Both durum wheat varieties, Ude and Dembi have demonstrated resistance to leaf rust at adult plant in the field and seedling stages in the greenhouse.

In general, durum wheat varieties showed better resistance than bread wheat. This might be associated with the fact that, most of the durum wheat genotypes were developed from local landraces, as Ethiopia is the centre of genetic diversity of this species. In effect, indigenous pathogens with high complimentary genetic diversity might co-exist with a wider range of durum wheat genotypes (Tesemma and Bechere 1998). This idea is also in agreement with previous reports, which stated that, most of 'he commercial durum wheat cultivars exhibited stable resistance to wheat rusts across seasons in hot spot areas of Ethiopia and they could be exploited in wheat breeding programs (Efrem et al. 199.-5).

Table 8 Response of wheat cultivars to dominant and /irulent races of wheat leaf rust at seedling stage in greenhouse in 2013 growing season.

Races Cultivar TKTT THTT PHTT Mekelle -1 3 3 2+ Mekelle- 2 3 3- 2 Mekelle- 3 3 3 3 Mekelle -A 3 3 3 Picaflor 3 3+ 3 Digalu 3- 2+ 2 Dashin 3+ 3+ 3 Ude 1 + 2 2- Dembi 2- 2- 2- Local cultivar (Shahan) 4+ 4 4 Morocco( Susceptible check) 4+ 4 4 "+ ”=slightly larger than the normal uredinia; “ slightly smaller than the normal uredinia

In contrast, as bread wheat is not indigenous to Ethiopia, cultivars are developed through selection and crossing programs using genetic materials introduced from abroad, mainly from CIMMYT. As a result, br:ad wheat cultivars in Ethiopia have a narrow genetic base (Hailu 1991). The narrow genetic base makes bread wheat varieties highly selected and break their resistance by the new' race (s) after short period of releasing.

Conclusions and Recommendations

Leaf rust is one of the most important foliar diseases of wheat in Tigray Region in general and southeastern zone in particular. The pathogen is capable to produce new races that attack previously resistant varieties and develop rapidly under optimal environmental conditions which results in a serious yield loss. Hence, monitoring the diseases and its races is great importance for sustainable wheat management programs. Considering this, field survey of wheal leaf rust was carried out in the major wheat growing areas of southeastern zone of Tigray in 2013 growing season. The result of this survey revealed that, the prevalence and intensity of leaf rust varied across districts, wheat types and varieties, altitudes and growth stages of the crop. A total of 108 wheat fields were inspected and 88% of the fields exhibited the symptoms of leaf rust. The highest and lowest intensity of the disease were recorded in Wukro and D/Temben Districts with mean incidences and severities of 63.2% and 37.3% and 7.4% and 4.1%, respectively. The overall mean incidence and severity of the disease were reached 48.4% and 18.2%, respectively.

The survey was carried out at altitude ranges of 1801-2654 masl. in which the incidence and severity of the disease was relatively high at lower altitudes as compared to higher altitudes. The lowest

136 incidence of 29% was recorded at altitude ranges of 2300-2654, while the highest incidence of 59.9% was recorded at altitude ranges of 1801-2300 m. Similar tendencies were observed regarding the severity of the disease in which, the lowest 5.6% and the highest 21.4% severities were noted at altitude ranges of 2300-2654 and 1801-2300 m.a.s.l in the same order. Race analysis provides essential information in determining the direction of research and breeding programs before the pathogen becomcs a threat to wheat production. Study on the variability of the pathogen resulted in the identification of 22 races from 40 isolates. Race TKTT was the most virulent race making all Lr genes except Lr9 ineffective. Raccs PHTT and PHRT were found to be the most commonly encountered races, detected eight and six times, respectively and followed by THTT and FHRT races detected four times each in the population of leaf rust. In general, 16 races namely TKTT, THTT, PHTT, PHRT, FHRT, RHTT, PJTT, MHTT, FHTT, PGRT, FGTT, MCST, RCJT, PCRR, FGRT, and MGJT of P.triticina were the most important races of leaf rust in the study area as they produce 68.8- 93.8% virulence on differential hosts. In contrast, the remaining races BBBT, BBQR, CBBT, LBBM, LBDC, and MBBR showed low virulence on 18.8% to 31.3% Lr genes. The ‘L’ group races, LBBM and LB DC were found to be the least virulent, each producing virulence on 18.8% of Lr genes. The present study indicated that, all the identified races of P.triticina have shown genetic uniqueness as compared to the previously identified races in Ethiopia. The variation over location, wheat varieties grown, md environmental conditions, especially temperature might be contributed for the uniqueness of the races.

Approximately, 81% of the Lr genes were ineffective to more than 55% of Puccinia triticina isolates. High virulence frequencies (>72.5%) have been found on the resistance genes Lr2c, Lr26, Lrl, Lr3, Lrl 6, Lr3ka, Lrl 1, Lr30, LrB, LrlO, Lrl4a and Lrl 8. However, Lr genes 9, 24 and 2a were found effective lo 100%, 95% and 82.5% of the tested isolates. Hence, these genes are among the most important genes, which could be used as sources of resistance to wheat leaf rust. Evaluation for wheat varieties for their resistances is very important in integrated leaf rust management. In this study, 10 commonly grown wheat varieties were evaluated against three virulent and dominant races of leaf rust at seedl ng stage in greenhouse. These varieties showed broad infection types from 1 (small uredia surrounded by necrotic area) to 4 (large uredia without chlorosis). Generally, the study confirmed the presence of wider range of virulence among the population of leaf rust races in wheat, indicating the presence of genetic diversity among the raccs. To conclude, not all the tested bread wheat varieties have adequate resistances for leaf rust populations, indicating the need for incorporating more effective genes in to the target wheat cultivars. However, durum wheat varieties, Ude and Dcinbi showed resistance to leaf rust population in the field and greenhouse. Hence, they could be important sources of leaf rust resistant genes for this area.

Recommendations

■ Lea rust is highly variable even within a single cropping season, and breakdown the previously resistant varieties. Hence, it has to be surveyed regularly to determine its current status and to take action before the pathogen becomes a risk to wheat production; ■ Virulence has been observed on all of the Lr genes except Lr9. Thus, searching for new source of leaf list resistant genes is necessary to maintain leaf rust resistance; ■ The leaf rust resistant gene Lr9 was identified as effective gene to all leaf rust isolates. Hence, it should be utilized in breeding programe with other effective genes through gene pyramiding as the additive effects of these genes offer the cultivar a wider base of leaf rust resistance; ■ The results from both seedling test and field survey revealed that local cultivar exhibited susceptibility to leaf rust populations. Hence, Breeders and/ Plant pathologists should replace this cultivar by developing race non-specific resistant varieties; ■ The Ethiopian durum wheat landraces are potential sources of leaf rust resistance. Hence, their resistant genes should be exploited in wheat breeding programs; and ■ Fina ly, Plant pathologists and/or breeders should use this data as a base line during resistant variety development programme in the study area.

137 References

Abebe T, Woldeab G and Dawit W. 2013. Physiological races and virulence diversity of P. graminis pers. f. sp. tritici eriks. & E. Henn. On wheat in Tigray region, Ethiopia. ESci J. Plant Pathol. 2(1): 01-07. Afzal SN. Haque I, Ahmedani MS, Munir M, Firdous SS, Rauf A, Ahmad I, Rattu AR and Fayyaz M. 2009. Resistance potential of wheat germplasm ( Triticum aestivum L.) against stripe rust disease under rain fed climate of Pakistan. Pak. J. Bot. 41(3): 1463-1475. Ahmad S, Afzal M, Noorka IR, Iqbal Z. AkhtarN. Iltkhar Y and Karnran M. 2010. Prediction of yield losses in wheat (Triticum aestivum L.) caused by yellow rust in relation to Epidemiological factors in Faisalabad. Pak. J. Bot. 42(1): 401-407. Anteneh B. 2011. Evaluation of detached leaf assay for assessing leaf rust (Puccinia triticina Ericks and Henn.) resistance in wheat. PI. Sc. M.Sc. Thesis presented to the school of Graduate studies of Ambo. BoARD (Bureau of Agriculture and Rural Development). 2005. Progress Report on area and productivity of crops ia the region. CSA (Central Statistical Authority). 2009. Country level Agricultural products producer price index, (Ag-PPI). Addis \baba, Ethiopia. Pp 20-25. CSA (Central Statistical Authority). 2011. Report on area and crop production forecast for major crops (for private peasant holdings, meher season). Statistical bulletin. Addis Ababa, Ethiopia. Dagnatchew Y. 1967. Plant disease of economic importance in Ethiopia. Exp. Station Bull. No.50.Colege of Agri. Debre Zeit, Ethiopia: II.S.I.U. Dyck PL and Johnson R. 1983. Temperature sensitivity of genes for resistance in wheat to Puccinia recondita. Can. J. Plant Pathol.5: 229-34. Efrem B, Teslaye T, Mengistu H and Yeshi A. 1995 Breeding durum wheat for resistance to leaf rust and stem rusts in Ethiopia. In: Danial DL (ed.). Breeding for disease resistance with emphasis on durability. Proceedings of a regional workshop for Eastern, Central and Southern Africa. October 2-6, Njoro, Kenya. CIMMYT. Pp. 125-132. ENMA (Ethiopian National Meteorology Agency). 2013. Atlas of Tigray regional state. Meteorological data in Tigray: 1997-2013. Fetch TG and Dunsmore KM. 2004. Physiological specialization of P. graminis on wheat, barley, and oat in Canada in 2001. Can. J. Plant Pathology. 26: 148-55. Green GJ. 1975. Virulence changes in Puccinia graminis f.sp. tritici in Canada. Canadian Journal of Botany 53:1377-1386. Hailu G. 1991. Bread wheat breeding and genetic research in I ihiopia. In Ilailu, G., Tanner, D.G. and Mangistu H. (eds).Wheat production and research in Ethiopia: A historical perspectives. Addis Ababa. IAR / CIMMYT. pp. 73-93. Hussain S, Ilassan F and Kirmani MAS. 1980. Virulence in Puccinia recondita Rob.ex. January 1975 to December 1975. Ambo, scientific phytopathologieal laboratory. Kilpatric RA. 1975. New wheat cultivars and longevity of rust resistances, 1971-1975. U.S. Dep. Agric. Res. Serv. Long and Kolmer JA. 1989. A North American System of nomenclature for Puccinia triticina. Phytopathology 79:52.1-529. Knott DR. 1989. The wheat rusts breeding for resistance. Springer- Verlag, Berlin Heidelberg-New York- London-Paris-Tokvo. Kolmer JA. 2003. Postulation of leaf rust resistance genes in selected soft red winter wheat. Crop Sci.43:1266- 1274. Kuraparthy V, Chhuneja P, Dhaliwal IIS, kaur S, Bowden RL and Gill BS. 2007. Characterization and mapping of cryptic alien introgression from Aegilops geniculata with new leaf rust and stripe rust resistance genes Lr57 and Yr 40 in wheat. Theoretical and Applied Genetics 114:1379-1389. Long DL, Schfer JF, Roelfs AP and Roberts JJ. 1986. Vinilence and epidemiology P. recondita fsp.tritici in the United States in 1985. Plant dis.70:l 107-1110. Mamluk OF, El-Daoudi YII, Bekele E, Sotli MB, Ahmed MS, Maliir MA and Bahamish IIS. 2000. Wheat Leaf and Si:em Rusts in the Nile Valley and Red Sea Region ICARDA/NVRSRP, Cairo, Egypt. McIntosh RA, Wellings CR and Park RF. 1995. Wheat rusts: An atlas of resistance genes. CSIRO Publications. Victoria, Australia, McVey I)V, Nazim M, Leonard KJ and Long DL. 2004. Patterns of virulence diversity in P. triticina on wheat in Egypt and the United States in 1998-2000: plant disease 88: 271-279. Mengisti H, Getaneh WA, Rebeka D and Ayele B. 1991 Wheat pathology research in Ethiopia. In: wheat reseaich in Ethiopia. A historical perspective. Addis Ababa, IAR /CIMMYT. Roelfs AP 1982. Effects of barberry eradication on stem rust in the United States. Plant Dis. 66177-181.

138 Roelfs AP, Singh RP and Saari EE. 1992. Rust diseases of wheat. Concept and methods of disease management. Mexico, DF: C1MMYT.81. pp. 6. Schafer JF and Roelfs AP. 1985. Estimated relation between numbers of urediospores of P. graminis fsp. tritici and rates of occurrences of virulence. Phytopathology 75:749-750. Serbessa N. 2003. Wheat Stem Rust {Puccinia graminis f. sp. Tritici) Intensity and pathogenic variability in Arsi and Bale zones of Ethiopia. PI. Sc. M. Sc. Thesis presented to School of Graduate Studies of Alemaya University, Ethiopia. Seek M, Teng PS and Roelfs AP. 1985. The role of wheat leaves in grain yield and leaf rust losses. Phytopathology 75:1299 (abstract.) Sewalem AM, Heinz WD, Klaus P and Erich CO. 2008. Postulation of seedling leaf rust resistance genes in selected Ethiopian and German bread wheat cultivars. Crop Science society of America 48: 507-516. Singh RP. 1991. Pathogenecity variation of Puccinia recondita f.sp.tritici and P.graminis fsp . tritici in wheat growing areas of Mexico during 1988-1989. Plant disease 75:790-794. Stakman EC and Levine MN. 1962. Analytical key for the identification of physiologic races of Puccinia graminis tritici. (Processed) division of cereal crops and dis., USD A, Minnestota Agric. Exp.Sta.pp.7. Stubbs RW, Prescott JM, Saari EE and Dubin HJ. 1986. Cereal disease methodology manual. CIMMYT: Mexico. D.F. pp. 46. Torabi M, Nazari K and Afshari F. 2001. Genetics of pathogenecity of Puccinia recondita f.sp. tritici, the causal agent of leaf rust of wheat. Iran J. Agric. Sci. Abstract in English, 32: 635 Tesemma T and Bechere E. 1998. Developing elite durum wheat selections (composites) for Ethiopian peasant farm use: Raising productivity while keeping diversity alive. Euphytica 102: 323 - 328.

139 Durum Wheat Research and Achievements Shitaye Homma, Wasihun Legesse, Tebekew Damte, Zehara Mohammed, Mekuria Temtme, Ashenafi Gemechu EIAR, Debre Zeit Agricultural Research Center, P.O.Box 32, Debre Zeit, Ethiopia

Durum wheat is an indigenous crop in Ethiopia whereas bread wheat is introduced (Ayele Badcbo et al. 200()). Both crops are produced under rain fed conditions. Durum wheat grows at an altitude betweer 1800 and 2800 m in Vertisols and light soils in areas under different moisture regimes. The national average yield of wheat has reachcd is about 2.4 tons/ha in recent years (CSA 2013) which is very low compared to the rest of the world. The causes for the lower average yield are different biotic and a hotic stresses. Among the biotic stresses diseases, insect pests and weeds are the major ones where as drought, water logging, low soil fertility, and heat stresses are abiotic stresses that limit the potential of the crop.

The current expansion of local food factories increased the demand of durum wheat as a raw material for pasta production. However, the lower yield of the crop in the country and the need to meet the industrial quality is a challenge in the production and supply of durum wheat to the food processing industries. Due to this reason, the national production hardly meets the demand by the pasta producing factories, which led the country to import wheat from different countries in the world, which leads to the loss of a huge amount of foreign currency. Therefore, to improve the lower national average yield, to meet Lhe industrial quality, and to substitute import, undertaking research in priority areas would be of paramount importance.

The objective of the durum research project is to broaden the genetic base of durum wheat through collection, introduction, and hybridization, evaluate the yield performances of different genotypes under optimum/high moisture areas and low moisture areas. It is also aimed at studying the prevalence of races of rusts and reaction of commercial cultivars and level of damage under farmers' fields. The other objective is to evaluate the responses of introduced lines against rusts, select for further breeding, study, and recommend weed management practices suitable for smallholder and commercial farmers, maintaining and multiply true-to type quality seeds and promote them along improved packages.

Germplasm Enhancement, Variety Development, and Early Generation Seed Production Broadening of the genetic base through germplasm collection, introduction, and evaluation is important to have enough variability to selcct better cultivars. Crosses are made and segregating populations are evaluated to select desirable recombinants at different generations depending on the inheritance of trait of interest. Main and offseason were used to advance the generations. More than hundred F2-Fs populations were evaluated at Debre Zeit each year and selections were retained (Table 1) that showed good reaction to stem rust resistance and with phenotypic performance. Thousands of introduced durum wheat germplasm were screened against stem and leaf rust at Debre-zeit quarantine field and 425 lines with good reaction were promoted.

Table 1 Selections in a single year (2012/13) from some germplasm enhancement

Activity Locations Entries tested Entries selected Durum wheat hybridization and Debre Zeit 125 SH* and 11 113 SH and 11 selection SP SP Germplaiim introduction and screening Debre Zeit 3575 93 *SI1 refers to single head; SP refers to segregating population

Furthermore, the selections were advanced for evaluation at different stages of testing each year viz. preliminary observation nurseries, national variety, and adaptation trials. The experiments were

140 planted as multi-locations across the country depending on the breeding stage. The durum wheat research considers variety development for two moisture regimes i.e., low and optimum/high moisture environments and the testing locations represent such variability. An adaptation trial was conducted consisting of hitherto released varieties across the country and cultivars showing performance in yield and quality, disease resistant than the local varieties were rccominendcd for further use (Table 2). Seed {breeder, pre-basic and basic seed) of released varieties and elite lines (maintenance breeding) was multiplied every year required for supply to seed producers, demonstrations, pre-scaling ups and for future release or experimental purposes (Table 3). Community based seed production wras one means of addressing the seed demand where organized fanners selling the seed is means of generating incomc or farm households and the group as well.

Table 2 3rain yield performance of the durum wheat cultivars in adaptation trial at different locations

Entry Bichena Buei Chefe Donsa Denbi Debre Zeit Eteya Kokate Limu Mean Assasa 2032 4674 2350 4182.5 2337.5 3230 1750 2750 2913.3 Bakalchs 2245.5 5183 3167.5 4025 5000 4775 1500 3000 3612.0 Bichena 2294 5817.5 3477.5 2477.5 4325 3415 2250 2500 3319.6 Denbi 3690 4842 5130 2727.5 3810 2875 2500 2000 3446.8 Ejersa 2592 5525 3525 4067.5 3347.5 4295 2000 2250 3450.3 Flakit 2937.5 4304 5180 4302.5 3885 4060 1500 1500 3458.6 Ginchi 2898 4547.5 4947.5 3577.5 3432.5 4550 1750 3000 3587.9 Hitosa 2831.5 4190.5 2862.5 3182.5 2600 4130 2000 2500 3037.1 lllani 2451 2433 2635 2970 5217.5 3280 2250 2500 2967.1 Kakaba 3120 3762 3577.5 4187.5 4320 3455 1250 3500 3396.5 Klinto 2813 2333 3130 4337.5 4940 3585 2250 3000 3298.6 Kokate 3028.5 3994 5530 3895 2765 3465 2000 2750 3428.4 Lelliso 3517.5 2572.5 3552.5 3325 3095 3265 1500 2500 2915.9 Malefia 3219 4651.5 4280 2812.5 2185 4320 2500 3000 3371.0 Megenegna 3835.5 3274 3417.5 4272.5 2220 4565 2750 3500 3479.3 Mettaye 3189 2102 2977.5 4465 2165 5005 2000 2500 3050.4 Mossoto 3522.5 2340.5 3925 3590 2585 5355 1750 3500 3321.0 Obsa 2791 2807 3717.5 4095 2692.5 5530 2000 3750 3422.9 Oda 3933.5 4011 4590 3767.5 2415 3835 1500 2500 3319.0 Quamy 3801.5 3324.5 3377.5 3897.5 3065 3315 2000 2000 3097.6 Selam 3698.5 3682 3247.5 3522.5 3267.5 3820 1875 2250 3170.4 Tate 3095 3050.5 4367.5 3355 1665 3725 1750 2500 2938.5 Toltu 3344.5 3288 3977.5 3620 2757.5 4525 2500 2000 3251.6 Ude 2253.5 3834.5 3387.5 2930 3880 4490 2250 2500 3190.7 Werer 3324 4432 6750 3597.5 2457.5 4520 2000 1250 3541.4 Yerer 3482 3076.5 2892.5 2165 2132.5 4515 2500 4250 3126.7 Local 2351.2 3241.9 3121.9 3289.7 2572.1 3778.1 3740.4 2586.5 3085.2 Mean 2592.31 3418.35 3362.98 3389.9 2773.21 3876.4 3158.7 2612.2 3148 LSD (5%) 1069 1175 835 730 929 1072 685 898 458 CV (%| 30 25 18.61 15.75 24.5 20 15.86 25 30.6

Advanced lines were tested every year in national yield trials at various locations where widely- adapted, high yielding, tolerant to prevalent diseases and better than the existing cultivars will be proposed for release. Hence, two high yielding, good quality and disease resistant varieties were released in 2012 for optimum and high moisture areas namely mangudo (STJ3//BCR/LKS4/3/TER-3 ) and Mukiye (ICAJIHAN22). Other two candidate varieties (IDON-MD-2009-off/53/2009 and DSP2009_off.F4.2H.712_meh.lH.248) w'ere verified in 2014/15 and were evaluated by national variety release committee and one variety is expected for release. Demonstrations and scaling-up activities were carried out across four regions to promote the new varieties for immediate use by durum wheat farmers.

141 Table 3 Amount of seed produced (kg) in a single year (2012/13)

Variety Breeder seed Pre-basic seed Yerer 71 293 Ude 146 1566 Mangudo 263 608 Mukiye 256 505 Hitosa 81 221

Denbi 66 -

Kilinto - 105

Werer - 62

Assasa - 698

Quamy - 330

Ude - 381

DW/SR - 205

Ginchi - 589 Total 883 5563

Pathology Leaf, stem and stripe rust, stagnospora/septoria blotches, and fusraium head scab often hamper wheat production in Ethiopia. Various activities were conducted every year to deal with dynamism of the pathogens and diseases and release different options to reduce the wheat loss from major wheat diseases.

Key location durum wheat disease nursery: Annually, entries from different stages of durum wheat variety trials are assembled to form a key location durum wheat nursery (KLDN) with the objective of identifying genotypes with multiple disease resistance. Hot spots were selected i.e., Holctta ( septoria and scab), Bekoji (septoria and yellow rust), Sinana and Arsi Robe (stem rust and yellow rust), Debre Zeit (stem and leaf rust), and Meraro (Yellow rust). Ninety and 122 elite lines and commercial cultivars were tested in 2011/12 and 2012/13 cropping seasons, respectively. The highest stem rust (50S), leaf rust (90S), yellow rust (100S) and septoria index (1.44) or horizontal and vertical plant part infected (85) were noted on the check cultivars in 2011/12. Those lines or cultivars showing combined resistance were selected and the information was shared with breeders and the data used for decision making in variety promotion.

Wheat disease survey: Annually, a coordinated wheat disease survey is conducted across the major wheat producing regions of the country. The survey aimed at studying the resistance of improved wheat cultivars lo major diseases, monitoring the appearance and distribution of new rust races and observing the effectiveness of resistance genes under fanners’ fields. The wheat diseases survey was coordinated from Debre Zeit research centre. Survey routes across major wheat producing zones were identified and wheat fields were visited along the routes. GPS was used to locate the coordinates and altitudes. Different field notes were taken on wheat variety, crop growth stage, drainage condition, soil type, field size, and precursor crop. The incidence and severity of diseases were noted following standard procedures. A total of 126 fields were visited across three routes in 2011 and 2012. The three rusts, septoria, fusarium head blight, tan spot, root rot and smut diseases were noted. Fourteen stem rust samples have been collected from 14 different districts and sent to Ambo Plant Protection Research Center in 2012/13. Three stem rust races have been identified (TTKSK,TKTTF and TRTTF) from Debre Zeit mandate areas. The bread wheat variety Kubsa was susceptible to the three rust diseases. Generally, the durum wheat area is declining from lime to time, where only 6% of the total wheat area was covered by durum wheat in 2011/12. The dominant improved durum wheat cultivar was Kilinto which covered only 4% of the wheat area in the study fields.

Ethiopian wheat rust trap nursery: The Ethiopian wheat rust trap nursery is coordinated from Kulumsa research centre to monitor the prevailing virulence and effectiveness of resistance

142 genes at field levels. Annually, internationally designated differential lines, commercial cultivars and elite breeding lines are assembled and planted on-research stations across the country. Commercial cultivars and elite lines, which include stem, leaf and yellow' rust differentials were planted al selected locations for selecting best lines and commercial cultivars. Most of the bread wheat cultivars showed resistance reaction to stem rust in 2011/12. Almost all stem rust differentials showed moderately susceptible to susceptible responses to the pathogen at Debre Zeit. Under field conditions, it is often difficult to describe effective stem rust resistance genes. It seems no virulence for S ri0, Sr22, Sr24, and Sr26 at Debre Zeit conditions. Moreover, Thatcher NILs (near Isogenic Lines) exhibited susceptibility to leaf rust.

Phenotyping of international and national wheat germplasm against Ug99: Annually, large number of germplasm is sent to Debre Zeit for screening against Ug99 and other local races. The Ug99 (TTKSK) race is isolated, maintained and the purity is checked regularly in the greenhouse by inoculating appropriate differential sets. Additional bulk spores are collected from d jrum and bread wheat nurseries and temporarily stored at 4°C. Diseases assessments are taken 2-3 times using a modified Cob’s scale and data sent to partners and collaborators.

During main and off-seasons, 7984 in 2011/12 and 8338 in 2012/13 materials from different origin including bread wheat, durum wheat and triticalc have been evaluated against stem rust at Debre Zeit. Only 2-7% of bread and durum wheat materials were promoted for the next stages. About 642 germplasm have been introduced in 2012/13 for yellow rust resistance from foreign countries and evaluated at Jeldu (321 lines) and Meraro (321 lines). Eighty percent of the materials were susceptible to yellow rust and the rest were resistance/moderately resistance to yellow rust. About 3500 genotypes were screened against rust diseases during the 2013/14 off-season and 177 were selected and promoted. Moreover, survey was undertaken for diseases in 14 districts of central and Eastern Shcwa. The survey results showed highest incidences of leaf rust, yellow rust and septoria w'ith a record of 90, 70 and 74, respectively. Three races of stem rust (TTKSK, TKTTF and TRTTF) were identified. Under key location disease nursery, 35 elite lines of durum wheal showed combined resistance to major diseases across all key locations. Landraces collected from different locations w'erc evaluated for their reaction to yellow' rust at Meraro and Jeldu. Out of 361 collections, 22 of them have been identified as sources of resistance to yellow rust that will be used for future breeding program.

Maintenance of stem rust races: The three stem rust races on bread and durum wheat (TTKSK, TKTTF and JRCQC) were maintained in greenhouse.

Table 4 Entries tested at Debre Zeit (2011/12)

Oigin Bread wheat Durum wheat Triticale Total CIMMYT 1540 3258 4798 USD A 1788 1788 ICARDA 17 192 209 Canada 114 114 IBC 463 463 USA-NDU 90 90 Australia 522 522 Total 2020 5850 114 7984

Weed management Weed, if not managed timely, could highly affect crop growth and result in a huge yield reduction. The method of management could be cultural (hand weeding) and chemical or mechanical. People have different traditions and belief about w'ecd control. A study on weed management showed that both adopters and non-adopters of weed management reported the equal importance of broadleaf and grassy weeds in their wheat fields (Bekele et al. 2000). About 95% of adopters and 93% of non adopters controlled weeds by hand, whereas 87% of adopters and 61% of non-adopters used

143 herbicides. The average number of hand weeding for adopters and non-adopters was 1.7 and 1.8, respectively. Only 18% of adopters and 5% of non-adopters hand weeded based on the recommendation of Oromia Agricultural development Bureau. The percentage of adopters and non adopters who were aware of the recommended weeding frequency was 42% and 33%, respectively. The main constraints to frequency of hand weeding for adopters were labor shortage (74.5%) and lack of cash to hire labor (59.6%). Similarly, the main constraints for non-adopters were labor shortage (68.2%) and lack of cash to hire labor (60.6%). The shortage of labor and the high cost of labor was the main problem in managing weeds. Therefore, chcmical control of weeds could be one of the alternative options. Appropriate chemicals, its rate, and time of application should be studied for effective weed control. Hence, the potential of herbicides was studied on broad and grass weeds management of problematic weed species of durum wheat.

Entomology Pests of wheat in East Shewa and North Shewa Zones: Wheat fields were surveyed in September 2011 in East Shewa and North Shewa /ones (Amhara Regional State) with the objectives of determining the current species composition, distribution, and measuring the relative importance of wheat insect pests, and identifying natural enemies associated with wheat insect pests. Wheat fields were randomly selected at the interval of 5 to 10 km along the sides of roads and inspected in a crossed diagonal line to detect any insect pest infestation. The stage of the crop was between Zadoks’ 17 (seven leaves unfolded) in relatively high altitude areas and 37 (Hag leaf just visible) ia the lower altitude areas. However, two wheat fields around Tulurea were at flowering (Zadoks’ 61) stage. The tef epilachna, Chnootriha similes (synonym Epilcichna similes) (Coleoptera: Coccinellidae) was the only insect pest found in few wheat fields. Both the larvae and adults feed by scrapping the upper leaf surface; however, the overall infestation level was very low. Symptom of shoot fly damages i.e. dead heart was encountered only in two wheat fields in Gogle (Mojo) and Ejere areas. Moreover, localized rodent damage (probably by Arvicanthis sp) was recorded around Bolosilas e and Bologiworgis areas in North Shewa zone of the Amhara regional state. In these areas, crops are fenced with stone wall or live fences which usually harbor rodents and other small vertebrate animals.

Storage pests: Four routes were used in 2011/12. Villages were selected at 10 to 15 km interval, three to five households were randomly selected and from each household 100 to 150 gm of wheat sample was collected, and incubated at 20-25°C for three months. After three months of incubation period, the wheat samples were inspected for the presence or absence of storage insect pests. Live insect pest was not detected in any of the samples after three months of incubation period. Seed of Lolium sp., Avena fatua and Phalaris paradoxa was found in 84.6% of the samples. The weight of these grass weed seeds per 100 g of wheat grain ranged from 0.02 to 8.8 g. About 116 farmers were interviewed. Most farmers store wheat in sacks but few farmers in Minjar-Shenkora area use traditional gotera to store wheat and other crops.

Yield loss caused by Russian Wheat Aphid: The amount of yield loss in irrigated wheat caused by Russian wheat aphid (RWA) was estimated on black soil at Debre Zeit Agricultural Research Center. Wheat varieties Werer, Denbi. Kubsa, and Millennium were sown at the rate of 140 kg ha'1, under insecticide protected and unprotected conditions. There was non-significant difference (p< 0.05) among varieties and variety by insecticide interaction in plant height, number of grains per spike, hundred seed weight and grain yield. However, insecticide application significantly (p < 0.001) affected all the aforementioned variables. In 2012 season, the RWA population was high on unsprayed wheat i.e., 47, 56, 60, and 64 RWA per tiller on Denbi, Werer, Kubsa, and Millennium, respectively. There were at most two aphids per tiller on sprayed wheat. Percentage of infested tillers in unsprayed wheat was greater than 95%. The amount of yield loss varied between years. The yield loss was 69, 80, 90 and 93 on Werer, Denbi, Kubsa and Millennium, respectively, in 2012 off-season. The corresponding percentage values in 2013 off-season were 15.4, 30.7, 40.9 and 38.9.

144 Soc o Economics and Extension Baseline information of wheat in East Shewa Three of the major challenges facing the Eastern Africa sub-region are rising level of poverty, food insecurity, and high rate of unemployment (especially among the youth). To alleviate these challenges, the region has to clearly articulate its development agenda covering all key sectors of the economy. One of the key requirements of EAAPP is the generation of baseline data for each country which can be used for subsequent monitoring and evaluation. Therefore, baseline information was generated on wheat commodity in selected sites of Ethiopia as a starting point for the action research program on wheat based farming system in East Showa zone. The specific objectives were to identify and document current wheat production status, potentials and constraints as well as farmers' techno,ogy needs; and to provide socio-economic baseline data and information for project's future monitoring and evaluation activities at East Showa zone, Lumc district in Oromia Regional State. The district; was selected on the basis of its large wheat production potential, number of growers, potential for wheat production, accessibility, and representativeness of the farming system.Both primary and secondary sources were used to collect the relevant data. Primary data were obtained from randomly selected farm households. A structured questionnaire was used for the baseline survey. Using primary source, information vital to household's demographic and socio-economic characteristics, was obtained by interviewing the randomly selected farm households. Secondary data were collected from the district agricultural office, Cooperative office, and finance and Economic development office in 2012.

Wheat production status: 98% of the farm households participated in wheat production in 2010 cropping season where all respondent engage in the cultivation of at least one wheal variety. Some (39%) sample farm households participated in improved wheat seed production in producing 13.3 qu ntal of wheat seed with average pricing of 581.2 birr per quintal. Fanners grew various wheat varieties, malt and food barley, tef, faba bean, field pea, chickpea. Wheat covercd 1.01 ha of land while other crops such as tef covered 0.95 ha, field pea (0.06 ha), food barley (0.18 ha) and malt barley (0.27 ha), chickpea (0.68 ha), lentil (0.32 ha) and grass pea (0.38 ha) hectares. Wheat was second productive major cereal crop in the study district with mean productivity of 26.1 quintal h a 1. Variety preference and adoption: Two bread wheat varieties were grown in Lume district in 2010. Kubsa (HAR 1685) and Digalu were the most commonly grown varieties where Kubsa wheat variety had large area (0.76 ha) allocation followed by Digalu variety which had good resistance to wheat rusts. The possible explanation for Digalu expansion was its yield traits, disease tolerance, white seed color, good quality for food and market.

Marketing of crops and decision making in household: Large farm households (85%) participated in marketing of wheat production and sold their wheat for traders (84. 1%), consumers (5.9%); 72% of the sample fanners sold their crop produce in local markets and the remaining was sold in the secondary markets.

The marketed surplus for wheat was 1458 kg and wheat was the sccond crops in terms of quantity sold in the markets next to Tef crop, which was 1600 kg. The proportion of wheat farmers involved in marketing was very high as compared to other crops i.e., 84.1% of famers participated in wheat marketing. This implies that wheat has key role in generating income (cash crop) for farmers in the study area. The mean income of fanners from the sale of wheat was 8190 birr.

The m ajor mode of transport to market wheat was domestic animals in which 83.6% used donkeys, 4.5% used public means to transport with the average transportation cost of 36 birr. The rest (13.4%) sold by carrying either on head or head-load or walking on foot (trekking) taking an average distance of 6.6 km to the main market. The income from wheat covercd household expenses such as consumable product (45.7%), purchasing fertilizer (14.8%), land renting (7.4%), tax (12.3%), and saving (2.5%).

145 Cropping pattern, input use and yield: Wheat was grown in pure stand cropping system i.e., 100 % of the respondents grew wheat in pure stand, and 98% of the sample households used chemical fertilizer (DAP + Urea). The average seed rate for wheat amounts to 120.1 kg ha'1. Herbicidcs such as 2, 4-D were used for the control of broad leaf weeds by 38% of fanners and the rest used hand weeding and fertilizer application using family members (women and children) and hired labors. The fertilizer was applied as top dressing with the average application rate of 14.9 kg ha' 1 and 15% of sampled farmers applied manure in their wheat fields with average application rate of 1965 kg ha'1.

Analysis of impact of improved wheat production on farmers’ income: Gross margin analysis was used to know the impacts of wheat crop production on wheat producer farmers’ income level, i.e., evaluation of effect of adoption of improved wheat production and profitability of the wheat crop production by considering the variables costs of all inputs, seed amount, chemical amount and costs and the gross income obtained from a hectare of wheat crop production.

Gross Margin = Gross income - Variable cost; Gross Income - Average yield/area x Price/unit

The gross margin was calculated by considering the selling practice of each farmers and average market price of input and labor to estimate the cost and return per hectare of wheat crop production. The price of output and input was taken from each plot of wheat land and average price was taken to arrive fcr the input and output price for wheat crop production. Thus, from land preparation to harvestirg of wheat crop production on average it required 64 person days ha'1. Wheat harvesting was done manually using sickle by family labor and in some cases by hired labor. Manually harvested wheat was threshed by using farm animals such as oxen, horses, and donkeys. The average labor demand for harvesting wheat is 29.4-person days ha 1 with average cost of 1080 birr.

Wheat seed supply system: 66.7% of farm household used their own seed for planting, 16.7% from cooperatives and 16.7% from MOA. The sample farmers (93.8%) replace their wheat seed for major variety within 3 years and the remaining within 2 years.

Characterization of Seed Demand

The study was carried out in Aleltu and Minjar Shonkara districts. Aleltu district is moderately productive and self-sufficient in food crops. Crop production is rain-fed. The most important crops for sale are teff, wheat, barley, maize, and fababean. Minjar Shonkara district is located in North Shewa of zone of Amhara regional state. There is only one. meher harvest and the main crops planted are wheat, teff, sorghum, chickpea, lentil and vegetables.

The total area covered with improved wheat seed combined with fertilizer during 2010 was 107.1 ha with the maximum plot of 3.5 ha and with mean of 0.9 ha. The maximum amount of improved wheat seed used with fertilizer was 18390 kg with the maximum amount of 450 kg seed and with the mean of 154.5 kg seed. The total amount of yield obtained from the use of improved wheat seed combined with fertilizer was 256195 kg with the maximum yield of 9000 kg ha'1 and mean of 2152.89 kg h a1. Area covered with local wheat seed combined with fertilizer was 7.25 ha with the maximum plot of two ha and mean of 0.07 ha. The local wheat seed used combined with fertilizer was 835 kg seed with the maximum amount of 200 kg seed and mean of 9.48 kg seed; the yield obtained from the use of local wheat seed was 7600 kg with maximum yield of 3000 kg with mean yield of 86.4 kg.

The use of fertilizer combined with local and improved seed was the main mode of modem technology adoption in cereal production. Both improved and local wheat crops were not produced without fertilizer the study area; 99.2% of the respondents explained that yield of improved was better than the local wheat varieties during 2010. Only 0.8% of the farmers responded that improved seed is

146 not better yielder than the local wheat variety. The farmers in revealed that 95% of them were using improved wheat seed for the last five years where as only 5% used local variety. Different factors hinder not to use improved wheat seed i.e., 20% of farmers indicated high price impede them not use improved variety while 10% revealed that poor quality improved wheat seed when they purchase from suppliers. Thus, it is important to assess seed quality standard of the supplier. Others (33.3%) of the farmers mentioned that lack of finance hinder them not to use improved varieties for the last five years, 9.1% mentioned untimely supply of improved seed, 11.5% lacked trust on the seed quality of suppliers and the remaining did not use owing to lack of awareness and high price of related packages such as fertilizer.

Improv’d wheat variety preference is often thought to derive from better yield, nutritional value, disease resistance, and market demand. Samples farmers (41.9%) preference for improved variety depended on yield potential, 20.9% explained their preference depended on market demand, and 14.3% on disease resistance, 19.2% on feed value for their cattle and only 3.8% on nutritional value. Accordingly, 64.3% of the respondents preferred HAR1685 (kubsa), 16.1% preferred ET13 and 7.9% preferred Pavon, while 11.6% preferred Digalu. Farmers (84.4%) indicated initial seed source of improved varieties was services cooperative union through purchases. Some seeds were also obtained from Debre Zeit agricultural research centers while remaining from neighboring fanners through seed exchange and cash from the local market.

Market demand is the total volume of a product that will be bought by consumers, at a certain period, in a specific location. Among the factors that affect the demand for improved wheal seeds are farmer s perception of the yield or quality advantages of improved to local seeds, the price of seed, prices of other inputs, relative price of crops, farmer’s forecast of weather conditions, and the cost of reaching/ distribution/retail outlets. Demand for improved seeds can also be influenced by the effectiveness of promotional campaigns, the efficiency of distribution, and the availability of crcdit and other complementary inputs such as fertilizer. Farmers (66.7%) wanted to purchase improved seed in the month of June and others (35%) preferred to purchase in May. Farmers purchased on average 119 kg of improved wheat seed with maximum amount of 450 kg. Fanners (76.5%) preferred package size of 50 kg from seed sources. Respondent farmers (48.2%) indicated that purchase pricc of improv ed seed was affordable and 41.2% explained the price was expensive. Farmers (25%) indicated that purchased seed at an interval of two years, 42.9% at intervals of 3 years, and 14.3% at an interval of 5 years. Respondents (71.9%) presented their improved wheat seed demand to development agents working in their respective peasant associations, 24.5% presented to farmers’ cooperative unions, and 25% submitted to district bureau of agriculture. Reasons for changing variety was susceptibility to disease, changing to new variety, and seek for planting material. Respondents (57.3%) got improved wheat seed as per their request, 39.3% cannot get as their request and only 3.4% of the farmers never request for improved wheat seed. Many fanners (81.8%) explained that the most likely reason for farmers not to get improved wheat seed regularly was due to supply shortage, high price (18.2%) and untimely supply.

Many farm households (86.6%) did not change their plan of improved wheat seed request during the surve\ period, 10.9% changed their plan of improved wheat seed request, and 2.5% never plan for improved wheat seed request. About 50 % of the respondents change their wheat seed demand temporarily owing to change of trend of rainfall, 41.5% owing to high price of seed and 8.3% due to limitation of supply.

Farmers (94.2%) gave positive response to improved wheat seed they purchased and express that the wheat seed has got good quality where as 5% of the sampled fanners explained that the improved wheat seed had poor quality in 2010; 80% of the respondents purchased poor quality seed from service cooperative union while 20% bought poor quality wheat seed from their own exchange. The criteria to determine good quality seed was purity from admixture, diseases free (discoloration, fungal bodies, and odor) and proper field emergency.

147 Farmers (95%) stored their wheat seed for next planting season; 71.9% of the farmers stored their wheat seeds with bags, 17.5% store in a bin and 9.5% stored in earthen pots and sacks. Most (42%) respondents stored wheat seed for more than six months to one year, 26.1% from three to six months, 10.1% from two to three months and 20.2% stored for less than a month.

Saving seeds from the current harvest to plant the following season is the traditional practice. It works especially well for self-pollinating crops such as wheat. The farmer can continue to grow same variety year after year without spending money on new seeds. A wealthy farmer may choose to try new varieties from time to time, but an impoverished farmer will buy once and save seeds from year to year. Farmers (97.5%) saved wheat seed for next planting season, 29.9% saved for 3 years, 23.9% for 2 years and 29.1% for 1 year and 9.4% saved their wheat seed for a 4 to 7 years.

Most farmers (94.1%) sold their wheat grain for traders; 80.7% of their wheat grain to nearby markets while 18.5% delivered to their grain to cooperative stores and 74.2% of the respondents sold wheat grain at any time for family needs. Most farmers confirmed that there was no association between wheat gnin and seed price in study areas.

Farmers raise many problems associated with the wheat seed supply system i.e., insufficient wheat seed supply (30%), unavailability of new improved wheat variety (24.2%) and its insufficient introduction. Farmers (22.4%) also specified susceptibility of the existing cultivars to rust, high price of related input (fertilizer), poor quality seed supply, and untimely supply as major constraints that impede the production of wheat.

Pre-extension, Demonstration, Evaluation and Popularization

Demonstration and evaluation of rust tolerant, high yielding, and end use quality improved durum wheat varieties along with their updated agronomic practices with due participation of farmers and other stakeholders was conducted in 2013/14. The durum wheat research program in collaboration with research extension and other stakeholders was working on demonstration, evaluation and popularization of high yielding, rust tolerant and end use quality improved durum wheat technologies in 14-selected durum wheat-growing districts of Ethiopia. To avoid provision of seed free; the seeds were given to farmers on revolving basis. This was complemented by signing of the Memorandum of Understanding (MoU) between farmers, researchers, and district agricultural office representatives:

• Five high yielding and quality durum wheat varieties (Mangudo, Mukiye, Denbi , Hitosa and Werer) were popularized and demonstrated in a total of 14 selected districts of Ethiopia, where 4 districts from Oromia, 1-in Tigray, 4 in Southern and 5 from Amhara regional states.; • Eighteen quintals seed of four improved durum wheat varieties (Mangudo, Mukiye, Ude, and Assasa) were distributed for the pre-scaling up and demonstration activity during 2013/2014 cropping seasons. In the pre scaling up and demonstration activities, 27 and 78 farm households were participated for the 2013/2014 cropping seasons, respectively; • Training was given on quality seed production and management to 107 community based seed growers, DAs, and districts agricultural experts of which 10.28% (11) of them were female participants; • In the dissemination and diffusion of durum wheat technologies, 140 farmers, researchers, extensionists and agricultural experts were participated on PVS in Lemo and Aleltu districts. Two durum wheat varieties (Mangudo amd Mukiye) were selected for their siiperioritv in quality, yield and other agronomic traits; and

References Ayele Badebo, Solomon Gelalcha, Ammar K. Nachit M.M and Abdella O. 2009. An overview of durum wheat research in Ethiopia: Challenges and prospects. CSA (Central Statistical Agency). 2013. Agricultural sample survey 2013/14. Report on area Production of Crops of Ethiopia.

148 Yield and Yield Stability of Bread Wheat Genotypes in Lowland Irrigated Areas Desta Gebre1, Mihratu Amanuel1, Hailu Mengistu1 and Beakal Tadesse1 1 Ethiopian Institute of Agricultural Research, Werer Agricultural Research Center Department of Irrigated Cereal Crops Breeding P. 0. Box 2003, Addis Ababa, Ethiopia

Abstract Nine wheat genotypes introduced from ICARDA and eleven genotypes collected from National Crossing program (NCP Kulumsa), were evaluated separately side by side at Werer Research Center from 2010/11 to 2013/14, at Gewane from 2011/12 to 2012/13, at Amibera in 2013/14, and at Mehoni and Kobo in 2013/14 cropping seasons with the objective of identifying widely adaptable genotypes. The combined analysis of variances showed that the effects of year and locations were highly significant for all parameters studied in both experiments. The combined ANOVAs of grain yield revealed significant (P<0.05) genotype by year interactions for ICARDA genotypes but non-significant for NCP genotypes. The environment effect accounted for 52.3 % of total sum of squares (TSS) for ICARDA genotypes and 46.2 % for NCP genotypes. The GE accounted for 17.9% of TSS for ICARDA lines and 9.6% for NCP lines. The highest yielding genotypes from NCP were ETBW5955 (3268.3 kg h a 1), ETBW5963 (3226.1 kg ha1), ETBW5901 (3208.2 kg h a1) and ETBW5898 (3164.3 kg ha'1) and from ICARDA materials GONGGLASE-4 (3102.8 kg ha1), SABA/FLAG-1 (3027.3 kg" ha'1) and i n IQUE 96/FLAG-l CROW’S7BOW’S’-3 (2973.6 kg ha'1). Generally, the genotypes derived from the NCP possessed higher yield, better thousand kernel weight and were early maturing type in comparison with the ICARDA materials.

Introduction

The productivity and annual production of rainfed wheal docs not satisfy the wheat demands of the population, as a result the country is forced to import wheat grain from abroad with high foreign currencies. On the other hand, the country has 12 river basins with total areas of nearly 4 million cultivable lands suitable for irrigation (MoA 2011). The previous research works on wheat genotypes screening conducted by Werer Agricultural Research Center (WARC) from 1969/70 up to 1986/87 clearly showed that the yield of wheat genotypes could reach up to 4400 kg ha 1 (Mohammed 1994) indica.ing the suitability of lowland irrigated areas for wheat production. Currently, the government has given due attention for irrigated agriculture and many dams have been constructed for production of different commercial and food crops with the purpose of attaining food security and earning foreign currencies by exporting agricultural products to abroad. This has created a good opportunity for researchers to begin and work on development of stable and well adapted wheat varieties along with their production technologies for irrigated lowland areas of Ethiopia. Thus, the irrigated wheat research and development works was reinitiated after 20 years in 2006 with the aim of developing well adaptable and stable wheat varieties tolerant to heat and salinity along with their appropriate production technologies for irrigated lowland areas of Ethiopia.

Evaluation of genotypes at a number of test environments provides useful information to identify their yield performance and yield stability. Naser et al. (2012) quoting Gauch (2006) stated that in addition lo yield performance of bread wheat cultivars, it is essential that the yield stability of such cultivars be determined in order to make specific selections and recommendations to farmers. Similarly, Naser et a l (2012) quoting Ramagosa and Fox (1993) pointed out that measuring genotype by environment interaction aids to determine an optimum breeding strategy lo improve for specific or general adaptation strategy, which is related to the expression of yield stability under a limited or wide range of environments.

149 Therefore, the objective of this paper is to present the findings of the investigation on yield and yield stability of bread wheat genotypes evaluated under irrigation in different environments.

Materials and Methods

Nine wheat genotypes introduced from ICARDA and eleven genotypes collected from National Wheat Research Coordinating Center (Kulumsa), were evaluated separately side by side at Werer Research Center from 2010/11 to 2013/14, at Gewane from 2011/12 to 2012/13, at Amibera in 2013/14, and at Mehoni and Kobo in 2013/14 cropping seasons. The genotypes were evaluated using RCBD in three replications with the plot size of 10 rows of 3m length and 0.3m between rows (9nr). The planting dates were 12 November for Werer and Amibera, 14 November for Gewane, 15 December 2013 for Mehoni and 10 January 2014 for Kobo. Sowing was done by hand drilling at a seed rate of 80 kg ha'1 and DAP fertilizer was applied at ihe rate of 50 kg ha 1 whole at sowing time at all locations. DAP was applied five cm away from the seed and covered with soil followed by irrigation water. On the other hand, UREA fertilizer was applied at all locations in split at the rate of 100 kg ha 1 (half at tillering and the remaining half at booting stages). It was applied in both times after irriggting the trials around 5pm to minimize the loss of N due to leaching and volatilization, respectively. Both trials were irrigated every ten days intervals at Werer, Amibera and Gewane, and every seven days intervals at Kobo and Mehoni until physiological maturity time of the genotypes. The trials were weeded four limes using hand weeding and no herbicides were used. Data were collected before harvest from inner rows on phenotypic parameters such as days to heading (DH), days to maturity (DM), plant height (PH), spike length (SL), total tillers (NTT) and effective tillers (NET). The central eight rows of each genotype were harvested manually using sickles. After harvest, data on number of spiketes per spike (NSPS) and number of seeds per spike (NKPS) were taken. The harvested genotypes from the central eight rows of each plot were threshed and cleaned to determine 1000 kernel weight (TKW) and grain yield per plot (GY). All the collected data both from the field and laboratory were compiled and subjected to statistical analysis using SAS Version 9.0 software.

Results

Genotype by environments interactions The comb ned analysis of variances for both experiments is presented in the table 1 and table 2. The effects of year (Y) and locations (L) were highly significant for all parameters studied in both experiments except for TKW in the second experiment. The combined ANOVAs for ICARDA genotypes revealed highly significant (P<0.01) G x Y interactions for DM and significant (P<0.05) interactions for PH, SL and GY (table 1). Moreover, highly significant (p<0.01) G x Y interaction effects were observed for DH, DM, SL, NTT and TKW but non-significant for PH, NSPS, NKPS and GY in the materials derived from NCP (table 2). Highly significant (P<0.01) L x G interactions were observed in ICARDA genotypes for DM, PH, NET, SL, NKPS and GY. In genotypes derived from NCP, highly significant (P<0.01) L x G interaction effects were observed only for DH, DM, NTT and NET but non-significant for PH, SL, NSPS, NKPS, TKW and GY (NCP. The environment (E) effect was accounted for 52.27 % of total sum of squares (TSS) for ICARDA materials and 46.21 % for materials obtained from NCP. The GE was accounted for 17.92% of TSS for ICARDA materials and 9.63 % fcr materials obtained from NCP. Both the environmental and genotypes by environment interaction effects was higher on ICARDA materials as compared to the materials derived from NCP (fig. 1 and 2). The results also showed that there were crossover interactions in both trials. The stability analysis showed that among ICARDA materials, UTIQUE 96/FLAG-l CROW’S’/BOW’S1- 3, SABA/FLAG-1 and GONGGLASE-4 and NCP materials, ETBW5901, ETBW5955, ETBW 5510 and ETBW5963 were high yielder and stable over locations. The results of stability analysis showed more stab e over years for ICARDA materials such as GONGGLASE-4, SABA/FLAG-1, UTIQUE 96/FLAG-l CROW’SVBOW’S’-3, PBW343 and GIZA-168//SHUHA’S7DOBUC’S’) and for NCP materials ETBW5955, ETBW5510, ETBW5898 and ETBW5963 (fig 3 and 4).

150 Table 1 Combined ANOVA over years and locations for ICARDA materials (2010/11 to 2013/14)

Mean Square sv* df DH DM PH NTT NET SL NSPS NKPS TKW GY Year (Y) 3 ** ** ** ** ** ** ns ** ** ** Location (L) 4 ** ** ** ** ** ** ** ** * ** Replication 2 ns ns * ns X ns ns ns ns ns Genotyoes (G) 9 ** ** ** ns ns ** ns **** i t * Y xG 27 ns ** * ns ns * ns ns ns * L xG 36 ★ ** ** ns ** ** ns ** ns ** * SV - Sources of variation; df- degrees of freedom; DH-Days to heading, DM - Days to Maturity, PTI - Plant Height, SL - Spike Length, NTT -Total Tillers, NET - Effective Tillers, NSPS - Number of Spikelets per Spike, NKPS - Number of Seeds per Spike, TKW - Thousand Kernel Weight and GY- Grain Yield i able 2 Combined ANOVA over years and location for NCP genotypes (2010/11 to 2013/14)

Mean Square SV df DH DM PH NTT NET SL NSPS NKPS TKW GY Year 3 ** ** ** ** ** ** ** ns ** Location 3 ** ** ** •kit ** ** ** ** ** ** Replics:ion 2 ** ** ** ** X* ns ns * ns ns Genotypes 11 ** ** ** ** * ** ** ** ** ** ** Y x G 32 ** ns ** * ** ns ns ** ns ** ** L xG 31 ns ** ** ns ns ns ns ns

6000 -1 ETBW5 898

ETBW5

ETBW5 881 ■4 ETBW 5510 Fig 1 Interactions of NCP genotypes to the Environments (G x E)

Fig 2 Inte"actions of ICARDA genotypes to the Environments (G x E)

151 II ftiM <•» i Itll.lf IK I liill fm « litlict

2 * ! I I 1 " 1 J - I 1 “ I 5 % I I_____ o • 3 •i; •i t

•I M •I 5 - i l l -l> -»1 II IS II II III 'Min I I5F ID I iclit I (5? it) Fig 4 GGE biplot of stability analysis of NCP materials across Fig 3 GGE bi| ilot of stability analysis of location ICARDA mat trials across location (1- Werer, 2 - Gewane, 3 - Waidulal and 4 - Kobo) (1 - Werer 2 Gewane 3 - Waidulal, 4 - Mehoni and 5- Kobo)

Genotypic variability The combined analysis of variances for both experiments showed highly significant differences (P<0.0l among the ICARD’s genotypes for DH, DM. PH, SL, NKPS, TKW and GY. The combined ANOVA showed highly significant differences (P<0.01) for almost all parameters among genotypes derived from the NCP. The overall mean values of the traits studied are presented in table 3 and 4. The mean values for ICARDA materials ranged from 50.8 to 63.5 days for DH, 83.6 to 92.6 days for DM, 61.2 to 69.9 cm for PH, 6.4 to 7.8 for NTT, 6.2 to 7.5 for NET, 6.7 to 7.7 cm for SL, 13.4 to 14.5 for NSPS, 23.0 to 29.9 for NKPS, 29.4 to 36.3 g for TKW and 2346.0 to 3102.9 kg h a 1 for GY (Table 3). Similarly, the mean values for genotypes derived from NCP varied from 44.9 to 57.7 days for DH, 73.5 to 88.6 for DM, 66.1 to 78.2 cm for PH, 5.1 to 6.6 for NTT, 4.8 to 6.4 for NET, 6.2 to 9.0 cm for SL, 13.21 to 16.8 for NSPS, 31.4 to 41.3 for NKPS, 28.1 to 36.2 g for TKW and 2409.9 to 3268.3 kg for GY (Table 4). The highest yielding genotypes from NCP were ETBW5955 (3268.3 kg ha'1), ETBW5963 (3226.1 kg ha'1), ETBW5901 (3208.2 kg ha'1) and ETBW5898 (3164.3 kg h a 1) with the maturity date of 85.2, 76, 80.1 and 78.9 days, respectively. These promising genotypes had better TKW (Table4). The top performed genotypes among ICARDA materials were GONGGLASE-4 (3102.8 kg/ha), SABA/FLAG-1 (3027.3 kg/ha), UTIQUE 96/FLAG-l CROW’S7BOW’S’-3 (2973.6 kg ha'1) and 1994/95//TEVEE’S/TADINIA (2727.8 kg h a1). These genotypes had 31.6, 34.7, 33.4 and 36.3g TKWr and 83.8, 91.4, 86.2 and 83.6 maturity days, respectively. The yield performances of the NCP genotypes were better as compared to the ICARDA materials.

Table 3 Average performance of ICARDA's genotypes under different irrigated environmental conditions (2010/11 -2013/14)

3enotypes DH DM PH NTT NET SL NSPS NKPS TKW GY (cm) (cm) (9) (kg h a 1) PBW 343 57.0 86.5 69.7 7.1 6.8 7.4 14.3 27.3 34.3 2830.0 SANDALL-3 57.1 87.0 68.3 7.3 6.2 7.3 13.8 26.8 31.7 2879.8 NABUK-6 63.5 92.0 61.2 6.8 6.4 6.8 14.5 27.2 29.8 2346.0 SABA/FLAG-1 58.6 91.4 66.7 7.8 7.5 7.4 14.1 23.0 34.7 3027.3 QAFZAH-3 3/FLORKWA-2 63.2 92.6 65.0 6.6 6.2 7.5 14.0 27.7 30.5 2451.3 G1ZA-168//SHUHA 50.7 85.9 67.2 6.4 6.3 7.1 13.6 25.2 32.3 2800.9 'S’/DOBUC’S GONGGLASE-4 51.4 83.8 69.9 7.1 6.9 7.7 14.2 29.9 31.6 3102.9 GA'AMBC 60.9 91.3 68.7 7.4 6.9 7.2 14.4 24.0 29.4 2528.8 UTIQUE £6/FLAG-1 54.4 86.2 66.1 6.7 6.5 7.5 13.5 25.1 33.4 2973.6 ROW’S7BOW’S'-3 1994/95//TEVEES/TADINIA 50.8 83.6 68.5 6.4 6.3 6.7 13.4 26.2 36.3 2727.8 Mean 56.8 88.0 67.1 7.0 6.6 7.3 14.0 26.2 32.4 2766.8 CV (%) 13.7 3.5 6.8 35.0 25.8 8.5 10.8 12.6 8.7 24.1 CD (0.05) 4.9 1.9 2.9 1.6 1.1 0.4 1.0 2.1 2.0 424.6

152 Table * Average performance of NCP genotypes under different irrigated environmental conditions (2010/11-2013/14)

Gerotypes DH DM PH NTT NET SL NSPS NKPS TKW GY (cm) (cm) (9) ( kg ha-1) ETBW5898 53.2 78.9 68.2 5.23 5.23 7.72 14.32 32.16 35.86 3164.3 ETBW5877 53.2 81.5 73.8 5.44 5.36 7.71 15.30 35.17 36.15 2768.6 ETBW5881 57.7 86.9 77.9 5.30 5.16 6.22 14.02 34.59 31.82 2409.9 ETBW5510 53.6 80.1 66.9 5.51 5.37 7.75 14.63 35.59 32.05 3096.0 ETBW5556 53.1 88.6 67.8 6.63 6.38 9.00 16.75 41.25 28.13 3064.8 ETBW5901 51.8 80.1 71.8 5.84 5.70 7.39 13.87 33.96 34.46 3208.2 ETBY/5942 49.2 77.4 67.6 5.37 5.09 7.05 13.65 32.91 32.03 3061.2 ET6//5954 57.5 85.4 76.3 5.37 5.30 8.62 15.18 34.26 35.74 2796.6 ETBW5955 56.9 85.2 78.2 5.89 5.87 8.32 15.85 35.47 34.89 3268.3 ETB A/5958 44.9 73.5 66.1 5.60 5.46 6.48 13.21 31.41 35.02 2534.9 ETB A/5963 48.9 76.0 72.2 5.30 5.30 7.45 14.35 35.37 35.94 3226.1 Ga’ambo 52.7 80.9 76.3 5.06 4.83 7.85 14.13 34.82 35.29 2980.1 Mean 53 81 72 5.5 5.4 7.6 14.5 34.5 34.1 2958 CR 5%) 1.03 1.6 4.8 0.93 0.64 0.58 1.67 3.95 1.99 549.6 CV% 1.8 2.1 7.5 13.6 14.9 8.8 13.2 12.9 6.0 22.2

Discussion

The results showed that the performances of genotypes evaluated under both trials were significantly influenced by the test environments (years and locations), which proved the presence of wide variations among the test environments where the genotypes were evaluated. The environment effect accounted for 52.3% of TSS for ICARDA materials and 46.2% for materials from NCP which agreed with similar finding of Karimizadeh et al. (2012) and Hinsta et al. (2011) that testing environments are relevant. The ICARDA materials were much more influenced by the environmental factors (soil type, soil fertility and temperatures) in comparison with NCP genotypes. The probable reason may be latitudinal effect as the ICRADA materials were developed at higher latitudes. The GE accounted for 17.9% of TSS for ICARDA materials and 9.6% for materials from NCP showing that the genotypes responded differently to the test environments. The effects of L x G interactions for ICARDA and NCP materials significantly varied showing that there was high genetic variability between the ICARAD and NCP materials. Generally, the environment effect relative to G effect was higher in the presen: study which may create the selection process more complex. The higher GE interaction effect relative to the G effect for grain yield of genotypes in ICARDA materials indicated that the genotypes exhibi ed both additive and crossover type of interaction. Similar findings were reported by Ezalollah et al. (2012), Naheif (2013) and Naser et al. (2012). This implies that multi-location trials are imporlant to identify and select high yielding and stable varieties for wide as well as specific environments.

Generally, the mean yield performances of the materials derived from the NCP were relatively higher and more stable as compared to the ICARDA materials over years and across locations (Fig 1 and 2). It may be attributed to the genetic materials inherited from the parental lines as they arc more adaptable to tropical areas (latitudinal differences). The stability analysis over locations showed that among ICARDA materials UTIQUE 96/FLAG-l CROW’S’/BOW’S’^ , SABA/FLAG-1 and GONGGLASE-4 and among NCP materials ETBW5901, ETBW5955, ETBW 5510 and ETBW5963 were high yielding and stable. It was revealed that there is a possibility of selecting widely adaptable candidate varieties for all locations from both ICARDA and NCP genotypes. The results of stability analysis over years showed that among ICA RD A materials GONGGLASE-4, SABA/FLAG-1, UTIQUE 96/FLAG-l CROW’S’/BOW’S’^ , PBW343 and GIZA-168//SHUHA,S7DOBUC,S,) and among NCP materials ETBW5955, ETBW 5510, ETBW5898 and ETBW5963 were more stable (Fig 2 and 4). These may be due to less variation of major weather factors observed such as mean temperature, relative humidity, evapo-transpiration and soil temperature of the growing season over years. The combined analysis of variances for both experiments showed highly significant differences

153 (P<0.01) for most of the traits among the ICARD’s genotypes and for almost all traits among the NCP genotypes indicating the presence of sufficient genetic variability among the tested materials. The highest overall mean yield was recorded from NCP genotypes such as ETBW5955 (3268.3 kg ha'1), ETBW5963 (3226.1 kg ha'1), ETBW5901 (3208.2 kg ha'1) and ETBW5898 (3164.3 kg ha'1) with the maturity date of 85.2, 76, 80.1 and 78.9 days, respectively. Among the ICARDA materials, GONGGLASE-4 (3102.8 kg ha'1), SABA/FLAG-1 (3027.3 kg ha'1), UTIQUE 96/FLAG-l CROW’SVBOW’S’-3 (2973.6 kg ha'1) and 1994/95//TEVEE,S/TADINIA (2727.8 kg ha'1) were high yielding genotypes with TKW of 31.6, 34.7, 33.4 and 36.3 g TKW and 83.8, 91.4, 86.2 and 83.6 maturity days, respectively. Generally, the genotypes derived from the NCP possessed higher yield, better TKW and early maturing in comparison with the ICARDA materials.

Conclusions and Recommendations

The test environments (years and locations) significantly affected yield and yield stability of the genotypes evaluated under both trials. The higher environmental effect (52.3% of TSS) for ICARDA materials than the NCP materials (46.2% of TSS) clearly showed that the ICARDA materials were more influenced by the environmental factors (soil type, soil fertility and temperatures) probably due latitudinal effect as the ICRADA materials were developed at higher latitudes. The effects of L x G interactions for ICARDA and NCP materials varied significantly showing that there was high genetic variability between the ICARAD and NCP materials. The stability analysis over locations and years for both irials showed that there were some genotypes among ICARDA and NCP materials relatively high yiel Jer and stable. Thus, the is a necessity to evaluate more genotypes across locations and over years for identifying and selecting wide as well as location specific adaptable varieties possessing the required grain yield and quality parameters. Moreover, from the studied materials, yield and yield stability :>f genotypes derived from the NCP were better than the ICARDA materials. Hence, the results demands wheat breeders to give attention for strengthening the local wheat crossing program and reduce dependency on introduced wheat genetic materials.

References

Ezatollah F, Reza M, Mostafa A and /ahra V. 2012. GGL biplot analysis of genotype x environment interaction in wheat-barely disomic addition lines. Australian Journal of Crop Science 6(6): 1074-1079. Hinsta Gebru, Abraha Hailemariam and Tesfaye Belay. 2011. Genotype-by-environment interaction and grain yield stability of early maturing bread wheat genotypes in the drought prone areas of Tigray, Northern Ethiopia. J. App. Sci. Technology. 2(5), 51-57. Mohammed J. 1994. Performance of wheat genotypes under irrigation in Awash Valley, Ethiopia. African Crop Sciencc Journal 2(2): 145-151. Karimizadeh R, Mohammadi M, Sobaghnian N, Hosseinpour T and Shafazadeh MK. 2012. Analysis of genotype and genotype-by-environment interaction in durum wheat in warm rainfed areas of Iran. Crop Breeding Journal 2(2):71-78. Naheif EM Mohamed. 2013. Genotype by Environment interactions for grain yield in bread wheat (7’.aestivum L.). Journal of plant breeding and crop science 5(7): 150-157 Naser Sabaghnia, Mohtasham Mohammadi and Rahmatollah Karimizadeh. 2012. Interpretation of genotype-by environment interactions in multi-environment trials of bread wheat using cluster analysis. Natura Montenegrina podoorica. 1 1 (3):511-523. MoA (Ministry of Agriculture). 2011. National Rice Research Development Strategy of Ethiopia.

154 Evaluation of Bread Wheat Genotypes for Yield and Yield Components in Irrigated Lowland Areas

Desta Gebre1, Mihratu Amanuel1, Beakal Tadesse1 and Hailu Mengistu1 1 Ethic pian Institute of Agricultural Research, Werer Agricultural Research Center; Department of Irrigated Cereal Crops Breeding; P. 0. Box 2003, Addis Ababa, Ethiopia

Abstract Eight wheat genotypes introduced from ICARDA were evaluated along with the standard check at Werer Research Center from 2010-2013, and at Gewane from 2012-2013 cropping seasons with the objective of identifying well adaptable candidate bread wheat varieties for release. Combined analysis of variance showed that the effect of year (Y), locations (L), genotypes (G) as well as Y x G and L x G interactions showed significant differences (P<0.05) for yield and most of the parameters measured. The highest overall mean yield was obtained from NEJEMAII-14 (2977 kg ha"1), ADEL-6 (2931 kg ha'1) and HAALLA-44 (2724 kg ha'1) with early maturing days of 81, 82 and 81 days, respectively. Thus, the three genotypes were higher yielder, stress tolerant and were selected as candidate varieties to be released for irrigated areas of Afar Region. Fianlly, NEJEMAII-14 and ADEL-6 were recommended for release and use by wheat growers under irrigation.

Introduction

The major production constraints in the irrigated lowland areas arc the lack of adaptable & stable wheat varieties along with its production packages. Since year to year climatic conditions are projected to become more variable due to climate change (IPCC 2007), widely adapted cultivars are crucial to buffer unpredictable climate stresses such as drought, heat and salinity. The main causes of yield variation that can be reduced by breeding were identified as lodging, heat, and early maturing cultivars (Saulescu et al. 1998). The purpose of this paper, therefore, is to present the findings of the investigation on yield and yield components of bread wheat genotypes evaluated under irrigation in different locations.

Materials and Methods

Eight vheat genotypes, which were introduced from ICARDA, were evaluated along with the standard check at Werer Research Center from 2010/11 to 2012/13 and at Gewane from 2011/12 to 2012/13 cropping seasons. The minimum, maximum, and mean temperatures as well as relative humidity, evapo-transpiration, and soil temperatures of the experimental periods for Werer (Amibera Distric ) were collected (Tables 1 and 2). The genotypes were evaluated using RCBD in three replica.ions with plot size of 10 rows of 3 m length and 0.3 m between rows (9 nr). The planting dales were 3-10 November for Werer and 6-12 November for Gewane. Sowing was done by hand drilling at a seed rale of 80 kg ha'1 and DAP fertilizer was applied at the rate of 50 kg ha ! whole at sowing time at both locations. DAP was applied five cm away from the seed and covered with soil followed by irrigation water. On the other hand, UREA fertilizer was applied at all locations in split at the rate of 100 kg ha'1 (half at tillering and the remaining half at booting stages). It was applied in both times after irrigating the trials around five pm to avoid the loss of N due leaching and volatilization, respectively. Irrigation was applied every ten days intervals at Werer and Gewane until physio ogical maturity time of the genotypes. The trials were weeded three times using hand weeding and no herbicides were used.

155 I able 1 Mean, minimum (Min) and maximum (Max) temperatures in °C at Werer Research Center during the experimental periods

Months 2010/11 2011/12 2012/13 Min Max Mean Min Max Mean Min Max Mean October 17.80 35.50 26.7 21.80 34.90 27.3 17.80 33.90 27.0 November 16.20 32.50 24.4 23.10 33.20 26.0 16.40 33.60 25.7 Decembei 16.20 32.30 24.3 14.90 31.90 24.0 15.60 32.70 24.0 January 16.10 31.90 24.0 16.00 33.10 24.2 12.70 32.50 23.8 February 18.90 34.30 26.6 16.60 34.50 26.2 16.40 34.20 26.0 March 24.40 34.30 29.4 19.00 36.60 28.7 21.40 36.40 28.8 Mean of 6 Vlonths 18.27 33.47 25.9 18.57 34.03 26.0 16.72 33.88 25.9 Mean of (Nov - Feb) 16.85 32.75 24.8 17.65 33.18 25.1 15.28 33.25 24.9

Table 2 Muan values of RH, evapo-transpiration, soil temperature at 5 cm depth of Werer Research Center during the experimental periods

Months 2010/11 2011/12 2012/13 RH Evap* Soil RH Evap Soil RH Evap Soil (%) (mm) T°5cm (%) (mm) T°5cm (%) (mm) T°5cm October 47 265.1 33.7 45 261.1 29.9 53 219.9 34.2 November 52 199.7 30.9 52 197.1 30.8 48 218.6 34.5 December 52 178.5 28.4 48 201.3 29 52 195.6 32.5 January 55 217.8 30.9 47 255.3 29.7 52 199.9 31.3 February 44 234.3 33.0 42 274.7 30.7 37 223.5 34.2 March 45 226.8 31.2 36 324.9 32.1 49 251.2 36.0 Mean 6 months 49.2 220.4 31.4 45.0 252.4 30.4 49 218.1 33.8 Mean (Nov-Feb) 50.8 207.6 30.8 47.3 232.1 30.1 47 209.4 33.1 * RH relative humidity; Evap - evapo-transpiration

Data were collected before harvest from inner rows on phenotypic parameters such as days to heading, days to maturity, plant height and spike length. The central eight rows of each genotype were harvestec. manually using sickles. After harvest, data on number of kernels per spike were taken. The harvestec genotypes from the central eight rows of each plot were threshed and cleaned to determine thousand kernel weight (TKW) and grain yield per plot. The field and laboratory’ were compiled and subjected to statistical analysis using SAS Version 9.0 software.

Results

Combined analysis of variance showed that the effect of year was highly significant (P<0.01) for all parameters studied except TKW. The cffect of locations was significant (P<0.05) for most of the parameters. Significant differences (P<0.01) were observed among the genotypes for most parameters studied except spike length. Genotype by year and genotype by location interactions were also significant (PO.Ol) for most traits (Table 3).

The mean performance of the genotypes for the studied traits ranged from 49 to 57 days for DH, 81 to 90 days 1'or DM, 63 to 72 cm for PH, 6-7 cm for SL, 30-38 for NKPS, 26-35 g for TKW and 1894 - 2977 kg ha'1 for grain yield. The highest yielding genotypes were NEJEMAH-14 (2977 kg/ha), ADEL-6 (2931 kg ha'1) and HAALLA-44 (2724 kg ha ‘). These genotypes were also early maturing with maturity date of 81, 82 and 81 days, respectively (Table 2). The yield performances of these three genotypes were high across locations and over years.

156 Table 3 Combined ANOVA for grain yield and yield components (2010/11 - 2011/12

Source of Mean squares variatior df DH* DM PH SL NKPS GY TKW Year (Y) 2 * ** kit ** ** ns Location(L) 2 ** ** ** ** ** * * Replication 2 ns ns ns ns ns ns ns Treatment(G) 8 ** ** ** ns ** * ** YG 14 ** ** ** t* ** * ** LG 14 ** ** ■kit ** ** * ** DH - days to heading; DM - Days to maturity; PH -plant height (cm); SL - spike length (cm); NKPS - number of kernels per spike; GY- grain yield (kg ha'1); TKW- thousand kernel weight (g)

Table 2 Mean performance of wheat genotypes under irrigated areas (2010/11-2012/13)

Genotypes DH DM PH SL NKPS GY TKW ADEL-6 51 82 65 6 31 2931 33 MORSUD-22 55 88 70 7 34 2224 31 HAALLA-44 49 81 63 7 34 2724 35 NADIA-15 57 90 71 7 30 1894 28 NEJMAH-6 53 81 65 7 32 2279 30 PRINIA-2/2* KAR-2 56 86 65 7 38 2046 26 PAVON-76 (check) 54 83 71 V 35 2191 31 NEJMAH-14 52 81 72 6 35 2977 31 SOLA. AN-1 52 87 69 7 32 2409 33 Mean 53 84 68 7 33 2408 31 CV(%) 5.7 3.7 7.9 10.3 12.4 11.9 9.3 LSD (5%) 2.5 3.3 4.5 0.6 2.9 465.0 1.9

Table 3 Mean yield (kg ha-1) performance of genotypes across locations and years

Genotypes 2010/11 2011/12 2012/13 Werer Werer Gewane Werer Gewane ADEL-6 2064 4011 3675 1987 3692 MORSUD-22 2400 4385 1826 1663 2307 H A A LJU 4 2320 4365 2758 1658 2909 NADIA-15 2381 4332 1562 1382 1381 NEJMAH-6 1900 4291 2722 1521 2800 PRIN A-2/2* KAR-2 1865 3693 1982 1391 2482 PAVON-76 (check) 2487 4388 2518 1369 2090 NEJIVAH-14 2567 4595 3545 1879 3116 SOLAJAN-1 2306 4167 2825 1554 2922 Mean 2254 4247 2601 1600 2633 CV (%) 26.4 12.9 16.9 8.4 12.0 LSD 5%) 1029.6 944.5 707.54 213 817.8

Discussions

Sufficient genetic variability was observed among the genotypes evaluated under the lowland irrigated areas, which could be utilized for future breeding works. Mihratu (2014) reported similar finding. Earliness in DH and DM observed for the three promising genotypes revealed the possibility of us ng these materials in the breeding program for improving earliness without affecting the yield. Saulescu et al. (1998) reported similar result. The high environmental stresses of the testing sites such as high mean temperature ranged from 24.8 to 25.1 °C), low RH (47.3 to 50.8 %), high evapo-

157 transpiration (207.6 to 232.1%) and high soil temperature (30.1 °C to 33.2°C) could affect the yielding potentials of the genotypes. The findings of Reynolds et al. (2012) also support the present results. The soil PH was relatively high for Werer site (8.1) as compared to Gewane site (7.62) but the mean temperaiure of Gewane was higher than Werer site (24.95 °C) (Werer unpublished data). Generally, the combined effects of these all stresses greatly affccted the yield performances of the genotypes. Despite these environmental stresses, the mean yield performance of the three promising genotypes over locations and years was higher showing their potential to adapt over wide ranges of environmental conditions. Two of the varieties ADEL-6 and NEJMAH-14 were released for production in irrigated environments.

Conclusions and Recommendations

Environmental condition such as high temperatures, evapo-transpiration, soil temperature, soil pH and low relative humidity could affect the overall performances of bread wheat genotypes. Despite these environmental stresses, three genotypes performed better across locations and years and were verified for release and two of the varieties (ADEL-6 and NEJMAH-14) were released for production. The result of the present study put forward to evaluate more bread wheat genotypes under different environments. Wheat stresses such as high temperatures combined with high evapo-transpiration, soil pH, salinity, and relative humidity should be give attention in developing high yielding and stress tolerant v/hat varieties for irrigated environments of Ethiopia. Lack of high yielding and stress tolerant in such areas is bottleneck to enhance wheat production in such areas and due attention should be given.

References

IPCC (Intergovernmental Panel on Climate Change). 2007. “Fourth Assessment Report of the Intergovernmental Panel on Climate Change: The impacts, adaptation and vulnerability (Working Group III).” Cambridge University Press, New York. Reynolds VIP, Pask AJD and Mullan DM. 2012. Physiological Breeding 1: Interdisciplinary Approaches to Improve Crop Adaptation. Mexico, D.F.: CIMMYT. Saulescu NN, Ittu G, Balota IM, Ittu M and Mustatae P. 1998. Breeding wheat for lodging resistance, earliness and tolerance to abiotic stress. Wheat prospects for global improvement 181-188.

158 Closing Remarks Fentahun Mengistu (PhD) Director General, Ethiopian Institute of Agricultural Research, Addiss Ababa, Ethiopia

It is needles to talk about the importance of wheat in Africa. A crop once that was considered an exotic crop has now labeled as one of Africa’s strategic commodity. It is clear why that; because wheat is a universal bread. These days, the demand for wheat is growing faster than for any other food crop in Africa. Despite this, however, unfortunately Africa has the lowest per capita wheat production. For instance, Sub-Saharan Africa now grows less than 10% of the wheat it could. Consequently, in 2011 Africa spent >14 b US$ to import 40 m tons. Likewise, SSA imported 12.3 MMT in 2010, is forecasted to import 35 MMT by 2050.This trend of increasing reliancc on imports is not however, sustainable, and can even threaten nutritional and economic security. This is happening while import dependence can be reduced through domestic production in SSA. Data shows that several countries in SSA could achieve wheat yields exceeding 6 t/ha (compared to global average of 3t/ha), and several countries including our East African region can profitably produce wheat under rainfed condition. The limiting factors arc not agro-ecological, they are rather socio cultural, institutional and policy impediments. Therefore, increased measures are needed to decisively increase wheat production, productivity, and self-sufficiency in our region. Domestic production, however, needs ensuring relative competitiveness since it would be affected by a fall in domestic yield otherwise. This calls for investment in R&D to increase yields and reduce production and marketing costs. This means that it requires developing suitable technologies and varieties (disease resistant) and increasing productivity among smallholders, puttingin place effective and sustainable wheat seed systems, making wheat value chains work more effectively, and above all fostering regional co-operation to develop the wheat industry.

EAAP0 can be taken a role model to fostering regional co-operation, collective action and regional integration that also helps realizing the macro development goals such as CADDP and NEPAD. ASARECA, etc. Looking at the workshop program over the last three days you have discussed progresses of research findings on wheat that includes variety, management, insect, disease and weed pest management, pre-and post-harvest implements, seeds, VCs, technology demonstration and adoption studies results. In the course of reviewing these valuable topics, I am sure you have identif ed several valuable technologies worth transfer lo wheat growers and users. In our thus far efforts I knowr we have already managed raising the wheal yields. For instance, in Ethiopia over the last years wheat production has increased 400% , and today the national average wheat yield reached 2.4 t ha'1 and a new record harvest of > 4MMT in 2013/14.EAAPP’s contribution was substantial in this. Not only that, we have progressed very well in building research capacities: WRCoE. This is anothc great achievement. Therefore, standing at this point some of our way forward should be the following. As we are approaching the closing of the lsl phase of EAAPP (we arc hopeful that EAAPP will continue: we need to do our assignments in this regard), wc need to consolidate the thus far findings and pack them up for scaling up. We need to share and exchange technologies and products across project countries; and transfer innovations emanated from the project to end users. Besides, we need to continue building the CoE. What is important is we need to understand that EAAPP was simply a showcase for regional collaboration serving the broader African Agenda as CADDP and

159 NEPAD, ASARECA, FARA, ctc. This means that we need to continue to collaborate even when we we will be in no EAAPP situation. There are several reasons for this continued collaboration. Primarily, it is in the best interest of African governments meeting development objectives, i.e. regional integration. Second, East African region is a rust hotspot; rust menace requires concerted effort on broader fronts. Research in Ethiopia and Kenya is already helping the world by serving a rust screening site. Hence, collectivc action brings about increased synergy in innovations, for sharing knowledge, experience, and skills. Thirdly, collaboration ensures cost effectiveness, economies of scale, market integration, and seed trade harmonization. I therefore look forward to a sustained partnershio and collaboration amongst all of you.

In the closing, I would like to thank the organizers of the meeting. 1 am also grateful to all of you for coming here and make fruitful deliberations for good of all of us and our communities.

160 List of participants

Name Country Designation Institutions Telephone contacts Email address 1. Abebe Atilaw Ethiopia Director. Seed System EIAR +251-911 103 486 [email protected] 2. Abiy Girma Ethiopia Accountant EIAR - KARC +251-920 806 616 [email protected] 3. Adane Choferie Ethiopia Breeder EIAR -KARC +251-911 840 982 [email protected] 4. Adugna Wakjira Ethiopia Deputy Director General EIAR +251-917 812 190 [email protected] 5. Aqegnenu Mekonnen Tessema Ethiopia Wheat breeder Sirinka ARC +251-912 415 130 [email protected] 6. Aklilu Nigussie Ethiopia Agric Economics, Extension & EIAR - WARC +251-911 914 003 [email protected] Gender +251-925 921 664 7. Alemayehu Assefa Ethiopia Research Coordinator - EAAPP EIAR +251-911 203 869 a [email protected] 8. Ashenafi Tariku Ethiopia Assistant Researcher II EIAR +251-913 092 400 [email protected] 9. Ashenati Gemechu Ethiopia Plant pathologist EIAR - DZARC +251-926 897 678 [email protected] 10. Assaye Legesse Ethiopia Task Team Leader World Bank +251- [email protected] 11. Ayalneh Tilahun Ethiopia Wheat breeder South Agricultural Research +251-911 975 285 [email protected] Institute 12. Azeb Hailu Ethiopia Associate Researcher I Tigray Agricultural Research +251-933 091 548 [email protected] Institute +251-344 408 031 13. Balcha Yaie Ethiopia Wheat breeder EIAR - KARC +251-934 893 948 [email protected] 14. Bedada Begna Ethiopia Agricultural Extension EIAR - KARC +251-920 067 109 [email protected] 15. Bekele Abeyo Ethiopia CIMMYT +251-920 722 647 [email protected] 16. Bekele Hundie Ethiopia Plant pathologist EIAR +251-931 475 002 [email protected] 17. Bekele Negash Ethiopia ICM Process Representative EIAR -KARC +251-913 054 367 [email protected] 18. Berhanu Arbissie Ethiopia Wheat Regional Research Tigray Agricultural Research +251-914 731 625 b [email protected] Coordinator Institute 19. Daniel Kasa Ethiopia Pathologist EIAR +251-917 850 880 [email protected] 20. Dawit Asnake Ethiopia Wheat breeder EIAR - KARC +251-911 916 376 [email protected] [email protected] 21. Dawit Habte Ethiopia Principal Investigator EIAR - KARC +251-916 453 137 [email protected] 22. Desta Gebre Ethiopia Wheat breeder EIAR - Werer Research Centre +251-912 149 441 [email protected] 23. Eshetu Lemma Ethiopia Field Research Assistant EIAR +251-911 750 796 [email protected] 24. Friew Kelemu Ethiopia Wheat mechanization Focal Person EIAR +251-911 341 468 [email protected] 25. Gebreyes Gurmu Ethiopia Land and Water Research Deputy EIAR +251-911 380 140 [email protected] Director 26. Genet Abaineh Ethiopia Accountant - EAAPP EIAR - KARC +251-920 166 118 [email protected] 27. Girma Mengistu Ethiopia Crop Researcher Oromia Agricultural Research +251-114 707 116 [email protected] Institute +251-911 726 214 28. Girma Moqes Ethiopia Director EIAR - HQ +251-911 112 602 [email protected] 29. Girma Tezera Ethiopia Training Expert EIAR +251-911 897 607 [email protected]

161 30. Habtemarian Zegeye Ethiopia Wheat breeder EIAR -KARC +251-910 955 505 [email protected] 31. Hussein Sareta Selamo Ethiopia Associate Researcher EIAR - Kulumusa +251-912 027 271 [email protected] 32. Kassahun Zewdie Ethiopia Researcher EIAR-HRC +251-911 186 486 [email protected] 33. Mathewos Ashamo Ethiopia Plant breeding Areka Agricultural Research +251-911 781 100 [email protected]

34. Mechal Tadele Ethiopia M&E Specialist EIAR +251-911 203 867 [email protected] 35. Mekonnen Mekuria Ethiopia T & D Specialist Ministry of Agriculture - EAAPP +251-910 141 930 [email protected] 36. Mekuria Temtme Ethiopia Plant breeder EIAR - DZARC +251-912 802 855 [email protected] 37. Mihratu Amanuel Ethiopia Wheat breeder EIAR - Werer Research Centre +251-911 005 935 [email protected] 38. Muluken Bayable Tadege Ethiopia Wheat breeder Amhara Regional Agricultural +251-910 114 610 mulukenbaya2012@ g mail, com Research Centre 39. Rehima Mussera Ethiopia Deputy Director, Agric Econ, EIAR +251-911 384 268 [email protected] Extension & Gender 40. Shitaye Homma Ethiopia Wheat breeder EIAR - Debrezeit +251-934 962 992 [email protected] [email protected] 41. Tadesse Dessalegn Ethiopia WRCoE Coordinator WRCoE +251-918 769 629 [email protected] 42. Tadesse Sefera Ethiopia Centre Director EIAR - Kulumusa +251-911 366 069 [email protected] 43. Tamirat Negash Ethiopia Wheat Entomology Researcher EIAR - KARC +251-913 218 246 [email protected] 44. Tariku Gosaye Ethiopia Procurement Coordinator Ethiopian Institute of Agricultural +251-911 413 329 T [email protected] Research (EIAR) 45. Tebkew Dambe Ethiopia Researcher EIAR - Debrezeit +251-114 338 555 [email protected] 46. Terefe Fitta Ethiopia M&E Specialist & ESS Focal Ministry of Agriculture +251-911 808 596 [email protected] person 47. Tesfay Gebrekirstos Gebrenariam Ethiopia Plant Pathologist Tigray Agricultural Research +251-914 761 733 [email protected] Institute 48. Tesfaye Haregewoin Ethiopia Planning, Monitoring & Evaluation EIAR +251-923 210 717 [email protected] 49. Tesfaye Solomon Ethiopia Wheat Value Chain Principal EIAR +251-912 068 192 [email protected] Investigator 50. Teshome Beyene Ethiopia Gender Specialist Ministry of Agriculture - EAAPP +251-911 721 663 [email protected] 51. TibebeAdmasu Ethiopia Procurement Finance Property EIAR - Kulumusa +251-911 755 125 [email protected] Management Process 52. Tolesa Alemu Ethiopia Researcher EIAR +251-911 488 299 [email protected] 53. Wasihun Legesse Ethiopia Wheat Research Coordinator EIAR - DZARC +251-923 975 859 [email protected] 54. Worku Denbel Bulbula Ethiopia Insect pests and diseases ElAR-WRCoE, Kulumusa +251-911 867 018 [email protected] Monitoring & Surveillance Focal [email protected] person 55. Wubishet Alemu Ethiopia Plant pathologist Oromia Agricultural Research +251-913 091 568 Institute 56. Yemane Giima Belaineh Ethiopia Molecular Biologist Haramaya University +251-933 405 510 [email protected] 57. Yitaye Alemayehu Ethiopia National Coordinator EAAPP +251-912 072 945 [email protected] 58. Zerihun Tadesse Tarekegn Ethiopia National Wheat Coordintaor EIAR - KARC +251-911 003 056 [email protected] 59. Githiri Charles K. Kenya Sub-county Agricultural Extension Ministry of Agriculture, Livestock & +254-720 350 435 [email protected] 162 Officer Fisheries 60. Kinyanjui Catherine Kenya 1 & D and Agribusiness Specialist EAAPP - PCU +254-722 249 669 catherinekinyan|[email protected] 61 Macharia Gethi Kenya Centre Director Kenya Agricultural and Livestock +254-722 328 963 mqethi@qmail com Research (KALRO) [email protected] 62. Maina Jedidah Kenya Research Specialist EAAPP - PCU +254-722 374 886 [email protected] 63. Munene Macharia Kenya Principal Research Officer Kenya Agricultural and Livestock +254-721 297 490 [email protected] Research (KALRO) 64. Muriuki N. Jane Kenya National Project Coordinator EAAPP-PCU +254-722 323 202 [email protected] 65. Nasirembe W. W Kenya Research Officer - Mechanisation Kenya Agricultural and Livestock +254-733 812 953 [email protected] Research (KALRO) 66. Wandera Foustine Peter Kenya EAAPP Desk Officer Kenya Agricultural and Livestock +254-721 737 708 [email protected] Research (KALRO) +254-204 418 301 67. Wanyera Ruth Kenya Principal Research Scientist Kenya Agricultural and Livestock +254-721 216 016 [email protected] Research (KALRO) 68. Elanga Anthony Tanzania Wheat Focal Person Uyole Agricultural Research +255-784 354 064 [email protected] Institute 69. Katunzi Justa M. Tanzania Principal Agricultural Officer /T&D Ministry of Agriculture, Food +255-753 904 861 [email protected] Security and Cooperatives 70. Mongi Rose Tanzania Plant breeder Uyole Agricultural Research +255-755 765 079 [email protected] Institute 71. Ngailo Jerry A. Tanzania Sr. Agricultural Research Officer Uyole Agricultural Research +255-784 906 728 [email protected] Institute 72. Nkori Kibanda Tanzania RRCoE Coordinator ARI-KATRIN +255-784 419 422 [email protected] 73. Agum Winfred A. Uganda Entomologist NARO - NARL +256-777 780 872 [email protected] 74. Akulumuka Vincent Uganda PM-EAAPP ASARECA +256-772 026 105 [email protected] 75. Atukunda Apophia Uganda ESS Specialist ASARECA [email protected] 76. Gidoi Robert Uganda Socio-Economist NARO-BUGIZARDI +256-772 479 617 [email protected] 77. Kabugo Florence Uganda PA-EAAPP ASARECA +256-772 488 567 [email protected] 78. Nakanwagi Josephine Uganda Student NARO +256-774 208 901 nakanwagi [email protected] 79. Ojangole Stephen Uganda M&E Specialist NARO +256-772 669 610 sojangole(5)gmail.coin 80. Wasukira Arthur Uganda Research Officer NARO - BUGIZARDI +256-782 427 527 awasukira(Q)gmail.com 81. Wobibi Stephen Mutento Uganda Senior Crop Technician NARO - BUGIZARDI +256-779 071 960 [email protected]

163