Impact of Irrigation on Poverty and Environment in Ethiopia Draft

Impact of Irrigation on Poverty and Environment in Ethiopia

Draft Proceeding of the Symposium and Exhibition held at Ghion Hotel, Addis Ababa, Ethiopia 27th -29th November, 2007

Compiled by: Seleshi Bekele; Makonnen Loulseged; Aster Denekew

International Water Management Institute P.O.Box 2075, Colombo, Sri Lanka University of Natural Resources and Applied Life Sciences, Muthgasse 18, A- 1190 Vienna, Austria

Haramamaya University, P.O.Box 138, Dire Arba Minch University, P.O.Box 21, Arba Dawa, Ethiopia Minch, Ethiopia

Ethiopian Institute of Agricultural Research Austrian Research Centers GmbH, A-2444 P.O.Box 2003, Addis Ababa, Ethiopia Seibersdorf, Austria

Ministry of Water Resources Ministry of Agriculture and Rural Development P.O.Box 5673, Addis Ababa, Ethiopia P.O.Box 62345, Addis Ababa, Ethiopia

IWMI receives its principal funding from 58 governments, private foundations, and international and regional organizations known as the Consultative Group on International Agricultural Research (CGIAR). Support is also given by the Governments of Ghana, Pakistan, South Africa, Sri Lanka and Thailand.

Awlachew, S.B.; M. Loulseged.; Yilma, A.D.; 2008. Impact of irrigation on poverty and environment in Ethiopia.

/water management/irrigation management/socioeconomic impact/environnemental impact/water right/

ISBN : ISBN :

Please direct inquiries and comments to: [email protected]

Table of Contents

Acronyms and Abbreviations ...... vii Acknowledgement...... ix Welcoming address...... xi Seleshi B. Awulachew...... xi Akissa Bahri, IWMI ...... xiii Opening Address...... xvi H.E Dr. Leopohld Moll ...... xvi H.E Ato Adugna Jebessa ...... xvii Theme one: Status Quo Analysis, Characterization and Assessment of Performance of Irrigation in Ethiopia...... 1 Statues quo analysis, Characterization and Assessment of Performance of Irrigation in Ethiopia ...... 2 Comparison of irrigation performance based on management and cropping types ...... 14 Irrigation and Rain-fed Crop Production System in Ethiopia ...... 27 Basline Servey of Irrigation and Rainfed Agriculture in...... 37 Blue Nile Basin ...... 37 Assessment of Design Practices and Performance of Small Scale Irrigation Structures in South Region ...... 47 Across System Comparative Assessment of irrigation performance of community managed scheme in Southern Ethiopia...... 59 Analysis of irrigation systems using comparative performance indicators:...... 77 A case study of two large scale irrigation systems in the upper Awash basin ...... 77 Discussion on Theme 1: Status Quo Analysis, Characterization, and Assessment of Performance of Irrigation in Ethiopia ...... 93 Theme two: Irrigation Impact on Poverty and Economy...... 95 Rural poverty and inequality in Ethiopia: does access to...... 97 small-scale irrigation make a difference?...... 97 An Assessment of the financial viability and income impact of small scale irrigation in Ethiopia ...... 117 Importance of Irrigated Agriculture to the Ethiopian Economy: Capturing the direct net benefits of irrigation ...... 127 Investment in irrigation as a poverty reduction strategy: Analysis of small-scale irrigation impact on poverty in Tigray, Ethiopia...... 156 The Impact of Small Scale Irrigation on Household Food Security: The Case of Filtino and Godino Irrigation Schemes in Ada Liben District,...... 179 East Shoa, Ethiopia...... 179 A Comparative Analysis of the Technical Efficiency of Irrigated and ...... 193 Rainfed Agriculture: A case of Awash and Rift Valleys of Ethiopia...... 193 Impact of irrigation on livelihood and food security in the modern Hare river irrigation scheme in Southern Ethiopia...... 210 Impact Assessment of Rainwater Harvesting Ponds:...... 223 The Case of Alaba Woreda, Ethiopia...... 223 Discussion on Theme 2: Irrigation Impact Poverty and Economy ...... 256 Theme three: Irrigation Institutions and Support Services ...... 259

v Status of Irrigation Institutions and Support Services in Ethiopia ...... 260 Does Access To Small Scale Irrigation Promote Market Oriented Production In Ethiopia?.. 262 Small Scale Irrigation Interventions for System Productivity and ...... 282 Natural Resource Management in Ethiopian Highlands: ...... 282 Institutions, Management Practices and Challenges of Small-Scale Irrigation Systems in Ethiopia: A Case Study of Two Modern Smallholders Irrigation Systems in Western Oromia, Ethiopia ...... 298 Water Rights and the Processes of Negotiations Among Irrigators in ...... 323 West Shewa Zone: The Case of Indris Scheme in Toke Kutaye District...... 323 Technical and Institutional Evaluation of Geray Irrigation Scheme in West Gojjam Zone, Amhara Region...... 342 Discussion on Theme 3: Irrigation Institutions and Support Services...... 354 Theme four: Environmental and Health Impact of Irrigation ...... 356 Case study review of investigated irrigation projects in Ethiopia ...... 357 Environmental Impact Analysis of Two Large Scale Irrigation Schemes ...... 370 in Ethiopia ...... 370 GIS And Remote Sensing Integrated Environmental Impact Assessment of Irrigation Project In Finchaa Valley Area...... 400 Entomological studies on the impact of a small-scale irrigation scheme on malaria transmission around Zeway, Ethiopia...... 418 Malaria Transmission In Ziway, Eastern Shewa Zone, As Related to Irrgation Development439 Discussion on Theme 4: Environmental and Health Impact of Irrigation ...... 455 Group discussion and findings ...... 461 Way Forward and Synthesis Plan...... 461 Exhibition ...... 461 Closing...... 462 List of Participants ……………………………………………………...463

Acronyms and Abbreviations

ADA Austrian Development Agency ADB African Development Bank ADLI Agricultural Development-led Industrialization AHFSI Aggregate Household Food Security Index AIP Amibara Irrigation Project AMU Arba Minch University ARC Austrian Research Centers ARIS Annual Relative Irrigation Supply ARWS Annual Relative Water Supply BOKU University of Natural Resources and Applied Life Sciences CA Comprehensive Assessment CSA Central Statistical Authority CGIAR Consultative Group on International Agricultural Research DAP Diammonium Phosphate Fertilizer DPPC Disaster Prevention and Protection Commission EARO Ethiopian Agricultural Research Organisation EIA Environmental Impact Assessment EIAR Ethiopian Institute of Agricultural Research EPA Environmental Protection Authority ESDA Ethiopian Sugar Development Agency FAO Food and Agriculture Organization of the United Nations FDRE Federal Democratic Republic of Ethiopia FFW Food-for-Work GDP Gross Domestic Product GMs gross margins GSDID Gobu Seyo District Irrigation Desk HU Haramaya University IAR Institute of Agricultural Research IDP Irrigation Development Program IDP Irrigation Development Strategy IFAD International Food and Agricultural Development IIPE Impact of Irrigation on Poverty and Environment IWMI International Water Management Institute MoA Ministry of Agriculture MoARD Ministry of Agriculture and Rural Development MoFED Ministry of Finance and Economic Development MoWR Ministry of Water Resources

vii MSI Medium Scale Irrigation NDVI Normalized Difference Vegetation Index OIDA Oromia Irrigation Development Authority PA Peasant Association PASDEP Plan for Accelerated and Sustained Development to End Poverty PRSP Poverty Reduction Strategy Paper REST Relief Society of Tigray RWH Rain Water Harvesting SNNPR Southern Nation and Nationalities People Region SSI Small Scale Irrigation TUs Tertiary Units UNDP United Nation Development Project USDA United States Department of Agriculture WDR Water Delivery Ration WSDP Water Sector Development Program

viii

Acknowledgement

This proceeding is the output of three days symposium and exhibition organized by International Water Management Institute (IWMI), University of Boku, ARC Sieberdorf, Haramaya University (AU), Arbaminch University (AMU), and Ethiopian Institute for Agricultural Research (EIAR). There were over 126 participants representing institutions from Government, NGOs, private sector, irrigators and students supported by the project who have contributed to the success of this symposium.

The International Water Management Institute is thankful to Austrian Government for allocating the necessary resource to undertake this research which will provide comprehensive and valuable input to the irrigated agriculture sub sector in Ethiopia in particular and Sub Saharan Africa in general.

The authors and co-authors who presented pepares during this symposium also acknowledge the support and assistance provided by various parties in the successefull accomplishment of their respectieve research work.

x Welcoming address

Seleshi B. Awulachew Head, International Water Management Institute (IWMI) East Africa and Nile Basin, Addis Ababa, Ethiopia

Your Excellency Ato Adugna Jebessa, State Minister, Ministry of Water Resources

Representatives of partner institutions for Impact of Irrigation on Poverty and Environment from Univeritaet Bodunkultur Wien, ARC Sibersodrf, Arba Minch University, Haramaya University, Ethiopian Institute of Agriculture

Delegates from various federal and regional government offices Academic and research institutions

International donor communities International and regional institutions Non governmental organizations Private sector representatives Irrigation scheme operators

Ladies and gentlemen:

Thank you very much for accepting our invitation on the Impact of Irrigation on Poverty and Environment symposium and poster exhibition and welcome. The symposium and exhibition are organized towards the final period of the project which has been implemented during the last three and half years

The three days symposium and exhibition are organized in order to share the generated knowledge by the researchers to the wider audience and stakeholders in the irrigation sector.

The symposium is organized with the objectives of: ¾ Bring together and share experiences among government (policy makers, technical experts), NGOs, private sector, international donors and financial institutions and related stake holders that are working on irrigation, socio-economy and environment; ¾ Disseminate and Share the results generated during the project implementation period; ¾ Carry out focused discussion to explore opportunities and mechanisms through which the uptake of knowledge, application, and dissemination and out scaling of findings could be enhanced. The symposium is organized under four themes that include: 1. Statues quo analysis, Characterization and Assessment of Performance of irrigation in Ethiopia: 2. Irrigation Impact on Poverty and Economy 3. Irrigation Policy, Institutions and Support Services) 4. Environmental and Health Impact of Irrigation

In total 28 papers will be presented. In addition a number of posters will also be exhibited. We will be having a number of plenary discussions and group work deliberations. You have the details of the program in your bags.

It is my hope that you will enjoy this event and we can make useful discussions that can lead to a good synthesis of knowledge building on the result of the project and wealth of experience and wisdom of the participants towards the support of sustainable development irrigation in Ethiopia.

In order to start and officially open the symposium I will like to request the representative of the various institutions sitting on the podium to make opening remarks. After the opening remarks, his Excellency Ato Adugna Jebessa will make an official opening speech of the symposium.

Accordingly, first I will like to invite: - Prof. Willibald Loiskandl, BOKU - Prof. Belay Kassa, HU - Dr. Tarkegn Tadessa, AMU - Dr. Solomon Assefa, EIAR - Dr. Akissa Bahri, IWMI

Thank you!

xii Welcoming address

Akissa Bahri, IWMI Head, International Water Management Institute (IWMI) Africa Region, Accra, Ghana

Your excellency, Mr. Ato Adugna Jebessa, State Minister of Ministry of Water Resources

Dr. Leopohld Moll, Director, Austrian Development Cooperation

Prof. Willibald Loiskandl, Universität für Bodenkultur Wien

Distinguished representatives of partner institutions for Impact of Irrigation on Poverty and Environment from Universität für Bodenkultur Wien, ARC Siebersdorf, University of Natural Resources and Applied Life Science, Arba Minch University, Haramaya University, Ethiopian Institute of Agriculture

Ladies and gentlemen,

It is with great pleasure that I welcome you all on behalf of the International Water Management Institute and on my own behalf to this symposium on “Impact of Irrigation on Poverty and Environment in Ethiopia” and to this opening session.

This two-day symposium is the output of a collaborative research project sponsored three years ago by the Austrian Development Agency (ADA) and implemented by the International Water Management Institute (IWMI), Austrian Institutions, Universität für Bodenkultur Wien (BOKU, Vienna), Austrian Research Centers (ARC Siebersdorf), the University of Natural Resources and Applied Life Science and Ethiopian Institutions: Arba Minch University (AMU), Haramaya University (HU), the Ethiopian Institute of Agricultural Research (EIAR), the Ministry of Water Resources (MoWR), the Ministry of Agriculture and Rural Development (MoARD) and the Regional Bureaus for Irrigation and Agriculture.

The International Water Management Institute, member of the CGIAR system, is the leading international scientific research organization on water, food and environment, with an overall mission of “improving the management of water and land resources for food, livelihoods and nature”. In Africa, IWMI conducts research in three sub-regions; the Nile Basin and East Africa, West Africa and Southern Africa. Water scarcity, poverty, low productivity, health issues, water quality, endemic droughts and floods and transboundary conflicts in water management, along with land degradation are some of the critical issues Africa faces. IWMI works closely with Africa-wide sub–regional organizations and many national and agricultural research systems to study the land and water management challenges facing poor rural communities and to develop innovative approaches, tools and interventions that can improve food security, livelihoods, health and ecosystem services.

Improved land and water management is essential for sustainable development and poverty reduction in sub-Saharan Africa including Ethiopia. I would therefore like to mention two key research programs, the Comprehensive Assessment of Water Management in Agriculture (the CA) (Comprehensive Assessment of Water Management in Agriculture. 2007) and the

xiii Investment in Agricultural Water for Poverty Reduction and Economic Growth in Sub-Saharan Africa (a collaborative program of the World Bank, FAO, IFAD, ADB and IWMI, in partnership with NEPAD, 2007) whose findings should be considered along the results of the present project.

• The CA has critically evaluated the benefits, costs and impacts of the past 50 years of water development, the water management challenges communities are facing today, and solutions people have developed. The results of the CA will support better investment and management decisions in water and agriculture in the near future and over the next 50 years.

• The synthesis report of the Agricultural Investment Study analyses the contribution to date of agricultural water management to poverty reduction and growth in the sub- Saharan Africa, the reasons for its slow expansion and apparently poor track record, as well as the ways in which increased investment in agricultural water management could make a sustainable contribution to further poverty reduction and growth.

Investment in irrigation is needed to reduce poverty in rural areas. Eighty-five percent of sub- Saharan Africa’s poor live in the rural areas and depend largely on agriculture for their livelihoods. The Nile basin is home to an estimated 175 million people and more than 330 million people live in the NBEA sub-region. The Nile Basin and East Africa sub-region is far from being homogenous in terms of agro-ecology, socio-economic development, historical and political background. The Sub-region, despite having significant water and land resources, has the highest proportion of people living below the poverty line and is the most food insecure sub-region in the world. Dependence on rainfed agriculture, coupled with high rainfall variability, is one of the main causes of food insecurity. Drought is a frequent and recurrent event throughout much of the region, the impacts of which are made worse by HIV-AIDS and war. The majority of the people, over 70%, depend on subsistence agriculture. However, the resource base of land and water is not well utilized, nor appropriately managed, and is degrading very rapidly. Water-related diseases are common and a major cause of the relatively low life expectancy in the region.

Agricultural growth is therefore clearly key to poverty reduction; it can also help drive national economic growth. Ethiopia relies on agriculture for a large part of its GDP (44%). Raising agricultural productivity is the most viable option for reducing poverty, and irrigation development can enhance economic development. Irrigation schemes can facilitate multiple uses of water that combine agriculture with livestock, fisheries, and other income-generating activities to enhance rural incomes and sustainability. Investment in irrigation is also needed to keep up with global demand for agricultural products and adapt to changing food preferences and societal demands, to adapt to urbanization, industrialization, and increasing allocations to the environment and to respond to climate change. Climate variability and extreme events will require water resources development, large water storage facilities, further irrigation development, and changes in the operation of existing schemes.

Your excellency, Distinguished participants, Ladies and gentlemen,

I hope that at the end of this symposium, we would gain insights into how various irrigation strategies have impacted on agricultural output and hence the reduction of poverty as well as improvement in the environmental conditions of Ethiopia. This should then help us to better understand the role that irrigation can play in Ethiopia’s development process and hence improve rural poverty, achieve gender equity and protect Ethiopia’s environment.

xiv

It is my hope that this symposium will constitute a major step that will lead to better natural resources management in Ethiopia.

On behalf of IWMI, I would like to thank you all for coming and look forward for a fruitful and rewarding symposium.

Thank you

xv Opening Address

H.E Dr. Leopohld Moll Director, Austrian Development Cooperation Addis Ababa, Ethiopia

His Excellency Dr. Leopohld Moll, Director Austrian Development Cooperation in his oral speech said that the Austrian Government is pleased to support t the Government of the Federal Republic of Ethiopia in its effort to fight poverty and improve the well being of its people. The Austrian government recognizes the significance of irrigated agriculture in mitigating the impacts of climate variability and attaining food security in Ethiopia. He said, it is my sincere belief that the out puts of the research results of the IIPE project provides veritable information that can assist policy makers to make appropriate policy related decsions in the subsector. Finally, he wished participants a successful two days deliberation.

xvi

Opening Address

H.E Ato Adugna Jebessa State Minister, Ministry of Water Resources Addis Ababa, Ethiopia

Dear delegates of government and non governmental organizations

Representatives of academic and research organizations

Dear invited guests and participants

Ladies and Gentlemen,

It is indeed my great pleasure to be here with you today for the opening of this important workshop on impact of irrigation on poverty and environment research in Ethiopia.

As you all know there is a global consensus to fight poverty and improve human well being through appropriate measures that can target the issues related to poverty. Accordingly, in the year 2000 the MDG have set quantitative targets to be achieved by the year 2015 for the reduction of poverty, i.e. improvement in health, education, and the environment and other dimensions of human well being, particularly in sub-Saharan Africa where about 25% of the world poor live.

The economy of Ethiopia is significantly agricultural based. Access to reliable water is a fundamental factor in influencing poverty and economy. In this country, recurrent droughts have caused serious failures in agricultural production that have resulted in mass starvation and loss of human and animal life, not to mention the devastation of the natural environment. A more satisfactory outcome can be achieved through the development of the country's water resources for growth of productivity and irrigation, wherever opportunities exist.

Population explosion and food insecurity are serious twin problems that must be addressed simultaneously on priority basis. Despite several attempts to address the situation, the problem of environmental degradation, agricultural productivity and food shortage remains critical. Unless they are seriously tackled they are threats to the country and can worsen poverty in the face of the rapidly growing population.

The Ministry of Water Resources in an effort to develop and utilize the countries water resources in a systematic way has prepared policy, strategy and development plans for effective and efficient water use. Through the river basin master plan studies all the potentials for irrigation developments are identified and there exists over about 3.5 million hectares of land suitable for irrigation development. In addition, improving water management in the rain fed systems as well has significant scope to increase productivity of agriculture

Recognizing these facts, while small scale irrigation and rainfed agriculture water management are undertaken by regional government, medium and large scale irrigation developments have been given significant attention by the federal government and the Ministry of Water Resources xvii

in it's development program. Accordingly, 487,000 ha of land are planned to be irrigated during the PASDEP period (2009/10) in addition to the existing one. Some of these projects are ready for detail design and construction and some are already under construction. This clearly shows, substantial investment in irrigated agriculture is needed to meet targets for poverty alleviation, food security and economic growth. I would like to also stress that some of the large and medium scale irrigations are and will be designed to benefit the smallholder farmers, as the government of Ethiopia strongly committed to eradicate the rural poverty. Notable example is the Koga irrigation development which is designed to develop about 6,000ha with over 7,000 (?) beneficiaries. As the development of large schemes for small holders is new experience for the country, it is important for research institutions to support such endeavors by undertaking adoptive and applied research which can support this and future development in sustainable manner

While many Sub-Saharan countries and Ethiopia alike committed to water infrastructure development, lending for irrigation in Sub-Saharan Africa has declined considerably over the past few decades. But still, there are reports that indicate many Bank-financed irrigation projects had produced satisfactory outcomes and the outputs of this particular research are expected to prove that the benefits of investment have reached the poor. However, it is clear that there are issues to be addressed and constraints to be overcome if investments in agricultural water or irrigation development are to achieve viability and sustainability.

Among other factors, capacity building stands as a key factor to obtain diversified expertise and to increase knowledge that is required for sustainable irrigation and drainage. There are pertinent issues with regard to technology, material and equipment selection and even methodologies in engineering design of irrigation projects. In this regard, researchers are expected to contribute significantly towards scientific, practical, and multidisciplinary solution for the prevailing irrigation development constraints and be able to advice decision makers. At this juncture, I would like to mention the MOU signed with International Water Management Institute (IWMI) is having such an objective of initiating conceptual and practical research on high priority areas in collaboration with Ethiopian partners, including issues that would have impact on the development.

Another important area is to ensure that all the partners public, private, civil society as well as donor/lending partners, have sufficient information from our data base to allow them understand the benefits of irrigation development from both social and economic perspective. I hope studies like the Impact of Irrigation on Poverty and Environment for which this symposium is organized can help to enhance such understanding and synthesize the new knowledge generated.

Finally, I would like to thank IWMI, BOKU, ARC Sieberdorf, and Ethiopian Institutions such as Arba Minch University, Haramaya University and Ethiopian Institute of Agricultural Research and other collaborating institutions for the initiative that they have taken in support of irrigation development in general and conducting the related research. The Ministry of Water Resources would like to reiterate its commitment to collaborate and closely work with all stakeholders that are interested in promoting and developing the water resources of this country.

Finally, for those of you who came from abroad wishing you a pleasant stay in Addis Ababa, I wish you all, success in your deliberation and I declare the workshop open.

Good bless you I thank you xviii

Theme one:

Status Quo Analysis, Characterization and Assessment of Performance of Irrigation in Ethiopia

Papers 1. Status quo analysis, characterization and assessment of performance of irrigation in Ethiopia 2. Comparison of irrigation performance based on management and cropping types 3. Irrigated and rainfed crop production systems in Ethiopia 4. Baseline survey of irrigation and rainfaid agriculture in Ethiopian part of Nile Basin

Statues quo analysis, Characterization and Assessment of Performance of Irrigation in Ethiopia

Seleshi B. Awulachew and Aster Denekew Yilma International Water Management Institute for Nile Basin and East Africa [email protected]

Abstract investment cost, etc, are some of the important attributes of the database. In This paper first looks in to the background addition, schemes that are operational and on major challenges of Ethiopia with respect failed are identified in the database. Based to poverty. It discuses the root cause of on the broad database, performances of the poverty and its vicious cycle nature, the schemes are highlighted. Furthermore, the interlink of population growth, the scarcity various sites that are used in the detail study of land and natural resources, the extension and the selection criteria for the impact of of agriculture in to marginal land, the irrigation on poverty and environment decreasing productivity, inability to invest project and the specific characteristics of and deepening of poverty and further these sites are described. aggravation as a result of various shocks such as drought, flood, war, etc. The paper Key words: poverty, water scarcity, also looks in to the importance of the broad database, GIS, irrigation, water resources. agricultural water management in general and irrigation in particular with respect to 1. Introduction increasing productivity and capability to break the vicious cycle and opportunity to Ethiopia is mainly agrarian nation and the reverse in to virtuous cycle that can help rainfed system has always played a central eradicate poverty and develop the poor role in Ethiopian society. Dependency on economy. The paper also looks in to how rainfed system has put more than 80% of the poor management of water resources and society at the mercy of meteorological impacts of variability of rainfall and related variability. drought affecting the socio-economy and the overall wellbeing of the country to the Ethiopia's agricultural population extent that significant population became dependent on imported food. Results of 90 broad assessment of water resources, 88 database of irrigation development and 86 potential, characterization by typology and 84 major performance in Ethiopia are 82 presented. Key water resources information 80 related to each of the 12 river basins in 78 Percentage of population Percentagepopulation of Ethiopia is summarized. Details of existing 76 irrigation and future potential are also 1979-1981 1989-1991 1999-2001 2003 2004 Year captured in the paper. A geographic information system (GIS) database Figure 1: Ethiopia’s agricultural population describing irrigation by typology, region and (data extracted from the World Bank location, scheme size, type of structures, Development Indicator WB 2006b) water source, number of beneficiaries, 2

Figure 1, shows the dependency of of Ethiopia and eradicate poverty, the policy Ethiopian population on agriculture and in and interventions should focus on 25 years the agricultural population reduced Agriculture as entry point. The current rural only from 89% to 81%. Agriculture in development policy and strategy of the Ethiopia is dominated by small holder government clearly stipulates this as production of cereals under rainfed priority. condition, accounting a total area of approximately 10 million hectares. Building further on the above factors, the According to Central Statistics Authority poverty situation in Ethiopia is a vicious [CSA, 1995-1999], within agriculture, some cycle in nature and requires key entry points 60 percent of the output is from crops, with for intervention. The following figure is a livestock and forestry producing 30 percent schematic example showing poverty is and 7 percent respectively. Crop production linked to and aggravated by various by area is predominantly cereals (84.55 demographic, biophysical, production percent) followed by pulses (11.13 percent) system, productivity and other socio- and others (4.32 percent). Five crops economic factors. account for almost all cereal production: maize (15.75 percent), teff (Eragrostis tef) Scarcity of land (25.78 percent), barley (12.29 percent), in the sorghum (12.39 percent) and wheat (10.76 settlement areas Population Deforestation, percent). According to Mulat et al (Mulat et growth land & water al 2004), agriculture remains the main degradation activity in the Ethiopian economy. It is the most important contributor to the country’s Inability to Agriculture GDP: accounted, on the average, 65.5%, invest to pushed in to or improve marginal land 52.7% and 47.1% of the GDP during 1960- land 1973, 1974-1991 and 1992-2002, productivity respectively. Poor health, Poor malnutrition productivity, Despite the above mentioned facts, there are food insecurity Deepening a number of factors that led to failure of poverty achieving food security in Ethiopia. The major causes for food insecurity in Ethiopia can be associated to the following: - Population growth and associated Figure 2: The vicious cycle of poverty and inadequate resource base to support aggravating factors in Ethiopia - Lack of growth of production and productivity Socioeconomic development and - Vulnerability to climatic variability civilization of human being is closely - Political instabilities and war associated to ability to utilize and control - and poverty water resources. Water serves as a positive input for many activities and play negative The main development objective of the roles. Positively, it serves essential Ethiopian Government is poverty biological needs, as basic element of social eradication. Hence, the country's and economic infrastructure, and as a natural development policies and strategies are amenity contributing psychological welfare. geared towards this end (MOFED: PASDEP Water also serves in negative roles such as 2006). As Ethiopia’s economy and majority flooding and diseases transmission. In of people’s livelihood is dependent on Ethiopia, as in all societies, there has always Agriculture, to develop the socio-economy been a struggle to reduce the negative/destructive impacts of water and 3

enhance its positive/productive impacts, but linkage to rainfall variability, the with limited focus and capacity. These economy of the country will continue to efforts have been increased since the past be severely affected. Particularly, this three to four decades and more so during the could be more severe under the strong last few years. However, the ability to use impact of climate change and and enhance the positive role of water and to variability. WB (2006) and IWMI reduce its negative impacts, in Ethiopia in (2007) describe the impact of costs general has been low. 1/3rd of growth potential of Ethiopian economy. The impact of this can be 2. The Importance of Agricultural Water shown from the recent information. Management and Irrigation in Ethiopia According to MOFED (MoFED 2006), GDP Growth of Ethiopia in 2002/3 was It is essential to increase agricultural -3.3% during the drought year while the productivity in order to eradicate poverty, previous and latter years were positive. improve the economy, and reduce In 2004/5, GDP growth was 11.9% and degradation. Irrigation and improved 2005/6 was 10.6%, which brings the agricultural water management practice is three year average down to 6.4%. important in Ethiopia for the following - Improved agricultural water major reasons: management and irrigation can increase - Population in Ethiopia is rapidly productivity of land, water and labor. increasing (over 80,000,000 currently), The following figure based on Central land holding size particularly in Statistical Authority data and Mulat et al highland areas is decreasing (2004) shows the crop productivity and substantially. Intensification and productivity growth for the period of the increasing productivity of land and last two decades for major crops in labour is essential to produce enough Ethiopia. However, recent data of food, particularly from the limited 2004/5 onwards and predicted available land. productivity data according to MOFED - Agriculture is primarily rain fed (2006) shows there has been increase in dependent. Unless the rain fed system is productivity of cereals. The increase is upgraded through improved water mainly attributed to increased input use management, recurrent drought and dry (seed, fertilizers, and pesticides) and spell continue to affect productivity and improved water management for hamper agricultural production agriculture in certain areas. The strategy - Ethiopia’s economy is strongly to achieve the future targeted result dependent on rain fed based agriculture, focuses to use intensification (irrigation, and rainfall variability impact costs the vertisol management, seed, fertilizer, economy significantly. Therefore, unless pest control) and expansion. agriculture is de-linked from the strong

Cereal Yields in Ethiopia (1980-2001) Cereal Yield t/ha Pulses yield in t/ha 1.6 Oilseed Yield t/ha) 1.4 1.2 1 0.8 0.6

Yield in ton/ha 0.4 0.2 0 1980 1983 1986 1989 1992 1995 1998 2001 Year

Figure 3: Crop yield in Ethiopia for the period of 1980-2001 (Data source: Mulat 2004)

Productivity (Q/ha) plan

40 34.84 35 31.57

30 25.85 22.91 25 20.03 20 14 .3 5 15

0 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10

Year

Figure 4: Cereal productivity and productivity Plan of Ethiopia (Data source: MOFED, PASDEP 2006)

important measure that may be pursued in - Contrary to the first bullet above, there are Ethiopia to cope with complex problems considerable land and water resources in various relatively remote parts of the leading to poverty and insufficient food country. The constraining factors for production development however are low infrastructure that includes accessibility such as roads, and The major sources of growth for Ethiopia is communication; unregulated water still conceived to be the agriculture sector, resources; no settled people to develop the as it is expected to be insulated from drought resources; lack of capital; and lack of shocks through enhanced utilization of the knowledge and capacity. Improving on these water resource potential of the country can enhance development of these resources. (through development of small scale Particularly, the irrigation development irrigation, water harvesting, and on-farm through improved infrastructure is an diversification), coupled with strengthened linkages between agriculture and industry 5

(agro-industry), thereby creating demand for the tarmac Addis-Assab road opened the agricultural output (MOFED 2006). Awash Valley to ready markets in the hinterland as well as for export (Metaferia, Irrigation development, including large and 2004). Although certain aspects of the medium scale irrigation development, as development during the pre-Derge era have public schemes, commercial farming and for wrong doings in terms of property and land small holders are getting importance under rights, there has been remarkable emergence the current government, particularly since of irrigation development and establishment 2004. of agro industrial centers. Teshome (2003) has reviewed and discussed the land tenure 3. Irrigation Status in Ethiopia system in the various regimes in Ethiopia. These establishments were highly motivated 3.1 History private sectors, which are both export and domestic market oriented. There is no documented history of water management for agriculture. Remnants of During the Derge era, all private farms were millennium old water storage structures for nationalized to establish the so-called state non-agricultural use around Axum in Tigray farms, thereby ending the embryonic private show the oldest usage of water in a sector. The government pursued the controlled manner. Certain, non-irrigation development of medium and large-scale related technologies to conserve water and irrigation schemes in a number of river soil have been practiced by Konso people in basins in addition to expansion in the Awash the South, at least for the last four hundred Valley. The Amibara Irrigation Project in years. However, there is no well- the Middle Awash, Alwero Irrigation Project documented resource material on water use in Gambella, Gode-West Irrigation near for irrigation in Ethiopia. Modern irrigation Gode town, the Omorrate Irrigation scheme development in Ethiopia is not having in Southern Omo, the Tana Beles, the centuries old history. There is no written Fincha Suger State, etc are some of the history on how Ethiopia has used irrigation expansions, most of which are suspended technologies to secure agricultural currently. production, as the vast country with small population had adequate natural resources Following the downfall of the Derge, the base and rainfall to produce the food current government withdrew from the requirements without the need to develop expansion of State Farms and further irrigation. construction of medium and large-scale irrigation (Metaferia, 2004). This has been Private concessionaires who operated farms the trend until the aftermath of the 2002/3 for growing commercial crops such as severe drought that has caused about 15 cotton, sugarcane and horticultural crops Million population under extreme food started the first formal irrigation schemes in shortage. Not only the government hesitated the 1950s in the upper and lower Awash to expand medium and large scale irrigation Valley. In the 1960s irrigated agriculture but also it has interrupted finalization of the was expanded in all parts of the Awash above 5 major irrigation projects started in Valley and in the Lower Rift Valley. The the former regime. On the other hand, the Awash valley saw the biggest expansion in government indeed provided certain view of the water regulation afforded by the attention on small scale irrigation mostly in construction of the Koka dam and reservoir the food insecure areas. Nevertheless, in the that regulated flows with benefits of flood water sector development program (WSDP) control, hydropower and assured irrigation it was identified to expand large and water supply. In addition, the construction of medium scale irrigation by about 147,000 ha 6

and small scale irrigation by about 127,000 ha. As strategy of developing irrigation 3.2 Water Resources, Irrigation Typology sector, the plan of the government targets to and Existing Schemes develop a total of additional 274,612 ha of land which brings the total irrigated area of Ethiopia has 12 river basins. The total mean about 478,000 ha by 2015. Despite ignoring annual flow from all the 12 river basins is the medium and large-scale sector for a estimated at about 122 BMC (WRMP, decade long, recently after the development 1999); Figure 5 and Table 1 show the river of the water sector development program, basins and distributions of water resources there is a growing attention to the irrigated in various basins. The water resources agricultural sector. The revised strategy even distribution shows slightly higher values as plans to put more irrigated land in short extracted from recent master plan studies. period of time.

Figure 5: Ethiopia’s River Basins

Table 1: Water resources distribution by river basins of Ethiopia Estimated ground River Basin Area (Km2) Runoff (Bm3) water potential (Bm3) Tekeze 82,350 8.2 0.20 Abbay 199,812 54.8 1.80 Baro-Akobo 75,912 23.6 0.28 0.13 Rech/yr Omo-Ghibe 79,000 16.6 0.42 (.10) Rech /yr Rift Valley 52,739 5.6 0.10 Mereb 5,900 0.65 0.05 Afar /Denakil 74,002 0.86 - Awash 112,696 4.9 0.14 Aysha 2,223 - - Ogaden 77,121 - - Wabi-Shebelle* 202,697 3.16 0.07 Genale-Dawa* 171,042 5.88 0.14

Total 1,135,494 124.25 2.86 Source: IWMI Working paper 123 (Awulachew et. al, 2007)

In addition, Ethiopia has also 11 fresh and 9 associations that have by laws and saline lakes, 4 crater lakes and over 12 traditional schemes – developed and major swamps or wetlands. Majority of the managed by community tradition Lakes are found in the Rift Valley Basin. and usually characterized by non For details refer Awulachew et al (2007). fixed structures and practiced The total surface area of these natural and traditionally. artificial lakes in Ethiopia is about 7,500 o Medium scale: Schemes exceeding km2, representing about 0.67% of area of 200ha but less than 3,000ha Ethiopia. Most of the lakes except Ziway, o Large scale: schemes exceeding Tana, Langano, Abaya and Chamo have no 3,000ha surface water outlets, i.e. they are endhoric. Lakes Shala and Abiyata have The latter two are mostly public schemes, concentrations of chemicals. owned and managed by the government, and in certain cases by large communities. 3.2 Irrigation Typology There are also irrigation typologies that are The irrigation schemes in Ethiopia are not clearly captured in policy and strategy divided according to the following typology: documents. These are o Small scale: These are schemes less o Water harvesting based irrigation; than 200ha. Two major categories e.g. Household based minute under this are modern schemes irrigation; which usually have fixed or o Ground water irrigation; improved water control/diversion o In-situ Agricultural Water structures and water users Management. 8

etc. For details refer Awulachew et al (2007).

Based on this the following map is one of 3.3 Existing Irrigation Schemes the products of the database, showing irrigation distribution in Ethiopia based on One of the objectives of impact of irrigation attribute of typology and regions. Note also on poverty and environment project is to that the map is not showing the complete list develop GIS database of irrigation schemes of the irrigation schemes, as some of the to understand the spatial distributions and geo-referencing information is missing and their characteristics. Accordingly, a database the map represents 107 schemes of complete have been developed for about 790 modern large and medium scale irrigation and some irrigation schemes having various attributes small-scale irrigation. such as name, administrative locations, georefenece, type of irrigation, typology,

e TIGRAY k An e Irrigation Schemes ger z eb R e iv e R Atbar r a R i iver v Large Scale e

r AFAR Medium Scale

AMHARA Small Scale Rivers BENISHANGUL Ab ba y R Lakes -GUMUZ iv er DIRE DAWA

ADDIS ABEBA G HARARI i b e Baro River R i v e OROMIA r GAMBELLA G oje SOMALI b R iver SOUTHERN Wa REGION bes G heb e ele r na Ri e le ver v i R iv R e r

0165 330 660 · Kilometers Figure 6: Existing Irrigation schemes distributed in the regional states of Ethiopia

According to MOFED (2006), with respect are also many additional scale development to irrigation development, within the projects under construction invested by program period of PASDEP 2004/2005 to regional governments, donors and NGOs 2009/2010, pre-design studies will be and private sector. carried out for 17,988 hectares, full-fledged design studies will be undertaken on

464,051 hectares, and construction works 3.4 Irrigation Potentials by River Basins will be completed for 430,061 hectares. Currently, actual implementation projects at Tendaho and Kessem totaling about 90,000 In Ethiopia, under the prevalent rain fed ha in the Awash Valley, 7,000ha Koga agricultural production system, the irrigation development in Blue Nile River progressive degradation of the natural Basin are actually near completion. There resource base, especially in highly 9

vulnerable areas of the highlands coupled basins. The total potential irrigable land in with climate variability have aggravated the Ethiopia is estimated to be around 3.7 incidence of poverty and food insecurity. million ha (Awulachew et al 2007). Table 2 Water resources management for agriculture and Figure 7 show the irrigation includes both support for sustainable development potential by river basins in production in rain fed agriculture and Ethiopia. Detail characterization of the irrigation (Awulachew et al 2005). potentials by basins is provided in Awualchew et al (2007). Currently, the MoWR has identified 560 irrigation potential sites on the major river

Table 2: Irrigation Potential in the River Basins Basin Catchment Irrigation potentials (Ha) WAPCOS, 1995 Area (Km2) (Respective recent master plan studies) Small Medium Large Total Total Irrigable % Irrigable scale scale scale Drainag Area Area of the e Area (Ha) Country (km2)

Abbay 198,890.7 45,856 130,395 639,330 815,581 201,346 1001000 27 Tekeze 83,475.94 N/A N/A 83,368 83,368 90,001 317000 8.5 Baro-Akobo 76,203.12 N/A N/A 1,019,523 1,019,523 74,102 985000 26.5 Omo-Ghibe 79,000 N/A 10028 57900 67,928 78,213 445000 12 Rift Valley 52,739 N/A 4000 45700 139,300 52,739 139000 3.7 Awash 110,439.3 30,556 24,500 79,065 134,121 112,697 205000 5.5 Genale-Dawa 172,133 1,805 28,415 1,044,500 1,074,720 117,042 423000 11.4 Wabi-Shebele 202,219.5 10,755 55,950 171,200 237,905 102,697 200000 5.4 Danakil 63,852.97 2,309 45,656 110,811 158,776 74,102 - - Ogaden 77,121 - 77,121 - - Ayisha (Gulf 2,000 - 2,000 - - of Aden) Total 1,118,074.53 3,731,222 982,060 3,715,000 100 Note: The national water resources master plan (WAPCOS, 1995) was a desk study without significant field investigation.

Figure 7: Irrigation Potentials in Ethiopia by River Basins

The complete database is developed for be the policy of market economy precluding existing irrigation development and irrigation government involvement in such economic potential. We trust that this database creates activities added with the complexity of the important information system and a foundation projects that were under establishment at for complete and comprehensive database that remote areas with low infrastructure, can be updated continuously for irrigation insufficient labor and market linkage. development in Ethiopia. The database is also However, the wisdom of the decision, for made available to regional irrigation abandonment of development schemes on development bureaus and federal institutions which hundreds of million have been invested, for use and further updating. This information remains to be questionable. It might be wiser system establishes a public good and any to finalize the schemes and settle smallholders interested institution or individual can receive of the area and/or encourage private operators a copy. The available formats for sharing to take over under an attractive/ acceptable include GIS products, Microsoft Excel or arrangement. On the contrary, private Microsoft Access database categorized per initiatives to takeover and finish some of the typology, river basins and regions. schemes - Meki-Zeway, Belbela & Wedecha, Alwero - either have not been accepted or have

failed of their own accord until recently 4. General performance of the irrigated whereby the former two have attracted the systems attention of flower farmers. According to

MCE (2004) some of the schemes have been The performances of the existing irrigation turned over to party affiliated companies with schemes are highly variable. Some of the limited success. These projects represent schemes from all typologies in terms of water priority schemes for rehabilitation and use efficiency, productivity, sustainability are completion. performing very well, while some are not performing efficiently, interrupted while under Besides the suspended schemes a total of construction, abandoned after implementation, 26,347 ha are transferred from public to or transferred from public to private or private or communal developers. The community and their performances are not operation of this transferred schemes are known. Many successful schemes are variable. Some are successful, some are failed providing increased income, higher after transfer and the performances of some are productivity, significant job opportunity and not known. MCE (2004) and Awulachew et al considerable contribution to the economy. The (2007) discuss these. existing irrigated schemes are estimated at about 2% of the total agricultural land but 5. Conclusion contribute over 5% of the agricultural production. On the other hand, there are also a This paper, which is related to the wider number of schemes with critical problems impact of irrigation on poverty and leading to complete abandonment or under environment research project, provided performance, missing the targets of information and database on the water performance in terms of land area developed, resources of Ethiopia, potential of number of beneficiaries or sustainability. The development, extent of existing development assessment related to the database focusing on irrigation development. It also development, see also Awulachew et al (2007) discussed irrigation development categorized reveals that 17% to 22% of schemes in the by various river basins and regions. Amhara, SNNPR, Oromia and Tigray, Discussions were also made on schemes that particularly small scale irrigation schemes fall are non-operational or transferred to under this category. community and private sector and their

implication on performance. A number of medium and large irrigation Specific database is also developed for existing schemes, with a total area of 44,050 hectares, irrigation schemes having a number of that were under construction, during the attributes. The developed database has previous government, were suspended by the information about the existing irrigation present one. The underlying reason seems to

12 schemes and potentials. The database under MoFED (2006). Ethiopia: Building on GIS environment, maps their spatial Progress A Plan for Accelerated and distribution using point maps from those Sustained Development to End Poverty schemes for which geo-refenced data is (PASDEP) (2005/06-2009/10) Volume I: available. Main Text Ministry of Finance and Economic Development September, Addis It is obvious that Ethiopia is extremly Ababa. dependent on rain fed agriculture; its majority of population are dependent on agriculture MoWR (Ministry of Water Resources). 1999. without limit to move out of the sector, Water Resource Management Policy agriculture being at low productivity, rapid (WRMP), Addis Ababa: Ethiopia. population growth and lack of innovation to maximize the benefit of the combination of MoWR (Ministry of Water Resources). 2002. population, land and water. Hence, most of the Water Sector Development Program population are poor and agriculture and overall (WSDP), Addis Ababa: Ethiopia. economy is vulnerable and remains very weak against the shocks of the climatic variability. Mulat Demeke, Fantu Guta and Tadelle Fered. The last five years attentions towards 2004 Agricultural Development In development taking the rural development Ethiopia: Are There Alternatives To Food policy and strategy, the water sector policy, the Aid? Department of economics, Addis irrigation development strategy, the PASDEP Ababa University. Unpublished Report actions are encouraging and hoped to accelerate development endeavors. Ethiopia’s WAPCOS (Water & Power Consultancy challenges towards development are immense Services (I) Ltd.). 1995. The National and require significant actions and efforts Water Resources Master Plan, Addis addressing the various problems from various Ababa: Ethiopia. sectors that speed up rapid development. References World Bank (WB) 2006a. Ethiopia: Managing Water Resources to Maximize Awulachew, S. B., Yilma, A. D., Loulseged, Sustainable Growth: Country Water M., Loiskandl, W., Ayana, M. and Resources Assistance Strategy. The World Alamirew, T. 2007. Water Resources and Bank, Washington DC Irrigation Development in Ethiopia. IWMI Working Paper 123 World Bank (WB) 2006b. Africa Development Indicators 2006. International Bank for Awulachew, S.B., Merrey, D.J, Kamara, A. B., Reconstruction and Development. Van Koopen, B., De Vries, F. Penning, Washington DC and Boelle, E., 2005. Experiences and Opportunities for Promoting Small-Scale /Micro Irrigation and Rainwater Harvesting for Food Security in Ethiopia, IWMI Working Paper 98, 2005

IWMI (International Water Management Institute) 2007. Ethiopia: Agricultural Water Management Policy Briefs. Issue 1, 2007

MCE (Metaferia Consulting Engineers). 2004. The World Bank. Assessment of experiences & opportunities on medium & large-scale irrigation in Ethiopia. Draft Report

Comparison of irrigation performance based on management and cropping types

Mekonen Ayana and Seleshi Bekele Awulachew Arba Minch University, Arbaminch, Ethiopia International Water management Institute for Nile Basin and East Africa [email protected]

physical structures, management and Abstract operation of the system at all levels will bring substantial improvement in the Although performance evaluation of performances of cotton producing schemes. irrigated agriculture has gained momentum since late 1980s worldwide such attempt is 1. Introduction rarely carried out in Ethiopia. The aim of this study was to assess the performance of 7 Irrigation is highly expected to play a major irrigation schemes some of which are role in the realization of Ethiopian food expected to contribute much to the national security and poverty alleviation strategy. economy. Sugar cane is grown by three of Irrigation enhances agricultural production these schemes whereas cotton is grown by and improves the food supply, income of three schemes and the remaining single rural population, opening employment scheme grows tobacco. With regards to opportunities for the poor, supports national management types both government agency economy by producing industrial crops that and community managed schemes are are used as raw materials for value adding considered. The scheme level values of industries and exportable crops. From this water supply performance indicators show important viewpoint irrigation projects are that there was no constraint of water widely studied, planed and implemented availability and supply at scheme level. In throughout the country. However, little or no general, schemes that grow sugar cane were attention is given to the monitoring and found to have attained higher outputs per evaluation of the performance of already units of land and water used which ranges established irrigation schemes. Whether from 7794 – 10834US$/ha and 0.24 – 0.55 traditional or modern, public agency or US$/m3 respectively. On the other hand, community managed many of the existing whether state farm or community managed, irrigation systems are deteriorating in their schemes that grow cotton have shown low physical structures, operation and output per units of land and water, i.e. 310 – management. 385 US$/ha and 0.01 – 0.05 US$/m3 Performance assessment is used to identify respectively. Large productivity the present status of the scheme with respect performance differences have been observed to the selected indicators and will help to between irrigation schemes with same identify ‘why the scheme is performing so’ cropping and management types. From which in turn imply means of improvement. scheme level performance values it is not Of course performance evaluation needs simple to identify the area where and what is relevant and reliable data which is rarely going wrong which is responsible for low measured in Ethiopia. performance. Generally, problems casing According to Clemmens, A.J. and Molden. low productivity derive both in management D.J. (2007) two major approaches to and deterioration of physical structures. performance evaluation are to consider, how Hence investment on improvements of well service is being delivered and the

14 outcomes of irrigation in terms of efficiency operation of facilities, maintenance and and productivity of recourses use. To proper application of irrigation water and measure these performances a number of agronomic activities (Small L. and Svendsen indicators have been proposed and tested in M. 1990). Facilitation and execution of these different parts of the world (Molden, D. et activities requires proper coordination of six al. 1998; Kloezen W.H., 1998; Burton, M. et functional processes of irrigation, i.e. al. 2000; Lorite, J. et al. 2004, Bos, M.G. et personnel management and support, al. 2005, Vandersypen et al., 2006). IWMI’s equipment management, financial minimum sets of performance indicators management and accounting, and resources were used by many researchers to compare mobilization. Planning, design and different irrigation schemes. Comparison construction of irrigation schemes are helps to identify ‘who is doing what right’ mainly dealing with creation of physical and what lesson can be learnt or who can be infrastructure to facilitate the capturing of a benchmark for a particular activity. The water from its source and transportation up objective of this study was to assess the to the farm level. These physical facilities performance of 7 irrigation schemes in need to be properly operated to ensure the Ethiopia based on management and capturing, allocation and delivery of water at cropping types using IWMI’s performance the right time and adequate quantity. indicators. Maintenance activities are designed to ensure the capabilities of physical 2. Technical background on infrastructure to deliver the intended amount performance assessment of water over the project life time. Application of water to the field is the core Performance can be simply defined as “the activity of irrigation which is designed to level of achievement of desired objectives” disperse the incoming stream from higher (Mohtadullah, K., 1993). Indicators are used level canal over the field thereby storing in to measure performance. An indicator is the crop root zones. Substantial some number that describes the level of improvements in the performance of such a actual achievement in respect of one of the complex system is not possible by making objective of irrigation system. Indicators are big improvements at only one level within used to simplify the otherwise complex the system (Clemmens A.J. & Molden D.J., internal and external factors affecting the 2007). Physical or management performance of irrigated agricultural system. improvements may need to be made at all Performance can be measured from process levels before substantial improvements in and output points of view. Process measures the performance can result. of performance relate to a system’s internal operations and procedures whereas output Gorantiwar S.D. and I.K. Smout (2005) have measures of performance examine the summarized performance measures quality and quantity of the system’s final proposed by various researches into output (Small, L. and M. Svendsen, 1990). allocation type and scheduling types. While quoting the value of certain Allocation types performance measures are indicators, at a particular irrigation system those which need to be attained primarily and time, it means that all other factors and during the allocation of the resources at the processes are ignored or neglected. The fact planning and operation stages. Productivity that an indicator services as a guideline for and equity are performance measures under further decision making it should be allocation type category. Scheduling type carefully chosen, measured and interpreted. performance measures consists of irrigation scheduling, i.e. temporal and spatial Irrigation performance, whether bad or distribution of irrigation water to the users. good, is the result of verities of activities This measures adequacy, reliability, such as planning, design, construction, flexibility, efficiency and sustainability.

15 Scheduling should be such that water equivalent) per water supplied to the scheme deliveries need to be adequate both in during the season. Land productivity on the planning and operation, reliable, flexible and other hand is production per unit of land sustainable. The same authors grouped these cultivated. two categories of performance measures into: economic (productivity), social 3. Materials and Methods (equity), environmental (sustainability) and management (reliability, adequacy, 3.1. Description of the schemes efficiency and flexibility). Conveyance efficiency is used to compare This study uses six government owned the amount of water delivered at the turnouts irrigation schemes for detail investigations of the main irrigation conveyance network which are believed to have large to the total amount of water delivered into contribution to national income and one the irrigation scheme. Its measurement is community managed irrigation schemes. important in that water allocation plans are The schemes are geographically located in developed using estimated efficiencies of south, east and central parts of the country. water flow at various stages and time. Table 2 gives brief information on the Deterioration of efficiency over the years schemes. For details on the characteristics of will reduce the performance of the irrigation the schemes, see Girma and Awulachew scheme over this period. Gorantiwar and (2007). I.K. Smout (2005) categorized the importance of efficiency in two ways: 3.2. Performance indicators Firstly, appropriate optimum allocation plans cannot be developed if proper The performance indicators adopted in this consideration is not given to efficiency. study are: Inaccurate or simplified estimates also have 1. Irrigation water delivery a major influence on other performance performance parameters such as productivity, adequacy, a. Conveyance efficiency (EC ) equity and reliability. Secondly, the b. Annual relative water supply inspection of efficiencies over space and (ARWS) time at different levels enables the irrigation c. Annual relative irrigation supply authorities to learn which part of the scheme (ARIS) is inefficient, where it is inefficient and how d. Water delivery performance or it is deteriorating. water delivery ration (WDR)

Productivity is related to output from the 2. Output performance indicators system in response to the input added to the a. Output per harvested area system and there are several indicators of (tons/ha) productivity. The primarily output of the b. Output per harvested area scheme is the total crop yield or its (US$/ha) economic equivalence per units of land or c. Output per command area water used. Hence, most often the (US$/ha) productivity is expressed in terms of land or d. Output per water supplied water supplied to produce a certain level of (US$/m3) output. Water productivity deals with the amount of production (mass or monetary

These indicators were measured using the following mathematical descriptions: Water flowing out of the system E = (1a) C Water flowing into the system Total volume of water sup plied ARWS = (1b) Total volume of crop water demand Total volume of irrigation water diverted ARIS = (1c) Total volume of irrigation water demanded Volume of water actually delivered WDR = (1d) int ended volume of water tobe delivered Pr oduction(tons) Outpout per harvested area (tons / ha) = (2a) Irrigated cropped area (ha) Local value of production(US$) Outpout per harvested area (US$ / ha) = (2b) Irrigated cropped area (ha) Local value of Pr oduction(US$) Outpout per command area (US$ / ha) = 2c) Command area (ha) Local value of Pr oduction(US$) Outpout per water sup plied (US$ / m3 ) = (2d) Diverted irrigation sup ply (m3 )

Data used in this study are emanating from different sources. Sugar estates have their own records regarding annual production, water diverted to the schemes and meteorological data.

Table 1: Sources of important data Data Irrigation schemes Metahara Wonji Finchaa Hare Sille Bilate Metro Data Estate Estate Estate NMSA NMSA Tessema, 2006 Production Estate Estate Estate Belete, Aklilu, Tessema 2006 2006 Water supply Estate Estate Estate Belete Aklilu ARWS calculated calculated calculated Belete Aklilu Tessema ARIS calculated calculated calculated Belete Aklilu Tessema Efficiency measured measured Belete Aklilu

17 Table 2: Characterization of selected irrigation schemes Scheme name Hare Sille Bilate Metahara Wonji Finchaa Latitude 6º 30´ to 6 º 38´ N 5º 49' to 5º 55' N 6048’ to 6050’N 8º 21' to 8º 29' N 8º 21' to 8º 29' N 9º 30' to 9º 60' N Longitude 37 º 33´ to 37º 37´ E 37º 26' to 37º 29' E 3804’ to 3805’ 39º 12' to 39º 18' E 39º 12' to 37º 10' to E 39º 18' E 37º 30' E Average annual rainfall 830.7 748 734 659.6 832 1300 (mm) Average annual ETo (mm) 1651.2 1540 1958 1596.5 Predominant soil types Sandy loam to clay soil Silty loam and clay Sandy to loam Sand to clay loam Clay, light soil black heavy loam clay Water source Hare River Sille River Bilate River Awash River Awash River Finchaa River Water availability Scarce in some periods Scarce in some sufficient abundant abundant abundant periods Irrigated area (ha) 1962 1082 870 11058 7279.8 8500 Main crops Banana, cotton, maize, Banana, cotton, Tobacco, maize Sugar cane Sugar cane Sugar cane, fruit trees, sweet potato, maize horticultural vegetables crops Year first operational 1996 1957 1962 1966 1954 1991 Type of management community government Government/ government government government private Land ownership private government Government government government government Method of water abstraction Gravity (inundation, Gravity Gravity Gravity (weir) Pump Gravity/ weir weir) (inundation) (barrage) Water delivery infrastructure Open channel Open channel Open channel/ Open channel Open channel Open pipelines channel/pipe Predominant on-farm water Furrow, basin Furrow, basin furrow furrow Furrow Sprinkler/ application method furrow

18 4. Results and Discussions irrigation scheme. Through filed measurements it was evidenced that the 4.1. Water delivery performance canal losses more than 50% of water over 5 km canal distance from the diversion point. Water delivery performances considered are As the physical conditions of canal in Hare conveyance efficiency, annual relative water irrigation scheme is bad, the losses are supply, and annual relative irrigation supply mainly attributed to seepage from the canals. and water delivery ratio. The results of Moreover, even if they are closed, points of conveyance efficiency measurements given unauthorized water turnouts contribute also in Fig. 1 show that there is a high water loss to low conveyance efficiency because of especially in community managed Hare leakages.

120

100

80 Metahara 60 Wonji Hare 40

Conveyance efficiency (%) Conveyance efficiency 0 0 2000 4000 6000 8000 Distance from headwork (m)

Fig. 1: Variation of conveyance efficiency along the canals of some schemes

Conveyance efficiency underlies spatial observed to be reasons for rapid decline of variations based on the conditions of the conveyance efficiency in Hare irrigation canal and management system. Farm units schemes which are located along the canal segment Previous studies in Wonji indicated that with low conveyance efficiency tend to seepage losses in the tertiary canals account suffer from unreliable and untimely supply to about 40% and contributed to rising of of water. groundwater level to 0.94m below the These problems have been observed in surface (Habib, 2005). community managed irrigation schemes such as in Hare. To this effect, farmers The values of water delivery performance, located at tail-end of the canal (>7km) are i.e. annual relative water supply (ARWS) limited in their crop diversification and and annual relative irrigation supply (ARIS) forced to grow relatively water stress are given in table 3. These indicators are resistant crops such as cotton and sweet evaluated as optimal if their values would be potato. Not only bad conditions of physical equal to one. Less or greater than one would structures but also leakage through mean under or over supply of water unofficial points of water turnouts are respectively. ARWS relates the total volume

19 of water applied (irrigation plus total water than theoretically forecasted. Wonji rainfall) to the volume of water required by scheme is characterized by lower values of the crops. It can also be used both as a all water supply and delivery performance measure of adequacy and seasonal timelines indicators compared to other schemes. The (Levine 1982 and Meinzen D., 1995, In: cost involved in pump diversion might have Kloezen W.H & G.-R. Carlos, 1998). contributed to efficiency of resources use in Wonji The annual relative irrigation supply (ARIS) on the other hand is the ratio of the volume 4.2. Production per unit area of irrigation water delivered to the volume of irrigation water demanded (net irrigation Values of crop production per units of water requirement). It indicates also the harvested land in the studied irrigation extent to which the water supplied was schemes are presented in Table 4 and figures adequate to satisfy the water demand. The 2-4. As can be seen from table 4, production value of this indicator is nearly unity in varies from 122 to 174 tons per hectare in Wonji irrigation scheme and range from sugar cane producing irrigation schemes and 1.46 to 2.05 incase of other schemes from 0.50 to 3.56 tons per hectare in cotton indicating that the amount of water supplied producing schemes. The productivity of at scheme level exceeded the estimated crop tobacco varied between 0.45 to 1.55 tons per water requirement. hectare. The average sugar cane production in Metahara, Wonji and Finchaa is Table 3: Values of water delivery respectively 162.3, 147.1 and 136.5 tons per performance indicators (2005/06) hectare. This shows that Metahara has Scheme Values of water supply produced more cane per units of area. name performance With a standard deviation of 13.7, the ARWS ARIS WDR productivity variation is higher in Finchaa Hare 1.22 2.05 1.07 followed by Metahara and Wonji sugar Sille 1.66 1.46 0.95 estates. Huge differences between minimum Bilate 1.86 2.00 1.30 and maximum productions in table 4 show Metahara 1.45 1.59 1.03 inconsistencies that exist in the management Wonji 1.11 0.95 0.62 practices as well as practically attainable level of productivity under the existed The water delivery ration (WDR) is an condition. Compared to sugar cane indicator that relates the amount of water producing schemes, large coefficients of delivered to the amount of water needed to variation (Cv) in cases of cotton and tobacco be delivered, i.e. total water supplied to the producing schemes have been observed scheme divided by gross irrigation water indicating high productivity variation from requirement. According to the values of this year to year. The reasons could be indicator, Wonji and Sille irrigation schemes inconsistencies in the agricultural practices, were found to have delivered less amount of management system and input supplies.

20 Table 4: Output per units of harvested land (from 1998/99 – 2005/06 except for Hare scheme) Scheme Crop grown productivity (tons/ha) Cv name minimum maximum mean SD (%) Metahara Sugar cane 152.6 173.8 162.3 9.3 5.6 Wonji Sugar cane 137.2 152.8 148.1 5.6 3.4 Finchaa Sugar cane 123.5 169.0 138.3 13.4 9.7 Average (sugar cane) 137.2 165.2 148.6 9.52 6.5 Hare1 Cotton 0.70 2.20 1.30 0.65 50 Sille Cotton 0.50 2.40 1.09 0.79 72 M. Sedi Cotton 1.57 3.56 2.51 0.67 26.7 Average (cotton) 0.92 2.72 1.67 0.65 43 Bilate Tobacco 0.47 1.55 0.90 0.33 36.9 1. Compared between different villages in the scheme season 2005/06 have been considered. This The productivity of cotton presented in this type of calculation was made taking into study, when compared to the optimum yield, account the farm gate unit price of sugar, i.e. 5.4 tons/ha (Aklilu, 2006), it is evident cotton and tobacco in the year 2005/06 as that there is a room for improvement. Melka 491.2, 1069.77 and 1962.79 US$ per tons Sedi was found to perform better than other respectively. The results presented in figures cotton producing farms. 2 and 6 show that outputs per units of land To compare the outputs of schemes and and water are by far large in sugar producing productivity of different crops per units of schemes than cotton and tobacco farming land and water supplied, monetary schemes. equivalents of the production during the

12000 10834 Output per harvested area 10000 9379 Output per cropped area

7794 ) 8000 6104

6000 5697 5314

Output (US$/ha Output 4000

2077 2000 1500

385 464

310 376 172 0 Sille Hare M.Sedi Bilate Finchaa Metahara Wonji Scheme name

Fig. 2: Output per units of harvested and irrigable area (2005/06)

The difference between schemes in terms of 2005/06, the proportions of harvested to output per irrigable area is less compared to total cultivated area in Wonji, Metahara and output per harvested area. In the year Finchaa are 49, 65 and 73% respectively.

21 Even if the cane production per harvested Finchaa. Sugar produced per hectare per area in 2005/06 was higher in Metahara month was also greater in Wonji followed (Fig. 4) than other cane producing schemes by Metahara and Finchaa which may be output per units of harvested land (US$/ha) attributed to the differences in the cutting was higher in case of Wonji than Metahara ages of the cane. (Fig. 2). This is because the end sugar productivity was higher in Wonji than Fig. 3 shows the relationships between the others, i.e. 21.92, 15.98 and 15.77 tons of size of area harvested and the corresponding sugar per hectare of harvested area cane production during the last 8 years respectively in Wonji, Metahara and (1998/99 – 2005/06).

1400

1200 Wonji

Metahara 1000 Finchaa tons)

3 800

600 Cane ( * 10 ( Cane 400

200

0 0 1000 2000 3000 4000 5000 6000 7000 8000 Area harvested (ha)

Fig. 3: Comparison of harvested area and cane production (1998/99 – 2005/06)

Annually harvested area and hence total area has increase from 9911.5 ha to cane production is greater in Metahara sugar 10145.9ha in Metahara. This is an increment estate followed by Finchaa and lowest in of about 2.4%. However, the size of Wonji. The regression coefficients (r2) of the harvested area was variable from year to relations given in figure 3 are 0.36, 0.004 year without showing linear increase. and 0.88 for Wonji, Metahara and Finchaa respectively. This shows that both harvested Figure 4 and 5 show the deviation of annual area and cane production was not production of sugar cane and cotton from significantly increased in Metahara and the overall average production of the Wonji. In case of these schemes, the points schemes involved in producing the same showing the relationships between areas crop during the last 9 years. While the harvested and cane production are productivity of Wonji is consistently close to concentrated at almost same area. Within the the average line, the productivity of period 1998/99 – 2005/06 the total cropped Metahara scheme is greater than the average

22 and that of Finchaa scheme is lower than the government managed, the management mean productivity of the schemes. setup and conditions of physical structures From the two cotton producing schemes, under which they are operating is different. Melka Sedi was found to consistently The more than 40 years old irrigation produce more than average production. On infrastructure in Sille farm and less the contrary the productivity of Sille scheme motivated and unskilled staff as well as low is below average in all 9 years considered input services are contributed to low except in one year, i.e. 2001/02 (Fig. 5). productivity of the farms. Although both belongs to the state farms,

10 Metahara 0 Wonji Finchaa -10

-20 Deviation from mean (tons/ha)

-30 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 Year

Fig. 4: Deviation of annual sugar cane production from mean, i.e. 149 tons/ha

2.0

1.5

1.0

0.5 Sille 0.0 Melka Sedi -0.5

-1.0

Deviation from mean (tons/ha) fromDeviation mean -1.5

-2.0 1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 Year

Fig. 5: Deviation of annual cotton production from mean, i.e. 1.97 tons/ha

23 Although adequate or more water than 4.3. Output per units of water supply required is supplied to the scheme, the output obtained in Sille and Hare schemes Water productivity has been defined as the are very low. It may not be the total amount amount of output produced per unit of water of water diverted to the scheme which is so involved in the production, or the value important to evaluate the influence of added to water in a given circumstance irrigation water on production rather (Molden et al. 1998). It was calculated by adequacy and uniformity of its distribution dividing the value of agricultural production on the cropped field. obtained from a unit area of land by volume of irrigation water supplied during the production season. Fig. 6 shows the productivity of water for different irrigation schemes.

0.7

) 0.6 3 0.55 0.5 0.48

0.4 0.29 0.3 0.24 0.2

Productivity (US$/m 0.1 0.05 0.01 0 Sille Hare M. Sedi Metahara Finchaa Wonji Scheme name

Fig. 6: Output per unit water supplied

Output (US$) per units of water supply (m3) gained is more in case of Wonji. This may varies between 0.01 in Sille, 0.05 in be due to the differences in the cutting ages community managed scheme and 0.55 US$ adopted by the schemes. Results given in in government managed irrigated sugar cane Fig. 6 are also influenced by the value of farm. Water productivity in other areas was crops grown, irrigation management and found to be 0.04 – 0.56 US$ (Merdun, weather conditions such as contribution of 2004), 0.03 – 0.91 US$ (Molden et al. rainfall. Proper irrigation scheduling that 1998). Differences have also observed not takes into account the contribution of only between different schemes, cropping rainfall during growing season will have and management types but also between the improving effect on water productivity. The same cropping and management type in difference in the water productivity between different schemes. The results indicate that sugar cane producing scheme is attributed to Wonji irrigation scheme was found to be the management practices. Despite adequate efficient in economical use of irrigation water deliveries at the scheme level (Table water followed by Finchaa and Metahara. In 3) in Sille and Hare irrigation schemes, both terms of cane productivity Metahara was land and water productivity is low compared found to produce more. However, sugar to Melka Sedi scheme, producing the same

24 crop, i.e. cotton. It is not the total amount of productivity that ranges from 123.5 - 173.8 water diverted to the scheme which so tons per hectare of harvested area, 7794 – important to influence the production, rather 10834 US$ per harvested area (2005/06) and its adequacy, uniformity and proper 0.24 – 0.55 US$/m3 of water supplied spreading over the cropped field. (2005/06). On the other hand schemes that grow cotton have relatively low productivity 5. Summary and Conclusion and high variations that ranges from 310 – 2077 US$/ha in community managed and The assessment of irrigation performance in state farm. Output per units of water seven irrigation schemes using output and supplied varied from 0.01 – 0.29 US$/m3 of water supply performance showed that there water supplied to the scheme. is a tremendous difference between the Cotton growing schemes are characterized schemes in their output performance. This is by high productivity variations between true even for the same cropping and seasons. This could be due to management types. Government agency inconsistencies in the management systems, managed schemes that grow sugar cane have input services and inability to minimize the got higher productivity that ranges 123.5 - influences of climate conditions through 173.8 tons/ha, 7794 – 10834 US$ per adoption of effective irrigation scheduling. harvested area and 0.24 – 0.55 US$/m3. On Huge variations between outputs of same the other hand schemes that grow cotton crop type in different schemes reveal that have relatively low productivity that ranges there is a room for improvement in the from 310 US$/ha in community managed productivity of land and water. However, scheme to 385 US$/ha in state farm. The answer to the question, ‘which one is doing water productivity of these schemes is what better and why?’ need the examination respectively 0.05 and 0.01 US$/m3. It is of internal process indicators. evident that as the management setup, staffing, capacity, and availabilities of Low productivity of irrigated agriculture in resources are different, not all schemes schemes such as Hare and Sille is possibly under similar management and cropping attributed to poor conditions of the irrigation types have similar performance. Then, there infrastructure, inadequate management is a huge difference in the attainment of the capacity and skills, lack of proper operation primary objective of irrigation, i.e., and on-farm water management practices increased outputs. and procedures, lack of incentives and hence The scheme level values of water supply low motivation to improve performance. performance indicators (ARWS and ARIS) Investment on improvements of physical revealed that there were no water supply structures, management and operation of the constraints during the season. That means system at all levels will bring substantial the water supplied during the season improvement of performances of these (2005/06) could meet the forecasted crop schemes. and irrigation water demand in all schemes considered. However, it should be noted that Scheme level values of water delivery and the scheme level values does not give any supply performance indicators presented in clue how efficiently, adequately, uniformly, this paper are based on data sets of one year. timely and reliably the water was distributed It doesn’t show also how adequately, within the farms. It is evident that measuring uniformly, efficiently and timely the water these indicators requires intensive field data distributed over the field and field units which need to be generated from field level throughout the season. Hence the scheme measurements. level performance indicators are of use for strategic thinking and don’t serve as such Government agency managed schemes that operational purpose, because they don’t grow sugar cane have got higher indicate exactly where the problems

25 responsible for low performance of the Kloezen, W.H. and C. Garce’s-Restrepo. system lie. The next study should focus on 1998. Assessing irrigation performance assessment of performance based on internal with comparative indicators: The case of processes indicators such as adequacy, the Alto Rio Lerma irrigation District, uniformity, reliability, efficiency and Mexico. IWMI Research Report No 22. sustainability. Colombo Sri Lanka.

References Mohtadullah K. (1993): performance of irrigation systems. 15th- congress on Aklilu Adagn. 2006. Assessment of irrigation and drainage. ICID . CIID, 74 irrigation system at Sille state farm. – 83. MSc-thesis. Arba Minch University, Ethiopia Molden, D., R. Sakthivadivel, C.J. Perry, C. de Fraiture and WimH. Kloezen (1998): Belete Bantero. 2006. Across-system Indicators for comparing performance of comparative assessment of medium- irrigated agricultural systems. IWMI scale irrigation system performance. Research Report No 20, Colombo, Sri MSc-thesis. Arba Minch University, Lanka Ethiopia. Small, L., M. Svenddsen. 1990: A Bos, M.G., M.A. Burton and D.J. Molden framework for assessing irrigation. (2005): Irrigation and Drainage Irrigation and Drainage system. 4, 283- Performance Assessment. Practical 312. Guidelines. CABI Publishing, CAB Tessema B. 2006. Assessment of hydraulic International, Wallingford, UK performances and irrigation potential of Bilate tobacco development irrigation Clemmens, A.J., D.J. Molden. 2007. Water scheme. MSc. Thesis. Arba Minch uses and productivity of irrigation University, Ethiopia. system. Irrigation systems. 25, 247-261 Vandersypen, K., Bengaly, K., Keita, A., Gorantiwar, S.D., and I.K. Smout. 2005. Sidibe, S., Raes D., Jamin, J.-Y. 2006. Performance assessment of Irrigation performance at tertiary level irrigationwater management of in the rice schemes of the Office du heterogeneous irrigation schemes: 1. A Niger (Mali): Adequate water delivery framework for evaluation. Irrigation and through over-supply. Agricultural water Drainage Systems. 19: 1-36. management, 83, 144-152.

Habib D. 2005. Study on hydraulic Yercan, M., M. Dorsan, M.A. Ul. 2004. performance of some irrigation canals at Comparative analysis of performance Wonji sugar state. Proceedings of 9th criteria in irrigation schemes: a case sysmposium on sustainable water study of Gediz river basin in Turkey. resources development. Arba Minch Agricultural Water management. 66, University. 52-57. 259 – 266.

Irrigation and Rain-fed Crop Production System in Ethiopia

Tilahun Hordofa1, Michael Menkir2, Sileshi Bekele2, Teklu Erkossa1 1 Ethiopian Institute of Agricultural Research, Addis Ababa, Ethiopia 2 International Water Management Institute (IWMI) Subregional Office for the Nile Basin and East Africa, Addis Ababa, Ethiopia [email protected]

Abstract minimum cultivated land by small-holder peasant farmers at the national level was Crop production is a function of water, 10.7 and 6.6 M ha, respectively. Increased nutrient, climate and soil environment. cultivable area by private small-holder Provided that all other requirement are farmers could not seem to contribute to the satisfactorily for proper growth and increased production. Total irrigated land by production, rainfall rarely meets the time private peasant farmers ranged between 66 with required amount of water application and 147 thousand hectares for the last one for plant growth. As a result average yield decade. During the last decade the area of agricultural crops under rain-fed under irrigation was steadily increasing for agriculture is low compared to irrigated most of the large scale schemes. Particularly agriculture. This study assesses irrigation Fincha’a and Metehara farms are and rain-fed agriculture system in significantly increasing while MAAE farm connection to its potential productivity has shown only a slight increase. At UAAIE under existing practice. While the rain-fed farm, crop production shows a decreasing areas considered in this study are the trend. Productivity of banana at Sille farm aggregate at the national level, five was decreasing despite its increasing in land systematically selected Medium and large area. scale irrigation schemes were selected based on cropping patter, geographic and agro Key words: Irrigation, Rain-fed, crop ecological representation These are production, productivity, large scale scheme, Fincha’a, MAAE, Metehara, Sille, and private small-holder UAAIE which are located at three river basins, viz, Nile (Abbay), Awash and the 1. Introduction Rift valley basins. Data were collected using pre-formulated checklists, through series of Crop production is a function of water, interviews and discussions; and from nutrient, climate and soil environment. published and unpublished documents. The Complex relationship existed between these result indicated that crop production was factors and the crop as a consequence of the undulating under rain-fed agriculture and as involvement of biological, physiological, a result the performance of rain-fed physical and chemical processes. However, productivity remained low and stable for efficient crop production and optimum yield most crops. Although crops grown by small- can be achieved only when the water supply holder private farmers are different, cereals is precisely in fine tune with the biological occupy about 74 per cent followed by pulses needs provided that the crop is well supplied and oil seeds with small proportion. During with the required nutrients and well adapted the last one decade, the maximum and to the prevailing environment. On the other

27 hand, there is no crop without water, what so The dependence on rain has significantly ever needs are in place. Therefore, timely affected the life of the people in particular supplied with an adequate amount of and economic development of the country in precipitation and/or irrigation could play a general. Improving this sector contributes to major role in increasing agricultural crop improve the productivity of agriculture and production. thus the generation of higher incomes, Rainfall rarely meets the time with required promotion food self-sufficiency and amount of application for plant growth. As a improving health condition of the people. It result average yield of agricultural crops also increases and diversifies production of under rain-fed agriculture is low compared raw materials for industries and promotion to irrigation, which is the application of of export item. controlled amount of water at specified time of application. In Ethiopia, traditional rain- Irrigation is one means for a good farm fed agriculture is the dominant form of husbandry, better land utilization and stable farming in which the peasant farm and higher crop production. Sustained households contribute the largest proportion growth and dynamism in agriculture is a of the total agricultural production. Out of fundamental necessity to meet the increasing the total land area of 112.3 M ha, about 16.4 demand for food and other products in view M ha are suitable for the production of of the growing population. Irrigated annual and perennial crops. Of the estimated agriculture will play a major role in reaching arable land, presently about 10 M ha is used the broader development vision of the annually for rain-fed crops (CSA, 2006). country in achieving food security, poverty alleviation and improvement in the quality The pattern and intensity of rainfall in the of life. The main objective of irrigated country is quite variable in which most of agriculture is to provide plants with the highlands receive between 510 to sufficient water to prevent stress that may 1530mm of rain annually and in typical arid cause reduced yields or poor quality of and semi-arid areas generally receive less harvest (Haise and Hagan, 1987; Tayler, than 500mm to about 750mm rain, 1965). respectively. However, rainfall in most cases is unreliable and erratic and moreover, Ethiopia is one of the few African countries productivity is constrained by several endowed with relatively abundant water interlinked factors such as unpredictable resources, favorable climate and potentially climate (flood, frost, pest etc.) small and huge irrigable land. The annual stream flow fragmented land holding, land degradation, and groundwater resources are estimated limited technological inputs, etc. About 80 around 122 and 2.6 billion m³, respectively. percent of the population lives in the About 83 percent of the total runoff is found highlands. Over population in these area in the basins of large rivers such as the caused shortage of land and thereby pushing Abbay (Blue Nile), the Baro Akobo, the the farmers onto lands with fragile soils and Omo Gibe and the Tekez (MoWR, 2002). steep slopes in which the land becomes exposed to erosion and eventually turns out While the potential benefits of irrigation are to unproductive state. On the other hand great, the actual achievement in many there is a huge tract of arable land in the low irrigated areas of the country is substantially lands which could be utilized for agricultural less than the potential. According to MoWR production. However, rainfall is either not (2005), Ethiopia is estimated to have 3.7 M sufficient or not dependable in amounts and ha of potentially irrigable area with the timing. As a result, crops suffer from severe available surface water resources and the soil moisture stress and drought. land irrigated through the development of traditional and modern irrigation schemes are estimated to be about 386,603 hectares,

28 which is about 10 per cent of potentially finance) is of paramount importance. This irrigable land. According to the report, in the presumes that the limited resources are modern irrigation there were 466 small, 102 efficiently used so that the benefits per unit medium and 9 large irrigation schemes with out puts are optimized. Therefore, this study the total area coverage of 28,939, 71,924 was aimed to assess irrigation and rain-fed and 49,675 ha, respectively have been agriculture system in connection to its developed by Government, potential productivity under existing nongovernmental organizations and private practice. investors. However, the major crops produced with irrigation are industrial and 2. Study Area cash crops with small proportion of food crops. The major part of food crops The study area encompasses rain-fed areas produced in the country come from rain-fed in the country in general and five agriculture. systematically selected Medium and large scale irrigation schemes based on cropping This gross underdevelopment has spurred patter, geographic and agro ecological the Irrigation Development Program (IDP) representation These are Fincha’a, MAAE, to put additional hectares of land under Metehara, Sille, and UAAIE which are irrigation within its 15-year plan period of located at three river basins, viz, Nile 2002-2016 (MoFED, 2006 ). Therefore, a (Abbay), Awash and the Rift valley basins. strategy that ensures economically The main features of the study sites are profitable, ecologically sustainable and presented in Table 1. socially acceptable use of the available resources (land, water, climate, labor,

Table 1 Main features of selected irrigation schemes Irrigation Schemes Features Fincha MAAI MSF UAAIE Sille SF River basin Abbay Awash Awash Awash Rift valley Latitude 9o 18’N 9o 46’N 8o 52’N 8o 37’N 6o 1’N Longitude 37o 14’E 40o 38’E 39o 54’E 39o 43’E 37o 37’E Altitude (m) 1450 750 950 1100-1200 1280 Tmax (0C) 24 34 33.1 32.6 32 Tmin (0C) 7.4 18 17.2 15.3 17 Rainfall (mm) 1300 500 550 500 729.6 Crops Sugarcane Cotton Sugar (fruits) vegetables cotton, and fruits maize and banana Source of water Fincha’a River Awash River Awash River Awash River Sille River Method of Sprinkler Surface Surface Surface Surface irrigation (furrow) (furrow) (furrow and (furrow) basin)

scale irrigation schemes. In addition, published and unpublished documents

(series of agricultural sample survey reports 3. Methodology (CSA) were reviewed and analyzed.

In this study status of rain-fed and irrigated 4. Results and Discussion agriculture in terms of area, production and productivity will be assessed at national and Rain-fed Farming System River basin level. Primary data (crop Ethiopian farming is mainly dependant on production and productivity, irrigated and rain-fed smallholder agriculture system as a rain-fed area etc) were collected using pre- means of food and income for its population. formulated checklists for time span of at Virtually all food crops come from rain-fed least ten years. Series of interviews and agriculture system. The farming societies are discussions were held with different principally private peasant. Crops grown stakeholders including; representatives of under rain-fed farming system were diverse water user associations and beneficiaries, and many within a cropping season. subject matter specialists from zonal and However, the agriculture system is district BoARD (agronomists, irrigation and dominated by cereal based productions since natural resource experts, extension it was produced in large quantity as personnel etc) for the small scale irrigation compared to other crops (Fig. 1). Pulses and schemes and the corresponding rain-fed oil seeds occupy small proportion in contrast agriculture system. Similarly interviews and to cereals. The major cereal crops gown discussions were also held with enterprise, were maize, sorghum, tef, wheat and barley. farm and unit managers of the selected large

Chat Hops 1% Fruits 0.2% 0.4% Sugar cane 0.2% Root crops 2% Coffee Oil seeds Vegetables 2.6% 7% 1%

Pulses 12%

5 074%

Fig. 1 Area coverage of rain-fed crops

30 Cereal production and productivity under due to an increase in cultivable area along rain-fed agriculture is shown in figure 2. The with the favorable conditions including crop production was undulating reaching adequate rainfall that might have occurred low during 1995/96 and 2005/06; and pick during the cropping season. Moreover, during 2001/02. As shown in Fig. 3, area visual observation elucidated that the under rain-fed agriculture was highest occurrence of sufficient rainfall (adequacy during 2005/06 cropping season and where and reliability) through-out the growing as production is low as compared to 200/01 season enables the private peasant farmers to cropping season. get bumper harvest, indicating rain-fed During the last one decade, the maximum agriculture to depend strongly on rainfall and minimum cultivated land by small- availability. In absence of sufficient rainfall, holder peasant farmers at the national level there is always low agricultural production was 10.7 and 6.6 M ha, respectively (fig. 3). thereby creating food shortage and food Agricultural production could be increased insecurity for its population and also dipping through increasing cultivable area or using poverty. Consequently, the performance of improved seed and cultural practices rain-fed agricultural production and (fertilizer, irrigation etc). The increased productivity remained low and stable for cultivable area during 2005/06 by private most of the years. The low level of small-holder farmers could not seem to performance of rain-fed agriculture could contribute to the increased production. The not only be attributed to erratic nature of production rather seems to fall as the area rainfall but also deteriorating soil fertility increased. There can be several reasons for and slow adoption and/or lack of appropriate low production. Simple reason for increased technologies. crop production during 2001/02 could be

3500.00 35.00

3000.00 30.00 2500.00 25.00 2000.00 20.00 1500.00 15.00 Production 1000.00 Productivity 10.00 500.00

0.00 5.00

-500.00 Tef Tef Tef Tef Tef Tef Tef Tef Tef Tef 0.00 maize wheat maize wheat maize wheat maize wheat maize wheat maize wheat maize wheat maize wheat maize wheat maize wheat Barley Barley Barley Barley Barley Barley Barley Barley Barley Barley sorghum sorghum sorghum sorghum sorghum sorghum sorghum sorghum sorghum sorghum 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01 2001/02 2003/04 2004/05 2005/06 Year

Production, '000 ton Productivity, q/ha

Fig. 2 Production and productivity of rain-fed agriculture

The above statements generally point out survived with uncertainty of rainfall and low that those small-holder private farmers level of technology aggravating the

31 occurrence of poverty and food insecurity. farmers use irrigation at small scale level to With the existing rate of production and enable them increase crop production and as productivity, it is going to be a challenge a means of raising income. Small scale and threat to the country to feed the ever irrigation not only increase crop production, rapidly growing population. Therefore, due but also improves cropping intensity and attention should be given to improve the reduces the effect of erratic rainfall. The production system to alleviate poverty and practice of irrigation may not be possible for secure food self-sufficiency. every farmer and could not be possible to expand the area. As shown in figure 3, during the last one decade, the total irrigated Irrigation Farming System land by private peasant farmers ranged between 63 and 175 thousand hectares Irrigation enables farmers to improve crop which is 0.8 to 1.8 per cent of the total area production and intensification thereby covered under rain-fed agriculture, sustaining and improving livelihoods and respectively. food security. In Ethiopia, private peasant

12000000 Rainfed Irrigated

10000000 10680886 10268921 10435370 9892611

8000000 9361560 Area (ha) Area 8210000 8016310

6000000 7948530 8072360 6649500

4000000 6 0 0 0 0 3 3 0 0 2000000 8114 8169 6317 6364 6821 8464 17573 146871 12064 12149

0 95/96 96/97 97/98 98/99 99/00 00/01 01/02 03/04 04/05 05/06 Year Fig. 3 Rain-fed and irrigated area under small-holders

The area under small scale irrigation has been appreciably expanding, particularly

after 2001/02 though apparently shown to decline during 2003/04.

Table 2. Irrigated area under small-holder private farmers (ha) Year Cereals Permanent Teff Barley Wheat Maiz Sorghum Total Pulses Oil seeds Others crops e 2005/06 7895 7441 6116 3558 15596 73509 3950 1740 30170 37502 6 2004/05 7756 7212 4614 2940 9756 59052 6832 1915 16143 34361 7 2003/04 7835 5647 3805 3257 11199 61406 5217 927 21578 31516 3 2001/02 1244 8016 4750 5383 12778 93615 9896 726 35855 33359 7 8 2000/01 5650 3680 1330 1896 15910 45770 3880 290 9240 22500 0 1999/00 5230 4630 1780 1856 11510 42850 2890 430 13390 21580 0 1998/99 2890 3580 2570 1365 7240 30110 2660 ** 8830 21350 0 1997/98 1154 4230 2540 1214 3690 34560 1920 ** 7350 18860 0 0 1996/97 4790 ** 800 1369 18970 39220 ** ** 3650 21470 0 1995/96 4490 1400 1890 2181 ** 55050 5300 ** 4890 1933 0 ** not reported The main agricultural produces under small scale irrigation followed by sorghum, tef, scale irrigation includes cereals, pulses, oil barley and wheat (Table 2). seeds and permanent crops like fruits, chat, coffee, hopes and sugarcane. Permanent In contrary to small scale irrigation, Medium crops occupy the largest share next to and large scale irrigation schemes in the cereals. Pulses and oil seeds occupy (5 %) country has been found to produce mainly insignificant part compared to permanent cash and industrial crops. The schemes crops. The major irrigated pulse crops considered this in study are government include field pea, horse bean, haricot bean owned enterprises. Medium and large scale and chick pea. Major irrigated oil seeds irrigation schemes are usually expected to include neug, linseeds and sesame. Others contribute to the national economic growth like vegetables and root crops are produced and alleviation of poverty and food security. in small proportion. Among cereals crops, However, the contribution of these schemes maize covers the maximum area under small to food crops production is almost none. The major crops grown under medium and large

33 scale irrigation include Sugarcane, cotton, cotton production shows a decreasing trend fruits and vegetables. Figures 4 to 6 show mainly the replacement of some of the area, production and productivity of cotton field by banana. This was attributed Medium and large scale irrigated agriculture mainly to market problems since there are schemes. During the last one decade the area no competent purchasers of the products in under irrigation was steadily increasing for the area, and transportation cost is too high most of medium and large scale schemes to move to other areas. Consequently, the (Fig.4). Particularly Fincha’a and Metehara farm has become price recipient, instead of farms are significantly increasing while negotiating for a reasonable price. MAAE farm has shown only a slight Productivity is increasing better at MAAE increase. Mainly due to re-occupation of farm only while at Metehara and sille farms part of the lands by the local people, Sille the trend is stable. Productivity has been farm has shown a decline in irrigated land declining at UAAIE and Fincha farms, but area. . Similarly, crop production, has shown attempts are made to compensate by an increasing trend for most of the irrigation increase in cultivated land area. While the schemes (Fig. 5). Sugar cane production is productivity of banana at Sille farm was increasing significantly at Fincha’a and decreasing despite its increasing in land Metehara while banana production is area, the overall productivity of Sille farm is increasing significantly at Sille farm. unstable, which may be among the reasons However at UAAIE farm, crop production for continuous change of crop types. shows a decreasing trend. At Sille farm

12.00

10.00

MAAE Sille 8.00 Fincha UAAIE Metehara

6.00 Area in '000 ha

4.00

2.00

0.00 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Year

Fig. 4 Trend of irrigated area under medium and large-scale irrigation schemes

34 9000

MAAE ('00 q) Sbanana (q) 8000 Scotton (q) Smaize (q) Fincha ('000 t) 7000 UAAIE ('00 q) Metehara ('000 t) 6000

5000

Production 4000

3000

2000

1000

0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Year

Fig. 5 Trend of agricultural production under medium and large-scale irrigation schemes

120.0

100.0 MAAE (q/ha) Sbanana (t/ha) SCotton (q/ha) 80.0 Smaize (q/ha) Fincha (00q/ha) UAAIE (t/ha) Metehara (000q/ha)

60.0 Productivity

40.0

20.0

0.0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Year

Fig. 6 Trend of productivity of irrigated agriculture under medium and large-scale schemes

35 5. Conclusion References

In Ethiopia, crop production is dominated by CSA (Central Statistical Agency) 2006. traditional rain-fed agriculture. However, Statistical Abstract. Federal Democratic rainfall is unreliable and erratic in nature. As Republic of Ethiopia Central Statistical a result, production and productivity is low Agency, Addis Ababa, Ethiopia. and stable. During the last one decade, the Haise, H.A. and Hagan, R.M. 1967. Soil, maximum and minimum cultivated land by Plant and evaporative measurement as small-holder peasant farmers at the national criteria for scheduling irrigation. Pp 577 level was 10.7 and 6.6 M ha, respectively. – 604. In: Irrigation of Agricultural Crops grown are mainly cereals occupying lands. R. M. Hagan, H.R. haise and T.W. about 74 per cent followed by pulses and oil Edminster (ed), Monog. Am. Soc. seeds in small proportion. The major cereal Agron. Medison, WI. crops gown were maize, sorghum, tef, wheat MoFED (Ministry of Finance and Economic and barley. However, rain-fed agriculture Development). 2006. A plan for does not seem to meet the ever rising Accelerated and sustained Development demand for food as population increased. to End poverty (PASDEP): Ethiopia, Hence, requires special focus on ways of building in progress. Volume 1 main increasing production and productivity. report. Federal Democratic Republic of Private peasant farmers use irrigation at Ethiopia, Ministry of finance and small scale level to enable them increase economic development, Addis Ababa crop production and as a means of raising MoWR (Ministry of Water Resources). income. Area under small scale irrigation has 2002. Water sector development been appreciably expanding, particularly programme 2002-2016, Volume II: Main after 2001/02. During the last one decade, Report. Federal Democratic Republic of the total irrigated land by private peasant Ethiopia, Ministry of Water Resources, farmers ranged between 63 and 175 thousand Addis Ababa. hectares which is 0.8 to 1.8 per cent of the MoWR (Ministry of Water Resources). total area covered under rain-fed agriculture, 2005. Irrigation (Potential) in Ethiopia. respectively. Cereals and permanent crops Irrigation and Drainage Development occupy the major part followed by Studies Department, Federal Democratic vegetables and root crops. Republic of Ethiopia, Ministry of Water The contribution of medium and large scale Resources, Addis Ababa. irrigation to food crops is almost negligible. Taylor, S. A. 1965. Managing irrigation The schemes produce mainly industrial, water on farm. Trans. ASAE; 8:433-436 exportable and cash crops of which sugarcane, cotton, fruits and vegetables are the major once. In most cases, the area under medium and large scale irrigation schemes was steadily increasing except at MAAE and Sille farm. At UAAIE production and productivity were declining while Sille farm production is generally declining and productivity was unstable. Crop productivity was declining at Fincha’a farm, better at MAAE farm and stable at Metehara.

Basline Servey of Irrigation and Rainfed Agriculture in Blue Nile Basin

Kebede Tsehayu FAO-Nile Project Entebbe, Uganda [email protected]

Abstract around 370 million people. The Nile basin countries are Burundi, the Democratic This paper discusses issues and challenges Republic of Congo, Egypt, Eritrea, of agricultural developments in the Blue Ethiopia, Kenya, Rwanda, Sudan, Tanzania Nile Basin of Ethiopia. The crop land in the and Uganda. basin can not sustain the population unless agricultural productivity increases. Due to The FAO-ITALY supported project climatic factors and low yield rainfed “Information Products for Nile Basin Water agriculture can not support the high Resources Management” is implementing a population in the basin. Up to date the land basin wide survey of current and future water under irrigation is very small. Like most use in rainfed and irrigated agriculture. The Nile Basin countries though agriculture Project “Information Products for Nile Basin dominates the economy of Ethiopia much Water Resources Management” is intended to was not done in the irrigation agricultural strengthen the ability of the governments of the resources development. ten Nile countries to take informed decisions with regard to water resources policy and

management in the Nile basin. The project is Irrigated agriculture is the largest draw on supported by the Government of Italy and the waters of the Nile in Egypt and Sudan. carried out under the umbrella of the Nile Basin But the others 8 Nile countries agriculture Initiative, of which Italy is a full partner. The is mainly rainfed and they are not using project will make extensive use of regional even 2% of Nile water. Ethiopia is expertise. It will also draw on the specific contributing more than 85% of the Nile experience and knowledge residing at FAO, water annual discharge and yet is not using especially in the fields of agricultural water use even 1 % of it. In near future the water and water productivity. scarcity in agricultural development of the 2. Objectives Nile Basin can be affected by ever increasing population, unpredictable The survey is intended to provide stakeholders climate, soil infertility, uncertainty of and decision makers in the Nile Basin with a surface water allocations, unexplored thorough assessment of the linkage between groundwater resources, low water agriculture and water in the basin. The survey availability, infrastructure etc. Agriculture is expected to address the following issues: is by far the main user of water in the Nile • What are the opportunities for enhancing basin and therefore requires due attention in productivity in agriculture? future investments. Ethiopia does not achieve food security until it utilizes Nile • Create Agricultural production database at district level providing information on area, water for irrigation. yield, and production for the main commodities for a baseline year from 1. Background (2000);

The Nile river basin covers an area of • Analysis of the current agricultural approximately 3.1 million square productivity under rainfed and irrigated conditions; kilometers and with a total population of

• Rapid export growth through production of production is insignificantly low in these areas. high value agricultural products; different agro ecological considerations might lead to the improvement of agricultural • Decentralization to shift decision making production. the following map shows the main closer to community to improve agro ecological zones of the basin. one can see responsiveness and service delivery; clearly the topography in the east is not easy for • Increase proper and modern water farming (figure 1). regional specialization resources utilization. based on suitability and provisions of compatible packages for the different agro ecologies that will lead to the improvement of 3. Rainfed Agriculture small-scale agricultural production are the basic development directions. • A baseline survey (for the year 2000 and onwards) of agricultural water use covering all aspects of agriculture is in progress. The survey includes a detailed review and maps of population, land use, cropping patterns, estimate of water use in rainfed and irrigated agriculture, and an assessment of current water development facilities.

• Projections for 2030 of demand for food and other produce in the basin, with estimates of arable land expansion in irrigated and rainfed agriculture, yields of major crops, cropping patterns, and cropping intensities will be executed. Results of a set of scenarios on water development for agriculture will be based on analysis by administrative units.

• The study will focus at national and sub- national levels with particular attention to the main crop production systems. Agricultural water productivity will be analyzed at district level. Results will be aggregated to country level with the aim to build up a comparable picture of water Figure 1. Agroecological Zone of the Blue productivity in rainfed and irrigated Nile Basin agriculture across the basin. Production and productivity are very low in rain fed agriculture is the major livelihood of the basin and the use of improved the country, which is characterized by low productivity enhancing inputs and productivity associated with underdeveloped, technologies is extremely low. The low input systems and highly degraded natural expansion of irrigated agriculture is among resources. peasant agricultural lands on the other factors is constrained by the rugged plateaus of the northern and south-central part of ethiopia have been degraded because of topography and terrain. Most of the regions over-utilization for centuries. one cannot hope have poor market access due to that ethiopia can maintain her food production underdevelopment of infrastructure as through the traditional farming method of whole and agricultural markets in highland agriculture. absolute dependence on particular. Natural resources degradation rainfed agriculture and low productivity (soil erosion and deforestation) are associated with underdeveloped, low input extremely sever in the basin, which makes systems of production and degraded natural it a necessity for all development resources base has rendered these areas highly interventions to be oriented towards vulnerable to even minor shocks. irrigated crop 38 sustainable natural resources management control, data were eventually selected as per the as a matter of priority. principle “best data source so far available around the reference year 2000”. All data are 3.1 Crop yield distribution of Blue Nile presented as they are and occasionally represent Basin different time periods. It is to expresses types and distribution of cropping systems. Map of major crops (teff, wheat, maize etc). Crop data from rain fed and irrigated agriculture was collected from Nile Countries through national consultants. Major crops were provided in terms of crop production, acreage and yield. The only district based agricultural survey was carried out in the year 2001/02 (1994 Ethiopian Calendar). However the survey data from 1994 (1987 Ethiopian calendar) provides also useful information on area, production and productivity even if this survey was zone based.

The crop production data is for both seasons ("Meher" and "Belg") for all holdings. The "Meher" is the major season while the "Belg" is the small season. Crops harvested between the months of September and February are considered "Meher" crops while those harvested between the months of March and August are considered "Belg". Private peasant Figure 2. Wheat distribution of the Blue holding is predominant while large-scale Nile Basin commercial production is negligibly very little. In this rainfed production dominated small scale farming system, the share of irrigated agriculture is also very little. The distribution of some major crops yield was displayed on (Figure 2,3,4). Countries like Egypt, Kenya have higher yield than Ethiopia. Especially Egypt which has used all its crop land for irrigation agriculture attains the highest yield in the Nile Basin. The yield for Blue Nile of Ethiopia is quite low in the range of 0.3 ton/hectare to 1.5 ton/hectare for major cereal crops.

The following maps present the spatial distribution of the dominant crops found in the Basin. It aims to provide insight in the cropping distribution of the farming systems, in particular in relation to the natural conditions determined by topography and climate. The data presented are the result of an intensive database compilation on reported cropping statistics. As such they combine findings from multiple data sources. After a process of quality 39

Figure 3. Teff Distribution of the Blue Nile Blue Nile of Ethiopia covers an area of Basin about 367,000km2 and a population of 31 millions according to 2005 UNPD figure. The Nile Basin in Ethiopia stretches over a very wide geographic area with diverse agro ecological conditions. The Nile basin comprises of the following contributing basins: The Abay river basin The Tekeze river basin The Baro Akobo river basin The Mereb river basins (partially)

The Nile Basin of Ethiopia known as Blue Nile contributes more than 85% of the annual discharge of the Nile being a big contributor, Ethiopia to date use only 1% of it.

According to the Master Plan Studies of Ministry of Water Resources, Blue Nile Basin has 165 Potential irrigation sites with total area of 2,126,700ha. The existing irrigation schemes have been collected as point data and it is difficult to calculate the area. Figure 4. Maize Distribution of the Blue Nile Basin

4. Potential and Existing Irrigation 4.1 Blue Nile and GIAM Schemes

According to Global Irrigated Area Map The fundamental goal of achieving (GIAM) most irrigation lands are situated sustainability irrigation Agricultural system in Egypt and Sudan. The rest upper Basin is complex and initiates many new countries have minor land under irrigation. questions that require further studies, which Ethiopia Blue Nile land under irrigation is in turn requires multi-disciplinary so small compared to its potential area. collaboration and funding. The According to the Master Plan Studies of investigation and further study programs Ministry of Water Resources, Existing would be essential for development of irrigation is 57,561 ha and the potential is improved methods and technology for 1,774,676 ha. Ethiopia in order to feed her improving agricultural productivity, and to fast growing population the irrigation alleviate future scarcity of water. The potential of the country has to be assessment of methodologies for developed. sustainable development is also quiet essential. In addition, the goal requires continuing research and gathering field data to assess management approaches. So this project output supports and benefit the Nile Countries to some extent.

Figure 5. Potential and Existing Irrigation Schemes of the Blue Nile Basin

Table 1. Existing and potential irrigation Summary in the basin Sub-basin Administrative Units Area Mean Water Existing Potential (km2) Resources Irrigation Irrigation (*106 m3/yr) (ha) (ha) Blue Nile Amhara, Benishangul 199,812 49,000 37,347 719,088 (Abbay) Gumuz, Oromia Baro SNNPR, Oromia, 75,912 23,237 12,315 486,299 Akobo Gambella, Benishangul Gumuz Tekeze/At Tigray, Amhara 86,510 8,191 7,899 569,289 bara Total 362,234 80,428 57,561 1,774,676 topography. But there are some pocket areas which have potential for irrigation. The irrigation potential of Ethiopia is estimated between 1.8 (Blue Nile Basin) and 3.7 (for the whole country) million hectares out of which only 5% is under irrigation (Irrigation in Africa by Food & Agriculture of the UN programme latter called AQUASTAT). As it is displayed in (Figure 5) Evapotranspiration map in general the west marginal area which is in

reddish brown color though it is flat and convenient for irrigation, it has high evaporation. The eastern and central part of the basin with bluish color is characterized by low evapotranspiration and rugged

income exports. A stronger performing agricultural sector is fundamental for Africa's overall economic growth. A constantly growing agricultural sector is crucial for addressing hunger, poverty and inequality. More than 70 percent of the total population and the majority of the extreme poor and undernourished live in rural areas. A healthy agriculture sector means more jobs, more income and more food for the poor.

Nile Countries mainly depend on export of primary agricultural products. The gross domestic product (GDP) of Nile Countries is very low as compared to other developing countries. Among the Nile Countries Egypt has a GDP of greater than 80 billion USD while the rest have a GDP of less than 20 billion USD according to African Development Report (Graph 1). Figure 6. Evapotranspiration Map of the When we see the composition of the GDP Blue Nile Basin for countries like Democratic republic of Congo, Brundi, Tanzania, Ethiopia and 5. Agricultural Trade of Nile Basin Rwanda agriculture contributed above 40% countries of the GDP (Graph 2).

Agriculture dominates the economies of most African countries, providing jobs,

Gross Domestic Product (GDP), 2003

100 84.4 80 US $ billion 60

40 17.8 14.4 10.3 20 5.7 6.7 6.3 0.6 0.8 1.7 0 i a a d pt d ni n run DRC gy za u E thiopia Kenya Sudan n B Eriteria Uganda E Rwa Ta

Graph 1 Gross Domestic Product (GDP)

Composition of GDP, 2003

100% 24 Services 32 36.5 38.6 Share 80% 49.2 47.4 39.7 46.4 17 61.4 64.7 Industrial 60% 19 21.9 16.4 Share 10.7 24.8 21.2 40% 34.6 Agricultural 57 49 24.7 41.8 19.6 45 Share 20% 41.6 35.5 32.4 16.1 13.9 15.8 0% a a n a ndi C pt d d u R ny D e an Egy thiopia K w Suda gan Bur Eriteria U E R Tanzania

Graph 2 Composition of GDP

For Egypt, Eritrea, DRC and Rwanda the import of agricultural commodities is higher than export of agricultural commodities. For Countries like Ethiopia and Sudan the difference between export and import of agricultural commodities is not so much even though import is a little bit high. For Kenya, Uganda, Tanzania and Brundi their export of agricultural commodities is higher than importing (Figure 7).

Graph 3. African Countries are highly dependant on single agricultural commodity

Figure 7. Agricultural trades in the Nile countries

At present, a characteristic of too many Over the past 30 years, agricultural imports African countries is a relatively have outstripped agricultural exports, undiversified economy with little industry making the region a net agricultural and manufacturing and exports dominated importer since 1980, Indeed, as population by one or two raw commodities (Graph 3). growth in Africa outpaces food production, Often a single, primary agricultural imports and food aid are required to make commodity is the major source of export up the difference In the mid-1990s, out of earnings, creating a source of uncertainty the world total of 32 million victims of because of their low income elasticity of disasters receiving relief assistance from demand and their declining and volatile the World Food Programme (WFP), 21.5 terms of trade. Overall, Africa's agricultural million were living in Africa. In 2000 sector accounts for about 20 percent of total Africa received 2.8 million tones of food merchandise exports, declining from more aid, which is more than a quarter of the than 50 percent in the 1960s world total. In 2001, the number of people suffering from food emergencies ranged between 23 and 28 million.

0.21% 2.03% Ethiopia export Value 0.85% 8.93%

Africa 43.14% Americas Asia Australia Europe Nile country 44.40% Oceanic

044% Graph 4. Ethiopian agricultural import is mainly from America and Europe. The import from Africa and within Nile Countries is insignificant

Ethiopia_ Import Value 0.00% 1.42% 1.10% Africa Americas 34.41% Asia Australia Europe Nile country Oceanic 0.51% 57.08% 5.47%

Graph 5. Ethiopian agricultural export is mainly to America and Europe. The export to Africa and within Nile Countries is insignificant

Increasing trade and market opportunities 7. Major sources of information and locally, regionally and internationally data for the survey contribute to agriculture's ability to grow expand incomes and reduce poverty and 7.1 Irrigation (potential) in Ethiopia both food insecurity. Trade often introduces existing and possible schemes, new, more productive and more FDRE, Ministry of water resources, sustainable production technologies, September 2005, Addis Ababa processing systems and related services. Trade provides opportunities to produce 7.2 Tana Beles Integrated rural higher value products. For many development project rehabilitation producers, expanding agricultural trade and restructuring study, December opportunities locally and within the 1993 region is an important first step for taking advantage of potential new access 7.3 Tekeze basin integrated development international markets. However, master plan project second phase according to a recent paper, 1 trade report, vol. No. 4, June 1997 continues to be marked by overwhelming dependence on traditional overseas 7.4 Ethiopian Agricultural sample markets in industrial countries, although enumeration 2001/2002 (1994 E.c), (admittedly from a low base), there has Federal democratic republic of been substantial growth in intra-regional Ethiopia, central, Agricultural Census trade within Africa. If we exclude commission (Amhara, Tigray, unrecorded, often informal-sector, trade Oromia & Benishangul gumuz across the porous borders, on average, regions) only 10 percent of exports of countries belonging to each African Regional 7.5 Abby River basin master plan study. Economic Organization are destined within itself. 7.6 Tekeze river basin master plan study.

6. Recommendations 7.7 Baro Akobo river basin master plan study Unless the level of food production increases it will be hard to sustain the 7.8 National statistical survey 1994/95 ever increasing population (1987 Eth. Cal.)(Estimates of area, production and yield) • Increase export growth through production of high value agricultural 7.9 Agro ecological zones Ministry of products to get foreign earnings Agriculture and natural resources ----

Rather than depending on single 7.10 Woody biomass inventories and commodity it is advantages to diversify strategic planning project (WBISPP) raw commodities. The potential area in the Blue Nile Basin need to be developed before water shortage becomes an issue.

Community based irrigation projects should be encouraged by the government

Assessment of Design Practices and Performance of Small Scale Irrigation Structures in South Region

Robel Lambisso World Vision Ethiopia [email protected]

Abstract planning, institutional & operation problems are lack of adequate community Uneven distribution of rainfall in the country in general and in South region in particular participation, water right conflict among up makes irrigation the best way to enhance streamers and down streamers (12%), food production. Development of small- market outlet problems (8%), proper scale irrigation schemes is the best handing over problem (31%), lack of proper alternative as they require minimum training and the like. During the time of the investment & their gestation period is study, 18 % the irrigation schemes of the comparatively low. The southern region as region were not operational because of part of the country has been implementing aforementioned and related reasons. such schemes. Despite remarkable Key Words: Small scale irrigation; achievements, some of the implemented Performance of Structures; Design Practices schemes have totally failed and some are performing below their capacity. A case study considering 26 existing small scale 1. Introduction rd irrigation works (about 1/3 of the total The majority of population of Ethiopia is schemes) in the south region is carried out dependent on rain fed agricultural for this research work and attempt is made production for its livelihood. However, to understand the causes of the major estimated crop production is not close to problems that are related to the design fulfill the food requirements of the country. consideration of the different components of One of the best alternatives to consider for the structure and identify the gap in reliable and sustainable food security knowledge between the current design development is expanding irrigation practices and performance of the structures. development on various scales (whether The pre and post construction institutional small, medium or large) and options aspects, planning aspects, social aspects, (diversion, storage, gravity, pumped, etc). economic aspects, operation & utilization For countries like Ethiopia where the aspects are also given due attention. Some of principal component of project development the physically observed problems of the (finance) is a constraint to incur huge existing irrigation structures considered for investment for irrigation, small scale analysis are: main canal siltation (50%), irrigation can be an alternative solution to sedimentation of the headwork (42%), enhance food production. This is of course problem of seepage through foundation without undermining the strategic (4%), main canal seepage (33%), scouring importance of developing medium to large of downstream bank (8%), drying of rivers scale irrigation schemes to feed the (12%), damage on impervious and flexible expanding population in the foreseeable apron(19%), change of river course(12%), future. Development of small scale irrigation damage of under sluices(27%) and damage through river diversion, constructing micro on CD works(4%). Likewise, some of the dams, water harvesting structures, etc may

be considered as pragmatic approach in the ♦ To conduct inventory on the success and contemporary Ethiopia for ensuring food failure cases of small scale irrigation self-sufficiency. schemes in the region. ♦ To categorize irrigation structures based Small scale irrigation structures, owing to on different problem parameters their relatively small investment cost, ease ♦ Investigate the causes of failure for of construction, simplicity of operation & selected structures in the region with maintenance have been a strategic target of regard to hydrologic, hydraulic design, the country for achieving sustainable food structural design & implementation security and self sufficiency. A number of aspect. such schemes have been designed and ♦ Investigate the existing schemes with constructed in the previous years. However, reference to planning, institutional and while some schemes are performing operational problems successfully, it has been observed in various ♦ Formulate a systematic database on reports that some of the schemes have failed attributes of irrigation structures in the to serve the purpose for which they are region using dBASE format and Arc intended. In line with this, recent study View GIS software. The database is report for the Amhara region (Asfaw, 2004) basically important for efficient follow has been used as a bench mark to conduct up, evaluation of projects and to assess related study in the southern region. what has been done in the subject so far. To this end, this research aims to evaluate ♦ Investigate selected successful structures the design practices and performances of in the region and draw lessons from small scale irrigation structures in Southern their success (i.e. whether any unique Nations Nationalities and Peoples Regional design practice had been adopted or not) Government. In addition to the hardware ♦ Investigate failed structures and draw problems, institutional, planning, social and lessons to devise better design practices economic problems contributing to the and suggest methods of rehabilitation. failure are also highlighted.

The out puts of the research are believed to highlight the problems of the existing 3. Methodology of data collection and irrigation schemes and show future direction for planning, design and operation of analysis irrigation projects. In this research, desk study is made on 2. Objectives existing small scale irrigation systems and practical field visits were conducted to 2.1 General objective 15(Fifteen) sites found in the region. In the desk study, the available relevant data on The primary objective is to investigate the 11(eleven) existing irrigation structures and cause of failure of existing small scale post implementation review reports on the irrigation schemes of the region to learn a status of existing schemes in the south lesson and generate knowledge on practices region is collected from the regional and performances for future practical irrigation authority. From the study and application and compile set of design reports, the current design practice recommendations for planning, design, for irrigation structures is examined and implementation and operation of irrigation from the practical visits undertaken on 15 schemes. sites, the extent and frequency of hardware and software problems on the schemes is 2.1.1 Specific objectives assessed. In the analysis, the frequency of each problem among the sites, type of problem in each site and problem ranking

has been done. Better insights on hardware of the region geo-referenced and digitized and software aspects of the problems have for the same purpose. been acquired via interviews with Some canals like the one shown in the figure community and technical personnel using do not have proper design. In this case, the relevant structured questionnaire. canal does not have a sustaining bed slope To help build the geographic database, and it also does not have side berms. inventory is made, missing coordinates were recorded using GPS instrument, inserted in spreadsheets as dBASE (dbf-IV) and imported to the drainage and rainfall maps Fig-1 Silted main canal of Satame irrigation project 4. Results and Discussion Sustaining bed slopes should be provided by

conducting proper profile leveling activities. 4.1 Main Canal Siltation The soil bank immediately beside the canal is frequently washed by the rain water and Out of the 26 sites considered for the deposited in to the canal. Hence, deposition analysis, 13 are observed to have main of large soil mass should not be allowed canals highly charged with sediment. In beside the canals and adequate berms should other words, 50% of the sites are seen to be provided. suffer from the problem. The problem As additional solutions to the problem, prevalence indicates the level of attention to proper procedures of design of de-silting be given. The silt load is observed to come basins are also forwarded along with either along with the river water (suspended FORTRAN program written for the same and bed load) or as a run off from upstream purpose. Vortex vanes and excluder tunnels nearby catchement. are also recommended in the context of The figure below shows a lined canal experimentation and further research. completely filled with sediment.

4.2 Headwork Sedimentation

This was observed at 11 of the 26 sites. Accordingly, some 42% of the sites are affected with this problem. Headwork sedimentation refers to the overall submergence of the weir proper, wing walls and appurtenant structures such as gates due to settlement of sediment and the bed level rise of the river channel. Since the phenomenon results in the course change of rivers, in some cases like Hao diversion it is seen that the main canal is completely washed away by the river water changing its course. To minimize the problem with headwork sedimentation especially with high flow conditions and movement of sediment laden water, provision of simple intakes of gabion or rock fill may provide a solution (Novak et al, 2001). Similar recommendation has been forwarded in the study at Amhara region

(Asfaw, 2004). Despite the failure, the of side intake with simple masonry Lenda project of Bilate river is an example obstruction across the river. If the stream is characterized by bends, locating the intake It is recommended to practice use of clay at the external bend will give better lining and plastic lining to reduce the performance (Arved et. al, 1993). This is seepage of irrigation water at the canal. mainly owing to the reason that the inner However, this may not always be successful sides of bends are liable to deposition of as it has been tried and failed in Wamole sediment. The figure below illustrates the irrigation project of Sidama Zone. case: The issue of canal seepage can be managed by giving due emphasis to the properties of the soil along the canal route & command area. In line with this, proper techniques of clay lining and plastic lining should be tried to reduce seepage problems in irrigation canals. The coefficient of permeability is the most important factor that should be taken in to consideration. According to the USBR recommendation, canal lining should be carried out for soils with permeability coefficient value greater than 0.8m/day Fig-2 Failed Hao diversion weir (EVDSA, 1990).

4.3 Main Canal Seepage 4.4 Seepage through Foundation

The problem of main canal seepage is The foundation seepage is also observed in observed on 9 of the 26 schemes considered the region as being one of notable problems. for the purpose of the analysis. This means The analysis of water seepage under the that the problem is prevalent on 35% of the foundation is very important and must be schemes. given due attention. Failure to apply the The seeping water is seen to ooze through proper subsurface flow theories and the underneath of the soil and hence practices may result in complete failure of significant quantity of irrigation water is lost the structure (Garg, 1989). During the prior to arriving to the distributing survey, 1 of the 26 sites (4%) is seen to fail watercourses. Besides the loss of valuable due to this problem. At the observed site, the diverted water, the seepage moisture in the water is totally flowing beneath the vicinity houses is also a serious problem in foundation of the structure resulting in the the area. According to the key informants of total failure of the scheme. As physically Gidabo and Gelana irrigation schemes, the observed, the structure does not even have schemes have never been meaningfully used up stream and downstream cut offs to for irrigation since their construction 10 control the sub surface flow which induce years back. This is mainly because of the failure due to uplift pressure or failure due to reason that community members living piping or undermining. The figure below along side of the main canals do not allow shows the situation: diversion of water due to the excessive seepage flow in to their living houses.

Fig-3 Failed Hazembara diversion weir

4.5 Damage on Intake gate and Sluice (8%) are observed to have damaged wings gates and eroded banks. The following figure taken from Lenda Out of the 26 sites considered for the irrigation project constructed across the analysis, 8 are observed to have highly Bilate river may depict the phenomena. damaged and broken intake gates. In other words, around 31% of the sites are seen to suffer from the problem. Similarly, 7 of the 26 sites (27%) have damaged sluice gates. The figures possibly indicate how the schemes are performing under difficult condition. The cause for the problem of gates can be mainly attributed to improper scheme operation. Provision of adequate training, proper handing over and follow up of the users is indispensable. In addition, users organized in water user association should take initiatives to generate some resource to carry out some maintenance works of this sort. Fig-4 Failed Side intake constructed on Bilate River 4.6 Damage on Headwork The stability of the weir proper can be Damage of the weir proper is observed in 1 ensured by correctly following the proper of the 26 sites that is Lenda (4%) and 2 sites procedures of the structural analysis after

identifying the various forces acting on the structure.

In some cases like Lenda, it was observed portion of the structure may end up in the that irrigation structures with in earth quake total collapse of the structures. prone areas are designed with out Accordingly, proper design of both considering the earth quake forces. The impervious and pervious aprons are required force due to earthquake should be to control the excessive upstream energy and considered where necessary for the stability control structures should be provided to analysis. manage the problem of receding jump. The The most important forces to consider are: figure below shows the phenomena: Forces due to surface flows, forces due to sub surface flows, forces due to self weight and external forces like earth quake and silt pressure. The downstream wing walls can be failed due to excessive downstream bed and bank erosion which results because of excessive energy carried by the water coming from upstream. In addition, as seen at Lenda project, outflanking of protection structures can also cause failure of wing walls. Hence proper estimation of the magnitude of flood, proper geological investigation and proper Fig-5 Failed Goche diversion weir protection measures are vital in this case. 4.8 Drying of River Flows 4.7 Damage on Downstream Bed Drying out of flow was observed at the 3 of This problem is observed on 5 of the 26 the 26 sites considered for the analysis. In (19%) of the surveyed sites and it is other words, about 12% of the schemes were attributed to improper hydraulic design that seen having no water in their river channels. arises from poor knowledge of the energy Some of the sites such as Balle have got dissipation and impact of sediment on the seasonal rivers. One of the reasons causing structure (Baban, 1995 & Novak et. al, drying out of flows is improper estimation 2001). The impervious floor is designed in of lean flows. The current design practice all cases to reduce the surface flow action follows the procedure of float method to that causes scouring due to unbalanced estimate the lean flow. It may not be wrong pressure in the hydraulic jump trough. to use the method for preliminary works. Generally speaking, except very few sites, However, the method is approximate enough end sills are not seen at the constructed to over estimate the actual flow in the river structures. These could have played and is not reliable for important works with significant role in controlling receding considerable investment. In case of over jumps and hence reducing erosive power of estimation, the actual flow may not be the flowing water (Chow, 1959). In the obtained to satisfy the CWR and diversion schemes where this problem downstream water demands which may prevails, abrasion and scouring of the result in complete diversion of water in to impervious floor is commonly observed. In irrigation canals letting the downstream dry. addition, flexible aprons are seen to be completely washed away by the energy of The caseof Ufute may be taken as a good flowing water. The prolonged occurrence of example for this situation. During the abrasion and scouring of downstream survey, people living downstream of the scheme were seen complaining because of

shortage of water for their livestock. However, interviews revealed that some other sites like Jelaka became dry due to significant water use far upstream from the irrigation scheme. This situation has created some level of conflict among the far up streamers & irrigators and the phenomena showed lack of adequate social work during feasibility study and follow up. The river drying problem can be observed in the following figure:

Fig-7 Scouring problem on an aqueduct at Lenda Project

The features of design of cross drainage works can be summarized in the following main categories (Arora, 1996):

A. Hydraulic Design This usually involves the following: ♦ Determination of the maximum flood discharge and high flood level Fig-6 Jelaka Diversion Weir ♦ Fixation of water way of the drain ♦ Determination of canal water ways 4.9 Damage on CD Works ♦ Determination of uplift pressure on the floor of drain CD- works are structures carrying ♦ Design of bank connections discharges of a natural stream across a canal intercepting the stream. When a B. Structural Design canal is to be taken to the watershed, it This usually involves the following: crosses a number of natural streams in the ♦ Design of the cross section of the distance between the headwork and aqueduct trough command area. As one of CD structures, ♦ Design of piers and abutments an aqueduct is observed at Lenda project ♦ Design of foundations site. However, design was not prepared for the structure and was just constructed At Lenda, the calculation of the adequate based on experience of masons. Hence water way is not carried out to allow safe serious scouring is observed both at up passage of the drainage water beneath the stream and downstream bed and bank of main canal. This could be carried out using the structure. The following figure depicts the formula forwarded by Lacey. For large the case: drains, the wetted perimeter may be taken

equal to the width of the river; however, for small drains like that of Lenda a contraction of up to 20% can be allowed (Arora, 1996). However, failure to carryout the calculation of the waterway may endanger the entire structure and in extreme case may even flood the nearby areas.

In the case of Lenda, determination of implementation process by providing labour waterway, protection works of bed and and local materials and have already bank, design of foundation and developed sense of ownership. consideration of scour conditions are totally On the other hand, the situation at Goche ignored resulting in serious erosion scheme of Hadiya zone is the opposite of problems as indicated in figure-7 above. what has been discussed. Practically, the Accordingly, it is recommended to do detail users had no involvement with the hydraulic and structural design of cross implementation of the project and a sort of drainage works rather than simply deploying induced development or top down masons to build the structure. development is observed there. The intended beneficiaries did not show any interest with 5. Planning, Institutional, Social & the scheme mainly because it is not Economic Problems addressing their real problem. They were requesting for development of a nursery at The above problems with such schemes are the place where the major portion of the not only attributed to problems of design command area of the scheme is found. and construction. The software aspects of Accordingly, the scheme has now failed to planning, institutional social & operational fulfill the purpose to which it is intended and and economic problems are also crucially simply the structure is located there. important. In the following section the Similarly, less participation of intended highlights of each problem are presented: beneficiaries is also observed at Ameka irrigation scheme in Hadiya zone. The 5.1 Planning Problems beneficiaries could not participate adequately in all the phases of planning, implementation and operation mainly for The planning process in the development of reasons of location of the command area far irrigation projects can be viewed in the light away from their village and unsuitability of of community willingness and participation. the command area soil. The location of the Accordingly, good performance of the command area at distant place and high schemes is directly related to the level of plasticity of the command area soil involvement of community members in the developed reluctance among intended users planning process. of the scheme. Hence, although the project In line with this, the schemes in the region is completed, no one in the area is interested can be categorized in to two: to use it. The implementing government ♦ Schemes implemented with due agency should have repeatedly consulted the involvement of stakeholders local people starting from the project ♦ Schemes implemented with out(less) inception and site selection rather than participation of stakeholders implementing the scheme with out their involvement. In addition to this, suitability As observed during field visits, strong of the command area soil should have been community participation was involved at tested in laboratory during the feasibility Ufute and Doje schemes that are study. implemented by WVE in Kembata zone. 5.2 Institutional Problems The schemes are functioning satisfactorily and users are also happy with them. In the implementation of irrigation schemes, Community interview revealed that the various institutions are involved in the beneficiaries clean silt from main canals process of planning, design, implementation with out the orders given by leaders of user and operation & evaluation. association. From the very beginning, the However, in some of the schemes built by beneficiaries adequately involved in the NGOs and GOs the expected level of participation of various institutions is not

observed. Schemes of Satame, Hazembara, Establishing WUA is a task to be carried out Lezembara, Hao, Jelaka & Goche can be during the planning process or right at the cited as projects having no design and beginning of implementation. This handing over documents. Preparation of important activity is carried out by the proper design is the due responsibility of the implementing institution and the respective implementing institution (WVE) and the cooperative promotion office. Failure to regional irrigation authority. The absence of establish legally instituted WUA and elect proper design document also resulted in leaders results in problem of proper creating problem to proper completion and operation of schemes. Some older projects handing over of the schemes to concerned implemented by LWF & WVE can be cited stakeholder institutions. At this point, it is as examples where strong WUA is not worth to mention that the main reason not to established. At such schemes, activities of handover irrigation schemes built by WVE silt cleaning from main canal and minor at Omosheleko wereda in Kembata zone is maintenance on gates could not be carried absence of proper design documents. The out because of lack of community and implementing institution may phase out resource mobilization in organized manner. from the area up on completion of In some projects like Jelaka, social problems agreement period and if the schemes are not like conflict were also observed due to water handed over to the stakeholders, they may rights. In the following sections, the issue of totally abandon. conflict is discussed in general and specific It is clearly seen that the handing over terms. problem largely results because of failure of Conflicts due to water rights are of two concerned institutions to discharge their due types: responsibilities. ♦ Conflict among irrigators, which is not Like wise, the cooperative promotion office common in the region as an institution at zone or wereda level has ♦ Conflict among irrigators and up stream a responsibility to properly organize the settlers or down stream settlers. The irrigation water users and follow up the conflict with up stream settlers is mainly collection of periodic contributions by users due to diversion of same river or to conduct maintenance works and help tributary for other irrigation or some sustain the project. However, almost in all other purpose as seen at Jelaka scheme. the visited projects irrigation water users are The riparian law that says ’first come just aware of this but have not yet started first served’ does not seem to work well contributions. in such conditions. The conflict with The institutional problem is also manifested down stream settlers mainly occurs by the absence of follow up during the during low flow seasons when irrigators operational phase of the implemented completely divert the stream flow to schemes. The performance evaluation of fields. This case is observed in Ufute irrigation projects built by NGOs is the duty scheme. of zonal disaster prevention and The issue of downstream water rights is preparedness office and the woreda council. something that should be discussed in detail In addition to periodic reports given to these during the feasibility study and failure of offices, these government institutions should doing these results in creating operation and observe the field level situation of individual utilization problems discussed above. schemes and should try to curb problems prior to total failure of the projects. 5.4 Economic Problems Most users of visited irrigation schemes in 5.3 Social and Operational Problems the region are characterized by low income The planning and institutional problems can condition at household level. Users are also be reflected in the proper operation and usually heard of complaining against the utilization of the implemented schemes.

high rates of costs of production inputs such production, purchasing on the market and as fertilizers and seeds. As indicated in the through effective food transfer programme. table below, 25 schemes of the 26 Small scale irrigation can lead to availability considered in the analysis generate no of food at household level through increased resource to help sustainability of schemes by productivity, stable production and hence conducting minor maintenance and increases income leading to alleviation of rehabilitation activities. The reason for this economic problems at household level. It is partly economic problem. However, this also appears that development of small scale does not mean that irrigation schemes are irrigation schemes helps the country attain not changing the life of beneficiary food self sufficiency at national level households. As an example, in cash crop producing areas of the region such as 6. Inventory and GIS based database Sidama, users are observed to fetch building considerable amount of money. During Regional level inventory has been made to interview with beneficiaries of kedoboga collect the information on existing schemes. irrigation scheme, it was revealed that an The information is collected from field individual is earning up to Birr 2000 per survey & study and design reports and other quarter of a hectare of land per one season reports concerning the topic. All the of cash crop. information pertaining to the irrigation Generally speaking, the country’s food projects are inserted in spreadsheet as situation is characterized by food insecurity dBASE and then attached to regional maps at both micro and macro levels. The major using Arc-View GIS [Fig-8 below]. area of concern is the availability of food at household level. The country is even not The most important advantage of the able to produce most of its food digitized maps is that they can be used as requirements in normal years. tools to store and retrieve information in Ethiopia's national target is to achieve food Arc-View GIS windows. Information of an security both at national and household individual scheme can be easily accessed by levels. To achieve this, specific programmes clicking on the point theme representing the are needed to address the two sides of the scheme. Like wise, any additional food security equation: availability of food information can be stored in the through increased production and storage, spreadsheets attached with the point themes. and access to food through family

Fig-8 Distribution of irrigation schemes and drainage basin

Analysis. Prentice Hall Inc. Englewood 7. Overall Conclusion cliff, New Jersey In the past, a considerable effort has been Ethiopian Valley Development Studies exerted by government and various agencies Authority (1990) Preliminary Water to promote irrigation development in the Resource Development Master Plan for south region. However, what this research Ethiopia. Volume 2, Water and Power has revealed is that a considerable number of Consultancy Services, Addis Ababa schemes (18%) have already totally failed Garg, S.K. (1989) Irrigation engineering and because of the various problems discussed in Hydraulic structures. Khanna Publishers the preceding sections & a significant Delhi number are performing below their capacity. Henry, M. and James, M.W. (1980) Applied Accordingly, the government and various Hydraulics in Engineering. other agencies involved in the subject need Hartmut K (1995) Water Management, Use to revise the approach towards irrigation and Conflict in Small ScaleIrrigation. development by: Nairobi, Kenya ♦ Integrating local research with modern IWMI (2004) Experiences and Opportunities irrigation development by creating for Promoting Small Scale Irrigation and mutual relationship between research Rain Water Harvesting for Food Security and irrigation development in Ethiopia. Addis Ababa ♦ Promoting irrigation development Kinori, B.Z and Mevorach, J. (1984) Manual activities based on local knowledge and of Surface Drainage Engineering. community participation Volume2, Elsevier edition, Amsterdam Novak P., Moffat B., Nalluri C. and ♦ Building the capacity of the technical Narayanan R. (2001) Hydraulic personnel involved in the subject so that Structures. Third Edition, SPON PRESS, better skills can be gained in planning, London and Newyork design and implementation of projects Nagar A.L, Das R.K. (1997) Basic Statistics. ♦ Integrating software and hardware Third Edition, Oxford University Press, aspects of irrigation schemes rather than Delhi focusing largely on design and Punmia B.C, Pande B.B. Lal (1992) construction alone as hardware Irrigation and Water Power Engineering components Mehra offset press, New Delhi Failure to consider these elements may result Rozgar, B. (1995) Design of Diversion in subsequent failure of other schemes. Weirs. John Wiley and Sons, England

Ramachandra Rao A. and Khaled, H. (2000)

Flood Frequency Analysis. CRC Press, References Newyork Co-SAERSAR (2000) Headwork, Main Arora, K.R. (1996) Irrigation Water Power canal and Reservoir Pond Design Report and Water Resources Engineering. for Lasho Small Scale Irrigation Project. Standard Publishers and Distributors, Awassa Delhi Co-SAERSAR (1998) A Review on Arved J. Raudkivi (1993) Exclusion and Irrigation Performance. Awassa Removal of Sedimentation fromDiverted Co-SAERSAR (2000) Maintenances and Water, Volume 6. A.A Balkema Rehabilitation Strategies forSmall Scale Publishers, USA Irrigation Projects. Awassa Asfaw Afera (2004) Appraisal of Current Co- SAERSAR (1996) Study Report and Design Practices of River Diversion Design of Diversion Weir for Eballa Structures in the Amhara Region. Addis Small Scale Irrigation Scheme. Awassa Ababa Co-SAERASAR (2004) Inventory report Asit K., Biswas (1997) Water Resources: on the status of 24 irrigationSchemes. Environmental Planning, Management Awassa and Development. Mc Graw Hill, Co-SAERAR (1998) Design Report of Lamo Newyork Small Scale Irrigation Project. Awassa Chow, V.T. (1959) Open channel hydraulics. WVE-Shone ADP (2001) Study and Design McGraw Hill Report of Lenda IrrigationScheme. Daniel P., Jery R. and Douglas A. (1981) Hosaina Water Resource Systems Planning and

Across System Comparative Assessment of irrigation performance of community managed scheme in Southern Ethiopia

Belete Bantero1, Mekonen Ayana2 and Seleshi Bekele3 1 2 3 UNIDO, AMU and IWMI-NBEA [email protected]

Abstract highly accessible, moderately accessible, less accessible, very less accessible, poorly The water users located at the upstream of accessible and water scarce zones. the irrigation system have more access to Sometimes the tail-end, which is water than those located at the downstream characterized by water scarce zone, is also of the system. Moreover, the irrigation found to be affected by water logging. activity of tail-enders is seriously affected in Since, available water and demand for water both water scarce and water abundant are not continuously monitored and periods due to under and over irrigation managed, the situation outlined paves the respectively. Not only water but also the path for potential conflicts among water situation of landholding affects the users in response to visible livelihood productivity and income of the farming differences. Despite disruption of community. The goal of this study was to downstream users from irrigating their field, assess the spatial variation of irrigation significant loss of scarce resource by the performance and to evaluate its effects in upstream users have resulted in detectable terms of performance of agricultural environmental threat such as water logging, production (intensification and sodicity (10.44meq/l, k (H.C.) 0.00279cm/hr productivity), income and resources base as ) and salinity problems in the area. well as the environment. Across-system performance assessment study sponsored by 1. Introduction International Water Management Institute (IWMI) was done on the basis of simple There is a common perception that water illustration of the approach followed for the users located at the upstream of the assessment and different levels of water irrigation system have more access to water accessibility along the canal reaches. The than those located at the downstream of the level of availability of irrigation water or system. Moreover, the irrigation activity of accessibility to the farm is affected by the tail-enders is seriously affected in both water proximity of the farm to the water source or scarce and water abundant periods due to to the water carrying canals. The study under and over irrigation respectively. In confirmed that as one gets away from both Hare irrigation scheme there are three the water source and the canal the diversion sites very close to each other accessibility of water becomes less and less, which are planned to serve a wide range of unless proper system for water allocation is users. Before the establishment of the in place and practiced. In view of that, six scheme, only limited numbers of farmers zones can be differentiated based on the which are close to the water source were condition of water accessibility. These are using traditional methods to irrigate their

lands. It is common that after the o to evaluate the effect of being at the establishment of irrigation scheme the upstream, midstream or downstream number of population in the command area regions of irrigation canal system in increases through migration of peoples from terms of performance of agricultural elsewhere. Intended is to get the benefit of production (intensification and irrigation that the early settlers are enjoying. productivity), income and resources Land acquisition of new comers is most base likely possible at downstream of the canal reaches in the system. Not only these late 2. Description of the irrigation scheme settlers will become the victims of water shortage but also the whole irrigation system 2.1. Hydrometeorology of the area will get under pressure because initially designed capacity may not met the increased The climate of the area is characterized by demand through time. Unless the available mean maximum and minimum temperature water and demand for water are of 30.3 and 17.40C respectively, annual continuously monitored and managed, the rainfall of 843mm and potential situation outlined above could be one of the evapotranspiration of about 1644mm. Mean potential causes of conflict among water monthly distribution of these parameters are users. shown in Table 1. The rainfall distribution There are complaints among four involved pattern is bimodal with first and maximum Kebeles regarding unequal distribution of peak in April to March and second peak in water among the users in the scheme. October. The area is characterized by high Especially, during dry seasons Kolla Shara potential evapotranspiration rate that ranges Kebele which is served by the upstream from 112mm in July to 180mm in March. diversion structure is diverting water Consequently, except in April and May, the without considering the share of three other evaporative demand of the area is greater Kebeles (Chano Dorga, Chano Chalba and than the amount of natural rain. This means Chano Mile) that are feed by the rest two that there is a negative climatic water downstream diversions. There are visible balance in the area. This calls for livelihood differences among the supplementary water application to the crop beneficiaries of the scheme. On the other fields through irrigation to sustain crop hand there are poor and food self- production. The warmest months of the year insufficient households living in the system. are February and March while the coldest Not only water but also the situation of are November and December. landholding affects the productivity and The maximum flow hydrograph of Hare income of the farming community. Provided River which is the source of water for the that water is sufficiently available, the direct scheme shows also two distinct peaks that benefits of irrigation, in terms of increased occur in May (5.60m3/ s ) and October farm output, will tend to accrue in (4.53m3/s ). The low flow hydrograph proportion to the size of landholdings, with between the two peak rainfall periods, i.e. large holders benefiting more than from May to October, is almost consistent smallholders, and smallholders benefiting ranging from 1.26 to 1.62m3/s. The low flow more than the landless. declines during the months from December The objective of this study was to assess the to April. spatial variation of irrigation performance in Hare community managed irrigation scheme in southern region of Ethiopia. The specific objectives are: o to assess the water distribution performance of the scheme Table 1: Mean monthly values of hydro-meteorological parameters

Rainfall PET Temperature ( 0C ) Hare River Flow (m3/s) (mm) (mm) Maximum Minimum Maximum Low flow January 28.1 139.5 31.7 16.3 1.31 0.69 February 27.9 140.0 32.9 17.1 1.49 0.64 March 64.0 179.8 33.0 18.3 1.95 0.63 April 144.1 141.0 30.8 18.2 4.06 0.85 May 140.5 136.4 28.9 17.9 5.59 1.33 June 63.1 120.0 28.1 17.9 3.79 1.25 July 43.4 111.6 27.7 17.9 3.81 1.26 August 53.2 124.0 28.5 18.0 3.54 1.31 September 78.1 135.0 30.1 17.8 4.26 1.33 October 110.6 136.4 29.8 17.7 4.53 1.62 November 59.3 138.0 30.6 16.0 2.65 1.16 December 31.1 142.6 31.1 15.7 2.13 0.93 Total/Mean 843.2 1644.3 30.3 17.4 3.26 1.09

To see the relationship between demand and (Fig. 2), its total diversion will likely have available water two scenarios have been negative effects on these ecosystems. The considered; (i) with 75% dependable water balance (water supply minus demand) diversion and (ii) with 100% diversion of is positive from April to August. the available water. The results are presented in Fig.1. In both scenarios the Fig. 1: Mean decade water balance (water available water cannot meet the water demand minus available water from the demand for irrigation during seven source) from September to August. consecutive months of the year viz. from 2.2. Background of Hare irrigation late September to March. Feedback from the scheme users also revealed that these identified periods to be water scarce times. Total 2.2.1 Irrigation Scheme diversions of water from the river course to the canals have been observed in the months Hare irrigation scheme encompasses three of December and January. The available diversion systems the upstream diversion water in these months can only irrigate an with control gate, the midstream traditional area which is 24% of actually irrigable land diversion and downstream diversion weir. (2224 ha). As the river pass through These diversion points are respectively diversified bushes to end in Abaya Lake designated as D1, D2 and D3 (Fig. 1). which is located downstream of the scheme The upstream and midstream diversions and

3000

2000

1000

-1000

Water balance (liter/s) Water -2000

-3000 I-F II-J I-A I-A I-N I-S I- J I- J I- J I-D I-O I-M II-F I-M III-J II-A III-J II-A III-J II-N II-S II- J II- J II-D II-O II-M III-F II-M III-A III-A III-N III-S III-D III-O III-M III-M 61 Growing season

75% Diversion of available water 100% Diversion of available water

their delivery systems were established in Chano Dorga, Chano Chelba and Chano the year 1993 while the downstream Mille. Fig. 2 shows the diversion points and diversion weir was implemented in 1996. It the canal systems delivering water to the was meant to serve four villages which are respective Kebeles. locally called “Kebele”, viz. Kola Shara,

Fig. 2: Layout of Hare Irrigation Scheme

diversion is accomplished by a simple open The upstream diversion (D1) of water is cut in to the river bank which is supported accomplished by simple concrete intake by diagonally arranged temporary barrier structure. The earthen unlined canal is (stones, plant rests) and act as divide wall. receiving water from the intake structure to Unlined earthen canal which is running on irrigate Kolla Shara command area. The the left bank of the river takes water from 6km long main canal was initially this diversion point to irrigate Chano Dorga trapezoidal in shape and now it is irregular. area. The main canal runs over 7.1 km The canals pass through villages and are distance and has got irregular shapes. In subjected to damages by human and animals some areas the canal bed has developed to which resulted in bank breaks, overtopping, gorges while in other locations shallow weed growth, accumulation of silt and depth and wide surface areas are the wastes in the canals. characteristics of the canal. There is not as Downstream of D1 is D2 the point at which such well designed secondary canal. Hence traditional diversion of water is practiced. field channels arranged by the community That means there is no headwork and water take of water which is directly used to

irrigate the corresponding fields. There are well as the significant change in their about more than 60 major outlet points standard of living. along the main canal. Since there is not Those who complained and protected the control structure at the head some farm irrigation canal from reaching their field at fields are affected by flood during rainy that time are now the one who are straggling seasons. to bring the water to their field are still relatively poorer. Downstream of the above mentioned intake The main canal taking of water from the points (D2) there is a diversion weir (at D3) weir is of two types. A certain length is to convey water it to a partially masonry masonry rectangular canal and the major type lined main canal. The structure was portion of the main canal is trapezoidal provided with a flow control mechanism pitching. The other irrigation infrastructures though its performance has deteriorated such as the turn outs, the division boxes, the from time to time. The weir and its delivery road crossings, the drops, the head and cross infrastructure were constructed by technical regulators are all constructed with masonry and financial support of Chinese and reinforced concrete, all are lined and government. It was planned to serve large plastered except in some area pointing work. portion of Chano Mille and some part of The main canal has got a length of 5.33 km. Chano Chalba irrigable areas. The There are 7 secondary canals which when construction of the weir was started without totally summed have got a length of 12.95 feasibility study and awareness of the users. km. The longitudinal slope of the canal Hence it was accompanied by complaints alignment is 0.1 % which was ensured from the users and even from local and through 13 drops. regional authorities. The reasons were luck of awareness, imbalanced weir location, and Group of farmers take water from secondary demand for small dam that could ensure the canals to their field plots through the outlet balance between water supply and demand structures prepared for same purpose. The over the growing season. During excess water from the canals and from the construction phase the communities have runoff joins the main drainage system. Part realized that the implementation of the of the drainage water joins the Hare River system will bring about the dissections of and the remaining flows in to Lake Abaya. It their farming field for delivery systems and is only the main drainage canal which is for the access roads etc. and became more functional at present and serves to remove reluctant. In spite of these resistances the immense amount of excess water at the Chinese contractors have implemented the down stream section of the irrigation diversion weir. scheme. While the construction work was under final As far as water allocation is concerned there phase, some parts had started to give partial is a water users committee (WUC) which is services but the discontented farmers partly responsible for fair distribution of water started destroying the irrigation channels, among the users. WUC also organizes particularly plowing over the secondary and maintenance activities. Whenever irrigation tertiary canal systems and dismantling of the is required, each water users group submit structures. But gradually the community request for water to the WUC. Once the realized that they have made a lot of request is made to the WUC, then this mistakes during construction while they committee prepares a tentative time have observed a lot of positive impacts in schedule up on which irrigation can be their life due to the intervention. made. Particularly, the primary user of this weir, Chano Chalba, is the leading Kebele with respect to their production competence as

2.2.2 Socioeconomics sweet potato and also cotton. Those farmers who have better access to water have Prior to construction of irrigation scheme in preference of growing banana mainly due to the area local farmers used to irrigate their its marketability on central market. As it is lands on their own traditional ways. This suitable for fruits trees the area is known for accounts to only some 300 hectares. After the provision of fruits such as mango, the implementation of first and second phase avocado, papaya to the central market. construction in 1993 1996 respectively more Farmers grow crops such maize and sweet than 1031 ha and 1336 ha of land have been potato for own consumption. Cotton is an developed. The number of beneficiaries has alternative crop for tail-enders as it also increased from time to time. withstands water stress conditions. The principal crops grown in the command area are banana, maize, mango, avocado,

Table 2: Demographic feature of Hare irrigation scheme

Number of Households Number of Population Name of Villages

Male Female Total Male Female Total

Kola Shara 800 164 964 2358 2474 4832 Chano Dorga 413 20 433 1403 1363 2766 Chano Chalba 751 175 926 2339 2713 5052 Chano Mile 821 102 923 3950 3074 7024 Total 2785 461 3246 10050 9624 19,674

As per the information in 2006 the total distributed in the three of the Kebeles except number of beneficiary households is about in Chano Dorga which has the least number 3246 out of which 14% are female headed. of households (Tab. 2). The relative proportion of female and male Table 3 shows the land use patterns of the of the total population in the command area command area. From the total irrigable area is almost equal. More female households of 2224ha in the scheme the highest are found in the command areas of upstream proportion is in Chano Mille (32.2%) and modern diversion canals viz. in Kola followed by Chano Chalba (29.2%), Kola Shara and Chano Chalba commands. The Shara (27.7%) and Chano Dorga (10.9%). number of households is almost equally

Table 3: Land use patterns in the command area Type of land use Kola Chano Chano Chao Total Shara Dorga Chalba Mille Total area, ha 800 745 799 900 3244 Cropped area (annual), ha 251 282 199 496 1228 Cropped area (perennial), ha 391 120 450 400 1361 Total cropped area, ha 642 402 649 716 2409 Irrigable area at the moment, ha 617 242 649 716 2224 Irrigated area at the moment, ha 617 242 649 454 1962 Area occupied by infrastructures, ha 100 43 90 164 489 Forest area, ha 20 200 40 10 270 Grassland, ha 38 100 20 10 188 Average landholding, ha 0.97 1.50 1.50 1.25 1.42

Except in Chano Dorga Kebele, the largest been inspected. The capacities of the canal proportion of cropped area is covered by systems have been measured at different Perennial crops such as Banana and other reaches. fruit trees. Irrigable areas in the first three More than 800 GPS points has been taken Kebeles have already been developed under for evaluating the scheme performance with irrigation. However in Chano Mille which is respect to its proximity to the watercourse relatively far from the headwork, the and main canal reach. irrigated area is only 63% of irrigable land. Measurements of flows in the canals have been conducted after the canals are 2. Methodology maintained, i.e. after the removal of sediments, weeds and other barriers in the Like other schemes in the country, there is canals. Since this study was conducted no any kind of record available in Hare during the out set of the rainy season right community managed irrigation scheme. after they maintained the canal section to Hence, useful information for the execution start irrigation, while they were not yet of this study was generated through opened their many illegal outlets particularly measurements, observations and interviews. the canal of traditional diversion to Chano Discussions have been held with water Dorg. Accordingly, the measurements for development committee, community elder this diversion have been only taken along groups Kebele administration member, the canal at 5 points (300, 1804, 3802, 5390 development agent/extension agent, farmers, & 6730 m). Then again, for the other two female headed households etc. The scheme diversions, measurements have been taken has been frequently visited to examine the safely in both directions along the canal and operations, the conditions and functions of laterally at several points. For instance, for irrigation systems, agronomic practices of the u/s diversion, Kola Shara along the canal farmers, cropping patterns etc. These visits at 370, 1600, 3290, 4190 & 4889 m were conducted with the accompany of positions, while for modern diversion 334, different water users and operators group, 1100, 2230, 2810, 3860, 4030, 5560 & viz. water development committee members, 5890m. For the lateral flow performance administrators, local elders, model farmers, investigation, the traditional diversion was male and female farmers. An attempt was not considered as it has no as such properly made to understand the system and collect managed or working secondary canals. data necessary to measure the performance Thus, the other two diversions lateral canal indicators. performance was evaluated in two The main canal and the secondary canal, categories that is comparison between the tertiary canals and field canals including the secondary lined canal and the secondary drainage lines of all the three schemes have earthen canal of the Modern versus the

secondary earthen canal of the Modern Cum Across-system performance assessment was Traditional. done on the basis of illustration given in It is common to measure the discharge at figures 3 and 4. The level of irrigation water the intake and the application point to availability or accessibility to the farm is estimate the losses. Even though such affected by the proximity of the farm to the method is capable of giving the general water source or to the water carrying canals. nature of the water conveying structures As one gets away from both the water performance, by this study we assumed such source and the canal the accessibility of methods are really less important as it hardly water becomes less and less unless proper locate the apparent position of the system for water allocation is in place and significant loss. In contrast the method we practiced. According to figure 4, seven applied here, measuring within short interval zones can be differentiated based on the and projecting the loss per unit length found condition of water accessibility. These are to be a relevant technique to get the desired highly accessible zone, moderately result. Accordingly by linking the actual accessible, less accessible, very less efficiency and the distance from the intake accessible, poorly accessible and water point the linear correlation coefficients have scarce area. Some times the tail-end which been obtained. is characterized by water scarce zone can also be affected by water logging.

4000 3500 SC - 1 3000 2500

2000 1500 1000 Turnout

500 Distance along MC from Water Source (km) Water source (0) 500 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 1000 Distance from MC (m) Distance from 1500 2000 2500 SC - 2 3000 3500 4000

Fig. 3: Simple illustration of the approach followed for the assessment

All relevant data were collected along main defined in this case with respect to canal (MS) and secondary canals (SC) that proximity to water source that decreases are functional during the season. from the head to tail end of the canal Accessibility of water to a farm unit which systems. The assessment was carried out is measured in amount and timely delivery is following two directions (Fig. 3), i.e., (i)

along the main canal that receives water to water than secondary canal 2 (SC-2) due from the main source and (ii) along to its relative closeness to the water source secondary canals that takeoff from the main along the main canal. Likewise there are canal. The hypothesis here is that the differences along the secondary canal itself secondary canal 1 (SC-1) have more access as one goes from head to tail.

Table 3: Summary of information in Fig. 4 Zones Distance ranges (m) along Water accessibility Main canal Secondary canal 1 0 - 500 0 - 250 high 2 500-1000 500-750 moderate 3 1000-1500 570-1000 less 4 1500-2000 1000-1250 Very less 5 2000-2500 1250-1500 Poorly accessible 6 2500-3000 1500-1750 Water scarce

Areas that fall in the regions of poorly resistant compared to common crops grown available and water scarce zone are in the area. characterized by critically limited water availability and hence depend more on rainfall. Water logged areas are located at end of the systems. The owners of such lands are suffering from shortage of water during irrigation and flooding during off- irrigation periods. These areas are covered by cotton which is relatively water stress

Fig. 4: Simple illustration for different levels of water accessibility along the canal reaches

4. Results and Discussions smaller and smaller. This is exactly what is observed around successful irrigation 4.1. General Practices schemes. The main advantages of irrigation practice lay on provision of opportunity for One of the most important problems that intensification of cropping. Under exist in and around the small- and medium- decreasing size of landholdings in irrigated scale irrigation schemes in the country is agriculture, intensification of cropping discrepancies between design specifications coupled with productivity improvement is of the systems and expectations from the the way to enhance food production. No same. No reference is usually made, if at all doubt that better access to inputs and available, to the design documents while technologies contribute to improvement of operating and managing the schemes. productivity. Intensification of cropping is With increasing number of population the mainly determined by the type of crops size of the landholdings in an area becomes selected and availability of water.

12 Kola Shara 10 Chano Dorga Chano Chalba 8

Cropping Intensity (ha/yr) Intensity Cropping 2

SC-1 SC-2 SC-3 SC-4 0

Distance along SC (km) 0-0.50 0-0.50 0-0.50 0-0.50 Distance along MC (km) 0.5-1.0 1.0-1.5 1.5-2.0 0.5-1.0 1.0-1.5 1.5-2.0 0.5-1.0 1.0-1.5 1.5-2.0 0.5-1.0 1.0-1.5 1.5-2.0 0-1.5 1.5-3.0 3.0-4.0 4.0-6.0

Fig. 4.1: Cropping intensity along the canal reaches (the product of average area cultivated per season and frequency of harvest per season)

Figure 4.1 shows the spatial variation of curve coincides with the previous two up to mean cropping intensity that was practiced a certain distance beyond which rapid raise by one household. It is evident that the is taking place. Farmers located here, i.e., 4 intensity of cropping decreases along the – 6 km away from the diversion point along main canals from upstream to downstream this canal, are practicing higher crop incase of communities in Kola Shara and intensification compared to those located in Chano Dorga which are served by the the middle and head regions of the canal. upstream and midstream diversions These framers are trying to convert the respectively. On the contrary, there is slight challenges of flooding and shallow decreasing trend of intensity as one goes groundwater depths in the tail regions of the down along the secondary canals. Under canals to opportunities in that they are Chano Chalba condition, the trend of the

adopting multi-cropping system or intercropping.

140

120 Kola Shara Chano Dorga 100 Chano Chalba

40 Irrigation Interval (days) 20 SC-1 SC-2 SC-3 SC-4 0 Distance along SC (km) 0-0.50 0-0.50 0-0.50 0-0.50 0.5-1.0 1.0-1.5 1.5-2.0 0.5-1.0 1.0-1.5 1.5-2.0 0.5-1.0 1.0-1.5 1.5-2.0 0.5-1.0 1.0-1.5 1.5-2.0 Distance along MC (km) 0-1.5 1.5-3.0 3.0-4.0 4.0-6.0

Fig. 4.2: Irrigation Interval (days the users have to wait until the arrival of the next irrigation water)

As can be viewed from Fig. 4.2 the 4.2. Water delivery performance frequency of getting water by the users decreases from head to tail-ends of all canal The values of some performance indicators systems. It varies from 7 – 30 days, from 7 – are given in the table 4.1. As far as most of 90 days and from 15 – 120 days in Chao these indicators are concerned, Chano Dorga Dorga, Chano Chalba and Kola Chara which is served by traditional diversion (D2) respectively. Irrespective of the location of is found to perform better than others. The diversion point with respect to each other, productivity of land and water is higher in those users close to water source get Kola Shara followed by Chano Dorga. The frequent access to water. The communities later is characterized by greater values of using traditional diversion get frequently water supply performance indicators, i.e. water compared to Kola Shara and Chano RIS, RWS and WDR. Chalba that use respectively simple diversion structure and diversion weir.

Table 4.1: Values performance indicators Performance Indicators Kola Shara Chano Chano Dorga Chelba Output per water consumed (birr/m3) 0.56 0.44 0.35 Output per cropped area (birr/ha) 4400 3464 2736 Water delivery ratio (WDR) 0.56 1.09 0.71 Relative irrigation supply (RIS) 1.40 2.70 1.78 Relative water supply (RWS) 1.18 1.79 1.36

Community in Kola Shara, Chano Dorga Dorga command areas. From midstream to and Chano Chalba are served by the downstream no decreasing trend both along upstream, midstream and downstream main canal and secondary canals rather diversions respectively. variation among the canals in terms of productivity is visible. The lower areas of 4.3. Output performance tail ends are usually characterized by shallow groundwater tables (o.6 – 2m below According to figure 4.3 the productivity of the surface) which are likely to contribute to banana decreases from upstream to the water requirements of perennial crops. midstream rapidly in Kola Shara and Chano

100

90 Kola Shara 80 Chano Dorga Chano Chalba 70 60 50 40 30 Banana yield (qt/ha) Banana yield 20 10 SC-1 SC-2 SC-3 SC-4 0 Distance along SC (km) 0-0.50 0-0.50 0-0.50 0-0.50 Distance along MC (km) 0.5-1.0 1.0-1.5 1.5-2.0 0.5-1.0 1.0-1.5 1.5-2.0 0.5-1.0 1.0-1.5 1.5-2.0 0.5-1.0 1.0-1.5 1.5-2.0 0-1.5 1.5-3.0 3.0-4.0 4.0-6.0

Fig. 4.3: Productivity variation of banana along the canal reaches

100 90 80 70 60 50

40 Kola Shara 30 Chano Dorga

Yield performance (%) performance Yield 20 Chano Chalba 10 SC-1 SC-2 SC-3 SC-4 0

Distance along SC (km) 0-0.50 0-0.50 0-0.50 0-0.50 0.5-1.0 1.0-1.5 1.5-2.0 0.5-1.0 1.0-1.5 1.5-2.0 0.5-1.0 1.0-1.5 1.5-2.0 0.5-1.0 1.0-1.5 1.5-2.0 Distance along SC (km) 0-1.5 1.5-3.0 3.0-4.0 4.0-6.0

Fig. 4.4: Yield performance (actual yield/potential yield) irrigation along the canal reaches

Yield performance which is the ratio of shows more variation between upstream, actually harvested yield to potentially midstream and downstream in Chano harvestable yield of crop varieties can be an Chalba that is found in command area of indication for how agronomic practices and modern weir diversion. Except Chano other inputs (water and agrochemicals) were Chelba the yield performance of the other effective to exploit the yielding potentials of villages is the reflection of irrigation the crops selected. The indicator shows also intervals (Fig. 4.2). Those villages that have a decreasing tendency towards midstream. got water in shorter intervals have registered Almost equal yield performance is observed better yield performed than villages with in Chao Dorga which is located in the longer irrigation intervals. command area of traditional diversion. It

80000

70000

60000 Kola Shara 50000 Chano Dorga 40000 Chano Chalba

30000

20000

Annual income of a HH a (birr) of income Annual 10000

Distance along SC (km) 0-0.50 0-0.50 0-0.50 0-0.50 0.5-1.0 1.0-1.5 1.5-2.0 0.5-1.0 1.0-1.5 1.5-2.0 0.5-1.0 1.0-1.5 1.5-2.0 0.5-1.0 1.0-1.5 1.5-2.0 Distance along SC (km) 0-1.5 1.5-3.0 3.0-4.0 4.0-6.0

Fig. 4.5: Average annual income of Household along the canal reaches

The ultimate goal of irrigation development round access to water be it from canal or is to improve the livelihood of the farming groundwater is greater than farmers with community. Its achievement is largely limited access to water (Fig. 4.5). Annual depends on integration and coordinated income of households in Kola Shara and operation of nested systems of irrigation Chano Dorga command areas is variable which was indicated by Small and Svendsen irrespective of proximity to the canal. Most (1992). As the production increases the often, resources are efficiently and income of the farmers is likely to increase. effectively utilized when they become Other factors such marketability of the scarce. Similarly households that have less produce and market access influences the access to water tend to use it more total income of the households. The annual efficiently than those having access to income of households which have year abundant water resources.

160000

140000

120000

100000

80000

60000

40000 Resources base (birr) Resources base 20000

0 0 10000 20000 30000 40000 50000 60000 70000 80000 Annual income (Birr)

Kola Shara Chano Dorga Chano Chalba

Fig. 4.6: Relationship between annual income from agricultural production and resources base (monetary values of important possessions; example house, chattels etc.)

Improved agricultural production coupled household improves with the level of with remunerative selling of the production income. There are farmers who have will result in improved income of the managed to construct house in the nearby farming community. This together with towns, buy taxis and small transporters etc. better social facilities enhances the betterment of farmers’ welfare. Fig. 4.6 shows how the resources base of a

Table: Chemical properties of the soil of different Kebeles EC PH SAR ESR ESP (%) Texture K cm/hr (μmohs/cm) (meq/l) (meq/l) (meq/l) Kola 2060.8 8.35 10.44 5.87 85.44 Clay 0.00279 Shara Chano 166.4 7.82 6.94 6.94 71.61 Sandy 0.015 Dorga Clay Loam Chano 147.2 7.69 5.3 1.63 61.94 Loam 0.0882 Chalba Chano 170.24 7.47 0.5 0.49 32.7 Clay 0.00576 Mile(D)

4.5. Environmental Indicators has a potential to cause depletion of the important micro nutrients such as iron, From suitability aspect based on FAO frame manganese, and zinc work, the soil of Kola Shara with its SAR, 10.44meq/l value and the soil of traditional irrigation users, SAR value of 8.84 both 4.6. Social Problems appear within 8 to 18meq/l hence characterized in moderately suitable range Although there are official rules and (i.e. class 2 level). However, along with its agreements that regulate water distribution clayey nature and very small hydraulic among the users, there exist unofficial conductivity value calculated, that is diversions, unauthorized use of water, 0.00279cm/hr, which is quite less than vandalism and stealing of water by the threshold level of 5cm/hr, hence the soil upstream user kebeles particularly by Kola category as per critical limit for sodium Shara as the controlling mechanisms are the tolerance could be n2 (that is permanently local river materials that can easily be not suitable range). Nevertheless, salinity subjected for deviation. Non-permanent level (EC, 2060.8 μmohs/cm) that is control materials instead they use stones and sufficiently less than10, 000 μmohs/cm) plant rests, as they are easily shifted from preserved the area soil in the moderate their initial position (see figure 4.14) along suitability range. Still, the soil is categorized with the absence of observers assigned to under good to injurious, suitable only oversee the proper functioning of the permeable soils and moderate leaching delivery system aggravated the disruption of coupled. The extent with in this range is the downstream users from irrigating their harmful to more sensitive crops. This has field. Such problem of unlawful diversion been also confirmed from the actual survey also exists by Chano Chalba KA, the one done in the area; they grow mostly cotton using the modern diversion, because of such which is salt tolerant crop. From acidity and gateway mentioned above the downstream alkalinity point of view also, the soil of Kola users Chano Mile KA are highly disrupted to Shara is categorized under alkaline category irrigate their field. as its pH value, 8.35 is greater than 8, which

74 Arrangements of Stone, gravels, plant rests to facilitate diversion of water

Diversion canal

River Flow

A Main River Course

Fig. Simple illustration showing how farmers manipulate the arrangements of traditional water diversions based on the conditions of water availability (with decreasing availability of water in the rivers and increasing demand for water the adopted diversion arrangement by farmers shift from A to B and to C

mainly at Kola Shara KA looking into FAO frame work and the facts in the area. 5. Conclusion Thus, in order to maximize the resourceful At Hare irrigation scheme, it has been use of the present water supply at Hare identified that the availability of water, the irrigation scheme, a matched and designed poor water management and high demand better water management of the irrigation are the main actors. From the water scheme is not only capable of reducing the availability versus demand analysis effect of naturally occurring low flow and undertaken within ten days interval for the subsequent yield reduction, but also the whole year, out of P=38, f (D>Q, 0.00)= 20 increased efficiency augments the times, s (D

Another critical social problem is continued inconsiderate use of irrigation water at upstream i.e. ample loss of water which is disrupting downstream users from irrigating their field, as well as visible environmental threat to the up stream irrigators themselves such as water logging coupled with continual malaria out break and other waterborne diseases, sodicity (SAR, 10.44meq/l, clayey soil, hydraulic conductivity calculated, 0.00279cm/hr) of their soil (in permanently not suitable range (n2)) and salinity problem,

Analysis of irrigation systems using comparative performance indicators:

A case study of two large scale irrigation systems in the upper Awash basin

Abdu Beshir1 and Sileshi Bekele2 1Oromia Water Works Design and Supervision Enterprise, 2IWMI-NBEA [email protected]

Abstract consumed are in the order of 21017.44, 23791.21, 0.74 and 2.3 for NuraEra and 20074.97, 13916.03, This research was conducted to introduce the 1.4 and 1.2 for Wonji, respectively. . concept of evaluating the countries large scale irrigation systems and using the IWMI’s minimum NuraEra spent 0.36 percent of its income in the set of indicators for the same purpose. This was year of analysis and that of Wonji is 0.17 percent done by selecting two irrigation systems in the for operation and maintenance of the delivery Upper Awash Basin. NuraEra state and Wonji system. FSS was 100% for both of the farms. And estate farm were selected for their relatively better gross return on investment was 84.7% for NuraEra organization and management, nearness to weather and 76.8% for Wonji. station and their representative nature of the large scale irrigation systems in the country. Background

The necessary primary and secondary data to Rapid increases in the world's population have calculate the nine indicators were collected which made the efficient use of irrigation water vitally are measurement of canal capacity and pump important, particularly in poorer countries, where capacity, and total yields, farm gate prices of the greatest potential for increasing food production irrigated crops, area irrigated per crop per season or and rural incomes is often to be found in irrigated per year, crop types, production per season or per areas. It has therefore become a matter of serious year, incomes generated from water fee and concern in recent years that, despite their very high cropping pattern. costs, the performance of many irrigation schemes has fallen far short of expectations (FAO, 1986). The indicators used are output per cropped area, output per command area, output per irrigation The country’s irrigation efficiencies are generally diverted, output per water consumed, relative water low, of the order of 25 to 50%, and problems with supply, relative irrigation supply, water delivery rising water tables and soil salinisation are now capacity, financial self sufficiency and gross return emerging. As is usually the case, these problems on investment. only emerge some 10 to 15 years after project

inception. More detailed information on the long- The result of the analysis shows that the ratio of term operational behavior of the existing schemes RWS, RIS and WDC for NuraEra are 4.8, 6.6 and often results in optimum design at the planning 1.4, for Wonji estate 1.36, 1.4 and 0.77 stage of the new schemes, leading to lower capital respectively. The four agricultural indicators; and lower overall costs, the avoidance of costly output per cropped area, output per command, output per irrigation supply and output per water

remedial works in the long term and achieving consumption, where total consumption includes maximum returns on water (Woodroofe, 1993). water depletion from the hydrologic cycle through process consumption (ET), other evaporative losses State run farms, which include large-scale irrigation (from fallow land, free water surfaces, weeds, systems, were reiterated as major components of trees), flows to sinks (saline groundwater and seas), efforts to develop the country’s agricultural sector, and through pollution. notably in the Awash Valley. However, productivity of these top–down managed systems We are interested in the measurement of production over the decades has been disappointing—the farms from irrigated agriculture that can be used to have been beset by a number of environmental, compare across systems. If only one crop is technical and socio-economic constraints. The large considered, production could be compared in terms scale systems in the Awash basin and elsewhere of mass. The difficulty arises when comparing suffer from water management practices that have different crops, say wheat and tomato, as 1 kg of resulted in rising ground water tables and secondary tomato is not readily comparable to 1 kg of wheat. soil salinisation where large tracts of land have When only one irrigation system is considered, or gone out of production (EARO 2001; Paulos). irrigation systems in a region where prices are

similar, production can be measured as net value of Even if evaluating the existing irrigation systems is production and gross value of production using an old phenomenon in the other parts of the world; local values. it has not been tried in Ethiopian large-scale farms. Hence, this study attempts to introduce the concept of comparative performance indicators as a tool to Output per cropped area evaluate the performance of large-scale irrigations Pr oduction in the Upper Awash valley. = Irrigated Cropped area The indicators (1)

Nine indicators are developed related to the irrigation and irrigated agricultural system. The Output per unit command area main output considered is crop production, while Pr oduction the major inputs are water, land, and finances. = Command area Indicators of Irrigated Agricultural Output (2) The four basic comparative performance indicators Output per irrigation supply relate output to unit land and water. These Pr oduction “external” indicators provide the basis for = comparison of irrigated agriculture performance. Diverted irrigtion supply Where water is a constraining resource, output per (3) unit water may be more important, whereas if land is a constraint relative to water, output per unit land may be more important. Output per unit water consumed Output per unit of irrigation water supplied and Pr oduction output per unit of water consumed are derived from = a general water accounting framework (Molden et Volume of water consumed by ET al, 1998). The water consumed in equation 4 on (4) page 21 is the volume of process consumption, in this case evapotranspiration. It is important to Where, distinguish this from another important water accounting indicator—output per unit total

Production is the output of the irrigated area Irrigation demand in terms of gross or net value of production measured at local or world prices, = the crop ET less effective rainfall.

Irrigated cropped area is the sum of the areas Relative irrigation supply is the inverse of the under crops during the time period of irrigation efficiency Molden et al (1998). The term analysis, relative irrigation supply was presented to be Command area is the nominal or design area consistent with the term relative water supply, and to be irrigated, to avoid any confusing value judgments inherent in the word efficiency. Diverted irrigation supply is the volume of surface irrigation water diverted to the Both RWS and RIS relate supply to demand, and command area, plus net removals from give some indication as the condition of water groundwater, and abundance or scarcity, and how tightly supply and demand are matched. Care must be taken in the Volume of water consumed by ET is the interpretation of results: an irrigated area upstream actual evapotranspiration of crops. in a river basin may divert much water to give adequate supply and ease management, with the Five additional indicators were identified in this excess water providing a source for downstream minimum set for comparative purposes. These are users. In such circumstances, a higher RWS in the meant to characterize the individual system with upstream project may indicate appropriate use of respect to water supply and finances. available water, and a lower RWS would actually be less desirable. Likewise, a value of 0.8 may not Relative water supply and relative irrigation supply represent a problem; rather it may provide an are used as the basic water supply indicators: indication that farmers are practicing deficit irrigation with a short water supply to maximize Re lative water supply returns on water. Total water supply = Crop demand Water delivery capacity(%) Canal capacity to deliver water at system head (5) = Peak consumptive demand (6) (7) Where: Where: Total water supply

= Surface diversions plus net ground water draft Capacity to deliver water at the system head plus rainfall. = the present discharge capacity of the canal at the Crop demand system head, and

= Potential crop ET, or the ET under well-watered Peak consumptive demand conditions. = the peak crop irrigation requirements for a monthly period expressed as a flow rate at the Irrigation supply head of the irrigation system.

= only the surface diversions and net groundwater Water delivery capacity is meant to give an draft for irrigation. indication of the degree to which irrigation infrastructure is constraining cropping intensities by

comparing the canal conveyance capacity to peak consumptive demands. Again, a lower or higher Description of the Study Area value may not be better, but needs to be interpreted in the context of the irrigation system, and in The mean annual rainfall of Awash Basin varies conjunction with the other indicators. from about 1600mm at Ankober; in the highlands north east of Addis Ababa top 160mm at Asayta on Financial Indicators the northern limit of basin. Addis Ababa receives 90% of its annual rainfall during the rainy period The two financial indicators are: March to September. At Dubti the same over all proportion is received during the two rainy periods, distributed 30% and 60% respectively. The mean annual rainfall over the entire western catchment Gross return on Investment (%) is 850mm and over the head waters of Awash, as gauged at Melka Hombole it is 1216mm.The production = annual and monthly rainfall are characterized by Cost of irrigation structure high variability (Halcrow, 1989).

(8) Wenji Estate farm The estate was constructed by HVA in the early Financial self sufficiency 1960s .It comprises 5925ha of sugar and is operated by the ESC (Halcrow, 1989). The estate is located Revenue from irrigation service fee = in the central east part of the main Ethiopia rift Total O & M expenditure system, 107 km southeast of Addis Ababa. It is found between 8030’ to 8035’N and 39020’E grid (9) and at an altitude of 1540 m a.s.l Average Maximum and Minimum temperature are 27.60C Where, and 15.30C, respectively. The surrounding topography is steep on all sides north and east being Cost of irrigation infrastructure considers the bound by river awash the west and the south by the cost of the irrigation water delivery system border drain which protects the estate from the run referenced to the same year as the SGVP, off and sedimentation from those steep slopes (Mukherji, 2000). Irrigation supplies are provided Revenue from irrigation, is the revenue by continuous electrical pumping from the Awash generated, either from fees, or other locally into a settling basin at the head of the main canal. generated income, and Night storage reservoirs cover a surface area of 60ha (Halcrow, 1989). Total O & M expenditure is the amount expended locally through operation and management

Nuraera state farm Canals and irrigation layouts have not been The farm is found in the eastern showa formally designed so that their capacities are not administration zone. It is divided in three sub listed. farms. From the total land owned by the farm Water use efficiencies would appear to be in the which is 3,277ha, 3,069 ha is covered by crops the range of 20%-30% (Halcrow, 1989). Irrigation of remaining 208 ha by infrastructures. The farm has citrus is generally by small basin around tree bases. been developed on an adhoc basis over several years and water abstraction is by gravity.

Main gate on the main canal at Nuraera Diversion Weir at Nuraera

Abstraction is measured by staff gauges in the main returned to Awash. The northern end of the canal canal from gravity offtakes irrigation is generally used to outfall into Lake Beseka. This has been undertaken 24 hours a day. Wastage of water is stopped by a cut-off-drain which is partly inevitable but it appears that little, if any flow is

constructed with the objective of leading the flows decided to measure the discharge at various gage back to Awash. heights and take the average discharge to calculate the amount diverted for each season. Methodology Maintaining the abandoned BOC took much of our After reconnaissance survey and consulting the time before engaging in the measurement. officials of the Upper Awash agro-industry enterprise, one farm was selected among the three farms of the enterprise. NuraEra farm was selected for its nearness to the weather station, relatively better organization and management and availability of secondary data. Moreover the other two farms are medium scale farms.

The other farm selected for its attributes of nearness to Addis Ababa and wheather station and the availability of data were Wonji sugare estate. Secondary data collections were started in February 2005 in collaboration with the farm officials. Preparation for work at Nuraera

Primary data collection In the case of Wonji showa estate farm, the pumps The job of primary data collection which is in the main pump station are more than 40 years measurement of canal capacity and pump capacity old. Their Design capacity went on reducing was started in April 2005 in collaboration with the through these years. Therefore the current capacity department of hydrology in the Ministry of Water of each pump should be measured. That can be Resources. done either by measuring the water depth over the cippoleti weir and use the formula to calculate the The upper Awash Water Administration center discharge or by measuring the flow on site. installed a BOC on the main canal at around 200m from the diversion weir for the purpose of updating The measurement was taken at 0.2*the depth of its rating curve for the reason that the flow water over the weir crest measured from the surface characteristics changes due to high rate of of the water and the width of the crest was divided sedimentation resulting in change of canal by 0.2m interval( see appendix B). morphology.

The plotted rating curve was used to find out the total water diverted using the daily records of the gauge height taken by main gate operator who is an employee of the center.

But due to the shortage of material and manpower, the center has been using the old rating curve for the purpose of knowing the total amount of water diverted. Due to the high siltation rate and the changing nature of the hydraulic condition in the main canal , it will be hard to get a reliable data using a rating curve plotted two years ago. For this reason it was

good number of years record of the production and prices of crops. For instance, wonji have a 50 year data of production.

The only problem in the case of the needed secondary data was the record of money spent for operation and maintenance of the irrigation systems

Results and Discussion

Measuring discharge from cippoleti weir at Wonji Nuraera After measuring the discharge at various working gauge heights, the average discharge was estimated Secondary data collection to be 3.1 m3/s. And this discharge was used to estimate the amount of water diverted in the year The Secondary data included total yields, farm gate 2003/04. prices of irrigated crops, area irrigated per crop per Based on the collected data of planting and harvest season or per year, crop types, production per of each crop for the year of analysis, the season or per year, incomes generated from water agricultural practice of the farm is divided into two fee and cropping pattern. seasons. The first season of cropping goes from January to July, the amount diverted in these Climatic data of Wonji farm were collected from months was. the nearby station at wonji while that of NuraEra Season A 3.1*212*24*60*60 = 56,782,080 m3 were collected from the head office of Ethiopian Season B 3.1*153*24*60*60 = 40,979,520 m3 National Meteorological Agency in Addis Ababa. Total amount of water diverted for the year 2003/04 Since both farms have a planning and program = 97,761,600m3 department run by skilled manpower, they kept

Total output of NuraEra farm for the year 2003/04 Crop Production in Unit price in Total Output Area (ha) Qt. birr/Qt. in Birr Orange 772 254,580.6 112 28,497,753.48 Manderine 181 35,844 87.50 3,133,836.55 Guava 55 1,307.2 32.20 42,038.27 grape Vine 11.18 538.3 400 215,304.00 Mango 40.35 9,395 117.75 1,106,250.65 Tomato 224.53 35,115 40.60 1,424,968.73 Onion 54.12 4,790 136.50 653,925.34 Green chilies 5.30 182. 134.60 24,494.91 Cabbage 0.55 23.6 40.60 959.31 Carrot 0.30 22.2 40 889.64 Beet root 0.45 36 40.00 1,444.40 Beans(local) 2.00 1,463.8 61 89,323.52

Bobbybeans(export) 149.50 6,221.8 1,383 8,604,265.49 Okra 0.60 3 172 532.92 Maize 350.00 17,002.6 120.20 2,043,368.86 Pop corn 58.60 385.7 483.30 186,411.74 Cotton 735.00 18,061.8 1,494.30 26,989,446.28 Total 2,640.5 384,973 73,015,214

As it is clearly seen the farm used wider cropping pattern. And based on the planting and harvesting dates of each crop, the whole year is divided into two seasons and their CWR and irrigation requirement was calculated using CROPWAT 4.2. The result is as shown below:

CWR and IR of each crop SEASON A SEASON B Crop water Net irrigation Crop water Net irrigation Crop Area(ha) requirement requirement Area(ha) requirement requirement (mm/season) (mm/season) (mm/season) (mm/season) Tomato 220.5 760.68 613.44 224.53 712.92 446.05 Onion 47 490.79 389.56 54.12 489.76 243.78 Vegetables 8.15 515.89 413.39 7 493.45 242.48 Beans 150 427 330.55 151 424.51 178.94 Maize 408.6 684.37 495.58 408.6 537.53 416.32 Cotton 735 999.13 634.58 Citrus* 1008 1334.31 873.46 Mango* 40.35 1973.22 1512.36 Grape 11.18 1057.43 596.58 wine* Total 2628.78 845.25 * Perennial crops

To calculate the first four indicators grose 2. Output per unit command area production was taken instead of SGVP because of production the fact that there is no price difference in the two = irrigation systems and no common crop is grown in Command cropped area both the farms. Sugarcane is not grown in NuraEra farm but it is the only crop in Wonji sugar estate. Out put per unit command area 1. Output per croppedarea = 73,015,214.08 / 3,069 = 23,791.21 birr/ha production = 3. Output per irrigation sup ply Irrigatedcroppedarea production = Diverted irrigation sup ply = 73,015,214.08 / (2628.78+ 845.25) = 73,015,214.08 / 3474.03 = 73,015,214.08 /97,761,600 = 21,017.44 birr/ha = 0.75 birr/m3

Then, the net crop water requirement and the net irrigation requirement (IR) are computed for each 760.68* (220.5/2628.78) + 490.79* (47/2628.78) + irrigated crop and for each growing season (option 515.89* (8.15/2628.78) + etc. 2 in CROPWAT main menu). The crop coefficients provided with CROPWAT program are used (input: = 1,030.70 mm/season planting dates and growth length in days). The outcomes were: In the same way, the total net irrigation

requirements are computed. The efficiency used is The total net crop demand for season A is (refer 40%. This percent was set basing the estimation of table 4.2): Halcrow(1989) and allowing for the improvements

of the years between the year of study and the year CWR * (area / area ) + CWR * tomato tomato total onion of analysis. (area onion/ area total ) + etc. =

SEASON NCWR* NIR** A(Jan-July) 1,030.70 693.7 B(Aug.-Jan) 560.5 369.3 Total 1591.2 1063 *Net crop water requirement **Net irrigation requirement

Output per unit water consumed Pr oduction = Volume of water consumed by ET

Where Volume of water consumed by ET is the actual evapotranspiration of crops.

= 73,015,214.08 / (1,030.70*2628.78+ 560.5*845.25)*10-3*104

= 2.3 birr/m3

5. Relative water supply = (Irrigation derived + total precipitation) / CWR

Diverted irrigation in mm Season A = 56,782,080 / 2628.78*10 = 2160 mm Season B = 41,247,360 /845.25*10 = 4879.9 Total = 7039.9 RWS = (7039.9+566.9)/1591.2 = 4.8 6. Relative irrigation supply = Irrigation applied / irrigation requirements = 7039.92/1063 = 6.6 The scheme irrigation requirement was calculated with CROPWAT (option 4 in main menu) using the climate data, cropping pattern, planting dates, and area. For 2003/04, the scheme irrigation requirements were:

Scheme irrigation requirement of NuraEra farm

Jan. Feb. March April May June Jul. Aug. Sept. Oct. Nov. Dec. IR in l/s/ha 0.37 0.39 0.69 1.13 1.2 0.77 0.75 0.73 1.47 1.4 1.17 0.59

Peak irrigation requirements occur in September, 1.47 l/s/ha. Peak demand is 1.47 * cropped area for that month =1.47 *2639.78= 3880.48 l/s. As explained above the canal capacity was measured using bank operated cable. The result was 5.4 m3/s

7. Water delivery capacity: 5,400 lit. / 3,880.48 lit. = 1.4 or 140%

8. Gross return on investment =Production / cost of Therefore Gross return on investment irrigation infrastructure = 73,015,214.08 birr/ 80,218,006birr = 91% The cost of the distribution system can either be estimated from original costs, or estimated by using 9. Financial self-sufficiency present costs of similar types of infrastructure = Revenue from irrigation/Total O&M expenditure development. Since it was not possible to get the design documents of the two projects, the document Where, Revenue from irrigation, is the revenue of other projects with more or less similar structural generated, either from fees, or other locally condition were taken. After studying the documents generated income, and Total O&M expenditures are of the similar irrigation projects in Awash Basin, the amount expended locally through O&M plus the Angelele and Bolhamo design document were outside subsidies from the government. used for their better similarity and nearness to NuraEra and Wonji to estimate the cost of Even if the Upper Awash Water Administration distribution system per hectare. Center collects water fee at the rate of 3 birr per 1000 m3 of water, no money was spent for All the delivery system of the NuraEra farm is earth maintenance and operation in the year of analysis. canals .Therefore the cost of earth canals per Rather the farms perform some clearing and hectare in these documents is found to be 1497.9 4 maintenance work whenever they feel necessary. birr Thus the income generated from the farms doesn’t The formula to calculate the present worth is. go to the purpose of operation and maintenance.

Pr esent Net Worth(PNW ) However the intent of IWMI indicator number 9 is to see how much of the cost of operation and

= (Initial cos t / ha) × (1 + r)n maintenance work is generated locally. In this context we can say that both farms are 100 % self

sufficient. It was learnt from the farm record that Where: r is interest rate, which is taken from the the amount of money spent for operation and design document of Angelele Irrigation Project maintenance of the delivery systems for the year and n is years from construction time. 2003/’04 was 260,317.23 birr.

= 260,317.23/260,317.23 PNW = 1497.94(1+0.1) 30 = 100% = 26138.2birr/ha

In this respect, the management of these two farms The cost of delivery system of NuraEra is a little bit different to that of normally practiced. = 26138.2 birr/ha *3069 ha In order to have additional information for = 80,218,006 birr evaluating that part which is related to maintenance

and operation, one more indicator is developed and

added in this paper. This indicator atleast will give applied each furrow is cut off as it reaches the end us additional insight on the relative focus given to of the furrow ,which is blocked ,and ponds up the delivery systems. within the furrow .The furrow length for the part of the field (anjir ) depends on the gradient available Percent allocated to O & M and three lengths 32m,48m,and 64m are being used currently. The advantages of this system are that Total O & M exp enditure = there is no run off and the entire water applied to Pr oduction the field incorporated into the soil. As such, in addition to the infiltration characteristics of the soil, = 260,317.23/ 73,015,214.08 the size of the inflow stream and the gradient of the = 0.36 % furrow become important variables to control the rate of advance of water front to the end of the Wonji Sugar estate furrow, which determines the cutoff time for the inflow. (Mukherij, 2000) The only crop grown in this farm is sugarcane. The kind of irrigation practiced at Wonji may be termed ‘’blocked end furrow irrigation system’’. Water Production and area of Wonji estate for the year 2003/04 Command Irrigated Farm gate area(ha) area(ha) Production(qtl.) price(birr/qtl.) Income (birr) 5,929 4,110 7,446,581 11.08 82,508,117.5

The main pump station at Wonji has a total of eight = Out put / command area pumps with a design discharge of = 82,508,117.48/5929 (6*750+2*500)5500 lps .According to Mukherji the = 13,916.03 birr/ha two vertical shaft pumps are working with overall efficiency of 68% and the other six horizontal shaft 3. Out put per irrigation supply pumps with an efficiency of 50% of their designed = Out put / irrigation diverted capacity. = 82,508,117.48/75320000 = 1.1 birr/m3 Crop water Irrigation requirement requirement 4. Out put per unit water consumed mm/year mm/year =out put/Volume water consumed by ET 1,667.1 1,054.33

The result of our measurement almost agrees with that of Mukherji, The current capacity of the two vertical shaft pumps is 240.3 lps and the horizontal Volume of water consumed by ET = 1667.1*4110*10 m3 shaft pumps are with a capacity of 488 lps. 3 Therefore the maximum discharge that can be = 68,517,810 m pumped for irrigation purpose is 3,165.34 lps allowing 10% for factory use. Out put per unit of water consumed = 82,508,117.48/68,517,810 1. Out put per cropped area 3 = Out put / irrigated cropped area = 1.2 birr/m = 82,508,117.48 / 4,110 = 20,074.97 birr/ha 5. Relative water supply = (Irrigation derived + total precipitation) / 2. Out put per command area CWR = (1,466.08 + 817.3)/1667.1

= 1.36 = 1,466.08/1,054.33 = 1.4 6. Relative irrigation supply = Irrigation applied/ irrigation requirements

For 2003/04, the scheme irrigation requirements were: IR in Jan. Feb. March April May June July Aug. Sept. Oct. Nov. Dec. l/s/ha 0.4 0.58 0.94 0.93 0.87 0.65 0.31 0.23 0.73 1 0.94 0.85

Peak irrigation requirements occur in October, 1 l/s/ha. Peak demand is 1 * cropped area for that month = 1 * 4110 = 4110 l/s.

7. Water delivery capacity: 3,165.34/4110 = 0.77 or 77% Comparative analysis of the two irrigation systems 8. The same procedure was applied to Wonji to calculate gross return on investment except that The four basic comparative performance indicators irrigation structure costs are included since all relate output to unit land and water. These the necessary structures are found in Wonji “external” indicators provide the basis for estate farm. The present initial cost per hectare comparison of irrigated agriculture performance of the delivery system is found to be 26137.5 (Molden, 1998). birr. The values of the four indicators for the respective Gross return on investment farms are as tabulated below. = 82,508,117.5 birr / 235,727,927.5 birr = 35%

9. Percent allocated to O&M = 139,330.46/ 82,508,117.48 = 0.17%

summary of the four indicators Output per Output per Output per Output per water Site cropped area, irrigation command, birr/ha consumed birr/m3 birr /ha supply,birr/m3 NuraEra 21,017.44 23,791.21 0.75 2.3 Wonji 20,074.97 13,916.03 1.4 1.2

output per cropped and command area Output per irrigation supply and water consummed 25,000.00 2.5

20,000.00 2

15,000.00 1.5 10,000.00 1 5,000.00 0.5 0.00 Output per cropped Output per command, 0 area, birr/ha birr/ha NuraEra Wonji Output per irrigation Output per w ater supply,birr/m3 consumed birr/m3 88

As clearly seen in the first table and chart, NuraEra achieved the required performance. NuraEra show a is higher in output per cropped area and output per generous supply of water. Using the rainfall in a command. This is attributed to the high value and better way than that of the NuraEra contributes for export crops grown in the farm and also the higher this higher value of output per irrigation in Wonji. cropping intensity. In other words NuraEra is better in land productivity and Wonji show better performance in In the case of output per irrigation supply Wonji water productivity. has the higher value (1.4) implying that Wonji is better in effective utilization of water. But as the indicators are comparative we can’t say that Wonji

Result of RWS and RIS Relative water Relative irrigation Water delivery Site supply supply capacity NuraEra 4.8 6.6 1.4 Wonji 1.36 1.4 0.77

RWS,RIS and WDC 7

6 5

4 NuraEra 3 Wonji 2

1 0 RWS RIS WDC

Higher values of RWS and RIS indicate a satisfies the operation and maintenance needs. more generous supply of water (Molden, Rather as Molden et al. (1989) explained 1998). In our case, NuraEra show this fact financial self-sufficiency tells us what percent which means productivity to land is more of expenditures on O&M is generated locally. important. In the Wonji farm we have more If government subsidizes O&M heavily, constrained supply of water so that financial self-sufficient would be low, whereas productivity per unit of water is more if local farmers through their fees pay for most important. These two values can be minimized of the O&M expenditures, financial self- by adjusting the release of water from storage sufficiency would be high. or diversion with the available rainfall. NuraEra and Wonji are 100% self sufficient The lower values of RWS and RIS of Wonji meaning all the expenditures for operation and estate farm is not mainly the result of good maintenance generated locally from the management rather the decreasing capacity of income of these two farms. Interms of percent the pumps limiting the supply. As the result of allocated to operation and maintenance NurEra the measurement conducted showed, 35 and 52 is better. This is because of high silt deposit percent decrease in the capacity of the rate in the delivery systems. With out frequent horizontal and vertical pumps has been clearing of the system normal operation of the witnessed respectively. With this capacity, farm can’t be carried out. the pumps are not enough to satisfy peak irrigation demand. This will get worse unless a GRI of the NuraEra and Wonji are 84.7% and major rehabilitation work of the pump house is 76.8% respectively. NuraEra show a higher done. rate of return. This is mainly because of the lower investment cost and higher productivity On the other hand high value of RWS and RIS of the farm. But it is hard to say that this is of for NuraEra. This result was expected by system is better than that of Wonji because it is the researcher at the end of reconnaissance taking much money for its maintenance. survey of the area because of two main Consequently its sustainability is in big reasons: one the 34 km main canal is not well question of such systems maintained and the absence of the necessary irrigation structures in the line of the canal expose the system for very low efficiency, Conclusion second the farmers around the state farm steal water from the main canal. Consequently much The higher values in RWS and RIS combined water is lost through its way to the field. with the lower ratio of output per water consumed in NuraEra shows that the availability of water is not a problem. In addition it can be inferred that the issue of water management is in jeopardy. The main reasons for this problem are the meager attention given by the farm for irrigation water management, the low water rate and the effect of unregistered users.

As to Wonji farm the low value of WDC can be early warning indicator. This is meant to Unauthorized diversion from the main canal say that the pump capacity is deteriorating and at NuraEra is not in position to satisfy peak irrigation demand. This problem may manifest itself by A high value of financial self-sufficiency does reducing production in near future. This has not automatically indicate a sustainable system big implication on the production of sugar. as the O&M expenditures might be too low to WDC of NuraEra farm seems in a better meet the actual maintenance needs (Molden, position but the increasing sedimentation and 1998).Thus the high value of FSS of the two the absence of silt excluders put its farms doesn’t show that the expenditure sustainability in question.

Both farms spent less than one percent of their Guideline: Irrigation and Drainage Paper. income for operation and maintenance of No.58. FAO, Rome. delivery systems. This by itself shows the very Ethiopian Agricultural Research Organization low attention given to the irrigation systems. (2001). Agricultural water management: But in relative terms NuraEra was better. Brief Profile. (Unpublished document) Food and Agriculture Organization The rate of gross return on investment is high (1996).CROPWAT and CLIMWAT for both of the farms. software packages: Irrigation and drainage paper No 49.FAO, Rome. Furthermore, as explained in the literature Food and Agriculture Organization review part, there are varies types of (2002).Crops and Drops. FAO working institutional designs of the irrigation systems paper, Rome. adopted in different part of the world. This part Food and Agriculture Organization (1989). has not been given the proper attention in our Guidelines for designing and evaluating country. As a result, the irrigation delivery surface irrigation systems. FAO Irrigation systems are left unattended for many years. and drainage paper no 45.Rome. The case of the two irrigation systems is not an Food and Agriculture (1986).Organization, exception. The management of the farms, operation and maintenance of irrigation having their main objective on productivity, schemes .FAO irrigation and drainage didn’t give the proper attention for their paper 40,Rome. irrigation delivery systems. This can be seen Food and Agriculture (1999). Modern water by the fact that the two farms have no separate control and management practices in irrigation department despite being very irrigation. Water report 19, Rome profitable ones. Moreover no permanent Food and Agriculture (1998). Crop budget is allocated for irrigation system evapotrinspiration-guidelines for maintenance. computing crop water requirements. FAO irrigation and drainage paper 56,Rome. This research managed to show some of the Halcrow and Partners(1989).Master plan for problems that our large scale irrigation systems the development of surface water resources are in. The lack of awareness in the farm in the Awash basin. Government of officials on one side and policy problem in the Ethiopia, Addiss Ababa. other left the irrigation systems with lots of Hunt, Robert C. 1988. Size and the problems. Water logging, deterioration of structure of authority in canal irrigation canal capacity and salinity are coming faster. systems. Journal of Anthropological Research 44:335–355. References Majumdar,D.K.(2002).Irrigation water management .Prentice-Hall of India.New Admasu, (2002). Integrated water Delhi. management and land management Manz, David, H. (1992). Performance research and capacity building priorities for evaluation of irrigation conveyance Ethiopia. Proceedings of systems using dynamic simulation models. MoWR/EARO/IWMI. International Water Canada. Management institute. Addis Ababa. Molden,D.J,Sakthivadivel,R,perry,C.J,andCha Clarke D. 1998. CropWat for Windows: User rlotte de Fraiture(1998). Indicators for Guide. Version 4.2. University of comparing performance of irrigated Southampton, UK. Agricultural systems. Research report 20. Clyma, W., Latimon, D.L and Reddy, M.N. International Water Management Institute. (1982). Irrigation water management Sri Lanka. Mrema, G. C. (1984). around the world. Paper prepared for Development of small-holder irrigation in presentation at the ninth technical Tanzania: Problems and prospects. In conference on irrigation, drainage and Proceedings of African Regional flood control. Jackson, Mississippi. Symposium on Small Holder Irrigation, 5– Douglas L. V., Juan A. S. (1999). Transfer of 7 September 1984, Harare, Zimbabwe. Irrigation Management Services.

Mukherji, Jp. (2000). Rehablitation, Optimization, and expansion of Agriculture and Factory. Interim report. Vol. 1, Ethiopian Sugar Industry Sh.Co., Wonji- Shoa sugare Factory Mrray-Rust, D. H., and W. B. Snellen. (1993). Irrigation system performance assessment and diagnosis. (Jointly published by IIMI, ILRI, IHE), Colombo, Sri Lanka:

International Irrigation Management

Institute.

Oad, R., Mcorrik, P.G, and Clyma, W.

(1988).Methodologies for interdisciplinary

diagnosis of irrigation systems. Project

report 93.Colorado state university,

Colorado.

Pery,C.,J.(2001). Charging for irrigation

water: The issue and the options, with a

case study from Iran. Research report

52.International water management

institute.Sri Lanka.

Shanan, Leslie. (1992). Planning and

management of irrigation systems in

developing countries. Agricultural Water

Management. The Netherlands: Elsevier.

Sharma, D.N, Ramchard Oad, and Sampath,

R.k.(1991).Performance measures for

irrigation water delivery systems.ICID

Bulletine 40:21-37.

Small, L. E., and M. Svendsen (1992). A framework for assessing irrigation performance. International Food Policy Research Institute Working Papers on Irrigation Performance No. 1. International Food Policy Research Institute. Washington, D. C. Withers, B.; and Vipond S. (1974). Design and practices. London: BT Batsford

Limited.

Woodroofe,R. (1993). Monitoring and

evaluation of irrigation projects.

Technical manual.Ministry of water

resources. Addis Ababa

WORLD BANK/UNDP (1990). A proposal

for an internationally –supported program

to enhance research on irrigation and

drainage technology in developing

countries. Irrigation and drainage volume1

Yesuf, K. (2004). Assessment of small-scale irrigation using comparative performance indicators on two selected schemes in upper Awash river valley. A thesis submitted to the school of graduate studies. Alemaya University.

Discussion on Theme 1: Status Quo Analysis, Characterization, and Assessment of Performance of Irrigation in Ethiopia

Chair person: Dr. Deboorah Bossio Rapporture: Makonnen Loulseged

The chairman for this session introduced the theme and the floor was opened for questions, comments and suggestions.

Questions and Discussions irrigation development should be undertaken in an integrated manner in Marginal environments are means of living order to ensure its sustainability. for people who are habitats of the area. Unless there are other means of income we can not Why the presenter did not see with respect avoid (stop them) growing crops on marginal to integrated watershed management. lands. The environment has not been benefited from technologies promoted under optimum Ans: Irrigation schemes are operating conditions. Then, how can we say agriculture in within the environment of complex marginal environment causes poverty? watershed activities. Sedimentation is the cause of failure in many schemes as Ans: Marginal lands with steep slopes a result of poor watershed management. are being farmed, on the other hand Therefore, due attention should be given there is plenty of irrigation potential in to sound watershed management low land areas where communities can practices. make use of it. Do you have bench marks from other The data base is organized for modern countries for your comparison of performances irrigation schemes. Do we have a plan to set up a of irrigation schemes in Ethiopia? data base for traditional irrigation schemes in Ethiopia? Please give attention in collecting GIS Ans: As the conditions of irrigation data for the traditional irrigation systems. schemes are different it will be un fare to compare their performance with other Ans:IWMI is willing and interested to countries; for example a scheme in collaborate with other institutions in Ethiopia bench marked with a scheme setting up the database for traditional located somewhere in US. Instead we irrigation. can bench mark better performing schemes with in the country. For When we talk about challenges in irrigation example Methara under the same schemes, the impact of land degradation is not environment with other scheme located captured in the study? Why irrigation projects somewhere in Ethiopia only focuses on the dam and command area but not the upper catchment; so also the research The scope of irrigation development is results? beyond putting physical assets on the ground. This is about 25% of the tasks. The remaining Ans: There are many intrinsic problems 75% covers a complex and wide range of tasks associated with irrigation and drainage. that we need to put in place and achieves In SNNP alone there are about production and income objectives. The seventeen schemes that have failed. In experience in the last 2-3 decades shows line with the policy statement, future irrigation schemes are not fully or properly

utilized due to inadequate attention being given management is the main cause for low water to skill in O&M, skill in irrigation water and productivity of rainfed agriculture in Ethiopia. irrigation extension, markets, credit services, The contribution of rainfall variability and poor input supply, institutional arrangement in farming practices to the low productivity of managing schemes, catchment treatment, etc. rainfed agriculture should be studied These issues were discussed in the past. comparatively such study would have paramount However, policy makers and practitioners are contribution towards water allocation among not responsive to these complex issues. Now competing water users. thanks to researchers (IWMI) and others that we have good evidence to feed into an informed Ans: Climate variability has a seasonal decision to focus on the issues beyond impact on the agricultural production construction. Future success in irrigation depending on the timing and amount of development will depends on willingness to rainfall in each season. Poor farming draw lessons from our experiences and practice has also a negative impact on proactively engage in rectifying the issues agricultural production. However their mentioned above. relative impact has not been studied in this project as it is beyond the scope of Ans: Acceptable suggestion this study.

Attempt has been made to compare Melka Why factors that are responsible for the Sedi irrigation scheme with Sile and perhaps performance at scheme level were not with others. Because of differences in Climatic identified? What was the purpose of comparing factors (rainfall, etc), soil salinity, age of the number of irrigation schemes unless these scheme. How about issues of management? factors are identified?

Ans: The performance of each scheme Ans: Factors that have contributed to is identified in terms of its technical low performances of community performance. However, the evaluation managed irrigation scheme (Hare) and and establishment of indicators on the state farm (Sille) are mainly internal management process at scheme management and deterioration of level has been found difficult. As a physical infrastructures. It is however, result comparison between schemes has hardly possible to identify the details of not been easy. indicators of external performance indicators such as output and water Rainfed agriculture has low production and supply at scheme level. Instead we need productivity due to erratic nature of rainfall in to consider on farm process indicators Ethiopia. But some downstream water users (accepted international indicators). argue that poor farming practice and

Theme two:

Irrigation Impact on Poverty and Economy

Papers 1. Rural poverty in Ethiopia: Does small scale irrigation make a difference 2. An assessment of the financial viability and income impact of SSI in Ethiopia 3. Importance of irrigated agriculture to the Ethiopian Economy: Capturing the direct net benefits of irrigation 4. Investment in irrigation as a poverty reduction strategy: The case of SSI and its impact on poverty in Tigray, Ethiopia

Rural poverty and inequality in Ethiopia: does access to small-scale irrigation make a difference?

Regassa E. Namara, Godswill Makombe, Fitsum Hagos, Seleshi B. Awulachew International Water Management Institute [email protected]

Abstract Key words: Rural poverty, FGT indices, Small scale irrigation Ethiopia is an agrarian society in a land of drought and floods. Agricultural production, 1. Introduction which is the source of livelihood for eight out of ten Ethiopians, is extremely vulnerable to Farmers in rural Ethiopia live in a shock-prone climatic conditions. The causes of rural poverty environment. Agricultural production, which is are many including wide fluctuations in the source of livelihood for eight out of ten agricultural production as a result of drought, Ethiopians, is extremely vulnerable to climatic ineffective and inefficient agricultural marketing conditions. The causes of rural poverty are many system, under developed transport and including wide fluctuations in agricultural communication networks, underdeveloped production as a result of drought, ineffective and production technologies, limited access of rural inefficient agricultural marketing system, under households to support services, environmental developed transport and communication degradation and lack of participation by rural networks, underdeveloped production poor people in decisions that affect their technologies, limited access of rural households livelihoods. However, the persistent fluctuation to support services, environmental degradation in the amount and distribution of rainfall is and lack of participation by rural poor people in considered as a major factor in rural poverty. decisions that affect their livelihoods. However, Cognizant of this reality the successive the persistent fluctuation in the amount and Ethiopian governments and farmers have made distribution of rainfall is considered as a major investments in small scale irrigation schemes. factor in rural poverty. Small-scale farmers are This paper aims to assess the efficacy of these the largest group of poor people in Ethiopia. investments in reducing poverty based on data Their average land holdings are smaller, their obtained from a survey of 1024 farmers drawn productivity is low and they are vulnerable to from four major regional states of Ethiopia. The drought and other adverse natural conditions. Foster, Greer and Thorbecke poverty measures Cognizant of this reality the successive were used to compare the incidence, depth and Ethiopian governments and farmers have made severity of poverty among groups of farmers investments in small scale irrigation schemes. defined by relevant policy variables including Despite efforts to reduce poverty in the country access to irrigation. In order to explore the over the past decade, farmers, herders and other correlates of rural poverty and their quantitative rural people remain poor. Poor people in rural significance, logistic regression model was areas face an acute lack of basic social and estimated. The main conclusion of the study is economic infrastructure such as health and that the incidence, depth and severity of poverty educational facilities, veterinary services and is affected more by the intensity of irrigation use access to safe drinking water. Households (as measured by the size of irrigated area) than headed by women are particularly vulnerable. mere access to irrigation. Alternatively, there Women are much less likely than men to receive seems to be an economy of scale in the poverty- an education or health benefits, or to have a irrigation relationship. voice in decisions affecting their lives. For them,

poverty means high numbers of infant deaths, meaning of poverty. There is great variation in undernourished families, lack of education for the manner in which poverty is being defined children and other deprivations (IFAD). and measured in developing countries (May,2001). Poverty is a persistent feature of The impact of drought on the overall macro- socioeconomic stratification through out the economy of Ethiopia is very significant. There world. Over the last twenty five years the is very strong correlation between hydrology understanding of poverty has advanced and and Ethiopia’s GDP performance. It is widely become more holistic. From having been accepted that the Ethiopian economy is taken understood almost exclusively as inadequacy of hostage to hydrology due to the so far income, consumption and wealth, multiple insignificant infrastructural development in the dimensions of poverty and their complex water sector (World Bank, 2006). Oftentimes, interactions are now widely recognized. These Ethiopia is ravaged by droughts, leading to include isolation, deprivation of political and dramatic slow downs in economic growth. The social rights, a lack of empowerment to make or development of water storage facilities which influence choices, inadequate assets, poor health could be used, among other things, to develop and mobility, poor access to services and irrigation is seen as a way of reducing Ethiopia’s infrastructure, and vulnerability to livelihood dependence on the annual availability of rainfall failure. (UNPD, 2006; World Bank, 2006). Often distinction is made between absolute and In Ethiopia, the persistent correlation between relative poverty. Relative poverty measures the rainfall and GDP growth is striking and extent to which a household’s income falls troubling. The effects of hydrological variability below an average income threshold for the emanate from the direct impacts of rainfall on economy. Absolute poverty measures the the landscape, agricultural output, water- number of people below a certain income intensive industry and power production. These threshold or unable to afford certain basic goods impacts are transmitted through input, price and and services. Absolute poverty is a state in income effects onto the broader economy, and which one’s very survival is threatened by lack are exacerbated by an almost complete lack of of resources. Consideration is also necessary of hydraulic infrastructure to mitigate variability the dynamics of both chronic1 and transient and market infrastructure that could mitigate poverty, and of the processes which lead people economic impacts by facilitating trade between to escape from or fall into and remain trapped in deficit and surplus regions of the country. poverty (Carter et al. 2007). Another related concept is equity, which is usually understood Evidences from elsewhere indicate that initial as the degree of equality in the living investments in water resources management and conditions of people, particularly in income multipurpose hydraulic infrastructure had and wealth, that a society deems desirable or massive regional impacts with very large tolerable. Thus equity is broader than poverty multiplier effects on the economy. Therefore, it and is defined over the whole distribution, not is possible that irrigation investments in Ethiopia only below a certain poverty line. The meaning may have contributed to poverty reductions of equity encapsulates ethical concepts and among other people than the irrigators, who are statistical dispersion, and encompasses both direct beneficiaries of the investment. However, relative and absolute poverty. Hence, ideally any in this paper we limit ourselves to the poverty impact of small-scale irrigation development on 1 the direct beneficiaries or farmers. Chronic poverty is an individual experience of deprivation that lasts for a long period of time. In this Definition of concepts sense the chronic poor are those with per capita income or consumption levels persistently below the poverty line during a long period of time. Transient Before addressing the rural poverty and poverty is associated with a fluctuation of income irrigation nexus, it is important to clarify the around the poverty line.

assessment of how irrigation can affect poverty significance in predicting poverty. Section fives must consider impacts on these varied gives some policy conclusions and implications. dimensions of poverty and their interactions. For example, it must consider whether changes are 2. Methodological Issues in absolute or relative terms, and whether they are long lasting or transient. Similarly, it must Data sources encompass the other dimensions of poverty beyond income, consumption and wealth. In This study is part of a comprehensive study on order to understand the dynamics of poverty, the impacts of irrigation on poverty and one can draw on the notions of ‘capabilities’ and environment run between 2004 and 2007 in ‘entitlements’ that have received a good deal of Ethiopia implemented by the International Water attention (Sen 2000). Sen’s work belies the idea Management Institute (IWMI) with support from that income shortfalls are the main attribute of the Austrian government. The socio-economic poverty. He emphasises the importance of the survey data on which this paper is based is bundle of assets or endowments held by the gathered from a total of 1024 households from poor, as well as the nature of the claims attached eight irrigation sites in 4 Regional states to them, as critical for analyzing poverty and involving traditional, modern and rain fed vulnerability. systems. The total sample constitutes 397 households practicing purely rainfed agriculture Nevertheless, while recognizing that poverty is a and 627 households (382 modern and 245 multidimensional phenomenon consisting of traditional) practice irrigated agriculture. These material, mental, political, communal and other households operate a total of 4953 plots (a aspects, the material dimensions of poverty household operating five plots on average). Of expressed in monetary values is too important an the total 4953 plots covered by the survey, 25 aspect of poverty to be neglected (Lipton 1997). percent (1,250 plots) are under traditional Given the fact that there is ‘a lack of consensus irrigation, 43 percent (2,137 plots) are under regarding the measurement of other forms of modern while the remaining 32 percent (1,566 deprivation’, the approach followed in this paper plots) are under rainfed agriculture. The data is ultimately grounded on the notion of some was collected for the 2005/2006 cropping minimum threshold below which the poor are season. categorized (Lipton 1997). There is growing recognition that poverty may adequately be Poverty indices defined as private consumption that falls below some absolute poverty line. This is best When estimating poverty using monetary measured by calculating the proportion of the measures, one may have a choice between using population who fall below a poverty line (the income or consumption as the indicator of well- headcount) and the extent of shortfall between being. Most analysts argue that, provided the actual income level and poverty line (the depth information on consumption obtained from a or severity of poverty). The poverty line is household survey is detailed enough, usually based on an estimated minimum dietary consumption will be a better indicator of poverty energy intake, or an amount required for measurement than income for many reasons purchasing a minimum consumption bundle. (Coudouel et al. 2002). One should not be dogmatic, however, about using consumption This paper analyses the state of poverty and data for poverty measurement. The use of inequality among sample farm households with income as a poverty measurement may have its and without access to irrigation. It also analyses own advantages. In this paper we estimate the correlates of poverty. Section two presents poverty using income adjusted for differences in the data collection and analytical methods. household characteristics. Section three shows the results of poverty As for the poverty measures, we will be profiling, while section four assesses the concerned with those in the Foster-Greer- determinants of poverty and their quantitative Thorbecke (FGT) class. The FGT class of

poverty measures have some desirable properties (such as additive decomposibility), and they At α = 2 equation 1 measures poverty severity include some widely used poverty measures or squared poverty gap. This takes into account (such as the head-count and the poverty gap not only the distance separating the poor from measures). Following Duclos et al. (2006), the the poverty line (the poverty gap), but also the FGT poverty measures are defined as inequality among the poor. That is, a higher weight is placed on those households further away from the poverty line. α 1 ⎛ g ()p ; z ⎞ P ()z; α = ∫ ⎜ ⎟ dp (1) We calculated these indices using DAD4.4 0 ⎝ z ⎠ (Duclos, J-Y et al., 2006)

Inequality indices where z denotes the poverty line, and α is a To assess the income inequality among the nonnegative parameter indicating the degree of different farm household groups, we calculate sensitivity of the poverty measure to inequality the Gini coefficient of inequality and Decile among the poor. It is usually referred to as ratios. Gini coefficient of inequality is the poverty aversion parameter. Higher values of the most commonly used measure of inequality. The parameter indicate greater sensitivity of the coefficient varies between 0, which reflects poverty measure to inequality among the poor. complete equality, and 1, which indicates The relevant values of α are 0, 1 and 2. complete inequality (one person has all the income all others have none). The decile At α =0 equation 1 measures poverty incidence dispersion ratio presents the ratio of the average or poverty head count ratio. This is the share of consumption or income of the richest 10 percent the population whose income or consumption is of the population divided by the average income below the poverty line, that is, the share of the of the bottom 10 percent. This ratio is readily population that cannot afford to buy a basic interpretable by expressing the income of the basket of goods. rich as multiples of that of the poor.

At α =1 equation 1 measures depth of poverty In summary the analysis of poverty and (poverty gap). This provides information inequality followed four steps. First, we have regarding how far off households are from the chosen household income as a welfare measure poverty line. This measure captures the mean and this was adjusted for the size and aggregate income or consumption shortfall composition of the household. Second, a poverty relative to the poverty line across the whole line is set at 1075 Birr (1USD=9.07Birr), a level population. It is obtained by adding up all the of welfare corresponding to some minimum shortfalls of the poor (assuming that the non- acceptable standard of living in Ethiopia poor have a shortfall of zero) and dividing the (reference). The poverty line acts as a threshold, total by the population. In other words, it with households falling below the poverty line estimates the total resources needed to bring all considered poor and those above the poverty line the poor to the level of the poverty line (divided considered non-poor. Third, after the poor has by the number of individuals in the population). been identified, poverty measures such as Note also that, the poverty gap can be used as a poverty gap and squared poverty gap were measure of the minimum amount of resources estimated. Fourth, we constructed poverty necessary to eradicate poverty, that is, the profiles showing how poverty varies over amount that one would have to transfer to the population subgroups (example irrigators Vs poor under perfect targeting (that is, each poor non-irrigators) or by characteristics of the person getting exactly the amount he/she needs household (for example, level of education, age, to be lifted out of poverty) to bring them all out etc.). The poverty profiling is particularly of poverty (Coudouel et al. 2002). important as what matters most to many

100

policymakers is not so much the precise location income for the former is almost doubles that of the poverty line, but the implied poverty of the latter group. The income difference comparison across subgroups or across time. widens with the size of irrigated area. Lastly, we analyzed income inequality among Comparison of the per capita income for the sample households. lower 10% of income distribution for irrigators and non-irrigators shows that the per 3. Household income distribution capita income for the irrigators is three times that of non-irrigators. This difference is also The income distribution differentiated by access influenced by the size of cultivated area to irrigation and irrigation use intensity is shown The gap in mean per capita income between in table 1. A close scrutiny of the table shows poor and non-poor households is substantial the following interesting results: irrespective of access to irrigation. Even The mean per capita income of rainfed though the mean per capita income of poor farmers is below the poverty line. people with access to irrigation is higher than Interestingly also the mean per capita income that of the poor without access to irrigation, values up to the eighth income decile is lower the difference seems to be insignificant. than the assumed poverty line. However, the The Gini index of income inequality values mean per capita income for irrigators and the suggests that income inequality is higher overall sample is higher than the poverty line among households with access to irrigation as and the gap between mean per capita income compared to those with no access. The values and poverty line widens in proportion to the for the decile ratios also indicate that income size of irrigated area. inequality is lower among the rain-fed Comparison of the mean per capita income farmers. for the richest 10% of irrigators and non- irrigators shows that the mean per capita

Table 1. Distribution of per capita income by income deciles for irrigators and non-irrigators Deciles Rainfed Irrigators Overall 1st 2nd 3rd 4th quartile a quartile quartile quartile First 38.5 114.5 72.6 90.8 80.4 116.5 233.4 Second 166.8 331.0 236.4 274.6 242.6 362.9 520.0 Third 285.6 503.0 391.0 385.5 466.7 538.9 708.2 Fourth 401.6 648.2 526.8 509.3 584.3 658.8 960.2 Fifth 514.2 850.9 651.5 673.4 813.8 864.0 1268.3 Sixth 617.0 1127.9 842.0 827.0 1035.8 1270.1 1641.1 Seventh 774.2 1507.0 1099.5 1112.8 1245.2 1766.1 2295.2 Eighth 984.5 2067.9 1542.5 1481.8 1729.0 2506.7 3294.9 Ninth 1379.2 3231.7 2425.7 2033.6 2374.8 3889.2 4796.6 Tenth 4152.5 8736.3 7096.5 6395.6 7447.2 9352.3 10212.0 Mean 930.7 1908.3 1487.3 1369.6 1613.7 2230.9 2492.7 Poverty line 1075 1075 1075 1075 1075 1075 1075 poor 486.2 498.5 492.8 503.4 527.0 525.2 602.9 non-poor 2688.4 3497.1 3290.5 2980.4 3123.2 3998.4 3718.5 % poor 77.1 58.5 65.7 66.3 58.9 53.9 41.8 Gini coefficient 0.499 0.546 0.547 0.507 0.515 0.537 0.503 Deciles ratio 11.6 26.9 20.7 14.8 20.1 22.6 16.4

Given the assumed poverty line, the proportion reduction impact of access to irrigation is very of poor households among those households much influenced by the size or irrigated area. with no access to irrigation is higher than those The relationship between poverty and irrigation who have access to irrigation. The poverty

101

and other relevant factors will be analyzed in 32.2% to lift their per capita income level to the more detail in the succeeding sections. poverty line or alternatively to move them out of absolute poverty, while the income of rain-fed 4. Poverty profile poor farmers should be increased by 42.5% to lift them out of poverty. The higher poverty Rural poverty and irrigation severity value for rain-fed poor farmers also indicates that inequality among the poor rainfed Table 2 shows the incidence, depth and severity farmers is higher when compared to irrigating of poverty by access to irrigation, irrigation poor farmers. Similar interpretations hold for typology, and extent of irrigated area owned by tables 3 through 6 as well. those who have access to irrigation. As expected the poverty incidence, depth and severity values However, note that the incidence of poverty are lower for farmers that have access to among the sample households is still higher irrigation. While the interpretation of the irrespective access to irrigation. When incidence values is straight forward (i.e., it comparing irrigation scheme types, the poverty indicates the proportion of poor people in the situation is worse among irrigators belonging to sample), that of the depth and severity is not. traditional scheme. Poverty indices are also The depth of poverty for irrigators is about 0.322 responsive to the size of irrigated area. Poverty as compared to 0.425 for those without access to incidence for households belonging to the first irrigation. The interpretation is that the per quartile of irrigated area is about 65.8%, which capita income of farmers with access to decreases to 40.3% for those in the fourth irrigation needed to be increased on average by quartile.

Table 2.The effect of irrigation on incidence, depth and severity of poverty Variables Incidence (α = 0 ) Depth (α = 1) Severity (α = 2 ) value SD Value SD Value SD Access to irrigation Irrigators 0.585 0.0197 0.322 0.0140 0.226 0.0125 Non-irrigators 0.771 0.0211 0.425 0.0161 0.283 0.0144 Irrigation Scheme Type Traditional Schemes 0.661 0.0303 0.404 0.0234 0.297 0.0216 Modern Schemes 0.537 0.0255 0.270 0.0169 0.181 0.0148 Size of Irrigation area No irrigation 0.792 0.0191 0.466 0.0160 0.333 0.0154 1st quartile (0.66) 0.658 0.0374 0.351 0.0259 0.230 0.0220 2nd quartile (1.87) 0.586 0.0436 0.299 0.0298 0.203 0.0254 3rd quartile (3.56) 0.524 0.0390 0.268 0.0246 0.171 0.0209 4th quartile (7.92) 0.403 0.0450 0.177 0.0246 0.104 0.0181

It is true that the exact magnitude of the of poverty line due to several reasons (Coudouel calculated poverty incidence, depth and severity et al.2002). To avoid the potential bias that values is influenced by the level of the chosen might be created due to the use of in appropriate poverty line. This is particularly true when one poverty line, we have plotted a graph depicting considers the fact that the different regions of the relationship between all the realized income Ethiopia are expected to differ in the magnitude per capita and the corresponding poverty incidence values1 and the results are shown in Figure 1 and 2.

Figure 1 shows that baring the results for the extreme low values of income per capita, at all of the realized per capita income (plausible poverty lines), the poverty incidence is higher among farmers with no access to irrigation. The vertical line indicates the assumed poverty line (1075 Birr).

102

Figure 1. Poverty incidence curves for irrigators and non-irrigators under different poverty line assumptions.

1. We have used DAD 4.4 to generate these curves.

Figure 2 Shows poverty incidence for different irrigated area categories. The figure indicates that poverty incidence is very responsive to the size of irrigated area.

103

Ethiopia due to population pressure. Any farther Poverty, farm size and livestock holding expansion is possible only in fragile lands or The effect of farm size and livestock holding on important natural resources enclaves. The the incidence, depth and severity of poverty is relationship between livestock holding and shown in table 3. The incidence depth and poverty is generally as expected: poverty severity of poverty among farmers in the higher incidence is lower among farmers with highest farm size category is significantly lower. livestock holding. However, it should be noted that the room for expanding farm size is limited in most parts of

Table 3. The effect of farm size and livestock holding on poverty incidence, depth and severity Variables Incidence (α = 0 ) Depth (α = 1) Severity (α = 2 ) Value SD Value SD Value SD Farm Size 1st quartile 0.789 0.0249 0.524 0.0216 0.400 0.0211 2nd quartile 0.700 0.0288 0.360 0.0204 0.235 0.0181 3rd quartile 0.600 0.0313 0.291 0.0201 0.183 0.0164 4th quartile 0.531 0.0312 0.260 0.0194 0.163 0.0157 Livestock holding 1st quartile 0.657 0.0230 0.407 0.0231 0.299 0.0217 2nd quartile 0.669 0.0295 0.383 0.0212 0.260 0.0182 3rd quartile 0.654 0.0299 0.353 0.0205 0.231 0.0172 4th quartile 0.607 0.0308 0.272 0.0190 0.164 0.0155

104

significantly lower among farmers whose Poverty and cropping pattern substantial proportion of cultivated area is devoted to vegetables and root crops. This Table 4 depicts the influence of cropping pattern suggests that poverty among smallholders can be on poverty indices. It is interesting to note that reduced through diversifying crop production by as the proportion of cultivated area devoted to including high value crops such as vegetables. cereals increases the value of the FGT poverty However, it is also important to note that most indices increases. This is particularly important of the farmers who grow vegetables and root because most of the sample farmers grow low crops had access to irrigation. value staple cereal crops. On the other hand, the incidence, severity and depth of poverty is

Table 4. The effect of cropping pattern on poverty incidence, depth and severity Variables Incidence (α = 0 ) Depth (α = 1) Severity (α = 2 ) Value SD Value SD Value SD Crop area shares: cereals 0.0 - 0.25 0.575 0.0319 0.385 0.0257 0.307 0.0239 0.25-0.50 0.630 0.0290 0.334 0.0196 0.218 0.0170 0.50-0.75 0.641 0.0303 0.290 0.0190 0.175 0.0152 0.75-1.0 0.780 0.0259 0.441 0.0203 0.299 0.0185 Crop area shares: vegetables No vegetables 0.766 0.0158 0.440 0.0126 0.308 0.0117 0.0 - 0.25 0.455 0.0399 0.178 0.0195 0.091 0.0130 0.25-0.50 0.368 0.0495 0.179 0.0313 0.125 0.0291 0.50-0.75 0.263 0.1011 0.096 0.0528 0.062 0.0440 0.75-1.0 0.258 0.0786 0.181 0.0603 0.145 0.0537 Crop area share: root crops No root crops 0.661 0.0161 0.366 0.0117 0.252 0.0105 0.0 - 0.25 0.645 0.0435 0.329 0.0291 0.210 0.0239 0.25-0.50 0.667 0.0786 0.411 0.0592 0.295 0.0518 0.50-0.75 0.0 0.0 0.0 0.0 0.0 0.0 0.75-1.0 0.0 0.0 0.0 0.0 0.0 0.0 Crop area shares: fruits No fruits 0.671 0.0176 0.351 0.0120 0.233 0.0109 0.0-0.25 0.523 0.0377 0.296 0.0257 0.203 0.0217 0.25-0.50 0.738 0.0480 0.471 0.0383 0.345 0.0351 0.50-0.75 0.625 0.1211 0.398 0.0930 0.297 0.0855 0.75-1.0 0.903 0.0531 0.668 0.0647 0.675 0.0672

the study, the lowest poverty incidence was Poverty and geographic characteristics observed in East Shewa and the highest in North Table 5 shows poverty indices by geographic Omo. The observed low poverty incidence rate location of the sample households. The poverty in East Shewa is not surprising given the fact incidence is generally higher in all of the that the zone is relatively well developed in Ethiopian regional states. It is relatively lower in terms of services and infrastructure, thus Oromia and Tigray regional states and higher in providing relatively better marketing conditions Southern Nations Nationalities and Peoples and employment opportunities. We have also states2. When comparing the zones included in assessed poverty according to which basin the sample irrigation schemes or farm households 2 However note the regional differences in poverty line and the non-representative ness of the sample

105

belong. It was found that poverty is significantly lower in Awash and Denakil basins.

Table 5. Headcount, depth and severity of poverty among sample households Variable Incidence (α = 0 ) Depth (α = 1) Severity (α = 2 ) value SD Value SD Value SD Sample total 0.657 0.0148 0.362 0.0107 0.248 0.0010 Zones North Omo 0.871 0.0285 0.626 0.0286 0.506 0.0302 Arsi 0.648 0.0460 0.268 0.0261 0.145 0.0222 Awi 0.717 0.0438 0.390 0.0325 0.264 0.0279 Raya Azebo 0.565 0.0351 0.299 0.0227 0.193 0.0178 East Shewa 0.455 0.0387 0.177 0.0192 0.092 0.0132 West Shewa 0.727 0.0347 0.417 0.0260 0.286 0.0227 West Gojam 0.664 0.0418 0.330 0.0277 0.207 0.0231 Basins Abay 0.707 0.0227 0.388 0.0166 0.262 0.0145 Awash 0.535 0.0301 0.219 0.0162 0.120 0.0127 Denakil 0.444 0.0500 0.272 0.0362 0.204 0.0310 Rift Valley 0.871 0.0284 0.626 0.0286 0.506 0.0302 Tekeze 0.704 0.0440 0.369 0.0302 0.235 0.0256 Region Amara 0.693 0.0299 0.368 0.0215 0.236 0.0188 SNNP 0.871 0.0285 0.626 0.0286 0.506 0.0302 Oromia 0.607 0.0233 0.293 0.0148 0.182 0.0123 Tigray 0.580 0.0343 0.323 0.0237 0.220 0.0200

The incidence of and severity of poverty is profound effect on poverty. In fact there are no higher in rural than urban areas (52 per cent and poor people with post secondary education. 36 per cent, respectively). Poverty is uniformly Poverty is also highly associated with family distributed throughout the country’s rural areas. size . The poverty incidence is almost 90% An exception is the region of Oromiya, where among households having 10 members or more. the level and intensity of poverty is significantly Contrary to our expectation, the poverty lower. incidence is relatively lower among female headed households. Poverty incidence is also Poverty and household demographic and lower among younger households. socioeconomic characteristics Table 6 presents the state of poverty among sample farmers by their demographic and socioeconomic characteristics. Education had a

106

Table 6. Household socioeconomic and demographic characteristics Variables Incidence (α = 0 ) Depth (α = 1) Severity (α = 2 ) value SD Value SD Value SD Education No education 0.677 0.0186 0.364 0.0132 0.243 0.0114 Elementary 0.649 0.0295 0.356 0.0209 0.241 0.0182 Secondary 0.539 0.0465 0.295 0.0333 0.215 0.0311 Post secondary 0.0 NA 0.0 NA 0.0 NA Household Size 1 person 0.348 0.0703 0.177 0.0445 0.122 0.0390 2-4 persons 0.529 0.0278 0.277 0.0181 0.183 0.0153 5-9 persons 0.727 0.0183 0.399 0.0139 0.275 0.0126 10 + persons 0.885 0.0408 0.581 0.0401 0.435 0.0411 Gender Male 0.664 0.0162 0.368 0.0118 0.254 0.0105 Female 0.626 0.0370 0.330 0.0257 0.221 0.0220 Household age group 15 through 24 0.561 0.0658 0.301 0.0463 0.212 0.0419 25 through 34 0.592 0.0347 0.310 0.0239 0.211 0.0215 35 through 44 0.665 0.0292 0.359 0.0412 0.245 0.0187 45 through 54 0.710 0.0320 0.315 0.0225 0.315 0.0225 55 through 64 0.680 0.0381 0.359 0.0268 0.236 0.0232 65 through 74 0.686 0.0460 0.358 0.0322 0.233 0.0278 75 + 0.646 0.0691 0.364 0.0491 0.248 0.0430

5. Determinants of rural poverty: the role of as education of different household members, access to irrigation number of income earners, household composition and size, and geographic location. Poverty and poverty changes are affected by The regressions will return results only for the both microeconomic and macroeconomic degree of association or correlation, not for variables. Within a microeconomic context, the causal relationships. simplest method of analyzing the correlates of poverty is to use regression analysis to see the Empirical Model effect on poverty of a specific household or The discussion in section 3 has relied largely on individual characteristic while holding constant descriptive results, exploring relationships all other characteristics, which is the focus of between variables without holding the effect of this section. In these regressions, the logarithm other factors constant. However, correlations of consumption or income (possibly divided by among key variables potentially could obscure the poverty line) is typically used as the left the relationship between poverty and a single hand variable (Qiuqiong et al.2005). An factor of interest. Consequently it is useful to alternative framework transforms the continuous analyze the impact of the relevant variables on income variable into binary variable using poverty holding all other factors constant. This poverty line as a cutoff value (Anyanwu 2005). implies the need to separate the effects of The resulting dummy variable indicates whether correlates. We approach this problem through a household is poor (i.e., the household’s income the application of multivariate analysis, using is less than the poverty line) or non-poor (i.e., logistic regression. The dependent variable is a household’s income is more than the poverty discrete variable which takes a value equal to 0 line). In this paper we follow the later approach. for non-poor, if a household had per capita The right-hand explanatory variables span a income equal to or more than 1075 Birr and 1 large array of possible poverty correlates, such for poor if a household had a per capita income

107

less than 1075 Birr (which is considered her as a the conditional probabilities are calculated for poverty line). each sample household, the partial effects of the continuous individual variables on household The explanatory variables considered in the poverty can be calculated using model were household heads’ personal characteristics (age, gender, educational ∂Λ(β ' X i ) achievement, etc), household demographic = Λ()()β ' X i []1− Λ β ' X i ∂X characteristics (household size and its square), i household wealth (farm size, livestock holding), the nature of farming system (share of grains in the total cultivated area, size of irrigated area), The partial effects of the discrete variables will and location (zones to which the household be calculated by taking the difference of the belong). See table 7 for details of the variables mean probabilities estimated for respective included in the model. discrete variables at values 0 and 1.

In the model, the response variable is binary, Alternatively, we present the change of the odds taking only two values, 1 if the rural household ratios as the dependant variables change. The is poor, 0 if not. The probability of being poor odds ratio is defined as the ratio of the depends on a set of variables listed above and probability of being poor divided by the denoted as X so that: probability of not being poor. This is computed as the exponents of the logit coefficients ( e β ) Prob()Y = 1 = F(β ' x) and can be expressed in percentage as [100( e β - 1)]. Prob()Y = 0 = 1− F(β ' x) (1) Before presenting the model results we wish to

give a brief description of the variables included

in the model (See table 7). There is significant Using the logistic distribution we have: association between poverty and access to

irrigation. Irrigating households have also β 'x β 'x Prob()Y = 1 = (e 1+ e ) = Λ()β ' x significantly higher farm size, family (2) size, and years of schooling. They also devote significantly lower area to the cultivation of food Where Λ represents the logistic cumulative grains than the non-irrigators. The proportion of distributions function. Then the probability female headed households is relatively higher model is the expression: among farmers without access to irrigation.

E[]y x = 0[]1− F()β ' x +1[]F()β ' x (3)

Since the logistic model is not linear, the marginal effects of each independent variable on

the dependent variable are not constant but are dependent on the values of independent variables. Thus, to analyze the effects of the independent variables upon the probability of being poor, we calculated the conditional probabilities for each sample household. Once

108

Table 7. Description of variables included in the model Variables Irrigators Non-irrigators t-statistic χ 2 Proportion of poor (Y=1=poor, 0 other 56.8 76.6 NA 41.578*** wise) (%) Proportion of female (%) (X1) 14.8 19.6 NA 4.051* Zones (Number) (X2) North Omo 55 55 NA NA Arsi 109 30 NA NA Awi 55 53 NA NA Raya Azebo 107 100 NA NA East Shewa 108 57 NA NA West Shewa 110 55 NA NA West Gojam 83 47 NA NA Irrigated area (Timmad) (X3) 3.02 NA NA NA Farm Size (Timmad) (X4) 6.87 5.90 8.321*** NA Area share of grains (%) (X5) 64.33 91.21 234.085*** NA Livestock holding in TLU (X6) 3.78 4.20 2.708 NA Family Size (number) (X7) 5.63 5.34 3.569* NA Age of household head (years) (X8) 45.99 44.84 1.386 NA Years of schooling (X9) 2.34 1.65 11.389*** NA

The logistic regression analysis is fitted to and years of schooling significantly reduce the strengthen and clarify the descriptive results of probability of being poor, while increases in the preceding descriptive sections. family size and area share of food grains in the total cultivated area significantly increases the Empirical results probability of being poor. The relationship The model results are summarized in table 8. between poverty and family size is non-linear. The likelihood ratio χ 2 statistic is used to test Family size increases the probability of being poor up to a certain point beyond which any the dependence of rural poverty on the variables successive addition of a family member included in the model. Under the null hypothesis contributes to the reduction of poverty. This (Ho) where we have only one parameter, which confirms the usual inverse U relationship is the intercept ( β o ), the value of the restricted between poverty and family size (World Bank log likelihood function is -666.39, while under 1991, 1996; Lanjouw and Ravallion 1994; Cortes 1997; Szekely 1998, Gang et al. 2004). the alternative hypothesis ( H1 ) where we have all the parameters, the value of the unrestricted Livestock holding size, which is usually log likelihood function is -453.64. The model regarded as a measure of wealth (Shiferaw et 2 al.2007), had the expected sign but not χ statistic is highly significant, indicating that statistically significant. Contrary to our the log odds of household poverty is related to expectation female headed households had lower the model variables. With regard to the chance of being poor as compared to male predictive efficiency of the model, of the 1024 headed households. Concerning location effects, sample households included in the model, 822 or the probability being poor for sample 80.3% are correctly predicted. households from North Omo and West Shewa is significantly higher, whereas the probability of The results of the parameter estimates of being poor for households from East Shewa and determinants of poverty generally agree with the Raya Azebo zones is significantly lower. descriptive results of the preceding section. Of the twelve variables included in the model, nine We assess the magnitude of the effect of were found to have a significant impact on changes in statistically significant and policy poverty. Increases in farm size, irrigated area

109

relevant variables on household poverty based ***Statistically significant at p<0.01; on the partial effects of the respective variables **Statistically significant at p<0.05; on conditional probabilities (Table 9). The *statistically significant at p<0.1. partial effects of continuous variables were calculated using equation 4, while those of the In logit model analysis, it is marginal effect discrete variables were calculated by taking the values and elasticities that have direct economic difference between the mean probabilities interpretation not the estimated coefficients. estimated at the respective values (0 and 1) of Looking at the marginal effect and elasticity the discrete variables. The partial effects thus values presented in table 8, the irrigation calculated from the logistic model show the variable comes third or after area share of grains effect of change in an individual variable on the and family size variables in quantitative probability of being poor when all other importance with respect to poverty reduction. exogenous variables are held constant. Rural poverty is highly responsive to the cropping pattern. A unit increase in the Table 8. Parameter estimates of determinants of proportion of area of grain crops increase the poverty model probability of being poor by 0.41% or a 1% a Variables Estimateincrease in SEthe proportione β of100( areae βdevoted-1) to Constant -1.018grain crops 0.913increase 0.361 the probability-63.9 of being Size of irrigated area -0.354***poor by 0.44%. 0.117 This 0.702 implies -29.8that changing the Area share of grains cultivation 1.942***crop mix managed0.433 6.970by farmers 597 towards high Irrigated area-by-area share of grain 0.291*value crops0.156 such as1.338 vegetables 33.8 would have a Farm size -0.202***profound effect 0.026 on 0.817 rural poverty. -18.3 Irrigation Livestock holding in TLU -0.039technology 0.025facilitates 0.961 the cropping -3.9 pattern shift Family size 0.724***process. A one 0.146 timmad 2.064 increase 106.4 in irrigated area Square of family size -0.022*would reduce 0.012 the probability 0.979 -2.1of being poor by Age of household head -0.0500.075% . In 0.035 other 0.951words, a -4.9 1% increase in Square of age of household head 0.001irrigated area 0.000 would 1.001 reduce the 0.1 probability of Level of education of HH head -0.116***being poor 0.032by 0.2%. 0.890 Increasing -11 the household Sex of the household head(=Male) 0.438*member by 0.246 one person 1.549 would 54.9 increase the Zones: probability of being poor by 0.15%. North Omo 2.248***Alternatively 0.440 a 1% 9.470increase in 847 the family size Arsi 0.663*would increase0.378 the probability1.940 94 of being poor by Awi -0.1611.21%. 0.353 0.852 -14.8 Raya Azebo -0.569*Another significant 0.296 0.566policy relevant -43.4 variable is East Shewa -1.353***years of schooling. 0.309 0.258An unit increase -74.2 in year of West Shewa 1.107***schooling decreases 0.357 3.026 the probability 202.6 of being West Gojam (reference) poor by 0.0245. Note: Restricted log likelihood value [Log(L0)]=- 666.3848 Unrestricted log likelihood value [Log (L1)]=- 453.6428 Log likelihood value ( 2 χ ()df =9 ) = −2 log L0 − − log L = 425.4841 []()() ()1 *** % of correct prediction=80.3 Number of observation=1024 a The parameters were estimated using maximum likelihood methods. They are un-weighted

110

Table 9. Marginal effects of the significant access to irrigation (See panel a and b of Figure variables 3). In the past due mainly to the demand for Determinants Marginal Elasticity irrigated land exceeding the supply and due to effects also partly to the egalitarian policies followed Irrigated area in -0.0747 -0.20 for rural development, the irrigated land is Timmad rationed in Ethiopia. In an effort to reach many Area share of 0.4089 0.44 people the irrigated plots distributed to farmers grain crops are often far below an economic size that is Farm size in -0.0426 -0.40 sufficient to warrant the full engagement of Timmad farmers in irrigated production business. Family size 0.1526 1.21 Consequently, irrigated farming is considered as Years of -0.0245 -0.07 a second best option by farmers. schooling Gender (Male) 0.0865 0.02 Rural poverty is also very responsive to Zones cropping pattern changes (see panel c and d of North Omo 0.3113 0.06 Figure 3). Reductions in area share of food Arsi 0.1240 0.02 grains and increases in the area share of high Awi -0.0346 -0.01 value crops such as vegetables significantly Raya Azebo -0.1268 -0.04 reduces rural poverty. Two major variables that East Shewa -0.3156 -0.07 allow the change to high value crops are access West Shewa 0.1948 0.05 to irrigation and proximity to the demand centers thus allowing easy marketing. Figure 4 ( panel a and b)show that poverty is highly related to The interesting results contained in table 10 can family size and level of education of the be graphically depicted. Poverty is more household head. responsive to the size of irrigated area than mere

111

1 1 0.9 o 0.9 0.8 0.8 0.7 0.7 0.6 0.6 0.5 0.5 0.4 0.4 0.3 0.3 0.2

0.2 po being of Probabilitity 0.1 0.1 0 0 No 1st 2nd 3rd 4th Irrigators Non-irrigators irrigation quartile quartile quartile quartile Access to irrigation

Predicted probability of being poor being of probability Predicted Size of irrigated land

Panel a Panel b

o 1 0.9 0.9 0.8 0.8 0.7 0.7 0.6 0.6 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2 Probability of being poor being of Probability Probability of being po being of Probability 0.1 0.1 0 0 No grains 0-0.25 0.25-0.5 0.5-0.75 0.75-1.0 le le e e le i til ti ar Area share of grains quart quar quartil qu rd h 1st 2nd 3 4t No vegetab Area share of vegetables Panel d

Panel c

Figure 3. Irrigation, cropping pattern and poverty

112

0.9 1 0.8 0.9 o o 0.7 0.8 0.6 0.7 0.6 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2 probability of being po being of probability Probability of being po being of Probability 0.1 0.1 0 0 1 Person 2-4 5-9 10+ No Elementary Secondary Post Persons Persons Persons education secondary Household Size Education status

Panel a Panel b

Figure 4. A graphical illustration of the influence of education and household size on poverty.

113

6. Conclusions and policy implications From the results presented in this paper, the following conclusions may be made: In Ethiopia agriculture and even the There is significant difference in performance of macro-economy is taken incidence, depth and severity of poverty hostage by the amount and distribution of between households with access to rainfall (Reference).The unreliable rainfall irrigation and those without. However, pattern in many parts of the country forced the poverty incidence among the sample the farming population to adopt a risk-averse households is still unacceptably high behavior, the behavior that limits the irrespective of access to irrigation, capacity of farmers to innovate and adopt indicating that poverty deeply entrenched farming technologies with potential of in rural Ethiopia. boosting yield and income. For instance, the Poverty indices are responsive to successive Ethiopian governments have tried irrigation typology and irrigation to enhance the productivity of agriculture intensity. Among the irrigation the two through modest investments in agricultural irrigation typologies studied the poverty research and extension, mainly focused on situation is relatively milder among seed and fertilizer technologies3. Several modern irrigation scheme users. evaluation studies of these programs have Poverty indices were found also to be underlined that the seed and fertilizer responsive to the irrigation intensity as technologies were mostly successful in areas measured by the size of irrigated area. endowed with relatively ample moisture Poverty incidence is significantly lower (reference). It was based on this revelation among households with higher irrigated that the government, NGOs and farmers have area size. Due to demand outstripping the made investments in agricultural water limited supply of irrigation service and management such as small-scale irrigation due to considerations for equity, irrigation schemes to extricate the agricultural sector plots are rationed in Ethiopia. The limited and the economy at large form the shackles differentiation observed in the size of of unreliable rainfall. The main goals of irrigated land among sample farmers is these investments in small-scale irrigation due to the prevalence of informal irrigable schemes were reducing food insecurity and land markets. This calls for an incidence of rural poverty. This paper investigation to determine a minimum assessed whether the developed irrigation irrigated area that needed to be allotted to schemes have lived up to the expectation of a household for sustained poverty significantly reducing rural poverty and also reduction and food insecurity eradication. inequality. Poverty incidence is also related to the cropping pattern, indicating that mere The study was based on the extensive data access to irrigation would not bring the set generated form a total of 11 small-scale desired results. Poverty situation is more irrigation schemes (i.e., 7 modern schemes sever among farmers devoting significant and 4 traditional schemes), sampled from proportion of their cropping land to food four major regional states of Ethiopia. For grains (cereals, oil seeds and pulses) comparison purposes a sample of adjacent irrespective of access to irrigation. villages with no access to irrigation was also Vegetable growers are better off in terms sampled. All in all 1024 farming households of poverty situation. The implication is were randomly sampled from the selected that irrigation project planners should irrigation schemes and rain-fed villages. It consider the crop mix in future irrigation consisted of 627 irrigating households (of development plans. which 382 are modern scheme irrigators and Income inequality among households 245 are traditional schemes irrigators) and with access to irrigation is worse than that 397 purely rain-fed farmers. It is to be noted of those with out access. The implication that even those households with access to is that even though accesses to irrigation irrigation do manage rain-fed plots. Only few moves up the mean income, farmers have farmers were found to be purely irrigators. different capacity in making better use of the available irrigation water and therefore irrigation widens the income

3 See for instance the evaluation reports of the recent Extension Package Projects

114

gap4. However, the main policy concern Lipton, M., and M. Ravallion. 1995. Poverty in Ethiopia is reducing absolute poverty at and policy. In Handbook of development this moment. economics, Vol. III, ed. J. Behrman and Finally, our study confirms that while the T. N. Srinivasan. Amsterdam: Elsevier. income inequalities among households Lanjouw, P., Ravallion M.1994. Poverty and without access to irrigation are lower, it household size , policy research working was found that inequality among rainfed paper 1332, World Bank, Washington poor farmers is higher than those with DC. access to irrigation!!! May J. 2001. An elusive consensus: References definitions, measurement and analysis of poverty. In: choices for the poor. Anyanwu, JC. 2005. Rural poverty in Lessons from National Poverty Nigeria: profile, Determinats and Exit Strategies, Grinspum, A. (ed.). UNDP. Paths. African Development Review, Vol 17(3): 435-460 Qiuqiong H., Scott R, David D., Jikun H. 2005. Irrigation, Poverty and Inequality Carter M.R., Little D., Tewodaj Mogues, in Rural China" . Australian Journal of Workneh Negatu. 2007. Poverty Traps Agricultural and Resource Economics, and natural Disasters in Ethiopia and Vol. 49( 2): 159-175 Honduras. Ravallion M. 2003. “Assessing the Poverty Coudouel, A. , Hentschel, J., Wodon, Q. Impact of an Assigned Program.” In 2002. Poverty Measurement and Francois Bourguignon and Luiz A. Analysis, in the PRSP Pereira da Silva (eds.) The Impact of Sourcebook, World Bank, Washington Economic Policies on Poverty and D.C. Income Distribution: Evaluation Techniques and Tools, Volume 1. New Cortes, F.1997. Determinants of poverty in York: Oxford University Press. in Hogares, Mexico, 1992, Revista Mexicana de Sociologia, Vol(59)2: 131- Sen A. 2000. Social exclusion: concept, 160. application, and scrutiny. Social Development Papers No. 1. Office of Duclos J-Y, Abdelkrim Araar and Fortin, C. Environment and Social Development, 2006. DAD: a soft ware for distributive Asian Development Bank analyses. Szekely, M. 1998. The economics of Duclos J-Y, Abdelkrim Araar.2006. Poverty poverty, inequality and wealth And Equity: Measurement, Policy And accumulation in Mexico, St.Antony’s Estimation With Dad. Springer 233 Series, Macmillan, New Yrok. Spring Street New York, Ny 10013 & International Development Research Tarp F, Simler K, Matusse C, Heltberg R, Centre Po Box 8500, Ottawa, On, Dava G.2002.The Robustness of Poverty Canada K1g 3h9. Profiles Reconsidered

Gang, I.N. K.Sen, Yun, M-S. 2004. Caste, UNDP. 2006. Human Development Report Ethinicity and poverty in rural 2006. Beyond scarcity: Power, poverty India(see:www.wm.edu/economics/semi and the global water crisis. New York. nar/papers/gang.pdf) Jyotsna, J., Ravallion M, 2002. Does Piped World Bank. 2006. Ethiopia: Managing Water Reduce Diarrhea for Children in water resources to maximize sustainable Rural India? Journal of Econometrics growth. A World Bank Water Resources Assistance Strategy for Ethiopia. The World Bank Agriculture and Rural Development Department. Report NO. 36000-ET. Washington D.C. 4 See which studies agree with our findings and which ones do not?

115

World Bank 2005. Water Resources, Growth and Development. A working paper for discussion prepared by the World Bank for the Panel of Finance Ministers. The U.N. Commission on Sustainable Development 18 April 2005.

World Bank. 1991. Assistance strategies to reduce poverty: a policy paper, World BankWashington DC.

116

An Assessment of the financial viability and income impact of small scale irrigation in Ethiopia

Godswill Makombe1, Fitsum Hagos1, Regassa E. Namara2 and Seleshi Bekele Awulachew1

1 International Water Management Institute (IWMI) Subregional Office for the Nile Basin and East Africa, Addis Ababa, Ethiopia 2 International Water Management Institute (IWMI), Africa Regional Office. Accra, Ghana. [email protected]

compared to projects from other regions Abstract (Inocencio et al., 2007). In a study of 314

Recently, there has been very little irrigation schemes of which 45 were from SSA, 51 development in sub-Saharan Africa. The from Middle East and North Africa, 41 from main reasons cited for this lack of interest in Latin America and the Caribbean, 91 from developing irrigation in sub-Saharan Africa is that irrigation projects are expensive and South Asia, 68 from South East Asia and 18 perform poorly compared to projects from from East Asia, Innocencio et al (2007) other regions. However, when classified into success and failure projects, the sub- showed that when average costs are Saharan Africa success projects’ investment considered, establishment costs for irrigation costs are not significantly higher than from projects in SSA were significantly higher at other regions. African countries like an estimated USD14, 455 per ha compared Ethiopia, which has embarked on an to USD6, 590 for non-SSA projects. They agricultural led development program, further analyzed establishment costs by aspire to use irrigation as a development defining “Success” and “Failure” projects. strategy with small scale irrigation playing a In defining these they used 10 percent key role in rural development. This study economic internal rate of return as a cut off evaluates the financial performance of small point. Those projects that achieved less than scale irrigation using O & M and investment 10 percent economic internal rate of return recovery, and the ability to replicate the were classified as “Failure’ projects. The investments. It is concluded that the justification for using 10 percent was that systems are financially viable and provide a this is the cut off point used for evaluating low cost development option for rural areas. public projects.

1. Introduction After applying this classification, they found that for the “Failure” projects in SSA the Recently, there has been very little irrigation establishment costs averaged USD 23,184 development in sub-Saharan Africa (SSA). compared to USD 10,624 for non-SSA The main reasons cited for this lack of projects, whereas for the SSA “Success” interest in developing irrigation in SSA is projects the average was USD 5,726 that it was believed that irrigation projects in compared to USD 4,603 for non-SSA SSA are expensive and perform poorly projects. This difference was not

117

statistically significant showing that for classification system uses the size of “Success” projects, the SSA projects are not command area irrigated as follows: more expensive than their non SSA 1. Small scale irrigation systems <200ha counterparts. Their analysis also shows that 2. Medium scale irrigation systems (200- the performance of both non SSA and SSA 3000ha) projects has improved over time. 3. Large scale irrigation systems (>3000) Commenting on this performance The second classification uses a mix of the improvement, they conclude that: “The history of establishment, time of degree and speed of improvements have establishment, management system and been deeper and faster in SSA than in non- nature of the structures as follows: SSA, so that the difference in unit cost and 1. Traditional schemes: These are SSI project performance between SSA and non- systems which usually use diversion SSA, which used to be significant in earlier weirs made from local material which decades, has been reduced to the extent that need annual reconstruction or from there is no significant difference in the latest small dams. The canals are usually decade”. (Inocencio et al, 2007; pp 42) earthen and the schemes are managed by the community. Many are Some countries in Africa, have a renewed constructed by local community effort interest and some like Ethiopia have recently and have been functional for very long become interested in the role that irrigation periods of time, some were recently can play in the development process. constructed with the aid of NGOs and Agriculture plays a major role in Ethiopia government. contributing more than 44 percent to GDP 2. Modern schemes: These are SSI over the period 1996 to 2006 (Government systems with more permanent of the Republic of Ethiopia, 2006a). Most of diversion weirs made from concrete agriculture’s contribution is based on hence no need for annual smallholders who produce cereals under reconstruction and small dams. The rainfed production. This leaves the primary and sometimes secondary performance of the Ethiopian economy canals are made of concrete. They are exposed to the vagaries of nature by community managed and have depending on how good the rainfall season recently been constructed by is (World Bank, 2006). The Ethiopian government. government, in its agricultural led 3. Public: These are large scale development program, aspires to use operations constructed and managed irrigation as a major development by government. Sometimes, public component. Currently less than 5 percent of schemes have out growers whose the potentially 3.5 million ha of irrigable operations are partially supported by land is developed. The government of the large scheme. Ethiopia aspires to develop about 430,061 4. Private: These are privately owned ha within the planning period of the Plan for systems that are usually highly Accelerated and Sustained Development to intensive operations. End Poverty (PASDEP) which spans the Given our interest in SSI, which is years 2005/05 to 2009/10 (Government of distinguished from large scale irrigation by the Republic of Ethiopia, 2006b). This the farm level scale of operation, we planning document aspires to strongly therefore prefer to identify SSI irrigation develop and support small scale irrigation systems using the second classification (SSI). system and we study the first and second categories of this classification. Werfring Irrigation development in Ethiopia is (2004) describes the typology of SSI in classified using two systems. The first Ethiopia in detail.

118

Given the strong support envisaged for SSI (Innocencio, 2007). Verdier (1992) gives development during the PASDEP planning estimates of O & M for earth work (canals, period, it is important to provide insights drains, feeder roads with no tarmac) as 2 into the viability of SSI investment in order percent of investment, and concrete to inform investment decisions. structures (river diversion, weir and inlet) as 1 percent of initial investment. In this 2. Objectives study O & M is assumed to be 10 percent of annual replacement cost. The main objective of the study is to provide The first index shows farm level ability to a contribution that can be used to partially recover O & M costs. If farmers cannot answer the question whether investment in recover O & M, it renders the scheme non- SSI is a viable option for the proposed financially viable. The second index shows agricultural led industrialization whether farmers recover both initial development strategy by assessing the investment and operation and maintenance financial viability of existing SSI. costs. Ideally, in a financially viable Supplementary to the main objective we also scheme, both investment and O & M should estimate the importance of agriculture to be recovered. The third index shows rural smallholders by estimating how much whether farmers can recover both initial income is derived from agriculture investment and operation and maintenance compared to off farm sources. PASDEP costs and still have the potential financial aspires to develop and support SSI but the capacity to reinvest in a similar SSI system, current farm level contribution of irrigation in other words, could the schemes to the rural households is not known. This potentially financially perpetuate paper also aims to estimate the income themselves. impact of irrigation to smallholder producers 4. Data collection 3. Methodology Data were collected on the initial investment The methodology we use is partially based or establishment costs for the small scale on that used by Huang et al (2006) to irrigation systems. During the growing evaluate benefits and costs of irrigation season May 2005 to March 2006, plot level systems in China. We use gross margin data were collected from ten SSI schemes. analysis to estimate agricultural income for Data were collected on cropping patterns, irrigators and non irrigators. Based on the areas under crops, yields, marketed output, gross margins, we estimate the income inputs, and input and output prices. Since contribution of agriculture in general and farmers usually grow at least two crops, irrigation in particular to the household. sometimes three on the irrigated plots, the In order to assess the financial viability of cropped area is summed across seasons. On SSI we define three indices; each of the schemes, a random sample of 50 1. O & M index = GMI / O & M. farmers was selected. A random sample of 2. Financial performance index = GMI / 50 non irrigating farmers was also selected (I + O& M.). from each site as a control. Data were also 3. Replicability index = GMI – (I + O & collected on non-agricultural income so as to M) / (I + O & M). estimate the contribution of agriculture and Where: GMI = Gross margin from irrigated irrigation to household income. production, O & M = Operation and During the summer most of SSI systems maintenance costs, I = annual replacement grow cereals like teff, maize and barley cost all on a per ha basis. Annual under supplementary irrigation given that it replacement cost is computed as initial rains during the summer. During the winter investment divided by project lifetime. farmers grow a variety of vegetables Project lifetime is assumed to be 30 years including onions, tomatoes, and leafy green

119

vegetables like spinach under full irrigation. systematically collected and kept in a central Rainfed farmers’ production is primarily location. The data, if available, could be based on the staple cereals teff, wheat, found for different schemes in different barley and sorghum. Both rainfed and locations, for instance in different ministries. irrigating farmers also grow perennial crops Sometimes data were found in one location, like mango, banana, sugar cane which are like a ministry, but different departments. sometimes intercropped with seasonal crops. Furthermore, some data is kept at federal Data were collected on all crops grown on a level, whereas other data are kept at regional sampled farm. and sometimes district level. Some regional During data collection we took cognizance authorities pass information to the federal of the fact that most of the cereal production level, for instance to the Ministry of Water is kept for home consumption. The Resources, but some simply do not. If a computation of gross margins was based on donor is involved in the project data collected on yields and the prices of establishment, sometimes the donor keeps marketed output. For instance, if a farmer the records, if at all. This made the process sold half of the wheat yield, we assumed of collecting establishment costs for even a that the prices realized in the market would small sample of SSI systems quite an have also been realized by the farmer if the arduous and time consuming task. Given rest had been sold. Although it is possible this, even though we started off with 10 that if more produce is put on the market, schemes, three traditional and 7 modern, we prices tend to reduce we also argue that the could only collect accurate investment shadow price attached to the retained output information on all the modern schemes and by the farmer has to be higher than the only one traditional system. The Hare market price, assuming a rational farmer modern scheme was excluded from the would sell if their shadow price is lower sample for the reason of perennial crops as than the market price. mentioned earlier so the final sample, for the Most farmers, both irrigating and rainfed financial viability analysis, was made up of mainly grow seasonal crops but some do one traditional and six modern schemes. grow perennial crops. Data on the input- Even though the data is still not centralized output relationships of seasonal crops were for modern schemes, the likelihood that it easier to collect than those of perennial exists and that it can be accessed is higher crops. For instance, there could be some for modern schemes than for traditional ones perennial crops intercropped and spaced since the modern schemes are usually built within a seasonal crop. The area was better with some form of government involvement estimated for the seasonal crop than for the at regional or federal level. The data for perennial crop. Even though in some cases most traditional schemes is very difficult to of modern irrigation schemes where most come by. Bruns (1991) notes that there is a the area was under a perennial crop like serious lack on information on SSI. banana, the input-output relationships were still much easier and accurate for seasonal 6. Income levels and dependence on crops since the operations on the perennials agriculture and irrigation are not as regular and consistent as on the seasonal crops. Based on these two Table 2 summarizes the cropped area and observations, this analysis only includes the incomes for the sample farmers. The sites with seasonal crops. average irrigated area for all the sample irrigated systems is 0.71 ha, but is slightly 5. Challenges in data collection higher for modern schemes at 0.76 ha while it averages 0.58 ha for traditional systems. During the process of collecting the data on The highest irrigated area is at Endris establishment costs, we realized that the data modern irrigation system at 1.07 ha while on small scale irrigation systems is not the lowest is at Haiba modern irrigated

120

system at 0.35 ha. Average cropped area for income. In table 1, total income is the sum the rainfed farmers is 1.41 ha. of agricultural income, irrigated and non Table 2 also shows the extent to which the sample households depend on agriculture for

Table 1. Income dependence on irrigation for sample schemes.1 Region Site name Site type Area (ha) Income (Irrigation/Rainfed) Agricultural income Total3 % of total Irrigated2 Rainfed (USD) Rainfed Irrigated + irrigate d Oromia Endris Modern 1.07 1.88 603 83 36 [n=42] (1.20) (1.46) (975) Traditional 0.65 1.40 471 90 38 [n=41] (0.81) (1.19) (1198) Rainfed N/A 2.75 360 90 N/A (n=55) (1.47) (288)

Wedecha Modern 0.46 1.13 771 92 44 Belbella [n=51] (0.35) (0.68) (493) system Traditional 0.56 0.94 570 98 45 Filitino (0.41) (0.47) (404) [n=52] Rainfed N/A 1.36 468 99 N/A [n=57] (0.59) (315) Gologota Modern 0.96 N/A 713 88 86 [n=52] (0.41) (721) Rainfed N/A 1.39 304 71 N/A [n=55] (1.21) (203) Amhara Zengeny Modern 0.85 0.76 552 91 58 [n=49] (0.77) (0.39) (633) Rainfed N/A 0.81 261 90 N/A [n=47] (0.46) (326) Tikurit Traditional 0.47 0.90 576 95 55 [n=55] (0.49) (0.86) (639) Rainfed N/A 1.18 277 96 N/A [n=42] (0.76) (164) Tigray Haiba Modern 0.35 0.73 346 82 34 [n=47] (0.28) (0.43) (297) Rainfed N/A 0.72 240 80 N/A [n=53] (0.37) (251) Golgol Modern 1.03 1.86 1100 80 52 Raya [n=26] (0.60) (1.15) (1071) (Kara Rainfed N/A 1.59 247 100 N/A Adishu) [n=22] (0.89) (320) All Modern 0.76 1.16 650 87 53 [n=42] (0.71) (0.96) (728) All Traditional 0.58 1.09 536 95 45 [n=42] (0.59) (0.87) (806) All Irrigated 0.71 1.13 616 89 50

121

[n=42] (0.68) (0.93) (753) All rainfed N/A 1.41 318 88 N/A [n=42] (1.12) (278) 1 ( ) = sdev; 2 = Gross irrigated area summed over the cropping seasons. 3= Total income (agricultural + non agricultural) irrigated sample while it appears to be irrigated plus non agricultural income. slightly lower at traditional irrigated Income from agriculture and that from schemes at 45 percent. Given the significant irrigation is expressed as a percentage of contribution of irrigated agriculture to total income to show the income dependence income, it is essential to establish if the on agriculture and irrigation. For instance, systems are financially viable both in the the average annual income at Endris modern long and short term. irrigation scheme is 603U SD. On average 83 percent of the 603 USD is derived from One of the concerns raised by the World agriculture (irrigated and non-irrigated) and Bank (2006) is that, given the national 36 percent of the 603 USD is derived from dependence on rainfed agriculture, the irrigated agriculture. Agricultural income at performance of the economy is directly Endris modern irrigated system is about 500 related to the quality of the rainfall season. USD and income from irrigated production We have demonstrated that this statement is is about 217 USD, meaning irrigated even truer for rural smallholders whose agriculture contributes slightly more than 43 incomes are a direct function of the quality percent of agricultural income, even though of the rainfall season, given their high gross irrigated area is less than rainfed area. income dependence on agriculture. This makes agriculture a highly significant Irrigation, if it uses stored water, can be used contributor to income for the smallholders, to de-link the performance of the national much more so than the dependence on economy, and more so the incomes of the agriculture depicted at national level. It also rural poor smallholders from the quality of shows that irrigation, when made available, the rainfall season. can play a significant role in contributing to the income of rural households particularly 7. Investment levels for sample if we take into cognizance the small areas of schemes irrigation developed per household. Table 2 summarizes the investment levels For all irrigated systems, agricultural income for the sample SSI schemes. Constant 2006 constitutes about 90 percent of income, prices were used to make the figures while it appears to contribute a slightly comparable since the cropping data came higher proportion on traditional irrigation from 2005/2006 growing season and the systems. The lowest contribution of schemes were established at different times. agricultural income is at Gologota rainfed The exchange rate of 1USD = 8.69 Birr system at 71 percent while it is highest for which prevailed in 2006 (CIA, 2007) was Golgol Raya rainfed farmers at 100 percent. used to convert the expenditures in Birr to This may be explained by the fact that USD. The average per ha initial investment Golgota is close to the capital city, several cost is estimated at 2090 USD per ha. This towns and public schemes which offer estimate does not include possible employment opportunities whereas Golgol contribution by the community on the form Raya is several kilometers from the capital of labor and other materials. The data show city and also has neither towns nor public that the systems are low financial investment schemes in its proximity, hence has limited irrigation projects as this is slightly under 40 off farm employment. percent of the figure quoted by Inocencio et al (2007) for success projects in SSA. Irrigation contributes significantly to income Annual O & M costs were estimated as 10 at an average 50 percent for the whole percent of annual replacement costs.

122

Table 2. Investment levels for small scale irrigation schemes in constant 2006 prices Region Scheme Scheme Year Command Investment (USD/ha) name type established area (ha) Initial Annual O& Investment replacement M Oromia Indris1 Modern 19804 382 744.966 24.83 2.48 Gologota1 Modern 19624 850 870.537 29.02 2.90 Wedecha Modern 19905 150 3436.898 114.56 11.46 Belbella Traditional 19905 85 2544.659 84.82 8.48 system2 Filitino Amhara Zengeny1 Modern 19974 270 1071.8010 35.73 3.57 Tigray Haiba2 Modern 19974 250 2087.5211 69.58 6.96 Gol Gol Modern 20034 104 3864.5212 128.82 12.88 Raya (Kara Adishu)3

Water source and delivery system: 1 River diversion and gravity, 2 Small dam and gravity, 3 Deep well and pressurized drip and sprinkler. Year established: 4= actual, 5=based on feasibility study Sources of Investment figures: 6. Indris Irrigation project. RID-OFFICE for C.Z. 1991. Porject proposal report. Information Brochure. Idris Irrigation Development project, 7. Average of sample modern schemes, 8. East Shoa Water Mineral and Energy Resources Department (1998). Goa Worka Small Scale irrigation Project Proposal Final Draft , 9. East Shoa Water Mineral and Energy Resources Department (1998). Filtino Small Scale irrigation Project Proposal Final Draft , 10. Personal communication Yenew Desalegn, Irrigation expert. Zengeny irrigation scheme, 11. Co-SAERT (1993).List of irrigation sites constructed by Co-SAERT from 1987-1992 E.C. Unpublished., 12. Raya Valley ground water Development report. Unpublished. Note: Investment data for the traditional schemes at Endris and Tikurit t which are included in the income analysis, table 1, were not available. Table 3 summarizes the results of the Investment costs differ by site and region. It financial viability analysis. Based on the is beyond this paper to establish the reasons for the variations but this might partly three indices defined above we get some depend on the water source and delivery insights into the financial viability of system. The scheme with the highest irrigated schemes. At the onset, we have to establishment cost, Gol Gol Raya, has deep point out that this analysis only provides wells as water source and uses a pressurized insights into the financial viability of these drip and sprinkler system. We also note that systems because it is based on one year’s for the two Wedecha Belbella systems, data. Given the variation of agricultural which are close to each other, the traditional performance from year to year, ideally more system investment is lower, most likely than one year’s data on gross margins would reflecting less concrete infrastructure provide better insights. If more than one installed on the traditional system. year’s data is available, one could do many scenario analyses, one of which could be to 8. Assessing the financial viability of use both the performance and replicability SSI indices with a flow of gross margins and investment costs.

123

Table 3. Performance ratios for small scale irrigation systems Region Scheme name Distribution within Performance Indices (%)

Index O & M Financial Replicability Category

Oromia Indris <0 5 5 11 GE 0 < 1 0 5 14 GE 1 95 90 74 Mean 111 10 9 Gologota <0 10 10 10 GE 0 < 1 0 0 2 GE 1 90 90 88 Mean 245 22 21 Wedecha Modern <0 2 2 8 Belbella GE 0 < 1 0 6 10 system GE 1 98 92 82 Mean 70 6 10 Traditional <0 0 0 4 GE 0 < 1 0 4 12 GE 1 100 96 84 Mean 68 6 5 Amhara Zengeny <0 0 0 0 GE 0 < 1 0 0 2 GE 1 100 100 98 Mean 141 13 12 Tigray Haiba <0 0 0 17 GE 0 < 1 2 17 23 GE 1 98 83 60 Mean 65 6 5 Gol Gol Raya (Kara <0 0 0 19 Adishu) GE 0 < 1 4 19 19

GE 1 96 81 62 Mean 39 4 3 All Modern <0 3 3 10 GE 0 < 1 1 7 11 GE 1 96 90 79 Mean 119 11 10 All <0 2 3 9 GE 0 < 1 1 6 11 GE 1 97 91 80 Mean 111 10 9

124 that takes into account the time value of more cover O & M costs while the lowest money for both costs and returns flows in percentage covering investment costs is at computing either an Internal Rate of Return Golgol Raya at 81 percent. This shows that and or Net Present Value. Alternatively, farmers have the ability to pay for both O & instead of computing an internal rate of M and investment costs. return, one could also assume that farmers could borrow money at a certain interest rate, The replicability index asks the question if for instance the rate at which the government farmers were to pay for the current scheme borrows for development projects, annualize and to concurrently invest in a similar one, the cost flows by the interest rate and then could they manage it, in other words, could evaluate whether the farmers earn a return the systems potentially financially perpetuate higher than the interest rate. We do themselves. The answer to this question is understand that such analyses would be more yes they could manage. The lowest informative than the one done here. percentage of farmers with this ratio greater However, we believe that in the absence of than one is at Haiba with 60 percent, data to achieve such, our analysis is followed by Golgol Raya with 62 percent. informative, even though at best, it gives us We do understand that this analysis evaluates the performance of the systems for one year, what could happen; otherwise the income say emulating the first year of the project. earned from irrigation is subject to many competing family needs which generally do In our suite of indices, the first index shows not include reinvestment. This is just a whether the systems recover operation and simple way of evaluating financial viability. maintenance costs. If it is negative, irrigated From this simple analysis, the SSI systems income is less than O & M, if it lies between are financially viable and could also 0 and 1, O & M is only partially recovered potentially financially perpetuate themselves. and if it is greater than 1 O & M is fully Adams (1990) notes that, if they can be recovered. The same interpretation applies viable, SSI provide a low cost, low for the financial performance ratio where technology alternative to development. instead of O & M the sum of annual replacement cost and O & M is used to 9. Conclusions determine cost recovery. The means for these indices shown in table 3 show that The financial analysis shows that SSI most of the schemes recover O & M. It is projects in Ethiopia are very low investment possible that we may have underestimated O ventures. From the three indices we used, & M, at 10 percent of investment, however, we conclude that the systems are financially the degree to which most of the schemes viable. However, it is important to note that recover O & M leaves a lot of room for O & only one year’s data has been used in this M to increase substantially but still being analysis and therefore is missing the recovered. The lowest O & M index mean variability in returns that is characteristic of value is 39 for Golgol Raya and the highest agricultural production. The financial is 245 for Gologota. The low O & M index viability performance is in line with the at Gologol Raya can be explained by the fact observations made by Inocencio et al (2007) that the deep well water source combination of improved performance of recent irrigation with drip and sprinklers requires more projects in SSA. This makes investment in maintenance than the diversion weir and SSI a potentially viable low investment, gravity flow used in the other systems. development alternative. We show the degree to which Ethiopian Of importance is also the percentage farmers depend on agricultural income, and distribution of farmers between the ratios specifically on irrigated income and how this across systems. For instance, at Endris varies by location. The analysis shows that modern irrigation scheme, 95 percent of SSI development has potential for improving farmers fully cover their O & M costs, and the well being of the poor farmers through its 90 percent cover both investment and O & M significant impact on incomes. costs. In comparison at Golgol Raya 96 percent fully cover O & M costs while 81 It is important to note that all of the schemes percent cover investment plus O & M costs. evaluated in this study use diversion weirs, In general, at all the schemes, 90 percent or except one which uses underground water

125 and another using a small dam, thus their Department. Ministry of Finance and performance dependends on the quality of Economic Development. November, the rainfall season. The schemes using 2006a. diversion weirs although cheap to establish, Government of the Republic of Ethiopia., cannot achieve one of the government’s 2006. A Plan for Accelerated and objectives of de-linking national economic Sustained Development to End Poverty. performance and farmers’ incomes from the Ministry of Finance and Economic quality of the rainfall season. The financial Development. November, 2006b. viability of these SSI systems however, Huang, Q., Rozelle, S., Lohmar, B., Huang, provides insights into the fact that stored J., and Wang, J. (2006). Irrigation, water could also be potentially used for SSI agricultural performance and poverty to de-link irrigation performance for the reduction in China. Food Policy (31) 30- quality of the season. However, this needs to 52 be evaluated against the investment costs for Inocencio, A., Kikuchi, M., Tonosaki, M., the stored water. Maruyama, A., Merrey, D., Sally, H., de Jong, I., (2007). Costs and performance Finally, given the experience of collecting of irrigation projects: A comparison of data on establishment costs for SSI, we sub-Saharan Africa and other developing conclude that data management and regions. Colombo, Sri Lanka: centralized systematic data collection of SSI International Water Management investment and production data is definitely Institute. 81 pp. (IWMI Research Report one area where there could be significant 109). improvement in Ethiopia. Well organized Verdier,Jean.1992. Maintenance des data collection assists analyses that help Périmètres irrigués. Ministère de la inform decision making for policy makers. coopération et du développement, Paris. Werfring, A., 2004. Typology of Irrigation in References Ethiopia. A thesis submitted to the University of Natural Resources and Adams, W.M., (1990). “How beautiful is Applied Life Sciences Vienna. Institute small? Scale Control and Success in of Hydraulics and Rural Water Kenya Irrigation”, World development 18 Management in partial fulfillment of the (10):1039-1323. degree of Diplomingeieur. Central Inteligence Agency (CIA) (2007) World Bank., 2006. Ethiopia: Managing World Fact Book. water resources to maximize sustainable Bruns, B. (1991). Intervention in Small growth. A World Bank Water Resources Scale Irrigation: Some principles in Assistance Strategy for Ethiopia. The Assessment. World Bank Agriculture and Rural Paper presented at the Irrigation Studies Development Department. Report NO. center at Andalas University, West 36000-ET. Washington D.C. Sumatra, October 24, 1990. Published in Indonasian as “Intervensi pada Irigasi Slala Kecil: Beberapa Prinsip Penilaian”: pp 6-13. https://www.cia.gov/library/publications/ the-world-factbook/geos/et.html Accessed July 12, 2007. Government of the Republic of Ethiopia., 2001a . Water Sector Strategy, Ministry of Water Resources, Water Sector Development Programme, 2002- 2016. Irrigation Development Programme. Government of the Republic of Ethiopia., 2001b. Ethiopian Water Sector Strategy Government of the Republic of Ethiopia., 2006. National Accounts Statistics of Ethiopia. National Economic Accounts

126

Importance of Irrigated Agriculture to the Ethiopian Economy: Capturing the direct net benefits of irrigation

Fitsum Hagos1, Godswill Makombe1, Regassa E. Namara 2, and Seleshi Bekele Awulachew1 1International Water Management Institute (IWMI) Subregional Office for the Nile Basin and East Africa, Addis Ababa, Ethiopia 2International Water Management Institute (IWMI), Africa Regional Office. Accra, Ghana. [email protected]

Abstract total national GDP was about 5.7 and 2.5 percent during the 2005/06 cropping season. Irrigation development is seen as one of the means to reduce poverty and promote As a result of expansion, by the year economic growth. While a lot of effort is 2009/2010 the expected contribution of exerted towards irrigation development, little attempt is done to quantify the smallholder managed irrigation to national contribution of irrigation to national income economy, assuming that exiting cropping in Ethiopia. This study is an attempt to pattern, and the average gross margin values quantify the actual and expected for different crop categories are still valid, is contribution of irrigation to the Ethiopian expected to increase from USD 262.3 national economy for 2005/06 cropping million in 2005/2006 to about USD 414.2 season and 2009/10 using adjusted net gross million in 2009/2010, which accounts to margin analysis. about 5.5 percent of the agricultural GDP and 2.3 of the overall GDP for the same Our results show that irrigation in the study year. On the other hand, the contribution sites generates an average income of about coming from the large scale sugar growing USD 323/ ha. This compares to the estates in 2009/2010 is estimated to be USD calculated gross margin for rainfed which is 217.5 million which amounts to 2.9 and 1.2 USD 147/ha. This indicates that after percent of the agricultural and overall GDP accounting for annual investment respectively. Similarly the contribution replacement cost net gross margin from coming from large scale commercial farms irrigation is more than twice higher than growing crops other than sugar cane is gross margin from rainfed agriculture. On expected to increase to USD 35.8 million in the contribution of irrigation to national 2009/2010 which accounts to 0.4 and 0.2 economy, in 2005/06 smallholder irrigated percent of the agricultural and overall GDP agriculture contributed about 262.3 million respectively. This implies that large scale USD. This accounts for about 4.46 percent commercial farms will contribute about 3.3 of the agricultural GDP in 2005/2006 and and 1.4 of the agricultural and overall GDP 1.97 percent of the total overall GDP. The respectively. In summary, our results total income earned from large scale indicate that under conservative estimates schemes is estimated to be about 74.0 the future contribution of irrigation to million USD. This accounts for about 1.26 agricultural and overall GDP will be about 9 percent of the agricultural and 0.5 percent of and 3.7 percent respectively. the total GDP respectively. Overall, the contribution of irrigation to agricultural and When some of the assumptions related to cropping pattern, input and output prices,

127

and efficiency levels are relaxed, the Ethiopia is said to have an estimated contribution of smallholder managed irrigation potential of 3.5 million hectares irrigation to agricultural and overall GDP (Awulachew et al. 2007). However, the total will vary between 4 to 6 and 1.8 to 1.9 estimated area of irrigated agriculture in the percent respectively. Similarly, the country in 2005/2006 was 625,819 ha, contribution from large scale irrigation to which in total constitutes about 18 percent agricultural and overall GDP will be in the of the potential (MOWR, 2007). range of 3 to 6 and 1.2 to 2.5 percent respectively. Overall, the future contribution Irrigation is expected to contribute to the of irrigation to agricultural GDP will be in national economy in several ways. At the the range of 7 to 12 percent while the micro level, irrigation could lead to an contribution to overall GDP will be in the increase in yield per hectare and subsequent range of about 4 percent. To enhance the increases in income, consumption and food contribution of irrigation to national security (Bhattarai and Pandy, 1997; economy, besides increasing the presence of Vaidynathanet al., 1994; Ahmed and physical water infrastructure, however, there Sampath, 1992; Lipton et al. 2003; Hussain is a need to: i) improve provision of and Hanjra, 2004). Furthermore, Hussain agricultural inputs, ii) promote high value and Hanjra (2004), based on their studies in crops through the extension system, iii) Asia, indicated that irrigation benefits the create good market conditions, and iv) poor through higher production, higher increase the efficiency of small and large yields, lower risks of crop failure, and schemes. higher and year round farm and non-farm employment. Irrigation enables 1. Introduction smallholders to adopt more diversified cropping pattern, and to switch from low Unreliable rainfall, recurrent drought and value subsistence production to high-value limited use of the available water resources, market-oriented production (Hagos et al., coupled with heavy reliance on rain-fed 2007). subsistence agriculture, have contributed adversely to the economy of Ethiopia. In Macro level impacts manifest themselves fact, the World Bank (2006) estimates that through agricultural impacts on economic unmitigated hydrological variability growth. At the aggregate level irrigation currently costs the economy over one-third investments act as production and supply of its growth potential and leads to 25 shifters, and have a positive effect on percent increase in poverty rates. Hence, economic growth. Studies in Asia show that enhancing public and private investment in agricultural growth serves as an “engine” of irrigation development has been identified economic growth, and irrigation-led as one of the core strategies aimed to de-link technological changes are the key drivers economic performance from rainfall and to behind productivity growth in the enable sustainable growth and development agricultural sector (Hussain and Hanjra (World Bank 2006; MoWRa, 2002; MoWR, 2004; Alagh, 2001; Dhawan, 1988). Other 2002b; MOFED, 2006). In the government effects of irrigation on changes in the policy documents, irrigation development is environment and other social impacts have identified as an important tool to stimulate been reported in the literature such as on the sustainable economic growth and rural economic value of wetlands (Barbier and development and is considered as a corner Thompson, 1998); employment impact of stone of food security and poverty reduction irrigation (Berck and Hoffman, 2002) and (MoWRa, 2002; MoWR, 2002b; MOFED, non-farm sector benefits from irrigation 2006). investment (Bhattarai, et al. 2003).

128

The methodological approaches applied to 2002b; MOFED, 2006) and associated capture these diverse impacts of irrigation changes in cropping patterns. In so doing, varied from linear programming, to we tried to determine how much irrigation is regression models, to partial equilibrium contributing and will contribute to national models, to economy–wide models such as income relative to rain fed agriculture. This input-output models, Social Accounting method of adjusted gross margin analysis Matrices (SAM) and Computable General accords with the System of Environmental Equilibrium (CGE) Models. For instance, and Economic Accounts (SEEA) Bhattarai and Pandy (1997) used a linear recommendations (UN, 2003) and provides programming technique to isolate the impact a “best estimate” of the change in GDP of irrigation from other factors (such as road generated by irrigation at the farm gate and market) on crop production and (Doak, 2005). However, it should be noted productivity in Nepal. Vaidynathan et al. that a large number of estimates and (1994) used regression analysis at the assumptions are required to estimate the aggregate level to assess the difference in impact on GDP, and the results should be land productivity between irrigated and un interpreted with caution. In addition, the irrigated lands in India. Ahmed and Sampath increased output from irrigated farms will (1992) used a partial equilibrium model that have different multiplier effects in the wider incorporates demand and supply shifts to economy, so the total impact of irrigation on assess the impact of irrigation on efficiency GDP is likely to be higher than the farm gate and equity in Bangladesh. Makombe (2000) impact. used a similar partial equilibrium model to estimate the impact of irrigation induced For estimating the contribution of technological change in Zimbabwe. Bell smallholder irrigation we relied on data and Hazel (1980) used SAM and a semi collected during the 2005/6 season from input-out model to measure the magnitude eight smallholder irrigation schemes in and incidence of regional downstream four regional states in Ethiopia, namely effects of the Muda irrigation project in Amhara, Oromia, Southern Nations Malaysia. While there are various studies Nationalities and Peoples Regional that have tried to capture the diverse impacts Government (SNNPR) and Tigray. The data of irrigation, there are, however, few studies collected included command area, actual that attempted to capture the direct cultivated area cropping pattern, output contribution of irrigation to the national types and value, input use and input economy. One such study is by Doak et al., expenditure and information on the level of (2004) and Doak (2005) which develop a operation, (e.g. fully operational, medium, simple methodological framework to low or not operational). Moreover, we used measure the economic value of irrigation to secondary data gathered from selected large the New Zealand’s National economy. Our scale commercial farms in the Awash and study builds on the approach followed by Nile Valleys and price data and production Doak et al., (2004) and Doak (2005). data from the Central Statistical Authority (CSA). The objective of this study, hence, was to estimate the net contribution of irrigation to The research report is presented as follows. GDP at the farm gate. This study attempted Section two outlines the methodology used only to capture the direct benefits of to value the contribution of irrigation to the irrigation to national economy for a given national economy followed by presentation year (2005/2006) using a farm gate value of data sources in section three. In section approach and made forecasts on its future four we present an overview of the contribution based on the projected annual contribution of agriculture to national growth-rate of irrigated areas in the National economy while section five outlines the Irrigation Development Program (MoWR, hitherto irrigation development in Ethiopia

129

followed by, in section six and seven, by affected area provides an initial estimate of presentation of agricultural production and the GDP change (at the farm gate) likely to cropping pattern, both rainfed and irrigated occur as a result of irrigation. respectively. In section eight, we present the envisaged future expansion of irrigated In the Ethiopian context, farmers use full agriculture based on the National Irrigation irrigation to grow crops during the dry Development Program (IDP). Sections nine season when crop production from rain is and ten present the results of the valuation, not possible. This implies that households current and future, to national economy. In get additional income from irrigation to that section eleven, we conduct sensitivity what farmers get during the main cropping analysis to take account of possible changes season. Under small scale irrigation system, in cropping patterns and crop cover, in input irrigation does not replace rain fed and output prices and improvements in agriculture but supplements it. Large scale efficiency levels. The final part concludes schemes, however, are under full irrigation and draws some policy recommendations. throughout the year. We made adjustments in the methodology to take account of this 2. Methodology in valuing the difference between small and large scale contribution of irrigation for the irrigation. Hence, for a given farmer i under national income smallholder agriculture,

The methodology calculates the contribution NetIncomei = Netmarginirrig + Netmarginrf of existing irrigation to gross domestic (2), product (GDP) by taking into account the where the total income constitutes of income alternative rainfed production from the same from rainfed and income from irrigation. area of land. The method adopted follows a “with minus without” irrigation approach, The gross margins (GMs) were determined adjusted for changes in farm type and scale. for farm types in each of the schemes and aggregated to a scheme scale throughout Following Doak (2005) the formula is: Ethiopia based on the data obtained from the Farm gate GDP due to irrigation = GDP household surveys and secondary sources. with irrigation – GDP without irrigation. The gross margins are those for the 2005/06 = (irrigated land use mix * (irrigated Gross season and are defined as the revenue Margin – fixed costs)) – (rainfed land use generated from the activity less the direct mix * (rainfed Gross Margin – fixed costs)). costs of producing the revenue. The Gross Margins were adjusted to account for the (1) differences in overheads (fixed costs) of land uses with and without irrigation, and A gross margin is the total revenue also for differences in shadow prices of associated with a particular production labor and oxen in irrigated and rain fed (income) less the costs that clearly vary in systems (for the small scale schemes). direct proportion to the level of production - Shadow prices of labor and oxen were the direct or variable costs associated with estimated from the production data by first the enterprise. Gross margin analysis is an estimating labor elasticity, which was used accepted tool commonly used in the to estimated the marginal value of labor, in a evaluation of farming enterprises. It has production function framework (for details been used for the evaluation of the costs and see Jacoby, 1993). benefits of irrigation in cost benefit analysis. Assessing the change to the gross margin The “without irrigation” land use is that per unit area as a result of irrigation and then which would now exist if irrigation had not scaling this appropriately by the total been developed, rather than if irrigation was

130

no longer available for that particular land. years) and O & M was assumed to be 10 This was estimated based on GM of rain fed percent of annual replacement for small agriculture around the scheme or average scale schemes and 50 yrs and 5 percent for GM value for all rainfed, if data for adjacent large scale schemes (Inocencio et al., 2007). rainfed plots were not available. The value of irrigated production and the value of In estimating the future contribution of production from the rainfed use that would irrigation to national economy, we used be most likely if there was no irrigation were information about the expected growth of derived from the survey data for each the irrigation sector during 2005/2006 to scheme. For the large scale schemes, we 2009/2010 based on the country’s Irrigation explored the dominant rainfed production Development Program (IDP) (MoWR, 2006; type and estimated average gross margins World Bank, 2006; MOFED, 2006). These per ha from the household survey. policy documents outline how irrigation is expected to develop over the planning The assumption here is that all of the now period. The details are provided in section irrigated lands would have been under some eight of this paper. sort of rain-fed farming had it not been converted to irrigation plots. There are also A complex issue related to the calculation of some other possible scenarios. It is possible the future contribution of irrigation to the that some of the current irrigated lands are national economy is how to address the hitherto uncultivated lands or new impact of increased output on price. Gross openings5. If this is true, the methodology margin calculations generally assume that a we adopted may underestimate the true change in output has no effect on prices. contribution of irrigation development While for small-scale changes at the without considering the environmental costs individual farm level this may well of such changes. It may be that the current approximate the truth7, the large-scale land irrigated land may have been used for use changes generated by irrigation on the grazing livestock6. The direction of bias on national scale are believed to be sufficient to our estimation depends on whether the gross have some measurable effect on output margin per unit area from livestock rearing prices. Lipton et al. (2003) state that if is greater or less than the gross margin per irrigation leads to increases in staples or unit area for cropping under rainfed. While a non-staple food output then this may result meaningful analysis should take account of in lower prices for staples and food in these diverse scenarios, the lack of data on imperfectly open economies or if there are livestock productivity under pastoral significant transport costs from food-surplus production in Ethiopia and environmental area to towns or food deficit areas. For crops costs of land use change made it impossible. that are largely dependent on the local Hence, the approach described above (in market and for which there is little equations 1 and 2) was used to assess the opportunity to develop large-scale export current and future contribution of irrigation markets increases in production tend to have to the national economy. a dramatic effect on price (Doek, et al., 2004). A complicating factor in assessing For the fixed cost, we calculated an annual the impact of future irrigation-driven replacement cost all on per ha basis. Annual increases in output on price is also that replacement cost was computed as initial growers of annual crops are very flexible in investment divided by project lifetime (25 the combinations of crops that they choose to grow (Doak et al., 2004). If, for example, 5 The development of Finchaaa Sugar estate is a case in point. 7 Even at the small scale, we observe increases in 6 The development irrigation in the middle and crop output of tomato and onion leading to lower Awash is a case in point. crashes in prices.

131

potatoes are in over supply, growers would This study made use of both primary data on switch to another crop which proves more smallholder production collected from profitable. The farmer is, therefore, able to household surveys and data from various choose the most profitable product to secondary sources. The household survey produce, and to increase the value of the was part of a comprehensive nationwide product e.g., by producing at a time of the study on the impacts of irrigation on poverty year when price is highest, or by increasing and environment run between 2004 and the quality of the product (for example, 2007 in Ethiopia. It was a component of the through improved fruit size). There is also Impact of Irrigation on Poverty and the possibility that as irrigation expands, it Environment (IIPE) research project run by tends to get more government support e.g. the International Water Management extension and hence intensification can Institute (IWMI) with support from the increase. This upside potential has by and Austrian government. The survey, which large been included in the analysis. We investigated the impact of irrigation on suggested possible scenarios in changes in poverty and irrigation’s contribution to cropping patterns. However, it is difficult to national economy, addressed a total sample exactly forecast the possible future changes size of 1024 households from eight in cropping patterns. The crop combinations irrigation sites in 4 regional states involving and gross margins used in the analysis are, traditional and modern and rainfed systems. therefore, only indicative of a range of The total sample comprised 397 households possible crops with similar outcomes. practicing purely rainfed agriculture and 627 households (382 modern and 245 To quantify the price effect of irrigation traditional) practice irrigated agriculture. development we assumed different price These households operate a total of 4,953 scenarios based on certain assumptions plots (a household operating five plots on about demand growth and output growth. In average). The data collected include the light of all these considerations, we demographics, asset holdings, access to assumed different price changes in price of services, plot level production and sale and the major produce when assessing the input use data (distinguished between impact of future irrigation driven increases irrigated and rain fed), constraints to in output. This is described in detail in agricultural production and household section ten of this report. perceptions about the impact of irrigation on poverty, environment and health and other Finally, there are a host of multiplier effects household and site specific data. The data expected to manifest themselves with were collected for the 2005/2006 cropping irrigation development, including expansion season. We used part of this comprehensive of the off-farm sector, provision of inputs to dataset for the analysis here. industry and better nutrition for rural households. These effects are not captured We also used secondary data from various in this study. Our calculated GDP sources. From the large scale schemes we represents, at best, the return to producers’ gathered data on investment cost/initial labor and capital (including capital tied up in capital outlays, cost of production, output land). It is also worth noting that the high and revenue among others. From official income sector of irrigation (emerging flower documents such as the policy documents of farming and capital intensive commercial the government (MoWR, 2006; World farms are not included in our assessment. Bank, 2006; MoFED, 2006) we gathered Our method therefore underestimates the developed and projected irrigation true contribution of irrigation to GDP. development plans and we used land utilization and crop cover data from the 3. Data sources Central Statistics Agency (CSA, 2006). Furthermore, for specific data on future

132

expansion and new development plans on Agriculture is the main stay of the Ethiopian sugar estates we used the revised master economy. It is major contributor to the plan of the Ethiopian Sugar Development national economy both in terms of income, Agency (ESDA, 2007). The plans for the employment and generation of export development of small scale irrigation are revenue. Its contribution to GDP, although prepared by the regional governments and showing slight decline over the years, has are compiled by the Ministry of Agriculture remained very high, about 44 percent. From and Rural Development that oversees the among the sub-sectors of agriculture, crop development of the sub-sector. production is major contributor to GDP accounting for about 28 percent in 2005/06. 4. Contribution of agriculture to The most important crops grown and their national economy area coverage are described in section six and seven.

Table 1: Contribution of Agriculture to GDP (in 000 Birr) (1995/96-2005/06) Agri Crop GDP at Current Agricultural contribution contribution Year Market Prices GDP Crop GDP to GDP to GDP

1995/96 53,597,593 28,613,235 17,286,203 0.53 0.32

1996/97 55,520,011 28,767,766 16,764,422 0.52 0.30

1997/98 53,391,285 25,214,701 14,505,336 0.47 0.27

1998/99 57,368,203 25,397,662 15,500,013 0.44 0.27

1999/00 64,397,933 28,444,382 17,713,717 0.44 0.28

2000/01 65,687,343 27,750,560 16,333,285 0.42 0.25

2001/02 63,461,569 24,460,704 13,135,220 0.39 0.21

2002/03 68,898,037 26,207,930 14,963,341 0.38 0.22 2003/04 81,754,514 32,229,991 19,746,954 0.39 0.24

2004/05 98,397,946 42,196,370 27,349,050 0.43 0.28 2005/06 115,589,480 50,893,906 32,246,432 0.44 0.28 Source: FDRE (2006).

irrigation land (Awulachew et al. 2007)8. 5. Overview of irrigation development The total estimated area of irrigated in Ethiopia agriculture in the country is about in 2005/2006 was 625,819 ha, out of which Ethiopia is said to have an estimated 483, 472 is from the traditional irrigation, irrigation potential of 3.5 million hectares of 56032 ha is from modern small scale,

8 Other estimates put it in the order of 3.7 million hectares (MoWR, 2002; World Bank, 2006).

133

86,612 ha is from modern medium and large schemes are not considered, the irrigated scale schemes. Out of the total irrigated land area covers a minimum of about 1.6 area, 197,250 ha is covered by the so-called percent of the total cultivated area. modern schemes while the remaining area There is high spatial variability in water traditional schemes (MoWR, 2002). The resources endowment and development in total and modern irrigated area account for the country. Hence, ninety percent of the about 17 and 5 percent of the potential country’s water resources development respectively. The total cultivated land area, occurs in four river basins (World Bank, rainfed included, in 2005/06 was about 2006). Much of the formal irrigation 12.28 million hectares (MOFED, 2006). The developments are located in the Awash total current irrigated land area, hence, Basin, where about 50 medium- and large accounts for about 5 percent of the total scale irrigated farms are located (Fig. 1). cultivated land. When the traditional

Irrigation Schemes by River Basin

MEREB GASH Existing Irrigation Schemes

TEKEZE Large Scale DENAKIL Medium Scale Small Scale Rivers

AYISHA Lakes ABBAY AWASH

BARO AKOBO OGADEN

WABI SHEBELE OMO GIBE

RIFT VALLEY

GENALE DAWA

0150 300 600 · Kilometers

Fig.1: Existing Irrigation Schemes in various river basins in Ethiopia (Source: Awulachew, et al. 2007)

In terms of regional distribution, Afar and crops, stimulant crops9 and sugar cane. Oromia have the bulk of the share in Cereals are the dominant food crops irrigated agriculture accounting for 45 and covering 58 % of the land area10 and 87 % 31 percent of the total irrigated area. of the volume of grain production11 (See Fig Amhara, SNNPR and Tigray account for 8, 2). The major cereal crops include: teff 7 and 5 percent of the total irrigated area (Eragrostis tef), barley (Hordeum vulgare), respectively (Awulachew, et al. 2007). wheat (Triticum durum), maize (Zea mays), . sorghum (Sorghum bicolour) and finger 6. Agriculture production and

cropping pattern 9 Stimulant crops consist of Chat, coffee and hops. Based on Central Statistics Agency’s 10 2005/06 agricultural sample survey (CSA, Total cultivated land area during 2005/06 cropping season was estimated at 12.28 million 2006), the major crops during the main rainy hectares. season (a.k.a meher season) are cereals, 11 Total volume of agricultural produce during pulses, oilseeds, vegetables, root crops, fruit 2005/06 cropping season was 133. 1 million quintals. A quintal is equivalent to 100 kgs.

134

millet (Eleusine coracana). Teff, maize, sorghum made up 25, 17, 16 and 16 percent sorghum and wheat took up 22, 15, 14 and of the grain production in the same order. 14 percent of the grain crop area, respectively. Maize, wheat, Teff and

Crop cover rainfed (CSA data)

cereals pulses 2%0%0% 7% 1%2%0%1% oil seeds 12% vegetables root crops fruit crops chat 75% coffee hops sugar cane

Fig 2: Crop cover during the Meher season of 2005/2006 (Source: CSA (2006))

Pulses grown in 2005/06 covered 12.7 % of Root crops covered more than 1.6% of the the grain crop area and 9.5 % of the grain area under all crops in the country. Potatoes, production. Faba beans (Vicia faba), field sweet potatoes and taro covered 36.5, 29.7 peas (Pisum sativum) and chick-peas (Cicer and 15.1 percent of the area to the root crop arietinum) were planted on 4.5, 2.29 and total. The same crops and onion contributed 1.98 percent of the grain crop area. The 33.7, 30.6, 12.9 and 13.2 percent to the root production obtained from the same crops crop production in the same order. was 3.8, 1.4 and 1.6 percent of the grain production. More than 45 thousand hectares of land is under fruit crops in Ethiopia. Bananas Oilseeds comprised 7.8% of the grain crop contributed about 62.4% of the fruit crop area and 3.6% of the production to the area followed by mangoes that contributed national grain total. Neug (Guizotia 12% of the area. Nearly 4.3 million quintals abyssinca), linseed (Linum sativum) and of fruits was produced in the country in sesame (Sesamum indicum) covered 3.0, 2.1 2005/2006. Bananas, Papayas, mangoes and and 2.0 of the grain crop area and 1.1, 0.9 oranges took up 49.4, 16.6, 12.8 and 11.8 and 1.1 of the grain production. percent of the fruit production, respectively.

Vegetables took up 1.1% of the area under The area and production of chat and coffee all crops at national level. Of all the area are larger than that of fruits since they earn a under vegetables 69 and 19 percent was considerable amount of cash for the holders. under red peppers and Ethiopian Cabbage Chat (Catus adulis) and coffee shared 1.24 (Brassica carinata), respectively. As to and 2.39 percent of the area under all crops production of vegetables, 39.8 and 40.2 in the country and 1.2 and 1.7 million percent was that of the same crops. quintals of produce was obtained from these

135

crops in the same agricultural year between traditional and modern irrigation respectively. schemes while looking into cropping composition. We clustered crops into Sugar Cane is grown on about 19 thousand different categories; namely, cereals, pulses, hectares of land in the country, yielding 16.1 oil seeds, spices, vegetables, fruits and million quintals of produce by the peasant others and calculated area cover (as holders. percentage of the total area) for these different crop categories in the different 7. Irrigation typologies and cropping systems. The dominant crop categories patterns under traditional irrigation system, in terms of the percentage area covered are: cereals In the Ethiopian context, the irrigation sub- (55%), vegetables (11%), fruits (11%), sector is classified as small (less than 200 pulses (10%), spices (8%), oil seeds (5%), ha), medium (200 to 3000 ha) and large- and others (0.2%) (Figure 3a). In the modern scale (over 3000 ha) schemes (MoWR irrigation systems, in the order of 2002a; Awulachew et al., 2005). Small scale importance, the dominant crops are: cereals irrigation schemes are considered as (67%), vegetables (17%), fruits (4%), pulses traditional if the diversion weirs are made (3%), spices (0.2%), oil seeds (0.4%), and from local material which needs annual others (5 %) (Figure 3b). reconstruction. The canals are usually earthen and the schemes are managed by the Under the rainfed agricultural system the community. Many are constructed by local dominant crops are: cereals (77%), pulses community effort and have been functional (16 %), vegetables (1.3 %), fruits (1%), oil for relatively longer periods of time. On the seeds (1%), spices (0.4%), and others (3.3 other hand, small scale schemes are %) (Figure 3c). considered as modern schemes if they have more permanent diversion weirs made from concrete, and the primary and sometimes secondary canals are made of concrete. They are generally community managed and have recently been constructed by government or NGO. Werfring (2004) and Makombe et al. (2007) describe the typology of small scale irrigation in Ethiopia, the former in more detail. Small and medium scale schemes grow cereals as main crops. During the main rainy season most of small and medium scale irrigation schemes grow cereals like teff, maize and barley, with little or no supplementary irrigation, under rain fed conditions. During the dry season farmers grow cereals and a variety of vegetables including onions, tomatoes, and leafy green vegetables like lettuce under full irrigation. Farmers also grow perennial crops like mango, banana, sugar cane which are sometimes intercropped with seasonal crops.

From our survey data, we present below the composition of crops under irrigated and rain fed conditions. We made the distinction

136

Crop cover in traditional irrigation schemes Crop cover under moder irrigation schemes

Cereals Cereals 4% 5% 0% 11% Pulses 8% Pulses 21% Oil seeds Oil seeds 11% Spieces Vegetables Vegetables 0%3% 55% Spices 67% Fruits 5% Other Others 10% Fruits

Fig. 3a: Dominant crops under traditional irrigation system (n= 1240) Fig. 3b: Dominant crops under modern irrigation system (n= 2092)

Crop cover under rainfed system 1% 3% 1%1%0%

16% Cereals Pulses Oils seeds Fruits Spices vegetables Other

78%

Fig. 3c: Dominant crops under modern rain fed system (n= 1533)

The figures above show that there is traditional and modern systems. There is difference in the relative importance of the also noticeable difference in the share of crop categories under different systems. land taken by vegetables and fruits between Cereals and pulses are dominant under the modern and traditional irrigation rainfed system while vegetables and fruits systems. Vegetables take more land area cover about 2 percent of the land area. under the modern systems compared to that While cereals still remain dominant under of traditional systems while more land area the irrigation systems, covering about 61 is covered with fruit tress under the percent of the land area, vegetables and traditional system indicating perhaps the fruits become important under both

137

difference in age between the two type of and Finchaa schemes grow sugar cane, systems. while the Amibara and Upper Awash Large-scale irrigation schemes, on the other schemes grow cotton and Fruits and hand, grow mainly sugar cane, cotton and vegetables respectively (see Table 2) fruits and vegetables. Wonji/Shoa, Metehara

Table 2: Large scale schemes under irrigation and type of cropping Region Scheme name Major crop Area 2005/06 (in ha) Afar Amibara ( Middle Awash ) Cotton 6448 Oromia Finchaa sugar plantation Sugar cane 7185 Oromia Metehara sugar plantation Sugar cane 10145.9 Oromia Upper Awash Vegetables & fruits 6017.34 Oromia Wonji/Shoa sugar plantation Sugar cane 4094 Source: ESDA (2007); MOFED (2006);

8. Future expansion of irrigation in the Awash basin which will mainly development in the country involve expansion of the already existing large scale schemes or development of new The Irrigation Development Program (IDP) ones (see Table 3). About 90,000 ha of as set out in the government’s Plan for irrigation land will be developed in Kesem Accelerated and Sustained Development to and Tendaho to grow sugar cane while there End Poverty (PASDEP) document (2005/06- are planned expansions in the already 2009/10) envisages the expansion of existing sugar plantations. Overall, by the irrigation in the country by an additional year 2010 there will an additional 122,000 430,061 ha by the year 2010 (MoWR, 2006; ha of irrigated land developed to grow sugar MOFED, 2006). This will consist of mainly cane (ESDA, 2007). medium and large scale schemes. Accordingly, 39 significant irrigation There are also parallel plans to develop projects are planned to be implemented 98,625 ha under small scale irrigation by the during the PASDEP period. These include regional governments (Atnafu, 2007). The World Bank project around Tana (100,000 total extension to irrigated area by the year ha); Anger Negesso Project in Oromia 2009/10 compared to 2005/2006 will be in (49,563 ha); Humera project in Tigray the range of 528,686 ha. This implies that (42,965 ha); Kessem Tendaho in Afar further development will extend the irrigated (90,000 ha); Upper Beles in Benishangul area to about cover 33 percent of the Gumz (53,000 ha) and Ilo-Uen Buldoho irrigated potential and about 9 percent of the (32,000 ha) in Somali (MOFED, 2006; total cultivated land area. These plans are MoWR, 2006). Most of these irrigation used as indicative targets for future schemes will be community managed irrigation development for valuing the future schemes to be used by small scale farmers. contribution of irrigation to the national Exceptions are the schemes to be developed economy.

Table 3: Future development plans of large scale schemes Region Scheme name Main crop Future expansion/development until 2010 (in ha) Oromia Finchaaa Sugar cane 12000 Afar Kesem Sugar cane 40000 Oromia Metehara Sugar cane 10000 Afar Tendaho Sugar cane 50000 Oromia Wonji/Shoa Sugar cane 10000 Source: ESDA (2007)

138

9. Value of irrigation to national modern schemes compared to the traditional economy ones. The relatively longer irrigation experience and, hence, acquired improved The contribution of agriculture to national irrigated crop management practices of economy is estimated on the basis of the farmers working and the composition of estimated production during the Meher crops grown under traditional system may (main rainy season) and the Belg seasons also contribute to this difference. The (small rainy season) (BOFED, 2006). We development of modern irrigation schemes assume that the contribution from irrigation is a recent phenomenon in Ethiopia. is included in the production during the Belg season, although not explicitly stated in the There are also huge inter-scheme differences document. Farmers use full irrigation to in income within the same typology which grow crops during the dry season when crop could be attributed to relative difference in production using rain is not possible. This irrigation experience and access to market. implies that household’s get additional When it comes to medium scale irrigation income from irrigation in comparison to schemes, the average income from modern farmers who can only grow during the main irrigation schemes was USD 400/ha. We do rainy season. Under small scale irrigation not know the corresponding figure for a system, irrigation does not replace rainfed traditional medium scale scheme as we did agriculture but supplements it. Large scale not have such a scheme in our sample. We schemes, however, are under full irrigation assumed in this study that the average net through out the year. income from a traditional medium scale scheme is USD 400. Based on the net gross margin calculations (see table 4), irrigation in the study sites generates an average income of about Birr 2800 /ha, which is equivalent to USD 323/ ha12. This compares to the calculated gross margin for rainfed which is USD 147/ha. This indicates that after accounting for annual investment replacement cost, net gross margin from irrigation is more than two times higher than gross margin from rainfed agriculture.

When we disaggregate net income by irrigation typology, we also see a strong difference between the categories. Average income from small scale but modern schemes is about USD 355/ha while from small scale traditional is about USD 477/ha. This may sound counter intuitive in the sense that schemes with permanent structures and well lined canals should have led to better returns. The reason for higher margins for traditional schemes may have to do with high average investment cost of

12 1 USD was equivalent to 8.67 ETB in 2005/06 prices.

139

Table 4: Gross margin calculation from small and medium scale irrigation schemes GM/ha GI Scheme Scheme Area rain GM/ha minus Total name scale Typology (in ha) AFC fed irrigated FC income Modern/ Indris Medium traditional 382 74 429 1850 1776 678571 Gologota Medium Modern 850 303 1068 7596 7293 6199193 Modern/ WBS Medium traditional 685 200 1485 2603 2402 1645656 Tikurit Small Traditional 102 91 1353 4140 4050 413081 Zengeny Small Modern 270 222 1971 3375 3152 851160 Haiba Small Modern 250 437 1585 2795 2358 589537 Golgol Micro Raya irrigation Modern 104 1372 1710 2240 868 90280 Modern/ Hare Medium traditional 1345 159 646 950 791 1064206 Source: Own calculation

Taking the average income from irrigation to be Birr 2.27 billion (about 262.3 smallholder managed small and medium million USD). This accounts for about 4.46 scale irrigation schemes in the country and percent of the agricultural GDP in the total hectarage for both categories, we 2005/2006 and 1.97 percent of the total calculated the total income driven from GDP.

Table 5: Gross margin calculation from large scale irrigation schemes Area Average Annual GM Total Scheme (in investment recovery Total GM per Net income name Main crop ha) cost/ ha cost/ ha (in million) ha income Amibara* cotton 5358 11418 228 13.8 1212 984 5270343 Finchaa sugar cane 7185 62672 1253 184.4 2943 1689 12137261 Metehara sugar cane 10146 9303 186 303.7 32649 32463 32936660 fruits & 99396889 Upper Awash vegetables 6017 3793 76 62.9 16594 16518 Wonji/Shoa* 47040510 * Sugar cane 4094 35987 720 439.3 12210 11490 * Based on 2004/2005 estimate ** Average investment cost for Wonji is taken as the average for Metehara and Finchaa

When it comes to valuation of the any possibilities to practice rain fed contribution from large scale schemes, we agriculture. Before netting out the followed strictly the approach outlined in contribution coming from rainfed, the section two. Hence, in calculating net average income from large scale schemes income from large scale schemes we was USD1456/ha. There are strong deducted the contribution of rainfed from differences in GM between the schemes, the net income obtained under irrigation to however. As we did not have data from account for the income foregone for not rainfed in and around the large scale using the land under rainfed production. The schemes, we used rainfed data from the rationale behind this is that irrigation in the medium and small scale sites. The large scale schemes is full season devoid of calculated average gross margin per ha from

rainfed agriculture, as indicated earlier, was irrigation to national income is still very USD 147. Taking this value into account, small compared to the 28 percent the netted income from a hectare of contribution of crop production. Regional irrigation under large scale schemes is USD comparisons could shed some light on this. 1308. When we differentiate the large scale In the Sudan, for instance, irrigation schemes into sugar plantation and other crop contributes about 50 percent of the crop growing plantations (i.e. predominantly production while almost all agriculture in vegetables and fruits and cotton growing Egypt is irrigated (FAO, 1997). schemes) the average net income is USD 1782.5 and USD 998.9 respectively. Taking 10. Projecting future contribution of the all large scale schemes in the country, irrigation differentiated by their cropping pattern, and the average income from the selected In this section we present the projected learning sites, the total income earned from expansion of irrigated agriculture vis-à-vis large scale schemes amounts to Birr 641 rain fed agriculture and the contribution of million (ca 74.0 million USD). This the former to agricultural GDP. To set the accounts for about 1.26 percent of the future scenario we used cropping patterns as agricultural GDP and 0.5 percent of the total observed in our empirical results and GDP respectively. When only the improved projected cropping patterns of the PASDEP system is considered, it contributed to about (2005/06-2009/10) document (see Table 7). 1.26 and 0.5 percent of the agricultural GDP The projected irrigation development, both and GDP respectively. Overall, the small-and medium scale and large scale contribution of irrigation to agricultural schemes is taken into account in setting the GDP and total national GDP was about 5.7 future scenario. and 2.5 percent during the 2005/06 cropping season. This shows that the contribution of

Table 6: Cropping pattern under different systems (% area covered by) by small and medium scale irrigation

Crop category Area under Rain Area under Area under Average area fed system (%) Traditional Modern Cover under irrigation (%) irrigation (%) irrigation (%)

Cereals 77 55 67 61

Vegetables 1 11 21 16

Perennials/fruits 1 11 4 7

Pulses 16 10 3 6

Oil seeds 1 5 0.4 3

Spices 0.5 8 0.3 4

Others 3 0.2 5 2.5

141

Source: Own calculation

In projecting future scenarios we assumed Ethiopian economy is expected to grow at that the cropping pattern of the large scale an average of 7.3 percent through out the sugar plantations to be the same. We ruled PASDEP period. Agriculture, the major out reductions of irrigated land due to sector of the economy is also expected to salinity or other environmental damages in grow at an average rate of 6.2 percent those sugar plantation for lack of data that (MOFED, 2006, p. 55). Agriculture’s share clearly shows the magnitude of the problem to the economy will show slight reduction or how effective are the ameliorative from 46.2 percent in 2004/2005 to 43.9 measures undertaken by these schemes. On percent at the end of the planning period. the other hand, we assumed that the Taking the baseline situation (2005/06), cropping pattern in the smallholder managed Ethiopia’s GDP will grow to Birr 153.2 large, medium and small scale irrigation billion while agricultural GDP will grow to schemes to be the same as depicted in Table Birr 64.7 billion both at 1999/00 constant 6. The land cover statistics of the irrigation, basic prices. all typologies considered, and rainfed systems are also given in Table 7. We For the assumptions about the IDP relaxed this assumption later in the differentiated into small-medium scale & sensitivity analysis as it is realistic that large scale we used MOFED (2006) and farmers will shift to high paying crops as MoWR (2006), as indicated in section eight. they gain experience and the market As the national IDP indicates the country’s situation likely to improve. irrigation coverage will increase from the current 625,819 ha to 1.15 Million hectares. Table 7: Land use assumptions for future Accordingly, there will be 638,129 ha of irrigated areas (2005/06-2009/2010) small scale irrigation, both traditional and Land use Area with Area improved, 328,485.9 ha of smallholder irrigation without managed medium and large scale irrigation (in 000 ha) irrigation schemes and 122000 ha of large scale (in million schemes dedicated for sugar plantations and ha) 35511 ha of large scale commercial farms Cereals 809.2 9.2 dedicated to growing of vegetables, fruits and cotton. Pulses 1.6 Oil seed 119.4 1.2 Taking all the envisaged areal expansion, crops crop cover assumptions as indicated in Vegetables 212.2 Tables 7 and 8 and the average gross margin Fruits 99.5 0.419 by crop category (table 9), we calculated Cotton 0.043 that the contribution of smallholder Sugar cane 122.0 0.060 managed irrigation to national economy to Coffee 0.734 increase from USD 262.3 million in Floriculture n.a. 0.002 2005/2006 to about USD 414.2 million in Tea 0.0038 2009/2010, which accounts to about 5.5 Other 86.2 0.039 percent of the agricultural GDP and 2.3 of Total 1326.5 12.65 the overall GDP for the same year. On the Source: MOFED (2006) and own other hand, the contribution coming from calculation; n.a.= no data available the large scale sugar growing estates is in 2009/2010 is estimated to be USD 217.5 The PASDEP document also outlines the million which amounts to 2.9 and 1.2 projected development of the economy for percent of the agricultural and overall GDP the whole planning period. Accordingly, the respectively. Similarly the contribution

142

coming from large scale commercial farms It is realistic to assume that there could be growing other crops other than sugar cane is either upward or down ward movements in expected to increase to USD 35.8 million in prices of agricultural inputs and outputs. 2009/2010 which accounts to 0.4 and 0.2 Furthermore, the efficiency of farmers is percent of the agricultural GDP and overall also expected to improve with time as they GDP respectively. This implies that large gain irrigation experience and experiment scale commercial farms will contribute with various technologies and combinations. about 3.3 and 1.4 of the agricultural GDP Hence it is important to relax these and overall GDP respectively. This shows assumptions and see the effect of these that the bulk of the contribution is expected changes on irrigation’s contribution to to come from smallholder managed national income. This section presents the irrigation systems. In summary, this results of the sensitivity analysis. indicates that under conservative estimates the future contribution of irrigation to 11.1 Simulating changes in cropping agricultural GDP and overall GDP will be in patterns under smallholder managed the range of 9 and 3.7 percent respectively. irrigation schemes This estimation is based on the projected areal expansion, current cropping patterns To simulate the effect of such change in and prices. These results are likely to change cropping pattern on the agricultural GDP we when some of the assumptions were allowed set the following scenarios: Scenario 1 to change as shown below. involves 10 percent increase in area coverage of vegetables and fruits (10 percent 11. Sensitivity analysis decrease in area for cereals) while areas for pulses and oil seeds and other crops remain In projecting the future contribution of the same; Scenario 2 assumes 10 percent irrigation to national economy or increase in area of vegetables and 5 percent agricultural GDP our assumptions were in fruits (15 percent reduction in area for rigid: only a change in area expansion was cereals cetaris paribus) and Scenario 3 assumed. However, it is realistic to assume assumes 10 percent increase in area for both that there will be various changes associated vegetables and fruits (20 percent reduction with irrigation expansion. For instance, in area for cereals) and finally scenario 4 given the significant difference in the gross assumes a 25 percent increase in area of margin between different crop categories, vegetables and fruits (i.e. 25 percent farmers will benefit economically from reduction in area for cereals ceteris paribus). growing more vegetables and fruits than The outcomes of these scenarios were growing cereals. Hence it is realistic to compared against the baseline scenario assume that farmers will gradually shift to where we assumed that there will be only high value crops. Prices of inputs and aerial expansion (Table 11). outputs cannot be taken to remain constant.

Table 8: Estimated average gross margin for different crop categories Average Gross margins by crop category Birr/ ha - rain fed Birr/ ha - irrigated Cereals 1282.32 1720.84 Vegetables 3421.2 Fruits 2754.9 Pulses & oil seeds 1481.45 1558.19 Sugar cane 4528.7 Cotton 709.6 others (hops, chat, etc) 1254.8 2891.8

143

Source: Own calculation

As can be seen from Table 8 (See also Table cereals, Birr 3421 from vegetables, Birr 10a and 10b in Annex) there is significant 2755 from fruits, Birr 1558 from pulses and difference in the gross margin between oil seeds, and Birr 1719 from growing other different crop categories. On average crops such as spices and stimulants. farmers get Birr 1720 per ha from growing

144

Table 9: The effect of change in cropping pattern on the projected contribution of small holder managed irrigated agriculture to Agricultural GDP (Net gross margin in BIRR) Contribution to Contribution to Total NGM AgGDP in GDP in Relative Crop type (in Million USD) 2009/2010 (%) 2009/2010 (%) change (%) Baseline 315.2 4.22 1.78 Scenario 1 327.9 4.39 1.85 17 Scenario 2 335.8 4.5 1.9 28 Scenario 3 340.6 4.56 1.92 34 Scenario 4 3.84.5 4.67 1.97 45

As can be seen from the simulation results, economy, we set various scenarios: baseline the contribution from smallholder managed scenario GM net of annual investment irrigation schemes to Agricultural GDP recovery cost; 10 percent increase in price of increases to about 4.5 percent or even more vegetables and fruits ceteris paribus when these various changes in cropping (scenario 1); 15 percent increase in price of patterns are assumed. 15 and 10 percent vegetables and fruits ceteris paribus increase in the area of vegetables and fruits (scenario 2); 10 and 15 percent increase in (25 percent reduction in the area of cereals) price of cereals ceteris paribus (scenarios 3 lead to about 45 percent increase in the and 4); 10 and 15 percent increase in the contribution of smallholder irrigation to price of pulses and oil seeds ceteris paribus agricultural GDP as compared to the (scenarios 5-6); and 10 and 15 percent baseline scenario. This is an important result increase in price of other crops ceteris indicating that the contribution of irrigation paribus (scenario 7 and 8). The simulation could be maximized if smallholder farmers results are reported in table 10 below. shift their cropping pattern to high value crops. Hence, the extension system could These simulation results show that a 10-15 play an important role in providing and percent increase in the price of fruits and promoting high value crops. vegetables leads to 15-23 percent increase in the relative contribution of smallholder 11.2 Simulating changes in crop prices irrigation to agricultural GDP. An equivalent increase in the price of cereals leads to 22- The factors that influence price change 32 percent increase in the relative could be related to overall demographic contribution of the sub sector. One the other change and improved economic hand, the same level of increase in prices of performance (through increased demand) pulses, oil seeds and other crops did not and increase in supply of output. It is yield significant change in their reasonable to assume that the population of contribution. The relative higher Ethiopia will continue to grow in the contribution from cereals comes from the foreseeable future while there could be bigger share cereals have on land cover differences in opinion about the prospects of claiming about 61 percent of the cultivated economic growth in the country. The area under irrigation, Hence, vegetables and prospects point towards improved economic fruits are economically more attractive. This performance, however. For this exercise, implies that an increase or decrease of prices hence, we assumed that demand factors will of vegetables and fruits will have a stronger play a more significant role in influencing relative impact on the contribution of the price of outputs.. To simulate the effect irrigation to national economy compared to of these changes in prices of output on the the price change of cereals. contribution of irrigation to national

145

Table 10: the effect of change in output prices on the projected contribution of small holder managed irrigated agriculture to Agricultural GDP Relative Contribution to AgGDP in Contribution to GDP change Scenarios Description 2009/2010 (%) in 2009/2010 (%) (%) GM net of investment Baseline recovery cost 4.22 1.8 10 % increase in price of Scenario 1 vegetables & fruits 4.37 1.85 15 15 % increase in price of Scenario 2 vegetables & fruits 4.45 1.88 23 10 % increase in price of Scenario 3 cereals 4.44 1.87 22 15 % increase in price of Scenario 4 cereals 4.55 1.92 32 10 % increase in price of Scenario 5 pulses and oil seeds 4.25 1.79 3 15 % increase in price of Scenario 6 pulses and oil seeds 4.26 1.80 4 10 % increase in price of Scenario 7 other crops 4.25 1.79 3 15 % increase in price of Scenario 8 other crops 4.26 1.80 4

11.3 Simulating changes in input prices 47 respectively. In projecting the impact of irrigation on national economy, one needs to Fertilizer is the most important input for consider the effect of changes in input prices smallholder farmers working under on the gross margin. To simulate such an irrigation. The average cost of fertilizer effect, we determined the impact of the varies by type of crop category. Cereals and following scenarios: 10, 15, 25 and 35 vegetables are major consumers of fertilizer percent increase in the price of fertilizer. with average expenditure per hectare of Birr Given the current trends in fertilizer prices, 287 and Birr 403 respectively. Pulses and oil it seems realistic to assume that fertilizer seeds, other crops and fruits reported prices will increase. expenditure per hectare of Birr 238, 161 and

Table 11: Effect of changes in fertilizer prices on contribution of smallholder contribution to agricultural GDP Crop category Contribution to AgGDP Contribution to GDP Relative change Baseline 4.22 1.78 Scenario 1 4.17 1.76 -5 Scenario 2 4.14 1.75 -8 Scenario 3 4.08 1.72 -14 Scenario 4 4.03 1.70 -19

As can be seen from the simulation results, scenario if there is a 10 percent or more the contribution from smallholder managed increase in price of fertilizer. A 35 percent irrigation schemes to agricultural GDP does increase in price of fertilizer, while fall significantly compared to the baseline assuming other things remain constant, for

146

instance, leads to a 19 percent reduction in costs and net gross margin (see Table 5). its contribution to agricultural GDP These differences could be attributed to compared to the baseline scenario. differences in structure of investment and management and, hence, efficiency of the 11.4 Improvement in efficiency of schemes. The lack of relevant information smallholder managed schemes on initial investment expenditure has also made the analysis difficult. In schemes Besides exogenous changes in prices and where we couldn’t get data on initial changes in cropping patterns, farmers are investment costs we used data related to also expected to gain irrigation experience initial capital outlays. The huge differences and improve their efficiency in using land in annual investment recovery costs and net and water. This is also expected to lead to gross margin could partly be attributed to increase in gross margin. We, hence, lack of reliable data. explored what happens to irrigation’s contribution if gross margin of smallholder In simulating the future contribution of large agriculture increases to that of irrigators in scale schemes, we set certain assumptions the traditional schemes. The simulation based on the differences in net gross margin results show that the contribution from between the three major sugar growing smallholder managed irrigation will increase schemes. Since there will be emerging to about USD 475.5 million that accounts to schemes, e.g. Kesem and Tendaho on 90000 6.4 and 2.7 percent of the agricultural GDP ha of land, in sugar cane production, we and overall GDP in 2009/2010. This has also need to set certain assumptions about these an important policy implication; government schemes performance. We first assumed that and extension support through education and the net gross margin for Finachaa, Metehara training contributes to improved efficiency and Wonjo/Shoa respectively applies to the and increased contribution of irrigation to new schemes (Scenario 1-3); Kesem & national economy. Tendaho perform on the average of the three exiting schemes (scenario 4); all schemes, 11.5 Projecting the future contribution of existing and emerging, perform like large scale plantations Finachaa (scenario 5); all schemes perform like Metehara (scenario 6), all perform like In projecting the future contribution from Wonji/Shoa or all perform on the average of large scale commercial plantations, we the three (scenarios 7 & 8) . Finally we tested various scenarios. First we need to assumed a 10 and 15 percent increase in the differentiate between large scale smallholder price of sugar while the average gross managed and large scale commercial margin works in all schemes (scenarios 9 plantations. The former category was and 10). For details see Table 12 below. covered in the proceeding sections while in Following these scenarios, the contribution this section we focus on the large scale from large scale plantations to agricultural commercial production. The major GDP, ranges from less than 1 percent for expansion in the state owned commercial scenario 5 (worst scenario) to about 6 plantations involves predominantly growing percent in scenario 2 (best scenario). The of sugar cane for sugar production. There is intermediate outcomes lie somewhere in no information on future expansion plans of between contributing about 3 percent to the fruits and vegetables and other crop growing agricultural GDP. These results show that large scale commercial farms. Our focus in the structure of investment and the way this section, hence, will be on sugar these schemes are managed may have a plantations. Worth noting is that in the significant bearing on their contribution to existing sugar plantations there is huge national economy. difference in annual investment recovery Table 12: Projected contribution of large scale sugar estates to Agricultural GDP

147

Scenarios Description Contribution to Contribution % Change AgGDP to GDP

Baseline average NGM for LSS 2.9 1.2 assumed

Scenario 1 Kesem & Tendaho 1.5 0.65 - 140 performs like Finchaa

Scenario 2 Kesem & Tendaho 5.8 2.46 290 performs like Metehara

Scenario 3 Kesem & Tendaho 2.9 1.22 0 performs like Wonji/Shoa

Scenario 4 Kesem & Tendaho 3.4 1.4 50 achieves average performance

Scenario 5 All perform like Finchaa 0.32 0.13 -258

Scenario 6 All perform like 6.1 2.5 320 metehara

Scenario 7 All perform like 2.16 0.9 -74 Wonji/Shoa

Scenario 8 All perform like average 2.87 1.2 -3

Scenario 9 10 percent increase in 2.87 1.2 -3 baseline NGM

Scenario 10 15 percent increase in 2.87 1.2 -3 baseline NGM

Taking these scenarios into account the percent respectively. Similarly, the contribution of smallholder managed contribution from large scale irrigation to irrigation to agricultural and overall GDP agricultural and overall GDP will be in the will vary between 4 to 6 and 1.8 to 1.9 range of 3 to 6 and 1.2 to 2.5 percent

148

respectively. Overall, the future contribution margin from irrigation is more than twice of irrigation to agricultural GDP will be in higher than the gross margin from rainfed the range of 7 to 12 percent while the agriculture. When disaggregated by contribution to overall GDP will be in the irrigation typology, average income from range of about 4 percent. small scale modern systems is about USD 355/ ha while from small scale traditional 12. Conclusions and recommendations systems it is about USD 477/ha. This difference in net income between the Irrigation development is quite a recent traditional and modern systems could be phenomenon in Ethiopia. While the country attributed to differences in the investment has huge potential for irrigation only about 5 cost and relative irrigation experience. We percent of this potential is currently used. also found huge inter-scheme differences in Irrigation development is identified as an average income within the same typology important tool to stimulate sustainable which could be attributed to differences in economic growth and rural development and relative irrigation experience and access to is considered as a corner stone for food market. When it comes to medium scale security and poverty reduction in the irrigation schemes, the average income from country. To this effect a comprehensive modern irrigation schemes was USD 400/ha. national Irrigation Development Strategy Taking the average income from (2005/06-2009-2010) has been developed smallholder managed small and medium and is being implemented with the aim of scale irrigation schemes in the country and establishing small, medium and large scale the total hectarage for both categories, we irrigation schemes, either for use either calculated the total income driven from under smallholder managed schemes or irrigation to be Birr 2.27 billion (about 262.3 large scale commercial plantations. In spite million USD). This accounts for about 4.46 of this, there is little attempt to measure the percent of the agricultural GDP in actual and expected contribution of 2005/2006 and 1.97 percent of the total irrigation to the national economy. Hence, overall GDP. the objective of this study was to estimate the net contribution of irrigation to GDP at On the other hand, the average income net the farm gate following an adjusted gross of annual recovery cost from a hectare of margin analysis approach. Studies of this irrigation under large scale schemes is USD kind could be instrumental in comparing the 1,308. Taking the all large scale schemes in actual and expected direct benefits of the country, differentiated by their cropping irrigation with the actual and expected costs pattern, and the average income from the of irrigation expansion to guide policy selected learning sites, the total income makers in irrigation development. However, earned from large scale schemes is estimated there is need for caution. This study does not to be Birr 641 million (ca 74.0 million capture the all the multiplier effects of USD). This accounts for about 1.26 percent irrigation. Doing that requires more data of the agricultural and 0.5 percent of the than is presently available. However, this total GDP respectively. Overall, the first attempt can be extended to more precise contribution of irrigation to agricultural and analysis of economy wide effects of total national GDP was about 5.7 and 2.5 irrigated agriculture development. percent during the 2005/06 cropping season. When only the improved system is Our results show that irrigation in the study considered, it contributed to about 1.26 and sites generates an average income of about 0.5 percent of the agricultural GDP and USD 323/ ha. This compares to the GDP respectively. Our result show that the calculated gross margin for rainfed which is bulk of the contribution to national economy USD 147/ha. This indicates that after comes from the smallholder managed accounting for replacement cost, net gross irrigation schemes, most importantly from

149

the traditional schemes. The same results percent reduction in the area of cereals) also show that the contribution of irrigation leads to about 45 percent increase in the to national income is still very small contribution of smallholder irrigation to compared to the 28 percent contribution of agricultural GDP as compared to the crop production and role of irrigation to baseline scenario. This is an important result national economy in some countries such as indicating that the contribution of irrigation the Sudan and Egypt where in the former could be maximized if smallholder farmers irrigation contributes to about 50 percent of shift their cropping pattern to more high the crop production while in the latter value crops. Hence, the extension system almost all agriculture is irrigated. could play an important role in providing and promoting high value crops. Taking all the envisaged areal expansion, Likewise, simulation results on the effect of exiting cropping pattern, and the average price change show that a 10-15 percent gross margin values for different crop increase in the price of fruits and vegetables categories, the expected contribution of leads to 15-23 percent increase in the smallholder managed irrigation to national relative contribution of smallholder economy is expected to increase from USD irrigation to agricultural GDP. An equivalent 262.3 million in 2005/2006 to about USD increase in the price of cereals leads to 22- 414.2 million in 2009/2010, which accounts 32 percent increase in the relative to about 5.5 percent of the agricultural GDP contribution of the sub sector. One the other and 2.3 of the overall GDP for the same hand, the same level of increase in prices of year. On the other hand, the contribution pulses, oil seeds and other crops did not coming from the large scale sugar growing yield significant change in their estates in 2009/2010 is estimated to be USD contribution. The relatively higher 217.5 million which amounts to 2.9 and 1.2 contribution coming from cereals is percent of the agricultural and overall GDP attributed to the bigger share cereals have on respectively. Similarly the contribution land claiming about 61 percent of the coming from large scale commercial farms cultivated area under irrigation. This implies growing crops other than sugar cane is that an increase or decrease of prices of expected to increase to USD 35.8 million in vegetables and fruits will have a stronger 2009/2010 which accounts to 0.4 and 0.2 relative impact on the contribution of percent of the agricultural and overall GDP irrigation to national economy compared to respectively. This implies that large scale that of cereals. Hence, vegetables and fruits commercial farms will contribute about 3.3 are economically more attractive and could and 1.4 of the agricultural and overall GDP yield more value to the economy if more a respectively. In summary, our results more land is shifted from cereal production indicate that under conservative estimates to production of vegetables and fruits. the future contribution of irrigation to On the other hand, increase in price of agricultural and overall GDP will be in the fertilizer leads to reduction in the range of 9 and 3.7 percent respectively. contribution of irrigation to national economy. Accordingly, a 35 percent Furthermore, we also relaxed some of the increase in price of fertilizer, while assumptions to check the sensitivity of our assuming other things remain constant, for results to model assumptions. We assumed instance, leads to a 19 percent reduction in various changes in cropping patterns, small holder irrigation’s contribution to changes in input and output prices and agricultural GDP compared to the baseline improvement in levels of efficiency. scenario. Our results from the simulation exercise in relation to shift in cropping patterns show Besides changes in prices and cropping that a 15 and 10 percent increase in the area patterns, improved efficiency is found to of vegetables and fruits respectively (i.e. 25 impact significant increased in the

150

contribution of irrigation to national conditions, and iv) increase the efficiency of economy. Our simulation results show that small and large schemes. the contribution from smallholder managed irrigation will increase to about USD 475.5 million, which is 6.4 and 2.7 percent of the agricultural and overall GDP in 2009/2010, when all smallholder irrigation farmers perform to the level of traditional irrigators. This has also an important policy implication: government and extension support through education and training contributes to improved efficiency and increased contribution of irrigation to national economy.

Furthermore, when we simulated the effect of changes in efficiency levels of existing and emerging large scale sugar plantations and changes in the price of sugar we found that the contribution from large scale plantations to agricultural GDP, ranges from less than 1 percent for worst scenario to about 6 percent in the best scenario. The intermediate outcomes lie somewhere in between contributing about 3 percent to the agricultural GDP. These results show that the structure of investment and the way these schemes are managed may have a significant bearing on their contribution to national economy.

In summary, taking these scenarios into account the contribution of smallholder managed irrigation to agricultural and overall GDP will vary between 4 to 6 and 1.8 to 1.9 percent respectively. Similarly, the contribution from large scale irrigation to agricultural and overall GDP will be in the range of 3 to 6 and 1.2 to 2.5 percent respectively. Overall, the future contribution of irrigation to agricultural GDP will be in the range of 7 to 12 percent while the contribution to overall GDP will be in the range of about 4 percent. To enhance the contribution of irrigation to national economy, besides increasing the presence of physical water infrastructure, there is a need to: i) improve provision of agricultural inputs, ii) promote high value crops through the extension system, iii) create good market

151

References Environmental and Resource Economics. 22(1-2): 133-156. Ahmed, A. U and Sampath, R.K., (1992). Bhattarai, M., Baker, R., Narayanamoorthy, Effects of irrigation-induced A.(2003) . Who benefits from irrigation technological change in Bangladesh rice investments in India? Implication of production. American Journal of irrigation multiplier estimates for cost Agricultural Economics 74(1): 144-57. recovery and irrigation financing. A Alagh, Y. K. (2001). Water and food workshop paper presented at the ICID security in South Asia. International regional meeting in Taipei, Taiwan, 10- Journal of Water Resources 12 November, 2003. Development. 17(1): 23-36. Bhattarai, M., and Pandy, S. (1997). The Atnafu, Teshome (2007). Current and Future economics of wheat production in the Plans of Irrigation and Drainage rice-wheat system in Nepal. A chapter in Development in Ethiopia. Irrigation and a book series in P.S. Teng et al. (Eds). Drainage Development Studies 1997. Application of systems approach Department. Ministry of Water at the farm and regional levels 1:45-57. Resources. Addis Ababa. Kluwer Academic Publishers. IRRI, Awulachew, S. B., Aster Denekew and ICASA. Netherlands. Mekonnen Luelseged (2007). Status of Central Statistical Agency (CSA) (2006). Irrigation Development in Ethiopia. Agricultural Sample Survey 2005 / 2006 International Water Management Institute (1998 E.C.). Report on Area And (IWMI). Addis Ababa. Production Of Crops. Statistical Bulletin, Awulachew S.B., Merrey D., Kamara A., 361. Addis Ababa. Pp. 118. Koppen B.V., de Vries F.P., Boelee E. Dhawan, B. D. (1988). Irrigation in India’s and Makombe G. (eds). 2005. agricultural development: productivity, Experiences and opportunities for stability, equity. Sage Publications India, promoting small-scale/micro-irrigation New Delhi. India. and rainwater harvesting for food Doak, M. (2005). Value of irrigation in New security in Ethiopia. IWMI Working Zealand. A paper presented on OECD Paper. No. 98. IWMI (International Workshop on Agriculture and Water: Water Management Institute (IWMI), sustainability, markets and policies, 14- Colombo, Sri Lanka. 86 p. 18 November, 2005. Adelaide, Australia. Bahtia, R., Monica Scatasta and Rita Cestti Doak, M., Irene Parminter, Gerard Horgan, (2003). Study on the multiplier effects of Rachel Monk, and Graeme Elliot. (2004) dams: Methodological issues and The Economic Value of Irrigation in New preliminary results. Paper presented for Zealand. MAF Technical Paper No: the Third World Water Forum, Kyoto, 04/01. Japan. March 16-23, 2003. Ethiopian Sugar Development Agency Barbier, E. B. and Thompson, J.R. (1998). (ESDA) (2007). Revised Master The value of water: Flood plain vs. large Schedule for New and Expansion projects scale irrigation benefits. Ambio 27(6): and Annual Implementation Plan 434-443. (2006/07). Project and Development Bell, C. and Hazel, P. (1980). Measuring the Services. Addis Ababa. 143 Pp. indirect effects of an agricultural FAO (1997) Irrigation in the near east investment project on its surrounding region in figures. Food and Agriculture region. American Journal of Agricultural Organization of the United Nations. Economics 62(1): 75-86. Rome, 1997. Berck, P. and Hoffman, S. (2002). Assessing http://www.fao.org/docrep/W4356E/w43 the employment impact of environmental 56e00.HTM. accessed 04.12.07. and natural resource policy. FDRE (2006). National Accounts Statistics of Ethiopia. National Economic Accounts

152

Department. Ministry of Finance and for Accelerated and Sustained Economic Development. November, Development to End Poverty (PASDEP). 2006. Unpublished. (2005/06-2009/10). Volume I: Main Hagos, F., Godswill Makombe, Regassa Text. Ministry of Finance and Economic Namara, and Seleshi Bekele Awulachew Development (MoFED). September, (2007). Does access to small scale 2006. Addis Ababa. 229 pp. irrigation promote market oriented MoWR (Ministry of Water Resources). production in Ethiopia?. Submitted to (2002a). Water sector development Agricultural Economics. programme 2002–2016. Irrigation Hussain I. and Hanjra, A. (2004) Irrigation development program, Main report. and poverty alleviation: Review of the MoWR, Addis Ababa, Ethiopia. 142 pp. empirical evidence. Irrigation and MoWR (Ministry of Water Resources). Drainage, 53: 1-15. (2002b). Water sector development Jacoby, H. G. (1993). Shadow Wages and programme 2002–2016. Irrigation Peasant Family Labour Supply: An development program. MoWR, Addis Econometric Application to the Peruvian Ababa, Ethiopia. 57 pp. Sierra. The Review of Economic Studies, MoWR (Ministry of Water Resources). Vol. 60, No. 4., pp. 903-921. (2006). Five year Irrigation development Lipton, M., Litchfield, J and Faures, J-M. Programmme (2005/06 – 2009/10). (2003). The effects of irrigation on MoWR, Addis Ababa, Ethiopia. 57 pp. poverty: a framework for analysis. Water Vaidynathan, A. Krishnakumar, A. Policy, 5: 413-427. Rajagopal, A. Varharajan, D. (1994). Makombe, G. (2000), The economic impacts Impact of irrigation on productivity of and performance of irrigation in land. Journal of Indian School of Zimbabwe. Dissertation. Colorado State Political Economy 6(4). University. Department of Agricultural Werfring, A., 2004. Typology of Irrigation and Resource Economics. Fort Collins, in Ethiopia. A thesis submitted to the CO 80523. University of Natural Resources and Makombe, G., Fitsum Hagos, Regassa Applied Life Sciences Vienna. Institute Namara, and Seleshi Bekele Awelachew of Hydraulics and Rural Water (2007). A comparative analysis of factors Management in partial fulfillment of the affecting the financial performance of degree of Diplomingeieur. smallholder irrigated and rainfed World Bank (2006). Managing water agricultural production in Ethiopia. sources to maximize sustainable growth: Unpublished. A World Bank Water Resource Ministry of Finance and Economic Assistance Strategy for Ethiopia. Development (MOFED). (2006a). Agriculture and Rural Development Ethiopia: Building on Progress. A Plan Department. World Bank.

153

Annex I

Table 10a: Gross margin of rain fed crops Mean area Gross value of GM per ha rain Crop type (in timad) output/ mean area GM rain fed fed Wheat 1.3 857 366.7 1089.6 Teff 1.8 806 437.8 956.9 Barley 1.2 507.7 429.2 1493 Maize 1.7 609.2 341 816.7 Finger millet 1.3 717 652.6 2055.4 Sorghum 2.4 754.7 644.6 1053.8 Chick pea 1.3 498.4 337.9 1031.7 Lathyrus 1.2 654.8 592.2 1970.8 Bean 1.2 725 334.3 1078.5 Lentil 0.6 309.4 250.4 1584.8 Nug 2.2 1132.3 939.6 1709.9 Grass pea 1.6 568.2 237.9 599.2 Eucalyptus 0.9 271.3 259.7 1185.6 Hops 0.6 214.1 136.5 956 Other 1.1 495.4 407.8 1553.6

154

Table 10b: Gross margin of irrigated crops

Mean area Crop type (in timad) GVOUT GM irrigated GM per ha Wheat 1.4 1291.2 745.3 2077.7 Teff 1.5 904.7 514.2 1418.4 Barley 1 687.2 513.1 2052.3 Maize 1.6 864.7 610.4 1575.1 Sorghum 1.7 720.4 610.8 1480.7 Cotton 2 394.9 354.8 709.6 Chick pea 1.4 966.3 452.3 1256.5 Lathyrus 1.3 390 252 800.1 Bean 1.1 641.1 265.8 1012.6 Lentil 1 1505 1309.4 5237.6 Nug 1.4 717 538.5 1516.8 Grass pea 2.2 916.4 420.2 771.8 Pea 1.5 1467 617.7 1625.5 Linseed 2.1 445.2 252.7 493 Pepper 1 952.2 833.7 3437.9 Potato 0.9 574.6 420 1812.5 Sweet potato 2 498 397 814.3 Cabbage 0.8 821 662.3 3230.8 Onion 1.5 3112 2699.4 7415.8 Tomato 1.5 1506.1 1017.7 2765.4 Shallot 0.9 1873.5 1016.2 4471.7 Papaya 1 679.5 625.4 2399.9 Banana 1.9 534.1 475.8 995.9 Mango 1.2 711 652.2 2211 Guava 0.7 1038 933.7 5412.8 Coffee 0.7 8407.4 8341.4 45210.6 Sugar cane 0.8 1001.3 869.52 4528.77 Eucalyptus 0.8 5191.9 5113.7 26808.3 Hops 0.9 441.4 340 1456.1 Chat 0.6 1221.9 1098.5 7323.7 Enset 1 400.5 258.3 1051.2 Other 1 505.6 441.4 1736.2

155

Investment in irrigation as a poverty reduction strategy: Analysis of small- scale irrigation impact on poverty in Tigray, Ethiopia

Gebrehaweria Gebregziabher1 and Regassa E. Namara2

1 Norwegian University of Life Sciences (UMB), N1432, Ås, Norway; and Mekelle University, Ethiopia 2International Water Management Institute, Africa Regional Office, Accra, Ghana [email protected]

Keyword: Tigray, Irrigation, Poverty reduction, Abstract Matching, Propensity Score

The regional government of Tigray has invested in millions of Birr to develop irrigation schemes 1. Introduction as a strategy of poverty reduction. The study was based on a representative sample of 613 farm Ethiopia is one of the poorest economies in the households (331 irrigators and 282 non- world (Hagos 2003) and Tigray is its poorest irrigators) drawn using three stage stratified and most severely food insecure region as sampling with probability proportional to size. compared to the other regions of the country The main aim of this paper is to study the impact except to SNNPR (Federal Democratic Republic of irrigation on household income, therefore, to of Ethiopia (FDRE) 1999). Poverty reduction in contribute to the scant literature on irrigation- Tigray is a core policy agenda of the Ethiopian poverty reduction nexus in Ethiopia, which government in general and the regional policy makers can use it as an input to make government of Tigray in particular. A general informed policy decisions in their future consensus was reached that an increase in endeavors. We found that farming income is agricultural production and poverty reduction more important to irrigating households than to should come mainly through agricultural non-irrigating households, while off-farm intensification and adoption of technologies that income is negatively related with access to improve soil moisture to use more productivity irrigation. We also found that irrigating enhancing inputs. The use of productivity households’ average income is above the enhancing inputs (such as fertilizer and high regional average, while non-irrigating yielding variety) depends much on availability households’ average income is 50 percent less of moisture in which case, investment in than the average income of irrigating irrigation becomes crucial. Despite the role of households. Although there can be other factors, irrigation in easing the effect of rainfall which may contribute to the difference in uncertainty on agricultural performance, income, these results are inline with our Ethiopia in general having an immense irrigation expectation and supports the decision of the potential, has remained dependent on rain-fed Tigray government to use irrigation as a poverty and less productive agriculture, which resulted reduction tool. We have used a stochastic in food insecurity and sever poverty. To this dominance analysis and found that the results end, the Ethiopian government in general and are consistent. This result differs from a the regional government of Tigray in Tigray has focused on rural investment on small-scale previous study by Pender et al. (2002), which irrigation as a key poverty reduction strategy. argues that irrigation has less impact in Since the establishment of the Commission for agricultural yields than expected, reducing Sustainable Agricultural and Environment returns to investment in modern irrigation. Rehabilitation of Tigray (CoSAERT) in 1995, 54 micro-dams; 106 river diversion; and a

156

number of spate irrigation projects were Hagos et al. (2006) is the only recent research constructed with a total irrigation capacity of output from Tigray which deals with the impact 3700 hectares benefiting 19,000 households of small-scale water harvesting (ponds and (Abraha 2003). In addition to the government’s shallow-wells) on household poverty. This study effort, non-governmental organizations such as is the first of its kind anywhere in Ethiopia in Relief Society of Tigray (REST), have invested addressing and comparing three irrigation in irrigation projects. According to Abraha systems (i.e., earth dam, river diversion and (2003), a micro-dam project, to irrigate 100 shallow wells) under different agro-ecological hectares, is estimated to cost about 5.84 million settings. Furthermore, it has made an effort to Birr (1US$=8.65 Birr), while a river diversion address the complete pathways and layers that project that can irrigate 45 hectares costs 1.17 could exist between irrigation and poverty million Birr, in which case investment per reduction. hectare is estimated at 58,390 and 25,896 Birr, in dam and river diversion projects, respectively. Accordingly, the main objectives of this paper are: In spite of the high optimism and the amount of 1) To study the impact of small-irrigation on resources committed to develop irrigation, household income in Tigray, so that policy Pender et al (2002), argued that in Tigray makers can use the research outcome to make “irrigation has contributed to intensification of informed policy decision. To this end, we land use and to change in crop choice, but has investigate irrigation’s impact on household been associated with less adoption of fertilizer income. We also test whether irrigation has an and improved seeds and less improvement in effect on off-farm employment and income yields than expected. As a result, it appears that diversification. the returns to investment in modern irrigation so 2) This paper seeks to contribute to the empirical far have been relatively low”. On the other hand, literature on irrigation-poverty reduction given the experience that irrigation has been an linkages, through a better understanding the enabling factor for the use of other productivity pathways of irrigation-household income and enhancing agricultural inputs (Dhawan 1988), poverty reduction from the experience of Tigray, and the high expectation from irrigation as anti- Ethiopia. poverty program, the findings of Pender et al. (2002) seem to be paradoxical, which attract the To achieve the main objectives specified above, attention of researchers, policy makers and we develop an analytical framework that depicts financing agencies. the linkage between irrigation-household income and poverty reduction (see Figure 1). The The existing literature and empirical studies framework shows how the linkage works dealing directly with irrigation-poverty linkage between four inter-linked systems (which are: are not only dominantly of Asian origin, but they irrigation, socio-economic, household are few, recent origin and polarized. On the characteristics and agro-climatic systems). other hand, although there are many studies, which indirectly deal with the linkages of The structure of the paper is as follows: section irrigation and household income as a proxy of 2 reviews related literature, while in section 3; household wellbeing or poverty, most of them we describe the conceptual framework that are like a by-product of a general analysis of the captures the pathways. In section 4, we briefly phenomenon of agricultural growth and/or discuss data, sampling procedure and the study poverty (Saleth et al. 2003). Literature review area. Section 5, is dedicated to discuss the pertinent to the linkages of irrigation-household empirical method. In this section, we have income and poverty reduction is presented in the discussed factors that determine participation; next section. In general, we note that there is a hence, we identify varibles that are used to knowledge gap whether small-scale irrigation match participants with non-participants. contributes to increase household income and Furthermor, we have brifley discussed the poverty reduction. To our best knowledge, advantges and limitations of PSM as an

157

estimation method. We use section 6 to presents section and time series data to study the effect of results and discussions followed by conclusion investment in irrigation in poverty reduction in in section 7. India, where they found that investment in irrigation as compared to investment in rural 2. Literature review letracy was more effective poverty reduction instrument, but since they used a single equation Among the existing literature on irrigation and and highly aggregated data it makes it difficult its impact on poverty reduction, some are based to capture the layers and linkages between on empirical research, which focuses on specific irrigation, agricultural growth and poverty locations. These types of literatures use primary reduction (Saleth et al. 2003). or secondary data and are methodologically rigorous. On the other hand there are literatures, Furthermore, the success stories of China’s food which are based on perceptions and logic based self sufficiency in the 1960s and 1970s, was arguments (e.g Lipton and Litchfield, 2003; ), attributed to a massive investment in irrigation while the third type of literature is based on (Huang et al., 2005; and Huang et al., 2006) project evaluation, which mostly is based on the implying that irrigation plays an important role interest of funding organization (Hussain and in poverty reduction. Huang et al. (2005) has Hanjra, 2004). Among these, one of the studies used household level cross sectional data to that attempt to deal with irrigation poverty apply a multivariate analysis method, where it linkages is (Hussain and Wijerathna, 2004), found a strong positive correlation between which is a wide-ranging study that covers six access to irrigation and household income, major Asian countries (i.e., Pakistan, India, leading to poverty reduction and equitable Bangladesh, China, Vietnam and Indonesia). income distribution. Although highly aggregated and review based, Hussain and Wijerathna (2004) argued that As mentioned above, the literature on irrigation irrigation reduces poverty both directly and and its impact is polarized. For example, unlike indirectly, where the direct impacts are realized to the above stated literature, different studies through labour and land augmentation effect that which manly used aggregated data (e.g., ultimately translates to improved productivity, Rosegrant and Evenson, 1992; Jin et al.,2002; employment, income and consumption, while and Fan et al.,2000 ) have found negative and/or the indirect impact is realized through enhanced weak relationship between irrigation and local economy and improved welfare at macro agricultural productivity implying negative or no level (Hussain and Wijerathna, 2004). impact on household income and poverty reduction at large. According to Rosegrant and Regardless of the methodologies applied, most Evenson (1992), for example the effect of of the studies carried to investigate the impact of irrigation on agricultural productivity in India irrigation on poverty reduction are classified as was found negative. Moreover, Jin et al. (2002) comparative analysis, such as before and after, uses aggregated nation wide data of China’s with and without or more or less comparisons major crops but cannot find a relationship Hussain and Hanjra (2004) is one of the between irrigation and total factor productivity descriptive/comparative type study, which (TFP). On the other hand, Fan et al.(2000) has attempts to study the irrigation-poverty linkage, made a comparative analysis of impact of public and argued that access to irrigation reduces expenditure in irrigation, research & poverty. Furthermore, Hussain et al. (2006) has development, road, education, electrification and used primary data to make a comparative rural telephone networking, where “investment analysis of irrigation impact on household in irrigation was found to have the least impact income in the marginal areas of Pakistan, where on both production and poverty alleviation” (Fan it concludes that small-scale irrigation is et al.,2000). Most of the studies that used positively correlated with household income and aggregate data could not identify a positive then reduces poverty. Similarly, Bhattarai and contribution of irrigation to poverty reduction, Narayanamoorthy, (2003) has used both cross implying that the direct effect of irrigation could

158

be undermined by other factors which could we developed a conceptual framework (see have been observed at household and/or plot Figure 1) as a guide to our research. Figure 1 level. illustrates the basic relationship capturing the major pathways and layers inherent in irrigation- In general, the lack of consensus regarding the poverty linkage, which helps to net out the linkages between irrigation and poverty impact of irrigation on poverty reduction. The reduction seems to mirror the general debate framework makes clear that the impact of regarding the role of investment in agriculture. irrigation comes through its multi-dimensional For instance, Christiaensen et al. (2006) argue effect, such as its effect in input use, crop that although the majority of poor people in intensity, land and labour productivity. developing countries, especially in Sub-Saharan Africa (SSA), depend directly on agriculture for In rural areas where most of the people depend their livelihood, there is no common view about on agriculture for their food and income, water the role of agriculture in economic development and food security are closely related (FAO and poverty reduction. For example, the dual 2003). In the framework, the impact of irrigation economy model inspired by Lewis in the 1950s, on household income and poverty reduction is argue that resources have to be diverted from captured through two major pathways (i.e., land agriculture to the industrial sector, while a and labour productivity). Irrigation enhances the positive view that emerged in the early 1960s use of agricultural inputs (such as fertilizer and argue about investment in agriculture and its HYV), which in turn improves the productivity contribution to economic growth and poverty of land and labor (especially, agricultural labor) reduction is more than an equal amount of ultimately resulting in high household income investment in non-agriculture (Christiaensen et and poverty reduction. Such an agricultural al. 2006). The experience of the Green performance could result either because of the Revolution in Asia, where traditional agriculture input use effect or simply due to the external was rapidly transformed substantiates the role of shock minimizing effect of irrigation. investment in agriculture in economic growth and poverty reduction (Christiaensen et al. For example, crop production in the highlands of 2006). Empirical evidences show that in areas Tigray requires more than 90 days (for where irrigation is widely used, agricultural vegetative and flowering), but usually the rain yields and household income are higher, and less stays effectively for about 60 days (during July- poverty and undernourishment are observed August), where agricultural crops are grown (FAO 2003). In the framework we made clear once a year, therefore, farmers are not willing to that the impact of irrigation comes through its invest in fertilizer and other agricultural inputs, multi-dimensional effect. because of the risk of crop failure. As a result agricultural productivity is low and poverty is 3. Conceptual framework high. FAO (1999) argued that higher productivity and production is associated with We hypothesize that irrigation had a significant high input use, therefore, the main constraint to impact on agricultural performance and poverty increase food production is limited uptake of reduction in Tigray. We assume that the effect of new technologies by risk-averse farmers. The irrigation on production is ultimately translated uncertainty caused by unreliable moisture to household income and poverty reduction. availability is the main factor behind the risk Although it may differ from location to location aversion behavior of farmers. Since the and irrigation technology/system, the pathways exogenous component of production uncertainty through which irrigation can impact on poverty is reduced with assured access to irrigation, we reduction are complex and diverse. Hence, if assume that production and income difference researchers and policy makers have to between irrigating and rain-fed households is understand how irrigation affects poverty, it is observed even if there is no difference in input essential to understand the complexity and use. diversity of pathways and linkages. Accordingly,

159

Another dimension through which irrigation can in areas where there is irrigation project, we impact on household income and poverty assume that more jobs and informal businesses reduction is through its spillover effect. The (such as family based petty trade) are created. economic integration (linkage) effect of Since irrigation creates demand for small scale irrigation on poverty reduction is important, but implements, credit, marketing and extension in most cases remains masked. As discussed services, every job created due to irrigation may above, irrigating households benefit directly trigger another job in the non-agricultural sector. through increased and stable income or because Figure 1 depicts the relationship between of the higher value of irrigated land. On the irrigation and poverty in more detail. The keys other hand, even landless laborers and small to the acronyms used in the figure are presented farmers who have no access to irrigation often in table 1. benefit through higher wages, lower food prices and a more varied diet (FAO 2003). Therefore,

160

Figure 1 Impact of Irrigation on Household Income and Poverty Reduction: Irrigation-Poverty Linkages

IRRIGATION SYSTEM LANDPR LABPR SOCIO-ECONOMIC SYSTEM

DAM I OX & LIVSTOCK I GWM ASSET R N GWT P R CREDIT RDIV I U LABOUR T LSIZE MARKET A O G T HOUSEHOLD CHARACTERSTICS R H I I AGRO-CLIMATE SYSTEM HHSIZE N N ALTITUDE

SEX C C AVRAFLL

EDUC O O CV AGE M M SOILTYPE COWORO E E LANQUALI

POVERTY REDUCTION HOUSEHOLD (POVREDU) INCOME

(HHINCOME)

161

4. Data and the study area

The data used in this paper was obtained from a six sites, two each are in the southern and North- survey made to study small-scale irrigation in west zones, while the others are one each in the Tigray region, Ethiopia as part of a PhD Eastern and Central zone of Tigray; therefore, study program. The study area covers six we believe that our data is representative of the communities (tabias), each of which consists of region of Tigray. about 4 villages. As presented in Figure 2, of the

Figure 2: Map of Tigray, Ethiopia and study sites.

The sample selection process involved three Golgol Raya uses pressurized tube stage stratified random sampling. First, all (drip/sprinkler) irrigation systems. tabias in the region having irrigation projects were stratified based on the type of irrigation. In the second stage, we stratified all farm Six sites were selected among which two of households in each tabia based on their access to them use micro-dam, two river diversions and irrigation. Access to irrigated plot through any the rest two use ground water as a source of other means (such as formal or informal land irrigation. Among the two ground water sites rental contract) was not considered in the Kara-Adi-Shawo irrigation project located in stratification process. Finally, we randomly selected 613 farm households (100 sample

162

households from each of the five tabias and 113 counterfactual income is not observed, resolving households from Kara-Adi-Shawo). The such missing data problem requires feasible proportion of sample households with and method of estimation that is based on economic without access to irrigation mirrors the theory. In other words, in studying the impact of proportion of total households in the respective irrigation, a methodological problem that is tabia. This approach enables us to collect frequently observed is the tendency to assume information about non-irrigating households every income and poverty difference observed who are comparable in basic characteristics to between households with and without access to the irrigators that can serve as counterfactual. irrigation solely attributed to the irrigation factor From the total of 613 sample households, 331 of (Dhawan 1988), therefore, to insure them had access to irrigation and 282 of them methodological rigorousness, estimating the were purely rain-fed cultivators. counterfactual is at the core of impact evaluation (Baker 2000). In line with this, we used We asked our respondents about their household matching method to form a counterfactual specific information. We have also collected against which comparison can be made. To data on farm input and output by asking each analyze the impact of small-scale irrigation in household head to recall his activities and Tigray, we consider irrigation as a treatment and production on a particular plot during the rain-fed as a control. A dummy variable I is used immediate past harvest year, that includes to denote access to irrigation, where (I=1) if multiple cropping, especially in irrigated plots. household i has access to irrigation, and (I=0) Data collection was carried during October- otherwise. Variables Y1 and Y0 represent December, 2005. Detailed plot level data was household’s income with and without access to also collected. A plot is defined as a distinct irrigation, respectively. Subscripts 1 and 0 management unit based on the type of crop indicate income with and without access to planted during the 2004/2005 agricultural year. irrigation, respectively. In line with this, the Plot size was not physically measured, but we impact of irrigation on income of household i is ask farmers to tell us in local measurement unit given by: (i.e., in tsimdi). Four tsimdi is equivalent to one ΔiiiYY=−10 Y [1] hectare. We have asked each respondent about For a household who have access to irrigation, the prices of input and output, but we have also randomly checked in the nearby market from we only observe Y1i , while for those who have which we calculated an average price for each no access Y0i is observed, implying that a product type in order to control the effect of household can not be in both situations at a, price difference. The empirical method of therefore, we only observe Y or Y , which can analysis is outlined below. 1i 0i be written as: ( ) 5. Empirical Method YIYii=+−101 IY i [2]

5.1. Estimation Method In Equation [2] if I = 1, (1-I) = 0, thusYIY= ii1 and the reverse is also true. When we say The difficulty in impact evaluation is, impact, we mean the change in income due to identifying the comparison group (the access to irrigation, thus by rearranging equation counterfactual). To make an impact evaluation, [2], we get we need to know what the outcome (in our case the income of households who actually have YYiii= 0 +Δ I [3] access to irrigation) would have been in the If household i has no access to irrigation, I = 0, absence of irrigation (i.e., the counterfactual). ΔiI = 0, therefore, YYii= 0 . Once the problem of identifying the counterfactual is resolved, the difference In summary, we draw three basic points about between the actual and the would be income is the whole process of examining the impact of the impact of irrigation. However, since the

163

irrigation. Firstly, the framework differentiates population irrespective of household’s access to between outcomes (Y1i and Y0i ) and impact (Δi). irrigation, i.e., E(Δi), while Equation (5) The former is simply about describing the estimates the average treatment effect conditional on access to irrigation, i.e., E(Δ ⎢I = outcomes (Y and Y ), while the second is i 1i 0i 1), which is ATT. The most common evaluation about impact (Cobb-Clark and Crossley 2003). context is one of ex-post evaluation, where we Secondly, the analytical framework allows for wish to know what change in outcomes an heterogeneity in impact as well as in income intervention delivered for those who were (income without irrigation). This point is very subject to the intervention (Cobb-clark and important in an empirical impact study and Crossley 2003). differentiates the analytical framework adopted in this study from other models which assume ATT could have give a policy idea about the homogeneity. The assumption of heterogeneity possible impact of irrigation if more investment is important, because in practice, all households is made to expand the program and more who have access to irrigation can not benefit households get access to irrigation. However, equally due to heterogeneous characteristics. the basic problem in estimating ATT is the Thirdly, the framework is restrictive, because it missing data problem. For example, in Equation assumes a Stable-Unit-Treatment-Value (5), ⎡ ⎤ is observed, while (SUTV). As explained in the second point, the EY⎣ 1i I=1⎦ impact of irrigation varies across households due EY⎡ I=1⎤ is missing. If we assume that the to their heterogeneous characteristics, and it ⎣ 0i ⎦ assumes that any impact is confined within that impact of irrigation is homogenous, it would household which implies SUTV, thus it rules out imply that Equation (4) is equal to Equation (5), the possible interaction effect, however, this i.e., ATE = ATT. Thus the missed data would may not be plausible assumption, because of the have been estimated by EY⎡⎤⎣⎦0i I= 0 in spillover effect of irrigation. Equation (4), because homogeneity assumes that

The assumption of heterogeneity is important to EY⎣⎡ 0i I=1⎦⎤ =EY⎣⎡ 0i I= 0⎦⎤ . However, since frame our analysis. According to Cobb-Clark different households have different and Crossley (2003), population average characteristics, they respond quite differently to treatment effect (ATE) [Ε(Δi)] and average the same treatment. Hence, the realistic treatment effect on the treated (ATT) assumption about the impact of irrigation is ⎡⎤EIΔ=1 are different, but are frequently heterogeneity, which invalidates the possibility ⎣⎦()i that the missed data ⎡⎤ in Equation estimated impact parameters, which can be EY⎣⎦0i I=1 specified as in equation (4) and (5), respectively. (5) can be approximated by EY⎡⎤⎣⎦0i I= 0 as in ATE=Δ= E E Y − Y = ()iii[ 10] Equation (4). Therefore, the basic question is, how can we estimate the income of those {EY⎣⎦⎡⎤10ii−= Y I1 Pr()} I =+ 1{} EY ⎣⎡ 10 ii −= Y I 0 ⎦⎤ Pr() I = 0 [] 4 households who actually have access to irrigation in the absence of irrigation. Since the objective of this paper is to estimate the average treatment effect on the treated One possibility to handle such a problem is to (ATT), Equation (4) is irrelevant. Hence, the use the income of households who have no average effect of the treatment (irrigation) on the access to irrigation to estimate what the income income of the treated (ATT) can be written as: of those households who have access to ATT=Δ==⎡⎤ E I1 E⎡⎤⎡⎤⎡⎤ Y − Y I == 1 E Y I =−irrigation 1 E Y would I = 1 have been in the absence 5 of ⎣⎦()iiiii⎣⎦⎣⎦⎣⎦10 1 0 [ ] irrigation which can be written as:

The difference between equation (4) and (5) is E⎣⎡Δiiiii I==1111⎦⎣⎤⎡ EY10 − Y I == ⎦⎣⎤⎡ EY 1 I =− ⎦⎣⎤⎡ EY 0 I = ⎦⎤ that Equation (4) estimates the average treatment [6] effect of irrigation on the income of the whole

164

using YI0ii,0,= X close to x drown from the Observed Missing households who have no access to irrigation, i.e., If we use income of non-participating household I =0 to serve as a comparison group for each to estimate the unobservable/missing data, equation (6) can be rearranged as: household {E[YI1ii],1,== X x} in the treated, i.e., I =1, therefore, the missing data is now estimable through the counterfactual as EI⎣⎦⎡⎤Δ==i 1 EY⎣⎦⎣⎦⎡⎤⎡⎤10ii I=−10 EY I = follows. [7a] Δ=iiYEY10 − [ i] [ 8] By subtracting and adding EY⎡⎤ I=1 to ⎣⎦0i But, since Equation (8) estimates homogeneous equation (7a) we get impact, while the heterogeneous impact (ATT) is estimated as: 11 EY⎡⎤⎡⎤⎡⎤⎡⎤10ii I=−1011 EY I = − EY 00 ii I =+ EY I =ATT=Δ==⎡⎤ E I1 Y − E Y I = Δ I 9 ⎣⎦⎣⎦⎣⎦⎣⎦⎣⎦()iiii∑∑()10[] [] [7b] I I By rearranging the above specification, we obtain 5.2. Selection procedure of participants

EY⎣⎦⎣⎦⎣⎦⎡⎤⎡⎤⎡⎤10ii−=+ Y I110 EY 0 i I =− EY 0 i I ==When we embark on impact evaluation, especially when non-parametric method is EI⎡⎤⎡⎤⎡⎤Δ=+110 EYIEYI =− = = ATTBIAS +employed, it is important to have clear ⎣⎦⎣⎦⎣⎦iii{ 00} understanding about the selection processes of [7c] project site beneficiary placement, Therefore, it is now clear that Equation (7c) administrative and institutional details of the suffers from bias because the income of program (Ravallion, 2005), both at household households with and without access to irrigation and plot level. would be different in the absence of irrigation, why identifying a counterfactual is the core of Accordingly, in Tigray, irrigation project sites impact evaluation. While experimental design were selected based on environmental and method is theoretically ideal for establishing a geological futures of the area, which includes: counterfactual, it is practically impossible, availability enough catchments area, sufficient hence, it has been shown that non-experimental reservoir, presence of sufficient command area, design methods, particularly the matching geological feasibility, short crust length. In the method is considered as the best solution in regional state of Tigray, it is a tradition to practice (Cobb-Clark and Crossely 2003). In consult the community and make sure that the Equation (7c), ATT=Δ E⎡⎤ I and project is accepted by the community before ⎣⎦i construction is started. Moreover, priority is given to drought prone areas. Accordingly, since BIAS==−={ E⎣⎦⎣⎦⎡⎤⎡⎤ Y00ii I10 E Y I } , the site selection criteria are related to therefore if a parametric regression method is topographical issues, whether a plot is irrigated applied, the assumption will be no selection bias (treated) or not depends on factors, such as in program placement, however, when the rainfall agro-ecology; slope of land, program is policy induced (such as placement to susceptibility to erosion, soil type and soil irrigation), it is purposive placement, then the quality. Commonly, irrigation projects found in outcome will depend on treatment status lowland areas with upstream catchments, implying selection bias (Ravallion, 2005). therefore, we assume that plots that become irrigated are steep sloped, and those which are Specifically, matching is used to estimate the susceptible to erosion. Furthermore, because of expected counterfactual {EY[ 0ii],1, I== X x} continuous soil erosion and/or sediment accumulation that take place prior to project

165

inception, potentially irrigable plots can be livestock ownership was used as a selection peroxide by their soil chrematistics. variable, it is also an outcome of access to irrigation; hence, we opt not to use it as selection The issue of household’s access to irrigation criteria. (i.e., whether a household is treated or not) is relevant after the project is constructed. 5.3. Why Propensity Score Matching (PSM) According to criteria used in the region, priority Method? is given to farm households who get land within the command area before the project was In practice, participation in anti-poverty program constructed. The standard irrigated plot size is cannot be randomly; hence, matching method is one tsimidi (i. e., 0.25 ha, that represents an area among the appropriate evaluation tools to assess a farmer can plough with a pair of oxen within a the impact of such social programs. To apply the day), hence, who had more than one tsimidi matching method, it is necessary to identify were lowered to one tsimidi considering that one households from the non-participant group that tsimidi of irrigable land is equivalent to 2 or 2.5 is similar in terms of observable characteristics. tsimidi of rainfed land (depending on the However, in practice, since exact matching is availability of land in each community), rarely possible, because the observable variables however, in most of the communities, the based on which the counterfactual is estimated withdrawal was done without compensation and individuals are matched can be many and because of scarcity of land. Farm households different in dimensions making matching who lost land because of water in the reservoir difficult, options for closeness in matching must are the next beneficiaries to get the standard size be considered (Rosenbaum and Rubin 1985), of irrigable land in the command area. Finally, hence, the Propensity Score Matching (PSM) given that the command area allows (i.e., there is method, which is based on the assumptions of unoccupied irrigable land), additional farm conditional independence and common support, households become beneficiaries among which has been used as one method to solve the poor households (i.e., who lacks livestock and problem of dimensionality. The idea of have more family size) and female headed estimating propensity score is used to balance households get priority. Although small in households who have access to the treatment number (49 households), we found that (irrigation) by choosing control households from households can access irrigable land through those who have no access to the treatment land rental market. Accordingly, we used (irrigation), but look like the treated households household head’s sex, family size, female and based on observable characteristics (Ho et al., male adult members of the household, plot size, 2007), therefore, are comparable to estimate the a dummy variable for type of land rented in impact of irrigation. “The use of PSM relaxes (1=irrigated, 0=raiinfed), number of plots owned the assumption of exogenous placement of anti- by the household and dummy variable for poverty programs, and it attempts to balance the ownership of land (i.e., whether the household distribution of observables, i.e., the propensity has rented in land or not) as matching variables score” (Ravallion, 2005), therefore, unlike the to estimate the propensity score. Usually, much theoretic randomization, PSM emphasizes households with more family size, especially on the matching variables and on the quality and with more dependent are considered as poor in quantity of data the rural areas of Tigray as elsewhere in Ethiopia. Plot number was used as matching As compared to parametric models, PSM is variable, because it proxies the probability of preferred, because it relaxes randomization. having land in the command area and the Furthermore, its simplicity in relaxing the probability of being considered as a poor. If a assumptions of functional forms that normally household owned more plots, the probability of are imposed by parametric regression models, having land in the command area is high, while such as OLS is an advantage. “PSM allows the the probability of being considered as poor to estimation of mean impacts without arbitrary get access to irrigation is low. Although, assumptions about functional forms and error

166

distributions. Furthermore, despite that minimize the level of bias. There are examples regression models use full sample, PSM is indicating that bias can be large in non- confined to matched one (i.e., the region of experimental impact evaluation among which common support), therefore, impact estimated (Lalonde, 1986; Glewwe et al., 2004; and Van with parametric models (i.e., based on full or de Walle, 2002) are few, but this does not mean unmatched samples) are more biased and less that non-experimental impact evaluation robust to miss-specification of regression methods can not be used(Ravallion, 2005). functions than those based on matched samples.”(Ravallion, 2005). The spillover effect is another methodological challenge of impact assessment, implying that Finally, our study takes the advantage of having eliminating selection bias by itself is not detailed survey data and full knowledge of the sufficient to identify the impact of treatment. program. Our knowledge about the The estimation method outlined in section 5.1 administrative and implementation procedure of assumes that the presence of irrigation project in irrigation schemes in the region helps us to the community affects only those who have identify proxy variables that determine program access to it (i.e., ATT), however, although they participation. We have collected a detailed data are not direct beneficiaries, those who get by administering the same questioner to both the employment through the project implementation participants and non-participants. The nearest also benefits fro the project. For example, neighbor and kernel matching methods were irrigation network construction and catchments used to estimate the average treatment effect of treatment brought a huge employment irrigation on the treated (ATT). We have opportunity for people inside and outside the checked that the common support and balancing command area. Furthermore, even after the properties were satisfied in our data, hence, the project completion, more labour get recruited remaining bias, if any, can be attributed to because of the labour intensive nature of unobserved characteristics (Jalan and Ravallion, irrigation. The benefits of lower food prices lead in press). to improved nourishment of the whole community. On the other hand, irrigation brings 5.4. Limitations negative externality, such as prevalence of malaria. In general, in the presence of positive There are three basic problems that confound spillover effect, estimated impact could be impact evaluation in general, which includes: downward biased, while it could be upward selection bias, spillover effect and biased if the negative affected is assumed in the data/measurement error (Ravallion, 2005) where estimation process. our study is not exceptional. The main concern of non-parametric impact evaluation is whether 6. Results and Discussion the selection (placement) process to participate in the program is full captured by the control 6.1 Descriptive results variables (Ravallion, 2005). Household characteristics and resource Since irrigation a policy induced program to endowments: reduce poverty, it is impractical to assume that The descriptive results are presented in tables 2 participation could be random, hence, it should and 3. There are no significant differences be emphasized that the concern about selection between irrigators and rain-fed farmers bias is that some of the variables that jointly regarding household demographic characteristics influence income and access to irrigation are and level of education. No significant difference unobservable making it difficult to claim that the is observed in farm size between the two groups. entire difference between the income of We noted that households who have access to households with and without access to irrigation irrigation hired more labor as compared to is attributed to irrigation (such a bias is specified households who have no access to irrigation. in Equation 7c). This indicates that we can only

167

Both irrigating and rain-fed households have the level of that of irrigators through various almost equal number of oxen, milk cows and mechanisms. This confirms the usual claims labor although we observed slightly higher made in the literature that consumption values in favor of irrigating households. expenditure is the preferred measure of welfare to income13. Although the average per capita Comparison of level and sources of income, income of irrigators and rain-fed farmers are consumption and poverty: above the official poverty line that of irrigating Irrigators had, more diversified income sources. households is almost double of that of the non- The irrigators had significantly higher non-crop irrigating households. The difference between farming income. The non-crop farming activities per capita consumption expenditure of irrigators are mainly related to livestock rearing including and non-irrigators is statistically significant. dairying, poultry and bee keeping. There is no significant difference in the magnitude of The observed income difference between the income obtained from off-farm activities two groups is also reflected in the poverty between the two groups, however, the incidence rate. The poverty incidence was contribution of off-farm income to non-irrigating calculated using a poverty line determined based households’ total income is about 18 percent on the estimated income required to access the higher than that of irrigating households’. This minimum calorie required for subsistence (i.e., might be due to the labor intensive nature of 2200 kcal) and other essential non-food goods irrigation. The implication is that households and services. The official national poverty line who have access to irrigation are more occupied is 1075 Birr in 1995/96 constant national in their own farm and have less off-farm average prices (Weldehanna 2004), however, the participation. regional poverty line (for Tigray region) was estimated at Birr 1033.5 (Hagos 2003). Our Although farming income constitutes, on study shows that poverty incidence (i.e., the average, about 72 percent of the total sample proportion of poor households) among irrigators household’s income, it contributes 76 percent of group is significantly lower than that of non- income of households who have access to irrigators. The poverty incidence among non- irrigation, and only 66 percent of income of irrigators is slightly higher than the regional households without access to irrigation. Given average for Tigray and significantly higher than the contribution of agriculture to the income of the national average (Table 2 and 3). In general rural households, such differences in the Tigray, Amhara and SNNPR are the worst proportion of farming income supports the regions in terms of poverty incidence and depth. argument about the role of investment in irrigation as a poverty reduction strategy. In 6.2. Model results general, the descriptive statistics makes clear that irrigators have less off-farm employment The propensity score matching allows for the and more cropping income. statistical comparison group to irrigation participants. Table 4 presents the logit Overall, the total income of non-irrigators is regression used to estimate the propensity scores only about 67% of that of irrigators. Thus the on the basis of which the matching was mean income for irrigators is significantly subsequently done. The logit regression suggests higher than that of non-irrigators. However, the that the probability of access to irrigation difference in the total household consumption increases as household’s ownership of land expenditure between the two groups is not that (both in size and number of plots) increase. significant. The consumption expenditure is higher for irrigators only by 8.6%. The implication is that even though the observed 13 However, beware that the consumption smoothing income gap between irrigators and non-irrigators is usually achieved through distress measures such as is huge, the non-irrigators where able to smooth drawing down on the stock of assets owned such as their consumption level and bring it almost to livestock, borrowing, etc.

168

Table 5 gives our estimates of average income method. Such an assessment is based on set gains, off-farm labor allocation and magnitude poverty lines, which ideally give the minimum of off-farm income difference based on Nearest income that is sufficient for an individual to Neighbor, Kernel and stratification matching fulfill a minimum level of consumption, methods. therefore, the individual’s standard of living is above the poverty line. The general principle of The overall average income gain due to setting a poverty line is that the individual participation in irrigated agriculture ranges whose income is above the poverty line is being between 3600 to 4500 Birr based on the adequately nourished and can fulfill the basic matching method adopted. The average income needs. We used a poverty line equal to 1033.5 gain estimated stratified matching method is Birr (Hagos, 2003). We assessed the impact of lower than that of the Kernel matching method access to irrigation on the cumulative and nearest neighbour methods. The nearest distribution of income and then poverty neighbour matching method is some what reduction by simulating multiple poverty lines. conservative since only 71 cases from the total We found that the stochastic dominance tests of 282 rain-fed farming households were judged confirm the results of propensity score matching to be comparable to irrigators when using this that investment in small-scale irrigation has method (Table 5). On the other hand, the significant impact on household income and stratification matching method is not restrictive. poverty statues. Results of the stochastic Although the Kernel matching method is dominance tests are reported in figures 3-5. marginally conservative as compared to Comparing the head count ratio (the first order stratification, but resulted in higher overall stochastic dominance tests), we found that income gain. 236 rain-fed farmers where poverty incidence is significantly low for estimated to be comparable to the irrigators households with access to irrigation. Similarly, when using the Kernel matching method. the second and third order stochastic dominance tests confirm that the depth and severity of The mean overall income gain due to poverty is lower for irrigating households. participation in irrigation calculated based on the whole irrigators and non-irrigators sample (i.e., 7. Conclusions without using PSM method) is 1413.07 (4278.445 minus 2865.377, see table 2). Thus Poverty reduction in Tigray regional state is a the use of the whole rain-fed sample as a core policy agenda of both regional government counterfactual would under estimate the impact and the federal government of Ethiopia. of irrigation on income and poverty. The bias is Investment in small-scale irrigation was about 2208.19 (3620.2614 minus1413.07) Birr. regarded as a key poverty reduction strategy and Moreover, the irrigators had lower off-farm many governmental and non-governmental income than non-irrigators but the difference is organizations have constructed different not statistically significant. irrigation systems with a total irrigation capacity of 3700 hectares benefiting about 19000 farming 6.3 Stochastic dominance analysis households. However, limited efforts have been made so far to assess whether investments in Although randomization is considered as a small-scale irrigation in Tigray have attained the powerful method of impact assessment, no stated objectives of poverty reduction, food single method is ideal implying that a security and overall socioeconomic combination of tools might be appropriate improvement in Tigray. In fact some of the (Ravallion, 2005). Accordingly we have used a limited efforts made to assess small-scale stochastic dominance analysis to check the irrigation systems are quite pessimistic (see robustness of our estimation of matching Pender et al. 2002). The main objective of this study was to robustly assess the link between 14 This number shows the minimum gain based on the public investment in irrigated agriculture and its estimation of stratified matching method.

169

impacts on income and household poverty in Interprise), Mekelle. Tigray. Baker, J. L. (2000). Evaluating the Impact of To analyze the welfare impact of small-scale Development Projects on Poverty:A irrigation, Propensity Score Matching method Handbook for Practitioners. The World has been applied to a data set generated from a Bank. random sample of 613 farming households (i.e., 331 irrigators and 282 rain-fed farmers) Bhattarai, M. & Narayanamoorthy, A. (2003). representing different agro-ecological zones of Impact of Irrigation on Agricultural Growth Tigray and irrigation system typologies. The and Poverty Alleviation: Macro Level main conclusions from the study are as follows: Analyses in India: Paper Presented at the There is no significant differences in Iwmi- Tata Annual Workshop in Anand, household demographic characteristics Gujarat, India, January, 27-29, 2003. between the irrigators and non-irrigators sample households Christiaensen, L.Demery, L. & Kühl1, J. (2006). Irrigators hired more labor indicating the The Role of Agriculture in Poverty relative labor absorption potential of irrigated Reduction an Empirical Perspective: Policy farming as compared to rain-fed farming Research Working Paper 4013, 4013, Irrigators had more diversified income World Bank sources Households with access to irrigation had Cobb-Clark, D. A. & Crossley, T. (2003). lower participation in off-farm activities again Econometrics for Evaluations: An indicating the labor absorption or on-farm Introduction to Recent Developments. The employment generation capacity of irrigated Economic Record, 79 (247): 491-511. agriculture The mean income of irrigators is Dhawan, B. D. (1988). Irrigation in India's significantly higher than that of rain-fed Agricultural Development: Productivity, farmers. There is also a difference (although Stability and Equity. New Delhi, Sage not statistically significant) in total household Publications India Pvt. Ltd. consumption expenditure between the two groups. Fan, S., Zhang, L. & Zhang, X. (2000). Growth The over all average income gain due to and Poverty in Rural China: The Role of participation in irrigated agriculture estimated Public Investments. Environment and using PSM method ranges between 3600 to Production Technology Division, 4500 Birr per household per annum, which is International Food Policy Research higher than the income gain estimated based Institute: EPTD DISCUSSION PAPER NO. on the whole sample (i.e., using the total rain- 66. fed farmers sample as a counterfactual or comparison group). Hence, the use of PSM FAO(1999). Poverty Reduction and Irrigated avoided the under estimation of the magnitude Agriculture, International Program for of irrigation impact on income. Technology and Research in Irrigation and Finally, the significant income gain has Drainage,Issues paper No. 1, January, Rome. significantly reduced poverty among farmers with access to irrigation. FAO(2003). Water and Food Security. The State of Food Insecurity in the World: Monitoring Reference progress towards the World Food Summit and Millennium Development Goals, Italy. Abraha, Z. (2003). Business Plan: First Draft, Executive Summary, Tigray Regional FDRE. (1999). Comprehensive and Integrated National State, Bureau of Water Resource Water Resource Management: Ethiopian Development (Irrigation Construction Water Resource Management Policy,

170

Ministry of Water Resources. Hussain, M., Hussain, Z. & Ashfaq, M. (2006). FDRE(2002). Food Security Strategy, Addis Impact of Small Scale Irrigation Schemes Ababa, Ethiopia. on Poverty Alleviation in Marginal Areas of Punjab, Pakistan. International Research Glewwe, P., Kremer, M., Moulin, S. & Journal of Finance and Economics, 6 (6). Zitzewitz, E. (2004). Retrospective Vs. Prospective Analyses of School Inputs: The Jalan, J. & Ravallion, M. (In press). Estimating Case of Flip Charts in Kenya. Journal of the Benefit Incidence of an Antipoverty Development Economics 74: 251– 268. Program by Propensity Score Matching. Journal of Business and Economic Hagos F.(2003) Poverty, institution, peasant Statistics. behavious and conservation investment in Northern Ethiopia. Department of Jin, S., Huang, J., Hu, R. & Rozelle, S. (2002). Economics and Social Sciences, Ås, The Creation and Spread of Technology and Agricultural University of Norway. Total Factor Productivity in China’s Agriculture. American Journal of Hagos, F., Kruseman, G., Abreha, Z., Linderhof, Agricultural Economics, 84: 916-930 V., Mulugeta, A. & G/Samuel, G. (2006). Impact of Small Scale Water Harvesting on Lalonde, R. J. (1986). Evaluating the Household Poverty: Evidence from Northen Econometric Evaluations of Training Ethiopia. Poverty Reduction and Programs with Experimental Data. The Environmental Management (PREM). American Economic Review, 76 (4): 604- 620. Ho, D. E., Imai, K., King, G. & Stuart, E. A. (2007). Matching as Nonparametric Lipton, M., Litchfield, J. & Faures, J.-M. (2003). Preprocessing for Reducing Model The Effect of Irrigation on Poverty: A Dependence in Parametric Causal Inference. Framework for Analysis. Water Policy, 5: Political Analysis 15: 199–236. 413-427.

Huang, Q., Dawe, D., Rozelle, S., Huang, J. & Pender, J., P.Jagger, Gebremedhin, B. & Haile, Wang, J. (2005). Irrigation, Poverty and M. (2002). Agricultural Change and Land Inequality in Rural China. The Australian Management in the Highlands of Tigray: Journal of Agricultural and Resource Causes and Implications. Policies for Economics, 49: 159–175. Sustainable Land Management in the Highlands of Tigray, Northern Ethiopia: Huang, Q., Rozelle, S., Lohmar, B., Huang, J. & Summary of Papers and Proceedings of Wang, J. (2006). Irrigation, Agricultural Workshop held at Axum Hotel, Mekelle, Performance and Poverty Reduction in Etiopia, 28-29 March 2002. Socio-economic China. Food Policy, 31 30–52. and Policy Research Working Paper 54. ILRI (International Livestock Research Hussain, I. & Hanjra, M. A. (2004). Irrigation Institute), Nairobi and Poverty Alleviation: Review of the Empirical Evidence, IRRIGATION AND Ravallion, M. (2005). Evaluating Anti-Poverty DRAINAGE, 53 ( ): 1–15 Programs. World Bank Policy Research, Development Research Group, Working Hussain, I. & Wijerathna, D. ( 2004). Irrigation Paper 3625. and Income-Poverty Alleviation:A Comparative Analysis of Irrigation Systems Rosegrant, M. & Evenson, R. (1992). in Developing Asia. International Water Agricultural Productivity and Sources of Management Institute (IWMI). Growth in South Asia. American Journal of

171

Agricultural Economics, 74: 757-761 Walle, D. V. D. (2002). Choosing Rural Road Rosenbaum, P. R. & Rubin, D. B. (1985). Investments to Help Reduce Poverty. World Constructing a Control Group Using Development 30 (4): 575–589. Multivariate Matched Sampling Methods That Incorporate the Propensity Score. The Weldehanna T (2004)The experiences of American Statistician, 39 (1): 33-38. measuring and monitoring poverty in Ethiopia. Poverty Analysis and Data Saleth, R. M., Namara, R. E. & Samad, M. Initiative (PADI).Mombassa, Kenya, at (2003). Dynamics of Irrigation-Poverty http://www.worldbank.org/afr/padi/ethiopia Linkage in Rural India: Analytical _paper.pdf, accessed on 13.07.07. Framework and Empirical Analysis. Water Policy, 5: 459-473.

Table 1. Key to the acronyms used in Figure 1 or conceptual framework. Accronym Description DAM Source of water for irrigation is micro-dam RDIV Source of water for irrigation is river diversion GWM Source of water for irrigation is groundwater :modern communal GWT Source of water for irrigation is groundwater: private manual IRRIG Access to irrigation LABOUR Household’s labor endowment CREDIT Access to credit ASSET Asset holding LSIZE Land holdings size OXEN Number of oxen owned LIVESTOCK Households livestock holding (TLU) MARKET Distance to nearest market INPUT Total expenditure o inputs used (Chemicals, fertilize, seed, etc) LANDPR Land productivity LABPR Labor productivity HHSIZE Household size in adult equivalent AGE Age of household head SEX Sex of household head EDUC Number of educated household members COWORO Consumer worker ratio ALTITUDE Altitude above sea level AVRAFLL Average rainfall CV Coefficient of variance of rainfall SOILTYPE Soil types LANQUALI Land quality AGRINCOME Income from agriculture (cropping income) OTHINCOME Income from other sources HHINCOME Total household income POVREDU Poverty statues based on income level

172

Table 2: Summary statistics of selected variables used in estimating the treatment effect Welfare indicators Irrigators Non-irrigators t-test (N=330) (N=282) ( Significance test of Mean SE Mean SE difference) Household characteristics and resource endowments Family size (number) 5.066 .120 4.681 .127 -2.206** Family size (adult equivalence) 4.495 .126 4.154 .129 1.884* Female adult members of the household 2.650 .078 2.482 .076 -1.5302 (number) Male adult members of the household 2.417 .082 2.199 .087 -1.817* (number) Number of plots the household 4.574 .109 3.351 .114 -7.736*** cultivated in 2005/06 Number of oxen the household own 1.260 .056 1.121 .066 -1.625 Number of milk cows the household .656 .057 .660 .075 0.043 own Farm size in hectare 5.018 .160 5.038 .203 0.082 Household members who can read and 1.426 .084 1.259 .081 -1.418 write (number) Income and consumption Total Household income in 2005/06 4278 364 2865 224 -3.178*** (Birr) Total Household consumption 3058.839 111 2817.016 106.248 -1.555 expenditure in 2005/06 (Birr) Proportion of farming income (%) .76 .011 .66 .016 -5.234*** Per capita income (Birr) 1230.097 169 799.304 63.477 -2.248** Per capita expenditure(Birr) 803.190 34 785.259 30 0.393 Poverty incidence (%) .44 .027 .56 .030 3.111***

Table 3. Poverty by region using poverty line based on Basket of Kcal Region Per capita consumption Poverty Index (%) Poverty Gap expenditure (Birr) (1999) (2002) 1999 2002 Tigray 903.60 0.58 0.56 0.17 Afar 1105.6 0.52 0.33 0.10 Amhara 917.2 0.57 0.54 0.16 Oromia 1184.0 0.35 0.34 0.08 Somali 1166.4 0.35 0.31 0.07 Benshangul-Gumuz 1026.8 0.48 0.47 0.13 SNNPR 945.5 0.57 0.56 0.18 Gambela 1223.5 0.42 0.34 0.09 Harari 1459.7 0.29 0.22 0.05 Addis-Ababa 1569.0 0.30 0.30 0.09 Dire Dawa 1397.1 0.25 0.29 0.07 National 1087.8 0.46 0.45 0.13 Source: FDRE (1999, 2002)

173

Table 4: Estimation of the propensity score to estimate the impact of irrigation on household’s income, off-farm labor participation, and off-farm income Variable Variable Description Coef. z accirri Access to Irrigation (1=yes, 0=no) plotsize Plot size in hectare -2049959(.0378145) -5.42 typland Type of rented in land (1=irrigated, 0=rainfed) -.4637671(.4452168) -1.04 hheadsex Household head sex(1=male, 0=female) .0742088(.2271787) 0.33 familysize Family size in number of people .0532757(.0483211) 1.10 femwl Adult female working labour .0183509(.1003034) 0.18 mamwl Adult male working labour -.0856737(.0919116) -0.93 plotnumber Number of plots operated by the household in .5404309(.0649462) 8.32 2005/06 production year ownrship Whether a household rented in land (1=yes, 0=no) -.1359277(.183545) -0.74 _cons Constant -1.079743(.4993401) -2.16 Notes: () = Std. Err.; *, **, *** Significant at 10%, 5% and 1%, respectively; Treatment is Access to Irrigation

Table 5: Impact of irrigation on household income, household labor allocation and off-farm income (Bootstrapped standard errors): Estimation results of matching method Impact on Matching Method Number of Number of Average Treatment effect t- Treated Control on the Treated (ATT) statistic Income Nearest neighbour 331 71 3940.604 (1348.995) 2.921** (Equal version) Nearest neighbour 331 71 3940.604(1677.466) 2.349** (random version) Kernel Matching 331 236 4405.777 (1382.702) 3.186** Metod Stratification 331 259 3620.260 (1516.378) 2.387** Off-farm Nearest neighbour 331 71 14.171 (20.037) 0.707 labour (Equal version) allocation Nearest neighbour 331 71 14.171 (23.227) 0.610 (random version) Kernel Matching 331 236 8.808 (14.145) 0.623 Metod Stratification 331 259 9.605 (9.958) 0.965 Off-farm Nearest neighbour 331 71 77.023(233.666) 0.330 Income (Equal version) Nearest neighbour 331 71 77.023(446.155) 0.173 (random version) Kernel Matching 331 236 -256.966(288.294) -0.891 Metod Stratification 331 259 -103.900(97.895) -1.061 Note: numbers in parenthesis are bootstrapped standard errors, ** significant at 5% level of significance, df = 8

174

Figure 3: First order stochastic dominance test to compare the incidence of poverty among households with and without access to irrigation

1 0,9 0,8 0,7 0,6 Irrigation users 0,5 Non Irrigation users 0,4 Poverty line 0,3 0,2 0,1 0 0,25 0,5 0,75 1 1,25 1,5 1,75 Head count ratio

Figure 4: Second order stochastic dominance test to compare the depth of poverty among households with and without access to irrigation

0,6

0,5

0,4

Irrigation users 0,3 Non Irrigation users poverty line 0,2

0,1

0 0,25 0,5 0,75 1 1,25 1,5 1,75 Poverty gap

175

Figure 5: Third order stochastic dominance test to compare the severity of poverty among households with and without access to irrigation

0,35

0,3

0,25

0,2 Irrigation users

0,15 Non Irrigation users Poverty line 0,1

0,05

0 0,25 0,5 0,75 1 1,25 1,5 1,75 Poverty gap sqaured

300 250

250 200

200 150

150 CV (%) RF (mm) 100 100

50 50

0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Months Figure 6: Average monthly rainfall distribution (RF) and Coefficient of Variance of rainfall CV) of Tigray (1956-2006) Source: Ethiopian Metrology Agency, Tigray branch office

176

Appendix 1: Logit model Estimates of Propensity Score Matching

**************************************************** Algorithm to estimate the propensity score **************************************************** The treatment is access to irrigation (accirri) access to irrigation 1=yes, 0=no Freq. Percent Cum. 0 282 46.00 46.00 1 331 54.00 100.00 Total 613 100.00

Estimation of the propensity score

Iteration 0: log likelihood = -422.93873 Iteration 1: log likelihood = -377.59997 Iteration 2: log likelihood = -376.19771 Iteration 3: log likelihood = -376.18939 Iteration 4: log likelihood = -376.18939

Logistic regression Number of obs = 613 LR chi2(8) = 93.50 Prob > chi2 = 0.0000 Log likelihood = -376.18939 Pseudo R2 = 0.1105 accirri Coef. Std. Err. z P>z [95% Conf. Interval] plotsize -.2049959 .0378145 -5.42 0.000 -.279111 -.1308809 typland -.4637671 .4452168 -1.04 0.298 -1.336376 .4088418 hheadsex .0742088 .2271787 0.33 0.744 -.3710532 .5194707 familysize .0532757 .0483211 1.10 0.270 -.0414319 .1479834 femwl .0183509 .1003034 0.18 0.855 -.1782401 .2149419 mamwl -.0856737 .0919116 -0.93 0.351 -.2658171 .0944697 plotnumber .5404309 .0649462 8.32 0.000 .4131387 .6677231 ownrship -.1359277 .183545 -0.74 0.459 -.4956693 .2238138 _cons -1.079743 .4993401 -2.16 0.031 -2.058432 -.1010546

Note: the common support option has been selected The region of common support is [.24230422, .98838866]

Description of the estimated propensity score in region of common support

177

Estimated propensity score

Percentiles Smallest 1% .2458666 .2423042 5% .2771301 .2439084 10% .3179111 .2447952 Obs 590 25% .4128555 .2452673 Sum of Wgt. 590 50% .5375359 Mean .5527539 Largest Std. Dev. .1792381 75% .6757212 .9483694 90% .8177568 .9499213 Variance .0321263 95% .8822696 .952905 Skewness .3062784 99% .9415825 .9883887 Kurtosis 2.282861 ****************************************************** Step 1: Identification of the optimal number of blocks Use option detail if you want more detailed output ****************************************************** The final number of blocks is 8.This number of blocks ensures that the mean propensity score is not different for treated and controls in each blocks ********************************************************** Step 2: Test of balancing property of the propensity score Use option detail if you want more detailed output ********************************************************** The balancing property is satisfied This table shows the inferior bound, the number of treated and the number of controls for each block Inferior access to irrigation of block 1=yes, 0=no of pscore 0 1 Total .1428571 36 2 38 .2857143 46 6 52 .3571429 38 39 77 .4285714 60 107 167 .5714286 49 93 142 .7142857 21 54 75 .8571429 9 30 39 Total 259 331 590 Note: the common support option has been selected ******************************************* End of the algorithm to estimate the pscore ******************************************* end of do-file

178

The Impact of Small Scale Irrigation on Household Food Security: The Case of Filtino and Godino Irrigation Schemes in Ada Liben District,

East Shoa, Ethiopia

Abonesh Tesfaye1, Ayalneh Bogale2, Regassa E. Namara3

1Ministry of Water Resources, 2Haramaya University, 3IWMI-Acra [email protected]

economy, the country has been a food deficit Abstract country for several decades, with cereal food aid averaging 14 percent of total cereal production Irrigated production is far from satisfactory in the country. The country's irrigation potential is (FAO, 2001). Irrigation is one means by which estimated at 3.7 million hectare, of which only agricultural production can be increased to meet about 190,000 hectare (4.3 percent of the the growing food demand in Ethiopia potential) is actually irrigated. The aim of this (Awulachew et al., 2005). However, in Ethiopia paper is to identify the impact of small-scale irrigated production is far from satisfactory irrigation on household food security based on (Woldeab, 2003). While the country’s irrigation data obtained from 200 farmers in Ada Liben potential is about 3.7 million hectares (WSDP, district of Ethiopia. Different studies revealed 2002), the total irrigated area is 190,000 ha in that access to reliable irrigation water can enable 2004, that is only 4.3 percent of the potential farmers to adopt new technologies and intensify (FAO, 2005). cultivation, leading to increased productivity, overall higher production, and greater returns It was claimed that Ethiopia can not assure food from farming. In the study area also about 70 security for its population with rain fed percent of the irrigation users are food secure agriculture alone without a substantive while only 20 percent of the non-users are found contribution of irrigation. Thus, the government to be food secure. Access to irrigation enabled of Ethiopia has prepared a water sector the sample households to grow crops more than development program to be implemented in 15 once a year; to insure increased and stable years between 2002 and 2016. this program production, income and consumption; and assigned a prominent role to the development of improve their food security status. The study irrigation in the country for food production concludes that small-scale irrigation is one of the (mowr, 2001). this paper reports the results of a viable solutions to secure household food needs study conducted to assess the efficacy of in the study area but it did not eliminate the food irrigation led food insecurity eradication and insecurity problem. poverty reduction policy objectives of ethiopia based on data collected from godino and filtino 1. Introduction small scale irrigation schemes found in ada liben district of the oromia regional state of ethiopia. 1.1 Background 1.2 Irrigation and Household Food Security: some empirical evidences Ethiopia is faced with complex poverty, which is broad, deep and structural (MoFED, 2002). Despite the importance of agriculture in its

179

Chamber (1994) based on some empirical diversification, enhance food self sufficiency, studies confirms that reliable and adequate increase rural incomes, generate foreign irrigation increases employment, i.e., Landless exchange and provide employment opportunity laborers as well as small and marginal farmers when and where water is a constraint. Nigigi have more work on more days concluded that the major contributions of of the year, which ultimately contributes to food irrigation to the national economy are food security. A study conducted in 10 Indian villages security, employment creation, and improved in different agro-climatic regions shows that foreign exchange earning. increasing irrigation by 40 percent was equally effective in reducing poverty (reducing food A study by IFAD (2005) states that in Ethiopia, insecurity) as providing a pair of bullocks, the construction of small-scale irrigation increasing educational level and increasing wage schemes has resulted in increased production, rates (Singh et al., 1996). Kumar (2003) also income and diet diversification in the Oromia stated that irrigation has significantly and Southern Nation and Nationalities People contributed to boosting India's food production (SNNP) regions. According to this study, the and creating grain surpluses used as drought cash generated from selling vegetables and other buffer. A study by Hussain et al. (2004) produce is commonly used to buy food to cover confirms that access to reliable irrigation water the household food demand during the food can enable farmers to adopt new technologies deficit months. The same study further added and intensify cultivation, leading to increased that during an interview conducted with some productivity, overall higher production, and farmers, it was disclosed that the hungry months greater returns from farming. This in turn opens reduced from 6 to 2 months (July and August) up new employment opportunities; both on farm because of the use of small scale irrigation. and off-farm, and can improve incomes, Moreover, the increase in diversity of crops livelihood, and the quality of life in rural areas. across the schemes and the shift from cereal- The same study identified five key dimensions livestock system to cereal-vegetable-livestock of how access to good irrigation water system is starting to improve the diversity of contributes to socioeconomic uplift of rural household nutrition through making vegetables communities. These are production, income and part of the daily diet. A study conducted by consumption, employment, food security, and Woldeab (2003) also identified that in Tigray other social impacts contributing to overall region irrigated agriculture has benefited some improved welfare. households by providing an opportunity to increase agricultural production through double According to a study carried out on five cropping and by taking advantage of modern irrigation schemes in Zimbabwe, the schemes technologies and high yielding crops that called were found to act as sources of food security for for intensive farming. the participants and the surrounding community through increased productivity, stable However, these studies were descriptive than production and incomes (Mudima, 1998). The analytical in that they did not formally account same study reported that farmers participating in for/ isolate the possible contribution of other irrigation schemes never run out of food unlike confounding variables such has their counterparts that depend on rain-fed household/village characteristics, and other agriculture. policies and interventions that might have as well contributed to the food security status Ngigi (2002) disclosed that in Kenya for the two differences between irrigators and non-irrigators. decades agricultural production has not been Moreover, the empirical works in this area are able to keep pace with the increasing population. very scant in Ethiopia in particular and in Africa To address this challenge the biggest potential in general. Thus, the study aims to contribute to for increasing agricultural production lies in the the small scale irrigation-food security literature development of irrigation. According to the and to provide policy conclusions and same study, irrigation can assist in agricultural

180

implications for future planning of irrigation local language. One week training was given to systems. the enumerators about method of data collection and the contents of the questionnaire. Data 2. Research methodology collection proper was started after pretest was conducted and modifications were made based 2.1. Study area, sample size and sampling on the feedback from the pretest. Secondary techniques information that could supplement the primary data was collected from published and unpublished documents obtained from different godino and filtino small scale irrigation schemes governmental and non-governmental are found in ada liben district and were organizations. constructed by oromiya irrigation development authority (oida) in 1996 and 1998, respectively 2.3. Method of data analysis (oida, 2000 ). the water source for godino irrigation scheme is wedecha dam, which has the capacity to irrigate about 310 ha. while the water The study employed both descriptive and source for filtino irrigation scheme is belbela econometric techniques. The descriptive analysis dam, which has a capacity of irrigating 100 ha. was performed using frequencies, means, and the irrigable land in the respective command maximum and minimum values. The areas is distributed to farmers by the econometric analysis employed the Heckman government. except few farmers who lease-in two-step procedure to identify the impact of additional irrigable land almost all farmers in the small scale irrigation on household food security area own quarter of a hectare. the major types of from among possible other household food crops grown by irrigation are onion, tomato, security influencing factors. potato and chick pea among others. Heckman two-step procedure: Evaluating the Out of the 45 Peasant Associations (PA) that are impact of a project/program on an outcome found in the Ada Liben district, two PAs namely variable using regression analysis can lead to Godino and Quftu were purposely selected biased estimate if the underlying process which mainly because of availability of irrigation governs selection into a project/ program is not schemes. To select sample respondents from the incorporated in the empirical framework. The two PAs, first the household heads in the two reason for this is that, the effect of the program PAs were identified and stratified in to two may be over (under) estimated if program strata: irrigation users and non-users. Then the participants are more (less) able due to certain sample respondents from each stratum were unobservable characteristics, to derive these selected randomly using simple random benefits compared to eligible non-participants sampling technique. Since the number of (Zaman, 2001). household heads in the two groups was To evaluate the impact of a program, a model proportional, equal number of sample is drawn commonly employed can be expressed as: from each group, i.e., 100 household heads were selected from each group. In total 200 Y = Xβ + αI + u (1) household heads were interviewed. Where Y is the outcome/impact, X is a vector of personal exogenous characteristics and I is a 2.2. Data collection dummy variable (I=1, if the individual participates in the program and 0 otherwise). The data required for this study was collected From this model, the effect of the program is from sample respondents using a semi-structured measured by the estimate of α . However, the questionnaire. The enumerators for the data dummy variable ‘I’ can not be treated as collection were selected on the basis of their exogenous if the likelihood of an individual to educational background and their ability of the participate or not to participate in the program is

181

based on an unobserved selection process variables on household food security. Thus, the (Maddala, 1983). Some studies have shown the Heckman two-step procedure takes the limitations of applying the classical linear following form: regression methodology to the analysis of K D * = γ Z + u (2) samples with selectivity bias (Heckman, 1979, i ∑ k ik i K = 1 Dardis et al. 1994, Sigelman and Zeng, 1999, Maddala, 1992). Application of the classical linear regression model does not guarantee S consistent and unbiased estimates of the Ci = ∑ β s X is + ε i Observed only if parameter. One solution to this problem in S =1 econometrics is the application of Heckman D* >0... two-step procedures. It is considered as an i (3) appropriate tool to test and control for sample Where the disturbances ui and ε i follow a selection biases (Wooldrige, 2002). bivariate normal distribution with a zero mean, variance σ and σ respectively, and The Heckman two step procedures involves two u ε equations. The first equation (i.e., the selection covarianceσ εu . Therefore, we define a or participation equation) attempts to capture the dichotomous variable D which takes a value 1 factors governing membership in a program. i when a household is an irrigator and 0 This equation is used to construct a selectivity otherwise. The estimator is based on the term known as the ‘Mills ratio’ which is conditional expectation of the observed variable, included as independent variable to the second household food security (C : equation known as response or outcome i) equation. If the coefficient of the ‘selectivity’ * term is significant then the hypothesis that the E(C i / Di > 0) = xβ + σ εuσ ε λ ()− γz (4) participation equation is governed by an unobserved selection process or selectivity bias Where λ is the inverse Mills ratio defined as is confirmed. Moreover, with the inclusion of λ(− γZ ) = φ(− γZ )()/(1−ϕ − γZ ); β and γ are extra term, the coefficient in the second stage the vectors of parameters which measure the ‘selectivity corrected’ equation is unbiased effect of variables X and Z, and are the (Zaman, 2001). Therefore, to evaluate the φ ϕ impact of small scale irrigation on household functions of density and distribution of a normal, food security, we use the Heckman two-step respectively. The expression of conditional procedure. expectation shows thatCi equals xβ only when the errors ε and u are non correlated, i.e., Specification of the Heckman two-step i i procedure: σ εu = 0 ; otherwise, the expectation of Ci is affected by the variable of equation 2. Thus, from expression 4 we find that: Let Z ik be a group of K variables which represent the characteristics of a household i * which influences the probability of participation Ci /Di >0=E(Ci/ /Di >0)+Vi =xβ+σεuσελ()−γZ +Vi (5) in irrigation agriculture measured by a latent * variable Di andγ k are the coefficients which WhereVi is the distributed error term, reflect the effect of these variables on the N 0,σ 1−σ λ λ − γZ probability of being an irrigation farmer, and ( ε ( εu ( ( )))) X is a group of variables which represent the is 3. Results and discussion characteristics of household i which determine household’s food security ( Ci ) and β s are the 3.1. Descriptive Results coefficients which reflect the effect of these

182

The variables included in the model are defined problem even in this seemingly better off part of in table 1. The dependent variable for the first the country indicating the depth of the problem. stage of the Heckman two-step procedure is participation in irrigation. This variable is a dummy variable (given a value of 1 if the household participates in the irrigation scheme 60 and 0 otherwise) for the second stage of the model household food security status is a 50 continuous variable measured by the annual 40 food expenditure in Birr of the household per Irrigators 30 adult equivalent. Before discussing the Non-irrigators econometric results, however, we present some 20 interesting descriptive results. 10 One of the pervasive features of food insecurity in Ethiopia is that it is usually seasonal. It 0 mainly coincides with the active agricultural y r er ul b J

% of households reporting food shortage food reporting households of % June mber tobe season or wet season. To this effect we have e c August O tried to see if there is discernable difference in Novem Sept the timing of food inadequacy between irrigators Months of food shortage and non-irrigators. Surprisingly, there is no difference regarding the timing of food shortages between irrigators and non-irrigators Figure 1. Incidence of reported food shortage by (See Figure 1). The food shortage months start months as early as June (which is the beginning rainy season and therefore agricultural activities in The irrigators and non-irrigators have slightly the study areas) and extends up to November different copping mechanisms in the advent of (which is the beginning of harvest season). No food deficit problem (See figure 2). None of the household from the irrigators group has reported irrigators have reported off-farm employment as food shortage in June. September is the most a coping strategy and also relatively fewer serious food shortage month among non- irrigators reported to have used credit as a means irrigators, while October is the peak food of copping with food shortage. It must be noted shortage month for irrigators. About half of the that using wage employment and consumption non-irrigators reported food shortage in the credit as a strategy to avert food insecurity is month of September. However, there is a stark considered as a distress measure or strategy in difference regarding the incidence rate of Ethiopia. Small animals (such as sheep, goats reported food shortage between the two groups. and chicken) is the most important copping The proportion of farmers reporting food strategy among both irrigators and non- shortage in every month is significantly lower irrigators. for irrigators group. It is interesting to note that irrigation has not eradicated the food insecurity

183

50 45 40 35 30 Irrigators 25 Non-Irrigators 20 15 10

% of sample farmers reporting 5 0 Sales of Sales of Small Off-farm Consumption Cattle Animals Employment Credit Food shortage copping strategy

Figure 2. Food shortage coping mechanisms

Based on how households adapt to the presence insecure) was calculated based on the estimated or threat of food shortages, the overall Coping cost of acquiring the recommended daily calorie Strategy Index (CSI) has been calculated for allowance, which was taken as 2200 kcal per each of the sample households and the resulting adult equivalent per day15. This cut-off value is values were averaged for irrigators and non- estimated to be Birr 900.0 per adult equivalent irrigators. It was found that the average CSI for per annum. Thus, households having food irrigator households is 11.4, while for non- consumption expenditure per adult equivalent of irrigators the corresponding value is 31.4. The less than Birr 900 are considered as food mean difference is statistically significant (Table insecure, while those earning more than Birr 900 2). The higher the CSI, the more food-insecure are considered to be food secure. Based on this is a household (reference). Therefore, based on indicator, again there is substantial difference in CSI the non-irrigator households are more food food insecurity incidence rate between irrigator insecure as compared to irrigator households. and non-irrigators households (see figure 3). Generally out of the 200 sample households 45 The calculated food consumption expenditure percent of them are food secure and 55 percent per adult equivalent values also confirms the of them are food insecure. food security status difference between irrigators and non-irrigators (table 2). The average food consumption expenditure per adult equivalent per annum for irrigation user households is 1322.4 Birr, while the corresponding figure for non-users is 774.4 Birr. The mean difference is statistically significant. Moreover, the total consumption expenditure (both food and nonfood) for irrigators is almost double that of non- irrigators.

The minimum food consumption expenditure per adult equivalent above which a household is considered to be food secure (alternatively 15 This cut-off value was calculated following Greer and below which a household is considered as food Thorbecke (1986) food energy intake method of measuring household food security

184

90 80 70 60 50 Irrigators 40 Non-Irrigators 30 20 % of the households the of % 10 0 Food Secure Food Insecure Household food security status

Figure 3. Household food security status differentiated by access to irrigation

When comparing other indicators of welfare 3.2. Econometric Analysis Results between irrigation and non-irrigators, statistically significant differences were detected Determinants of likelihood of access to (Table 3). For example, irrigators have small irrigation: The first stage of the Heckman model household size, higher level of education, large predicts the probability of access to the irrigation livestock holding size, and better quality scheme of a household. Among the observable (fertility) cultivable land. The irrigators had also hypothesized variables, those that significantly better access to extension and credit services influenced the probability of participating in (Table 4). In conclusion, the descriptive irrigation farming include nearness to the water analyses indicate that irrigators are better off in source, household siz size of cultivated land, terms of food security status and other welfare livestock holding, the quality of land owned by a indicators. But is this due solely to access to farmer and access to credit (Table 5). The irrigation? Other observable and unobservable relationship between household size and variables might have contributed to the observed participation in irrigation project is non-linear. food security status difference between irrigators As the size of a household increases by one adult and non-irrigators. Therefore, we know turn to equivalent, the probability of access to irrigation the presentation of Heckman’s two stage decreases by 30.4% but only up certain point regression model to show the impact of access to beyond which a unit increase in household size irrigation on food security while controlling for starts increasing the likelihood of participation the effects of other observable and unobservable in irrigation. As the size of cultivated area confounding factors. increases the probability of being an irrigator decreases. This may imply that irrigators tend intensify their cultivated land, while rain-fed farmers try to put more land under cultivation.

185

Irrigators have significantly more livestock than security. The coefficient of the land size variable their rain-fed only farmers. They also possess shows that as the household gets one more more fertile land. hectare of land food consumption expenditure of the household increases by 85 Birr. This result is Determinants of household food security: The consistent with the findings of Mulugeta (2002), significance of the lambda term in the second Ayalew (2003), Abebaw (2003) and Yilma stage of the Heckman procedure, confirms the (2005). presence of selectivity bias (Table 6). As expected, access to irrigation had significant Access to extension service and nearness to the impact on household food security. In the study water source are also found to have a positive area irrigation enable households to grow crops relationship with household food security. The more than once a year, to insure increased and positive effect of access to extension service stable production, income and consumption may indicate that in the study area, those thereby improving food security status of the households who get technical advice and household. This result is consistent with the training or those who participated in field finding of Abebaw (2003). The other variables demonstrations are well aware of the advantage that significantly enhance household food of agricultural technologies and adopt new security are experience (as indicate by farmers technologies and produce more, thereby age in years), access to extension service, and improving the household food security status. size of cultivated land. . The nearness to the water source may be a The relationship between household size and surrogate variable for access to irrigation. It has food security is non-linear (see the coefficients already been shown that to the irrigation for household size and its square variable). The scheme, significantly improves household’s food negative and significant coefficient of household security status. The possible other justification size reveals that larger household size leads to could be that the nearness to water source may food insecurity, but only up to a certain point. proxy the location of the farms in relation to the The coefficient of the variable indicates that as irrigation water source . Therefore, households the household size increases by one adult who are closer to the irrigation scheme do not equivalent the food consumption expenditure of incur much cost to access their farm so they can the household decreases by 391.9 Birr. This follow up the farm activity closely and result is consistent with the finding of Mulugeta frequently and may get a better yield. (2002) and Yilma (2005). Contrary to other similar studies (Belayneh, 2005), in this study 4. Conclusion and Implications female headed households had better food security status than the male headed households. The variables that significantly predict access to The coefficient of the variable shows that when irrigation are: household size, size of cultivated the head of the household is male, food land, livestock holding, farmers’ perception of consumption expenditure of the household soil fertility status, access to credit, nearness to decreases by 331.1 Birr. The possible the water source and household size square. The justification for this inverse relationship could variables that reduce the probability of access to be that though male headed households are in a irrigation are large household size, large better position to pool resource to increase cultivated area and access to credit. Rain-fed production, they might spent more money on farmers tend to have large cultivated area. The nonfood expenses rather than spending on food negative relationship between access to credit items to meet the household’s food needs. and access to irrigation may be explained by the fact that: (1) in Ethiopia, the institutional credits The regression result also shows that as the usually give priority to rain-fed agriculture, and cultivated land size increases, a household is (2) the demand for credit among farmers with able to increase and diversify the quantity and access to irrigation may be lower for they can type of crop produced, which may in turn lead to satisfy cash needs through sales from their increased consumption and household food irrigated crops.

186

The variables that increase the probability of household must be trained as to how to increase participation of farmers in irrigation farming production per unit area (productivity). include large livestock holding size, ownership of relatively fertile land and nearness to water Access to extension service is also positively source. Obviously, those households that are related to household food security. Extension situated near the water source are more likely to workers could play a key role in transferring participate in irrigation scheme. However, it knowledge to the rural people easily there by does not mean that placement of an irrigation improving production and consumption. scheme in the village is solely governed by Capacity building of the existing ones and hydrological considerations. It involves political training more extension workers might help process and power relations. address the issue.

In the study area the use of small-scale irrigation Reference contributes significantly to improve household food security. In addition to access to irrigation, Abebaw, Shimeles. 2003. Dimensions and access to irrigation, household size, sex of the determinants of food security among rural household head, size of cultivated land, and households in Dire Dawa, Eastern Ethiopia. An access to extension service significantly MSc Thesis Presented to the School of Graduate influence the food security status of a farm Studies of Alemaya University. 152p. household. Awulachew, B., Merrey, J., Kamara, B., Van The relationship between a household food Koppen, B., Penning de Vries, F., and security status and household size is non-linear Boelee, E. 2005. Experiences and (see the signs for the variables household size opportunities for promoting small scale micro and the square of household size). As the size of irrigation and rain water harvesting for food a household increases the per capita food security in Ethiopia. Working paper 98. expenditure decreases, but up to a point, after IWMI (International Water Management which the per capita food expenditure starts to Institute) Addis Ababa, Ethiopia. increase as the household size increases. Contrary to expectation, female headed Belayneh Belete. 2005. Analysis of food households are less likely to be food insecure as insecurity causes: the case of rural farm compared to male headed households. This households in Metta woreda, eastern needs further investigation, however, tentatively Ethiopia. An MSc Thesis Presented to the it may be explained by differences in the School of Graduate Studies of Alemaya expenditure behavior of male and female University. 130p farmers-female members of a farm household tend to spend more on food items to guarantee Chamber, R. 1994. Irrigation against Rural the food needs of the family before anything Poverty. pp. 32-33. In Socio- Economic else. Another possible explanation may be that Dimension and Irrigation, ed., R.K. Gujar. the male members of a female-headed household Jaipur. India: Printwell. may have gainful employment elsewhere thus contributing to household food security. Cho, S., D. Newman and J. Bowker. 2005. Measuring rural homeowners’ willingness to Size of cultivated land and household food pay for land conservation easements. Forest security are positively related indicating larger policy economics 7:757-770 farm size improves household food security. Households with large farm size are found to be CBE (Commercial Bank of Ethiopia). 2006. food secure; however, there may not be a International Banking Division. possibility of expanding cultivated land size any more because of increasing family size and Dardis, H., Soberon and D. Patro. 1994. degradation of the existing farm land. Therefore, Analysis of leisure expenditure in the United

187

States. The Proceeding of the American Council on Consumer Interest 39,194-200. MoWR (Ministry of Water Resources).2001. Irrigation development strategy (Component FAO (Food and Agricultural Organization of the of the water sector development program). United Nations), 2001. The state of food and Draft report. Ethiopia, Addis Ababa. agriculture. pp. 18-20. World Review Part I. Rome. Italy. Mudima, K.1998. Socio economic impact of smallholder irrigation development in FAO (Food and Agricultural Organization of the Zimbabwe: A case study of five successful United Nations).2005. Global information irrigation schemes. Private irrigation in Sub and early warning system on food and Saharan Africa. International Water agriculture. World food program. Special Management Institute. Colombo, Sri Lanka. report. Italy, Rome. Mulugeta Tefera. 2002. Determinants of household food security in Eastern Oromiya, Green, W. 2003. Econometric Analysis. 5th ed. Ethiopia: The case of Boke District of Prentice Hall. Inc, New York. Western Hararge Zone. An MSc Thesis Presented to the School of Graduate Studies Heckman, J., 1979. Sample selection bias as a of Alemaya University. 151p. specification error. Econometrica. 47(1):153- 162. Ngigi, S.2002. Review of irrigation development in Kenya. pp. 35-37. International Water Hussain, I., R.E., Namara, and S. Madar. 2004. Management Institute. Colombo, Sri Lanka. Water for food security for the poor. A collection of thematic papers. Asian Reilly, B. 1990. Occupational endogeneity and Development Bank. Colombo, Sri Lanka. gender wage differentials for young workers: An empirical analysis using Irish data. The IFAD (International Fund for Agricultural economic and social review. 21(3): 311-328 Development). 2005. Special country program phase II (SCP II) interim evaluation Singh, B., B. N. Singh, and A. Singh. 1996. report number 1643-ET. Ethiopia. Effect of mulch and irrigation on yield of Indian mustard on dry terraces in Alfisols. Kumar, D. 2003. Food security and sustainable Indian Journal of Agricultural Science 60 agriculture in India. pp. 1-2. The water (7):477-479 management challenge. Working paper 60. International Water Management Institute. Sigelman, L., and L. Zeng. 1999. Analyzing Colombo, Sri Lanka. censored and sample selected data with Tobit and Heckit models. Political analysis 8:167- Madalla, G.S. 1983. Limited Dependent and 182 Qualitative Variables in Econometrics. Cambridge University Press. United Teshome, W. 2003. Irrigation practices, state Kingdom. intervention and farmers Life-Worlds in drought-prone Tigray. pp. 2-53. Phd Madalla, G.S. 1992. Introduction to Dissertation, Wageningen University, the Econometrics. pp. 341. 2nded. Cambridge Netherlands. University Press. United Kingdom. Wooldridge, M. 2002. Econometric Analysis of MoFED (Ministry of Finance and Economic Cross Section and Panel Data. pp. 551-565. Development).2002.Sustainable development Massachusetts Institute of Technology. and poverty reduction program. pp. 1-87. London, England. Addis Ababa, Ethiopia.

188

WSDP (Water sector Development Program). Thesis Presented to the School of Graduate 2002. Water sector development program Studies of Alemaya University. pp. 134. 2002-2016, Volume II: Main report. Addis Ababa, Ethiopia. Zaman, H. 2001. Assessing the poverty and vulnerability impact of micro credit in Yilma Muluken. 2005. Measuring rural Bangladesh. pp. 34-36. A case study of household food security status and its BRAC. Office of the chief economist and determinants in the Benishangul Gumuze senior vice president (DECVP). The World Region: The case of Asosa Woreda. An MSc Bank.

Table 1. Definition of model variables Variable Variable Expected code type Variable definition Mean Std. sign

ACCIRRG Dummy Access to irrigation of the household Positive HEADAGE Continuous Age of household head in years 48.0 13.5 Positive HEADAGE2 Continuous Age of the household head square Positive HHSIZEAE Continuous Household size in adult equivalent 4.7 1.7 Negative HHSIZEAE2 Continuous Household size in adult equivalent square Positive EDUCATA Category Education of the household head /illiterate, Positive read and write, grade 1-4, grade 5-8 and grade >8/ SEXHEAD Dummy Sex of the household head (1=male, 0=female) Positive CUTLAND Continuous Cultivated land size in hectare 1.5 1.2 Positive LIVESTOC Continuous Total livestock holding in TLU 6.7 4.2 Positive DISMARKE Continuous Distance from the market place in km 6.7 2.1 Negative SOILFERT Dummy Farmers’ perception of soil fertility status (1= Positive fertile, 0= infertile) SUPPEX Dummy Access to extension service (1= access, 0=no Positive access) CREDIT Dummy Access to credit (1=access, 0=no access) Positive NEARNESS Continuous Nearness of households to water source in km 13.0 9.7 Positive

Table 2. Comparison of consumption expenditure per adult equivalent between irrigators and non-irrigators User Nonuser MD t - value Mean Std Mean Std Food consumption expenditure 1322.3 563.4 774.4 369.7 547.8 8.0***

Total expenditure 1,780.3 946.4 955.6 434.5 824.7 7.9***

Coping strategy index 11.4 13.9 31.4 16.1 19.93 9.1*** Source: survey result (2006) *** indicates significance level at 1 percent.

189

Table 3. Summary of descriptive statistics of sample households by access to irrigation /continuous variables/ User Nonuser MD t - value Mean Std Mean Std HEADAGE 46.8 14.4 49.5 12.5 2.7 1.4 HHSIZEAE 4.3 1.7 5.1 1.8 0.7 3.0*** DEPRATIO 0.4 0.1 0.5 0.1 0.0 3.1*** CUTLAND 1.5 1.5 1.4 0.7 0.1 0.9 LIVESTOC 7.3 3.4 5.0 2.6 2.2 3.6*** TOTPRODUC 13,689.1 21,706.8 2,255.4 3,487.0 11,433.7 5.2*** TOTEXPEN 1,780.3 946.4 955.6 434.5 824.7 7.9*** DISMARKE 7.3 2.2 6.1 1.9 1.2 4.0*** Source: Survey result (2006) *** indicates significance level at 1 percent.

Table 4. Summary of descriptive statistics of sample households by access to irrigation /discrete variables/ Variable User Nonuser Total χ2

EDUCATAGORY 0.007*** Illiterate 69 58 127 Read and write 1 13 14 Grade 1-4 3 7 10 Grade 5-8 15 15 30 Grade >8 12 7 19 SEXHEAD 0.6 Female 7 9 16 Male 93 91 184 SUPPEX 0.002*** Access to extension 67 45 112 No access to extension 33 55 88 CREDIT 0.01*** Access to credit 31 48 79 No access to credit 69 52 121 SOILFERT 0.001*** Fertile 93 67 160 Infertile 7 33 40 Source: Survey result (2006) *** indicates significance level at 1 percent.

190

Table 5. Estimation result of the Binary Probit model and its Marginal Effect Variable Coefficient Marginal effect

CONSTANT 2.634 1.050 (0.203) (0.203) AGEHEAD -0.861 -0.343 (0.248) (0.248) HHSIZEAE -0.764 -0.304 (0.021) ** (0.021) SEXHEAD 0.414 0.165 (0.438) (0.438) EDUCATAGORY -0.293 -0.117 (0.764) (0.764) DISMARKE -0.324 -0.129 (0.673) (0.673) CUTLAND -0.604 -0.241 (0.004) *** (0.004) LIVESTOC 0.362 0.144 (0.000) *** (0.000) SOILFERT 0.838 0.334 (0.019)*** (0.019) SUPPEX -0.427 -0.170 (0.169) (0.169) CREDIT -0.615 -0.245 (0.024)** (0.024) NEARNESS 0.403 0.160 (0.008)*** (0.008) AGEHEAD2 0.722 0.288 (0.302) (0.302) HHSIZEAE2 0.687 0.274 (0.034)** (0.034) Dependent variable Access to irrigation Weighting variable One Number of Observations 193 Logliklihood function -69.13 Restricted log likelihood -133.65 Chi squared 129.03 Degree of freedom 13 Significance level 0.00

Source: Model out put (2006) *** and** are level of significance at 1 percent and 5 percent respectively Values in parenthesis are p values

191

Table 6. Estimation Result of the Selection Equation and its Marginal Effect Variable Coefficient Marginal effect

CONSTANT 1553.936 1553.936 (0.000)*** (0.000)*** ACCIRRIG 576.882 576.882 (0.000)*** (0.000)*** AGEHEAD 14.918 14.918 (0.348) (0.348) HHSIZEAE -391.676 -391.676 (0.000)*** (0.000)*** SEXHEAD -331.133 -331.133 (0.001)*** (0.001)*** EDUCATAGORY 1.736 1.736 (0.930) (0.930) DISMARKE 13.567 13.567 (0.378) (0.378) CUTLAND 85.751 85.751 (0.058)* (0.058)* LIVESTOC -5.063 -5.063 (0.717) (0.717) SOILFERT -47.613 -47.613 (0.534) (0.534) SUPPEX 117.729 117.729 (0.069)* (0.069)* CREDIT -44.539 -44.539 (0.429) (0.429) NEARNESS 9.602 9.602 (0.009)*** (0.009)*** AGEHEAD2 -0.112 -0.112 (0.441) (0.441) HHSIZEAE2 25.607 25.607 (0.001)*** (0.001)*** LAMBDA -243.448 (0.041)** Dependent variable Total food (Total food expenditure per adult equivalent per annum) Number of Observations 193 Selection rule is: User =1 Log-L = -1395.69 Restricted (b=0) Log -L = -1489.70 R-squared = 0.58 Correlation of disturbance in -0.67 regression and selection criteria (Rho) Prob value = 0.00 Source: model out put (2006) *** ** and * show level of significance at 1percent, 5 percent and 10 percent probability level. Values in parenthesis are p values

192

A Comparative Analysis of the Technical Efficiency of Irrigated and Rainfed Agriculture:

A Case of Awash and Rift Valleys of Ethiopia

Dawit Kelemework Department of Economics, Haramaya University [email protected]

Abstract compared in relation to their respective frontiers. The marginal and average Ethiopia’s economy is heavily dependent on productivities of the important factors of the agricultural sector, which contributes production is also calculated and compared. 45% of the GDP, providing livelihood for 85% of the population and accounting for 1. Introduction 60% of the foreign exchange earning. Ethiopia has a total land area of about Ethiopia, one of the poorest countries in 113,000,000 hectares (Annual Report in the Sub-Saharan Africa, has been repeatedly hit Ethiopian Economy, 1999). The economy is by drought resulting in famine and the loss heavily dependent on the agricultural sector, of life of thousands of its rural citizens. The which contributes 45 percent to GDP, 60 country’s agriculture mainly depends on rain percent of the foreign exchange earnings and fed peasant farming which accounts for 96% provides livelihood to 85 percent of the of the food produced in the country. On the population (EEPRI, 2005). Of the arable other hand, it is estimated that the major land, only 40 percent is currently cultivated river basins of the country can irrigate about (Awulachew et al, 2005). As a result of the 3.5 million-hectare of land and at present importance of agriculture in Ethiopia’s only about 161,010 ha or 4.6% is irrigated economy, the government has embarked on around the major river basins. Though the an agriculture centered rural development expansion and better utilization of this program which is meant to spearhead the irrigation potential is unattested, the country’s economic development program production efficiency of the existing (Government of the Federal Republic of irrigation systems also needs attention. This Ethiopia, 2003). Irrigation development is paper compares the technical efficiency of viewed as an integral part of this economic rainfed and irrigated agricultural production development program as promulgated by the in Ethiopia. Using the stochastic production Ethiopian Water Sector Strategy (Federal frontier approach, the study concludes that Democratic Republic of Ethiopia, 2001). the existing irrigation systems are not that efficient and there is a need to make them It is estimated that the major river basins of operate near their production frontier. The the country can irrigate about 3.5 million- production frontiers of both irrigated and hectare of land. At present only about rainfed agriculture is estimated along with 161,010 ha or under five percent is irrigated the technical efficiency of each farmer in (Annual Report on the Ethiopian Economy). both groups and the two groups are The private sector accounts for 6,000 ha of the developed irrigated area (Amare, 2000).

Unpredictable rainfall coupled with a high

193

rate of population growth, makes Ethiopia what the government plans to develop by unable to feed its people. Even under 2016, showing that this irrigation sub sector favorable growth scenarios of rain fed still has great potential for contributing to agriculture, the country still faces a deficit of the Ethiopian’s government development food crops. objectives. The policies for economic development are During the former Derg17 regime, many formulated in an environment which can be small scale irrigation schemes were referred to as the “Ethiopian Paradox”. The collectivized. They were generally poorly Ethiopian Water Resources Management operated, managed and maintained and Policy (1999) states that Ethiopia is currently most, if not all, need for endowed with relatively higher amounts of rehabilitation (CRS, 1999). Many small- rainfall in the region and has a surface scale irrigation infrastructures, especially runoff of about 122 billion cubic meters of traditional diversion weirs, which tend to be water, excluding ground water. The same washed away by flash floods, need annual document also states that “…a number of upkeep. Siltation and damage within the studies made in the field confirm that if the canal system from flooding are also major country’s water resources are developed to concerns for small scale irrigation (ibid). cater for irrigation, it would be possible to The degradation of upper catchments and attain agricultural surplus enough both for watersheds in many areas does not help the domestic consumption as well as for situation (ibid). external markets.” pp VIII. However, even Because of the ambitious government plans given this estimated potential, Ethiopia to expand small scale irrigation in Ethiopia, continues to be a recipient of food aid. it is important to study, among other Irrigation development is therefore performance parameters, the production perceived as one of the strategies for efficiency of small scale irrigation schemes . alleviating the paradox. The government of Many believe that the existing irrigation Ethiopia has an irrigation development schemes are not operating efficiently, and strategy which aims to develop over 470,000 that much has to be done to improve their ha of irrigation by 2016. Fifty two percent efficiency. For example, CRS (1999) has of this development will be large and identified that the existing small scale medium scale schemes while the remaining irrigation schemes exhibited inefficient use 48 percent will be small scale schemes16 of water, leakage from unlined canals and (Federal Democratic Republic of Ethiopia, faulty usage of irrigation water. This study Water Sector Development Programme estimates the technical efficiency of small 2002-2016). Small scale irrigation scale irrigation in Ethiopia. development is therefore envisaged to play a critical role in the government’s strategy for addressing the food security situation in 2. Objectives Ethipoia and solving the paradox. However, The objectives of the study are to: the estimated area under small scale irrigation is owever, only 68,210 hectares in estimate and compare the technical 1996/97 (CSA, 1998), only 30 percent of efficiency of dryland and small scale irrigation farmers.

16 According to Awulachew et al (1999) compare the technical efficiency of different irrigation projects in Ethiopia are identified as small scale irrigation schemes and large-scale irrigation if the size of command area is greater than 3,000 ha, medium scale if it falls 17 Desalegn and Miheret, 2004 characterized the in the range of 200 to 3,000 ha, and small scale if Derg regime in Ethiopia as a Marxist-Leninist it is covering less than 200ha. unitary state with an ideologically driven state or command economy policy.

194

make recommendations that lead to construction was completed in 1997 (ibid). improved technical efficiency. At the time of study, there was no development agent assigned to the irrigation 3. The study sites project to provide extension advice, Batu Degaga Irrigation project is located at however, Yusuf (2004) wrote that there was 8° 25′ North latitude and 39° 25′ longitude, one development agent assigned by the in the upper Awash River Basin. The Woreda’s Irrigation Bureau to assist, advice, elevation of the project area is around 1350 organize and monitor the irrigation project meters (Yusuf, 2004). The total developed activity and the farmers in the association. irrigable area of the project is 140 ha of The Godino irrigation project is located in which 60 ha is currently under cultivation. East Shewa zone of the Oromia region, Ada Batu Degaga was established by World Liben Woreda. The water source is a big Vision Ethiopia, a non-governmental reservoir. The runoff of the surrounding organization in 1992. In 1993, a Farmers’ catchments areas supplying the Wedecha Water Users Association was formed and and Belbela streams are made to run to the was led by the selected administrative Reservoir and the water is distributed committee from the irrigation project. The through canals. Though a Water Users numbers of beneficiaries in the project were Association does exist, Oromia Irrigation varying from year to year but now there are Authority and Woreda authorities control 120 members (ibid). Extension advice at water distribution. Batu Degaga is being provided by six agricultural extension agents permanently At Batu Degaga and Doni farmers grow residing around the irrigation system. They vegetables like onions and potatoes. At are graduates of the newly established Godino a mixed farming system consisting agricultural training colleges. of vegetables, namely, chickpea, pea, onions, potatoes and cereals like maize, Doni irrigation project is located in the wheat and teff are practiced. upper valley of Awash River Basin and 33 Km North of Sodore Recreational Center. 4. Data collection Geographical location of the project is 8° The selected small scale irrigation schemes, 30′ N and 39°33′ E and the elevation varies namely, Doni, Batu Degaga, and Godino, from 1240m to 1280m above sea level. It which are located in the Rift Valley of has a different development path from Batu Ethiopia, were chosen due to their relative Degaga. Some 30 years ago, a private proximity to the capital, Addis Ababa. investor constructed a low head gravity weir Primary data were collected using the in Awash River and about 3 km of main household as the unit of analysis. From each canal for the scheme. During the Derg, the irrigation scheme, 50 randomly selected land was nationalized and distributed to households were beneficiaries of irrigation smallholders. A Producer Cooperative was and another 50 randomly selected established to administer and use the households belonged to a control group who scheme. However, scheme operation and did not have access to irrigation. The control maintenance was not good enough to keep it group is not far from the irrigation schemes. functional and after few years it almost They are just bordering the irrigable area. collapsed. Following the downfall of the Socio economic and production data were Derg, a group of individual farmers who collected for the sample households for a owned land within the boundary of the year, between March 2004 and February irrigation scheme started rehabilitating it and 2005. Production data was used to compute requested the assistance of CARE- the gross value of production per household. International in Ethiopia (a non- governmental organization). The request for rehabilitation was accepted in 1994 and the

195

The area planted was summed across that ui > 0, or is truncated above zero. This irrigation seasons where applicable. Data on error term provides deviations from the labor, family and hired, was collected for frontier. The negative sign in equation 2 each cropping season by operation and also along with positive values of ui cause summed across seasons. The number of negative deviations from the frontier for corrugated iron sheets of on the roof of the each observation. In their original paper house of the farmer was used as a proxy for Aigner at al.(1977) modeled this error term capital. Worku G. (1999) used the same as a half–Normal and also as an exponential methodology. Data were also collected on distribution. fertilizer applied, the number of oxen days In this study the frontier production function used for plowing, money spent on seed and is used for cross-sectional data as described pesticides, and the total number of by Jandrow et al (1982) to estimate farm irrigations. Household size and off farm level technical inefficiency. The computer income were collected to give an idea of the software FRONTIER Version 4.1 which household’s dependence on irrigation. gives the opportunity to specify the error The data were collected with funding from term as half-Normal and truncated half- the International Water Management Normal was used to estimate the production Institute (IWMI). function. The production function was specified as: 5. Methodology Value of output = (A, L, F, S, R, O, I, P) The study utilizes the stochastic frontier production function, as developed by where Aigner, Lovell and Schmidt (1977) and A = Total area planted (ha). Meeusen and Van den Broeck (1977), to estimate technical efficiency. L = labor used in person-days For a cross section of firms, the stochastic F = Fertilizer applied in kg frontier production function is given as: S = Amount of money spent on seed Yi = f ( Xi, b) + ei , i= 1,…,N………..1 (Ethiopian Birr, 1 USD = 8.65 Birr) where Yi is the output of the ith firm, Xi is R = number of sheets of corrugated iron of vector of inputs, and b is a vector of the roof of the house (what is the need of production function parameters. ei is an including this variable if we are going to error term made up of two components such take it out of the estimation model. that: O = number of oxen days for plowing. ei = vi - ui ...... 2 I = total number of irrigations during the In equation 2 the error term vi is assumed to year. be a symmetric disturbance that is P = amount of money spent on pesticides independently distributed as N(0, s2v). This (Birr) error term is thought to exist due to favorable and unfavorable external shocks In all the three irrigation centers, average out of the firms control and also to errors of land holding family labor days used in measurement. It is this part of the error term production are higher in the dry land setting that makes the frontier stochastic as firms than that of the irrigation. However, can temporarily be above the frontier if the irrigating farmers use much more hired value of vi is large enough (Aigner et al., labor. Output and off farm income are also 1977). very high in the irrigation setting as compared to the control group. Table 1 The error tern ui is assumed to be shows the mean values for the collected independent of vi and meets the condition data.

196

Table1: Mean values of collected data by irrigation system

Batu Degaga Doni Godino

Irrigation Dry land Irrigation Dry land Irrigation Dry land

Off farm income 187.8 96.2 1,790 1520 422 192.8 (Birr)

Household size 5.3 5.4 3.3 4.6 5.2 4.9

Irrigated land in h 0.55 1.53 0.44 1.78 0.36 0.67 (ha)

Family labor days 49.6 83.1 9.2 83.6 80.1 89.8

Hired labor days 276.6 52.8 235 43.2 23.2 26.1

Fertilizer (kg) 255.8 100.8 1,161 37.3 136 253.8

Seed (Birr) 858.2 173 2,988 177 293 305

Iron sheets 11.4 4.3 27.4 13.5 30 23

Number of irrigations 27.2 ---- 61.7 ---- 15 ----

Plowing oxen days 16.6 18.6 14.4 17.4 11.7 18.4

Pesticides (Birr) 459.3 80.1 2008 9.5 50 33.7

Value of output 16,374 4535.7 24,448 3,253 2,662 3294

either a partial translog or a Cobb Douglas 6. Model Specification using a one-sided generalized likelihood ratio test. The chow test is used to determine An important issue in this study is the choice of functional form and the distribution of the whether some of the schemes could be error term. It is not assumed that the pooled together and thus estimate on irrigation systems all have the same equation. The same is done for the dryland functional form which makes the data. comparison of technical inefficiencies The different Chow test showed that the data somewhat complex. We started with a from Doni and Godino could be pooled general translog functional form and then together. The likelihood ratio test also test whether the equations can reduce to showed that the equation for the combined Doni and Godino data could be reduced to a Cobb Douglas. The same test (LR test = 36.4) showed that the half normal

197

distribution of the error term was a better fit. The combined data for Doni and Godino has The same result was arrived for the dry land a sample size of 93. The variables included data for Doni and Godino. in the production function are the natural logarithms of output, land, number of The irrigation and dryland data for Batu irrigations, seed, fertilizers, pesticides, labor Degaga is therefore treated separately. and the number of sheets of corrugated iron Model specification tests for the irrigation (as a proxy for capital) and the interactions data showed that the equation for Batu of these variables. In the translog model the Degaga was a full translog with a truncated main variables with their interaction terms half normal error term. are 23 variables while there are 15 variables 7. Results that are believed to explain inefficiency of farmers. Estimation is done using FRONTTIER Version 4.1, a computer program for After some iterations, the likelihood ratio stochastic frontier production and cost test showed that the Cobb-Douglas function estimation, which was developed production function is the right specification by Tim Coelli for this data with 7 explanatory variables of the production function and 14 variables that

are believed to affect efficiency of farmers. 7.I Production function analysis The same test also showed that the half normal distribution of the error term is the

better distribution. 7.1.1 Irrigation Settings The final Maximum Likelihood Ratio estimates of Doni and Godino irrigation 7.1.1.1 Doni and Godino together schemes combined:

coefficient standard-error t-ratio constant 3.37 0.64 5.24 land 0.08 0.13 0.58 water 0.06 0.09 0.69 seed 0.34 0.11 3.20 fertilizer 0.28 0.08 3.39 pesticide 0.02 0.05 0.46 labor 0.36 0.10 3.50 roofing -0.005 0.04 -0.11

The dependent variable of this equation is the natural logarithm of output and all the explanatory variables are in natural logarithm form i.e the model is a double-log model. FRONTTIER Version 4.1 gives the level of inefficiency of each farmer and the mean efficiency of the system along with the production function estimates. Inefficiency effects (variables that are believed to explain farmers’ inefficiency) vary in number from scheme to scheme because variables that don’t have any variability were avoided from the model. 7.1.1.2 Batu Degaga

198

The data for Batu Degaga has a sample size of 51. The variables included in the production function are the natural logarithms of output, land, number of irrigations, seed, fertilizers, pesticides, labor and the number of sheets of corrugated iron (as a proxy for capital) and the interactions of these variables. In the translog model the main variables with their interaction terms are 23 while there are 13 variables that are believed to explain inefficiency of farmers. The different likelihood ratio tests showed that the production function for Batu Degaga can be represented by a full translog function while the error term has a truncated normal distribution. The Maximum Likelihood Estimation results of the production function for Batu Degaga irrigation scheme is: coefficient standard-error t-ratio constant 0.66 1.42 0.46 land 0.48 1.20 0.40 water 0.49 0.18 2.67 seed 0.19 0.07 2.76 fertilizer 0.03 0.72 0.04 pesticide -0.40 0.41 -0.96 labor 1.79 0.85 2.09 roofing 0.33 0.51 0.64 landsq -0.07 0.27 -0.27 fertsq 0.07 0.04 1.81 pestsq -0.04 0.03 -1.13 laborsq -0.18 0.17 -1.05 roofsq -0.04 0.15 0.29 landfert 0.09 0.30 0.30 landpest -0.04 0.07 -0.58 landlabor -0.11 0.35 -0.32 fertpest -0.03 0.09 0.33 fertlabor -0.04 0.17 -0.22 pestlabor 0.15 0.10 1.50

As was the case for Doni and Godino, is not an abundant resource in the irrigation elasticity of labor in Batu Degaga irrigation scheme. We can increase output in this scheme possesses the highest magnitude irrigation scheme by increasing the supply from among the main factors of production. of labor. Since this labor variable is defined A unit percentage change in the amount of as a labor day, if the farmer increases the labor days used will bring about a 1.8 number of labor days he or she spent on percentage change in the amount of output their farms by one percent, they can increase their agricultural output by 1.8 percent. produced. This change is also statistically different from zero. This implies that labor

199

On the other hand, if we can somehow change in seed also brings about a 0.18 increase the water supply of the irrigations percent change on output, though the and hence increase the number of times coefficients of both fertilizer and seed are farmers irrigate their land by one percent, not statistically significant. Pesticide has an we can bring about a 0.5 percentage increase unexpected negative sign. in their agricultural production. This 7.1.2 Dry land Settings coefficient of water is also statistically significantly different from zero. As per the results of the Chow test, the dry land data surrounding Doni and Godino A percentage increase in the size of land will irrigation schemes were estimated together also bring about a 0.48 percent increase in while those around Batu Degaga were output. This is in line with many other treated separately. studies whereby they confirm the very small size of land in Ethiopia being detrimental to 7.1.2.1 Doni and Godino dry land agricultural production. together The capital proxy (the number of sheets of The combined dry land data for Doni and corrugated iron of the roof of the house of Godino has a sample size of 100 and the the farmer) also showed that a percentage better representation of the data according to change in the capital of the farmer brings the likelihood ratio tests is a Cobb-Douglas about a 0.33 percent change in agricultural production function with a half-normal error output. distribution. The model has 6 explanatory variables of the production function and 11 Fertilizer and seed are the other factors of efficiency effect variables. production with a positive influence on output. A percentage change in the amount The final Maximum Likelihood Ratio of fertilizer used will bring about a 0.29 estimates of the model for the dry land percent change in output. A percentage agriculture for Doni and Godino combind

coefficient standard-error t-ratio constant 3.82 0.51 7.54 land 0.23 0.13 1.70 seed 0.60 0.11 5.57 pesticide -0.02 0.03 -0.77 labor 0.35 0.13 2.71 roofing 0.02 0.02 0.79 oxen -0.18 0.11 -1.62

In this dry land setting, increases in the are not using even local varieties of seed up usage of factors like labor, seed, land, and to their full potential. capital increases output, though at different As was the case for the irrigated agriculture, percentage increases. A percentage increase labor is not a very abundant resource even in the value of seed increases agricultural for the dry land agriculture. A unit output by 0.6 percent. Since close to all percentage increase in labor still increases farmers reported to use local varieties of output by about 0.35 percent and this seed, this coefficient showed that farmers elasticity coefficient is statistically different from zero. This may also be because of the

200

small family size of these areas from the sign of the ‘oxen’ coefficient. A unit national average or it could also be due to increase in oxen (defined as the number of the nearby high demand of hired labor by pairs of oxen used for plowing the land the irrigation farms. times the number of days they plow) will decrease output by 0.17 per cent. The use of The tiny holding size of land, as was the pesticide also has an unexpected negative case for irrigated agriculture, is restricting sign in the production function. However, agricultural output. A unit increase in the both the oxen and pesticide coefficients are size of land increases agricultural production statistically insignificant. by 0.23 percent and this coefficient is statistically significant. This could be due to Batu Degaga Dry land the very small size of farmers in the areas, The dry land data around Batu Degaga though it is a bit higher when compared with irrigation scheme has a sample size of 47. that of irrigating farmers. The estimation result showed that the better Though statistically insignificant, a representation of this data is a translog percentage change in the capital of farmers, model with a half-normal distribution of the as can be seen from the coefficient of the error term. The five main variables in the proxy variable, increases output by 0.18 per production function with their squared and cent. The statistical insignificance may come interaction terms make the total variables in from the fact that dry land farming is not the production function to be 16. There are capital intensive in Ethiopia. also 11 inefficiency-explaining variables in the model. Farmers seem to use more oxen days on their farm as can be seen from the negative The final Maximum Likelihood Ratio estimates for dry land agriculture in Batu Degaga: coefficient standard-error t-ratio Constant 13.0 1.57 8.23 Land 2.12 1.12 1.89 Pesticide 0.44 0.23 1.93 Labor -3.15 0.79 -4.0 Roofing 0.35 0.16 2.17 Oxen 1.45 0.90 1.61 Landsq 0.81 0.21 3.80 Pestsq 0.006 0.15 0.39 Laborsq 0.28 0.14 1.96 Roofsq -0.09 0.04 -2.38 Oxensq -0.43 0.34 -1.26 Landpest -0.09 0.06 -1.49 Landlabor -0.30 0.30 -1.00 Landoxen -0.08 0.336 -0.23 Pestlabor -0.03 0.05 -0.52

201

Pestoxen -0.06 0.05 -1.41 Laboroxen 0.23 0.33 0.69

Dry land farmers around Batu Degaga are agricultural output while that of capital highly constrained by their size of land. A brings 0.34 per cent increase. unit percentage increase in the size of land 7.2 Inefficiency Effects will bring about a 2.1 % increase in agricultural output, showing the relative 7.2.1 Efficiency of Dry land settings scarcity of land in the area. 7..2.1.1 Doni-Godino Irrigation Scheme Labor seems to be deployed excessively on There were 14 inefficiency effects that are this agriculture. A percentage increase in believed to explain the inefficiency of labor days will bring about a 3.1 percent farmers in Doni and Godino irrigation decrease of output. The redundant use of schemes. The maximum likelihood labor may not be surprising in the face of estimation of FRONTTIER Version 4.1 scarce land resource. gives the estimates of these variables with A percentage increase in the number of oxen that of the production function. In this days brings about a 1.5 per cent increase in computer program, the dependent variable is output. The coefficient is also statistically level of inefficiency (not efficiency). As a significantly different from zero. Increase in result, we expect the variable to have the the level of pesticide use by one per cent opposite sign of its effect on the efficiency. also brings about a 0.4 per cent increase in Technical inefficiency effects for Doni and Godino irrigation schemes are:

Coefficients standard errors t-ratios credit -0.09 0.19 -0.46 advice -0.26 0.16 -1.61 offfarm -0.00004 0.00003 -1.39 hhsize 0.07 0.04 1.89 gender 0.31 0.16 1.91 eduhh 0.03 0.02 1.75 age -0.01 0.01 -0.87 agesq 0.00006 0.0002 0.38 edumem -0.05 0.18 -0.28 extension 0.05 0.13 0.35 medslope 0.08 0.13 0.65 steeply 0.06 0.19 0.30 mdummy 0.49 0.21 2.36 tdummy 0.44 0.23 1.88 sigma-squared 0.20 0.03 5.91 gamma 0.0005 0.01 0.04

202

As can be seen from the negative sign of the Range of efficiency levels coefficients, farmers who get credit within Frequency the last three years and who are beneficiaries of agricultural advisory services, perform < 0.3 0 better in terms of efficiency than those who don’t. Since the coefficient on advice is 0.3 to 0.39 17 statistically significant, we can say that advice makes a tangible improvement in 0.4 to 0.49 28 efficiency of farmers. Farmers who have higher off-farm income 0.5 to 0.59 17 are also more efficient than those who don’t. This may be due to the fact that the extra 0.6 to 0.69 10 income may enable them to invest on improved technologies. It might also be the 0.7 to 0.79 9 case that farmers with high off-farm income, especially those in small local trades, are 0.8 to 0.89 more exposed to different ideas than those 5 who don’t have. 0.9 to 0.99 3 Males are found to be more efficient than females in Doni and Godino irrigation 1 4 schemes. The coefficient of gender is also statistically different from zero.

Farmers located at the middle and tail 7..2.1.2 Batu Degaga locations of the watercourse are less efficient than those at the head reaches. As The model for this irrigation scheme has 12 can be seen from the dummy variable for inefficieny-explaining variables. medium location of farms (mdummy) and Mean efficiency of the irrigated scheme is tail locations of farms (tdummy), farms at 76 %. That is without extra input, re- the head reach are more efficient. The allocations of the farmers’ resources can coefficients of both of these variables are increase output by 24%. statistically significantly different from zero. The other inefficiency variables were found Percentages of Technical Efficiency to be not statistically different from zero. Estimates for Batu Degaga Irrigation Scheme The mean efficiency of farmers of Doni and Godino irrigation schemes is 55.6 %. That is Efficiency Range Frequency we can increase output of these farmers by 44.4% by just re-allocating their input use. < 0.3 1

0.3 to 0.39 1 Percentages of Technical Efficiency Estimates for Doni and Godino Irrigation 0.4 to 0.49 6 Schemes Together 0.5 to 0.59 4

0.6 to 0.69 3

203

7.2.2.2 Batu Degaga dry land Areas 0.7 to 0.79 5 The final model for these farmers includes 0.8 to 0.89 16 11 inefficiency-explaining variables. The mean efficiency of the farmers is 65.6%, 0.9 to 0.99 15 implying that we can increase agricultural output of the farmers by 34.4% by 1 0 reallocating their resources. Percentages of Technical Efficiency

Estimates of Dry Land Farmers in Batu 7.2.2 Efficiency of Dry land Farmers Degaga

Efficiency Frequency 7.2.2.1 Doni and Godino dry land Range farmers: 0 There are 11 inefficiency variables in this < 0.3 model. The highest level of efficiency of farmers is exhibited in these areas. The 0.3 to 0.39 7 mean efficiency of these farmers is 79.8%. However, we can increase the output of the 0.4 to 0.49 2 farmers by 20.1% with the same level of inputs that farmers are using. 0.5 to 0.59 13 0.6 to 0.69 7 Percentages of Technical Efficiency Estimates for Dry Land Farmers Around 0.7 to 0.79 3 Doni and Godino 0.8 to 0.89 5 Efficiency Frequency Range 0.9 to 0.99 8 1 2 < 0.3 2

0.3 to 0.39 0 7.2.3 Comparison of Efficiency between 0.4 to 0.49 6 the irrigation schemes and rainfed agriculture 0.5 to 0.59 4 In two irrigation schemes dry land farmers happened to be more efficient than irrigation 0.6 to 0.69 8 farmers with respect to their own frontiers. 0.7 to 0.79 8 In Doni and Godino areas, the efficiency of irrigation farmers is 55.6 %. However, the 0.8 to 0.89 50 mean efficiency of farmers with no access to irrigation around these irrigation areas is 0.9 to 0.99 22 79.8 %. This may be due to the fact that low level of output of the dry land farming 1 0 system is forcing the farmers to allocate the small resources they have more efficiently than the irrigation farmers. The difference

204

between the dry land and irrigation farmers frontier and the frontier for irrigators is in these areas is more than 24 percentage much higher than that of dry land farmers. points. In Batu Degaga, irrigation requires 0.92 ha The mean efficiency of farmers in Batu of land, 22.2 days of irrigation, Br 299 Degaga irrigation scheme is 76% while dry worth of seed, 148.4 kg of fertilizer, Br land farmers around this scheme are 65.6 % 181.3 worth of pesticide and 244.7 efficient. The difference in efficiency of labordays to produce Br 8,103 worth of these two groups of farmers is more than 10 output; while dry land farmers in this area percentage points. are required 1.3 ha of land, Br 22.2 worth of pesticide, 121.5 labor days and 14 pairs of But we should take note of the fact that the oxen plowing days to produce Br 4,024 two types of farmers are facing two different worth of output. These show that irrigators frontiers. The irrigators are facing a higher face a higher frontier than dry land farmers. frontier than the dry land farmers and are on average more far from their frontier while dry land farmers are closer to their low 7.2.4 Marginal Productivities frontier. That is to say the availability of water for irrigators has pushed their frontier To compute for marginal Productivity of outwards and made them productive. And inputs we first non-linearize the estimated yet, the high inefficiency of these farmers production function and take the first indicates that there is even more potential to derivative of output with respect to the be exploited and the potential presented by specific input for which its marginal water isn’t yet exploited. productivity is to be determined. To compare the frontiers of irrigators and For Doni and Godino irrigation schemes dry land farmers, points on the frontier in together, the estimated production function each system are selected, specifically the is: average of the logarithmic transformations Lny = 0.34 + 0.78ln(land) + 0.6 lnwater + that were used to estimate the frontiers. 0.34lnseed + 0.28lnfert + 0.21lnpest These averages are then converted back to + 0.36lnlabor – 0.48lnroof original, non-logged, levels to give comparable input combinations on the Y = 0.34(land)0.78(water)0.6(seed)0.34(fert)0.8 frontiers of each system. Makombe et al, (pest)0 .21(labor)0.36(roof)-0.48 2001, used this methodology. The results of Taking the first derivative of Y with respect this evaluation show that Doni and Godino to each input and evaluating the resulting irrigation schemes require 0.77 ha of land, equation at the mean of regression variables 26.6 days of irrigation, Br 706.3 worth of gives the marginal productivity of each seed, 320 kg of fertilizer, Br 202 worth of input. pesticide, and 114.4 labor days to produce Br 5, 271 worth of output. On the other hand For the dry land farming around Doni and the dry land farmers surrounding these two Godino irrigation schemes require 1.22 ha of land, Y = 0.38(land)0.23(seed)0.6(oxen)- Br 194.4 worth of seed, Br 4.9 worth of 0.18(pest)-0.2(labor)0.35(roof)0.18 pesticide, 97.5 labor days, and 15.2 pairs of oxen plowing days to produce Br 2,591 Marginal productivity of inputs in Doni and worth of output. This implies that irrigators Godino areas, based on frontier regression and dry land farmers don’t face the same estimates:

205

Doni Godino Irrigated Doni Godino dry land Attribute Level marginal productivity Level marginal productivity Value of output 5,271 2,591 Land 0.77 32,529 0.2 18.6 Irrigations 27 687 --- Seed 705 15.8 194 0.05 Fertilizer 320 78 --- Pesticide 202 32.1 5 -0.66 Labor 114 97.6 98 0.05 Oxen --- 14 -0.2

All inputs have higher marginal productivity in surrounding dry land farmers are more the irrigated agriculture compared with dry efficient than the irrigating farmers, compared land agriculture. The result showed that any of course, with respect to their own frontiers. additional money spent on increasing land Both for the combined data for Doni and holdings, or to increase the number of times Godino as well as Batu Degaga irrigation farmers can irrigate their land, to supply scheme, among all the explanatory variables, fertilizer and pesticide have high return in the labor has the highest elasticity of output. In irrigation schemes of Doni and Godino. The Doni and Godino, a one-percentage change in irrigation schemes can also accommodate more the amount of labor days will bring about a farmers or the existing farmers should spend more than 0.36 percent change in output. We more time on agriculture since an additional can also increase agricultural production in laborday spent on the farm will bring about a these two irrigation schemes by more than 0.28 high return. percent if we increase fertilizer use by one percent. A percentage change in the value of seed also brings about a more than 0.34 8. Conclusion and recommendation percent change in output. In Batu Degaga, a 8.1 Conclusion unit percentage change in the amount of labor days used will bring about a 1.8% percentage The paper tried to analyze the level of change in the amount of output produced. On efficiency of farmers between irrigated and dry the other hand, if we can somehow increase the land farmers based on three irrigation schemes water supply of the irrigations and hence in Ethiopia. These schemes are Doni, Godino increase the number of times farmers irrigate and Batu Degaga irrigation schemes. their land by one percent, we can bring about a The empirical findings showed that 0.5 percentage increase in their agricultural inefficiency of farmers prevail in Ethiopia very production. A percentage change in seed also significantly, a result which is in conformity brings about a 0.18 percent change on output. with other efficiency studies of Ethiopian These coefficients are also statistically farming by Abay and Assefa (1996), Abrar significantly different from zero. (1998), Croppenstedt and Abbi (1996) and In terms of explaining the inefficiency of many others. The contribution of this paper in farmers, agricultural advices, existence of off- efficiency studies of Ethiopian Agriculture is farm income and the location of farms on the that it compares the efficiency levels of watercourse appeared to have significant irrigation and dry land farming. Though, there influence on the efficiency of farmers. Farmers are very few irrigation schemes in Ethiopia, at head reach are more efficient than farmers at much inefficiency is exhibited in the existing the middle and tail locations of the schemes. In fact in two of the irrigation watercourse. Males also happened to be more schemes, among three studied, their

206

efficient than females. The coefficients of income happens to increase the efficiency of these variables are statistically significantly farmers. different from zero. Moreover, availability of Since the marginal productivities of inputs in credit, number of years of education of the the irrigation schemes are very high, attention head of the household (in Batu Degaga only), should also be given to increase the availability existence of a member of the household who of these inputs. Government and other relevant completed primary school, age of the head of bodies should try hard to bring more land to the household and the slopes of farmlands have irrigations since the marginal productivity of their expected signs, though statistically land of the irrigated agriculture is tremendous. insignificant. The irrigation schemes of the lowland that The empirical findings also showed that there follows the Awash River should find ways to is significant inefficiency in the sampled attract more labor from the highlands where irrigation schemes. The mean efficiency of labor is expected to be abundant. The weather farmers of Doni and Godino irrigation schemes condition of this area along with the high is 55.6 %. That is we can increase output of demand for hired labor by the neighboring these farmers by 44.4% by just re-allocating large commercial and state farms has made it their input use. Mean efficiency of the Batu difficult for the smallholders to obtain as much Degaga irrigation scheme is 76 % implying labor as they want. The marginal productivity that without extra input, re-allocations of of labor in these schemes is very high. farmers’ resources can increase output by 24%. Fertilizer, pesticides and seeds should also be In Doni and Godino irrigation schemes dry better supplied to the irrigation schemes since land farmers happened to be more efficient the marginal productivity of these inputs than irrigation farmers with respect to their happened to be very high. own frontiers. The dry land farmers nearby these schemes have a mean efficiency of 79.8 %. However, irrigators are more efficient than References: dry land farmers in Batu Degaga area. Abay A. and A. Assefa. (1996): The Impact of Education on Allocative and Technical Efficiencies of farmers: The case of 8.2 Recommendations Ethiopian Smallholders. Ethiopian Journal In the face of resource constraint of many of Economics, V (1996) No. I farmers and their high inefficiency levels, Abrar S. (1998). Identifying the Sources of much attention should be given in affirming Technical Inefficiency of Ethiopian that farmers are using to the best of the little Farmers Using the (Single Stage) resource they have. Many farmers especially in the irrigated agriculture are performing way far Stochastic Frontier Inefficiency Model, from their frontier. In the irrigated agriculture, WP-1998-09/04 government and other relevant bodies should Amare H (2000). Integrated Water Resources facilitate credit facilities since those farmers Development and Opportunities for who were beneficiaries of credit are much Irrigated Agriculture (Summary), 2000. closer to the frontier. Agricultural advices Annual Report on the Ethiopian Economy, should also be given to farmers in a concerted 1999. Ethiopian Economic Association. manner since this variable was found to significantly affect the level of efficiency of Assefa A. and F. Heidhues. Estimation of farmers. Education has also a positive impact Technical Efficiency of smallholder in terms of farmers’ efficiency. Therefore farmers in the Central Highlands of government should intensify its efforts to Ethiopia, Ethiopian Journal of Agricultural expand education to the rural sector. The fact Economics I (1996) No I that families where there is at least one Awulachew, S.B., D.J. Merry, A.B. Kamara, member who finished primary school are more B. Van Koppen, F. Penning de Vries, E. efficient further justifies the need for Boelee, G. Makombe (2005). Experiences expanding education. Activities that create and opportunities for promoting small- more off-farm income to the rural sector are scale/micro irrigation and rainwater also things to be encouraged since off-farm

207

harvesting for food security in Ethiopia. from Chaj in the Upper Indus Basin. Colombo, Sri Lanka: IWMI. V 86p International Water Management Institute, (Working paper 98). Colombo, Sri Lanka Catholic Relief Services, Programmatic Jondrow J., K. Lovell, I. Materov, P.Schmidt. Environmental Assessment of Small-Scale On the Estimation of Technical Irrigation in Ethiopia, September 1999. inefficiency in the Stochastic Frontier Coelli T., P. Rao, G. Battese. Efficiency Production Function Model, Journal of Measurement Using Stochastic Frontiers Econometrics, 19(1982): 233-238. (1998) Kamara A.B., B. Van Koppen, L. Magingxa Croppenstedt A. and M. Abbi (1996). An (2002): Economic Viability of small-scale Analysis of the extent of the Technical Irrigation Systems in the Context of State Efficiency of Farmers Growing Cereals in Withdrawal: the Arabie Scheme in the Ethiopia: Evidence from three Regions, Northern Province of South Africa. Ethiopian Journal of Economics, V (1996), Physics and Chemistry of the Earth (2002). No. I Kloezen W. H. (1998). Measuring Land and Debertin L. D. Agricultural Production Water Productivity in a Mexican Irrigation Economics. Macmillan Publishing District, Water Resources Development, Company, New York and Collier Vol. 14, No 2, pp 231-237, 1998. Macmillan Publishers, London. Makombe G., M. J.Makadho and R. K. Federal Democratic Republic of Ethiopia Sampath (1998). An Analysis of the Water (2001). Ethiopia -Water Sector Strategy, Management Performance of smallholder 2001. -Ministry of Water Resources, irrigation schemes in Zimbabwe. Irrigation Water Sector Development Programme, and Drainage Systems 12: 253-263, 1998. 2002-2 016. Irrigation Development Makombe G., R. Meinzen-Dick, S. P. Davies Programme. and R.K. Sampath. An Evaluation of Bani Forsund F., K. Lovell, P. Schmidt . A Survey (Dambo) Systems as a Smallholder of Frontier Production Functions and of Irrigation Development Strategy in their relationship to Efficiency Zimbabwe, Canadian Journal of Measurement, Journal of Econometrics, Agricultural Economics, 49 (2001): 203- 13 (1980): 5-25 216 Government of the Republic of Ethiopia Ministry of Water Resources, 2000. (2003). Rural Development Policy Schmidt P., K. Lovell. Estimating Technical Strategies. Ministry of Finance and and Allocative Inefficiency relative to Economic Development. Economic Policy Stochastic Production and Cost Frontiers, and Planning Department. Addis Ababa, Journal of Econometrics, 9 (1979): 343- April, 2003. 366 Greene W.. On the Estimation of a Flexible Schmidt P., K. Lovell. Estimating Stochastic Frontier Production Model, Journal of Production and cost Frontiers when Econometrics, 13(1980): 101-115 Technical and Allocative Inefficiency are Gujarati N. D. Basic Econometrics (1995), Correlated, Journal of Econometrics, 13 Third Edition. McGraw-Hill, Inc (1980): 83-100 Gulilat B.. Water Resources Policy and Water Stevenson R.. Likelihood Functions for Code as Related to Irrigated Agriculture Generalized Stochastic Frontier (Summary), 2000. Estimation, Journal of Econometrics, Hussain I., R. Sakthivadivel, U. Amrasinghe, 13(1980) M. Mudasser and D. Molden, 2003. Land Sherlund S. M., C. B. Barrett, and A. A. and Water Productivity of Wheat in the Adesina. Smallholder Technical Efficiency Western Indo-Gangetic Plains of India and with Stochastic Exogenous Production Pakistan: A Comparative Analysis. Conditions, 1998. WP 98-15, Cornell INTERNATIONAL WATER MANAGEMENT University. INSTITUTE, RESEARCH REPORT 65. The Ethiopian Economic Association. Annual Hussain I., M. Mudasser, M. A. Hanjra, U. Report on the Ethiopian Economy. I Amrasinghe and D. Molden, 2004. (1999/2000) Improving Wheat Productivity in Pakistan: Yusuf K. (2004). Assessment of Small Scale Econometric Analysis Using Panel Data Irrigation Using Comparative Performance

208

Indicators on two Selected Schemes in Analysis of Own-Operators and Tenants Upper Awash River Valley. Msc THESIS, (The Case of Ethiopia). MSc Thesis, IRRIGATION ENGINEERING, ALEMAYA Economics Department, Addis Ababa UNIVERSITY. University. Worku G., 1999. Technical Efficiency of Cereal Producing Farmers: A Comparative

209

Impact of irrigation on livelihood and food security in the modern Hare river irrigation scheme in Southern Ethiopia

Judt Christine1, Loiskandl Willibald1, Ruffeis Dominik2, Hauser Michael2, Seleshi Bekele3, Hagos Fitsum3

1University of Natural Resources and Applied Life Sciences, Vienna, 2BOKU Research for Development Forum (DEV-FORUM); Department of Water - Atmosphere Environment, Institute of Hydraulics and Rural Water Management, Vienna, Austria, 3International Water Management Institute [email protected]

Abstract agricultural office, health centre, drinking The purpose of the undertaken study was to water points, school, electricity etc. evaluate the impact that the modern Hare river The major trigger was the introduction of a irrigation scheme had on household food new banana type so that farmers changed from security as well as on lifestyle changes of the food crops to cash crops to earn a higher population in the study site Chano Chalba. income. Following, the wealth situation of the This was done on the basis of the FAO food population ameliorated but less food crops are security pillars access to food, availability of produced and people become more dependent food, utilization of food and the overall factor on the local market. The infrastructure of the of food stability. RRA tools were used to study site developed in a positive way but still conduct a before-after comparison, considering education, especially on food issues, are a ten years period. The quantitative data was needed to have a sustainable repercussion and analysed using SPSS and/or Excel and simple to secure people’s health and food situation. statistical measures such as cross tabulations, Further positive changes on the food situation frequencies, percentages and means gave a could be able if the higher income was utilized visible overview of the outcomes. more efficiently and if the construction of the The modern irrigation scheme did not affect modern irrigation scheme had been more the livelihood and food situation directly but appropriate and by incorporating the farmer’s indirectly through other modernizations that requests. came with and after the construction of the modern main canal, e.g. road, merchants, Key-words: poverty, irrigation, livelihood, food security

1. Introduction According to the Aggregate Household Food Security Index (AHFSI)18, 1.1. Background on food security and established by the FAO (1995), Ethiopia irrigation had an index below 65, i.e. critical food In the 1996 Rome Declaration on World Food security status, between 1991 and 1993. Security, food security is defined as “Food that is available at all times, to which all persons have means of access, that is nutritionally 18 adequate in terms of quantity, quality and .The AHFSI calculates the "food gap" between the variety, and is acceptable within the given undernourished and average national requirements, the culture.” instability of the annual food supply and the proportion of Further, the FAO speaks of food security on undernourished in the total population. The index ranges household basis when all members of a from 0 to 100, with 100 representing complete, risk-free household can be supplied with sufficient and adequate food, whether through their own food security and zero, total famine (FAO, 1995). production or through buying of food.

210

In Ethiopia, irrigation has a long tradition order to have a significant effect in reducing (Kloos, 1990). One of the main targets of preschooler morbidity and improving child irrigation systems is to fortunate agricultural nutritional status.” production in qualitative as well as in Among the reasons for negative impacts were quantitative meaning (Mengistu, 2003). reduction in the household’s production of Harvests shall be enlarged so that people either food crops, increased food prices, loss of produce enough food for the non-harvest time household access to land due to changing or to sell their overproduction and earn some tenure relations and household expenditure money to buy food. Another opportunity to patterns favouring durable consumer goods. produce more food crops is irrigated gardening, an activity mainly done by women. 1.3. Statement of the problem High reliance of the people on their own food In Ethiopia, there has been a revival of production has a direct impact on local food irrigation during the last decades in order to availability and on accessibility. Some coping enhance rural development and food security strategies, which are eroding the household (FAO/WFP, 2006). Given that 85 percent of asset base, are used by the most vulnerable the people are employed in agriculture population to survive (FAO/WFP 2006). (Mengistu, 2003), developing this sector could help to reduce poverty and enhance food As IFAD stated in 2005, irrigation has the security of the majority of the Ethiopian potential to reduce food insecurity by several people. factors. “For the farmer, the soils, crops, As Lankford (2003) argues there must be a livestock, weather, water, nutrients, pests, positive balance of benefits against risks and markets, income, outgoings, shelter, transport, costs of irrigation. A more secure and fuel, property, family and social networks, and increased crop productivity, improved much more, all form part of the integrated planning and timing of start of the cropping environment in which he or she makes a season and extended harvest season, raised livelihood. Improving food security through number of jobs and income are some knock-on irrigation affects or is affected by each of these effects that show how irrigation facilitates aspects. A major challenge is to design economic transactions and improves meaningful integrated solutions to the real livelihoods and the wealth and infrastructure of problems faced by farmers” (IFAD, 2005). whole villages (Lankford, 2003). Therefore the consequences of irrigation 1.2. Irrigation and food security in systems both positive and negative have to be Ethiopia assessed so that policy makers, government Case studies undertaken by Thompson (1991), and other NGOs can react and, finally, burst Lall and Broadway (1994) and Adem (2001) the vicious circle of poverty and enhance food already highlighted the positive impacts of security. enhanced accessibility to water for irrigation. Besides an extended growing period, a higher 1.4. Objectives of the study variety in food and cash crops and, as a result, As irrigation development is often associated an increase of cash income, food shortages with cash crops, irrigation investments’ could be reduced. Nearby, the target contribution to food security is often communities improved the infrastructure of questioned. This study gives, among other their village and standard of living in general. things, an answer to this question, in case of Negative impacts of the projects were over Chano Chalba and the modern Hare river watering of some fields, overproduction of irrigation scheme. certain crops and competition for water from It is figured out how the modern irrigation nearby communities. canal changed the agricultural production and livestock holdings/population, income sources Kennedy et al. (1992) compared the effects of and expenditures, health situation, hygienic cash crop schemes on health and nutrition in standards, market situation, lifestyle of the six countries. They concluded that “increases households and the infrastructure of the study in income have to be accompanied by site. improvements in the health environment in

211

For the purpose of investigating the food, utilization of food and the overall factor perceptions of household food security, the of food stability. research focused on the four dimensions of Following, the established hypothesis food security: availability of food, access to concerning the four dimensions of food security, were:

Since the modern irrigation scheme was introduced, households have a better and 1. Availability: more sustainable availability to food than before (whether through their own

production and/or buying). 2. Access: The access to food and other assets are better developed today compared to ten years ago. (markets are better available; enough money to buy food) Since the construction of the modern canal, people have a more varying and balanced diet. Since the construction of the modern canal, people have more time for cooking 3. Utilization: and education thus people have a better knowledge of a balanced diet and related

effects. Since the construction of the modern canal, people have more meals a day, are not restricted in the choice of their food, and are healthier and physically stronger. 4. Stability: The food security situation is more solid and offers more stability throughout the year and periods of stresses, shocks and in terms of seasonality.

In concrete, the results of this case study 2. Description of the study area provide an insight of the impacts that the modern Hare irrigation scheme has primarily 2.1. General description of the on household food-security but also on the country environment, economy and society. It Ethiopia is a landlocked country in the horn of contributes to the project “Impact of Irrigation Africa. It comprehends 1 120 000 square Development on Rural Poverty and kilometres total land area, whereof 8 percent Environment”, coordinated by the International are farmed by smallholder peasants and about Water Management Institute (IWMI), the 3 100 000 hectares are fallow. The total area of University of Natural Resources and Applied grazing and browse is estimated to be up to Life Sciences Vienna (BOKU) and ARC 65 000 000 hectares, of which 12 percent is in Seibersdorf Research in association with the mixed faming, the rest in pastoral areas (MoA, BOKU Research for Development Forum 2000 as cited by Mengistu, 2003). (DEV-FORUM). Agriculture plays a central role in economic and social life. It is the leading sector in TheT externalities shall also be valuable for national economy, composing 40 – 50 percent other researchers and stakeholders and of GDP and around 90 percent of export influence the thinking on approaches to earnings. About 6 million people are irrigation development. The concerned chronically food insecure and depend on food population gets a better view of their present aid throughout the year, another 10 million are situation, impact of outside interventions on considered as vulnerable (FAO/WFP, 2006). their livelihoods and consequently are enabled Subsistence agriculture is almost entirely rain to make better-informed decisions about their fed and yields are generally low. The existing future livelihood strategies. irrigation potential is far from being reached (Mengistu, 2003). Thus the study focused only on the named study site, the gained results are just valid at The population is estimated to 76.5 million local level and cannot be generalized to a (World Bank, 2007) and is growing rapidly, larger population. with about 2.9 percent annually (Kebede, 2003). 38 percent of the population are underweight, another 47 percent stunted. Infant mortality rate is 107 per 1000, 47 percent of children

212

aged under five are malnourished and 10.5 and soil fertility is also good (Agricultural percent affected by wasting (FAO/WFP, Office, 2007). 2006). 2.2.4. Farming system and cropping season 19 2.2. Description of the study site From February to May, food crops (e.g. maize, sweet potato) are planted while cash crops are 2.2.1. Location of the study site and the both planted and harvested. During the modern irrigation scheme summer months, from June to September, food Chano Chalba is a Peasants Association that crops are harvested and, as it is rainy season, lies in the SNNPR, Gamo Gofa Zone, in the cash crops are also planted and harvested. South of Ethiopia. Hare River is one of the From November to January, when it is dry four main rivers draining to the nearby located season and the most critical time during the Lake Abaya. year, there are no agricultural activities undertaken (Agricultural Office, 2007). The Hare irrigation scheme comprises a total irrigable area of 2224 ha and there exist three The more recently introduced cash crops different irrigation schemes. In Chano Chalba, banana, mango, avocado and papaya have the the modern Hare right side irrigation scheme is advantage that they have to be planted only used (Bantero, 2006). The community has a once (trees!). From then on, they are less time common diversion weir with the neighboring consuming because weeding is not that Kebele Chano Mile. This weir is located next important compared to sweet potato or maize. to the asphalt road, leading both to Addis Moreover, these cash crops, especially the Abeba and Arba Minch. banana, can be harvested nearly every two months depending on the effort the farmer has 2.2.2. Climate taken. Chano Chalba, lying in the lowlands at 1169 m a.s.l. near to Arba Minch, has a tropical 2.2.5. Land holdings and distribution climate with maximum and minimum Chano Chalba comprises a total area of 799 ha temperatures between 30.3°C and 17.4°C, of which 649 ha are cultivated, in average respectively. landholdings from 0.5 to 3.50 ha. 199 ha are February and March are the hottest months cropped annual, 450 ha perennial, and all of with rare rainfall, while June to August and the agricultural land is irrigated. Forests cover November to January are more moderate with 40 ha, grassland 20 ha and 90 ha are occupied higher precipitation. The monthly average by others (Bantero, 2006). rainfall recorded from 1970 to 2006 at the Because of the high population growth, the Arba Minch Farm and Arba Minch University resettlement and return of soldiers, land station is in total 829.3 mm, respectively holdings became smaller compared to former (Bantero, 2006). times and it is not possible to increase the land size. The only possibility is to rent land from 2.2.3. Soil fertility poor people, their way to increase cash income Along the modern irrigation canal in Chano (Group discussion, 2007). Chalba, soil is sandy at the head of the modern main canal where soil fertility is poor. In the 2.2.6. Population and family middle parts, there is loam and soil fertility is constitution very good. At the tail of the canal, people As already mentioned, Ethiopia’s population is brought soil from another place to increase soil growing rapidly. In Chano Chalba, a total fertility. Beside this soil, there exists silty loam population of 2980 people was recorded in 1988 and increased to 6200 inhabitants in 2005 (Catholic Church Mission, 1988; Agricultural 19 During the realisation of this study, rare data on Office, 2007). Households increased from 445 the study site were available. Therefore, the in 1998 up to 732 households in 2005 information given in the chapter “Description of the (Agricultural Office, 2007). study area” is based on the statements from key informant interviews and group discussions.

213

The interviewed households consisted in not enough water in the canal, people cannot average of 8.5 members. irrigate their fields. The construction of a dam is seen as solution to this problem. Further 2.2.7. Problems and advantages of the problems in this part of the canal are siltation modern irrigation scheme and damage by farmers. The modern irrigation system was built up and paid by the Chinese Government in One of the main advantages of the concrete cooperation with the Ethiopian Government. canal is its robustness. Digging the destroyed Main purpose of its construction was to and blocked canals (e.g. after heavy rains) increase the amount of irrigation water, again and again to make the water flow was a enhance cash crop (cotton) and food crop time consuming activity. Besides, if the water (beetroot) production and to decrease water was standing too long, people as well as losses (Agricultural Office, 2007; Belete, livestock that drank from that water became ill 2006). easily. Unfortunately, the field canals are still traditional. During the construction, there was a disagreement between the Ethiopian and Chinese Government because the Chinese did 3. Methodology not build the main canal the way the Ethiopians wanted so the Ethiopian 3.1. Sampling procedure Government redrew every responsibility for Purposive sampling was used to examine the the project. However, the Ministry of impacts the modern irrigation canal had all Agriculture brought the material to construct over its longitude. Indicators for the sampling also modern field canals but unfortunately the were the household’s wealth situation, average Chinese had already stopped before the main size of land holdings, and the use of both the canal was finished. modern scheme today and the traditional before, so that the changes of a ten years Although the modern canal led to a lot of period could be deducted. The sampling advantages, one of the main objectives of the included 9 households each from the head and construction has failed because water loss is tail of the main canal while 11 are from the still a problem. Moreover, increased water middle (Table 1). The difference of two more usage by farmers and salinity are constraints in the middle is a result of the unreliable for adequate irrigation practices. sampling procedure of the Kebele officer that At the head of the modern canal, the field the researcher faced in the beginning. canals break when there is too much water, especially during rainy season. In the middle, the canal is lower than the fields and if there is

Table 1: Sampling scheme of households for household interviews Location of main Head Middle Tail farmland Wealth Category of 3 R 3 A 3 P 4 R 3 A 4 P 3 R 3 A 3 P HH Total HH interviewed 29 observation, transect walks and questionnaires from the selected households. 3.2. Data collection methods Maps and timelines were established in Due to time and finance limitations but to get cooperation with the participants and gave a reliable in-depth information, rapid rural visible overview about the construction of the appraisal tools were used. modern irrigation canal, availability of Primary data was collected from key irrigation water throughout the year and informants, local people, focus groups, experts, agricultural practices.

214

Rankings accompanied the household type (Hawesh banana). The only disadvantage interviews and were used to evaluate of the Asmara banana compared to the preferences or importance of main crops Hawesha banana is its higher water demand. planted, income sources and expenditures. Secondary data was obtained by In former times, farmers were mainly planting comprehensive literature review and from food crops (maize, teff, sweet potato, beetroot) offices operating in the area, to get the for their own consumption as well as cotton as necessary data on agricultural production data cash crop. In the meantime, nearly all of the and market prices. farmers changed totally to cash crop production (banana, mango, avocado), defined 3.3. Data management and analysis as crops that are mainly planted for selling. The collected data was recorded in a laptop to Some of the farmers still have a little piece of avoid data loss. Descriptive data was coded land where they plant at least some food crops and analysed in the light of the literature for the household’s own consumption while reviews. Quantitative data was introduced into others divide their plot into different parts to SPSS and/or Excel and simple statistical plant on one half cash crops and on the other measures such as cross tabulations, half food crops. frequencies, percentages and means were used to reduce the volume of data, making it easier Data from the Agricultural Office in Chano to understand. Chalba showed the increase in production both of cash crops and food crops during the last ten 4. Results and discussion years. The increase in food crops is explained with the growing population size, the higher 4.1. Availability of food cash crop production by the shift to monoculture of banana and mango. 4.1.1. Agricultural practices: From According to the statement above, the situation food crops to cash crops of food produced by the households for their Cropping patterns changed definitively since own consumption was better in the past. At the the modern irrigation system was built and time this study was carried out, cash crops since the Kebele got its own Agricultural supplied the households with the necessary Office in 1996 that introduced a new banana money to buy food from the markets. type (Asmara banana) in order to increase farmer’s income and to improve their The magnitude of the shift to cash crop banana livelihoods. is also shown in Table 2 below. For 27 of the This banana tree is smaller, easier to plant and 29 interviewed farmers (93.1%), banana is the harvest and harvest-outputs are higher; it needs most important crop today. On second rank less treatment thus is less labour- and time- comes maize (13 people, 44.8%), followed by consuming. The linear form of the fruit makes the recently introduced cash crop mango (8 it better transportable than the “older” banana people, 27.6%).

215

Table 2: Importance ranking of crops (CC = cash crops; FC = food crops); first three ranks Rank Today Before Maize Mango Cotton Maize Sweet Banana 1 (FC) (CC) (CC) (FC) potato (FC) (CC) 93.1% 44.8% 27.6% 37.9% 27.6% 10.3% Maize Mango Sweet Sweet Maize Cotton (CC) 2 (FC) (CC) potato (FC) potato (FC) (FC) 20.7% 44.8% 27.6% 17.2% 37.9% 17.2% Cotton (CC), Sweet potato (FC), Maize Maize Sweet Cotton (CC) 3 (FC) (FC) potato (FC) 24.1% Each 20.7% 34.5% 27.6%

Source: Household interviews/rankings, 2007.

Concerning the diversity of crops being grown, close to the fields and its dung was used as a there is less different kind of food crops but natural fertilizer. more different kind of cash crops grown on the Since changing to mono cropping banana, soil main field today. is loosing steadily its fertility (Agricultural In average, farmers are planting 2.14 different Office, 2007). Furthermore, the farmers are not kinds of food crops and 2.21 cash crops today, familiar with the use of industrial fertilizer and compared to 2.52 food crops and 1.72 cash have little knowledge and practice. The crops before. attempt of the Agricultural Office (1994) to keep soil fertility by applying chemical Reason number one for the change to mono fertilizer was abolished soon because of the cropping is the higher income farmers get from high costs and the bad introduction and the recently introduced cash crops banana and performance (Agricultural Office, 2007). mango. Although the new banana type needs Instead, the Agricultural Office started recently more water, it seems to be more resistant to give lessons on generating compost. While against drought, compared to maize. it was only 9 farmers out of 29 (31%) who Moreover, food crops do not grow properly were using organic fertilizer before, the next to the big banana trees that consume all of number increased to 24 (82.8%) who started to the water and give a big shadow. use compost recently to keep and increase soil Other recently introduced fruit trees, especially fertility. mango but also papaya, already show to be a big success, not like the attempt of a local 4.1.3. Use and practices of irrigation water cereal called “Taleba” where soil-fertility did Farmers in Chano Chalba were irrigating with not fit and the crop was abolished soon. a traditional system since long time ago. Today, all farmers in Chano Chalba are Since the farm size of the households has irrigating both their main field and garden. decreased, people have to make a more precise Regarding the irrigation time, the Head of choice on what to grow thus food crops have Water User Association who is in charge with declined for the benefit of cash crops and the coordinating the irrigation scheme, allows the related higher income. farmers to irrigate two or three times a year but More detailed, 14 farmers (48.3%) use more, 9 irrigation time also depends on soil type, crop (31%) less and 6 (20.7%) the same space on type and on the weather conditions. their plot to grow cash crops today. From the 29 interviewed people, 25 (92.6%) 4.1.2. Practices to keep soil fertility are irrigating temporary, i.e. two to maximum People had another cropping pattern before the four times during the year, and 2 persons modern canal was constructed. They changed (7.4%) use the irrigation water for their crops the crops on the field or even left a part fallow all over the year. No further explanations were so that it could regenerate; livestock was hold given to explain this case.

216

By contrast, ten years ago it was 15 farmers diet above all with micro-nutrients like (57.7%) who irrigated temporary while 11 vitamins and minerals. The left-over is sold by (42.3%) irrigated all over the year. the women and girls on the local market or on the street to earn some money that they can One of the main purposes of irrigation is to manage themselves. increase the agricultural production. But in case of Chano Chalba it is difficult to compare 4.1.6. Livestock the different crop types and, comparing the In Chano Chalba, oxen, cows, goats, chicken, information with the data given by the donkeys and sheep are and were kept as cattle. Agricultural Office it is hard to believe that Livestock is generally a household asset and half of the people stated that their harvest has needed to overcome periods of food shortages. decreased compared to ten years ago. Therefore, it is usually used for the households It is more reasonable and believable that 25 of own consumption and only in certain cases, the interviewees (86.2%) claimed that their mainly from wealthier farmers, sold. agricultural production is still influenced by Meat, eggs, milk and butter are incorporated in lack of rain. the daily diary but not consumed often. Milk is known to be healthy and therefore saved for 4.1.4. Storage of food crops and cash crops young children and pregnant or lactating In the past, when people were mainly growing women. maize as food crop, they stored the harvest in a barn next to the house. Today, the cash crops The given data on livestock changes are banana and mango are usually sold directly to controversial. According to the information of the merchants and the small amount of maize the Agricultural Office of the Kebele, livestock farmers still plant is often not necessary to population on household basis doubled since store. While all of the 29 interviewed farmers 1997. But 24 (82.8%) of the 29 interviewed had stocks ten years ago, it is only 16 (55.2%) people claimed that they have less livestock today. From the ones who still have maize today compared to ten years ago. stocks today, 7 (24.1%) said that their reserves The reason for the decrease of livestock has last longer, 8 (27.6%) shorter and for 1 farmer nothing to do with the construction of the (3.4%) it remained the same. modern canal but is a problem of the small landholdings and the high population density. 4.1.5 Home gardens Thus, the cattle have to be brought to the The importance of the home gardens lies on highlands or near to the lake while ten years the bigger variety of food for the household’s ago there was enough space near to the fields consumption and for women to get their own and the cattle’s dung was used as natural income by selling the surplus of these crops. fertilizer. Moreover, most of the animals died because of In the study site, only 15 (51.7%) of the 29 the sleeping-sickness that is transfused by the interviewed households had a home garden ten tse-tse-fly. That problem was removed five years ago while everyone (100%) is in years ago when they got nylon nets that caught possession of one today. One of the reasons is and killed the flies. Furthermore, the that their fields were nearer to their houses, so Agricultural Office started with livestock they did not need a garden but planted their vaccinations in Chano Chalba once a week. food crops directly on the field. Concerning the above mentioned diversity of Thus people have less livestock, the food, farmers are planting 4.43 different crop availability of livestock products like butter, types, mainly food crops, in their gardens milk, cheese and meat has declined. Prices at today. Ten years ago, the 15 households who the market are considered as too expensive so had a garden had only 1.46 different food there is less diversity of animal products, crops. especially in proteins, in their diet.

According to this result, the people have a 4.1.7. Income and expenditures bigger choice in food, especially in fruits like As can be seen from Figure 1, the wealth mango, avocado, papaya, and can enrich their situation of the Kebele developed in a positive

217

way. There are more “richer” and “average” decreased. households today while the category “poor”

Figure 1: Development of wealth categories in Chano Chalba, 1995-2006

Wealth Categories in Chano Chalba

Rich Average Poor

100% 89% y 80%

60% 52% 40% 38% 20% 6% % Wealth Categor % Wealth 5% 10% 0% 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Year

Source: Agricultural Office, February 2007.

• Income Since ever, cash crops were seen as main • Expenditures source of income in Chano Chalba. Due to the Income from cash crops is generally managed shift to cash crops, people have a higher and by the male household head, mostly in more constant income today, although the accordance with his wife. In female headed number of different income sources decreased households it is the women who decided how significantly from 3.17 to 2.24. to spend the income.

Selling of home made talla, arake or honey, Generally, people spend more money on food income sources ten years ago, is considered as purchase, housing, education and medical less important today. The production has expenses today than ten years ago while decreased for both selling and own expenditures on clothes as well as consumption. Reasons for the abandonment of governmental taxes play a minor role. Since beehives, was spraying of DDT several years electricity came recently, it was not of ago to kill the mosquitoes that transmit malaria importance in the ranking ten years ago, but also displaced the bees. neither expenses on mobiles, television and else.

218

Figure 2: Importance ranking of expenditures; (mentioned most often on rank 1 to 3)

Most important expenditures

before today

25 20 15 10

Frequency 5 0 food car, radio group housing medicine payments television, livestock vaccination

Source: Household interviews/ranking, 2007. with each other because if there are more The result that most of the people placed people who need to buy food from the market housing on first rank is connected with the fact but less who sell their food crops there, the that people were often moved to another place competition is higher and following prices on of the Kebele during the resettlement and that the markets increase. they could afford iron sheet roofs instead of grass roofs because of their higher income When people spent money on food purchase from the cash crops. Housing is seen as a big ten years ago, it was because they wanted a and important investment that is only made bigger variety. They bought food they did not once. Some also put it on first place because it produce themselves but that were considered was an expenditure of a high amount of money as healthy, like peas and beans. Others, on the that they had saved for a long time for exactly contrary, just bought salt and oil. Today people that purpose. do not have enough food crops so they need to When asking whether the households spend buy even part of their staple food. more, less or the same amount of their income to purchase food today, the answers were as 4.2. Access to food shown in Table 3: Money spent on food purchase, today compared to ten years ago 4.2.6. General changes in the Kebele’s infrastructure The infrastructure developed to a great extent Table 3: Money spent on food purchase, since the modern irrigation system was today compared to ten years ago constructed and brought a lot of benefits to the Kebele: the Agricultural Office, the Health Answer Frequency % Centre, a road, electricity, the school, a More 24 82.8 telephone station, potable water points, flour Less 4 13.8 mills, etc. The money for these improvements Same 1 3.4 was provided both by the Government and the inhabitants of Chano Chalba. Source: Household interviews, 2007. Farmers consider the road as most important

innovation because it made the access to the The reason for the bigger amount spent on markets easier, merchants came from Addis food purchase is first because people have less Ababa and people could sell their crops. That food crops and livestock, thus they produce is underlined by the 15 people (53.6%) who less themselves to eat, and second that food abnegated that they would have got the same prices are higher today. That stands in relation

219

income from cash crops ten years ago while 5 preferred dishes of the people it is eaten up to people (17.2%) answered in the affirmative 2 or 3 times a day. Moringa stenopetala is and 8 (28.6%) could not give an answer. related to the moringa oleifera which has its Explanations when abnegating never included origin in the North of the Himalaya and was the irrigation scheme but the lack of already used as a nutritious supplement by old knowledge about the value of the bananas and traditional healers in Asia and Africa to help the absence of the road, thus the missing sickly people. The fruits contain important merchants. minerals, vitamins, amino acids and enzymes. It has sufficient amounts of iron, zinc and 4.2.7. Markets and market prices copper as well as 17 to 18 amino acids that are the most important components for the Chano Market takes place every Monday and construction of proteins and the immune is the most important market for the people of system (Schulz, 2006). Chano Chalba. People are coming from the Studies in Senegal, Malawi and Tanzania surrounding Kebeles (Chano Mile, Chano confirmed its positive effect on child growth Dorga, Kola Shara), the highlands (Dorze, when consumed regularly and it was diagnosed Chencha), Arba Minch and from Lante, to that it decreased the proneness to cold- participate in this social event. infections, worms and certain skin allergies up The next bigger market people tend to go is in to 70%. Arba Minch, a 2 hours walk, while the Moreover, it is of special interest because it preferred Chano Market is reached within half stimulates lactation and it was proved that it an hour by all the inhabitants. removes hazardous materials from water The better transportation system to (Schulz, 2006). surrounding villages is connected to the new road because people could catch one of the Usually, people eat the same the whole day or minibuses that cruised between Lante and two different kinds of meals, depending Arba Minch if they had the necessary money. whether they are on the field or not. The typical daily meal consists of roasted maize or Although there is more food available on the beans, tea made from the coffee leave, markets today, both in quantity and quality, it “halakko”, served with the local bread or sweet is also a fact that people get less food for the potatoes. All around the day, people eat same amount of money. bananas, mangos and avocados, some fresh Data on retail prices at Arba Minch from 1996 made bread and drink water with lemon. until 2005 were looked up on the most important foodstuffs and show the steady People are used to eat maize, in best case increase. differently prepared, three to four times a day. 12 of the interviewed farmers (41.4%) stated 4.3. Utilization of food that they can afford to prepare their meals from different kind of ingredients and 3 households 4.3.6. Local food (10.3%) eat the same dish all over the day. Ethiopia is famous for “enjera”, a flat, pancake like bread made from teff. Enjera is eaten with 4.3.7. Number of meals different kind of sauces, that consist, The number of meals depends whether people depending if fasting day or not, in meat or spend the whole day on the field and how far vegetables. away the field is located. As mentioned before, The situation is different in Chano Chalba. the banana trees do not need so much There, maize is staple food and they usually treatment as the time-and work consuming make the local bread, or other dishes like cash crops. porridge, from it. Because of the higher price, That is why 7 (24.1%) of the 29 farmers eat enjera made from teff is mostly eaten for more meals today, 4 (13.8%) less and for 18 holidays, when people afford “better” food. interviewees (62.1%) there was no change in number of meals. “Halakko”, moringa stenopetala, grows in When it comes to the choice of food and nearly every garden and contributes a lot to the according to the FAO definition of food daily diet. Although it is not among the most security that “...food is acceptable within the

220

given culture”, only 7 people (25%) said that 5. Conclusion and recommendations they can eat whatever they want today while it • For the irrigation scheme and its was 21 (75%) ten years ago. The reasons are stakeholders again connected with the decreased food crop As can be seen from the results of Chano production and the increase of market prices. Chalba, irrigation requires not only governmental support. Management and 4.3.8. Healthiness of food and “food rehabilitation should be done from a bottom-up education” viewpoint and pro-poor, so that farmers “own” Concerning the healthiness of food, it is the irrigation schemes and make it more difficult to make a clear statement. As already efficient. mentioned, people have a bigger variety both Although it is difficult for a non-irrigation- from fruits in their garden as well as on the expert to say whether the dam, that is wished market. Although some products became more from the farmers of the study site, made sense expensive, they could be substituted by and provided the necessary water that is cheaper ones, and as it turned out during the needed during dry season or not, it is clear that interviews, people have quite a good people would make a bigger effort to sustain knowledge about food preparation, also due to the irrigation system. Water losses might also the lessons they got from the Health Centre. It be smaller if the field canals were made the depends on the people how important they modern way, from concrete. consider a healthy diet and how they spend their income. • For Chano Chalba and its inhabitants Although 9 (31%) of the 29 farmers stated to Chano Chalba developed step by step and is have more times of food shortages today still moving on since they got the modern compared to ten years ago, the author thinks irrigation scheme. The road, the Agricultural differently because all the people seen in the Office, the Health Centre, electricity etc., all study site did not show any observable signs of interventions improved the situation of the under nutrition and, furthermore, if people inhabitants to some extent. Only regarding have enough money to renew their houses, their food situation, it seems that the positive they should have enough money to eat, too. change has not taken place yet. In contrary, Information of the Health Centre also showed people are somehow restricted in their choice that there are no noteworthy problems with of food. The explanation of this problem from malnutrition and data gathered from the Arba the author’s point of view is both the missing Minch Catholic Church Medical Section education of most of the people as well as the showed that the low incidence of malnourished focus on other targeted values of the children that existed in former times also population. decreased. If there are still cases of Although the author would say that the food malnourished children, the UNICEF distributes situation of the study site is not in a bad state, food and supplements (Alimi) to the Health yet no one is starving or suffering from under Centre of the Kebele where women can get it nutrition and people look quite healthy, there is for free. no doubt that it could be much better. Regarding the people’s values, they have to 4.3.9. Health situation, sanitation decide themselves what is most important to and nutrition them. Reasons that the health situation improved in Questions pop up when thinking about what Chano Chalba was induced by the construction would happen if the market for the cash crop of the potable water pumps as well as by the banana changed and people lost their (only) lessons from the Health Centre on food source of income. Then there would be a lack preparation and water storage, the distribution of food crops grown, people would not have of mosquito-nets, a family planning model, the necessary money to buy food from the antenatal examination, free vaccinations and market where prices might rise even more. education on topics like hygiene in kitchens, A shift to producing again more and maybe FAO breastfeeding recommendations, toilets, different kind of food crops could be helpful, hand washing and cleaning practices. although farmers would have to accept a

221

smaller income. But as the experiences of implications for irrigation development. some of the interviewees showed, people who In: Physics and Chemistry of the Earth, Nr. plant food crops and cash crops to the same 28; 2003¸ pp. 817-825. extent are not worse off at all. MENGISTU A. Ethopia: Country Moreover, as already recognized by the Pasture/Forage Resource Profiles. Suttie farmers and experts, mono cropping has and Reynolds, July 2003. negative impacts on the agricultural area by Web: reducing soil fertility. But first, farmers have to http://www.fao.org/ag/agp/agpc/doc/counp be aware of the problem so that they give more rof/ethiopia/ethiopia.htm. [7.7.2006]. effort and place bigger emphasis on strategies ROBINSON W I, LIEBERG A T, TREBBI L. on keeping soil fertility. FAO/WFP Crop and Food Supply Assessment Mission to Ethiopia. 24th The ongoing strategy of family planning to February 2006; Web: slow down the rapid population growth of the http://www.who.int/hac/crises/eth/sitreps/E past years might show its success in the thiopia_fao_Feb06.pdf. [7.7.2006]. coming years and hopefully it will change the Rome Declaration on World Food Security and situation of landholdings in a positive way. World Food Summit Plan of Action. Also the lessons people got in sanitation and Rome, 1996. Web: health care practices should show a positive http://www.fao.org/docrep/003/w3613e/w impact in the livelihood of people. 3613e00.htm. [24.6.2006]. SCHULZ H. Moringa Oleifera - das References: vergessene Kind der Ayurveda. Web: http://www.ayurvedagesundheit.de/moring ADEM F. Small-Scale Irrigation and a_oleifera_das_vergessene_kind_der_ayur Household Food Security. A case study veda.html. [23.01.2006]. from Central Ethiopia. FSS Discussion Sustainable Livelihoods Guidance Sheets: Paper No. 4; Forum for Social Studies, Section 1. DFID, April 1999; Web: Addis Ababa, February 2001. http://www.livelihoods.org/info/guidance_ BANTERO B. Across systems comparative sheets_pdfs/section1.pdf. [25.6.2006]. assessment of Hare community managed The Evaluation. Ethiopia, Special Country irrigation schemes performance. Arba Programme Phase II, Report # 1643-ET, Minch University, October 2006. #April 2005, Office of Evaluation, IFAD, KEBEDE H. Country Presentation on an Rome; Web: Overview of Food Insecurity and FIVIMS- http://www.ifad.org/evaluation/public_htm related Activities in Ethiopia. IAWG – l/eksyst/doc/country/pf/ethiopia.pdf. Annual Meeting from Oct. 1-3, 2003. [9.7.2006]. KENNEDY E, BOUIS H, VON BRAUN J. THOMPSON J. Combining local knowledge Health and Nutrition Effects of Cash Crop and expert assistance in natural resource Production in Developing Countries. A management: Small-scale irrigation in Comparative Analysis Reprinted from Kenya. Acts Press, Nairobi, 1991. Social Science and Medicine, Vol. 35, World Bank. Country Brief. August 2007. No.5, IFPRI, 1992. http://go.worldbank.org/WA1RL12OL0, KLOOS H. Small Scale Irrigation and Food [8.10.2007] Production in Ethiopia. A review. In: Ethiopian Journal of Development Research Vol.12, No.2. Institute of Development Research, Addis Ababa University, April 1990. LALL A C, BROADWAY A C. An appraisal of the integrated rural development program in Allahabad. The Indian Journal of Economics LXXV, 1994, no.296, pp. 127-135. LANKFORD B. Irrigation-based livelihood trends in river basins: theory and policy

222

Impact Assessment of Rainwater Harvesting Ponds: The Case of Alaba Woreda, Ethiopia

Rebeka Amha ILRI, Addis Ababa, Ethiopia. [email protected]

Abstract statistically significantly. Results of analysis of qualitative information, consistent, with Rainfall in the arid and semi-arid areas is the Probit model results, also showed that generally insufficient to meet the basic labor requirement, economic problem to use needs of crop production. Thus, there is now simpler water lifting and watering increasing interest to the low cost alternative equipments, inability to easily understand generally referred to as ‘water harvesting’ the benefit of the technology and problems especially for small scale farming systems. related with the structure of the RWH In Alaba, even if government efforts of technology adopted were some of the major household level water harvesting schemes problems faced by households, and have a are wide spread, the performance obtained negative impact on the technology adoption was not assessed. Due to this reason, there rate. was a need to asses the impact of the existing rainwater harvesting systems in The Ordinary Least Square estimation of the Alaba Woreda. determinants of the value of crop production shows that adoption of RWH has a positive The study assesses the determinants of and statistically significant effect on value of households’ adoption of rainwater crop production, after controlling for input harvesting ponds, and its impact on use and other factors. This shows that RWH agricultural intensification and yield in ponds have direct and significant impact on Alaba Woreda, southern Ethiopia. Results value of crop production. We also find that are based on data collected from a survey of households with RWH technology use more 152 households and 1036 plots operated by labor and seed but less oxen power the households. Households were stratified compared with those households who have into those with rain water harvesting ponds not adopted the technology. Moreover, labor and those without from which equal number and seed inputs have positively significant of sample households ware drawn. Analysis impact on yield while the effect of oxen of descriptive information (mainly focusing power is insignificant. These results show on cropping pattern) and econometric that in addition to its direct impact, RWH methods are used. Analysis of qualitative has significant indirect impact on value of information supplemented the econometric crop production through its effect on results. intensity of input use.

The finding in the cropping pattern shows Labor requirements and cost considerations that, farm households have started to grow appear to be important factors that influence new crops (vegetables and perennial crops) household’s adoption of RWH technology. as a result of water availability from the This implies that research and development water harvesting ponds. Results of Probit interventions need to take account of the analysis on the determinants of adoption of labor and cost demands of the technology. rainwater harvesting ponds shows that The effectiveness of the technology household size, education status of adoption is mainly constrained by problems household head, ownership of livestock related to water lifting and watering (cattle, oxen and pack animals), homestead equipments, and accidents occurring due to plots and type of pond explained adoption 223

absence of roof cover and fence to the than 90% of the total agricultural output and ponds. This implies that support will be cultivate close to 95% of the total cropped needed to provide affordable but improved land dominate the sector. Agricultural water lifting and watering equipments, and production is highly dependent on the give training to farm households on vagaries of nature with significant construction and use of roof covers and variability in production and actual fences to the ponds. As households shift to production patterns (Demeke et al, 2005). high value but perishable commodities due to the RWH, emphasis needs to be given to Due to population increase in the highland marketing extension, especially in areas, more and more marginal areas are facilitating markets and market linkages to being used for agriculture which led to the farmers. degradation of the natural resources .One of the major challenges to rural development in Future intervention to promote RWH the country is how to promote food technologies need to provide due attention to production to meet the ever-increasing quality, rather than focusing on the number demand of the growing population. Rainfall of adopters. Households appear to neglect in the arid and semi-arid areas is generally the community ponds since they focus on insufficient to meet the basic needs of crop using cleaner water obtained from production. In degraded areas with poor household ponds and other sources of clean vegetation cover and infertile soil, rainfall is water. In this process the community ponds lost almost completely through direct are becoming a cause of health problems. evaporation or uncontrolled runoff. Thus, Thus, it is important that appropriate overcoming the limitations of these arid and attention be given to the community ponds semi-arid areas and making good use of the as well. vast agricultural potential under the Ethiopian context, is a necessity rather than Finally, it was found out that women are a choice. Thus, there is need for appropriate getting benefit from the technology interventions to address the prevailing adoption as any member of the family. constraints using suitable technologies for Their participation in the technology improved and sustainable agricultural adoption is mainly in watching the production. ponds. They also have contribution in With regard to agricultural water planning and decision making stage, and development, small scale irrigation seems to in giving support during construction, be preferred to large scale schemes. The maintenance and clearance of the pond. reason for the preference of small-scale Female headed households are being irrigation to large scale irrigation includes constrained to be beneficiaries due to the high capital requirement and cost of economic and manpower shortage. constructing large scale scheme which can only benefit a fortunate few but easy 1. Background adaptability of small scale irrigation (Turner, 1994). Ethiopia, like other Sub-Saharan African (SSA) countries, is an agrarian economy, There is now increasing interest to the low with a very small industrial sector. The cost alternative generally referred to as agricultural sector, on average, accounts for ‘water harvesting’ especially for small scale about 45% of the GDP, 90% of merchandise farming systems. Runoff, instead of being export earnings, 80% of employment, more considered as a problem, can be harvested than 90% of the total foreign exchange and used for different purposes, which earnings, 70% of the raw material supplies otherwise is lost and causes soil erosion. for agro-industries, and is also a major Various methods of rainwater harvesting are supplier of food stuff for consumers in the available, through which rainwater is country. Smallholders who produce more captured, stored and used at times of water 224

scarcity. Rainwater harvesting can be some of the major reasons for the limited broadly defined as a collection and coverage (Ngigi, 2003). concentration of runoff for productive After the fall of the military government, purposes like crop, fodder, pasture or trees both the Transitional Government of production, livestock and domestic water Ethiopia (TGE), established in 1991, and the supply (Ngigi, 2003). Federal Democratic Republic of Ethiopia (FDRE), established in 1995, have adopted Collection and storage of rainwater for an economic development policy to achieve different purposes has been a common food self sufficiency and sustainable practice since ancient times. The system was development, based on a strategy called used thousand years ago in many parts of the Agricultural Development-led world. There are also evidences indicating Industrialization (ADLI) , which gives more ancient churches, monasteries and castles in emphasis to improvement in agricultural Ethiopia used to collect rainwater from productivity. Besides, recognizing the rooftops and ground catchments. Birkas in problem of variability in the rainfall Somalia region and different runoff basins in distribution in the country, the 1995 strategy Konso are good examples of the traditional advocates for water centered sustainable rainwater harvesting practices in Ethiopia. rural development (Desta, 2004). Based on Moreover embankment and excavated this, several rain water harvesting ponds20 for agriculture use and water technologies have been constructed by supply, runoff farming and various types of regional states, NGOs, communities, and soil moisture conservation techniques for individual farmers through out the country. crop production could be mentioned as examples (Nega, 2004). To mitigate the erratic nature of rain fall in the arid and semi-arid parts of the country, In Ethiopia, promotion and application of which threatens the lives of millions of rainwater harvesting techniques as people, a national food security strategy alternative interventions to address water based on the development and scarcity were started through government implementation of rainwater harvesting initiated soil and water conservation technologies either at a village or household programmes. It was started as a response to level was adopted after 1991. The Federal the 1971-1974 drought in Tigray, Wollo and Government had allocated a budget for food Hararge regions with the introduction of security programs in the regions, an amount food-for-work (FFW) programme which equal to ETB 100 million and ETB one were intended to generate employment billion during the 2002 and 2003 fiscal opportunities to the people affected by the years, respectively. Of the total budget, most drought. Since then, however, the of it was used by regional states for the interventions have been extended to the construction of rainwater harvesting other parts of the country with very limited technologies including household ponds, in coverage. The low level of community collaboration with the Federal Ministry of participation and declining attention were Agriculture and Rural Development (Rami, 2003).

20 Even if government efforts of household According to (Nega, 2005) they are defined as follows. level water harvesting schemes are wide Pond: is small tank or reservoir and is spread in Alaba, the performance obtained constructed for the purpose of storing the surface was not assessed. Due to this reason, there runoff was a need to asses the impact of the Excavated pond: is a pond type constructed by existing rainwater harvesting systems in digging the soil from the ground Alaba Woreda to determine their Embankment pond: type of pond constructed effectiveness and sustainability. In addition, across stream or water course consisting of an there was a need to assess the condition of earthen dam. indigenous rainwater harvesting

225

technologies and practices in Alaba. Hence, quantitative and qualitative approach in this study is aimed to fill this gap of assessing the impact of RWH technology knowledge in the region. interventions on agricultural production. Finally, understanding the impact of the The purpose of impact assessment is to RWH technologies on agricultural determine the welfare changes from a given productivity and the determinant factors of intervention on individual, households and rainwater harvesting ponds, which affect institutions and whether those changes are productivity or level of yield, is a vital issue attributable to the project, programme, or for designing appropriate agricultural policy intervention. Impact assessments are development policies and strategies, as well often undertaken ex ante, evaluating the as technology interventions. Therefore, the impact of current and future interventions, outcome of this study may serve as a source or ex post, evaluating the impact of past of additional information which may be of intervention. It can also be made significant use to policy makers and concurrently within the project cycle planners during the designing and (Shiferaw et.al, 2005). Our focus in this implementation of RWH technology study is the ex post impact assessment. Ex strategies. post assessment attempts to understand the pathway through which observed impacts The study was conducted amid some have occurred and why interventions fail or limitations. One of the limitations is the succeed in attaining stated objectives. unavailability of base line data. Such data Hence, ex post assessments can inform would reflect the condition of the farm policy choices as to whether related planned household’s agricultural production process programme interventions should be pre-technology intervention, and would have discontinued, modified, improved or been helpful to compare more sustained in the future (Ibde). comprehensively and evaluate the relative effect of the technology intervention on Hence, this study is aimed at assessing the agricultural productivity overtime. The other impact of rainwater harvesting ponds on limitation of this study is related to the lack crop yield using a quantitative approach of accurate measures and valuation supplemented by a qualitative approach in techniques to include the environmental Alaba. In particular the study focuses on: benefits and costs that accrue from the RWH • Identifying the determinants of technology intervention. household decision to adopt rainwater harvesting ponds. 2. Literature Review • Examining the impact of rainwater harvesting ponds on crop yield, input Agriculture is the most water-demanding use and cropping pattern. sector, in addition to being a major source of • Assess the constraints and options to employment and a major contributor of the improve rainwater harvesting ponds national gross domestic product (GDP) of • Assess the differential impact of the many developing countries in Africa. technology by gender Agriculture in Ethiopia provides 86 percent • Derive policy implications to improve of the country’s employment and 57 percent the performance of the rainwater of its GDP. Rain fed crop cultivation is the harvesting ponds. principal activity and is practiced over an area of 27.9 million hectares (ha) of land The study is expected to identify problems (Gebeyehu, 2006). encountered, so that possible measures are taken when these interventions are Some empirical studies suggest that replicated in other parts of the Woreda or the irrigation has shown some positive impacts country. Besides, being an empirical study it in increasing agricultural productivity and will help to add to the empirical literature thereby increase the income of farm that uses the combination of both households, who participate in the irrigation

226

schemes (FAO, 1993). In the context of have to depend on rainfall that is highly farm households living in the Sub-Saharan variable both in distribution and amounts. African countries, irrigation has, however, By using underground spherical tanks proved costly and can only benefit farm having a combined capacity of 60 m3, households with large plots in addition to seasonal water for supplemental irrigation concerns related with the environmental and for an area about 400 m2 was guaranteed. health side effects of the schemes. With rainwater harvesting, farmers have diversified to include horticultural cash Large-scale dam and irrigation projects have crops and the keeping of dairy animals. For not been widely implemented in Ethiopia as instance households with supplemental they have often proved to be too expensive irrigation earn US$735(per ha) from cash and demanding in construction and crop compared with US$146 normally maintenance. Therefore, water harvesting earned from rain fed maize. This has tanks and ponds at the village or household contributed to food security; better nutrition level are proposed as a practical and and higher family income (RELMA-ineffective alternative to improve the lives of ICRAF, 2004). rural people at little cost and with minimal India has a long tradition of rainwater outside inputs. In theory, household water harvesting so much so that it is regarded as harvesting can be done mainly through the one of the dying tradition of the country21. effort of the individual farmer. Use of stored However, it has been reviving apace in rainwater could supplement natural rainfall many parts of the country, particularly in and make farming families less vulnerable to rain scarce areas. Derwadi village, a village drought and therefore less dependent on in the central state of Maharashtra, is one of outside help in harder times (Takele, 2002) such dry villages of India. A remote village with no assurance to drinking water, with The experience in China on the development farming being mainly rain fed based and of rainwater harvesting shows that since the agricultural production can’t meet more than 1980’s , Gansu, Sichuan, Guangxi, Guizhou three-month food of the village, Derwadi and Yunnan provinces adopted rainwater used to be a desperate village with no harvesting techniques. To date, rainwater employment opportunity for the community harvesting projects have been carried out in and where schooling is a distant dream for about 700 counties of 15 provinces in semi- the kids of the community. The villagers arid and humid areas covering two million established a link with an Indo-German km 2 and with a total population of 0.36 watershed Development NGO called billion. By the end of 2001, about 12 million Watershed Organization Trust (WOTR), water cellars, tanks and small ponds were which later assisted them to construct built with a total storage capacity of 16 contour trenches, farm and contour bunds, billion m3, supplying water for domestic use and check dams. A degraded land then for 36 million people and supplemental stared to provide adequate water both for irrigation for 2.6 million m2 of dry farming drinking and for irrigation, thus paving the land. This has helped the people access way for transformation of the lives of the water and engages in agricultural production villagers. They not only managed to hence improving food security and diversify from traditional pearl millet to alleviating poverty. Rainwater harvesting other host of crops ranging from various has also been known to benefit ecological vegetables to cotton, but also managed to and environmental conservation (UNEP, produce the crops in surplus and be able to 2005).

Impact of rainwater harvesting as shown in a 21 This document on India’s experience is case study of Mwala division, Kenya obtained from website indicates that harvesting runoff water for www.rainwaterharvesting.org/rural, where an supplemental irrigation is a risk-averting interesting account of experience with rainwater strategy, pre-empting situations where crops harvesting in more than 20 Indian villages is presented. 227

sell, perhaps for the first time, to big towns. conservation agriculture technologies which They managed to send their kids to school. included the use of rainwater harvesting. With the help of the NGO they also This was estimated to involve at least 50,000 managed to form self help association that hectares. The experience of Zambia shows enabled them to organize and carry out such that crop yields have on the minimum activities as construction of toilet, kitchen doubled. Maize yield rose from under garden and improved cocking devices. 0.5t/ha to above 2t/ha and cotton from 1.5t/ha to 3t/ha under conventional as The other experience with rainwater compared to conservation agriculture harvesting from India is Gandhigram village respectively. This has been attributed to of Gujarati state. This village is also one of improved rainwater harvesting made the water scarce areas of the country, possible by the planting stations and surface constantly suffering from acute water cover. Most farmers have diversified their scarcity both for consumption and cropping system to include crops such as production. Assisted by a local NGO called maize, beans and sunflower. Increased Shri Vivekanand Research and Training production at the household level in the last Institute, the community started to build five years has introduced the rapid re-birth communal dams- small and big- in 1995 so of a cash economy among the communities. as to store rainwater and use it during dry This has propelled private entrepreneurship season. A committee was formed from in agricultural related trading. Large and among the beneficiaries to oversee the small private entrepreneurs have emerged distribution of the water and maintenance of and are selling agricultural inputs and other the dams. They evolved an interesting household commodities as well as buying management mechanism where each off the crop. Most households are able to put household is asked to pay Rs 3 (equivalent up for sale 20-30% of their produce. The of $0.067) per month for water supply for ultimate effect is enhanced livelihoods consumption purpose, and Rs (UNEP, 2005). 250(equivalent to $5.56) per ha for irrigation purpose. The community managed not only Hatibu et al (2004) tried to quantify the to secure sustained supplies of water for effect on farmers’ income and living domestic consumption, but also was able to standards of different rainwater harvesting embark upon producing high value crops methods, taking two districts, Maswa from like ground nuts, wheat, onion and cumin. north and Same districts from Eastern parts, They managed to increase their agricultural of Tanzania. All types, viz. in-situ, micro yield and work availability has also and macro catchments and rainwater increased for land less laborers. As it has harvesting with storage are all practiced in become beneficial, the momentum for the two regions in descending order of rainwater harvesting continued in the village prevalence; in-situ is more prevalent in both as is evident from community’s interest to regions followed by micro and macro increase the number of dams by constructing catchments, with rainwater harvesting with new ones. Interestingly enough, they are storage being the least. The harvested now on the stage of forming a cooperative rainwater is used mainly to grow maize in for processing and marketing their Same area while it is used for rice in Maswa agricultural products. region. Good rainwater harvesting increases yield of maize (in Same area) by four fold of By the 1990’s, Zambia’s southern province rain fed yield level, and two fold for rice (in was recording unprecedented levels of food Maswa area)(Ibid). insecurity, hunger and general poverty. Government food, seed and fertilizer relief It is only recently that rainwater harvesting support become the norm rather than the has started to receive significant attention exception for many households. During the from Ethiopian government though it has a 2002/2003 season, over 12% of the farm long history. It has been regarded as one of households were estimated to have adopted the crucial tools to achieve food self-

228

sufficiency, and is being implemented on a Besides, rainwater harvesting is found to large scale particularly in water scarce areas have undesirable, but not unexpected, health of the country. As the phenomenon is quite side effects. For instance many people and recent, detailed study hasn’t been made. livestock have been drowned into the tanks However, some preliminary studies have and ponds, with often no fences and live been made on some parts of the country. saving mechanisms like ladder and ropes Rami (2003) is one of such studies, and is (Ibid). It is also cited by people living near basically an account of two weeks field visit the ponds as a source of malaria out break. in Amhara and Tigray regions. The However, it doesn’t mean that rainwater emphasis is mainly on rainwater harvesting harvesting didn’t have any positive effects implementation related problems in the on the community. It has enabled them to regions and the prospects of using it for the grow crops of short growing periods like stated objective of attaining food self- vegetables. And some have had good sufficiency. It has been found that RWH is experience, as is the case in Tigray region top of the agenda in the two regions, as is where, for instance, “a farmer and his wife the case at national level, with some times were able within a single season to pay their over ambitious plans of constructing wells old extension credit of more than 1000 Birr and ponds. through the planting and sale of vegetables (cabbages, tomatoes, beans and peppers) The success in attaining the planed amounts (Ibid). The upshot is that rainwater of tanks and ponds to be constructed and the harvesting is beset with challenges and can perceptions of the beneficiaries are found be an utter failure and end up in undesirable mixed. Shortages of required construction negative consequences if not cautiously raw materials, lack of timely dispersal of approached. However, it can play immense finance and shortage of skilled labor have role in helping attain food security if been among the factors inhibiting the implemented with thorough consultations attainments of the stated goals. This is with the beneficiaries and is accompanied evident from Amhara region where it once with other activities like afforestation and was planned to construct 29005 tanks made soil conservation and fertility enhancing of cement and plastic and 27955 wells were practices. excavated for the purpose but only 12614 tanks were constructed. Furthermore, the The econometric approach has some tanks constructed so far are found to be limitations in accurately and fully measuring substandard, many collapsed and majority the changes resulting from NRM leak and seep water, the main factor being interventions, especially those changes lack of experienced masons and supervisors which are non-quantifiable. Hence, as a and mismatch between the type of soil in the remedy to the shortcomings of the area and the tank construction method. The econometric approach, at present tanks were first tested in Adama area and ,researchers like Kerr et.al (2005) are implemented in the two regions, with advocating that better results could be basically different soil structures from obtained using an integrated quantitative and Adama area, without-taking into account the qualitative approach in assessing the impact specificities of the two regions (Rami, of NRM interventions. 2003). In addition, most of the construction was assigned to each Woreda as a quota Kerr et.al (2005) employed quantitative resulting in less attention being paid to analysis (as with and without design mainly quality as compared to number. Further, the employing instrumental variable approach) implementation tended to be top-down and also qualitative information to better approach, particularly in Amhara region, understand interest in relation to relevant and this has also contributed its share to the research questions, and to identify the problems (Ibid). projects’ unintended consequences in evaluating the performance of watershed projects in India. Specifically, the study tries

229

to identify: the successful projects, the the impact of land management and approaches adopted which lead to the investment on the value of crop production success and additional characteristics of in Uganda .The data for the analysis particular villages’ contribution to achieve obtained from a survey of 451 households. improved natural resource management, Selected regressors include several variables higher agricultural productivity, and reduced at the village, household and plot levels. The poverty. The results of the study show that study has shown that improvement in land in both of the states, participatory projects management can lead to higher productivity combined with sound technical inputs and lower land degradation. Participation in performed better as compared to technical assistance programs, pursuit of technocratic, top-down counterpart. certain livelihood strategies, investment in Evidence also found on the existence of irrigation, and promotion of more potential poverty alleviation trade-off during specialized production of cereals or export an effort to increase agricultural productivity crops are found to achieve “Win-Win” and conserve natural resources through outcomes, increasing agricultural watershed development. Particularly, the productivity while reducing land empirical result indicates the existence of degradation. The results of the study don’t strong evidence on the skewed distribution support the optimistic ‘more people-less of benefits towards largest land holders in erosion’ hypothesis, though the results are projects, which are more successful in both consistent with population induced conservation and productivity .The short- agricultural intensification’, as hypothesized term costs imposed on ‘losers’ (i.e. the poor) by Boserup. In addition it indicates the need may be substantial and projects would gain to make further research to identify from a greater focus on mechanisms to share profitable as well as sustainable land projects benefits (Shiferaw et.al, 2003). management options, as no land management practices except irrigation were Apart from the qualitative analysis approach found to be very profitable in the short-run used in the early periods, the literature on (Shiferaw et.al, 2003). quantitative analysis approaches for assessing the impact of natural resource Gebremedhin et al.(2002, 2000), have management policy or technology applied an econometric analysis to examine interventions can include the econometric the nature and impact of community approach (Shiferaw et.al, 2003).The woodlot and grazing land management’s commonly applied method in natural respectively; and identify the determinant resource management intervention impact factors of collective action and its assessment, i.e., the econometric approach, effectiveness, in Tigray, Ethiopia. Empirical is developed by linking the measures of results of the analysis indicated that, more current output, cost or profits directly to past collective action exists manage community research investments. In this approach, woodlots in areas with intermediate either a primal function, based on estimated population density. In relation to community production function, or a dual function, grazing land management, results from the using a profit or cost function and their regression analysis depict that, while related system of supply and factor demand population pressure has resulted in reduction functions are employed. In general, once the of violations of use restrictions of grazing econometric approach is adopted, the impact land in areas with low and intermediate level of the natural resource management of population density, intermediate technology or policy intervention is obtained population pressure has the tendency to by translating the parameter estimates of the reduce the development of use restrictions function used, into economic benefit value and the enforcement of penalties (Shiferaw et.al, 2003). (Gebremedhin et.al, 2000). Besides, while negative relationship has been observed For instance, Pender et al. (200l) employed a between communities access to market and structural econometric approach, to explore household’s contribution to collective

230

action, tree planting, and the survival rate of DATA ANALYSIS trees (Gebremedhin et.al, 2002). However, the result from both studies reveal that, the Qualitative approaches are increasingly used presence of external organizations is in conjunction with quantitative approaches negatively associated with the probability of and such combinations can enhance the community payment to guard, survival rate validity and reliability of impact evaluations. of trees, and collective action for grazing While quantitative approaches allow land management Gebremedhin et. al (2002, statistical tests for causality and isolation of 2000). programme effects from other confounding influences, quantitative methods excel at 3. Methods of the study answering impact assessment questions about ‘what’ and ‘how much’, whereas Sampling and data qualitative methods are preferred for exploring questions of ‘how’ and ‘why’. A The data for the analysis is obtained from a mix of quantitative and qualitative household and plot level survey in Alaba approaches is ideal because it provides the Woreda. The Woreda is located 310 km quantifiable impacts of the intervention as south of Addis Ababa and about 85km well as an explanation of the processes and southwest of the Southern Nations relationships that yielded such outcomes Nationalities and Peoples Regional (Shiferaw et.al, 2005). (SNNPR) state capital of Awasa. A semi- structured questionnaire has been employed Descriptive Analysis to interview household heads. This part mainly focuses on describing the A total of 152 households which are selected impact of rainwater harvesting ponds on the using a stratified sampling technique have cropping pattern. Cropping pattern of the been surveyed. Based on farming system farm household’s has been assessed based practiced, the 73 peasant associations in the on the farming system. Woreda are stratified in to two, namely 43 peasant associations with Teff/ Haricot Bean Livestock and 30 peasant associations with Econometrics approach Pepper/ Livestock farming system. From each stratum 2 peasant associations were Empirical model and econometric selected randomly and the households estimation within each of the four peasant associations were further stratified by adoption of RWH Since there is no predetermined model that technology. In the end, from each of the four can be used in the quantitative estimation, randomly selected peasant associations, a following Pender and Gebremedhin (2004), total of 38 households were randomly models for the use of inputs on each plot selected, where 19 of the farm households (from equation 2 up to equation 6); adoption adopting the technology and 19 farm of RWH ponds (equation 1); and the value households without the technology stratum. of crop production on each plot in 2005/06

(from equation 7 to equation 9) are adopted Moreover, interview has been done with in this study. experts working in the OoARD (office of Agricultural and Rural Development). Secondary data was also used from To identify the determinant factors that publications, books, articles etc. to influence the farm households’ decision to supplement the data. adopt RWH pond or to invest on various types of RWH ponds, a probit model is estimated. Hence, a RWHp dummy variable (where 1=household with RWH technology and 0=household without RWH technology)

231

is modeled as a function of village-level output and input demands, and reduces factors (XV), plot-level factors (Xp), problems due to outliers and non-normality household-level factors (Xh) and pond type of the error term found when using a linear which can be plastic covered or concert specification (Pender and Gebremedhin, basement (P).These can be written as 2004). follows: Thus, the use of inputs – Labor days/ha RWHp = f (XV, XP, Xh , P) ……. (1) (lnL), Oxen power days/ha (lnO), Seeds kg/ha (lnS), use of Fertilizer (F), and use of Where, Household- level factor (Xh) Manure/Compost (M/C), are modeled as a includes: function of explanatory variables including • Human capital (demographic village-level factors (XV), plot-level factors features) - age, household size, (Xp), household-level factors (Xh ) and the educational status. predicted value of adoption of rainwater • Physical capital - land holding, value harvesting ponds (RWHp).The models for of all assets owned, value of livestock the variable inputs can be written as follows: which includes oxen, packed animals, lnL = f (XV, XP, Xh, RWHp) ………. .(2) poultry, cattle etc. lnXK = f (XV, XP, Xh, RWHp ) …..…(3) • Social capital- membership in local lnS = f (XV, XP, Xh, RWHp) …….... (4) organization and associations. F = f (XV, XP, Xh, RWHp) ………. (5)

• Financial capital-households saving M/C= f (XV, XP, Xh, RWHp) …….. .(6) and credit access. Village-level factors (Xv) includes: Where, ln stands for logarithm • Indicators of agricultural potential: rainfall condition(here due to lack of The econometric model used depends on the adequate information at PA level, nature of the dependent variable. For use of during estimation, location dummies labor, oxen power and seeds on cultivated has been used in order to capture the plots, the least squares regression is used difference in rainfall, altitude, while the regression equations for the population density and other variable inputs, fertilizer and environmental factors for the four PAs manure/compost, Probit model is used since included in the study). the dependent variable is dummy variable.

• Household access to services and infrastructure: walking time from the Finally, in assessing the impact of RWH farm household’s residence to the ponds on agricultural output, the value of the nearest input/ output town market, agricultural output harvested from a plot is village market, Cooperative shops and modeled in three different alternatives. First, all-weather and seasonal road. a full model of the value of crop production from a plot is modeled as a function of Plot-level factors (Xp) - Natural capital village-level factors (XV), plot-level factors (X ) and household-level factors (X ). • Indicators of quality of the plot (size p h Besides, the use of variable inputs Labor of plot, slope of the plot, soil depth, (lnL), Oxen power (lnO), Seeds (lnS), soil type and soil fertility of the plot), Fertilizer (F), Manure or Compost (M/C) how the household acquired the plot, and the predicted value for adoption of the purpose for which the plot is used RWH ponds (RWHp) are included.A full and walking time from farm model of the value of crop production from household’s residence to the plot in a plot can be written as follows: hours.

LnY= f (lnL, lnO, lnS, F, M/C, X , X X In the crop production regression and input V P, h, RWHp) ……. (7) use regressions, a logarithmic Cobb-Douglas specification is used. This leads to a theoretically consistent specification for 232

However, in the second regression, These approach analysis the perception of household-level characteristics (Xh) and experts and farmers regarding the adoption of RWH pond (RWHp) are constraints and opportunities of RWH omitted. This is because the effect of these technologies. The qualitative information variables on production may be indirectly was gathered using an open-ended question through the use of inputs. Thus, the second - that was included in the questionnaire in structural model of the value of crop yield is order to augment the results of the modeled as a function of all factor inputs by econometrics analysis. excluding household-level factors (Xh) and adoption of RWH pond (RWHp) from the 4. Results and Discussions regression. Thus the second model of the value of crop yield from a plot is given as Impact on Cropping Pattern follows: LnY= f (lnL, lnO, lnS, F, M/C, XV, XP) As part of the assessment for the impact of ………………………………………… (8) RWH technology intervention on the farm household’s crop choice decision, the study The third model developed in this study for has employed a descriptive analysis of the the value of crop production is a reduced- crop mix for those with RWH technology in form equation, which includes all village- the different farming systems. Here, the crop level, plot-level, household-level types are classified into categories such as characteristics as explanatory variables and annual crops, perennial crops, vegetables, the predicted value for adoption of RWH spices, others and no new crops. As can be ponds. However, it excludes the use of seen from the table below, of the total inputs like Labor (lnL), Oxen power (lnO), number of the crop types sawn by all the Seeds (lnS), Fertilizer (F) and Manure or sample households (382 plots), 188 Compost (M/C) from the model. This observations are in the teff/haricot specification can avoid the potential for bean/livestock farming system category and endogenity bias. And also to examine the 194 observations are under the total effect of all factors on crop production, pepper/livestock farming system category. and whether it is a direct effect on production or indirectly through its effect on In the teff /haricot bean/livestock farming the use of inputs and adoption of RWH system, of the total 188 observations, 60.1% ponds. grow vegetables where as 4.3%, 6.9%, 4.3% represent annuals crops, perennial crops and The models for reduced- form specification spices, respectively. In the vegetable crop of the value of crop production from a plot category cabbage, onions and carrot account can be written as follows: 16.5%, 14.9% and 12.2%, respectively. On LnY = f (Xv, Xp, Xh, RWHp) ……….. (9) the other hand, in the pepper/ livestock farming system, of the total 194 In all cases, the least square regression was observations 67% is vegetables category used to estimate the value of crop where as 6.2%, 4.1%, 2.1% represent annual production.Generally, one important point crops, perennial crops and spices. In the that should be noted is that, for equation vegetable category which have great share 2,3,4,7 and 8 robust regression is undertaken from the different classifications cabbage, to avoid the hetroskedasticity problem that beet root, tomato, carrot and onion, account was observed during estimation. And also for 16.5, 12.9, 10.3, 9.8 and 8.8 percent, problem of multicolinearity and omission of respectively. variables has been checked. The result of the crop mix analysis imply Qualitative Analysis that, the shift in farm household’s crop choice decision towards highly priced and marketable agricultural products like vegetables and perennial crops or increment

233

in the number of harvesting per consumers immediately after harvested, year(intensification), could have a positive either their market value will decrease with impact on the farm households income as time or it might be a loss to the farm well as level of living. However, the level household. Besides, an examination of the and magnitude of benefit accrue to the farm type of crops grown under the vegetable household will significantly depend on category witnessed that most farm market and infrastructure accessibility. This households have concentrated on specific is because most of the crop categories sawn crops (tomato, cabbage, onions, and carrot) in farm households with rainwater and the production and supply of these crops harvesting technology are perishable; for in large quantities might reduce the price of example, vegetable represent the highest the commodities and there by affect the percentage of (60.1%) in Teff/Haricot bean/ economic feasibility of the technology. livestock farming system and (67%) in Thus, effort should be made to supply pepper/ livestock farming system. Hence, variety seeds to farmers so as to diversify unless these products are able to reach to the type of crops grown.

Table 1: Types of crop grown after start to use the technology based on farming system Category of crop types grown Total

Farming Type of Nothi Annu Peren Vegeta Spice Oth system crops grown ng als nial bles s ers new crops crops No new crop 40 40

Te ff/ Ha ric grown (21.3) Chat 1 (.5) 1 Coffee 12 12 (6.4) Banana 1(.5) 1 Sugarcane 1 (.5) 1 Avocado 2 2 (1.1) Papaya 4 4 (2.1) Onions 28 28 (14.9) Ginger(Jinjibl 1 (.5) 1 e) Pepper 6 6 (3.2) Carrot 23 23 (12.2) Tomato 7 (3.7) 7 Cabbage 31 31 (16.5) Chilli Pepper 2 2 (1.1) Kale 4 (2.1) 4 Sweet 1 (.5) 1

234

potatoes Garlic 3 (1.6) 3 Beet root 15 (8) 15 If other 6 6 specify (3.2) Total 40 8 13 113 8 6 188 (21.3) (4.3) (6.9) (60.1) (4.3) (3.2) No new crop 38 38

Pe pp er/ liv grown (19.6) Chat 2 (1) 2 Coffee 8 (4.1) 8 Orange 1 (.5) 1 Banana 2 (1) 2 Pineapple 1 (.5) 1 Avocado 2 (1) 2 Mango 1 (.5) 1 Papaya 2 (1) 2 Onions 17 (8.8) 17 Pepper 4 4 (2.1) Carrot 19 (9.8) 19 Tomato 20 20 (10.3) Cabbage 32 32 (16.5) Lettuce/'Selat 5 (2.6) 5 a'/ Kale 6 (3.1) 6 'Kosta' 4 (2.1) 4 Sweet 1 (.5) 1 potatoes Garlic 1 (.5) 1 Mandarin 1 (.5) 1 Beet root 25 25 (12.9) If other 2 (1) 2 specify Total 38 12 8 (4.1) 130 4 2 (1) 194 (19.6) (6.2) (67) (2.1) *The number in the bracket shows percentage value *The number out of the bracket shows frequency

235

Determinants of adoption of RWH pond, invest on technology interventions than input use and crop yield those with few physical resource. The positive correlation with oxen power may be Determinants of Households Decision to due to households focus on agricultural Adopt RWH Pond production. However, it should be noted that the significant explanatory variables have The estimation results of the Probit model insignificant effect in magnitude implying for the determinants of household’s decision its less importance to make policy to adopt RWH technology is presented in implication.

Table 2. As can be shown in the table, from Plot level factors the locational dummies, Ulegeba Kukke shows stastical significance at 10% level. Among the plot level factors, household No association has been found between decision to adopt RWH pond is more likely village level factors and technology in homestead plot. The result indicates farm adoption decision. household’s effort to fully utilize family Household human capital labor so as to meet the human resource Household size is positively correlated with requirement during construction and the adoption decision of rainwater utilization of water, thereby reduce the harvesting ponds at 5% level of significance. finance that could otherwise be needed for This means households with large family hiring labor. It can also show the capital size are more likely to adopt the technology constraint faced by households to buy since they can compensate costs involved in modern water lifting equipment. The most hiring labor for any activity that the interesting implication of this result is that, technology demands. This implies that the accumulated water is used to produce research and development interventions crops with high market value rater than used need to take account of the labor and cost as supplementary source of water during dry demand of the technology. Households who spells, as initially intended by government can read and write, and those who are when the technology was introduced as educated up to grade seven are more likely country level. Ponds with concrete basement to adopt RWH. The positive association have shown stastically significant negative with the technology adoption can occur with correlation with adoption of rainwater the expectation that they can understand the harvesting pond at 1% level. This implies benefit more easily and are more open to that the higher cost involved in pond access information than illiterate construction will result in less technology households. This implies that expansion of adoption decision. education in the woreda will have a positive impact in increasing the adoption decision Determinants of Agricultural Input Use rate. The estimation result for the agricultural Household physical capital endowment inputs of: labor person days per hectare, From the household physical resource oxen power days per hectare, seed - kg/ha, endowment indicators included in the fertilizer and manure or compost is model, oxen, cattle and pack animals have presented in Table 3. depicted positive correlation with adoption decision of the technology. This indicates Impact on use of Oxen Power that adoption of the technology requires large resources, thus households with a The estimation regression analysis also better physical resource are more likely to indicates that, adoption of rainwater harvesting technology has a negative stastically significant association with use of

236

oxen power, more likely due to lower use of depend on activities that don’t use oxen oxen power and more human labor on power as their source of livelihood. Farm homestead plots where the technology is households with saving have depicted mostly adopted. significant negative association with oxen power use, more likely households with The locational dummies of Ulegeba Kukke, saving are engaged in livestock production, Andegna Hansha and Hamata are positively trading or use the money for health associated with value of oxen power used expenditure and for some other purposes. relative to Mudda Dinokosa. From the The amount of oxen power used has shown household access to services and significant positive association with flat and infrastructure indicator, only nearness to moderately sloped plots in comparison to village market is significantly correlated steep plots. The result might indicate with more use of oxen power. Probably the farmers risk aversion behavior due to crop correlation could be because of the failure which could be caused by high runoff possibility to get more seed and fertilizer problem. Plots with medium soil depth are enabling them to use more oxen power in less likely to use oxen power compared to order to increase their agricultural plots with deep soil depth. Homestead plots productivity. Moreover, it is shown that have stastically significant negative medium rainfall condition is positively correlation at 1% level. This means, it is less correlated with the use of oxen power than likely that households will use oxen power low rainfall condition. on homestead plots. However, the likely use In the household level factors, household of oxen power is shown to be significantly size, heads who can read and write, and higher in crop land plots. An interesting those who are educated up to fourth grade result is found in the relationship between are positively associated with the use of plot size and oxen power use, where larger oxen power at 1% level of significance. This plot size is significantly associated with implies those households having large lower oxen power use. family size and educated members are more likely to use oxen power to utilize labor Impact on use of Seed available in the family to produce more output. From the household physical As expected the estimation of the regression resource endowment indicators, owned land analysis indicates that, adoption of RWH has shown positive correlation with the use pond has stastically significant association of oxen power at 5% level of significance, with more likely use of seed. This could which implies that more oxen power will be probably imply the impact of the RWH used by heads who own more land. In technology on crop production is indirectly addition, ownership of goats and sheep, and through its effect on intensity of agricultural beehive are stastically significant at 10% inputs. level. The significance might imply The regression result depicts that no household’s involvement in sheep, goat or evidence has been found between locational honey trading to get extra income and use dummies and amount of seed used. From the more oxen power in order to increase village level indicators, closeness to town agricultural production especially in cases and village market is significantly associated when the household has large land size. with more use of seed, probably the In relation to household head’s membership household heads are less likely to be in various associations, the study showed engaged in non-farm labor employment and that relative to households with heads a hence, more emphasis be given to crop member in association, households with production. heads not a member in associations are With respect to household size, large family negatively correlated with oxen power use. size is significantly associated with more use This might imply, non-members may of seed, probably indicating that the

237

members in the household utilize labor by correlation with labor input could be working in agricultural activity which because of either inability of the economy to demands more seed. From the education absorb the excess labor force in extended status, households with heads who can read families or constrained by transaction cost in and write, and those with formal education the labor market and there by the family up to fourth grade have shown positive members are compelled to engage in crop association with use of seed relative to production at the existing plot. Stastically illiterate headed households. Households significant negative correlation exists endowed with large sized land are between the age of the household head and significantly associated with more use of use of labor input. That means older-headed seed. No significant correlation has been households are less likely to supply labor. observed between social and financial factors, and amount of seed used. The result Furthermore, in relation to the household in the correlation between plot level factors physical resource endowment, ownership of and intensity in use of seed, more likely use more oxen power is likely to utilize more of seed is shown on cropland and homestead labor input than in cattle and pack animal plots. ownership. This is probably due to complementarity. An important point that Impact on labour use should be noted is the insignificant impact of this variables when consider the As anticipated the estimation of the magnitude. In relation to household head’s regression analysis indicate that, membership in local organization, the study adoption of RWH technology has a witnessed that, members in Edir and other positive stastically significant related local organization are more likely to use labor input than those who are members association with use of higher labor, in Edir only. In addition, households with most likely due to the higher level of saving are less likely to use labor input, labor requirement during watering , probably suggesting household’s construction an other activities involved. involvement in activities other than agriculture. As can be seen from the result of the regression analysis, location dummy of The result also shows a mixed correlation Hamata PA is associated with more likely between plot level factors and labor input use of labor input at 5% level of use. For instance, labor input use is significance. From the correlation between significantly greater on plots with flat and household access to infrastructure and medium slope than plots with steep slope, service indicators and use of labor input, perhaps indicating farmers risk aversion closeness to village market, town market behavior and their emphasis on short term and seasonal roads are associated with benefit. Since steep sloped plots are more higher intensity in use of labor input. exposed to soil erosion problem. More over, Probably household heads are engaged in less of labor input is used on inherited and farming activity by utilizing more seed, plots with medium soil depth. Homestead oxen and fertilizer use. Areas with high plots have stastically significant negative rainfall depict statistically negative association at 1% level. However, more use association with labor input use, suggesting of labor input is observed on cropland plots. the need for more labor input in areas where An interesting result is found in the there is low rainfall. relationship between plot size and labor The result of the regression analysis shows input use, where larger plot size is that, a farm household with large family size significantly associated with lower labor has stastically significant association with input use. use of more labor. Probably the positive

238

Impact on use of Fertilizer household’s nearness to village market, town market and seasonal road is more As can be seen on table 3, the adoption of likely to use manure or compost inputs. RWH technology is shown to have Probably this is due to the use of more labor insignificant impact on use of fertilizer seed input when the household is closer to suggesting that its impact on crop this services. In areas where there is high production isn’t seen indirectly through its rainfall, more use of manure or compost is effect on fertilizer input. observed.

From the village level factors, walking time Further more, from the household level to the nearest village market has a negative factors, households with large family are correlation with fertilizer use at 10% level of more likely to use manure or compost, significance. That means households closer probably due to the availability of labor to to the village market are more likely to use carry manure or compost to the farm land. fertilizer. No evidence has been found on the With respect to educational status, existence of correlation between the likely household heads with formal education up to use of fertilizer and factors like human, fourth grade are less likely to use manure or social and financial capital part of the compost relative to illiterate heads. Most household level indicators. Further more, likely this could be affected either by strong positive correlation has been found educated headed households positive between value of beehives and the likely use correlation with more likely use of fertilizer of fertilizer, which is perhaps due to there by reducing the likely use of manure or households focus on beekeeping activity compost , or these households are enabling them to buy more fertilizer using constrained by labor required to carry the incremental income. manure or compost to the farm.

In relation to the association between plot In relation to household’s physical resource level factors and the likely use of fertilizer, endowment, ownership of large sized land is crop land plots are shown to have positive correlated with less likely use of manure or association with the use of fertilizer at 1% compost, probably due to its high demand level of significance. Less fertilizer use is for labor input to carry manure or compost observed on homestead plots due to more to wider farm lands. Ownership of large possibility to use manure or compost than number of oxen is correlated with more buy fertilizer. In small plot size it is more likely use of manure or compost. Those likely to use higher amount of fertilizer engaged in livestock production as shown by which is mainly due to an increase in ownership of large number of cattle and efficiency when household’s own small beehives are less likely to use manure or sized plots. Moreover, plots closer to the compost. residence of the farm household have depicted significant correlation with more With respect to the financial capital part, likely use of fertilizer. households who have access to credit are more likely to use manure or compost input. Impact on use of Manure or Compost Probably due to the possibility of using the credit to buy seed, oxen etc. which might As can be depicted from table 3, adoption of lead to demand more manure or compost .In RWH technology is found to have addition, those with saving are also more insignificant impact on manure or compost. likely to use manure or compost. Probably No evidence has been found on the due to their preference to spent it on other existence of correlation between the use of things than on fertilizer by replacing it with manure or compost and the locational manure or compost. dummies. From the locational dummies,

239

Finally, in relation to the association impact shows that, households with RWH between plot level factors and the likely use technology are significantly correlated with of manure or compost, the result witnessed higher use of labor and seed but lower use of that, state owned and inherited plots are oxen power than those without the positively correlated with more use of technology. Intensity in use of labor and manure or compost. On the other hand, on seed input has a positively significant impact flat and moderately steep plots, households on yield while oxen power has insignificant are more likely to use manure or compost impact on yield. than on those steep sloped plots, probably to avoid risk of crop failure. Medium soil depth As can be seen from the structural model for is more likely to use manure or compost. the value of crop yield, in the village level Plots that are highly fertile are more likely to factors, seasonal road have negative stastical use manure or compost than those infertile significance at 10%. With respect to the once because it will be risky for the impact of plot fertility on value of crop household to use the input on infertile plot yield, households are more likely to produce than fertile once. Households are less likely more output in moderately fertile plots than to use manure or compost on cropland plots infertile once. As can be observed from the but more likely to use it on homestead plots, table, cropland and homestead plots are probably due to its closeness to the more likely to produce more yield. Besides, residence of the farm household. the result indicates the positive impact of use of labor, fertilizer and seed on value of crop yield. In the reduced model of crop Impact on Crop Yield yield, depicted in column 4 of table 4, village level factors, plot level factors, Table - 4 presents the full model of the value household level factors and household of crop yield (column-2). Here, variables rainwater harvesting technology adoption such as household level factors; household – decision were included in the regression and human, social, physical, and financial capital assessed with respect to their impact on the endowment; and adoption decision of RWH value of crop yield. technology that were included in the unrestricted OLS regression have been The village level factors don’t explain found to be jointly statistically insignificant. variation in the value of crop production. In column – 3 and column– 4 results of the Moreover, from the household level factors, structural and reduced models are shown household size has shown positive respectively. association with value of crop yield at 10% level of significance. This implies that The impact of adoption of RWH technology households having large family size are on crop production can be explained in two more likely to produce more output. With ways, directly or indirectly. The direct respect to the impact of household physical impact is, if the accumulated water is used capital endowment, greater ownership of to supplement the shortage of water during cattle has shown association with higher dry spell periods in rain fed crop production, value of crop yield (and stastically where as the indirect impact is through its significant at 10% level).From the plot level effect on intensity in use of agricultural factors included, state owned plot are more inputs. The estimation result of the study likely to produce more output than rented indicate that, adoption of RWH technology plots. Possibly indicating household’s high is shown to be positively correlated with future discount rate and become less likely value of yield at 1% level of significance. to invest on productivity enhancing This might imply that the direct impact of activities on rented plot. Plots with shallow the technology adoption on crop production and medium soil depth are less likely to is significant. An examination of the indirect produce more output than plots with deep

240

soil depth. It is also shown that, cropland non-farm activities. Credit access and saving and homestead plots are more likely to have a positive impact on manure or produce more output compared with compost input use. Household access to grazing, woodlots and spice plots. In services and infrastructure facilitates the addition, a negative significant association is movement of inputs to and outputs from observed between plot size and value of rural parts to towns, where large market is crop yield. available. The regression result shows an increase in yield when the household is As can be depicted from the result of the located closer to seasonal road and is reduced model, household family size is stastically significant. Households closer to positively correlated with value of yield at village market are able to use higher amount 10% level of significance implying that of seed, labor, oxen and more likely to use large family will produce more output. From fertilizer and manure or compost input. In the determinant factors of input use table, addition, households closer to cooperative households with large family size have shops and seasonal roads are more likely to shown significant association with use of use labor input and those nearer to town higher labor, seed, oxen and more likely use market are able to increase seed amount. of manure or compost. Intensity in use of labor has a positive impact on yield at 1% The result of the value of crop yield also level of significance. This suggests that shows that, state owned plots witnessed yield averages 11% higher per additional stastically significant association with higher labor a household uses. Moreover, average value of crop yield. Probably, suggesting yield increases by around 9% per additional that farmers are more likely to invest on seed amount used by the household. Even productivity enhancing activities on state though fertilizer isn’t significantly affected owned plots. It is also shown that shallow by household size, fertilizer is positively and medium soil depth has stastically correlated with value of yield at 1% level of significant association with lower yield than significance. That means yield is more likely on deep soil depth. Finally, crop land and to increase with more use of fertilizer input. homestead plots are shown to have positive Household age and education have association with value of yield. insignificant impact on value of yield. However, household age has a significant Perceptions of the constraints and impact on labor. Old age is negatively opportunities in adoption and use of associated with labor input use. Educational RWH technologies status has a positive impact on seed and oxen input use. Farmers were asked to rank the purpose for which the accumulated water was used Variations in resource endowment among based on the amount of water utilized in households will obviously have an impact each activity. As can be seen in table 5 on the level of crop yield either directly or below, households use the pond water for indirectly through their effect on the different purposes including as source of household’s demand for agricultural inputs. drinking water for animals and households. Of the factors, which are used to measure In addition to using the water for washing household physical capital endowment, cloths and cooking, households use the ownership of cattle has a positive impact on water for nursering some plants, for the value of crop yield. However, it has vegetable and fruit production. About 40.8% insignificant impact when consider the of households responded that they use the magnitude to make policy implication. water for vegetable production as a Households with saving are negatively supplementary during dry spell periods to be associated with labor and oxen inputs use. their first choice. In the second rank, 27.6% Probably they might prefer to be involved in of the households use the water for

241

nursering. About 23.7% and 18.4% of the livestock respectively. households use it for drinking and for

Table 5. The purpose of the pond water

Rank1 Rank 2 Rank 3 Rank 4 Freq(%) Freq(%) Freq(%) Freq(%) For HHH drinking water 7(9. 2) 15 (19. 74) 18(23.7) 2 (2.6) Drinking water for livestock 4(5. 3) 13 ( 17.11) 9(11.8) 14(18.4) Nursering 26(34.2) 21 (27.6) 12 (15.8) 1(1. 32) Vegetable production 31(40.8) 14 (18.4) 1 (1. 32) 3(3. 95) Spices production 2(2.6) 1 (1. 32) Fruit production 2 (2.6) Washing cloths and food 6(7.9) 10(13.16) 19 (25) 4(5. 3) cooking Total 76(100) 76(100) 59(77.6) 24(31.6)

Table 6 depicts cross tabulation of the type 47 households with plastic cover and none of RWH technologies adopted at plot level basement, 38.3% use metal Bucket for with their corresponding equipments used lifting and watering plants while 29.8% of for water lifting and application. As shown the households use big plastic container in the table, 65.3% of the households ‘Jerikan’. Besides, households with concrete represent those who adopted plastic-lined based ponds mainly use mental bucket RWH pond and those waiting for plastic followed by big plastic container, pulley and sheet. Concrete structures made of clay ‘commendary’ each accounting 20% of the and/or cement accounts 34.7%. Of the total households.

Table 6. Cross tabulation between type of RWH technology and type of water lifting equipments used Type of water lifting equipments used Total Pulley ‘Commendary’ Pot Tridle Jog 'Jerikan' 'Tanika' Bucket pump Ponds 2(4.3)b 7(14.9) 2(4.3) 1(2.13) 14(29.8) 3(6.4) 18(38.3) 47(65.3) covered with plastic and none covered basement % of Total 2.8 9.7 2.8 1.4 19.4 4.2 25 Ponds with 5(20) 5(20) 1(4) 5(20) 1(4) 8(32) 25(34.7) concrete basement % of Total 6.9 6. 9 1. 4 6. 9 1. 4 11.1 Total 7(9.7) 12(16.7) 2(2.8) 1(1. 1(1.4) 19(26.4) 4(5.6) 26(36.1) 72(100) 4) b Values in brackets are percentages. application equipments used in the total 72 In addition, the last raw of table 6 shows the plots with RWH technology. Thus, from the distribution of each type of water lifting and total households with RWH technology

242

majority of them (36.1%) use metal Bucket their ponds. This might result in lots of for lifting and watering plants followed by problems like accident on animals or kids, use of big plastic container (26.4%) and bad smell when the volume of water lowers ‘commendary’ (16.7%).The highest which could be source of malaria, high percentage in the use of metal Bucket for evaporation rate. Of the households with a water lifting and watering plants indicates cover for their ponds 33.3% and 26.7% of the difficulty for a farm household in terms them use wood (trees) and Satera of time as well as labor days required to respectively. Besides, 13.3% of them use irrigate the entire plantation in the plot. This Cob, wood with kenchibe and wood with difficulty is due to lack of capital for buying Sinkita each. On the other hand, with regard or renting simpler equipments which is a to those who use fence to avoid risks, 68.4% major detrimental factor affecting the rater of them use it while the rest 24 households of rainwater harvesting technology adoption. don’t use fence for their ponds. Most of the households use wood as a material to do the As can be seen on table 7 below, only 19.7% fence followed by using wood with kenchibe of the households that adopt the technology accounting 25% and 23.1% of them have a cover for their pond while 80.3% of kenchibe alone. them respond that they didn’t put a cover for

Table 7. If the pond has a cover and fence Does If yes, what Does the If yes, what are your are the pond the materials RWH materials have used ? pond used? fence to have avoid cover? risk? Freq(%) Freq(%) Freq(%) Freq(%) Yes 15(19.7) Wood 5(33.3 ) Yes 52 Wood(acacia 20 (38.5 (68.4) tree) ) no 61(80.3) Cob 2 (13. 3 no 24 Cob 2(3.85 ) ) (31.6) Total 76(100) ‘Satera’ 4(26.7 ) Total 76(100) ‘Kenchibe’ 12 (23.1 ) Wood and 2(13. 3 Cob and 3(5.77 ) ‘kenchibe’ ) ‘kenchibe’ Wood and 2(13. 3 Wood and 13 (25 ) ‘Sinkita’ ) ‘kenchibe’ Total 15(100 ‘Kenchibe’ and 2 (3.85 ) thorn ) Total 52(100 ) * Sinkita and kenchibe are kinds of bush trees. Satera is a grass material

Households with RWH technology were responses mentioned problems related with asked to list problems they encountered agricultural inputs and 9.47% cited problems during implementation and utilization of the related with health. Thus, problem of technology. These include problems related equipment for water lifting and application to RWH pond (33.7%), 37.9% of the total is shown to be the dominant one with frequency of responses represents problems 37.9%. related with lack of equipments, 5.76% of

243

Of the pond related problems, accident on related to RWH pond being the dominant animals and kids, absence of roof cover problems observed, 81.5% of the households followed by quick dry up of the accumulated responded that they need government water problems take the highest share of support or other organization to supply them 39.4, 36.8 and 14.4 percent respectively. with more simple modern materials either by The highest percentage observed in the sharing 50% of the cost or via long term accident could be due to absence of cover credit so that they can produce more. About for the pond, absence of fence to the pond, 40.8% of the households suggest support and wrong location of the pond which might from government to avoid waste of labor increase accident on kids due to closeness to power and time in the process of water the house. The high proportion of uncovered application; we need more simple modern ponds could be due to lack of finance or materials either in the market at lower cost may be due to less awareness given by the or via long term credit since the price of experts or probably due to weakness of the water lifting and watering equipments are households. Quick dry up of the pond water unaffordable at household level. could be related to the RWH technology or structural design of the technology which In addition, for problems related to RWH emanates from lack of extension workers ponds, governments or other organizations with the necessary skill about the technology help or credit to make them buy iron roof during construction or even lack of roof since other raw material don’t stay long and cover for the pond. the need for professional help on the need of having cover and fence to minimize risk Furthermore, of the problems related to accounts 38.1% each. On the other hand, equipments used during pond utilization, the 18.3% indicates the need to have continuous respondents mainly focused on the problem assessment to have positive impact on how of water lifting equipment and lifting of to use and produce in each season and will water from the pond representing (around help to give solution for problems that 78%). This is followed by problem of water household face. application by using heavy materials reducing interest to produce vegetables in a Households with RWH technology were wider place accounting around 42%.In asked to list benefits they get after they start summary, majority of the problems cited by to use the technology, and in general the respondent households revolves around two total frequency of responses (251) reported issues: those related to RWH ponds and the benefits sited by farmers are classified in equipment problems. to four major categories. As can be seen from Table 8, these includes new things Possible solutions were suggested by found after they start to utilize pond households with RWH technology to (48.21%), 39.4% of the total frequency of overcome the aforementioned problems. responses represents benefits related to Most of the solutions suggested focuses water supply or availability, 11.6% of the mainly on the need for government support responses mentioned benefits related with in terms of finance, arranging training or production side and 0.8% are those related experience sharing tour to household heads. to individual opinions. Lack of equipments needed and problems

244

Table 8. List of Benefits Se CATEGORY OF THE BENEFITS Total .N REPORTED o Water New Productio Individua supply things n side l for opinions 1 domestic use 33 (43.4) 33 (13.15) 2 new food varieties in our diet 47(61.7) 47(18.7 3) 3 Reduce consumption expenditure by 28(36.8) 28(11.1 producing what we used to buy from the 6) market 4 For animals especially for those who 37(48.7) 37(14.7 can’t go long distance to drink water. ) 5 It was able to get water for households 29(38. 2) 29(11. easily and timely 55) 6 Produce vegetable beyond home 26(34.1) 26(10. consumption and get money to be used 36) for different purposes by selling the remaining amount. 7 Helps to use water for permanent plants 6(7.8) 6(2. during the dry season e.g. Chat, Coffee, 39) Papaya etc 8 Enable us to produce more than once in 9(11.8) 9(3. a year by using the pond water during 59) dry spell period 9 create new job opportunity by 20(26. 3) 20(7.97 developing the habit of working in dry ) season and use their time better than before 10 Can avoid dry up of pepper nursering 14(18.4) 14(5.58 by using water in the pond ) 11 The negative side out weights positive 1(1. 3) 1(0.4 ) one because the pond construction isn’t dome well and it has no plastic cover 12 I’m glad that the pond isn’t covered by 1(1. 3) 1(0.4 ) plastic or cement basement because it will help not to create bad smell when small animals died Total 99(39.4 ) 121(48.2 29(11.6 ) 2( 0.8) 251(10 1 ) 0)

Of the new benefits observed, 61.7% of the vegetable beyond home consumption and households respond the existence of new sell the remaining to use the money for food varieties in their diet while 36.8, 34.1 different purposes and creation of new job and 26.3 percent are reduction in opportunity by developing the habit of consumption expenditure by producing what working in dry season and use their time we used to buy from the market, produce which isn’t known before respectively. In

245

addition, the existence of water in their environmental sustainability, promoting compound was seen as beneficial for gender equity and women empowerment. It animals especially for those who can’t travel is one way of improving the living long distance to drink water and help the conditions of millions of people, particularly household to get water easily and timely those living in the dry areas. Water scarcity instead of holding heavy material for a long especially for domestic and agricultural distance to fetch water with 48.7% and purposes compromises the role of women in 38.2% respectively. Finally, from the food production. Hence, provision of water production side, 18.4% of the households by promoting rainwater harvesting and responded that it is used to avoid nursering management technologies reduces the of pepper from being dried while 11.8% of burden on rural women and thus increasing them responded that it helps to produce their productivity. more than once in a year using the water during dry season and 7.8% use the water This part tries to see the participation of for permanent plants during the dry season. women in male headed households in planning and decision making stage, Finally, half of the sampled households construction, maintenance, clearance and were asked about the factors hindering watching stages. In addition, it will try to them from adopting the technology. Of address the question if women are benefited the total responses reported, reasons and in what terms, and the reasons if they mentioned related to lack of financial aren’t benefited from adoption of the capital problems represent 41.8% technology. Besides, female headed particularly related to poor economic households were asked if they are selected situation to cover cost involved in pond as beneficiaries and how they are selected, implementation. Besides, 17.2% of them and if not, why not. The constraints that they face to use RWH technology are also are related with lack of knowledge and considered. follow up on the technology and most people don’t think that it will give that Most households replied that there is equal much benefit. Where as, problem of raw responsibility among women and men to materials mainly due to unfair participate in planning and decision making distribution of raw materials needed to accounting for 85.5% of the total rainwater take out the water inside, plot/farm land harvesting technology adopters. This is due to small size land around the followed by 17.1% of households who have homestead and other reasons which mentioned that during planning, the women mainly includes less work initiation suggest the time for the work to provide a mentioned account for 10.7% each from better food service. With regard to construction, 57.9% of the households said the total responses reported. that, women participated directly (by supplying water) and indirectly (by Gender and RWH Technologies preparing food and coffee) for workers. And about 33% of the households suggested that, At present, there is a growing tendency women assisted by providing the needed raw towards the adoption of low cost and simple material (like stone, sand, cement from alternative water management technologies home to where they work etc) and removing like rainwater harvesting technologies. the soil from around the pond to a bit far RWH technologies have the potential to area. contribute towards the Millennium Development Goals (MDGs) with a view of In the case of women participation in eradicating poverty and hunger, provision of maintenance, clearance and watching, safe drinking water and sanitation, ensuring

246

72.4% of the households responded that they in expenditure by using vegetable produced mainly participate in watching kids and for home consumption and selling the animals from getting into the pond remaining. More over, 61.8% of the accidentally since they spent most of their households consider the time saved that time at home. This is followed by their would have been wasted in fetching water participation in cleaning the area of the pond and 22.4% on ability to eat different and accounting 55.3%. Women participation new food varieties. during the dry season to carry out soil or Generally, the result implies that women are sand that enters into the ponds in rainy getting benefit from the technology adoption season has taken 50% of the household’s as any member of the family. Their response. And about 30% of the households participation in the technology adoption is participated in maintenance by bringing mainly in watching the ponds. They also water, raw material, food service and have contribution in planning and decision protecting the pond from being destroyed. making stage, and in giving support during construction, maintenance and clearance of In relation to female headed households, the pond. Female headed households are 67.1% of the households who adopt RWH being constrained to be beneficiaries due to technology responded that they aren’t economic and manpower shortage. selected as beneficiaries whereas the remaining 32.9% replied that they are 5. Conclusions and Recommendations selected to be beneficiaries. Out of those households who responded that female- Conclusions headed households are not selected to be beneficiaries, 68.6% of them mentioned that Due to population increase in the highland the main reason is economic and manpower areas, more and more marginal areas are problem. Less interest and initiation due to being used for agriculture which led to the less participation in agricultural work degradation of the natural resources .One of account for 17.6% of the household’s the major challenges to rural development in response. About 16% of the households the country is how to promote food responded that bias exists towards male production to meet the ever-increasing headed households on the ground that the demand of the growing population. Rainfall ladies can’t go through the hard work, and in the arid and semi-arid areas is generally the same percentage for the reason that they insufficient to meet the basic needs of crop don’t have anyone to teach them about its production. In degraded areas with poor use and purpose indicating less knowledge vegetation cover and infertile soil, most of about the work. On the other hand, out of the rainfall is lost through direct evaporation those households who responded that or uncontrolled runoff. Thus, overcoming female-headed households are selected to be the limitations of these arid and semi-arid beneficiaries, 52% said that government or areas and making good use of the vast agricultural extension is voluntary to give agricultural potential under the Ethiopian chance for anybody depending on their context, is a necessity rather than a choice. working ability in agriculture. About 44% Hence, to alleviate these development replied that it depends on their capacity to constraints, the Federal government and cover cost involved in pond construction. Regional states, and NGOs working in Moreover, 36% of them responded that it is research and development, have invested their own initiation that matters. huge resource on rainwater harvesting technology. With regard to the benefits achieved by women from the adoption of the technology, In this study, methodologies including about 78% of the households responded that descriptive(cropping pattern), econometrics they are beneficiaries in terms of reduction and qualitative analysis are used to assess

247

the determinants of households’ adoption of weren’t grown previously. In addition, it rainwater harvesting ponds, and its impact indicates that effectiveness of the on agricultural intensification and yield in technology adoption is mainly constrained Alaba Woreda, southern Ethiopia. Interview by problems related to water lifting and has also been done with experts on rainwater watering equipments, and accidents harvesting ponds. occurring due to absence of roof cover and fence to the ponds. Generally, directly or The finding in the cropping pattern shows indirectly, labor requirements and cost that, farm households have started to grow considerations appear to be important new crops (vegetables and perennial crops) factors that influence household’s adoption as a result of water availability from the of RWH technology. water harvesting ponds. The crops are those which are highly priced and marketable ones Recommendations implying the potential of RWH technologies to enhance a farm household’s income. The benefit found from the high valued and However, the benefit depends on market and perishable commodities due to RWH, infrastructure accessibility, and depends on market and infrastructure diversification in the types of the crops. accessibility, and diversification in the types Results of Probit analysis on the of the crops. Thus, efforts should be made to determinants of adoption of rainwater assess various agricultural commodities as harvesting ponds shows that household size, well as giving emphasis to marketing education status of household head, extension, especially in facilitating markets ownership of livestock (cattle, oxen and and market linkages to farmers. pack animals), homestead plots and type of The impact of household RWH technology pond explained adoption statistically adoption on the value of crop yield has been significantly. found to be stastically significant. Therefore, to mitigate the erratic nature of rain fall in In accordance with government’s target, the the arid and semi-arid parts of the country, Ordinary Least Square estimation of the development and implementation of rain determinants of the value of crop production water harvesting technologies will be shows that adoption of RWH has a positive helpful to promote productivity and and statistically significant effect on value of sustainable intensification of the rain fed crop production, after controlling for input agriculture. use and other factors. This shows that RWH However, the success of the technology ponds have direct and significant impact on adoption is mainly constrained by problems value of crop production. We also find that related to water lifting and watering households with RWH technology use more equipments, and accidents occurring due to labor and seed but less oxen power absence of roof cover and fence to the compared with those households who have ponds. This implies that support will be not adopted the technology. Moreover, labor needed to provide affordable but improved and seed inputs have positively significant water lifting and watering equipments, and impact on yield while the effect of oxen give training to farm households on power is insignificant. These results show construction and use of roof covers and that in addition to its direct impact, RWH fences to the ponds. has significant indirect impact on value of Labor requirements and cost considerations crop production through its effect on appear to be important factors that influence intensity of input use. household’s adoption of RWH technology. Results of the qualitative information, This implies that research and development consistent, with the crop mix and interventions need to take account of the econometric results, also showed that labor and cost demands of the technology. households started to grow crops that

248

RESULTS OF ECONOMETRIC ESTIMATION

Table-2 Determinants of adoption of RWH pond (Probit)

Probit use of RWH technology Coefficient Z P>z Explanatory Variables (dF/dx) ‡ Peasant association dummy,cf., Mudda Dinokosa Ulegebba Kukke -0.0007837* -1.85 0.065 Andegna Hansha -0.0004302 -1.01 0.312 Hamata -0.0003513 -0.72 0.472 Household access to services and infrastructure Walking time to the nearest town market (in hrs) -0.0001269 -0.61 0.545 Walking time to the nearest village market (in hrs) 0.0001965 1 0.316 Walking time to the nearest cooperative shops (in hrs) 0.0001392 0.52 0.603 Walking time to the nearest all weather road (in hrs) 0.0002143 1.02 0.308 Walking time to the nearest seasonal road (in hrs) -0.0000296 -0.06 0.954 Rain fall condition, cf., low Medium -0.0004712 -0.84 0.401 High -0.000446 -1.46 0.145 Household size 0.000111** 1.96 0.05 Age of household head ( in Ln) 0.0002167 0.29 0.772 Education level of household head, cf., illiterate Read and write 0.0079635*** 3.25 0.001 Up to 4th grade 0.0018686 1.44 0.149 Up to 7th grade 0.00026301* 1.86 0.063 Up to 10th grade 7.41E-06 0.01 0.991 Household resource endowment Land owned (in ha) -0.000184 -0.85 0.395 Value of cattle (both local & cross bred cows, calves, heifers, yearling, bulls) 3.59E-07** 1.98 0.048 Value of oxen (local and breed) 5.24E-07** 2.2 0.027 Value of sheep and goat -4.44E-07 -0.72 0.472 Value of pack animals (donkey, horse, mule) 6.69E-07* 1.88 0.06 Value of poultry (both local & improved) 2.19E-07 0.64 0.519 Value of beehives (improved, modified, traditional) 3.85E-08 0.27 0.79 Value of all assets owned (plow set, farm equip, motor pump, radio,.. -3.23E-08 -0.33 0.74 Household membership in local organization, cf., members in Edir and other local organizations Membership in Edir only 0.0002847 0.7 0.487 Household membership in associations, cf., association members No membership in association -9.37E-06 -0.02 0.985 Household financial capital , 1= yes Household with credit Access,1= yes -0.0000753 -0.17 0.865 Household savings, yes=1 -0.0002764 -0.71 0.478 How household acquired the plot, cf., rented and share cropping Allocated by the state 0.5627719 0.00 0.997 Inherited 0.5999944 0.00 0.998 Slope of the plot, cf., steep slope Flat 0.0044407 0.00 0.999 Moderate 0.0686505 0.00 0.999 Soil depth of the plot, cf., deep Shallow -0.0002766 -0.32 0.751 Medium -0.0001365 -0.11 0.912 Soil fertility level of the plot, cf., low fertility High fertility 0.0141321 1.25 0.21 Moderate fertility 0.0010029 1.11 0.267 Purpose for which the land is used, cf., grazing ,woodlots and spice land Cropland -0.0002559 -0.33 0.74 Homestead 0.0695164*** 4.8 0.000 Plot size in ha (in Ln) 0.0005554 0.94 0.345 Walking distance from household's residence to the plot (in hrs) -0.00168 -0.72 0.472 Type of pond, cf., ponds with plastic cover and those without a cover Ponds with concrete basement -0.377571*** -4.54 0.000 Number of observations 1036 LR chi2 (41) 350.92 Prob > chi2 0.0000 Pseudo R2 0.6399 *** is significant at 1%; ** is significant at 5%; * is significant at 10% ‡Reported coefficients represent effect of a unit change in explanatory variable on probability of adopting RWH technology.

249

Table – 3 Determinant factors of input use during 2005/06 agricultural fiscal year

Whether Whether Ln (Oxen- Ln (Labor- fertilizer manure/compos Explanatory Variables Ln (Seed/ha) days/ha) day/ha) were used t were used Peasant association dummy,cf., Mudda Dinokosa Ulegebba Kukke -0.245172 0.15099* 0.058052 0.0655231 -0.0197904 Andegna Hansha 0.214534 0.203828*** 0.039733 -0.1935646*** 0.079232 Hamata 0.001953 0.168604** 0.172659** -0.1475076** -0.0190538 Household access to services and infrastructure Walking time to the nearest town market (in hrs) -0.104291** -0.016135 0.020109 0.206203 -0.0265866* Walking time to the nearest village market (in hrs) -0.125701** -0.072537*** -0.117138*** -0.0425217* -0.0363848** Walking time to the nearest cooperative shops (in hrs) 0.034241 -0.02963 -0.057824* -0.0280787 -0.0054926 Walking time to the nearest all weather road (in hrs) 0.040986 -0.011034 0.022569 -0.0090631 0.0078478 Walking time to the nearest seasonal road (in hrs) 0.184175 0.097555 -0.110871* 0.0753763 -0.129366*** Rain fall condition, cf., low Medium -0.084553 0.112657** -0.054333 0.0087776 0.0026803 High -0.091135 0.008501 -0.212387*** 0.0527761 0.2818222*** Household size 0.026266* 0.021049*** 0.043193*** -0.0024128 0.0094189* Age of household head ( in Ln) 0.125784 0.108762 -0.181818* -0.0654953 0.0254648 Education level of household head, cf., illiterate Read and write 0.230052* 0.231572*** -0.087174 -0.0931605 0.0654167 Up to 4th grade 0.257753* 0.192213*** -0.078671 0.0288443 -0.0862418** Up to 7th grade 0.083556 -0.024551 0.002305 -0.0171464 0.0307067 Up to 10th grade 0.071938 0.080617 -0.053017 -0.0293807 -0.0785635 Household resource endowment Land owned (in ha) 0.007845* 0.006203** 0.00167 0.0027194 -0.0037889** Value of cattle (both local & cross bred cows, calves, heifers, yearling, bulls) -1.73E-05 -5.90E-05 -6.98E-05*** 4.99E-06 -0.0000345** Value of oxen (local and breed) 4.28E-05 2.83E-05 4.82E-05* 0.0000103 0.0000485*** Value of sheep and goat 0.000167 0.000129* -9.97E-07 -5.99E-06 -5.83E-06 Value of pack animals (donkey, horse, mule) -0.000118 -0.000051 -8.93E-05** 7.97E-06 -5.84E-06 Value of poultry (both local & improved) -0.000809 0.000172 0.000323 -0.00039 0.0003529 Value of beehives (improved, modified, traditional) -0.00041 0.000376* 0.000197 0.0003235* -0.0004251*** Value of all assets owned (plow set, farm equip, motor pump, radio, ...) 3.62E-06 -1.66E-05 -2.19E-05 7.05E-06 -3.57E-06 Household membership in local organization, cf., members in Edir and other local organizations Membership in Edir only -0.215644 -0.115894 -0.210552*** -0.089469 0.0591204 Household membership in associations, cf., association members No membership in association -0.094869 -0.191782*** 0.042779 -0.0621948 -0.0014808 Household financial capital , 1= yes Household with credit Access,1= yes -0.137139 0.070683 -0.06814 0.0624094 0.056192* Household savings, yes=1 -0.072473 -0.327655*** -0.114424** 0.0126967 0.1128724***

250

Table – 3 continued

Whether Whether Ln (Oxen- Ln (Labor- fertilizer manure/compost Explanatory Variables Ln (Seed/ha) day/ha) day/ha) were used were used How household acquired the plot, cf., rented and share cropping Allocated by the state -0.506682*** -0.141824* 0.084312 -0.1988535*** 0.158752*** Inherited -0.382232*** -0.169708** -0.111456* -0.1364283** 0.1498123** Slope of the plot, cf., steep slope Flat -0.119189 0.530278* 0.446515* 0.1701381 0.3856669* Moderate -0.10287 0.51544* 0.547266** 0.1265144 0.2790531** Soil depth of the plot, cf., deep Shallow -0.021532 0.129045 -0.117212 -0.0475644 0.2127672 Medium -0.000324 -0.300583*** -0.315847*** 0.0428845 0.1378711* Soil fertility level of the plot, cf., low fertility High fertility 0.048873 0.101733 0.035063 -0.0829447 0.1586607** Moderate fertility 0.144556 0.089368 0.062933 -0.0517906 0.479061 Purpose for which the land is used, cf., grazing ,woodlots and spice land Crop land 0.419156*** 0.37224*** 0.614584*** 0.4647761*** -0.0924947** Homestead 3.09079*** -0.340097*** -0.472505*** -0.5890224*** 0.4247779*** Plot size in ha (in Ln) -0.180882 -0.912926*** -0.779754*** -0.2589599*** 0.539933 Walking distance from household's residence to the plot (in hrs) 3.312421 0.011153 -0.12605 0.2058507** -0.1616669 Adoption of Rain Water Harvesting technology (predicted value), 1=yes 3.312421*** -0.291091* 0.265723* 0.1043238 0.0748814 Constant 4.448353*** 4.83144*** 6.78531***

Number of observations 1036 1036 1036 1036 1036 F (41,994) 8.80 14.08 14.46 Prob > F 0.0000 0.0000 0.0000 R squared LR chi2 (41) 281.62 353.37 Prob > chi2 0.0000 0.0000 Pseudo R2 0.1964 0.3137

*** is significant at 1%; ** is significant at 5%; * is significant at 10% Reported coefficients represent effect of a unit change in explanatory variable on probability of use of the mean of the data Ln represents natural logarithm

251

Table – 4 Determinants factors of value of crop yield

Ln (Value of yield/ha) Reduced Explanatory Variables Full Model ‡ Structural Model ¶ Model Peasant association dummy,cf., Mudda Dinokosa Ulegebba Kukke -0.240465** -0.16942** -0.272749*** Andegna Hansha -0.091321 -0.05626 -0.101886 Hamata -0.332615*** -0.29741*** -0.387513*** Household access to services and infrastructure Walking time to the nearest town market (in hrs) -0.037325 -0.02798 -0.037513 Walking time to the nearest village market (in hrs) 0.039986 0.041098 0.01502 Walking time to the nearest cooperative shops (in hrs) -0.017744 -0.03863 -0.016557 Walking time to the nearest all weather road (in hrs) -0.020955 -0.01405 -0.020943 Walking time to the nearest seasonal road (in hrs) -0.13985* -0.16159** -0.083644 Rain fall condition, cf., low Medium 0.016212 0.01092 0.003531 High 0.10563 0.095822 0.08433 Household size 0.008924 0.015446* Age of household head ( in Ln) -0.1558997 -0.13447 Education level of household head, cf., illiterate Read and write 0.007438 -0.059152 Up to 4th grade 0.064804 0.110153 Up to 7th grade 0.058197 0.079857 Up to 10th grade 0.123428 0.107066 Household resource endowment Land owned (in ha) 0.00154 0.0031 Value of cattle (both local & cross bred cows, calves, heifers, yearling, bulls) 4.44E-05* 4.55E-05* Value of oxen (local and breed) -3.44E-05 -1.22E-05 Value of sheep and goat 9.65E-05 8.20E-05 Value of pack animals (donkey, horse, mule) 8.94E-06 -3.14E-05 Value of poultry (both local & improved) 0.000275 0.00021 Value of beehives (improved, modified, traditional) 4.64E-06 -3.61E-05 Value of all assets owned (plow set, farm equip, motor pump, radio, ..) -8.60E-06 -7.41E-07 Household membership in local organization, cf., members in Edir and other local organizations Membership in Edir only -0.12421 -0.14033 Household membership in associations, cf., association members No membership in association 0.133489* 0.077884 Household financial capital , 1= yes Household with credit Access,1= yes 0.084706 0.045664 Household savings, yes=1 0.01175 -0.000479 How household acquired the plot, cf., rented and share cropping Allocated by the state 0.285989*** 0.220717*** 0.175439** Inherited 0.14397* 0.09171 0.047545

252

Table – 4 continued Ln (Value of yield/ha) Explanatory Variables Full Model Structural Model Reduced Model Slope of the plot, cf., steep slope Flat 0.107935 -0.05085 0.157219 Moderate 0.213 0.052619 0.253161 Soil depth of the plot, cf., deep Shallow -0.342699** -0.2061 -0.276843* Medium -0.320594** -0.2085 -0.269564* Soil fertility level of the plot, cf.,low fertility High fertility 0.083002 0.12039 0.042061 Moderate fertility 0.10888 0.136898* 0.099062 Purpose for which the land is used, cf.,grazing ,woodlots and spice land Cropland 0.545698*** 0.53749*** 0.692927*** Homestead 0.22273* 0.273696*** 0.376867*** Plot size in ha (in Ln) -0.056483 -0.02842 -0.123963* Walking distance from household's residence to the plot (in hrs) 0.085783 0.101174 0.077678 Labor-day/ha (in Ln) 0.101176*** 0.110689*** Oxen-day/ha (in Ln) 0.018104 0.006066 Seed/ha (in Ln) 0.086711*** 0.086715*** Use of fertilizer,1= yes 0.164603*** 0.171696*** Use of manure/compost, 1= yes -0.115259* -0.11909* Adoption of Rain Water Harvesting technology (predicted value),1=yes 0.055424 0.510136*** Constant 6.686813* 6.272492*** 7.859654***

Number of observations 1036 1036 1036 F (46,989) 8.11 F(27,1008) 12.18 F (41,994) 6.14 Prob > F 0.0000 0.0000 0.0000 R squared 0.125 0.0967 0.0953

*** is significant at 1%; ** is significant at 5%; and * is significant at 10%. Ln= natural logarithm. ‡ Reported coefficients represent effect of a unit change in explanatory variable on probability of use of the mean of the data. ¶ Variables that were jointly statistically insignificant in the unrestricted OLS regression were excluded from the structural model

253

Reference: Kerr, J. M. and Kimberly, R. C., 2005. Demeke M. and Ferede T., (2005). The Evaluating Watershed Project Performance of Grain Marketing in Performance in India: A Practical Ethiopia: The Case of Addis Econometric Approach. In: Shiferaw, B., Ababa Central Market, in S. W. Omamo, Freeman, H. A., and Swinton, S. M. S. Babu, and A. Temu (eds) “The Future (Eds.) Natural Resource Management in of Smallholder Farming in Eastern Agriculture: Methods for Assessing Africa: The Roles of states, markets and Economic and Environmental Impacts. civil society”, IFPRI Eastern Africa UK: CABI Publishing in association Food Policy Network Report. with The International Crops Research Institute for the Semi-Arid Tropics Desta L., 2004. Concepts of Rainwater (ICRISAT), Patancheru 502 324, Andhra Harvesting and Its Role in Food Security Pradesh, India; 2005. – The Ethiopian Experience. Paper presented on a National Water Forum, Addis Ababa October 25-26, 2004. Nega H., 2004.Managing rainwater in dry Ministry of Water Resources. land areas of Ethiopia for improved agricultural development: The Ministry FAO. 1993. Irrigation Water Delivery of Agriculture experience. Paper Model. Food and Agriculture presented in the second General Organization; Water Reports No. 7 Assembly Proceedings of Ethiopian Rome, Italy. Rainwater Harvesting Association.

Gebeyehu A., 2006.Water for growth and Nega H., 2005.Training manual on poverty alleviation. Paper submitted to Rainwater Harvesting and Utilization Ethiopian Economic Association. Techniques. Addis Ababa, Ethiopia

Gebremedhin B., Pender J. and Tesfaye G., Ngigi, S.N., 2003.Rain water harvesting for 2000.Community Resource Improved Food Security. Promoting Management: The case of grazing lands Technologies in the Greater Horn of In Tigray, Northern Ethiopia. In: Jabbar Africa. Greater Horn of Africa M. A., Pender J., and Ehui S. (Eds.), Rainwater Partnership (GHARP).Kenya Policies for Sustainable Land Rainwater Association (KRA). Management in the Highlands of Ethiopia. Proceeding of a workshop held Pender J. and Gebremedhin B., 2004. at The International Livestock Research Impacts of policies and technologies in Institute (ILRI). Addis Ababa, Ethiopia, dry land agriculture: Evidence from 22-23 May 2000. Workshop Summary northern Ethiopia. In Challenges and Paper No.9. Strategies for Dry land Agriculture. CSSA Special Publication no.32. Gebremedhin B., Pender J. and Tesfaye G., 2002. Nature and determinants of Pender J., Nkonya E., Jagger P. and collective action for Woodlot Sserunkuuma D., 2001. An econometric management in northern Ethiopia. Socio- approach to NRM impact assessment: an Economic and Policy Research Working example from Uganda. In: Shiferaw B., Paper40. ILRI (International Livestock Freeman H. A. (2003Eds.) Research Institute), Addis Ababa, Ethiopia. Rämi H., 2003.Ponds filled with challenges. Water harvesting-experiences in Amhara Hatibu N., Senkondo M. , Mutabazi K. and and Tigray. UN OCHA-Ethiopia Msangi K. “Economics of Rainwater Harvesting for Crop Enterprise in Semi- RELMA-in-ICRAF, 2004. Sear Net Briefs. Arid Areas” downloaded from Supported by A Net work for “Green” www.regional.org.au/cs/2004/symposia/ Water Harvesting (GWH) in Southern 1/4/276_hatibu.htm , accessed in and Eastern Africa and South Asia. October 2004 Nairobi-Kenya.

254

Takele E., 2006. Review on the effectiveness Shiferaw B. and Freeman, H.A., 2003. and impacts of small- scale irrigation in Methods for Assessing the Impacts of Ethiopia. World Bank. Natural Resource Management Research. A summary of the Turner B., 1994. Small – Scale Irrigation in proceedings.ICRISAT, Patancheru 502 Developing Countries. Land Use Policy 324, Andhra Pradesh, India. 11 251 – 61.

Shiferaw B., Freeman, H.A. and Swinton, UNEP, 2005.Rainwater Harvesting and the S.M., 2005.Natural Resource Millennium Development Goals. Management in Agriculture. Methods for Assessing www.rainwaterharvesting.org/rural, An Economic and Environmental Impacts. interesting account on Rainwater CABI publishing harvesting Experience of India

255

Discussion on Theme 2: Irrigation Impact Poverty and Economy

Chair: Dr. Birhanu G/Medhin Rapporture: Micheal Menkir

The chairman for this session introduced the theme and the floor was opened for questions, comments and suggestions.

Questions and Discussions 2.7 Using chow´s test you were able to pool the data of Doni 2.1 What are the inefficiencies and Godino but not Batu variables, that the government Degaga. Does this mean policy makers can take up and features were behaving in the go for improvement same manner? Given that their location is different. 2.2 What is the level of inefficiency/efficiency within 2.8 Area expansion increases irrigated agriculture itself? agricultural production which Since it is obvious that the has some contribution to production level of farmers poverty reduction. However cannot fell on the frontier line. extensive agriculture has its How much is the efficiency or negative impact on the natural inefficiency difference between resource i.e degradation. Your irrigated and rainfed analysis is based on extensive agriculture? agriculture rather it is better to consider intensive agriculture 2.3 The year 2005/2006 was a high for land and water productivity rainfall year; so percentage development. Therefore how do contribution of irrigation on you see the natural resource GDP would be less than in a degradation and environmental draught year. deterioration in your poverty analysis Ans-yes analyses may have underestimated percentage 2.9 Efficiency issue should be seen contribution slightly. clearly the with respect to rainfall percentage contribution of availability. In Godino water is irrigation varies b/n good and abundant, rainfall rich. poor rainfall year Therefore irrigators are less efficient. However in Batu 2.4 How sustainable is use of Degaga water is pumped and it irrigation from experiences of has a cost. So this is incentive salinity? for higher efficiency. The area is also dry land with highly 2.5 Development especially in high variable rainfall. Therefore evaporation areas in large scale while considering efficiency irrigation schemes. It is water availability and rainfall suggested that future studies should be an important should consider this aspect. parameters.

2.6 What should be the size of the 2.10 Increase of water supply by 1 sample area of irrigated percent leads to 0.5 percent agriculture to be representative output what should be the limit to talk about its contribution the o f applying more water, since national economy (GDP) over application will lead to miss management and

256

inefficient water use. In the (Trzpanosomiassis), lowland future farmers may be able to livestock disease. pay for the water they use for irrigation (from experience of 2.17 In your conclusion less choice is other water scarce countries like put as a negative aspect but we Morocco, Jordan and Israel. Is found those who specialised there a possibility of using this (follow one cropping pattern) in terms of cost recovery and are the richest of the operational and management beneficiaries as they can buy cost of irrigation projects? their food crops.

2.11 While talking about the need for Ans- the livestock disease irrigation need for the countries happened some 10 years ago, GDP growth. Are we but now lack of grazing land is considering other sectors using the main reason. And less the same source like choice of food is related to the hydropower, water supply education status

2.12 Quality of water definitely 2.18 There is a confounding effect decreases as consumptive use between irrigation and rain fed (irrigation ) increases. so is not agriculture. Farmers with access important to consider the to irrigation could be making decrease in the value of same more money or income from volume of water in the future in their rain fed production. How calculating or equating can we deal with this problem? monetary value of water? 2.19 What are the differences 2.13 How do you say that irrigation between depth of poverty and time increases production? severity of poverty? What are Irrigation time usually depends the parameters that are required on the stream size a farmer is to address these two terms? receiving. 2.20 What is the optimal investment 2.14 How is it possible that farmers cost per hectare of irrigation located at the tail end of the projects specially small scale system are less efficient and at projects the same time dry land farmers are more efficient. Answer- 2.21 What are the inefficiency Found the comment valid but variables which can be taken by could not consider during the the government study due to the complexity of determining the volume of 2.22 Definition of technical water received by a farmer. efficiency. Due to inherent Answer , Tail end users are less nature of inefficiency every efficient due to water limitation farmers cannot fall towards the because of over abstraction of frontier line water by upstream users 2.23 There are different kinds of 2.15 In your recommendation you small scale irrigation which stated that households with irrigation systems are viable access to irrigation will remain from the 25 SSI poor what does this indicate? 2.24 Did you see the effect of 2.16 The reason for livestock supplementary irrigation in your absence is not only because of analysis less grazing land but mainly due to sleeping sickness

257

2.25 Have you considered the cost of the dam? Is the water free? 2.34 Do you see the size of irrigable land which should be allocated 2.26 No one mentioned about example in Tigray it is 0.2 ha if sustainability of the irrigation tit is more it is not manageable? projects. Salinity in middle So did you come across of such awash valley....? kind of analysis.

2.27 Why did you leave commercial 2.35 What is the limit for farmer like the one in Maki investment? what kind of Ziway area in your analysis marketing is essential to impact the GDP? The other problem is 2.28 Change in quality of water what the discrepancy between land is the economic impact of low and water availability quality water 2.36 Share cropping is widely 2.29 Input-output pricing are they practiced. farmers lease their incorporated in your analysis land. Do you consider this when you talk of impact of irrigation 2.30 How do you identify poverty in the beneficiaries of irrigation 2.37 What is the reason behind for schemes female household to be more food secured than male headed 2.31 Female Household are they household? included? what is the finding 2.38 In the Alaba presentation the with respect to Female positive impact of water households harvesting structures is shown what about its impact on health? 2.32 What is the difference between Is the adoption continued even depth and severity of poverty after its introduction by the government? 2.33 Interaction between rain fed and irrigation systems. How do you deal with confounding effect

258

Theme three:

Irrigation Institutions and Support Services

Papers 1. Status of irrigation institutions and support services in Ethiopia 2. Does access to small scale irrigation promote market oriented production in Ethiopia? 3. Multiple Effects of Small Scale Irrigation in Ethiopia

259

Status of Irrigation Institutions and Support Services in Ethiopia

1Tena Alamirew, 1Desalegn Chemeda, 1Tesfay Beshah, 2Sileshi Bekele 1Haramaya University, 2 IWMI-NBEA [email protected]

Beneficiaries lack skills and institutions to Abstract manage common property resources; consequently, irrigation infrastructure quickly Ethiopia is investing good amount of scarce falls into a state of disrepair. Substantial resources on irrigation development. But the numbers of the beneficiaries don’t feel that performance of many of the irrigation schemes they own/control the water. In the perception is often far from satisfactory with of the many irrigators, maintenance of the disappointing results of public investments. In headwork and main canal are the responsibility this study, the type and performance of of the ‘government’. But government irrigation institutions and availability of organizations are more focused on the support services were investigated taking three development of new schemes. Maintenance large and nine small scale irrigation schemes was observed to be the most serious problem from different parts of the country. (74%) that caused underperformance of many schemes. The main cause was attributed to It was noted that at the macro-scale, the lack of fund (37%) and poor organization and mandates the Ministry of Agriculture and planning. Rural Development and the Ministry of Water Resources in irrigation development were not Water shortage ensued water theft and clearly articulated and scrupulously tended. unauthorized canal breaching are said to be the The set up of irrigation institutions from major (78.1%) sources of conflict and water Federal to Woreda level is frequently changing shortage in the irrigation schemes under ensuing institutional memory lapse, investigation. duplication of efforts, and lack of accountability. The roles of Water Bureaus and Marketing problem caused by absence of the Agriculture and Rural Development communal planning, contractual farming, Bureaus with respect to irrigation development transport access and staggered production are not clearly defined to date. Irrigation system targeted for market was cited as the extension service was observed as an disincentive to expand irrigated crop unsatisfactory and the training of farmers in production. Supply of seed, pesticide and irrigated crop production was wanting. insecticide is also identified to be the major challenges in diversifying production. The Only the agro-industrial state schemes - availability of support service in terms of sugarcane and cotton farms - are relatively inputs (seeds, fertilizer, herbicides, fuel, farms well managed with little or no institutional and implements) was noted to be below support service problems. Many of modern satisfactory. The Bureaus of Agriculture are small scale irrigation schemes (86%) are said to be the major provider of support nominally managed by Water User Association services (35%). The service from research (WUA) with well crafted bylaws. However, institution was reported as minimal. Service many of them lack the authority to enforce Cooperatives organized by the Cooperatives them. The use of local courts to fine offenders Promotion Agency have stared to deliver was noted to be ineffective. Compared to fertilizer in their cooperative shops recently. formal institutions like water user association, traditional institutions were found to be better In conclusion, irrigation water management efficient for their penalty sanction mechanisms institution is to be established and empowered are stronger. with appropriate statutes. It is to be provided with extension and support services. The beneficiaries should be consulted in the

260

planning and they should be accountable for the public investment.

261 Does Access To Small Scale Irrigation Promote Market Oriented Production In Ethiopia?

Fitsum Hagos1, Godswill Makombe1, Regassa Namara 2, and Seleshi Bekele Awulachew1

1 International Water Management Institute for the Nile Basin and East Africa, 2 International Water Management Institute (IWMI), Africa Regional Office. Accra, Ghana. [email protected]

and volume of sale. Increased support to Abstract education can, thus, help in the long-term to transform traditional subsistence agriculture The study examined the extent and nature of into more market-oriented agriculture. market oriented production in irrigated Finally there are unobserved site specific compared to rainfed systems in Ethiopia. By effects, related to location and other doing so the paper identifies the role of covariates, which influence market irrigation in market-oriented production, participation and volume decisions. while at the same time highlighting the main constraints to market oriented development. Key terms: irrigation, change in cropping Our results indicate that irrigation mix, market participation, volume of sale, contributes significantly to increases in Probit and Truncated regression; Ethiopia, market participation, volume of marketed Africa. produce and, hence, income, by inducing shifts in farmers’ cropping mix. The impact 1. Introduction of commercialization of production on household food security is not direct and Irrigation development is expected to immediate mainly because of failures in the increase market participation of producers food market. (Rosegrant et al., 1995; MoFED, 2006). Higher yields, higher cropping intensity and While irrigation enhances market all year round farm production leads to production, there are series of factors that increased market-oriented production, pose serious constraints to market implying a shift in supply (marketable production. Land size, oxen holding, access surplus production) and perhaps food to market and means of transport were found security. Irrigation is also expected to lead to to be important determinants of market changes in crop mix (cash crop orientation) oriented production calling for policy which is expected to have far reaching interventions in land markets, access to consequences on household welfare (Joshi et productive assets and infrastructure al., 2003). Crop-switching as Hussain and development and policy measures to Hanjra (2004) noted involves substituting improve the performance of agricultural low yielding and low profitable crops with new high-yielding and more profitable markets. The study also found education has crops. Implicitly this implies switching from market promoting effect in terms of subsistence production to market-oriented increasing the probability of participation production (ibid.). There are reports, however, that indicate that increased market orientation may not necessarily ensure food security especially if the macroeconomic

environment is not conducive or there are highly vulnerable subsistence-oriented distorted trade policies or there is poor structure of the economy. It further indicated infrastructure development (Van Braun, that smallholder farmers, generally with less 1995) or social protection for food security than 1 hectare of land, account for about 95 is not provided through markets and percent of the agricultural output. In times of government interventions (de Janvry et good weather, roughly 75-80 percent of the al.1991). output is consumed at the household level (World Bank, 2006). Bhattarai and Pandy In risky environments such as Ethiopia, (1997) in their study in Nepal indicated that smallholder farmers, who constitute the bulk wheat production was economically more of the population, are often caught in profitable in locations with better access to production of low-risk/low-return food irrigation and rural infrastructure. They also grains. With insufficient cash funds, and found that farmers with access to irrigation unpredictable outcomes, they cannot afford and markets are found to be much more to take the risk of diversifying from responsive to changes in wheat prices than subsistence food production into potentially farmers without access to such higher-return ventures (such as growing infrastructure, indicating the cash crops for market), or of spending their complementarity between infrastructure limited cash on purchased agricultural development and access to market and crop inputs, because if they fail – either because productivity. Lapar et al. (2003) pointed out of crop failure, price collapse, or lack of that smallholders generally have inadequate demand – they will not have either the basic capital resources—including, physical and food they would otherwise have produced, financial resources, but also intellectual nor the cash to purchase it, and their families capital resources such as experience, will go hungry (MOFED, 2006 p.6). education and extension— which limits their Irrigation removes some of the risks ability to diversify production portfolios. associated with rainfall variability and Lapar et al. (2003) further indicated that the thereby increases the likelihood of using inability of smallholder producers to take purchased quality inputs due to the reduced advantage of economies of scale in risk of crop failure. Irrigation is, hence, production and marketing is a significant expected to remove or ease risk so that impediment to market participation. farmers can venture into an inherently high Smallholders are often disadvantaged due to risk-high return production pathway, which poor access to information and market- may have a significant effect on poverty precipitating services such as extension reduction (MoFED, 2006). visitation and credit assistance and these impediments often give rise to low rates of While irrigation development is expected to adoption of improved technologies that induce such changes, the realization of these could potentially increase productivity, effects cannot be taken for granted. This diversification and, hence, market could be especially true in countries like participation. In addition, poor infrastructure Ethiopia, where many of the preconditions often increases the transaction costs of for market production seem to be missing. smallholder market participation. The households’ orientation towards market production is often hampered by various However, there is little empirical evidence factors at the household and village levels, on market participation in developing by market access conditions and other countries, particularly in Africa. The limited institutional and policy factors. The World studies there are focus on smallholder Bank (2006) indicated that current limited producers’ decision to participate in coarse access to transportation and markets grain markets (Goetz, 1992) or in livestock undermines incentives for surplus markets (Lapar et al. 2003; Bellemare and agricultural production and reinforces the Barrett, 2006). To our knowledge there is no

263

study, which has systematically investigated and econometric approaches in parts three the role of irrigation in inducing market- and four. In section five the study site and oriented production in Sub-Saharan Africa. data description and some descriptive A sound understanding of the patterns of statistical summary results are presented. market oriented production and the Part six discusses the econometric results constraints it faces could contribute to the and part seven concludes and draws policy development of more appropriate policies conclusions. regarding institutional arrangements and the creation of adequate infrastructure, which 2. Modeling participation and supply could benefit a large mass of smallholder decisions producers. This study is an attempt in this direction. Specifically it aimed to: (i) We developed a simple conceptual examine the extent and nature of market framework that captures interactions, oriented production in irrigated sites in processes and outcomes that result from contrast to rainfed areas in Ethiopia; (ii) irrigation development. Unlike rainfed identify the determinants of market-oriented agriculture, irrigation development enhances production, including the role of irrigation in cropping intensity as households are able to the process, and (iii) draw implications of produce more than once in a year. Irrigation market oriented production on food security also opens new horizons for growing new and poverty reduction. crops which are not usually possible under rainfed conditions (Joshi et al., 2003; We used a unique dataset covering various Hussain and Hanjra 2004; Hussain, 2005; small and medium scale irrigation schemes, Huang et al., 2006). Furthermore, irrigation both traditional and modern. Corresponding development enhances increased use of data from rainfed systems were used as a purchased inputs by reducing the risk of control. We explored the differential impact crop failure and increasing returns to of irrigation development on market agriculture and, hence, increasing production as contrasted to rainfed systems. household’s willingness to use purchased In explaining a household’s decision to farm inputs such as fertilizer, herbicides and participate in the market we introduced the pesticides and also hired labor (Hussain and distinction between participation per se and Hanjra, 2004). These changes in cropping volume decisions (i.e. level of participation). intensity and shift in cropping choice Where participation investigated whether the (diversification) are expected to have far household produces and sells products to the reaching consequences on food security and market regardless of the amount (value) of poverty, not least through the market sale and the level of participation behavior of smallholder farmers (Pandey investigated the factors that influence the and Sharma, 1996; Hussain and Hanjira, quantity of sale. It is difficult to assume a 2003; Hussain and Hanjra 2004; Huang et priori that the factors that influence the al., 2006). household’s decision to participate in the market are different from the factors that Irrigation development is expected to trigger influence volume decisions. Hence, we also this host of processes. However, while tested whether the decision to participate irrigation is the necessary condition to and the volume of sale are made induce these changes, it is not as such a simultaneously using appropriate sufficient condition as there are various econometric techniques. factors that influence these processes. First we present the theoretical model that focuses The paper is presented as follows. Part two on the household’s decision to produce for presents a theoretical model for modeling the market before we present the possible participation and volume decisions followed factors that influence market participation by the presentation of testable hypotheses and volume decisions.

264

household covariates; and We consider market participation and supply μ si ~ N(0,σ si ) denotes random error. decisions in the context of traditional Probit Unlike the latent specification in the Probit and Tobit models applied to household model, the dependent variable in (2) takes production data (see Lapar et al. 2003). For on positive and zero values. When a zero each household, i , i = 1, 2,…N, assume value is observed, we assume this to imply that the observed data, namely yi = 1 if that the household in question, rather than possessing an excess of the marketable participation is observed and yi = 0 product, actually has a demand for the otherwise, is conditioned by a K-vector of commodity (that is, a negative supply). household-specific covariates, xi . The Hence, sales quantities are left-censored at decision rule is to participate when the zero. utility of doing so, say, U i ()xi exceeds utilityV ()x , which is the utility reaped 3. Hypotheses i i from some alternative enterprise ( e.g. to In this section we present, in the form of produce food crops). Taking Taylor-series testable hypothesis, various factors that expansions of these two utility functions influence the irrigation-market production around the point xi = 0, yields the linear nexus. In most rural economies, farm households model, yi = 1 if xi γ ≥ xi μ , yi = 0 if are dominant decision-makers when it xiγ < xi μ , where γ and μ are K-vectors comes to the management of land and water of first-order effects depicting the impacts resources. Farm households appear to on the two utilities of changes in the levels represent an extremely robust and dominant of the covariates. Subtracting the left-hand- decision-making unit in relation to side from both sides of the inequalities, production, consumption and market equating the result to a latent variable, Z i , exchange in the types of economies we and permitting the equality to hold with studied. Farm households, therefore, become the natural core units in our models and error, μ , we are left with i analysis. Various development Z pi = xi β p + μip , Z i ≥ 0 if yi = 1, Z i ≤ 0 , interventions, including irrigation otherwise. (1) development, may have changed their Here β ≡ γ − μ measures the difference decision-making environment, however, in p terms of their capacity to produce, access in allocating resources to either enterprise, markets and the prices and price variability i.e. food or cash crop production. they face in these markets. Supply decisions are modeled in a similar In a world with well developed markets, way. We assume that the quantity supplied households will participate in all factor and on the market is a linear function of another commodity markets when these factors are set of household characteristics, which may used in production and commodities are be the same as the set represented by the produced and/or consumed by the covariates xi, above. Specifically, the supply households, as long as factors and relationship is: commodities are imperfect substitutes and Z si = xi β s + μ si , (2) distribution of factors and commodities vary where Z denotes householdi ’s the volume across households. There will always be si gains from trade when trade is costless (zero supplied; xi denotes covariates relevant to transaction costs). Such a world favors specialization. Under such scenario, the supply decision; βs denotes a vector of irrigation development is expected to unknown parameters depicting the promote market oriented production relationship between supply and the

265

regardless of the households’ consumption and education are more likely to participate demand. in markets than households with fewer In the real world there are transaction costs endowments. causing there to be price bands where purchase prices are higher than selling H4. Smallholder farmers are often prices. Significant positive transaction costs disadvantaged due to poor access to and information asymmetries can lead to information and market-supporting services market imperfections (de Janvry et al., such as extension services and credit 1991). For an economy where there are both assistance and these impediments often give sellers and buyers of a factor or commodity, rise to low rates of adoption of improved in the so called two-sided markets, positive technologies that could potentially increase transaction costs and information productivity, diversification and, hence, asymmetries lead to non-participation and a market participation. “self-sufficiency orientation” for factors that are owned and used in production and Hypothesis three implies that poverty may commodities that are produced and limit households’ participation in markets. consumed by households. In general, we Besides, food insecure households may expect that the higher the transaction costs, allocate most of their resources to meet their the wider the price band and the larger share food demands, even if growing for the (%) of households that will be non- market is economically more rewarding. participating. An implication of this is that Hypothesis four implies that availability of market non-participation can be an indicator inputs and new technologies also facilitate of the size of the transaction costs in a market oriented production. In this case, the specific two-sided market. However, the functioning of input markets and extension distribution among households and services play an important role in facilitating substitutability of factors in production and increased adoption of new technologies commodities in consumption within (improved seeds, agronomic practices, etc) households may also influence the degree of by farmers. Adoption of new technologies non-participation. The higher the elasticity plays a critical role in farmers’ increased of substitution in production and market oriented production as technological consumption the higher we expect the change without increased commercialization probability of non-participation to be. seems unlikely because of the increased use Overall, significant market non-participation of purchased inputs and is a sign of significant market imperfections diversification/specialization are inherent in an economy. On the basis of these elements of most technological innovations broader perspectives, we developed some in agricultural production. Hence, policies to testable hypothesis. speed up commercialization and technological change move jointly in a H1. Smallholder producers with access to reinforcing way (von Braun, 1995). Hence, irrigation are more likely to participate in we propose that households with good markets than farmers under rainfed systems. access to services (input and capital markets) are more likely to participate. H2. Smallholder farmers with better access These hypotheses were tested to markets (i.e., close to larger markets) are systematically. The results are reported in expected to be much more likely to the subsequent sections. participate in the market than farmers without access to such infrastructure. 4. Econometric estimation H3. Households with better endowments such as labor, capital (including livestock), land and other resources such as information

266

Let the amount of crops supplied by a lognormal regression models (Wooldridge, household i be given by: 2002). Hence, nested (log-likelihood ratio test) and non-nested Voung test (Voung, ysi = x1β1 + μ1 (3) 1989) model test statistics were derived to y determine whether to use the Tobit model where si is the volume of sales supplied by formulation or either the Cragg or the household that is expected to depend on Wooldridge model. If these test results the vector x1 regressors outlined in equation showed that these were separate decisions, then we used the double hurdle model ysi (2). As is censored this can be estimated (Cragg, 1971) or Probit plus lognormal using variants of censored regression regression models (also known as models. The most often used model is the Wooldridge model) along with other Tobit model (Wooldridge, 2002). explanatory variables to explain volume The participation equation, whether the decisions of households. household decides to participate or not, is The Cragg model has the advantage that it given by: nests the Tobit model and a likelihood ratio test can be performed easily to determine if y = 1[]xδ + v > 0 pi 2 2 (4) the household market supply decision is best modeled by a one-step or a two-step x, y procedure. The difficulty in comparing the where ( pi )are always observed whereas Wooldridge model against the Cragg model y y = 1 si is observed only when pi . Eq. (4) is that they are not nested to each other. The can be estimated using variants of the binary same is true for Tobit model and choice model, in our case we used the Probit Wooldridge model. We used the Voung ()u ,v (1989) non-nested model selection test. model. We assumed that 1 2 is Following, Greene (2000) and Fin and x independent of with mean of zero Schmidt (1984) the restriction imposed by implying that x is exogenous, the Tobit model is tested against the Cragg and v2 ~ N()0,1 . model by performing a likelihood ratio test of the following. One of the assumptions in, and important limitation of, the Tobit model is a single L = 2(ln L + ln L − ln L ) mechanism determines the choice between probit truncatedregression Tobit

y = 0 versus y > 0 and the amount of pi pi …………………………………………(5)

ysi given yi > 0 . However, in reality participation decisions and volume decision where L is distributed as chi-square could be separate, and are influenced by with k degree of freedom ( K is the number different factors. Estimating these decisions of independent variables including a simultaneously while the decisions are constant). The Tobit model was rejected in separate may lead to inconsistent estimates favor of the Cragg model if L exceeded the and wrong conclusions. Alternatives to chi-square critical value. The likelihood censored Tobit have been suggested to allow ratio test statistics of chi2 (37) = 4574.21, p=0.0000, indicated that the restrictions the initial decision of y pi = 0 versus imposed by the Tobit model is rejected in y pi > 0 to be separate from the decision of favor of the Cragg model. Thus, the same household and farm characteristics did not how much ysi given y pi > 0 . These include have equal influence on both the Cragg’s double hurdle model (Probit plus participation decision and the decision for Truncated regression model) (Cragg, 1971) how much to sell. It also implies that the or Wooldridge’s model using Probit plus participation decision and volume decision

267

are not made simultaneously. However, Management Institute (IWMI) with support hypothesizing that a given variable is from the Austrian government. The socio- interrelated with the participation decision economic survey, which investigated the and not with volume decision or vice versa impact of irrigation on poverty and irrigation is difficult. Consequently, the three models contribution to national economy, addressed are estimated with the same variables. a total sample size of 1024 households from Once the Tobit model was rejected, the eight irrigation sites from 4 regional states Cragg model could be compared with involving traditional, modern and rainfed Wooldridge model using Voung’s non- systems (see Fig. 1 and Table 1A). The total nested model specification test. Voung’s sample constitutes 397 households non-nested model specification test is given practicing purely rainfed agriculture and 627 by households (382 modern and 245 V = n −1 2 LR (θˆ ,υˆ ) ωˆ → N(0,1) (6) traditional) practice irrigated agriculture. n n n n These households operate a total of 4,953 ˆ where LRn (θ n ,υˆn ) is the difference plots (a household operating five plots on between the log-likelihood values for the average). Of the total 4,953 plots covered by the survey, 25 percent (1,250 plots) are two models, θˆ andυˆ is the maximum n n under traditional irrigation, 43 percent likelihood estimators from the two models, (2,137 plots) are under modern while the respectively and V is distributed as a remaining 32 percent (1,566 plots) are under standard normal variable. The Voung test rainfed agriculture. The data collected statistic of (V= -27.858, p= 0.000), strongly include demographics, asset holdings, access indicated that the Cragg model dominates to services, plot level production and sale the Wooldridge model. The critical and input use data (distinguished between values (c) for the 1 and 5 percent irrigated and rainfed), constraints to significance level are 2.58 and 1.96, agricultural production and household respectively. Consequently the results perceptions about the impact of irrigation on presented below are derived from the Cragg poverty, environment and health and other model. household and site specific data. The data was collected for the 2005/2006 cropping Finally, we also corrected the standard season. errors for clustering effects by assuming that observations are not independent within the 6. Results and discussion cluster although they are independent between clusters, in this case the household Summary statistics (Rogers, 1993). This is fair assumption as management could vary across households We present a summary of some of the most but not within plots run by the same important variables here (for details see household. Table 1 below). Of the total households surveyed, about 54 percent of the 5. Study site description and data households participated in the market by description selling a product and earning an average of †††††† Birr 591 (SD 2169) . The gross value of This study is part of a comprehensive sales realized by households varies greatly nationwide study on the multiple impacts of irrigation on poverty and environment run †††††† between 2004 and 2007 in Ethiopia. It was a 1 US Dollar (USD) = 8.39625 Ethiopian component of the Impact of Irrigation on Birr (ETB) in May 2006. Poverty and Environment (IIPE) research project run by the International Water

268

as can be seen from the high variance. This The functioning of input markets is expected variation is also stronger between farmers to influence the functioning of output working in different irrigation types. markets through its influence on production. Households in traditional irrigation and in Hence, we wanted to understand whether modern schemes earn an average income of farmers faced any input access problem Birr 699 (SD 2679) and Birr 779 (SD 4090) during the 2005/06 cropping season. respectively from crop sales in contrast to Reporting on their experience of input rainfed Birr 476.10 (1076.1). It seems that access, about 53 percent of the households average gross value of sales from modern responded that they had no input access irrigation schemes is higher than those from problem, compared with 47 percent who traditional schemes. However, testing for indicated that they did. The problems equality of the mean, in sales between the included: high input prices (45.8%), three irrigation types, indicated that there is shortage of capital (high down payment, not a statistically significant difference (p-value member of service cooperatives and lack of 0.0001). However, a separate test for access to credit) (18 %), lack/shortage of traditional and modern scheme indicated supply of inputs (mainly pesticides and that the mean difference is not statistically herbicides but also fertilizer) (16 %), lack of different (p-value 0.5354). This indicates timely supply (10 %), shortage of equipment that there was no difference in mean value and materials and skilled labor to apply of sales between traditional and modern these inputs (2.4), and distance to input schemes, although average sales from both markets and lack of supply locally (1.9%). sources are higher than those obtained from The most important problems are, hence, the rainfed system. high input prices, lack of credit access and lack of availability of inputs in space and When asked about whether the households time. There is a significant difference in the faced any output market and marketing perception of the presence of input related problems about 59 percent responded that problems between farmers working in they did not face any problems while the rainfed systems and under irrigation remaining 41 percent said that they did. systems. On average more farmers under There is a difference in the perception of the rainfed systems seem to face input access presence of a market and marketing problems than those working under problems between farmers working in irrigation. rainfed systems and under irrigation systems. More farmers under irrigation Moisture stress and water shortages could systems seem on average to face market and pose serious constraints to agricultural marketing related problems than those production and, hence, to market supply of working under rainfed systems. The major agricultural outputs. Asked if households problems include: market problem (low faced any shortfall in rain during the demand and low selling price) (30.7%), production season about 61 percent of the distance to market, road and transport respondents indicated they did not, while the problems (23%), same product & peak time remaining 39 percent indicated that they did. supply (20 %), unstable prices (11%), lack Similarly, irrigation farmers asked if they of services (information, service faced water shortage during the irrigation cooperatives, high tax) (5.3%), high season, 73 percent responded that they did purchase prices of agriculture goods when not, while 27 percent of the respondents did. they want to buy them (4.4%), low supply and poor quality (3.3%), exploitation by We present the composition of crops under local traders (1%), and others (additional different irrigation systems. The percentage costs) (0.7%). values indicated the percentage of the plots covered by these crops (Figure 3). The dominant crops under traditional irrigation

269

system, in order of importance are: maize, participating in output markets. These wheat, teff, followed by horticultural crops results show that participation per se is not such as mango, potato, banana and tomato influenced by whether the household works (Figure 3a). In the modern irrigation under irrigation system or not. However, schemes, in the order of importance, the when we disaggregate by crop types, dominant crops are teff, maize, onion, farmers growing irrigated annuals and wheat, tomato, barley and potato (Figure irrigated perennials are more likely to 3b). participate in the market in contrast to In the rainfed agricultural system cereals are farmers that grew rainfed annual crops, with the dominant crops: teff, wheat, maize, marginal effects 0.21 and 0.29 respectively. sorghum, barley and pulses and oil crops. It is believed that this is because the rainfed Horticultural crops such as onion, potato and annual crops tend to be mainly food crops. perennial crops such as mango and gesho The participation of the farmers growing (local hops) each cover less than 1 percent rainfed perennials is found not to be of the total plots (Figure 3c). significantly different from those growing Finally we looked into the nature of market rainfed annuals perhaps indicating the production, i.e. whether households are inherently low scale of cash crop production really exercising shifts in their cropping in the former. Hence, the result strongly choice? Or is it just the produced surplus indicates that irrigation significantly which is supplied to the market? How are contributes to market participation by the quantity of sales and value of sales enabling farmers to grow crops that are correlated? We estimated a simple marketable although rainfed growers also correlation coefficient between quantity and sell crops for various reasons. value of sales. A calculated correlation coefficient of 0.18 indicates that there is low Various household characteristics and linear association between quantity and resource level endowment variables were value of outputs. Therefore, it could be that found to have a significant effect on any farmers are shifting to more valuable households’ decision to participate in the products as the crop composition also attest. market. From among the household characteristics education attainment of the Explaining market participation head of the household and family size were found to be significant in explaining market The results from the Probit regression model participation. The number of years of on factors that determine households’ education of the head was found to be market participation, are reported in Table 2 positively and significantly associated with below. The fitted binary choice model is the households’ decision to participate in the found to explain the observed variation with market implying that educated households the observed probability of 0.60 and are more likely to participate in the market. predicated probability 0.62. We also As education increases by a unit, the estimated the marginal effects for the Probit probability of participation increases by model and these are reported here. about 2 percents. On the other hand family Farmers working under different irrigation size was found to have a negative effect on management schemes may have different market participation indicating that probabilities to participate in the market. households with more family members are Households working in modern irrigation more likely to focus on food production to schemes were found, albeit at 10 percent meet family food requirements. This is level of significance, less likely to typical of economies where food markets are participate in the market compared to not well developed and, hence, households rainfed farmers. Similarly, farmers working choose to first be food-self sufficient, before under the traditional irrigation scheme are they produce for the market. From among found to have not significant difference in the household resource endowments, the

270

size of the operated land area has a positive problems were found to be the most likely and highly significant effect on the decision ones to participate. This may reflect a to participate in markets. A unit increase in reverse causality in that those who area of operated holding leads to a 23% participated in the market ones most likely increase in the likelihood of participation. to face input access problems. Households This result indicates that land holding size producing for the market were about 6 could be an important constraint to market percent more likely to face input access participation even if irrigation access is problems such as untimely availability of ensured. Other resources such as labor (both seeds, seedlings, and chemicals. This result female and male), oxen holding were found was reflected during the rapid appraisal to be insignificant in explaining market study which indicated that farmers had a participation perhaps indicating that these hard time getting vegetable seeds and resources may not pose as significant pesticides. This may call for reorientation of constrains to participation per se in rural the input supply system to meet the Ethiopia. requirements of the irrigation system.

Distance to the market where produce is Community (site) level effects were also sold and type of means of transport had also found to be significant in explaining significant effect on market participation. variations in the probability of participation. Market participation per se increased with These effects could be related to village distance to market where the products are level covariates (such as location of the site, sold. Although this sounds counter intuitive, agro-ecology and crop suitability factors, this may be related to the fact that irrigation experience, weather conditions households who manage to transport to and other external effects) which may distant but larger markets are likely to influence market conditions. So taking benefit from the high price differentials Debre Zeit (Wedecha Belbela systems) as a manifest in fragmented markets. As reference, we found that households in agricultural markets in Ethiopia, as in most Endris (marginal effect -0.20), Golgol Raya, rural economies of the developing world, are Haiba (marginal effect -0.17) and Hare not well developed, price effects are not (marginal effect -0.11) are less likely to easily transferred across locations. This participate in the market while households in implies that farmers need to select markets Golgotha are more likely to participate where their products can fetch good prices (marginal effect 0.28). Both the Wedecha and the incentive to take a product further and Golgotha irrigation schemes are located afield requires market knowledge as a close to the major markets, Addis Ababa and precondition. Below we test this theory by Nazareth, on a well established marketing determining if the value of output increases route for vegetables (see Fig. 1). However, with distance to the market where the output from the results we have here it is difficult is sold. Conversely, this may also suggest to attribute to one factor, e.g. distance to that irrigation schemes are not positioned market, as being the principal factor close to markets. Participation also seems to influencing market participation. It is likely increase with the use of donkeys as a means that the dummy variables confound various of transport in reference to use of human factors. Hence, we can only say that there power. Those who used donkeys are 6 % are site level covariates influencing market more likely to participate in the market participation. compared to those who used human power. Finally, although less expected plot level Access to input markets were also found to characteristics such as slope of the land and have significant effect on market soil quality were found to be significant in participation of households. The households explaining market participation. who reported to have faced input access Accordingly, households operating land

271

with medium (marginal effect 0.07) and may pose a serious constraint to market steep slope (marginal effect 0.08) were development. found to be more likely to participate than In line with the results in the probability those operating flat lands. One possible model, education and family size were also explanation could be that the slope of land found to be significant in explaining the may influence crop choice, so irrigated amount of sale. The education level of the annuals and/or perennials are grown on such head of household was found to be lands. Households operating medium positively and significantly associated with (marginal effect 0.06) and good quality high value of sale, implying that educated lands (marginal effect 0.08), i.e. with more households are more likely to be market productive soils, were found to be more oriented. This may be because they are well likely to participate. The effect of these plot positioned to choose high return crops and characteristics on market participation could introduce innovative technologies. In be through their influence on crop choice contrast to the negative influence of family and productivity. Below we will explore size on explaining market participation, here further if the same set of factors also affect family size was found to have a significant the level of participation, the volume of sale and positive effect on volume of sale. This made by households. suggests that once households have decided to grow for the market, the family size does Explaining volume decisions not negatively influence volume of sale. Furthermore, households’ resource The most important determinants of volume endowments, specifically the size of the decisions (measured by the value of sale) are operated land area and oxen holding, have reported below. But for the truncated model positive and highly significant effects on the we did not report the calculated marginal volume of sale. Farmers usually allocate part effects as the purpose of our analysis is not of their land to grow high value crops after confined to the sub population. Hence we they have allocated sufficient land to grow report the coefficients as indicated in Table food crops. Oxen holding increases the 3. chance of increasing operating land holding Households operating both modern and through informal land transaction such as irrigation schemes supply more to the sharecropping and fixed renting. Therefore, market than farmers working in the rainfed households endowed with more land and system. In line with the results from the oxen holding are more likely to sell more to binary choice model, farmers growing the market than households with smaller irrigated annuals and irrigated perennials land holding and no oxen. supply more to the market in comparison to farmers that grow rainfed annuals because, Distance to market where the output was as indicated above, the rainfed annual crops sold has significant effect on the volume of tend to be mainly food crops. The results sale strengthening our conjecture that here, hence, strongly indicate that irrigation households who are able to participate significantly contributes not only to market transport their produce further but to more participation but also to increased supply of attractive markets. In line with this, the produce to the market. This could be the volume of sale was found to be significantly result of increased cropping intensity and influenced by the choice of transport. In this diversification into more cash crops, mainly case, households who rent vehicles have horticultural crops. Households that higher volumes of sale compared to those reported to have faced shortages in rainfall using human power. Moreover, unlike the supplied significantly lower volumes of result in the Probit model, use of donkeys as produce, and hence, earned less from the a means of transport has a negative effect on market. This indicates that shortfalls in rain, the volume of sale indicating perhaps that higher volume of sale requires other means

272

of transport than pack animals or human influences crop choice, so irrigated seasonal power (e.g. ISUZUs, the famous small or perennials are grown on such lands. trucks which can operate deep in rural Households operating medium and good areas.) quality lands, i.e. with more productive soils, were found to be supplying higher In contrast to the participation decision, volumes of output, underlining that reported market related problems were production enhancing factors have also found to have no significant effect on the market participation enhancing effects. amount of goods sold. This implies that market and marketing related problems may 7. Conclusions and recommendations deter households from participation but once they have made the decision to participate The objective of this study was to examine they supply what they can. However, those the extent and nature of market oriented who reported input access problem were production under irrigated systems in also found to be supplying more produce to contrast to rainfed systems in Ethiopia. The the market. This may reflect, as argued study t identified determinants of market- earlier, a reverse causality in that those who oriented production, including the role of participated in the market are more likely to irrigation in the process, in order to face inputs access problems. It could also be understand the main constraints and related to the location of the irrigation opportunities for market oriented schemes in relation to input supply centers development. Based on the study findings and the orientation of the input supply we have drawn policy implications relating system of the country. Access to off-farm to institutional arrangements and the income was found to have a negative effect creation of adequate infrastructure, which on the value of sale perhaps indicating that could benefit a large mass of smallholder those who have access to off-farm income producers. do not consider it worth the effort of One of the most important findings of this growing for markets. study is that irrigation contributes to a significant increase in market participation, The same community (site) level effects volume of marketed produce and, hence, were also found to be significant in income. Farmers working under irrigation, explaining variations in the volume of sale. traditional or modern, supply more marketed So taking Debre Zeit (Wedecha Belbela produce and earn more income than farmers systems) as a reference, we found that operating under the rainfed system. The bulk households in Haiba, Hare and Tikurit of the contribution comes from irrigated supply low volumes of output while annual and perennial crops, which indicates households in Golgotha and Zengeny supply that farmers are shifting their cropping mix more output (i.e., more valuable), as a result of access to irrigation. Disentangling which specific site level While irrigation enhances marketed oriented variables are important in explaining market production, there are a series of factors that participation is something that needs further pose serious constraints to the process. inquiry. Households having on average relatively larger plots are found to be more market Finally, the same plot level characteristics oriented. This implies that those who have such as slope of the land and soil quality smaller plots on average have access were also found to be significant variables in problems and tend to focus on food explaining volume decisions. Accordingly, production. This is especially true with households operating lands with steep slopes households that have bigger family sizes. were found to supply more than those This calls for policy intervention in the area operating flat lands. One possible of easing land transactions and assisting explanation is that the slope of land

273

household’s to access important productivity The study also found that education has increasing assets such as oxen. market promoting effects in terms of increasing the probability of participation The study also shows while the impact of and volume of sale. Adequate support to market oriented production on income education can, thus, help in the long-term poverty is direct and immediate; households transform traditional subsistence agriculture are faced with a possible trade-off between into more market oriented and modern growing for the market and growing for agriculture. Finally there are unobservable home consumption. Growing for the market site specific effects that influence market may not ensure household food security in a participation and volume decisions. situation where food markets function Identification of the most important village poorly. Under this situation of market failure level effects requires further inquiry. households prefer first to be food self- sufficient and only then become involved in References market production. Another entry point for policy could, therefore, be to create the Bellemare, M. F., and Barret, C.B. (2006). necessary infrastructure and policy An ordered Tobit model of market environment to improve the performance of participation: Evidence from Kenya and food markets. Such measures could induce Ethiopia. American Journal of farmers to be more market oriented. Agricultural Economics, 88(2), 324-337. Bhattarai, M., and Pandy, S. (1997). The Market problems and input access problems economics of wheat production in the seem to be pervasive in Ethiopia, and more rice-wheat system in Nepal. A chapter in so in areas where irrigation-induced market a book series in P.S. Teng, et al. (Eds.). oriented production is high. The study 1997. Application of systems approach indicated that farmers face diverse market at the farm and regional levels 1:45-57. problem such as low demand and low Kuluwer Academic Publishers. IRRI, selling price, distance to market, road and ICASA. Netherlands. transport problem, same product and peak Bingen, J. Serrano, A. Howard, J. (2003). time supply, unstable prices, lack of services Linking farmers to markets: different (information, service cooperatives, etc) and approaches to human capital high tax. Similarly farmers reported that development. Food Policy 28, 405–419. they faced diverse input access problems the Cragg, J. G. (1971). Some statistical models most important of which were high input for limited dependent variables with prices, lack of credit access and lack of application to demand for durable availability of inputs in all seasons and sites. goods. Econometrica 39, 829-844. Transport problems seem to pose a serious de Janvry, A., Fafchamps, M. and Sadoulet, problem as well. Households who are able to E. (1991). Peasant Household Behaviour rent vehicles supply more to the market. with Missing Markets: Some Paradoxes Those unable to transport their produce are Explained. Economic Journal 101, unable to reap the benefits of better markets. 1400-1417. The implication of this evidence is that Fin, T. and P. Schmidt (1984). A test of the irrigation development and market Tobit specification against an alternative infrastructure development are poorly suggested by Cragg. Review of linked. Hence, there is a need to link Economics and Statistic, 66, 174-177. irrigation development with road Greene, W. (2000). Econometric Analysis. infrastructure development and 4th edition. Macmillan. New York improvements in other marketing services. Goetz, S. J. (1992). A selectivity model of There is also a need for reorientation of the household food marketing behavior in input supply system to fit the requirements Sub-Saharan Africa. American Journal of the irrigation system.

274

of agricultural Economics 74(2), 444- 452. Huang, Q., Rozelle, S. Lohmar, B. Huang, Rogers, W.H., (1993). Regression standard and J, Wang, J. (2006). Irrigation, errors in clustered samples. STATA agricultural performance and poverty Technical Bulletin 13, 19-23. STATA reduction in China. Food policy 31, 30- Corporation. 52. Hussain, I. (2005). Pro-poor intervention Rosegrant, W. Mark, Renato Gazmuri strategies in irrigated agriculture in Schleyer and Satya N. Yadav (1995). Asia. Poverty in irrigated agriculture: Water policy for efficient agricultural Realities, issues, and options with diversification: market-based guidelines. International Water Approaches. Food Policy 20(3), 203- Management Institute. Final synthesis 223. report. von Braun, J. (1995). Agricultural Hussain and Hanjira (2003). Does irrigation commercialization: Impact on income water matter for rural poverty and nutrition and implications for alleviation? Evidence from South and policy. Food Policy 20(3), 187-202. South-East Asia. Water policy 5, 429- Wooldridge, M.J., (2002). Econometric 442. analysis of cross section and panel data. Hussain I. and Hanjra, A. (2004) Irrigation MIT, Cambridge, Massachusetts, 752 and poverty alleviation: Review of the pp. empirical evidence. Irrigation and World Bank (2006). Managing water Drainage 53, 1-15. sources to maximize sustainable growth: Lapar, M.L., Holloway, G., Ehui, S. (2003). A World Bank Water Resource Policy options promoting market Assistance Strategy for Ethiopia. participation among smallholder Agriculture and Rural Development livestock producers: a case study from Department. World Bank. 91 pp. the Philippines. Food Policy 28, 187– 211

Ministry of Finance and Economic Development (MOFED). (2006). Ethiopia: Building on Progress. A Plan for Accelerated and Sustained Development to End Poverty (MOFED). (2005/06-2009/10). Volume I: Main Text. Ministry of Finance and Economic Development (MoFED). September, 2006. Addis Ababa. 229 pp. Pandey, V.K. and Sharma, K.C. (1996). Crop diversification and self sufficiency in food grains. Indian Journal of Agricultural Economics 51 (4), 644-651.

275

Fig. 1: Sample sites

276

Table 1A: Study sites Region Scheme name Typology Closer town Irrigation type Sample size Oromiya Endris System Small Ambo Modern 55 Traditional 55 Dryland 55 Oromiya Wedecha- Medium Debre Zeit Modern 55 Belbella System Traditional 53 Dryland 57 Oromiya Golgotha Medium Nazareth Modern 55 Dryland 55 Amhara Zengeny Medium Gimjabet Modern 55 Dryland 53 Amhara Tikurit Small Bahir Dar Traditional 83 Dryland 47 Tigray Haiba Medium Samre/ Mekelle Modern 54 Dryland 54 Tigray Golgol Raya Micro- Alamata Modern 53 irrigation Dryland 46 SNNPR Hare Medium Arba Minch Modern 55 Traditional 54 Dryland 55

277

Table 1: Summary statistics

Variable name Mean (Standard deviation in parenthesis) Overall Rainfed Irrigated Age of household head, years (n= 4915) 46 (15) 45.0 (15.09) 46.29 (15.05) Years of education of household head 2.0 (3.1) 2.05 (3.06) 1.98 (3.07) (n= 4900) Family size, (n= 4948) 5.7 (2.4) 5.54 (2.35) 5.87 (2.48) No. of Female adults (n= 4948) 1.4 (0.85) 1.34 (0.74) 1.39 (0.89) No. of male adults (n= 4948) 1.5 (1.0) 1.45 (0.94) 1.58 (1.07) Amount of income from non-farm, Birr 537 (3067) 705.54 (4125.69) 459.60 (2422.15) (n= 4948) Remittances, Birr (n= 4948) 243 (1973) 27.38 (646.05) 342.89 (2339.14) Number of oxen (n= 4923) 1.4 (1.2) 1.56 (1.07) 1.32 (1.21) Number of donkeys (n= 4923) 0.5 (0.9) 0.65 (0.94) 0.50 (0.86) Number of contacts of household with 1.6 (3.2) 2.36 (3.97) 1.25 ( 2.73) extension agent (n= 4948) Number of contacts of extension agent 2.5 (5.3) 3.79 (6.38) 1.89 (4.59) with households (n= 4948) Land area, ha (n= 4786) 1.4 (1.2) 1.34 (1.39) 1.42 ( 1.19) Distance to market where output was 7.6 (6.9) 8.36 (7.49) 7.18 (6.67) sold, km (n= 4947) Gross value of Sales, birr (n= 4948) 591 (2169) 476.10 (1076.1) 645.14 (518.65) Market problem Dummy (yes =1) (n= 40.7 37.7 42.1 4953) Input access problem Dummy (yes =1) 46.6 53.3 43.4 (n= 4953) Rain/water shortage Dummy (yes =1) 39.1 26.8 (n= 4953)

278

% of plots % of plots

20 25 18 16 20 14 12 15 10 8 10 6 4 5 2 0 0 f t ef t ff ze a t to a to e r on e at n y u g he go ta an a pe te to e ato ge m na a n mai w po m Ens ane oni pea a a e man ban to pep maiz whe onio om barl pot b t ab bana ck p chick c sorghu i sugar c ch

Fig. 3a: Dominant crops under traditional irrigation system (n= 1240) Fig. 3b: Dominant crops under modern irrigation system (n= 2092)

% of plots

30 25 20 15 10 5 0 f t f a y g s s o n te e ize um e an u ru g io h h nu e y n arl b ypt on w ma b l th gesho ma potato sorg ca la u e

Fig. 3c: Dominant crops under modern rainfed system (n= 1533)

279

Table 2: Determinants of the probability of participation (standard errors adjusted for cluster effects)

Dependent variable : Whether the household sells a product to the market or not (0/1) Variable description Coefficient Standard errors Female headed household (dummy variable male =0) 0.022 0.078 Age of the household head 0.001 0.002 Education level of head 0.017 .010* family size -0.026 0.016* Female adult -0.050 0.037 Male adult 0.038 0.032 Off-farm income 1.02e-06 7.69e-06 Remittance income 0.0001 0.00003 Oxen holding -0.013 0.025 Distance to the market where output is sold (in km) 0.016 0.005*** Means of transport (donkey) (reference= human) 0.151 0.090* Means of transport (horse) (reference= human) 0.228 0.171 Means of transport (mule) (reference= human) 0.115 0.199 Means of transport (vehicle) (reference= human) 0.059 0.118 Household’s contact with extension agent 0.003 0.011 Land area (in ha) 0.062 0.023*** Rain shortage (dummy 1= yes) 0.026 0.074 Irrigation water shortage (dummy 1= yes) -0.021 0.085 traditional scheme (dummy reference =rainfed) -0.050 0.084 Modern scheme (dummy reference =rainfed) -0.278 0.075*** Input access problem (dummy 1= yes) 0.171 0.065*** Marketing problem (dummy 1= yes) 0.042 0.059 Dry land perennial (reference dry land seasonal) 0.109 0.120 Irrigated seasonal (reference dry land seasonal) 0.582 0.070*** Irrigated perennial (reference dry land seasonal) 0.996 0.158*** Endris irrigation scheme (reference= Debere Zeit) -0.506 0.095*** Golgol Raya irrigation scheme (reference= Debere Zeit) -0.448 0.136*** Golgota irrigation scheme (reference= Debere Zeit) 0.921 0.192*** Haiba irrigation scheme (reference= Debere Zeit) -0.441 0.112*** Hare irrigation scheme (reference= Debere Zeit) -0.281 0.152* Tikurit irrigation scheme (reference= Debere Zeit) -0.020 0.128 * Zenegeny irrigation scheme (reference= Debere Zeit ) -0.038 0.184 Medium Slope (dummy reference= flat) | 0.190 0.0622*** Steep slope (dummy reference= flat) 0.215 0.102** Medium fertility (dummy reference= poor) 0.165 0.077** good fertility (dummy reference= poor) | 0.228 0.078*** _cons | -0.194 0.176 Number of obs = 3754 Wald chi2(36) = 300.17 Prob > chi2 = 0.0000 Log pseudo-likelihood = -2276.33 Pseudo R2 = 0.0969 *, **, *** significant at 10, 5 and 1 percent level of significance.

280

Table 3: Level of participation (Value of sale) Dependent variable: Value of sale (in Birr) Variable description Coefficient Standard errors Female headed household (dummy variable male =0) -3679.5 (473.846) Age of the household head 37.04 (79.00) Education level of head 2001.188 (486.2001)*** family size 2831.62 (689.7989 )*** Female adult -4731.888 (1960.598)** Male adult -2636.161 (1702.532) Off-farm income -1.124729 (.3445668)*** Remittance income -.5153223 (.7632368) Oxen holding -2780.298 (1304.623)** Distance to market 160.1715 (107.3297) Means of transport (donkey) (reference= human) -23025.99 6164.656*** Means of transport (horse) (reference= human) -2342.406 6569.129 Means of transport (mule) (reference= human) 7528.12 8567.152 Means of transport (vehicle) (reference= human) 19583.99 5675.037*** Household’s contact with extension agent -1264.534 (688.0457)* Land area 6722.116 (728.9046)*** Rain shortage (dummy 1= yes) -7626.328 (4227.934)* Irrigation water shortage (dummy 1= yes) -3272.534 (4214.353) traditional scheme (dummy reference =rainfed) 20030.93 (5983.328)*** Modern scheme (dummy reference =rain fed) 17768.58 (5768.995)*** Input access problem (dummy 1= yes) 13467.66 (3457.655)*** Marketing problem -4479.403 (3194.229) Dry land perennial (reference dry land seasonal) -11678.92 (10525.31) irrigated seasonal (reference dry land seasonal) 17526.23 (4057.36)*** irrigated perennial (reference dry land seasonal) 24931.89 (7034.23)*** Endris irrigation scheme (reference= deberezeit) -5600.18 (5543.335) Golgol Raya irrigation scheme (reference= deberezeit) -9191.548 (6577.918) Golgota irrigation (reference= deberezeit) 11853.08 (6385.049) * Haiba irrigation scheme (reference= deberezeit) -197758 (24293.52)*** Hare irrigation scheme (reference= deberezeit) -47682.18 (11743.92)*** Tikurit irrigation scheme (reference= deberezeit) -28460.84 (7124.578)*** Zenegeny irrigation scheme (reference= deberezeit) 39782.97 (9889.322)*** Medium Slope (dummy reference= flat) | 2169.702) (3414.557 Steep slope (dummy reference= flat) 26719.8 (6862.004)*** Medium fertility (dummy reference= poor) 11340.24 (6171.729)* good fertility (dummy reference= poor) | 12887.67 (6265.083)** _cons -106551.5 (13938.39)*** sigma _cons | 8990.92 (327.1781)*** Number of obs = 4610 (2086 left-censored observations at gvout<=0 2524 uncensored observations) LR chi2(31) = 294.28 Prob > chi2 = 0.0000 Log likelihood = -32017.082 Pseudo R2 = 0.0046 *, **, *** significant at 10, 5 and 1 percent level of significance.

281

Small Scale Irrigation Interventions for System Productivity and Natural Resource Management in Ethiopian Highlands: Benefits and Best-bets

Tilahun Amede1, Ayele Gebre-Mariam2 and Fabrizio Felloni3

1International Water Management Institute (IWMI)/ International Livestock Research Institute (ILRI), Addis Ababa, Ethiopia 2Private consultant, Addis Ababa, Ethiopia 3Evaluation office, IFAD, Rome [email protected]

Abstract visible with horticultural crops. There has been also a shift towards improved varieties Water scarcity became a common with access to irrigation. Farmers replaced phenomenon in Ethiopia with drought early maturing but low yielding varieties frequency of at least once in three years with high yielding varieties. Crop while the country owns a large irrigation diversification increased significantly, in potential that should be exploited some sites from three to about 15 species, sustainably. Various national and although this decision making process did international institutions are currently not favour legumes. The apparent effect was engaged in developing small scale irrigation on crop rotation, intercropping and land (SSI) schemes for poverty alleviation. A management with in the order of 79, 42 and monitoring and evaluation exercise was 35%, respectively. On the other hand, there conducted in 2004 and in 2006 in four is a decline in number of livestock per administrative regions of Ethiopia, namely household, but an increased number of Tigray, Southern regions, Oromia and draught oxen. The decline is associated with Amhara, to assess the benefits and reduced grazing area due to conversion of associated environmental effects of SSI dry season fallow to vegetable fields and an investments of the International Fund for increase in area enclosure in the sloppy Agricultural Development (IFAD). A landscapes. The shift from cereal to combination of participatory M&E tools vegetable-dominated cropping increased the namely, individual interviews, group competition for water between downstream discussions, key informants, review of and upstream users and between resource- relevant documents and field observations rich and poor farmers. The impact of were used. The mission was supported by an irrigation schemes should be evaluated in depth pre-mission socio-economic survey better on long term benefits than short term in three representative irrigation schemes. fixes, as farmers initiated long term Data from the sites indicated that 50 % of investments like planting perennial fruits, the respondents had improved food security bought calves and other retail trade and higher income, while 26% of the investments. The communities would respondents did not see any change on their benefit most from further integration of livelihoods. Crop yield under irrigation was livestock into the schemes by adopting feed by 35% to 200% higher than under rain fed sourcing strategies for dairy and fattening. conditions, with much higher benefit The paper also presented best-bets for obtained from high potential areas and in improved irrigation management in farms where external inputs (fertilizer, Ethiopia. improved seeds and pesticides) are accessible. The positive effect was more Introduction

282

stated in its Poverty Reduction Strategy Water, land and finance are becoming the Paper (PRSP) emphasized the importance of scarcest resources in the agricultural system improved water resource development and of Ethiopia, scarcity being severe in regions its utilization to achieve food security where population pressure is high, access to through enhanced use of small scale market infrastructure is low and irrigation. Since the early 1990s’ the federal environmental calamities are frequent. and regional governments of Ethiopia, with Inefficient water management in the rainfed financial assistance from donors, have been agriculture coupled with accelerated land attempting to upgrade traditional small scale degradation plays an important role in schemes, built small scale dams, diversions aggravating the recurrent food insecurity in and water harvesting ponds to respond to the country. In the recent years, drought these environmental calamities. However, became a common phenomenon, happening the performance of the irrigation systems in any part of the country at any time of the has been poor. There exists a substantial year, with a frequency of at least once in yield gap in irrigated farms between three years. Four different drought scenarios achievable and actual yield both in terms of were identified in the mixed crop-livestock yield per unit of land but also yield per unit systems of Ethiopia namely, terminal of water depleted. drought, intermittent drought, foreseeable drought and definite drought (Amede, et al., Moreover, there exists conflicting reports 2004a). In situations where agricultural regarding the agricultural benefits of small production is operating under these various scale irrigation and its impact on natural drought scenarios, with annual rain fall resource management (Tafesse, 2003; IFAD, variability of 40 to 50%, supplementary 2003; Kijne, FAO; Ersado, 2005). In irrigation became a necessity for food some sites, small scale irrigation has production, particularly for intensifying significantly increased crop yield, and systems through high yielding and input households using irrigation have higher responsive varieties and breeds. Currently, agricultural production than non- the growth in food production in Ethiopia is intervention communities (Ersado, 2005). In primarily due to expansion of agricultural others, small scale irrigation didn’t bring land while production pre unit of investment significant increase in crop yield and remained stagnant. livestock productivity directly but increased yield by about 26& by promoting increased On the other hand, Ethiopia owns a wide use of improved seeds and fertilizer (Pender range of irrigation opportunities with about and Gebremedhin, 2004). These differences 9.85 million ha of potentially irrigable on impact of irrigation on agricultural arable land, while only 3 to 5% of the productivity may have appeared because of potential is currently under irrigation (WCD, the fact that small scale irrigation for food 2000) accounting for approximately 3 per security is more than just technologies; but cent of total food crop production. Current comprises production, marketing, credit, yield from rain-fed land is only about 50% social, policy and institutional issues of the irrigated land, given all other inputs (Tafesse, 2003). It could be also because of remain the same, thanks to the recurrent differences in methodologies. In general, drought and limited adoption of water irrigation farming is expected to reduce management practices. If the country is to farmers’ exposure to variability in crop and achieve its stated aims of food self- livestock yields and therefore improve food sufficiency and food security, the current insecurity, especially in the more remote, production shortfalls call for drastic disadvantaged and poorer areas; to raise measures to improve production efficiency agricultural production and rural incomes of both irrigated and rain-fed agriculture. In where crop diversification and market- response, the government of Ethiopia as oriented agriculture can be promoted; and to

283

enhance the capacity of communities to perceived performance), but nevertheless demand for better services but also to test, some compromises had to be made due to modify and adopt improved technologies. accessibility. In Oromia and SNNPR the opportunity was taken to visit older and new The objectives of this paper are i) to schemes while in Tigray schemes were quantify the negative and positive effects of selected covering a range of remoteness small scale irrigation on small holder from Mekelle. farmers and their systems and ; ii) to identify the biophysical and socio-economic factors The studied schemes have been selected affecting the performance of small scale through a two-stage procedure. First a list irrigation schemes . was compiled of schemes constructed under SCP II and operated for at least three years Materials and Methods (i.e. the more “mature” sites for impact assessment). Next one scheme in each of Sites and regions SNNPR, Oromia and Tigray has been extracted by simple random sampling. Four major administrative regions of Ethiopia, namely Amhara, Oromia, Southern Data Collection Regions and Tigray were considered for the study. These areas are characterized by food The quantitative and qualitative date about insecurity, drought prone and very high the impact of small scale irrigation on human and animal population. Based on system productivity and food security was altitude, annual rain fall and average collected in two rounds; between 13th temperature, these sites fall into two of the September and 14th October 2004 and from major traditional agroecological zones 12 November to 2 December 2006, as part namely; i) Weinadega: mid highlands of an IFAD field evaluation mission. The between 1 500 and 2 300 mts above sea study team started the mission by level with wheat, teff, barley, maize, interviewing and discussing with IFAD sorghum, faba beans and chickpeas as stakeholders at different hierarchies starting predominant crops while cattle, donekeys from the federal ministry of water resources and small ruminants are the predominant in Addis Ababa down to the scheme site animals and; ii) Dega: highlands between team members. The data collection 2 300 and 3 200 meters with barely, wheat, considers interviews of farmers, community oilseeds, and lentils as predominant crops leaders, extension agents, key informants, and sheep and cattle being dominant district subject matter specialists, bureau livestock enterprises. heads and federal authorities. We used PRA tools including transect walks, community During the field work a total of 16 of the group discussions, sample measurements IFAD SCP II irrigation schemes, those and secondary data from actors at all levels. established in 1988?? in Tigray, 5 in The mission was supported by a pre-mission Oromia, and 2 in SNNPR were considered. socio-economic survey, which was carried In addition discussions were held in all four out to obtain in-depth understanding of three project regions, and with a total of 14 small scale irrigation schemes (Hizaeti Afras project woredas. Most of these woredas are in Tigrai, Nadhi Gelansedi in Oromia and considered by the Disaster Prevention and Dobena in Southern regions) in the period Protection Commission (DPPC) as food June-September (IE Preliminary survey, insecure (Fig 1). Since time did not permit 2004). Prior to the field trips a checklist was visits to all 58 schemes of the project, a prepared considering relevant agronomic, sampling approach was taken. A variety of natural resource management and livelihood scheme characteristics was identified indicators and considered during data (including distance from roads/markets; age; collection from the respective sites,

284

communities and stakeholders. The pre- irrigation engineers through a pilot survey in mission survey also considered three assessing various water resources, including information sources namely literature perennials rivers. The need for upgrading review, focus group discussion for the traditional irrigation schemes arose from communities and questionnaire for the fact that traditional schemes were interviewing selected households. A PRA underperforming despite years of experience tool has been administered for focus group of communities in irrigation management. discussions with WUAs consisting of men The results of the interviews indicated that and women. The focus group participants the reasons why communities demanded for were 15-20, with 5-6 women per meeting. A upgrading traditional schemes were 1) prepared checklist was used as a guide for continual distraction of canals by erosion, the exercise. A questionnaire was prepared siltation, animals and human activities; 2) for a household survey and 84 farmers in extremely high labor demand to maintain Dobena and 155 farmers in Nadhi Gelan and clean traditional irrigation canals; 3) Sadi and 136 farmers in Hizaeti Afras were considerable loss of irrigation water from interviewed. The interview was carried out the traditional canals due to seepage and by random walk from homestead to easy destruction of canals by heavy rains; 4) homestead: adult heads of household with the difficulties faced by the respective irrigated land were interviewed. In the case communities to build permanent crossovers of Dobena, all households were interviewed, over gullies to reach communities on the as the scheme users were fewer than the other sides of gullies and river sides, as it planned sample (100 to 150 respondents per was observed in Mumicha in Oromia; 5) scheme). most of the traditional command areas were those on flat lands and valley bottoms, but it During the main mission, a combination of excluded fields even with very slight slopes five participatory M&E tools were used to and; 6) there was very limited institutional assemble information and data namely, support in accessing marketable enterprises, individual interviews, group discussions, technological innovations, and integrated key informant interviews, review of relevant extension services to crop, livestock, land documents and field observations. On site, and water management. the information gathering techniques included quantitative (through structured 1. Impact on system productivity questionnaires) and qualitative (through semi-structured interviews) methods with The impact of irrigation schemes on system both individual informants and groups, and productivity could be seen from the observations combined with discussions. perspective of its effect on crop, livestock Secondary data and information from and labour productivity in the respective woreda, regional and federal institutions that sites. Though the time period was very short have had a stake in respective projects were to evaluate the impact of irrigation on the also carefully studied. productivity of some of the systems, as the schemes were 2 to 10 years old, there are Results and Discussion variable results emerging from different sites and regions. Scheme development The project sites were identified for SSI development either because the respective communities contacted the woreda officials through their local representatives (e.g. 1.1 Crop yield Nazre in Tigrai) demanding the development of irrigation schemes. In other In average, crop yield under irrigation was at sites, the schemes were identified by least by 35% higher compared to non-

285

irrigated farms (Table 1), with benefits and improved water management skills. being much higher in high potential areas Moreover, the survey conducted in the three and in farms where external inputs representative schemes indicated differences (fertilizer, improved seeds and pesticides) in crop yield among sites. For instance, in were accessible. Maize was one of the most Dobena, the yield of tomato, potato preferred crops farmers have been producing cabbages and onions were 102, 94, 82 and under irrigation across sites, accompanied 44 qt/ha while in Nadhi Gelan Sadi, it was by vegetables. Farmers have indicated a 35, 37, 24 and 38 qt/ha, respectively (data significant yield increase, particularly in not presented), which could be explained by drought sensitive crops like maize (Tables 1 difference in agricultural potential of the and 2), which otherwise could completely sites, particularly due to soil fertility and fail if the terminal or intermittent drought rain fall amount and distribution. In general, coincides with the drought sensitive stage of an increase in crop yield was accounted not the crop i.e. flowering period. Most farmers only to improved access to irrigation but with market access produced green maize also to associated services in extension and (about 20% of the area across Tigray) input delivery. Moreover, the current (COSART, 2001) with high stover quality distribution of water by water masters for fodder. A case study in Tigray showed followed the principle of rotational irrigation that in good years a farmer obtained about 6 for priority crops known to the majority and and 3.5 tonnes ha-1 of maize under irrigated exceptionalities are established only after and rain-fed conditions, respectively, with strong negotiation. For instance in Nazre, about 60% yield advantage. A comparable the order of priority for getting access to amount of yield advantage was displayed by water was faba beans, tomato, pepper, a farmer in Burka Woldiya, Oromia. Across onions, and spices. This form of bylaws may sites, crop yield under irrigation was by 35% limit farmer innovation and responsiveness to 200% higher than under rain fed of individuals to market demands. conditions (Tables 1 and 2), with much higher benefit in high potential areas and in They consumed 71% of the cereals, pulses farms where external inputs (fertilizer, and oil seeds, 26% of the vegetables and improved seeds and pesticides) are about 2% of the fruits while they sold the accessible. rest to generate cash income. An economic analysis done with 10 representative heads In drought prone sites, farmers replaced the of households indicated that their total gross early maturing but low yielding variety (e.g. earnings from the sales of these products Katumani) by a high yielding maize cultivar was EB 22,602. Their net cash income per (e.g. Awassa 511) thanks to access to household after deducting costs was EB irrigation and obtained a 200% grain yield 1,141. In addition each household retained increase (BOA, Wukro Woreda, 09/2004). A produce to a value of EB 1,181 which they similar trend was obtained with wheat and used for home consumption (IFAD, 2004). faba beans (Table 2). Farmers in Nazre (Tigray) indicated that they got up to 4x In few schemes (e.g. Belessa), irrigation had more onion yield today because of the no significant effect on crop yield for combination of access to irrigation, good various reasons. In some it was because of varieties, access to pesticides and better shortage of water at the critical crop stages, extension support. Crop yield increase was while in others it was because of poor substantial particularly for horticultural agronomic practices related to very low crops not only because of improved access population density, late weeding, lack of fertilizer application and absence of pest to irrigation but also associated trainings and control. improved flow of information in pest management, organic manure application 1.2 Crop diversification

286

became unattainable, as it was the case in Traditionally, most of the sites were cereal Dobena, Southern region. In the current growers, except for Chat in Harer and rain-fed cropping systems cereals and Onions and Pepper in few sites near to major legumes are grown in rotation, while in the roads. The primary cereal crops grown irrigable fields with vegetables and fruit under supplementary irrigation were maize, trees are grown primarily in rotation with wheat, and barley. With the development of cereals. This cropping practice limited the irrigation schemes farmers have shifted possibility of integrating nitrogen fixing, soil towards growing diverse crop, in some sites improving legumes like faba bean in the up to 10 new marketable crops, irrigation systems. It is partly because of the predominantly vegetables. The bureau of small land size of irrigable plots (about 0.20 agriculture played a key role in the ha per household) and partly due to market introduction of new crops through preference for selected crops. In situation establishing multiple demonstration plots in where crop rotation was not practiced the farmers’ fields. Access to irrigation and risk to deplete the soil in a very short time opening up of new market opportunities and the possibility of pest incidence is encouraged farmers to systematically obvious. For example, growing potato and allocate their land to various enterprises, tomato rotatively on the same land, without both in rain fed and irrigable fields. For a break crop, may create a favorable ground instance in Lalay agula, Tigray, farmers for pests like potato late blight that would allocate 30, 31, 25, 10 and 3% of their land make it difficult to grow both crops next to wheat, maize, pepper, barley, and teff, time. Hence, crop rotation as a component respectively under rainfed conditions, while of integrated pest and soil fertility under irrigation the land allocation was 35, management should be sought as it was also 10, 50, 2 and 1.5% for pepper, onions, a concern shared by practicing farmers maize, tomato and cabbage, respectively across sites. (COSART, 2001). Besides, almost all varieties of crops grown in the sites at the Despite the above mentioned concerns, time of the mission were improved varieties interviewees indicated that irrigation came along with the diversion schemes via brought in considerable changes on the the Woreda bureaus. For instance in Lalay farming system through improved crop agula, there has been a shift from the local rotation (cereals in the main rainy season maize variety Birhu to improved varieties, and vegetable in the off season using Katumani and Awassa 511 due to better irrigation), intercropping and improved land access to irrigation water. Moreover, the management (particularly terracing and use diversion of the canals helped the extension of organic manure) within the order of 79, to easily establish seed multiplication sites 42 and 35%, respectively. In the three for non-traditional vegetables and spices, sample sites the farmers who practiced mainly onions, potato and tomato and hence improved agronomic management of crops it facilitated the growing of non-traditional across the various practices after irrigation crops in the area (data not presented). scheme was developed were 22.2, 41 and 36% in Dobena, Geland Sedi and Hizaeti Cropping Sequence and Management Afras, respectively (Table 6).

Access to irrigation in most sites created an opportunity for double and in few cases for Livestock systems triple cropping. However, due to the decreasing Feed shortage was apparent short before the water availability and an overwhelmingly main rains, between the months of April and increasing demand for water by downstream June across sites. The decline in forage users, the possibility to expand irrigable land availability was associated partly to

287

conversion of dry season fallow to vegetable prevalence of new pests and diseases in the fields and an increase in area enclosure of system, excessive soil erosion and in some the traditional grazing areas in the sloppy cases soil salinity. landscapes, accompanied by frequent drought. Contrary to earlier reports, there is The majority of farmers across the regions a decline in the number of livestock per have been introduced to use of inorganic household as a consequence of introduction fertilizers only very recently, mainly through of schemes across the regions regardless of the extension systems. The potential effect agro ecology, but again there are also an of fertilizers on crop yield and farm income increased number of draught oxen (data not is much better understood than 10 years a presented). Other research findings also go, though this capacity building process reported that increased irrigation was may not have been explicitly associated with associated with a reduction in ownership of the irrigation projects. However, only about livestock but with increased adoption of 55% of the farmers across the sites 55% use technologies that enhanced productivity fertilizers (mainly DAP), particularly for (Benin etal. 2003).With access to irrigation maize (data not presented). The government semi-pastoralist communities (e.g. was their major input source for chemical Gedemmso, Oromia) have been converted to fertilizer, mainly through credit a crop-livestock system with significant arrangements. For instance, in Gereb koky, reduction in stock. Tigray, the typical use of inorganic fertilizers in the irrigated fields was about 50 In theory, the expansion of the irrigable area kg/ha, applied mainly to high value should have allowed farmers to produce vegetables and green maize. On the other more biomass all year round, partly as crop hand, perennial crops like coffee and chat residues and grasses on strips, borders and did not receive any inorganic fertilizer hilly patches. However, the biomass across the visited sites, partly because they produced from the vegetable fields was are commonly grown in fertile homesteads. rarely used as feed source as the livestock The increased use of irrigation and fertilizer, rejected to it unless there were no other feed however, did not attract much use of options in the system. Moreover, some improved seeds because either they were farmers complained to the mission that the unavailable or unaffordable to farmers. The local authorities did not allow them to grow local extension agents were multiplying forages using irrigation as the current seeds of only selected crops, mainly maize, bylaws established by the water user potato and onions. Hence most farmers in associations gave priority mainly to food, the sites still use their own seed (Table 3). fruit and vegetable crops. These bylaws may The source of seeds in three schemes was limit innovations in promoting livestock 55, 23 and 22% for own seeds, government enterprises (e.g. Dairy and fattening). With sources and purchased, respectively (IE increased vegetable production from Preliminary survey, 2004). While seeds for irrigated plots and subsequent income some cereal crops, vegetable and coffee are farmers, e.g. Amhara region, afforded to secured from multiple sources, seedlings for upgrade part of their stock with fewer but fruit trees are commonly purchased from the more productive breeds and in the process local market. released part of their crop land for pasture development (Benin, et al., 2003). Similarly, recent reports from the region (G. Medhin etal. 2003) indicated that there is Seed and fertilizer use decreased use of chemical fertilizer with irrigation after the credit service for According to farmers view across the sites, fertilizers was abandoned. Credit and decline in land productivity was strongly financial services are not yet sufficiently associated with soil fertility decline, addressing the needs of irrigation users to

288

move into sizable market oriented and in areas where there was no access to enterprises. As an effect, there is a shift irrigation earlier (Benin etal. 2003). Data towards use and management of organic from three sites also indicated that 34, 26 fertilizers namely manure, compost and crop and 16% of the respondents had access to byproducts in most sites, with an increasing more food, no change or obtained increased number of practitioners adopting these income, respectively (IE Preliminary survey, practices (IE Preliminary survey, 2004) For 2004). In some cases, vegetables became instance in Burka Woldiyaa six out of seven part of the daily dish of farmers (e.g. randomly interviewed farmers practiced Chelekot in Tigray). This should have a composting for the last two years. A farmer positive effect on household health, in Nazre, who conducted an informal particularly through the integration of experimentation described that on a 200 m2 calorie, vitamin and micronutrient rich farm land he got a maize grain yield of 5, 4 vegetables and fruit crops (Amede et al., and 1 qt from organic fertilizer, inorganic 2004b). However, food security has not yet fertilizer and no fertilizer, respectively been fully achieved in almost all sites due to (personal communication). the small land holdings, low soil fertility status and other calamities. However, 83% Impact on household food security of the interviewed farmers still consider lack of enough irrigation water responsible for The major effect of the irrigation projects on low crop yield in irrigated crops (Table 4). the communities was through the attitudinal change that they could produce for the Impact on natural resource management market and could buy their food from the income they generate on farm. This is not There were both negative and positive yet a common knowledge in the rain-fed impacts of the irrigation projects on the agricultural systems of Ethiopian Highlands. environment. The major negative impact was done during the construction phase Farmer interviews across the regions in the whereby new farm gullies were created and high potential areas (with at least 5 or more debris from the construction plots were months long growing season) and low placed on farmlands. potential areas (with four or less months long growing season) (Engida, 2001), Soil Conservation and land rehabilitation revealed that crop yield of cereals (e.g. maize) increased by about 70 and 20%, There are differences among regions, the respectively, not only due to access to longest physical structure being made in supplementary irrigation but also due to Tigray and the lowest in Oromia. In the increased support of government institutions selected sites between 21 and 54% of the in extension and input delivery. In general, households indicated that small scale crop diversification has significantly irrigation attracted soil and water increased, in some cases from only three conservation practices (Table 5). The crops before the construction of the scheme difference in performance was dictated by up to 15 crop species, encompassing various the historical view and understanding the vegetable and high value crops (e.g. status of land degradation in the regions and Gedemmso in Oromia). With access to the subsequent regional policies. In situation irrigation, intercropping and relay cropping where extensive soil conservation was made are also becoming common practices even in in the 1980s (e.g. Burke Woldiyaa and monocropping-dominated systems (Table 5). Mumicha in Oromia), erosion and runoff About 40% of the farmers produced more was considerably reduced with very limited food than before the scheme was active siltation seen on the valley bottoms constructed, particularly apparent in the and diversion canals (Personal drought-prone environments (e.g. Amhara) communication, 2004). In some regions (e.g.

289

Tigray), a considerable amount of work in considerably. It was valued by the managing the upper slope of the schemes communities as a means to recharge the was done through the construction of soil springs, as sources of bee forage, and also as conservation terraces and tree bund a strategy to restore indigenous trees like stabilizers, which was a necessity to sustain Dodonaea viscose and Olea africana. In the future performance of the schemes. This some cases farmers were allowed to graze was done by introducing regional policies their oxen in the protected areas during since 1992, whereby an obligatory 20 days ploughing times. However, the sustainability per year labour contribution by every of the enclosure would heavily depend on resident in the region was adopted. This the immediate benefits communities and labour was mostly used to construct and household will obtain and the strength of the rehabilitate terraces and landscapes. This local institutions. type of policy could reverse land degradation especially if it is done in a Labor availability participatory way. It could lead to community action with apparent economic In situation where the cropping season is and social benefits in terms of fodder, fuel doubled because of increasing access to wood, water and other resources. irrigation during the dry seasons, the pressure on household labor was apparent. However, although terraces are in place in The shift in systems from less labor most of the sites the effectiveness of these intensive cereals to labor intensive structures in reducing erosion, in improving vegetables (Table 6) caused an increasing water infiltration and in performing other labor demand. When a vegetable farmer was environmental services was not assessed. compared to a cereal farmer the demand for Participatory impact assessment of the labor was 1638 and 406 man days per ha, structures to display the positive and respectively, which was about 400 % higher, negative effects on the system could attract indicating that there could be a need for an investors and the interest of communities to additional labor through hiring, debo (local manage them sustainably. Moreover, labour sharing arrangement among age integrating niche compatible trees on the soil groups) or any other arrangements. Farmers’ bunds may reduce the pressure on cow dung, interviews in Lalay agula revealed that lined which is currently used as a cooking fuel, to canales and cemented diversions reduced the be used for soil fertility restoration. Burning pressure on farm labor, about 5 to 8 man dung is one of the practices aggravating land days per family per season, which otherwise degradation, particularly in the Amhara and used to be invested in cleaning and repairing Tigray regions, as it breaks the nutrient furrows after the main rainy seasons. This recycling in crop-livestock systems. has created a job opportunity for the land less youth across the regions. There is a huge land area which became under area enclosure in all regions though Upstream and Downstream the size and management modalities differ Relationships from region to region. It was done particularly to protect schemes and upper The presence of very few perennial rivers slopes from producing silts but also to aggravated by recurrent drought, and rehabilitate exhausted upper slopes. In most extensive awareness creation campaigns cases it was done in consultation and towards a shift to vegetable farming in agreement with the local communities. In almost all schemes incurred a considerable sites where area enclosure is practiced for at competition for water in the command areas least two or more years, like in Gereb koky, and beyond across regions, and caused with an area enclosure of 30ha, the shortage of water for down stream users vegetative cover of the system increased

290

(Table 4). Competition for irrigation water there is no visible change on the income and between upstream and downstream users, livelihood of the people, even four years between vegetable growers and chat after the irrigation infrastructure is in place growers, between farmers with big irrigable and being operational (personal plots and small plots, and between water observation). users and water managers have been surfaced during the various formal and In general, there are very few operational informal discussions. Farmers in two sites in SSI schemes in the country that could be Harer, Oromia indicated that there is less labeled as optimum because of the various water for down stream communities than difficulties facing them at different times ever due to the need for frequent watering of and scales, ranging from shortage of the vegetables in the upstream fields. The technologies to imperfect markets. An traditional system, which has a considerable innovation systems approach is required to Chat farming, demanded watering only once enable the schemes bring the expected in a month and used to release a impact on the livelihoods of the people, considerable amount of water to including the identification of various downstream users. On the other hand, challenges small scale irrigators are facing at recently introduced vegetable crops like farm, community, district and higher scales. tomato require water once in a week, There is also a need to look for success particularly in areas where the evapo- stories where combination of technological, transpiration is very high. The consequence policy, institutional and market interventions was that there was an emerging conflict made some irrigation investment worth between upstream and down steam users, investing to make the respective rural and in some sites (e.g. in Burka Woldiyaa) communities food secured and keen to the case is presented to the local court. protect the environment. Table 7 displays Additionally, some farmers started to divert interventions that made few irrigation the water to non-traditional farms using schemes success story. motor pumps and gravity, commonly without asking for consent with the a. Access to irrigation and the associated traditional water users. In some cases, these institutional services given by were people highly protected by local and governmental and non governmental regional authorities with very limited chance institutions helped farmers to improve for the small farmers to maintain the their income and enhanced their statuesque. capacity to shift towards market- oriented agriculture. Unfortunately, In situations where there is an absolute priority was given only to the crop water scarcity, it is only households residing sector while livestock, particularly on the source who could benefit from the dairy and fattening, could have limited water flows. Communities residing increased the benefits by much higher down stream tend to send their water master orders. to negotiate with upstream communities in times of critical demand. b. Although crop diversity is one key way of minimizing risk and exploiting Lessons Learned opportunities, too much diversity in the farmers’ fields may prevent them In areas where institutions and market from more efficiently developing their incentives are in place (e.g. Ziwai in the Rift production skills and creating Valley of Ethiopia), farmers have doubled or functional market links with tripled their incomes in a very short period specialized traders and consumers. of time. In other isolated, less accessible Too little diversity may again lead to areas (e.g. Belessa in the Amhara region), deteriorated market prices during the

291

peak season. Hence farmers’ should References be assisted to optimize the number and type of enterprises that are Amede, T., Stroud, A., and Aune, J., 2004a. managing and minimize market trade- Advancing human nutrition without offs for better decision making. degrading land resources through modeling cropping systems in the c. Although there is a better use of Ethiopian Highlands. Food and land based resource under irrigation, Nutrition Bulletin, Vol 25, No 4. farmers are still dependent on low input management, which may not Amede, T., Kimani, P., Ronno, W., and bring the expected quantity and Lubanga, L., 2004b. Coping with quality reflecting market demands. Drought. CIAT Occasional Publications Hence there is a need for Series No. 38. Pp 39. diversifying the inputs supply systems including credit IE Preliminary survey, 2004. Preliminary opportunities. Impact Assessment: Case studies of three selected small scale irrigation d. Water is a very scarce resource, and schemes. Interim Evaluation, IFAD. the demand for irrigation water is on the rise. Hence irrigation Benin, S., Pender, J. and Ehui, S. 2003. investments should be supported by Policies for sustainable development in water saving agronomic and the highlands of Amhara region. technical measures, including Overview of findings. In: Amede, T. mulching, tie-ridging, minimum (ed). Natural resource degradation and tillage, lining of canals and drip environmental concerns in the Amhara irrigation. region. Pp 185-207.

e. The current extension support on Bureau of Agriculture-Tigrai, 2004. Report irrigation agronomy is far from on achievements in seed multiplication, responding to farmers’ expectations. field demonstration and field trials. Participatory on-farm research on Unpublished. irrigation frequency, crop water demand, crop rotation, organic COSART, 2001. Study and design resource management, micro dose department feasibility study report on application of chemical fertilizers, agronomy and soils. Regional Bureau of management of perishable seeds and Water Resources, Tigrai. related issues should be promoted. The process should give farmers the Dessie, Shiferaw, 2004. Impact assessment chance to innovate. on Shoba irrigation scheme. Ministry of Water resources, Addis Ababa, Ethiopia f. It could be necessary to distribute demonstration fields to various farm niches Engida, Mersha, 2001. Agroclimatic and landscape positions to reflect field classification of Ethiopia. Ethiopian variability. Promoting the capacity of elite Journal of Natural Resources 2(2):115- farmers ‘like the Hirsha Cadre’s in Tigray’ 135. will enhance the scaling-up process to reach more farmers and communities sustainably. G.Medhin, B., Pender, J., Ehui, S. and Haile, M., 2003. Policies for sustainable land management in the highlands of Tigrai, Northern Ethiopia. EPTD workshop summary paper No. 14.

292

Gezahagn, Aynalem, 2003. Agricultural extension policy and strategy in the Amhara Regional state. In: Amede, T. (ed). Natural resource degradation and environmental concerns in the Amhara region. Pp 185-207.

OIDA, 2004. Assessment report, Burka Woldiyaa and Nedhi Gelan Sadi. Addis Ababa.

Pender, J. and B. Gebremedhin, 2004. Impacts of policies and technologies in dry land agriculture: Evidence from Northern Ethiopia. Crop Science Society of America and American Society of Agronomy. Challenges and Strategies for dryland agriculture. CSSA Special Publication no. 32.

WCD (World Commission on Dams), 2000. Assessment of irrigation options. Technical review IV.2. Cap town, South Africa.

UNOPS, 2003. Special country programme- Phase II. Supervision report.

293

Table 1. Comparison of yield of various food and vegetable crops of irrigated and non-irrigated farms in Lalay agula, Tigrai (SCPII scheme). N= 6

Crop type Yield (qt/ha)

Non –irrigated Irrigated % farms farmers increase Maize 25.75 43.75 69 Onions 102 172 68 Pepper 11 17.75 61 Tomato 200 300 50 Carrot 226 305.25 35 Potato 174.8 250 43 Cumin 4.25 7.50 76 Source: Woreda Bureau of Agriculture, Wukro, 09/2004.

Table 2. Crop yield before and after the irrigation project in Shoba, Oromia (SCPI scheme). n= 10, nd= not determined.

Yield (Qt/ha) Crop type Land area Before the After the (ha) project project crease Barley Nd 10 12 20 Wheat Nd 8 16 100 Maize Nd 15 25 66 Faba bean Nd 4 12 200 Source: Dessie, 2004

Table 3. Seed source for major crops in IFAD sites (Mean of 3 sites) SD

Seed source No of sample rce Own seed 708 1 Government 294 9 Purchase 284 1 Total 1286 .0 Source: IE Preliminary Survey, 2004

294

Table 4. Experiences in access to irrigation water in three schemes of IFAD sites in Ethiopia. Management practices Irrigation site Total Dobena N.GalanSadi Hizaeti Afras % % % % sample Users sample users sample users sample users Practice crop rotation 51 61.4 109 70.3 135 99.3 295 78.9 Mulching 68 43.9 7 5.1 75 20.1 Intercropping 52 62.7 106 68.4 158 42.2 Contour farming 46 29.7 12 8.8 58 15.5 Physical soil conservation 16 19.3 84 54.2 29 21.3 129 34.5 Biological soil conservation 9 5.8 5 3.7 14 3.7 Other change in farm 2 2.4 5 3.2 1 0.7 8 2.1

Table 5. Change in farming practice as an effect of the SSI schemes, n=374.

Experience in water supply Irrigation site Total Dobena N.GalanSadi Hizaeti Afras Count Col % Count Col % Count Col % Count Col %

Sufficient water not available 73 90.1 125 82.2 106 80.3 304 83.3

Sufficient water available 5 6.2 27 17.8 25 18.9 57 15.6

Poor management of water 3 3.7 - - 1 .8 4 1.1

Total 81 100.0 152 100.0 132 100.0 365 100.0

Table 6. Labor requirement of various vegetable and field crops (man days/ha) in irrigated farms of Northern Ethiopia.

Crop Pre-planting Cropping Post harvest Total operation operation operation Onions 110 255 10 375 Potato 24 253 16 293 Carrot 50 235 15 300 Cabbage 110 235 10 355 Shallot 50 255 10 315 Maize 14 40 10 64 Wheat 22 60 10 92 Barley 22 60 10 92 Teff 18 68 12 98 Beans 18 34 8 60 Source: COSART, 2001

295

Table 7. Best bet interventions for small scale irrigation schemes; lessons identified from selected schemes and communities that work best.

• Technology for • Where does it fit best? SSI • Irrigation water • Water-saving lines canals • • Multiple cross-over alignments • • Water reservoirs with closing lockers • • Furrow irrigation practiced • • More perennials than vegetables or cereals • • Market access is reliable • Local institutions function • Vegetable • Close to town markets farming • Reasonably good road network • • Rotational cropping practiced • • Water supply is adequate • • Managed by literate and young farmers • • High yielding • Water supply is adequate varieties • Better access to inorganic fertilizers • • Rotation with legumes possible • • Land shortage is not apparent • • Managed by resource-rich farmers • • Inorganic • Best used for maize fertilizer • Water supply adequate • • Is accompanied by compost and manure • • Higher Urea and less DAP • • • Organic fertilizer • Biomass is abundant • • Cow dung is not used for cooking • • Composting is practiced • • Preferably for fruit trees and vegetables • Pest management • Avoid growing potato and tomato together • • Uproot diseased plants • • Crop rotation practiced • • Field health sanitary practiced

296

Fig. 1. Food insecure and secured Woredas in Ethiopia (DPPC, 2004)

297

Institutions, Management Practices and Challenges of Small-Scale Irrigation Systems in Ethiopia: A Case Study of Two Modern Smallholders Irrigation Systems in Western Oromia, Ethiopia

Shimelis dejene1, Woldeab Teshome2, Godswill Makombe3, Seleshi Bekele3 , Krishna Prasad3 1 Jima University, 2 AAU, 3 IWMI-NBEA [email protected]

Abstract locations within irrigation systems. Nonetheless, many irrigators did not maintain these positive This paper examines the institutional changes for long. The constraints were scarcity arrangements that facilitate irrigation and unreliability of water and management and management and the present state of irrigation socioeconomic problems. These, in turn, were management and establishes where problems mediated by lack of: a) clearly defined and well have occurred in the operation of Gibe-Lemu and enforced institutions of land and water rights; b) Gambela-Terre Small-scale irrigation systems. technical problems in design and construction; c) The study employed the case study approach to inadequate institutional capacity of the local state tackle the research. Key informant and expert irrigation agency to coordinate and support interview, desk review of different documents decentralized management of irrigation; d) produced about the projects, group discussion, policy related problems; e) inadequate direct observation and structured interview organization of users for self management; and f) schedule were used to collect data. The study problematic social relation of power among proved the proposition that the government has water users. Finally, the paper draws a number uncritically supported the irrigation systems. of conclusions, using the theoretical notions like Enabling legal system of land and water rights, context, social requirement for use, social effects strong woreda level state irrigation agency, and social construction, about policy options and support services (irrigation extension) and well- requirements in the readjustment of the surveyed established water users associations through irrigation systems and in the design of irrigation which purposes of irrigation are achieved were projects of these types. not adequately planned and put in place. These shortcomings undermined irrigation Key words: Institutions, irrigation, management management, ultimately risked feasibility and practice and challenges sustainability of irrigated agriculture. Findings revealed poor record of accomplishment, in spite Institutions, Management Practices and of the difference between the two systems, in Challenges of Small-Scale Irrigation Systems in managing water distribution in terms of the three Ethiopia: A Case Study of Two Modern most important performance indicators: Smallholders Irrigation Systems in Western adequacy, reliability and equity in water Oromia, Ethiopia distribution. Water related conflicts are rampant but not settled yet. In addition, results indicated that irrigation had positively impacted irrigators’ Introduction livelihoods in terms of diversification and intensification of crop production, household Gibe Lemu and Gambela Terre small-scale income, housing and employment generation and irrigation systems were constructed to promote this social effect of irrigation was significantly household food security through effective and different between irrigation systems (due to equitable use of the available land and water difference in the institutional and socioeconomic resources. However, these projects were poorly context of the two irrigation systems) and performing and the area under irrigation is below

298

expectation. Nevertheless, more grounded and irrigation technologies) by Mollinga (2003) and detailed scheme-specific information on the elements of the social organization of innovation reason why it is so is and which help guide theory by Engel (1997) were used in developing future policy for smallholders irrigation the instruments for data collection; formulation promotion not available. It is increasingly being of the hypothesis and in the interpretation and recognized that poor performance is not only a explanation of the data collected and the facts in consequence of technical performance in design the irrigation systems. and construction, but institutional and management problems tended to be more The study employed the case study approach to common constraints to the success and tackle the research. There are ample reasons for exploitation of small-scale irrigation schemes in using the case study approach. The first reason is sub-Sahara Africa. In addition, in-depth the nature and objective of the research itself; understanding of technical and agronomic irrigation management practices and analysis of problems of irrigated agriculture is impossible institutional contexts and causes of the problem without understanding the social organizations are complex processes and therefore require (institutions, the policy environment and social detailed investigation and comprehensive relations) in which it is embedded. The general understanding. Secondly, the conventional objective of this study is, therefore, to examine questionnaire survey (structured interview and describe the reasons for the disappointing schedule) alone does not allow comprehensive performance of the target irrigation systems with understanding and adequate description of how a focus on the institutional arrangements that the schemes are actually managed and what facilitate irrigation management, the present state institutional and socio-economic variables and of irrigation management at the local and processes explain poor performance of the grassroots levels and establish where problems schemes. (gaps) have occurred. Methods of Data Collection

Specific objectives: Secondary Data Collection

• To analyze the institutional Secondary data was collected through desk arrangements and relationships that review of the regional and national irrigation affect management and 'performance' of policy statements, legal frameworks regarding the irrigation systems; irrigation land and water rights institutions, • To understand how users are organized proclamations and regulations, project write-ups, for self management of irrigation; project appraisal documents, different reports • To understand how the irrigation produced about the projects and past case study systems are managed and; papers on irrigation. • To identify and clarify the major challenges of the irrigation systems; and Primary Data Collection • To suggest possible options/strategies

for rehabilitation of the surveyed Relevant primary data were collected using irrigation systems and in the design of various instruments such as similar irrigation projects

• Key informant interview; conducted to Methodology generate general understanding of the irrigation systems. In addition, the The research work included both appraisals of information obtained through this tool theoretical assumptions on conditions of was also used for developing more possibility for successful irrigation as well as the focused questionnaire for the household analysis of empirical data obtained during field interview research. The socio-technical approach to • Interview (semi-structured) with irrigation (the social shaping perspective of

299

executive committee of water users and 2.2. Data Analysis officials and experts with relevant government agencies at regional, zone Both qualitative assessment and descriptive and wereda levels of accountability analysis techniques were used for data analysis. • Group discussion: 60 purposively The data generated through household interview selected irrigators, divided into six was analyzed by employing SPSS. The study groups (three in each irrigation system), employed descriptive statistical methods such as and each with group size of 10 members frequency, percentage, mean, and23 standard were involved. Irrigators from the head- deviation, X2-statistic, T-test and ANOVA/F-test end to the tail-end areas were included in for analyzing the data generated through the groups so the data generated reflects household interview. the actual situations and facts at all water levels. • Direct observation of events during Rsult and Dscussion many visits paid to the schemes • Household interview using structured Description of the Irrigation Systems questionnaire Gibe Lemu Lemu and Gambella-Terre irrigation systems are found Gobu Seyo District, East For the questionnaire survey, 65 sample Wellega Zone of Oromia Region. Gibe-Lemu is households were selected using the following located at 80km towards Addis Ababa from procedure: Nekemte, capital of East Wellega zone and 3Kms from Bako town. Gambela –Terre SSIS is a) First, the sample frame was obtained located at 12kms from Ano-town, the district from the executive committee of capital and 30km from Bako, the biggest town WUAs; providing access to markets for farmers. The b) Then, the beneficiary households were rainfall in both irrigation systems is unimodal. stratified into head-end, middle and The unimodal rainfall pattern dictates the single tail-end irrigators based on their cropping season. However, in recent years, the location in the farm layout of the pattern of the rainfall becomes uneven and irrigation systems and proportion was unpredictable with negative implication on food assigned to each group for inclusion in production. the sample; and c) The households were stratified on the The total irrigable command area of the Gibe- basis of their location with the basic Lemu irrigation scheme is 113 ha. A main canal assumption that there could be inequity having a length of 7kms conveys water into the in the distribution of irrigation water command area. The method of distribution to the and the benefits derived from irrigation main, secondary canals and TUs is continuous, as a result of weakness in water while it is rotational in the farm units as per the control, technical problems and lack of initial design of the project. However, currently, management structures that suit layout the method of supply to the TUs is rotational due of the irrigation infrastructures (see to the decline in the volume of water conveyed also Vermillion/IIMI, 1997:30-31); into the diversion weir. The method of d) 30 % (from each scheme) of the application to the farm units is rotational, while sample frame were selected using the method of application of water is furrow. stratified random sampling Seven days are one irrigation interval for each technique and participated as farm unit at the time of design (Korea Design respondents in the household Team, 1990). However, there is severe water interview.

300

scarcity in the scheme to day to supply water in constructed along the main conveyance canal to accordance with this interval. bring additional 70 ha of land, which was not considered in the initial design, under its Gambella-Terre SSI project was initiated (during command area so as to benefit up stream farmers the Derg Regime) in line with the political whose land holdings fall on the left and right interest of the Derg, i.e., to be used as instrument sides of the main conveyance canal. This for promoting collectivism through cooperative increased the command area from 80ha to 150 farming. Initially, it was designed to develop an ha. Nevertheless, there is a wide gap between the irrigable area of 80ha. In 1995, additional 34 supply of water to day and during the time of turnouts, division box and other structures were designing the project (1988).

Table3.1 Average distance from the market Round trip distance Gibe Lemu Gambela Terre (minute): Mean N St.De Mean N SD

From the main road: 55.20 25 40.56 116.75 40 64.22

From the market: 93.24 25 51.32 144 40 77.49

Source: - Field survey, *=household, **= Male headed households, ***=Female headed households

independent district irrigation extension office The average round trip distance from the main and the specialized development centers were asphalt road and the market place is different merged with the District Agriculture and Rural between the two irrigation systems. Irrigators in Development Department (DARDD) with no Gambela Terre have to walk longer hours than clear line of communication with the Branch and Gibe Lemu to access the nearest local market to sale their agricultural produces (Tables 3.2). In Regional Offices of the Irrigation Agency addition, there is no all-weather road connecting responsible for irrigation development in the irrigation system to the main asphalt road Oromia. The District Irrigation Desk (DID) has despite the fact that it is one of the material been created in the DARDD as a team in 2004 contexts for successful irrigation (Engel, 1997; and the full time irrigation extension workers Mollinga, 2003; Dillon, 1992: FAO, 2003). become multipurpose. Further, farmers in Gibe Lemu have long standing (more than half a century) tradition in practicing traditional irrigation while farmers in Now the Gobu Seyo District Irrigation Desk Gambella-Terre had no irrigation experience (GSDID) and the extension centers are before arrival of the new irrigation project. accountable for supporting user- based/decentralized management of irrigation The Institutional Setting of Irrigation and coordination of efforts of partners in the Management at Local and Grassroots Levels administration of irrigation in the district. Nonetheless, it has inadequate capacity to 3.2.1. Local Level Institutional Arrangements shoulder these responsibilities in terms of human resource development, technical units, structure Structure and Management Capacity of the and logistics in spite of the government policy District Irrigation Desk (DID) for capacity building and institutional development. The GSDID has been consisted of Oromia Irrigation Authority was responsible for only one team leader who is in charge of the SSI development in Oromia between 1999 and Desk. It operates only with 20% of the required 2004. Following the 2004 restructuring, the technical staff (table 3.2). Regarding

301

transportation means, the Desk is equipped with scheme levels. But, CTA (1999: 91-92) argue only one motor bicycle, one room (office), one that a necessary condition for more efficient and table and one chair which are exclusively used lasting management of smallholders' irrigation is by the Team Leader by the survey date (March, existence of management capabilities, which are 2005). In terms of structure, the GSDID lacks built through organizational and institutional organizational unit (development centers) with development at various levels; from the apex DAs fully responsible for irrigation at the through the middle level to the grassroots levels.

Table3.2 Manpower status of GSDID; as of March 2005 Discipline (positions) Required and Available Gap Approved Team leader 1 1 - Irrigation Engineer 1 - 1 Irrigation Agronomist 1 - 1 Community participation Expert 1 - 1 Water Harvesting Expert 1 - 1 Total 5 1 (20%) 4 (80%)

Institutional instability had adversely affected the human resource capacity and structure of the Stakeholders and Partnership in Irrigation local irrigation agency. For example, the Management irrigation sector institution in Ethiopia experienced reorganization in 1983, 1994, 1995, The policy framework for small-scale irrigation 1999 and 2004. This in turn has challenged development in Ethiopia states that management human resource development and affected of SSIS is a joint responsibility. In view of this, structure and existing human resource capacity the regional state irrigation agency identifies of the irrigation sector institution. It could not cooperative promotion and input supply desks, maintain its trained manpower at both district district and grassroots level administrative and and grassroots levels. The number of trained legal entities and farmers and their organizations professionals who were working on irrigation at as main stakeholders in the administration of the District Irrigation Office was reduced irrigation in the study area. In addition, in 2004, following the restructuring in 2004. This the regional government merged five concerned weakened irrigation expertise at the District district level government agencies in one level. Until the 2004 reorganization, there had institution; Agriculture and Rural Development been one trained DA (Diploma graduates, who Department (ARDD), with the assumption that received in-service training in irrigation) in each organizational proximity can provide a fertile irrigation system. These trained DAs were taken ground for collaboration. Although merging is a to district to work in other offices. However, no good opportunity, the five institutions did not full-time and trained DAs assigned after that. In work together in irrigation management as addition, the newly assigned DAs have multiple expected; the achievement has mainly been mandates and over stretched with many physical proximity of the agencies although the activities. Therefore, they are unable to social shaping perspective of irrigation undertake strict follow up of user-management technologies assumes that irrigation systems are of irrigation and could not deliver adequate socially constructed. irrigation extension services to farmers. The responsibility for coordinating partners fell on the District Irrigation Desk (DID). Nevertheless, it could not manage to do it owing to lack of well-established institutional and

302

functional framework for cooperation and successful irrigation include, among others, harnessing their efforts. Lack of adequate improved seeds that work under irrigation and involvement of the partners in turn adversely strong extension services (Engel 1997:147, affected irrigation management in many ways, Mollinga 2003: 24 and Dillon, 1992: FAO including: 2003). However, survey results revealed only 48% and 35% of the sample households ever • Necessary inputs for irrigation could not used seeds of improved vegetable and cereal be availed to farmers regularly; his crop varieties in Gibe Lemu and Gambela Terre responsibility fell on the input Supply respectively. Out of the sample households, 36% Desk. in Gibe Lemu and 27.5% in GTSSIS procured improved seeds of maize and potato from the • Irrigation has no research input; adaptive extension service. Nonetheless, the maize and crop varieties that work under the potato seeds they obtained from the extension situation of irrigation and watering service are not recommended specifically for frequency (irrigation agronomy) irrigation. In addition, seeds of carrot, onion, recommended for irrigation are non- tomato, chile and other vegetable crops that existent. Although there exists one big irrigators regarded as 'improved' seeds have research center of the Oromia mostly been procured from the market or shops; Agricultural Research Institute in the they are not specifically recommended for nearby area, it has not been supporting production under irrigation. Therefore, they did irrigation development through not suit the irrigation systems, for they are supplying relevant technologies. This is affected by disease. due to the fact that agricultural research policies and strategies are not adjusted to Overall, irrigators have not regularly been meet requirements of irrigation and supplied with these support services mainly none-existence of enabling institutional because the government policy on agricultural framework for harnessing efforts of the input supply, agricultural research and rural research system with irrigation; extension, gives more priority to those farmers registered in package program for rain fed • The district and village level agriculture. It tended to favor, in terms of both administrative and legal entities do not supply and timing of supply, rain fed agriculture play any meaningful role in water during the main rainy season. control and conflict resolution though the task has become more complex to be Institutional and Organizational Conditions addressed by ‘Kore Aba Laga’ and the within the Irrigation Systems simple informal rules alone (bylaws); and Land Distribution and Its Problems

• Water users not organized into Gibe Lemu and Gambela Terre Small-scale legally recognized entities in irrigation schemes were constructed to resolve accordance with the principles of the problem of farmland shortage, increase organization of cooperative societies. production and productivity and to improve This responsibility rests with the farmers’ livelihoods through effective and equitable use of the developed land and water cooperative promotion desk. resources. In view of this, ORLUA Proclamation Nevertheless, it did not discharge this No. 56/2002) states a maximum of 0.5 ha is responsibility. retained for each former landholder in the command area and each member is equally Support Services allocated 0.25 ha per household. Nonetheless, this has not been finished in practice in both The support services, in Molliga’s and Engel’s irrigation systems. Results indicated the whole words ‘the material conditions of possibility’ for command area has been owned only by 22.4%

303

(in Gibe Lemu) and 57% (in Gambela Terre) of individual households. Equal contributions are the intended beneficiary households. Former plot requested from all members who cultivated 0.5ha- holders continued to control and manage land 12ha of irrigable plots. Ali Seid (2002), Lema areas that fall in the range of 0.5ha to 12ha in (2004) and JICA and OIDA (2003) found similar Gibe Lemu and 2ha to 5ha in Gambela Terre problem in their study in North Wello and East irrespective of their resource capacity to fully use Shoa zone respectively. In addition, some farmers it for irrigation. This has been in spite of the fact are over supplied with water, while some others that the rest farmers are landless; tenants and/or obtained water, which is far short to meet their sharecroppers (specially in Gibe-Lemu SSIS). needs due to the guesswork in water allocation. Over supply has led to misuse in the context of The problem of landlessness and skewed severe water scarcity. distribution of irrigable land was more severe2. in Gibe Lemu. Out of the irrigators, 2o% did not 3. Organization of Users for Self-Management of possess own irrigable plots, while some rich Irrigation farmers were managing 9-12ha of potentially irrigable land. T-test also showed there was Organizational Set-up significant difference among households in Gibe Lemu; which is significant at the 5% level. In accordance with the federal and regional Landless farmers and those who own small plots policy framework for small-scale irrigation access irrigation land mainly through development in Ethiopia, "Kore Aba Lagas'' are sharecropping system, labor exchange and in charge of water allocation, distribution, exchange of ox for land (see appendix table). observing the water rights of members, conflict Similar study by JICA and OIDA (2003: 3-6) in management and coordination of maintenance East Shoa and Woldeab (2003) in Tigray also activities. documented that sharecropping system (leasing- in and leasing-out) is one of the common option Although there are many deficiencies in their available to land owners with low resource organization, the water users in both irrigation capacity and landless farmers and farmers with systems have created their own management smallholding. structures and crafted internal bylaws as one of the social requirements for better management. Fair distribution of irrigable land has not been Executive committees, sub-committees and achieved in both irrigation systems by the survey water user teams (WUTs) were formed at date. The government failed to achieve fair irrigation system and distribution levels [territory distribution of irrigable land because: a) land units (TUs)] to facilitate water control and redistribution issue was not dealt with during coordination of maintenance activities design and construction, b) GTSSIS was initiated in line with the political interest (collectivism) of All water users are organized into 6 WUTs the Derg regime in which case land redistribution (Water users teams) in Gibe-Lemu (group size was not an issue and; c) lack of policy and ranging 10-20 members and in Gambela Terre enabling legal system for redistribution for a into four WUTs "goxi" with the number of long period and a time lag between the issuing of members per WUT ranging from 17 to 44. the Oromia Rural Land Proclamation and Nonetheless, the group size of two WUTs in operational regulation. Gambela Terre is above the optimum range (20- 1. 30) for good management (See Woulter, 2002: Inadequate land tenure in turn has created Blank, 2002). In these WUTs it has been management difficulties in the schemes. Both observed, because of large group size, greater 'Kore Aba Lega', DAs and the local state socioeconomic differentiation and lack of mutual irrigation agency do not clearly know the actual understanding among users, which led to severe size of irrigable plot managed by individual problem of water distribution and conflict over households. Hence, they could not adjust water water. Similar study in Kenya showed that, the allocation and resource mobilization to amount of whole schemes or part of it was not operational, water used and irrigable area controlled by in all schemes consisting of groups of over 30

304

members (Woulter, 2002). With a membership Irrigation Management Practices within below 30, he observed no water distribution the Irrigation Systems problem in Kenya. To the contrary, the situation in Gibe Lemu contradicts with Woulter's Water Management findings. The number of members of WUTs is 10-20, which is below 30 but still there is water Water Allocation distribution problem and users could not settle water dispute themselves. This shows group size Water committees are in charge of water is not the only factor for social cohesion and allocation with little support from irrigation effective group performance in water agronomists and development agents. They distribution. allocate water and prepare rotational schedules every year in September. However, water users Viability of the Water Committees for Self- expressed that Water allocation made by the Management of Irrigation ‘Kore Aba Lagas has limitations in terms of both design and implementation. In terms of design, it The responsibility for running management of does not clearly define water rights of individual the irrigation systems was delegated to "Kore farmers and TUs due to the guess work in water Aba Laga" in the hope of enhancing allocation. Equal water supply period per turn is effectiveness, equity and responsiveness in allocated for all TUs and individual water users irrigation management and to ensure in spite of the variation in water requirements of sustainability. Nonetheless, they were not the different crops grown, area of irrigable plots organized in such a way they can ensure these managed by individual irrigators and water objectives of decentralized management, demands in different TUs. The major although good organization is one of the social impediments for proper allocation and requirements for good irrigation governance. scheduling of water distribution (as reported by They have deficiencies in their management committee members), include: structures. They have no recognized legal power and the roles, responsibilities and authorities of • Guess work in water allocation; the the different positions along the management water committee undertakes water structure are not clearly defined and even it is allocation and defines water rights of totally missing from the by-laws of the ‘Kore members not based on study on water Aba Lagaa’ in Gambela Terre. requirements of different crops, irrigable plot area possessed by individual The committee lacks transparent accountability irrigators and measurement of the yearly to users although it is one of the essential factors water supply due to capacity problem. for good irrigation governance. Constituencies This is because the local state irrigation (water users) accuse committee members for agency failed to provide satisfactory power abuse, selfishness, lack of commitment, technical assistance in undertaking these and for not observing the internal bylaws. water management tasks and in building Nevertheless, informants reported that they were their capacity; and. not held accountable through legal processes. One informant in Gibe Lemu, Name- • Water users are not willing to register Mohammad Shumiye, expressed the intensity of types of crops they grow (vegetables or the accountability problem using the following perennials) and area of their irrigable proverb: plots with the committee for clear definition of water rights in spite of the "Yegebere balesiltan yasyilign gebeya new’, law (bylaws). meaning, the committee members abuse the power and authority we vested on them and Water Distribution prioritize their interest and irrigation fields in water allocation and distribution"

305

The most important performance indicators in Gibe Lemu and Gambela Terre, 80% and 90% of the distribution of irrigation water include the sample households witnessed that they could adequacy, timeliness and equity in the supply of not obtain the quantity of water that can support water (World Bank, 2000). Table 3.3 shows irrigation over the plot area they manage. The users’ evaluation of performance of ''Kore Aba vast majority, 76% (in Gibe-Lemu) and 80% (in Legas'' in water distribution. The water Gambela Terre) of irrigators were not able to committees in both irrigation systems were obtain water in a reliable manner. Further, results found to be in efficient in managing water of chi-square analysis indicated distribution in terms of the three indicators. In

Table 3.4.Water users’ opinion about water distribution in Gibe Lemu and Gambela Terre irrigation systems Item Opinion by irrigation system and location Gibe Lemu (N=25) Gambela Terre (N=40) Count % Count %

Enough water is not obtained 20 80 36 90 Water is not reliable 19 76 32 80 Water distribution is unfair 21 84 33 82.5 Source: Field Survey

Access to adequate irrigation water and the to adequate and reliable supply of water is problem of unreliability of water has strong highly unlikely if the farmer's irrigable plot is in association with location of farmers' irrigable the tail-end area in Gambela Terre (appendix plots relative to the headwork. The difference Table1). Water is scarce and the problem of between locations was highly significant in unreliability is more severe in tail end areas in Gambela Terre (X2=10.6, X2-Prob. =0.005 (for both irrigation systems. adequacy) and P<0.005 for reliability). This implies there is a greater probability that access

Table3.4 Order of reasons why farmers do not obtain adequate water for irrigation, Gibe-Lemu and Gambela Terre SSIS Irrigation system Statistics Water Seepage Poor water Turn scarcity loss control abuses Gibe Lemu N 6 3 5 9 % Of farmers 24 12 20 36 Rank 2nd 4th 3rd 1st Gambela Terre N 20 2 8 9 % of farmers 50 5 20 22.5 Rank 1st 4th 3rd 2nd Source: Field survey, March 2005

Tables3.4 shows farmers ranking of problems problems (water scarcity and seepage water loss) that constrained the supply of adequate water in tended to be the least important constraints for a timely fashion. Water scarcity due to decline in not meeting water needs in the scheme, the quantity of water conveyed into the scheme indicating institutional and management and uncontrolled distribution were the prime factors responsible for scarcity and unreliability Problems are more relevant. Water users in Gibe of water. In Gibe_lemu, hydraulic and technical Lemu believe that the current volume of water

306

conveyed into the scheme can meet water strong system management. Water scarcity is the requirement in the command area with some most important reason for not obtaining the adjustment and adaptation of water allocation to needed amount of water for irrigated agriculture the change in water supply and if there had been in command areas of Gambella Terre SSIS

Table 3.5 Social groups that get more water by illegal means Groups Percentage of farmers giving the response All HHs (N=65) Gibe Lemu Gambela Terre (N=25) (N=40) 4. Farmers with large family size 89.5 60 42.5 Head-end farmers 89.5 76 82.5 Rich farmers who irrigate perennials 39.7 36 35 Source: Field survey

Alongside the above, results revealed that Water Causes of Water Scarcity Committees were not able to ensure equity in water distribution (Table3.5). Informants Water is scarce in the irrigation systems; reported that powerful and rich socioeconomic especially in Gambella Terre. Table3.6 shows groups, in their words, 'gulbetegnas'/’bully perceptions of irrigators about causes of water farmers’ have been benefited more. Head-end scarcity. Gibe and Dokonu rivers, which are farmers had better access to irrigation water water sources for Gibe Lemu and Gambela owing to their proximity to the headwork Terre, were diverted at 2 and12 locations (location advantage). They release water for the respectively. This decreased the quantity of down stream farmers once their fields saturated water conveyed into the schemes. Nevertheless, with water. Households with large family size the problem has not been addressed due mainly are more powerful (because of size) and often, to first, there was no enabling legal system, they exercise power to obtain water by illegal which clearly defines the water rights of the means. They also take advantage of the relatively upstream traditional irrigators and irrigators in large family size and/or labor in defending their the new irrigation projects. In spite of the general water rights. Rich farmers in the middle areas, constitutional rule, there are no formal especially in Gambela Terre, irrigate large areas operational rules and regulations for managing of tree crops which are not in the priority list and the relation between the upstream and do not releases water for the tail-end farmers. downstream irrigators in sharing the water from Nonetheless, the WUA committee could not the same river. In the second place, the regulate this distribution inequity owing among responsibility for addressing such problem (the others to resistance by the powerful groups. role who should do what) of the different stakeholders has not been defined by the survey Water Scarcity: Causes and Coping Measures date.

TABLE3.6 Perceptions of irrigators about causes of water scarcity by irrigation system Causes of water scarcity % of farmers giving the opinion All HHs Gibe Lemu Gambela (N=65) (N=25) Terre (N=40) Diversion of water by traditional irrigators 89.32 88 97.5

307

Seepage loss 47.74 52 52.5

Increasing number of users 63.14 80 62.5 Lack of strict water control 70.84 80 80 Source: Field Survey, March2005

Alongside water diversion by traditional crops as an adaptive measure to the irrigators, the problem of water scarcity has been problem of unreliability of water; mediated by abuses (uncontrolled distribution), • Night storage was constructed (in social incompatibility (in Gambela Terre) and Gambela Terre SSIS) to over come increase in the number of water users in Gibe water scarcity through rotational Lemu, against the declining quantity of water distribution day and night. Still the conveyed into the scheme(table3.6). Farmers in volume of water flow is far short of the middle areas, especially in Gambela Terre, water needs in the scheme; and irrigate large areas of perennial tree crops, which • Water users in Gambela Terre employed are not in the priority list of crops to be grown; a paid guard; to control water leading to scarcity the in tail end area. Further, distribution and to address the the designed irrigation season for the scheme coordination problems of ‘Kore Aba (Gambella Terre) is October to March every Laga’ . However, the guard could not year. However, the indigenous growing season adequately manage the distribution for rain fed agriculture in the area is May to because of the size of the irrigation December. Farmers start irrigated agriculture by system that needs control, which is the end of December. Nevertheless, by this time, beyond the capacity of one person to the volume of water flowing to the diversion control. weir has declined substantially to the extent that it cannot support irrigation over the command Overall, the problem has not been fully area of the scheme or dries totally (problem of addressed by all these means due to in built social incompatibility). Similar studies have also defects in the design and implementation of the documented that increasing number of users on adaptive measures. the limited irrigation water has led to scarcity, and even limiting the types of crops grown by farmers (Alula, 2001 and Freeman and S.Silim, 2002). Conflict Management

Coping Measures Water disputes persistently occur between irrigators in the new schemes and upstream The following coping measures were taken traditional irrigators and among irrigators within to over come water scarcity the irrigation systems. Further, results of household interview that the majority (56% in • Changing the duration of water delivery Gibe Lemu and 57.5% in Gambela Terre) of the for TUs in response to the change in sample households have faced conflicts arising quantity of water conveyed into the from water allocation and distribution (table3.8). scheme; Informants reported increasing number of water • Prioritizing crops to be grown; vegetable users in Gibe-Lemu (against the declinigng crops, which require frequent watering, quantity of water conveyed), water scarcity were given priority. Nonetheless, (from the source) and poor water control as irrigators did not observe this cropping major causes. pattern for not strictly implemented to supply water in a reliable manner to The number of claimants of irrigation has grow vegetables. Hence, irrigators increased over time, without being accompanied shifted from vegetables to perennial tree by institutional adaptation, led to competition and conflict over water. Similar findings are

308

demonstrated in studies conducted by Freeman and the multipurpose DAs. The and S.Silim (2002) and Alula (2001). With committees transfer cases of irrigators increasing number of users, conflicts arising who were found guilty of illegal water from water allocation became more common; abstraction to these entities. But they do water management became more problematic not give satisfactory response though the and the interval between watering of plots task of conflict management has been increased almost to "breaking point" (Alula, beyond the legal power and capacity of 2001). ‘Kore Aba Lagas’ the WUA committees; 2. Lack of incentive for the managing Powerful households and rich farmers who grow entities; i. e.; board members of the crops which are not in the priority list, such as WUA have no incentive for the time coffee, chat, and sugarcane, in the middle areas they spent in irrigation management. capture more water by illegal means (more Coupled with resistance it frustrated and serious problem in Gambela Terre); leading to discouraged their commitment to scarcity in the tail-end area and tough conflict undertake strict control of water between the two groups. Nevertheless, it was distribution.. beyond the capacity of ‘Kore Aba Laga’ to be 3. Problematic social relation of power contained. Diversion of the Source Rivers by among water users. Some members, traditional irrigators had also gave rise to especially the powerful households do external water disputes. The local irrigation not observe the group-based rules and do agency mentioned lack of legal frameworks as not usually give consent to be governed the main reason for not addressing the problems. by the WUA committee members. Mollinga (2003) has also proved in his In summary, ‘Kore Aba Lagas’ are ineffective, study in India that socioeconomic reluctant and less committed in taking care of the differentiation (social inconsistency) water rights of members and in resolving among water users had impeded conflicts. Table4.8 shows farmers’ ranking of emergence of viable water users problems that discouraged commitment of the organizations who can undertake water committees. The prominent gaps include: effective water control.

1. Lack of satisfactory support from the

local administrative and legal entities

Table3.7 Farmers' ranking of causes of poor water control and poor conflict management by water committees by irrigation system % of farmers and rank Reasons Gibe Lemu Gambela Terre N % Rank N % Rank WUA-committees are reluctant 7 28 3rd 12 30 3rd Resistance by water users 8 32 2nd 15 37.5 1st Lack of adequate external support** 10 40 1st 13 32.5 2nd Source: Field survey, March 2005

Farmers undertake canal cleaning and system Maintenance of the Irrigation Systems maintenance activities under the leadership of the water committee with the assistance of

309

multipurpose DAs. Most of the time members were functional by the survey date although there contribute labor for maintenance. Maintenance is was no clear evidence whether it had been carried out twice a year and very irregularly in functioning fully or partially. This is because Gambella Terre albeit the O and M manual irrigators in Gibe Lemu are more committed to prepared for the schemes recommends that it maintain and sustain the irrigation system in should be undertaken thrice a year. spite of the severe coordination problem. The most important reason they suggested for farmer In Gibe Lemu the majority (56%) of the commitment was the role of irrigation in the life interviewees stated that maintenance of the of farmers in the area and the high market value structures was very good; 36 percent said it was of horticultural crops produced using irrigation good and only 4 percent said very poor due to accessibility to the good commercial (table4.9). Evidences obtained from the DA opportunity in Bako Town. A review of impacts office and the GSDID also showed that more of irrigation management transfer by Vermillion than 75 percent of the water distribution canals (1997:19) came up with similar results.

Table 3.8. Users’ opinion about maintenance of the schemes

Description Number and percent of irrigators All HHS GIBE-LEMU Gambela Terre

5. Count % Count %

Very good 16 24.6 14 56 2 5 Good 22 33.8 9 36 13 32.5 Acceptable 4 6.2 1 4 3 8 Poor 12 18.5 - - 12 30 Very poor 11 16.9 1 4 10 25 Total 65 100 25 100 40 100 Source: Field Survey, March 2005

In Gambela Terre, conveyance and distribution opinions indicated poor coordination of canal networks deteriorated due to a number of maintenance (92%), breaching of canals reasons. The distribution and conveyance canals (87.2%) to extract water by illegal means and became flat in many areas and pockets of water damage from animals (98.5%) as the major ponds created at many points along the causes of damage and threats to safety of the conveyance and distribution canals (see the irrigation system. Culturally, livestock freely photo below). Results of survey on farmers’ graze over the command area for not all farmers

310

cultivate their irrigable plots uniformly. In turnout; implying technical problems in design addition, turnouts are far a part and not evenly and construction have contributed to the distributed in some areas. Hence, irrigators break deterioration of the scheme, in addition to the canals and extract water where there is no institutional and management weaknesses. Photo: Water pond created on the main water conveyance canal due to damage by livestock and lack of maintenance, Gambela Terre SSIS

Irrigated Agriculture: Livelihood Impacts One method to show the social effect of the and Threats to Feasibility and Sustainability intervention on diversification is through comparison of types of crops cultivated by Impact on Farmers’ Livelihoods farmers before and after irrigation. The types of crops and the number of farmers who grew a Irrigation had contributed towards improvement wide range of horticultural crops has of irrigators’ livelihoods through its effect on substantially increased after irrigation (table3.8). Chi-square analysis also revealed that the crop production. Irrigation brought about change production of potato (P<0.05), onion (P<0.05) in cropping pattern and increased production and and tomato (P<0.05) was significantly different farm income, improved housing and wage labor before and after the introduction of irrigation in employment. Gambela Terre area (table 3.8)

Table3.9 Comparison of agricultural diversification before and after Gibe Lemu (N=25) Gambela Terre (N=40) HHs growing the crop HHs growing the crop Crops grown Before After Before After X2-tatist. N % N % N % N % Maize 22 88 23 92% 15 50 23 23.59 0.835NS Potato 6 12.5 22 875 9 29.0 29 93.5 5.226** Onion 12 36 16 64 6 19.4 25 80.6 4.476** Cabbage 1 - 9 37.5 7 23.3 17 56.7 0.709 NS Pepper 14 58.3 13 54.2 11 36.7 19 63.3 0.660 NS Carrot 1 4.3 7 30.4 2 6.9 14 48.3 2.005 NS Chat 6 26.1 7 30.4 1 3.4 14 48.3 .967 NS Coffee 7 28.0 18 72 5 17.2 23 79.3 1.616 NS Sugarcane 5 20 20 80 1 3.4 10 34.5 .545 NS Mango 6 24 17 68 8 26.7 24 80 .384NS Tomato 3 12 21 84 6 20 21 70 4.802 * Source: Field survey, NS=Non-significant, **=Significant at P<0.05

311

The second most visible impact of irrigation was least. This is owing to inequity in the spatial and temporal diversification of production. In Gibe temporal distribution of irrigation water in the Lemu, the number of households who used to tail-end area. Furthermore, the proportion of grow twice increased from 8% before irrigation irrigators who grow twice a year was higher in to 88% after irrigation and in Gambela Terre, Gibe Lemu (88%) compared to Gambela Terre from 2.5% before to 70% after irrigation. Results (70%). The difference is attributed to more also revealed that it is significantly different severe problem of scarcity and unreliability of between locations in Gambela Terre (X2-Prob. = water and farmers biased ness towards rain-fed 0.000) and farmers in the tail-end area benefited agriculture in Gambela Terre (see also fig3.3).

Figure3.1 Comparison of agricultural diversification before and after irrigation in GLSSIS 25

%age of growers

before After

123456789101112 crops

N.B:1=maize,2-potato, 3=onion,4=cabbage,5pepper,6carrot,7=chat,8=coffee,forage,10=sugacane,11=mango and 12-mango

Alongside diversification and intensification of Findings of the study also revealed that the crop production, SSI had a positive impact on increased income from the sale of crops the income of farm households in 2004/05. produced using irrigation has enabled irrigators However, ANOVA showed that there was a to invest in household assets. Table3.11 shows significant difference (F=13.47, P<0.0001) in the that 17 corrugated iron roofed and 9 grass roofed net income of households between irrigation houses were built through income from system and between locations. The average irrigation. The number of corrugated iron roofed household net income from all sources in dwellings built in Gibe-Lemu is 3 times as large 2004/05 was relatively higher in Gibe Lemu as Gambella-Terre. In addition, the number of (Table3.8). This could possibly be due to the dwellings built by irrigators in the tail-end areas relatively better supply of water, better water of both irrigation systems was low as compared management and more commitment of farmers to the other two water levels. to irrigated agriculture in Gibe Lemu.

312

Figure3.2 comparison of production diversification before and after irrigation in GTSSIS

%age of growers 15 before After

0 123 4 5 6 7 8 9 101112 crops

NB:1=maize,2-potato, 3=onion,4=cabbage,5pepper,6carrot,7=chat,8=coffee,forage,10=sugacane,11=mango and 12-mango Figure3.3 Comparison of cropping intensity before and after irrigation in Gibe Lemu and Gambella-Terre

%age of irrigators 25 GLSSIS GTSSIS 20 Total

0 before after before after once ayear twice ayear Period

313

Table3.10 Household net income (Birr) from irrigated agriculture in 2004/05 by irrigation system, location and sex Name of the Location of farm plots Sex of HHH Statistic irrigation system Head-end Middle Tail Total Male Female Gibe Lemu Mean 946.23 797 1180 949.1 1011 534 N 8 9 8 25 22 3 St. Dev. 1145.9 1141 1727 1327 1396 586.02 Gambela Terre Mean 394.61 624 276 351 372.16 177.5 N 13 13 14 40 33 6 St. Dev. 579.3 509.6 370 484.3 504.62 199.2 F=13.47, P<0.000 Source: Field Survey, Gibe Lemu and Gambela Terre, March-April, 2005

The use of hired causal (seasonal) and permanent permanent laborers respectively (Table3.14). A labor was low in both irrigation systems as labor sum of birr 6705 was paid for seasonal and is not a major constraint. Irrigation created a permanent hired labor (in 2004/05). limited number, 7 and 218, of employment opportunities (in 2004/05) for causal and

Table3.11 Roof materials of dwellings built through income from irrigation Number and value of houses built Irrigation Roof materials of Number built Value of the houses built system respondents house (performance) Mean N SD Gibe Lemu Corrugated iron 13 3801 10 2534.512 Grass roof house 4 346.25 4 57.63 Total 17 3704.57 14 3462.31 Gambella- Corrugated iron 4 4975.4 4 3386.62 Terre Grass roof house 5 380 5 164.32 Total 9 2422.22 9 3190.52 Source:

Table 3.14 Employment impact of irrigation and cash paid for laborers in 2004/05 Over all GLSSIS GTSSIS Description Statistic Cash paid No of Total cash No of Total cash N (average) laborers paid laborers paid Permanent Sum 7 2805 3 1205 4 1600 labor Mean 1 467.5 1 401 1 533.33 N 7 6 3 3 4 3 SD - 382 0.00 288.67 0.00 166.04 Causal Sum 218 3900 38 3330 180 570 labor Mean 19.8 433.33 6.33 832.5 36 114 N 11 9 6 4 5 5 SD 38.2 970 11.62 1445.5 53.55 166.34

314

Risks to Feasibility and Sustainability of committed less to irrigated agriculture. In Irrigated Agriculture addition, the actual irrigated area was small compared to the potential (150ha) and it has The average plot size farmers allocated for continuously been declining during 2001- irrigation occupies only a small portion while the 2004/2005 (table3.15). Irrigation has totally land allocated for rain-fed agriculture represents collapsed in tail-end area that constitutes more the lion's share in Gambela Terre (appendix than 53% of the total command area. table4); implying farmers in Gambela Terre are

Table3.15 Estimates of actual irrigated area (ha) and its trend, 2001-2004/05)

IRRIGATED AREA (HA) (1994) 2001/02 (1995) 2002/03 (1996) 2004/05 IRRIGATI IRRIGAB 2003/04 ON LE LAND ARE % OF IRRIGATE % OF IRRIG % OF IRRIGATE % OF SYSTEM (HA) A TOTAL D AREA TOTAL ATED TOTA D AREA TOTAL (HA) AREA L GIBE 113 76.05 67.30 80 70.80 78 60.02 80.0 70.80 LEMU GAMBEL 150 58.5 39 56.75 37.83 69.5 52.4 48.27 32.18 A TERRE TOTAL 263 134.5 51.16 136.75 52 130.4 49.58 1324 48.77 5 Source: Gobu Seyo District Irrigation Desk (GSDID)

The impact of irrigation projects on diversified 2004/05, while the area planted to perennial and intensive irrigated horticulture and increased crops such as sugarcane and coffee had been production not maintained for long for increasing in Gibe Lemu (Figure3.4). In investment in these crops has become a risky Gambela Terre, irrigated area of maize and chile business due to frequent crop failures. In Gibe had been declining; while irrigated area of Lemu and Gambela Terre, 92% and 84% of the sugarcane and coffee was increasing during sample households have faced crop failure. 2001/02-2004/05 (figure3.5). Hence, the majority of irrigators do not plant their irrigable plots to these fast growing Farmers’ perception and ranking of cause shows vegetable crops regularly. They shifted to that water shortage, unreliable access to water perennial tree crops such as: a) Sugarcane, chat, and prevalence of vegetable diseases (Due to coffee and banana, in Gibe Lemu, b) Coffee, lack of adaptive seeds of crop varieties and chat, 'Gesho', mango in Gambela Terre and to knowledge of irrigation agronomy) were the cereal production (mono-cropping) under rain- prime constraints that threatened irrigated fed. Results of trend estimate revealed the actual agriculture and dictated the change in cropping irrigated area of the major vegetable crops pattern (Tables3.16). Water is unpredictable and (potato, tomato and chile) and maize has scarce due to increasingly become shrunk during 2001/02-

315

Figure3.4 Estimates of the trend in irrigated area of major crops in GIBE LEMU, 2001/02-2004/05 45

Area(ha)25

0 2001/02 2002/03 2003/04 2004/05 Year

maize Sugarcane" potato tomato onion hot pepper coffee mango banana carrot

Figure3.5 Estimates the trend in irrigated area of major crops in GAMBELA 30 TERRE

maize 20 sugarcane potato Area(ha) tomato onion 15 hot pepper coffee mango banana 10 carrot

0 2001/02 2002/03 2003/04 2004/05 Year due to problems embedded in water water from the source (hydraulic problem). management, definition of water rights, Results of similar work by Alula (2001) also enforcing group-based rules and social relations showed that unreliability of water supply and among water users and decline in the amount of increased interval between watering of plots due

316

to scarcity and poor ware management, affected the type of crop that could be grown, even apparently limiting the practice of vegetable production. Table3.16 Farmers' ranking of the reasons for under use of their irrigable land Statistics Water Unreliable Shortage of Vegetable scarcity access to labor diseases water GLSSIS N 7 9 4 6 % of farmers 28 36 12 Rank 2nd 1st 4th 3rd GTSSIS N 18 12 5 10 % of farmers 45 30 12.5 Rank 1st 2nd 4th 3rd Source: Field survey

Double cropping has been less feasible and season in Gambela Terre; Crop damage from unsustainable more in Gambela Terre. In livestock discouraged interested farmers to addition to the gap in institutional development engage them selves in irrigated farming. and support system and management and water scarcity problems, the various groups of Conclusion and Suggested Policy Options informants reported that double cropping is less feasible owing to the following socio-cultural The study used the socio-technical approach to problems in the area: irrigation technologies as conceptual frame in examining institutions, management practices • Farmers have limited or no experience in and challenges. The following conclusions are irrigation before arrival of the new project. drown from findings of the study using the Almost all (97.5%) of them own large area theoretical notions like context, social of rain-fed land as an option. It was initiated requirement for use and social effects: primarily to promote the collectivist interest of the Derg; 4.1. Although it is relatively better in Gibe • The second problem was incompatibility Lemu, there was poor record of accomplishments between the new cropping pattern and the in water management in terms of adequacy, indigenous cropping pattern and the projects timeliness and equity in the supply of water, growing season and farmers’ growing season conflict management and system maintenance. despite the fact that compatibility is one of Access to adequate and reliable irrigation water the social requirements for successful is more unlikely if the farmer's irrigable plot is in irrigation. Maize planted shortly after the tail-end area (more serious problem in harvesting vegetable crops is affected by Gambela Terre). It was mainly because of the disease owing to the short time frame lack of the social conditions (Social between harvesting of vegetable crops and requirements) of possibility for successful planting maize, and lack of cropping irrigation. The main irrigation agency has weak sequence studied and specifically management capacity to support WUA recommended for the irrigation system. management of irrigation although it is a Horst (1998: Woldeab, 2003) and FAO necessary condition for efficient and lasting (1986) also write, ‘incompatibility between irrigation management. The WUAs are not the project cropping pattern and farmers’ properly organized to run irrigation management. cropping pattern could lead to Users have problematic social relation. Enabling underutilization of irrigation water’ legal systems of land water rights institutions are • A culture of open grazing during the dry non-existent at the operational level. Efforts of

317

stakeholders were not harnessed in irrigation market stimulus (access and the good administration. These problems in turn have commercial opportunity at Bako Town), shortage drastically affected management and utilization of adequate rain-fed land and the problem of of the developed resources. Therefore, policies landlessness, experience and interest of farmers for future investment in smallholder irrigation in irrigation and the role of irrigation in the life development and for rehabilitation of the of farmers. This shows that irrigation should find irrigation systems considered by this study its appropriate socioeconomic and institutional should give due consideration to averting these location to work effectively. The policy problems. implication is that:

4.2 Sustainable water rights of users not ensured • Small-scale irrigation should be promoted in the irrigation systems. Water dispute (internal where it is most demanded; and; and external) is a major and undressed problem • Farmers’ priorities and interest, in both irrigation systems. It could no be compatibility of irrigation to the socio- addressed only through the general constitutional cultural environment and farming system of choice rule and the informal bylaws of water the area and the opportunities (cultural, users. The problem is found to be very complex institutional, economic) for irrigation should and beyond the capacity of users’ organizations be understood before intervention. and local and village level administrative and legal entities. The major constraints are 1) there 4.5. Irrigation has been a success in the first few has been no enabling legal system (operational years of project implementation. It has regulations) both at District and grassroots levels positively contributed towards increased which clearly define the water rights of diversification and intensification of production downstream and upstream users and rules which and livelihood improvement. Nonetheless, many govern construction of new diversions; and 2) farmers, what Engel (1997) and Mollinga (2003) lack of clear definition of responsibilities (who call `the human agents` did not maintain these should do what) for dealing with the problem. To practices for long. They, do not practice irrigated ensure sustainable water rights of irrigators, vegetable production regularly, discontinued it, facilitate shared use of water by downstream and shifted to perennial tree crops or returned to the upstream users and improve water management former cereal/mono-crop production under-rain- there is an urgent need for creating formal fed. The constraints that discouraged farmers operational regulation. participation were among others institutional and organizational weaknesses that led to poor 4.3. Such technical resources as improved seed irrigation management or the lack of what Engel (technology) that is adaptive to the situation of (997) calls `the social organization` to coordinate irrigation and knowledge of irrigation (extension and manage the irrigation systems. Therefore, service and capacity building for irrigators) have adequate institutional and organizational not been met. This problem has been a major development is crucial to enhance effectiveness impediment to feasibility of irrigation. Therefore, of irrigation promotion and to ensure policies for input supply, technology sustainability of the benefits of irrigation and the development (agricultural research) and rural irrigation systems. extension have to be adjusted to meet these requirements of irrigated agriculture in the 4.6. Expansion of traditional irrigation in the irrigation systems. upstream areas of the rivers that are water sources for the schemes is a major threat to 4.4. In spite of the lack of strong system sustainability of the irrigation systems. There has management, water scarcity and unreliability and been continuous decline in the quantity of water organizational and institutional problems, conveyed into the schemes. This led to acceptable commitment of farmers was observed progressive degeneration and collapse of and the impact of the implemented SSI on irrigation in the tail-end area of Gambela Terre, farmers' livelihood was also relatively higher in covering more than 53% of the command area. Gibe Lemu. This could possibly be due to Therefore, there is an urgent need for addressing

318

the problem through establishing and enforcing between the projects growing season and the necessary institutional/legal framework. the indigenous growing season (in Gambella Terre) 7. A culture of open grazing during the dry 4.7. One of the major factors for season in Gambela Terre underperformance of the Gambela Terre SSIS is water scarcity. Therefore, future fate Reference (sustainability) of the scheme should be determined through detailed hydrological study AESE (1995) Food Security, Nutrition and on the water source before embarking upon any Poverty Alleviation in Ethiopia: Problem and investment aiming at rehabilitation of the Prospects. Proceeding of the Inaugural and irrigation system. First Annual Conference of the Agricultural Economics Society of Ethiopia, 8-10June 4.8. In nut shell, change to sustainable 1995, Addis Ababa, Ethiopia. diversified irrigated agriculture and to double cropping not met in both irrigation systems. The Alula, P. 2001. Institutionalization of Irrigation challenges and sustainability constraints that in South Wello, In: Alula, Pankhurst (2001) need urgent intervention through developing and (Ed.). Institutions for Natural Resource enforcing appropriate institutional support Management: Thematic Briefings, FSS systems at all level, from the apex and grassroots (Ethiopia) and University of Sussex (UK). levels, include: Bagadion, Benjamin and Frances F. Korten 1. Institutional and management limitations (2004) Developing Irrigators Organizations: that led to scarcity and unreliability of Learning Process Approach. In: World Bank water (2004). Putting People First: Sociological 2. Prevalence of vegetable diseases because Variables in Rural Development farmers have not regularly been supplied with improved and adaptive seeds of Blank, H.G. Clifford M. Mutero and Hammord vegetable crops that work under Murray – Rust. 2002 (Eds). The Chaning irrigation; because extension service and Face of Irrigation in Kenya: Opportunities in put supply policy is biased both in for Anticipating Change in Eastern and terms of supply and timing of supply to Southern Africa. Colombo, Sri Manka: rain-fed agriculture. It is not adjusted to International Water Management Institute. meet the requirements of SSI at the grassroots level (the role of policy) Cernea, M. (2004) (Edit). Putting People First: 3. Expansion of traditional irrigation in the Sociological Variables in Rural upstream areas of the rivers that are Development. Washington, D.C: The World water sources for the schemes leading to Bank Group. water scarcity in the schemes. Nevertheless, there has been no enabling CTA(2001) Water Management: Proceedings of legal framework that facilitate the shared a CTA Seminar. Cordoba, Spain, 20-25 use of water by the two groups September 1999. Aniane, France: Louma 4. Weak institutional capacity of the local Productions. state irrigation agency to support decentralized management of SSI EWZFEDD (2004) Socio-Economic 5. Weak linkage among stakeholders of SSI Profile of Gobu Seyo District, Nekemte management both at the District and scheme levels; East Wellega Zone Water, Mineral and Energy 6. Problem of social incompatibility Resources Development Department between the new cropping pattern and (1998) Design Report of Gibe Lemu the indigenous cropping pattern and Irrigation Project.

319

Engel, P.G.H (1997) The Social Organization of North, C.D. (1990) Institutions, Institutional Innovation: A focus on Stakeholder Change and Economic Performance. New Interaction. Amsterdam: Royal Tropical York Institute (KIT). Odi (1995) Recent Trends in Irrigation FAO (1995) Irrigation in Africa in Figures. Management: Changing Directions for the FAO: Rome. Public Sector, Natural Resource Perspectives; Numbers, London: UK. FAO (1986) Irrigation in Africa South of the OIDA(2000) Smallholders Modern Irrigation Sahara. FAO Investment Center Technical Schemes Evaluation Report. Finfinne. Paper No.5, FAO: Rome. Oromia Rural Land Use and FAO (2000) Socioeconomic Impacts of Administration Proclamation Number Smallholders Irrigation Development in 56/2002. Finfinne, August 16, 2002. Zimbabwe. SAFR, Harare. Peter, H. Stern (1998) Operations and FDRE (2002) Food Security Strategy (FSS). Maintenance of Small-Scale Irrigation Addis Ababa, Ethiopia Schemes. UK: International Technology Publications Ltd. Fiona, H., John Thompson and Jules N. Prety (1996) Sustainable Agriculture and Food Peter, P. Mollinga (2003) On the Water Front: Security in East and Southern Africa: An Water Distribution, Technology and Empirical Analysis of Current Initiatives and Agriculture Change in a South Indian Canal Review of the Literature. Stockholm: Irrigation System. Orient Longman Private Swedish International Development Limited: New Delhhi. Cooperation Agency (SIDA). Vermillion, D.L. (1997) Impacts of Gibe Irrigation Project Design Note by Korean Irrigation Management Transfer: A Review Design Team (1990) of the Evidence, Research Report 11. Colombo, Sri Lanka: International Irrigation Lemma, Dinku (2004) Smallholders Irrigation Management Institute (IIMI). Practices and Issues of Community Management: The Case of Two Irrigation Webb, P. (1991) When Projects Collapse: Systems in Eastern Oromia, Addis Ababa, Irrigation Failure in the Gambia from a Ethiopia (MA Thesis) Household Perspective in: Journal of International Development Vol.3 No. 4, July MoWR (2000) Ethiopian Water Resources 1991. Management Policy Document. Woldeab, Teshome (2003) Irrigation Practices, State Intervention and Farmers Life-worlds Nigussie, Taffesse (2002) The Role of in Drought Prone Tigray, Ethiopia (PhD Irrigation Development in Enhancing Dissertation) Household Food Security: A Case Study of Three Small-Scale Irrigation Schemes in World Bank (2000) Case Studies in Southern Nations, Nationalities and Peoples' Participatory Irrigation Management. WBI Region. MA Thesis, Addis Ababa Learning Resources Series. Washington, University. D.C: The World Bank.

320

APPENDIX

Tables Appendix table1 Water users’ opinion about water distribution by irrigation system and location of farm plots Item Response %age of water users giving the opinion by irrigation system and (yes/No) location Gibe Lemu (N=25) Gambela Terre (N=40) Head Middle Tail X2- Head Middle Tail X2- Stati. Stat.

Enough water is Yes 12 8 - 1.32NS 7.5 2.5 - 10.6** obtained No 20 28 32 25 30 35

Water is received when Yes 16 8 - 4.5NS 15 5 - 9.8** needed No 16 28 32 17.5 27.5 35 Water distribution is Yes 16 - - 5.5* 12.5 5 - 9.6** equitable No 16 36 32 20 27.5 35 Source: Field Survey, NS= Non-significant, *= significant at P<0.1, **=Significant at p<0.005

Appendix table2 Major crops cultivated, estimates of irrigated area and trends (2001/02-2004/05), Gibe Lemu and Gambela Terre SSIS

DESCRIP Irrigated area in ha (2001/02-2004/05) TION OF GLSSIS GTSSIS CROPS 2001.02 2002/03 2003/04 2004/05 2001/02 2002/03 2003/04 2004/05 GROWN (1994) (1995) (1990) (1997) (1994) (1995) (1996) (1997) MAIZE 12 8 2.5 2.06 12.38 2.28 2 0.125 SUGARC 10.60 17 20 40 0.27 0.8 2.25 4 POTATO 30.00 24 13.25 7.24 26.58 16.55 14.60 10 TOMAT 4.85 3.5 2 1.76 1.25 0.61 1.00 0.7 O ONION 1.44 1 0.90 1.50 4.64 3.51 2.00 3.00 PEPPER 2.22 7.3 4.21 2.05 7.33 3.63 4.21 1 COFFEE 3.50 3.00 3 7.01 5.11 5.60 5.92 7.00 PAWPA 4.00 - 0.5 0.12 0.08 0.05 0.03 - MANGO 2.52 2.6 2.92 2.92 5.00 2.00 3.3 6.50 BANAN 4.00 3.12 2.93 3 0.74 1.8 1.20 1 A Source: Gobu Seyo Wereda Irrigation Desk

321

Appendix Table3 ANOVA of household net income from irrigation in 204/05 Sum of df Mean square F Sig. squares Amount Between groups 13570199 1 13570198.64 13.47 0.000 (Br.)*irrigation (combined) sys*location Within groups 1.51E+08 150 1007749.71 Total 1.65E+08 151 Amount Between groups 13570199 1 13570198.64 13.47 0.000 (Br.)*irrigation (combined) sys*Sex Within groups 1.51E+08 150 1007749.71 Total 1.65E+08 151

Appendix table 4 Average land holding by type of use Type of use Average plot size per household Gibe Lemu Gambela Terre Mean N SD Mean N SD Total land size (ha) 2.54 25 3.05 3.16 40 1.77 Irrigable area 1.08 20** 0.72 0.72 39** 0.99 Area under rain-fed 1.57 25 1.73 2.5 40 1.27 Source: Field survey, the sign `**` implies the rest sample irrigators (five in Gibe Lemu and one in Gambela Terre) do not have own irrigable land

Appendix Table5 Vegetable growers who faced crop failure, by irrigation system and location of plots Do you grow Ever faced problem of crop failure (Yes/No) vegetables every By location of farm plots (%) Irrigation system year using Head-end Middle Tail-end All HHs irrigation? (%)

Gibe Lemu (N=25) Yes 8 92 24 36 32 No 92 8 - 4 4

Gambela Terre(N=40) Yes 77.5 84.2 26.3 26.3 31.6 No 20 15.6 10.5 5.3 - Source: Field survey

322

Water Rights and the Processes of Negotiations among Irrigators in West Shewa Zone: The Case of Indris Scheme in Toke Kutaye District

Tesfaye Zeleke Axum University [email protected]

Abstract Conflicts in connection to irrigation water use and rights that have escalated over Though water rights are at the core of exploiting water resources for irrigation Years have been attributed to the decline in purposes, trivial concerns were offered to the the volume of water resources, institutional case of Indris irrigation scheme in Toke failures to address the causes adequately, Kutaye district in West Shewa. The historical week observance on governing water right background and development of the scheme rules and increasing demand of users. As a has been presented in a contentious manner. result, negotiation processes aiming to settle The augmenting number of competitors too disputes were repeatedly initiated either by paved the way for conflicts that recurrently users, committee members (elders) or courts. erupt out and inevitably lead to a succession The procedures pursed to narrow competing of negotiation processes. With the inception interests around the scheme confirmed the of such missing gaps, this research aimed to pragmatic applicability of the central scrutinize water rights and the processes of arguments of both cyclical and negotiations among irrigators along Indris developmental models of negotiation modern scheme, in Toke Kutaye district. To processes discussed comprehensively by maintain this objective, qualitative research Gulliver. methods were predominantly utilized as the main data generating tools in the field. Thus, in the face of increasing demands on a declining water resource, the findings of this The findings of the research depicted that research revealed out that concerned Indris scheme marked three significant individuals or relevant institutions need to phases in its historical development. In these exert further endeavor on the formulation of phases, explorations pertaining to water water policies that clearly stipulate specific rights and processes of negotiations were irrigation water entitlements of users. found to be at their immature ground. While Enforcements on the frame of references set the elements of the riparian doctrine of water on the water manual need to be rigorously rights preponderated during its initial phase, checked on practical implementations. the components of appropriative doctrine Awareness buildings on irrigation water right pronounced more at its middle age. A mix of claims, promotion of negotiated approaches ingredients from both doctrines interwoven in disputes and accentuation on customary with certain extra requirements determined rules of resource use constituted the the water right access of users since the dimensions seeking meticulous conversion of the scheme into a modern considerations in prospect. style. Multiple water right rules emanating both from the customary and formal water Introduction. acts have co-existed to direct the actions of users. In this regard, the theoretical Ethiopia being predominantly an agricultural orientations of legal pluralism in water right country, half of the GDP, close to 90% of paradigms proved to coincide with the export earnings and about 85% of people's pragmatic contexts of water users from the livelihood sources has come from agriculture scheme. (CRDA, 1994:20). Irrigated agriculture, complementary to the conventional rain-fed agriculture, has a history of more than one century in some parts of the country

323

(Dessalegn, 1999: 10; Woldeab, 2003:25). permanently use or transfer to the other Some indigenous schemes are said to have party. The third dimension hold a concern existed since the reign of Menelik II with the competing nature of users over (National Irrigation Policy Discussion Paper, irrigation water, associated conflicts and the 1990:2). successive chains of negotiation processes held to evade disputes. All the The impulse to promote irrigation schemes aforementioned dimensions have not been has been triggered by the recurrently investigated around Indris modern irrigation occurring droughts and worsened food scheme located in Toke Kutaye district of insecurity situations (FAO, 2005). Tsegaye West Shewa Zone. (1991:2) elicited though irrigation developments were taken to be among Therefore, the main objective of the paper is optional mechanisms to cope up with the to examine water rights and the processes of prevailing poverty conditions, productivity negotiations among irrigators diverting water through such systems has failed to meet the from Indris scheme located in Toke Kutaye anticipated targets being constrained by district of West Shewa Zone. It also multiple factors. Studies by Mokonnen specifically looks at: (1992; FAO,1978; Lemma,1994) for ¾ The stages in the historical instance indicated, mismanagement of background and development of irrigation practices could result in the Indris modern scheme. problem of disputes, soil salinity, water ¾ The nature of water rights and logging, canal seepage and expose people to processes of negotiations prior to various diseases. and post of 1986 E.C. ¾ The rules and by-laws that govern In West Shewa Zone of Oromiya regional water use rights, distribution and state where this research endeavored to management aspects. scrutinize water rights and the processes of ¾ How decisions and negotiations negotiations, several irrigation schemes have been made for water access and operate both in indigenous manner and rights in light with interactive factors application of modern irrigation like gender, economic status and technologies. Indris modern irrigation power. scheme is among the scaled up ones and ¾ The significances of water right located at a distance of 2 Km south of Guder negotiations for users (in relation to town. Initially, the scheme had been livelihood improvement). operating in an indigenous manner. It was ¾ Highlight conflict settlement promoted into modern system completely in mechanisms adopted by water users. 1986 E.C by the financial assistance of European Economic Commission. To address the above stated objectives and keep narrow the missing gaps in the three Despite access to water rights and processes thematic dimensions noticed around the of negotiations over irrigation water scheme, a qualitative research method was constrain the development of the practice in employed to gather first hand data. Hence, several regards; these concepts have been interviews, observations, and focus group overlooked greatly by scholars or any other discussions constituted principal methods in relevant institutions. In the case of this the field work. The application of all these particular paper, problems stemming from methods makes it easier to triangulate and three dimensions are apparent and do cross verify the generated data. To visualize override in connection to Indris modern certain themes with deeper insights at their irrigation scheme. On the first place, naturalistic settings in the field, photography insignificant searches have been made on the has also been used. As a whole, 53 (fifty historical development of the scheme. Even three) persons were contacted fro interviews the prevailing data regarding its historical and 21(twenty-one) individuals took part in background were not only scanty but also the focus group discussions. presented in a contentious way among the water users themselves. Secondly, what rests Secondary sources including books, journals, at the core of each resource exploitation research papers and official records were constitutes issues of the right to hold it reviewed to substantiate the data obtained

324 through primary means. preponderate instead of it. As a result, the water flowing through the ditches blocked Thus, the primary data interpreted in this back to the main river. research was generated through a field work that covered a period of time ranging As computed against its years of foundations between March 06-07-1999 E.C and April (legendary sources trace back as far as to the 03-08-1999 E.C over 30 days. Preceding the last decade of 19th century), the over all main field work, preliminary field visits were social and physical infrastructures of the made twice during the months of December district in general and Guder in particular and January 1999 E.C. At that state, legality require an enthusiastic need of integrated to field entry and creation of rapports with efforts. It was reported that sectors like district agricultural workers and few education, health, water supply and committee members were assured. On the sanitation, and communication are first preliminary field visit, the surroundings components that demonstrated encouraging of Indris modern scheme was observed. In progresses. the second round, the potential research settings were marked out and visited. Figures pertaining to the population number of the district should be looked suspiciously. Description of the Study Settings. This is because no consistent census has been conducted since the separation of this Toke Kutaye district constituted one of the district from Ambo. Any how, following the 21 Weredas in west Shewa Zone. It is a separation, the Economic Planning and newly established district for administrative Development bureau of Toke Kutaye district purposes being as a sub division of Ambo has initiated a sort of pilot assessment to district. Guder, some 12 kilometers west of estimate the total population. Based on the Ambo town, serves as the chief socio- estimated result, the total population residing economic, administrative, political and in both rural and urban Kebeles of the district cultural capital of the district. It is situated at is about 122,857 out of which males account a distance of about 137 kilometers away for 58,828 (47.88 %) and females constitute from Addis Ababa on the Addis Ababa- 64,029(52.11 %) (Economic Planning and Nekemte main road. Development document of Toke Kutaye district, 1999: 11). The district has a total area of 78,886.77 hectares (which is nearly about 788.88 in Sq Several ethnic minorities with the dominant Kilometers). Out of the total land only Oromo ethnic group co-exist in the district. 1,384.8 hectare has been under irrigated Official guesses state that the Oromos cover agriculture (Strategic Planning and more than 95 % of the ethnic composition. Management document of Toke Kutaye The Amhara, Gurages and few Tigrians do district, 1999:10). dwell in the district (Economic Planning and Development document of Toke Kutaye The climatic situation of the district is district, 1999: 13). categorized into three major divisions: the cool, the warm-temperate and the hot. These Three chief religions have been apparent in climatic divisions cover 20.99 %, 51.31 % the district: namely Christianity, Islamic and and 27.7o %, respectively. The selected Indigenous (i.e. people’s adherence to peasant associations and specific research Waqefaana-believe in one supreme God) sites chosen for this research fit in to the Christianity, especially the Orthodox sect has warm climatic division (Ibid, 1999: 14). showed a substantial dominance in the area (Ibid: 17). The rainfall distribution is bimodal. While the dry season normally lasts from the The economic source of the district depends months of October to February; the main on agriculture and its produces. Agriculture rains are received from May to September. accounted for more than 90% of the The highest concentration of the rain falls economy of the district. Also, irrigated during the months of June to August. At this agriculture covers almost 18 peasant season, irrigation activities have declined and associations out of the 33 PAs in the district. the rain fed agricultural productions The chief crops produced in the district

325 include Teff, Wheat, Barely, Maize, File forms middle users, Kilinto Sorghum, Beans, and Oil seeds like Nug. constitutes the lower beneficiaries of Irrigated agriculture generates a significant the scheme. amount of income from the vegetables and 9 These sites have also been potential fruits produced at least twice per year areas where both Holeta and Bako (Economic Planning and Development Agricultural Research Centers document of Toke Kutaye district, 1999: 23). conducted pilot demonstrations repetitively. To have the bird’s eye-view of irrigation 9 The accessibility of larger practices in the district, indigenous irrigation concentrations of experienced systems have said to operate since time irrigators in the villages was another immemorial in various sites in the district. justification. Information extracted from the district 9 Moreover, the selected villages are agricultural desk revealed out that a sum of comparably accessible for over 25 major and minor rivers have been transportations. Besides, the villages diverted for irrigation purposes. While there are well suited for fieldwork to are 6 indigenously operating irrigation sites explore deep information as each of in the district, comparably the modern ones them followed a nucleated settlement make up 8 in figures. All the schemes pattern in their respective localities. irrigated via indigenous or modern 9 Lastly, as a social anthropologist, the techniques fall below or equals to 200 social dramas reflected over water hectares, except the case of Indris which use rights and process of covers more than 200 hectares. According to negotiations observed among upper Dessalegn’s (1999:10) classification of and lower sites or between the irrigation schemes depending on their size, villages and the Agricultural operation and management; the irrigation Training Center created a passionate schemes pragmatically functioning in the feeling in the researcher’s district can be categorized as small scale understanding to offer priorities for category while exceptionally Indris modern these villages. irrigation scheme falls into the medium scale. Review of Literatures and Theoretical Frameworks. To depict the specific irrigation settings, peasants from 5 PAs (namely Imela Dawo, A brief review of the literature on water Dale Dawe, Ajo Bedo, Dhaga File and rights and the processes of negotiations Kilinto) do divert water for irrigation usages portraying the experiences of certain from Indris scheme. Additionally, institutes developing nations would be made in the like Agricultural Training Center, Schools first part pursued by certain cases in Ethiopia and Hormat Engineering Factory claimed at the last part. water for irrigation and other usages. Having extensively discussed with committee To begin with, socio-anthropological issues members, Development workers and are embedded in the operation of all experienced elders, the researcher have irrigation systems, small or large: people selected three irrigation practicing villages organize socially in order to secure water, out of three (3) PAs identified as Selam transport it, divide into usable shares, enforce Sefer, Dhaga File and Kilinto. The rules for its distribution, pay for it and Agricultural Training Center was also dispose of unused portions (Cernea, 1991: included in the study as there have been 43). frequent reactions and interactions between Vermillion (2000:57) quoting (Arriens et. al the peasants and the training institute. These 1996; Secker, 1996) asserted that with rising sites were selected mainly because; populations and diversified economies, competition for water is rapidly intensified in 9 The sites constituted upper, middle many developing countries, especially in and lower categories of beneficiaries Asia. Such conflicts over water resources are of Indris scheme. Selam Sefer and further aggravated by the social inequality, the Agricultural Training Centers economic marginalization and poverty represent upper users. While Dhaga (Blank et. al, 2002:113; FAO, 2005:5).

326

or use of water resources. Water rights may Blank et al (2002:123) explained the multiple also be constructed as contractual rights or sources of conflict over water use in the may be based on bodies of norms other than Upper Ewaso Ng’iro North Basin of Kenya, domestic legislations-namely customary law attributing to causal factors linked with water (Ibid). scarcity, inequitable water allocation and distribution, election of representative water Bruns and Meinzen-Dick (2000: 203) have users, failure to observe water by-laws. offered clues on diverse levels of rights in a Water conflicts associated with scarcity are specific source, such as ownership rights, attributed to over abstraction of water in the rights to participate in decision making upper reaches and latent conflicts are process (including decisions concerning attributed to inequalities inherent in social, allocation of water), rights to use without cultural, economic and political disparities rights to participate in decision making among the stakeholders. process, rights which may or may not be transferred, rights to use only for a specific There is a repetitive claim that underlines season or purpose, individual rights and laws and traditions controlling water use in community rights. developing countries are often inadequate, unsuitable introductions, ignored or Boelens and Davila (1998:88) presented that unenforceable. Fair, rigorous and swift in a given region, it is not unusual for several enforcement is important in maintaining or mechanisms to operate simultaneously and it improving adherence to water use laws and is also common to find mixtures of peasant rules (Barrrow, 1987: 70). and governmental mechanisms: ¾ Concession of water use rights, Many developing countries have non- granted by the state administration to western legal systems. Reflecting this individuals or groups of applicants; argument, Blank et al (2002:278) indicated ¾ Granting of formal or informal titles the Kenyan experience presenting that in pre- over socio-territorial waters by their colonial times, management of water was an inhabitants; integral part of the overall customary laws ¾ Agreements for permanent transfer and behavioral norms of each tribal society. of water rights from one right-holder Access to water was guaranteed for each to another such as in the case of individual by virtue of affiliation with a purchase and sale, rental inheritance, specific community (e.g. a tribe or clan) and barter or gifting; water use was regulated by the political ¾ Acquisition of rights and access to leaders of each community (chiefs, elders, water by force; in many regions of clan leaders). the world, power groups have expropriated water by coercive force Bruns and Meinzen- Dick (2000) denoted from peasants and indigenous water rights are not just an analytical peoples. abstraction. The term water right is broad including diverse kinds of and levels of According to Boelens and Davila (1998: 29) rights. Clearing it, Boelens and Davila there may be multiple bases for water claims. (1998:87) have discussed water rights as The two most widely recognized doctrines social relationships among humans and not for water rights are based on ownership or only between the user and the water; thus, possession of land along rivers, streams or they are rooted in the other components of over aquifers (riparian rights) and claims the peasant community’s normative system. based on historic water usage (prior appropriation). The chief features of riparian Water rights comprise formal rights rights embrace: embedded in official tittles, permits and Gives equal rights of use to owners seasonal irrigation schedules, less formal of land which borders on or touches rights based on customary patterns and rights a stream or across which a stream implicit in social norms and local practices flows. (Bruns and Meinzen- Dick, 2000:28; Cotula A riparian right is attached to land 2006: 10). Thus, water rights are considered ownership - a user can take up the as legal entitlements for the abstraction and/ right to use water at any time even if

327

he/ she had not done so before and to Negotiated approaches are a subset of the do so affects existing users. larger set of processes through which The right is usufructuary i.e. the disputes are waged. Disputants choose where owner does not own the water (the to pursue their claims, shopping among resource it self can belong to the available forums to deal with water conflicts. state or some other authority) only Negotiation among disputants can often the right to use it. generate more creative and appropriate solutions than those imposed by a court or One who enjoys riparian rights, therefore, agency decisions. Mediations, facilitations should receive flows from upstream land and convening forums are among ways in owners with out material change in amount which third parties can contribute more or quality and should ensure that down productive negotiation (Ibid). stream owners enjoy the same. On the contrary, according to the appropriative Generally speaking, Gulliver (1979) wrote doctrine the first settler or user of water from negotiations are vital to reconcile conflicts a stream acquires a right to continue the use among diverse interest groups over resource of that portion of water needed for the use. As he puts it, irrigation of his/her land. Prior appropriation rights may be summed up as “first in time ‘Negotiation involves interaction first in right”; the earlier appropriator has a between different claimants, not right superior to later appropriator. unilateral decisions made in isolation. It includes sitting around Meinzen-Dick and Nkonya (2005:5) added a table to craft an agreement, that water rights can be broadly classified as formal trading arrangements as public, common, or private property. Public well as less visible struggle over water rights are rights held by the state and access to water, as local people in which the government allocates rights to comply with or contest the ways in users (the case of Zimbabwe in 1990s and which state agencies or other Mozambique 1991 serve as best examples). users acquire and distribute water. Common water rights refer to communal Thus, negotiations are processes water rights where water can be used by of interactions between disputing people in ways that are specified by some parties whereby, without community. In most African customary compulsion by a third-party water law, water is considered as a adjudicator, they endeavor to community property and private ownership come to an independent, joint of water is not recognized. Private rights are decision concerning the terms of rights held by an individual or corporations. agreement on the issues between It is generally only use rights that are them. It proceeds through the recognized for individuals to use water in exchange of information between certain ways. the parties. Information is verbal and non-verbal including Rights to water may be negotiated in many evidence, argument, appeals to contexts, not just within communities but rules and ideology, expressions of also between communities and others sharing strength and proposals of terms water resources. Various strategies may be for agreement. Negotiation is a open to communities including direct action continuing process, influenced-but to acquire more water and restrict others’ not fully determined-by changes in access; litigating in court; participating in rules and laws. Thus, agreements planning and other formal administrative may mark major milestones, but procedures; lobbying to advocate their case usually lead to further negotiation to the public and politicians; and trying to about how the agreement is to be reach agreements with other water users and worked out in detail, how to with water management agencies are all part monitor compliances and respond of negotiating rules about who gets water to violations, and whether to revise (Bruns: 2005: 1). agreements (Gulliver 1979: 79).’

Water rights are dynamic, flexible and

328 subject to frequent negotiations because of industries, amongst specific issues the policy uncertain water supply, damages to intake emphasized: structures, and social, political and economic 1. Ensure the full integration of changes (Meinzen-Dick and Pradhan, 2006). irrigation with the overall framework Reflecting this argument, Bruns and of the country’s socio-economic Meinzen-Dick (2000: 202), put that farmers development plans, and more in Nepal for example, shop for and use what particularly with the Agricultural they believe is the best strategy available to Development (ADLI) Strategy. them in a specific situation. The strategies 2. Promote users based management of they use depend on the social relations irrigation systems taking account of between the stakeholders (such as power, the special needs of rural women in kinship, economic, political) as well as the particular. legal resources they have at their disposal. 3. Enhance indigenous irrigation schemes by improving water In the Ethiopian context, studies conducted abstraction, transport systems and on water rights or processes of negotiations water use efficiency. over irrigation water have appeared scanty. 4. Establish water allocation and Furthermore, policies, legislations and rules priority setting criteria based on issued in connection to water rights have also harmonization of social equity, been far from satisfaction. economic efficiency and environmental sustainability Yacob (2002) contended that in Ethiopia requirements. there does not yet exist a single body of rules 5. Recognize that irrigation is an pertaining to the use rights of and integral part of the water sector and management of water resources; which holds consequently develop irrigation grains of truth for water competitors in the within the domain and framework of Waiyto valley of Southern Ethiopia where overall water resources management his study offered a greater focus. (Ibid).

Zewde (1994:88) has also elucidated that At Oromiya regional state level, a there is no extensive legislation covering the proclamation (No.30/1999) is enacted in use of water in Ethiopia. But, there are order to reinforce the tasks of the Oromiya decrees that water is a national asset and that Irrigation Development Authority. In the it can be controlled only by the central proclamation, the duties and responsibilities government. Additionally, Studies conducted to be assumed by the authority with regard to by (FAO, 2005) reaffirmed that written how best to confiscate water by users are information on water use is not available. some how indicated. Accordingly, under article 6 (powers and duties of the authority) Ewnetu (1987:1) put that in Ethiopia enacted number 1 of the proclamation reads as: water rules appeared recently, but prior to initiate and submit policies, strategies, laws this the people were using customarily and and regulations of the authority. even today it is observed in many parts of the country. Offering a comprehensive analysis, McCornick and Seleshi (2004) have Lemma (2004:50) in his part found out that remarked on ‘Water use rights in Ethiopia’ there is no policy in the region as whole that correlating with the policy environment as entails about water right and entitlement. follows:

Currently, the Ministry of Water Resources “The relevant policy and has formulated a water policy embodying the legislative framework must irrigation component both at Federal and continue to be strengthened and Regional level, basing on the Agricultural allowed to evolve to Development Led Industrialization Policy ( accommodate the indigenous ADLI) of the country (MoWR: 1999). While arrangements and established the overall objective of the policy is to water-rights, and meet the new develop irrigated agriculture for the demands. The recent production of food and raw materials to agro improvements in the national

329

water policy framework has Negotiations involved two distinct though established the necessary interconnected processes going on foundation, and there is some simultaneously: a repetitive, cyclical one and evidence that communities are a developmental one. The cyclical process playing a more active role in the comprises the repetitive exchange of decision making with regards to information between the competing parties, allocation and management of its assessment, and the resulting adjustments water at the local levels, of expectations and preferences while the allowing for better integration of developmental process involved the indigenous water rights and movement from the initiation of dispute/ management systems conflict to its conclusion leading to some (McCornick and Seleshi, outcome with its final implementation 2004:8). (Gulliver, 1979: 82) Hence, the elements of the processual To wind up Ethiopia’s experience regarding models were chosen to be instrumental to irrigation water rights, studies that elicited analyze the processes of negotiations that irrigation water entitlements and negotiations water users apply to settle conflicting over it like Indris modern irrigation scheme interests while diverting irrigation water in were missing. the study localities.

Theoretical Frameworks. Ethnographic Accounts: Findings and The central arguments of legal pluralism and Results of the Research. processual models of negotiations were utilized as the theoretical tools to analyze Historical Background of Indris Scheme: water rights and the processes of negotiations Phases of Irrigation Development. at Indris modern scheme, respectively. The investigation on the historical Legal Pluralism. background of Indris scheme demonstrated its development in three distinctive stages; Legal pluralism begins from the recognition pre-conditions in the initial phases (prior to that multiple legal and normative 1972 E.C), operations in an indigenous frameworks coexist. The paradigm of legal manner (extending from 1972-1986 E.C), pluralism has important consequences for the and complete conversion into modern style conceptualization of the relationship between (post 1986 E.C). norms and behavior. It depicted the perspective of people’s experience with In the first phase of irrigation development water access and control in which individuals around Indris river, as confirmed by the draw up on a range of strategies for obtaining participants of the focus group discussion irrigation water. Thus, government, religious, held at Selam Sefer, the practice was and customary laws, development project introduced alongside with the introduction rules, and unwritten local norms may all of the Grinding Mill technology and address who should receive water, from essentially characterized as; which sources, for what purposes (Burns and ¾ Production was only to Meinzen-Dick, 2000; Meinzen-Dick and complement household Bakker, 2001). consumptions. That connoted market orientation was quite Therefore, the conceptual frame work of negligible. legal pluralism became indispensable in view ¾ The practice was carried out on of contemporary water rights paradigms. It small pieces of land (confined aimed to explore the different only in their gardens) by few conceptualization of water and water rights households alone. and the variety of legal statuses attached to ¾ The vegetables and fruits planted water (Ibid). were limited in variety. ¾ The involvement and advisory Processual Models of Negotiation. support of external institutions like the agricultural desk had

330

been missing. population. The gap grows wider pertaining to the gradual decline in the volume of water. In the second phase, the completion of As a result the water users, the district and digging the web of canals and building up zonal irrigation bureaus as well as the the indigenous dam that in fact lasts only Agricultural Training Center initiated a during the dry season marked a momentous proposal to scale up the working capacity of shift in the historical development of the the scheme. Amongst organizations scheme in numerous respects. To portray the requested, it was the European Economic highlight of some; Commission (EEC) that showed practical Remarkable increase in the number interest to donate( 3.5 Million Birr) for the of water users (250 HHs). construction of the modern dam that took Incorporation of larger hectares of over 2(two) years. Prior to the involvement land under irrigation practice (about of EEC, countless efforts were in place to 180 hectares at its upper limit). build up the dam in relatively strong way to Use of irrigated produces both for function for longer durations than the domestic consumption and market indigenous manner. supply. Formulation of agreed up on by-laws The transformation of the scheme into with the technical assistance of modern operation has induced further agricultural workers and other developments besides those noteworthy officials from the desk. In the first moves formerly attained. Through the phase of irrigation development, discussions held with the agricultural certain tacitly perceived normative development agents, the following points rules emanating from the broader were obtained: customs of the society had guided 9 The command area has grown to cover the actions of water users. Thus, 7-8Kms. In its indigenous operation, it such sorts of implicit normative rules was managed to reach users within were transformed into more or less limited radius from the source, perhaps written rules being interwoven with not more than 4-5kms away from the certain fragmented elements of the main source. Only sites closer to the prevailing irrigation guidelines. main canal got irrigation water Developed the custom of irrigating sufficiently. Thus, the conversion of the twice per a year which was not scheme augmented the command area by formerly patterned in the first phase. at least 2-3kms. Institutions like the Agricultural 9 The scale of the scheme grew from small Training Centre began to divert to medium range. It was about 180 water as its second best alternatives. hectares (categorized as small scheme) Previously, the institute used to of land covered through indigenous divert a substantial amount of water irrigation techniques that stretched to from Chole River. incorporate about 381 hectares (categorized as medium scale) of land in Along with the significant progress that its current state. The numbers of water accompanied the scheme’s development, users have also increased from 250HHs water users have also internalized the wide- to 1020HHs as a result of opportunities ranging worth of irrigated farming. The secured in connection to the promotion compounded effects of these great changes of Indris scheme. and the increasing recognition by 9 The need to apply environmentally practitioners, created the motive for the sound and scientifically proved varieties transformation of the scheme into its third escalate based on the calculations to gain phase i.e. from indigenous scheme to modern satisfactory benefits. system. 9 The institutional capacities and management systems of irrigation water Principally, the increasing demand from the became to be handled by Indris Water framers side and the discharging capacity of Users Association (WUA). The the indigenous scheme on the other hand association was formed with the mandate happened to be inconsistent to meet the to operate since the aftermath of the interests of the drastically growing transformation of the scheme. Problems

331

interlinked with water scheduling, water o Lack of support on financial and misappropriations (in the form of theft, technical skills had also their seepages or salinity) received more contributions. concern in an integrated approaches o Institutional weaknesses prevailing since the promotion of the scheme. at the three stages of the scheme’s 9 Irrigation production (fruits and development was another vegetables) inclined to focus on market constraining factor. Nearly in all orientations. In this third phase, circumstances institutional matters production for market supply were either taken for granted or outweighed household consumption. deliberately overlooked as if it could 9 Water users’ dependence on irrigation pose only miniature effects. practice augments across the three o Conflicts of diverse nature (among villages. The majority of them were water users themselves in their reluctant to engage in the practice in vicinity, water users of upper groups former phases of irrigation development. against lower reaches, or between ¾ A slight modification to the naming water users and other claimant of the scheme was also another institutes) somehow lingered the development with the promotion of developments relating with Indris the scheme. The indigenous scheme. diversion was named after the name of the river itself i.e. Indris. The Therefore, interplay of factors both from intent behind the new naming has within and outside the irrigation system has stemmed from irrigators’ exclusive interacted to induce these encouraging dependence on the water diverted changes. Above all, the internal motives of from the river. On a meeting of the water users striving for betterment on water general assembly, which is of course resource extractions interwoven with the the highest decisive body; it was financial donation of European Economic reached a consensus to name the Commission played pivotal role to attain its scheme as ‘Indris Fayiissa’ literally current state. mean Indris our savior. Intricacy in the chains of structures and institutions involved in the management, allocation and distribution of water have moved from a state of simple operations to the level of sophisticated webs of networks directed by the frameworks of Indris Water Users Association. The general assembly of users took its ultimate power to provide decisions for the pragmatic allocations and distributions of water via the committee functioning at three levels: Executive, Gooxii (territorial level) and Garee (team level). Indris Modern Scheme: constructed with 3.5 million Birr donated by EEC. Water Rights among Irrigators in the Basing on discussions held with the Study Area: Basics for Decision and agricultural development workers and Access. irrigation experts pertaining to factors that impeded the development of Indris scheme, The basics for water right access and state of affairs described below were decision have enjoyed a broader spectrum of extracted; considering land rights, historical o The policy environment regarding precedence/settlement in the area, financial irrigation schemes lacked strong as well as labor requirements to be basis. Dearth of specific directives maintained by users. At a time the irrigation and guidelines on pragmatic practice commenced in the area, land rights implementations represents one or possessions had been considered as a instance. factor to enable users’ access to irrigation

332 water. This signified and coincides with the depending on the total hectares elements of the riparian water right doctrine. of land holdings. In the second phase of the scheme’s 9 Payments aiming to compensate operation in an indigenous manner, historical services rendered by distinct precedence and settlement in the villages had institutions like credit and saving predominantly served as a parameter to services. decide and secure users access to irrigation 9 Payments imposed on water diversion on top-of land rights. With the users as a punishment in conversion of the scheme into modern style, recognition to their disobedient the decisions to admit or deny users for acts against the WUA by-laws. irrigation water rights rest on and determined 9 Annual fees collected from through their labor contributions and water users for maintenance and fulfillment of obligations imposed by the operations. general assembly. Water users were also e. Safeguarding the scheme and any noted to employ a mix of techniques for other assets of Indris WUA. creating access to water rights in the form of f. Attendance and participation on sharecropping, purchase or contractual meetings. arrangements through negotiated approaches. Certain interlocking factors do shadow the Informants in the three research villages have water rights of users. To outline few: denoted the most essential factors that should When the scheme functioned in an be considered basics for access and decision indigenous manner, water right issues since the conversion of the scheme into were not as such perceptive. As a modern style: result, the inclinations of the i. The applicant should be above 18 agricultural desk as well as water users years old and establish his/her own were far from addressing themes on family. water rights. It has been with growing ii. The applicant should have a well conflicts, dialogue and violation of defined residence in the PA and be rules that water rights got an registered as a member of the WUA of increasing concern. Indris scheme. Diminished endeavor on the iii. The water users need to own not only a enforcement of by-laws of the WUA. land but also make sure that a portion The poor application of directives of the land has to be suitable for stated on the manual made their irrigation. implications to carry insignificant iv. To reinforce their water rights access, effects. users have to meet their obligations The dynamic nature of water rights; which are clearly incorporated on the the basic dimensions considered to water use manual. The obligations to access irrigators’ water right have not be maintained by water users at the been fixed. Access to land rights fore front encompass. fundamentally determined users’ water a. Respect and materialize the by-laws right during the initial phase of the of the WUA: scheme’s development. This was b. Entire utilization of potentially accompanied sooner with the questions available irrigation plots. of settlement closer to the water canals c. The members of WUA are obliged to i.e. historical precedence in the area. conduct canal cleaning and Recently, the reconstruction of the maintenance once or twice a year. scheme also necessitates the d. Timely completion of financial fees: reformation of irrigators’ water right in the obligation to timely pay their several regards. The dynamic natures financial fees constitute among of water rights become more intricate compulsions imposed on water as its rules originate from multiple users. The financial payments origins. For instance, since the expected from water users basically promotion of Indris modern scheme, comprises of; irrigators could get accesses of water 9 Taxes collected by rights plausibly via: administrative structures i. Inheritance from parents or

333

relatives who had formally confirmed the conception that ‘water flows established their irrigation water in the direction of power’. Within unequal rights. power structures, different societal bodies ii. Purchase of the water rights of a define their strategies in order to defend and given user through contractual materialize their own interests in controlling agreements. the water. Thus, ill-treated distribution of iii. Water rights maintained through irrigation water is not only through share-cropping. consideration of economic status but also iv. Water rights acquired in mystification of the power via which the user association with land re- households or institute define its access. The distributions by the government. same holds true for investigations on gender matters where in women are losers and men The water right rules are multiple in origin defined their dominance. and integrative in function. The predominately governing water use rules are Alike to economic status, power interactions the combination of customary normative and gender dimensions, religious out looks dictations and formal-legal irrigation water and ethnicity have also carried similar effects management guidelines. These water right though not as affective as dimensions rules are inheritable, dynamic in the sense discussed above. In terms of religious out that they would be subject to revisions either looks, for instance, there were occasions for omissions or additions pertaining to the when water users in the same church turmoil environment. deliberately cover water right offenses committed against users who do not belong Webs of multiple factors have been to the category. The tendency to sympathize investigated to hold linkages with water or disregard one another based on ethnic rights either to promote or curtail users’ considerations only rarely noticed compared accesses and participations for decisions. to other dimensions. Rather, strong Economic status, power, gender, ethnicity interrelationships maintained through social and religion comprised amongst the pertinent ties and webs of networks (like associations factors that have in one way or another of Iquib/Idir, marriage, extended families, contributed either to facilitate or deter users’ work labor parties in the form of Debaree or access to irrigation water. Economic status, Debo) worked either to enhance or deter the for instance, affected the water right of users access of users in or out of the network at least as the economically better of, differently. Completes a range of tuitions imposed on them earlier in relation to the The pragmatic application of irrigation water economically weak. right occupied a prominent place in the Can purchase access to water rights struggle to enhance irrigators’ livelihood or through contractual agreements or income expropriated from the practice. entering share-cropping. In contrast, Failures to observe the water use rules in water users in an economically weak general and ones own water right in position lack to utilize such particular deteriorated users’ livelihoods. As opportunities. an instance, users with longer experiences of Develop a relatively strong channel of implementing the water user rules have been communication with the committee reported to progressively improve their members, extension workers and other livelihoods year after year while conversely institutes. those failing to apply the rules suffer from crisis imposed on them in the form of In a similar talk, whatever the fashions of its punishments. manifestations, power relations have prevailed among diverse entities where in an There existed few non-irrigators who failed institution or individual user either to acquire the opportunity to divert water legitimately or illegitimately tries to impose from Indris scheme due to varied reasons. their power weakening the water right of None of the contacted non- irrigators claimed those in comparably underprivileged that constraints interlinked with water rights positions. In consistence with this view, have never deterred them both from the Boelens and Davila (1998:445- 447) access or decisions. Rather distance from the

334 scheme/ source, decline in the amount of water users do frequently negotiate over a water for the lower stream users (Kilinto) range of issues to be implemented with the and geographical barriers for the upper supposition to promote the effectiveness of stream users (Selam Sefer) became major the entire system. These include negotiations hampering factors on top of other personal executed in relation to a range of fees, matters. categories of items to be irrigated, maintenance schedules and labor Negotiations and Dispute Settlement over contributions, total hectares of land to be Irrigation Water in the Study Area. irrigated or possibly water scheduling days and hours. On the other hand, a series of Conflicts of competing interests among processes of negotiations would be diverse categories of users are another undertaken successively at times of serious dimension to be looked at with co-relation to conflicts. Generally, the steps pursued to the water right of users. The results and initiate and finalize the processes of discussions of primary data generated during negotiations moves through state of the field work proved increase both in the procedures described below; intensity and severity of conflicts over irrigation water matching the three phases of i. Recognition of the grounds of the the scheme’s development. Conflicts occur dispute/ Causes: at this point the among users in a village, along the streams negotiators themselves or the of the scheme or between farmers and the mediators (elders, committee, Agricultural Training Center of Ambo Kebele or district court) examine the College. basis of the conflict. ii. Conduct assessments on the The principal causes for conflicts to erupt out prospective points to be has been embedded in the transgression of negotiated: having deeply examined negotiated water use rules like theft, turn what have instigated the conflict; abuses, failing to timely pay financial fees or either the negotiators or mediators deliberate ignorance of labor contributions move to sort out the promising for canal operations and maintenance and arenas where in negotiation would gradual decline in the volume of water be set in motion to bring out resource itself. Power abuse became to be remarkable agreements. considered as the second chief factor for iii. Persistent presentation of conflicts. negotiators’ respective cases and points of departure: the segregation These conflicts have been handled at distinct of points over which negotiation levels within the village through elders and produce relatively stable committee decisions, at Kebele courts or at interactions, ultimately invite district court. In case these conflicts emerge negotiators to present their cases to into disputes that stayed over a prolonged each other and the mediators in the duration of time, a succession of negotiation attempts to persuade the audiences processes and procedures would be elaborating that their argument employed to manage the case. The processes contained more reality than their involved could range from identification of opponents. There also appeared the the central predicaments to the final remarks presumption that negotiators who of disputants that assured their consents not aspired to gain much from the debate to replicate a similar disagreement once need to condemn the arguments of again. The protracted conflict between the their opponents. The presentation of Agricultural Training Center and the farmers, their respective cases usually takes or users in the upper and lower streams much of the time in the entire basically exhibited typical instances of the processes of negotiation. case in point. iv. Narrowing the gap perceptible between the interests of Of course, negotiations have been carried out negotiators: the succession of in dual senses: negotiations under normal appointments to heed the respective circumstances and negotiations conducted to cases of water users in dialogue smooth disputes. In its former meaning, enable to easily distinguish the gap

335

in the interests perceptible between parts of their quests submitted for either the teams or individuals in negotiations though commonly unfeasible to debate. At this instance, the acquire the entire range of their initial negotiators themselves or mediators pursuits. would engaged in the facilitation of negotiation processes or suggest the strategies pertinent to successfully Conclusion and Recommendations. narrow the competing interests of claimants. Conclusion. v. Provide the final decisions to confirm and attain consensus: with Debates over access to water rights for the minimization of the gap, comes irrigation usage and negotiations have been the concrete decisions to be accepted central among water users and institutions commonly by the former opponent claiming water from Indris scheme. These sides. debates over water right claims have been mainly attributed to rising competitions over These steps largely represent the standards at irrigation water, drastic population growth least to be pursued, details of procedures and shrink in the volume of water due to incorporated in each steps vary from village augmented diversions points on Indris River to village or depending on the general setting as well as failures to strictly observe rules of where the negotiation have been conducted. conduct for resource use. Disputes associated The implications and essential arguments of with use rights, like farmers against Ambo the processual models of negotiation Agricultural Training Center, or among the (cyclical and developmental) hold farmers themselves have predominantly coincidences with the issues at hand. In each prevailed in the research sites. step of negotiation, as participants of the focus group discussions revealed out, there Though the elements of both the riparian and exist recurrent exchange of views and appropriative doctrines of water rights information that enabled the opposing mirrors among the irrigators in the study categories to learn more about the sites, their know-how regarding water right expectations and responses to the particular entitlements appeared to be found at its quests embraced in the process. This notion infancy. Themes that center on irrigation specifically parallels with the elements of water rights tended to be overlooked under cyclical model of negotiation. The normal circumstances. Such topics become cumulative effects of each negotiation steps sensitive and an area of controversy only finally culminate in generation of the when the rules are frequently violated consensus sought by water users as their affecting the use rights of members. As a primary target. So, the ultimate achievement whole, water right themes for irrigation (i.e. bargaining over their irrigation water diversion gradually began to receive rights) elicited the central attributes of the substantial concerns corresponding to the developmental model of negotiation. The end phases of the development of Indris scheme. results of negotiations over water rights assured through repetitive and cyclical The theoretical frameworks employed as a processes of both information sharing and conceptual tool to examine the water rights learning depicted in each stage enabled users and processes of negotiations over irrigation to transform from a state of competing water among irrigators diverting water from interests to collaborating entities. Indris scheme proved to be in consistence with the experiences of other developing Though the ultimate aim to conduct nations reviewed in the literature. Hence, negotiation processes have been triggered both the arguments of legal pluralism with the motives to induce a real truce, advocated principally by Burns and sometimes the outcomes may not end in the Meinzen-Dick (2000) in water rights way expected to attain encouraging targets. paradigms and the processual models of Instead, it leads to additional rearrangements negotiations discussed by Gulliever (1979) for having chains of negotiations. Anyhow, were pragmatically apparent in the daily at the end of negotiation processes, the actions and practices of water users in Toke competing parties or users may gain certain Kutaye district of West Shewa Zone.

336

by Gulliever (1979) tested to bear reliability In essence, the applicability of legal with the way conflicts over irrigation water pluralism was verified via the multifarious have been settled or the processes gone sources of water rights and their integral through to negotiate and renegotiate for functionality. This fitted with the securing access to water. Though these conclusions of former studies carried out in models are thought to work at the same various developing countries. In that regard, instance, the cyclical model applies more to a study by Cleaver, et al ( 2005) grip minor conflicts erupting out among demonstrated the conception that claims to users or teams in a village. Conversely, the resources in general and water in particular developmental model of negotiation has been are made and enforced through both ‘formal’ utilized to justify the protracted disputes (local government, Water Users continued over a number of days or a couple Associations, tenure arrangements) and of months. The arguments of this same ‘informal’ ( customary practices, social model appeared instrumental to reconcile the relationships, norms of use and access) frame antagonistic interests of the Agricultural of references. Training Center and farmers as well as frequently occurring disputes among water Correspondingly, at Indris scheme, a users at Selam Sefer (as upper stream combination of rules for water rights beneficiaries) and Kilinto (as lower stream stemming out of the broader normative rules beneficiaries). as well as guidelines formally enacted for irrigation water management were essentially Recommendations available. In the efforts to reinforce water rights besides rules framed out indigenously, The views contended herein below demand significant reforms embodying the legal the concerted efforts of interested environment have been accomplished with stakeholders, users themselves, researchers the help of Toke Kutaye district agricultural or institutes to contribute their resources (in desk. While the indigenous rules exhibit the the form of time, know-how/ technical skills, pragmatic contexts of irrigators in their donations or material supplies) to narrow respective vicinities, the formal-legal missing gaps and thereby endeavor for better approach tends to emphasize the components transformation. Hence, the clues embrace: of irrigation directives adopted by the i. The policy pertaining to irrigation government to be implemented for inducing water management in general and change. The arguments extracted from these water right entitlements in particular dual approaches have evidenced to effectuate have been in progress time over the water use right claims and the processes time. This investigation indicated the of negotiations among irrigators in the study need to further coin out specific areas. water right guidelines by the relevant organs in an instructive way and In this research, it was realized that equally applicable for all users. customary rules were particularly effective in curbing conflicts while the formal guidelines ii. The chief stakeholders like the become strong to reinforce punishments at district agricultural desk need to moments either the rules are poorly provide special focus on enhancing interpreted or deliberately transgressed. As a the awareness level of users about result, combinations of these rules enable to the overall nature of water rights. have wider frameworks for the actions and Formerly, it was confirmed that behaviors admitting users to divert water. inadequate attention was paid to Therefore, the paradigm of legal pluralism programs that create awareness contains much relevance with the existing among users about water rights. As it realities of water users in the three selected promotes the empowerment of users research settings for this study (Selam Sefer, in that regard, there is an acute need Dhaga Fillee and Kilinto). to incorporate water right themes in the plans for actions to be In a similar manner, the arguments of implemented. processual models (both cyclical and developmental) of negotiations expounded iii. There prevailed the tendency to

337

increasingly undermine the role of conflicts of interests, negotiated customary rules and instead approaches have been declining over forcefully impose legally recognized time tending to carry little credits rules by various government among users. The study recommends organizations. Yet, the customary offering particular considerations on rules were by no means less in collective actions that uphold the effective than the formally laid enhancement of negotiated directives in numerous respects (for approaches over resource use instance, saving resources). Hence, (water). The institutions like WUA emphasis has to be given on mediating the access of users need advocating the merits of customary also undergo periodic restructurings rules that shines the wisdom and in a way to accommodate ever cumulative experiences/skills of growing demands of the users. practitioners. iv. The causes of conflict were Reference essentially embedded in failures either to adhere or loose Adams, W.M (1992). Wasting the Rain: implementation of the negotiated Rivers, People and Planning in Africa. rules of water use. In this regard, the University of Minnesota Press, findings of the study have indicated Minneapolis, London. the existence of a serious affair that Aureli, Alice and Brelet, Claudine (2004). must be addressed thoroughly. Women and Water: An Ethical Issue. Therefore, water users, technical Retrievedfromhttp://unesdoc.unesco.or experts, the institutes claiming water g/images/0013/001363/136357e.pdf. from the scheme or other 27/06/2007. UNESCO, Printed in stakeholders need to work hand-in- France. glove manner to reinforce the Barlett, F. Peggy (1980). Agricultural practical interpretation of rules on Decision Making: Anthropological the ground. Contribution to Rural Development. Academic Press, Inc. Barrow, Chris (1987). Water Resources and v. In the face of growing population Agricultural Development in the Tropics. and declining water volume, Co-Published in the United States with competitions over irrigation John Wiley and Sons, Inc., New York. diversion would inevitably augment. Bedru Beshir (2004). “Small Scale Irrigation Initiatives to supplement water from Users Peasant Horticulture in Dugda other perennial and potential rivers Bora and Adami Tulu Jido Kombolcha (like Heddee) have been proposed by Woredas, East Shewa Zone: Challenges users, the district agricultural desk and Opportunities.” MA Thesis. Addis and the Agricultural Training Center. Ababa University. Addis Ababa. As the materialization of this Blank, G. Herbert; Mutero, M. Clifford and proposition would only be realistic Murray-Rust, Hammond (2002). The with the financial assistances of Changing Face of Irrigation in Kenya: interested organizations (like EEC), Opportunities for Anticipating Change in the study suggests conducting further Eastern and Southern Africa. assessments or searches on the International Water Management likelihood to assure irrigation water Institute (IWMI). sources in addition to Indris Scheme. Boelens, Rutgerd and Davila, Gloria (1998). Searching for Equity: Conceptions of vi. Negotiated approaches over water Justice and Equity in Peasant Irrigation. rights and other themes in the Van Gorcum and Comp,B,V., The irrigation system, as a best option, Netherlands. have facilitated desirable Bruns, Randolhp Bryan and Meinzen- Dick achievements in inducing common S. Ruth. (2000). Negotiating Water understandings between disputants. Rights. International Food Policy Despite its remarkable role to settle Research Institute. LTDG Publishing.

338

Carney,Diana and Farrington,John (1998). Ervin M. Alexander (2005). Applied Natural Resource Management and Anthropology: Tools and Perspectives Institutional Change. Routledge for Contemporary Practice. Second Research /ODI Development Policy. Edition. Pearson Education, Inc. Sydney. Routldge, Taylor and Francis group. Ewnetu Akama (1987). “Water Law of London and New York. Ethiopia.” Submitted in Partial Cernea, M. Michael (1991). Putting People Fulfillment of the Requirement for the First: Sociological Variables in Rural Degree of Bachelor of Law. Addis Development. Second Edition. A World Ababa University. Addis Ababa. Bank Publication. FAO (1978). Irrigation Practice and Water Cleaver, Frances; Franks, Tom; Boesten, Management. Rome. Italy. Jelke and Kiire, Angela (2005).Water FAO (1990). National Irrigation Policy and Governance and Poverty: What Works Strategy Workshop: Discussion Paper. for the Poor? Bradford Center for Office of the National Committee for International Development. University of Central Planning. Bradford. http://www.bradford.ac.uk/acad/bcid/re FAO (2005). FAD’s Information System on search/poverty/projects/water/Water_Go Water and Agriculture: Land and Water vernance_and_Poverty_Final_Report.pdf Development Division. . Fuller, Chris (1994). Legal Anthropology, Cornish, Gez (1998). Modern Irrigation Legal Pluralism and Legal Thought. Technologies: For Smallholders in London: LSE Research Online. Developing Countries. Intermediate Available Technology Publications In Association at:http://eprints.lse.ac.uk/archive/000005 with HR WALLING FORD. 38 Cotula, Lorenzo (2006).Land and Water Gulliver, P.H. (1979). Disputes and Right in the Sahel: Tenure Challenges of Negotiations: A Cross-Cultural Improving Access to Water for Perspective. Academic. Agriculture. Sida and FAO. Havet, Jacques (1978). Main Trends of CRDA (1994). Follow-Up Workshop on Research in the Social and Human Small Scale Irrigation Schemes. Addis Sciences. Part Two/Volume Two: Legal Ababa. Science/Philosophy. UNESCO, Place de Darout Gum’a (2004). “The Socio-Cultural Fontenoy, Paris. New York. Aspects of Irrigation Management: The ILRI (2006). People, Live Stock and the Case of Two Communities- Based Small Environment: Sustaining Water and Scale Irrigation Schemes in the Upper Nutrient Productivity. Community Based Tekeze. Basin, Tigray Region.” MA Irrigation Management in Ethiopia: Thesis. Addis Ababa. AAU. Strategies to Enhance Health, Livestock, Dejene Aredo (1994). Female- Headed and Crop Production and Natural Households in Two Contrasting Regions Resource Management. in Ethiopia: Access to and Management Indris Water Users Association. “Irrigation of Resources. Ethiopian Journal of Water Use and Management Manual of Development Research 16 (1): 1-16. Indris Water Users Association”. Retrieved from http://www. Undated. Ilri.org/data/livelihood/proceedings/pape Katharina, Wallner (2006). “Field Parameter rs/4Dejene.htm. Evaluation to Support Environmental Dejene Negassa (2001). “Access to Resource Impact Analysis in Ethiopia”. A Thesis Strategies and Changing Household Submitted to the University of Natural Organization: A Case from Kamme Resources and Applied Live Sciences Village, Bate, Oromia Zone of Amhara (Vienna), Institute of Hydraulics and Region.” MA Thesis. Addis Ababa. Rural Water Management, in Partial Addis Ababa University. Fulfillment of the Requirements for the Dessalegn Rahmato (1999). Water Resource Degree “Diplomingeniever”. ILRI. Development in Ethiopia: Issues of Kerealem Salilih (2007). “Irrigation Sustainability and Participation. Forum Management and its Contribution in for Social Studies. Addis Ababa. Reducing Households’ Socio-Economic Poverty: The Case of Two Small-Scale

339

Irrigation Schemes in Blue Nile Basin of Policy. The Federal Democratic Amhara National Regional State”. MA Republic of Ethiopia. BERHANENA Thesis. Addis Ababa University. SELAM PRINTING ENTERPRISE. Keyes, G. Conard and Ward, Jim J. (1985). Ministry of Education (2002). “Agricultural Development and Management Aspects Production Practices, Opportunities, of Irrigation and Drainage Systems. By Constraints and Intervention Options in the American Society of Civil Engineers. Ambo Wereda (West Shewa Zone- Lakew Birke (1970). The Impact of Dogali Oromiya Region)”. Ambo College of Irrigation Project on the National Income Agriculture. of Ethiopia. PH.D Dissertation. The Ohio State University. Mokonnen Cherinet (1992). "Socio- Lemma Dinku (2004). “Smallholders’ Economic and Environmental Impacts of Irrigation Practices and Issues of Irrigation: The Case of Ziway State Community Management: The Case of Farm. MA Thesis. Addis Ababa. AAU. Two Irrigation Systems in Eastern National Research Council (1991).Toward Oromia, Ethiopia.” MA Thesis. Addis Sustainability: Soil and Water Research Ababa University. Priorities for Developing Countries. Marshall,Catherine and Rossman, B. National Academy Press. Washington, Gretchen (1995). Designing Qualitative D.C. Research. Second Edition. Sage Nigussie Taffesse (2002). “The Role of Publications, Inc.Thousand Oaks, Irrigation Development in Enhancing London, New Delhi. Household Food Security: A Case of McCornick, G. Peter and Seleshi Bekele Three Small Scale Irrigation Schemes in (2004). Water Use Rights in Ethiopia – Southern Nations, Nationalities and An Overview. Water for Socio-economic Peoples’ Region.” MA Thesis. Addis Transformation Forum, October 25-27, Ababa University. Addis Ababa. MoWR, Addis Ababa, Ethiopia. Oromia Irrigation Development Authority Meinzen-Dick, Ruth and Bakker, Margaretha (1991). Megeleta Oromia. Proclamation (2001). Water Rights and Multiple Water No. 30/1999. Oromia Regional State Uses. Framework and Application to Irrigation Development Authority Kirindi Oya Irrigation System in Sri Establishment Proclamation. 7th Year, Lanaka. Kluwer Academic Publishers. No. 4. Finfine. Printed in the Netherlands. Pant,Niranjan (1984). Productivity and Meinzen-Dick, Ruth and Letica, Nkonya Equity in Irrigation Systems. Grini (2005). Understanding Legal Pluralism Institute of Development Studies. Ashish in Water Rights: Lessons from Africa Publishing House, Punjabi, Bagh, New and Asia. International Workshop on Delhi. ‘African Water Laws: Plural Legislative Pereira, L.Santos and Gowing, John Frameworks for Rural Water (1998).Water and The Environment: Management in Africa’, 26-28 January. Innovation Issues in Irrigation and Johannesburg, South Africa. Drainage. Great Britain. Cornwall. Meinzen-Dick, S. Ruth and Rajendra, Punima, B.C. and Lal, B.B. Pande (1992). Pradhan (2006). Legal Pluralism and Irrigation and Water Power Engineering. Dynamic Property Rights. CGIAR Twelfth Edition. Laxmi Publications System Wide Program on Collective PVT.LTD. New Delhi. Action and Property Rights. International Rathgeber, Eva (2003). Dry Taps… Gender Food Policy Research Institute. and Poverty in Water Resources Washington, D.C. Management.FAO.RetrievedFromhttp:// Melesse Getu (1994). ‘Tsemako Women’s www.fao.org/docrep/005/AC855E/ac Roles and Status in Agro-Pastoral 855e02.htm. Production’. MA Thesis. Addis Ababa Rowland, J. R. J. (1993). Dry Land Farming University. Addis Ababa. In Africa. Macmillan Education LTD. Michael, A.M (1978). Irrigation: Theories Sandra, Postel (1992). Last Oasis: Facing and Practices. New Delhi: Vikas Water Scarcity. W.W. Norton and Publishing House. Company. New York, London. Ministry of Water Resources (1999). Scoones, Ian and Wolmer, William (2003). Ethiopian Water Resources Management “Introduction: Livelihoods in Crisis,

340

Challenges for Rural Development in World Bank (1998). Planning the Southern Africa”. Edited in “Livelihoods Management, Operation and in Crisis? New Perspectives on Maintenance of Irrigation and Drainage Governance and Rural Development in Systems: A Guide for the Preparation of Southern Africa”. Institute of Strategies and Manuals. International Development Studies, Vol. 34. No 3. Commission on Irrigation and Drainage. IDS III Bulletin. Washington, D.C. Seid Yassin (2002). “Small- Scale Irrigation Yacob Arseno (2002). Conflict Management and Household Food Security: A Case over Water Rights in Ethiopia: The Case Study of Three Irrigation Schemes in of Waiyto Valley in Southern Ethiopia. Guba Lafto Woreda of North Wollo Edited in “Transformation of Resource Zone, Amhara Region.” MA Thesis. Conflicts: Approach and Instruments”. Addis Ababa University. Addis Ababa. By Baechler, Spillmann and Suliman. Shimelis Dejene (2006). “Institutions, Peter Lang Ag Bern, European Management, and Challenges of Small- Academic Publishers. Printed in Scale Irrigation Systems in Ethiopia: A Germany. Case Study of Gibe Lemu and Gambela Zewedie Abate (1994). Water Resources Terre in Western Oromiya”. MA Thesis. Development in Ethiopia: An Evaluation Addis Ababa University. of Present Experience and Future Throne and Peterson (1977) Irrigated Soils; Planning Concepts. A Management Their Fertility and Management. Second Method for Analyzing a Key Resource in Edition. Tata McGraw-Hill Publishing Nation’s Development. Ithaca Press, Company LTD. New Delhi. Readings. Toke Kutaye Econmoic Planning and Development Department (1999). “Economic Planning and Development Document of Toke Kutaye District.” Toke Kutaye Agricultural Desk (1999). “Strategic Planning and Management Document of Toke Kutaye Agricultural Desk, 2002-2005”. Tsegaye Kassa (1991). “Small Scale Irrigation Schemes in Mendoyu and Goro Awrajas, Bale region, Ethiopia.” MA Thesis. Addis Ababa. AAU. Turral, Hugh (1998). Hydro Logic? Reforms in Water Resource Management in Developed Countries With Major Agricultural water Use: Lessons for Developing Nations. Over Seas Development Institute. Vermillion, L. Douglas (2000). Water Rights in the State of Nature: Emergent Expectations in an Indonesian Settlement. Edited in “Negotiating Water Rights”. Edited by Bruns and Meinzen- Dick. International Food Policy Research Institute. Vinent, Linden (1995): Hill Irrigation: Water and Development in Mountain Agriculture. Intermediate Technology Publication. Woldeab Teshome (2003). Irrigation Practices, State Intervention and Farmers’ Life - Worlds in Drought – Prone Tigray, Ethiopia. Wageningen, The Netherlands.

341

Technical and Institutional Evaluation of Geray Irrigation Scheme in West Gojjam Zone, Amhara Region

Gashaye Checkol and Tena Alamirew Haramaya University [email protected]

Abstract while there is no change in the water supply indicating that the sustainability of the The technical and institutional performance scheme is in doubt. The cooperative support evaluation of Geray Irrigation Scheme was made in order to identify management services that had been rendered to the practices for implementation to improve the beneficiaries in the past four years were system operation and the general health of found to be minimal. Moreover, there were the irrigation system. The evaluation was few indicators that show the production was made by looking into the selected market oriented. The evaluation clearly performance indicators such as conveyance revealed the fact that conflict resolution efficiency, application efficiency, water remains to be the duty of the Kebele delivery performance, and maintenance authorities and WUA has no legal right to indicators. The availability of institutional enforce its bylaws. In conclusion, the overall and support services were also investigated technical adequacy of the scheme is rated through a questionnaire administered to beneficiary farmers and other stakeholders. very poor requiring tremendous mobilization The results obtained showed that the main of the community to sustainably manage it. and tertiary canal conveyance efficiencies Proper institutional setup needs to be in were 92 and 82 percents respectively. Many place, and WUA needs to be empowered of the secondary and tertiary canals are more in order to enforce its by-laws. poorly maintained and many of the structures are dysfunctional. Application 1. Introduction efficiency monitored on three farmers’ plot located at different ends of a given Ethiopia is labeled as the ‘water tower of secondary canal ranges from 44 to 57 Africa’. Its geographical location and percent. Water delivery performance was endowment with favorable climate provide a only 71 percent showing a very substantial relatively higher amount of rainfall in the reduction from the design of the canal continent. Preliminary studies and capacity. Maintenance indicator evaluated in professional estimates put the nation’s terms of water level change (31.9%) and annual surface runoff to 122 billion m3, effectiveness of the infrastructures showed groundwater potential to 2.6 billion m3 and that the scheme management was in a very the average rainfall of 1090 mm (Daniel, poor shape. Dependability of the scheme 2007). However, the annual amount of evaluated in terms of duration and irrigation water resources in Ethiopia gives the wrong interval showed that the scheme is impression that ‘rainfall is adequate for crop performing below the intended level. The 47 production’. Rainfall distribution is percent of the land initially planned for extremely uneven both spatially and development is currently under irrigation temporally.

As the capacity of the country to store the excess water is in excess is very poor, most

342

of the water flows out carried by harvesting pond construction that has been transboundary rivers to neighboring aggressively pursued all over the country are countries. Consequently, the country suffers evidences for the government’s part and from water scarcity triggered hazards, such commitment to irrigation development. as repeated crop failure, food insecurity, However, given the dismal and undesirable drought and famine with increased experience on the performance of the frequency. About 52 percent of the irrigation schemes developed earlier, there is population or 214 districts of the country are no guarantee that the new schemes will known to be food insecure. The population deliver the anticipated benefits. requiring food assistance is increasing in absolute terms. In 2006, for instance, about Given the anticipated importance of 15 million people were food insecure mainly irrigation in food security and poverty because of rainfall variation (Daniel, 2007). reduction and the huge investments committed on irrigation infrastructure, it is To address the challenges of food insecurity imperative that the irrigation sector is and associated poverty, improving efficient and effective in yielding desired agricultural productivity occupies a central socioeconomic returns. However, the place in the ‘Agricultural Development Lead performances of many of the existing Industrialization (ADLI) strategy of the irrigation schemes have been far from present Ethiopian government. Fighting satisfactory. Several of the previously poverty and reducing food deficiency are at constructed irrigation projects are totally or the heart of ADLI. The government has partially abandoned. Most have not reached committed itself to the reduction of poverty their planned levels of productivities and by half, through the endorsement of the many are not financially or technically Millennium Development Goals (MDG). sustainable in their present forms. Irrigation development is one of the pursued strategic interventions in this regard. The For irrigation schemes to be sustainable, government now recognizes that it is mutual supportiveness of irrigation hardware implausible to expect any great degree of (irrigation infrastructure) and software agricultural production intensification under (institutions) are vital. Mutual rainfed dryland farming system. Hence, supportiveness is ensured when the tremendous efforts are underway to promote hardware is cost and labor efficient, easy to large, medium and small scale irrigations operate and robust, and yielding predictable through huge financial and labor results. The software is characterized by investments. individual/collective interest and management skill embodied in a lean Despite the endowment of Ethiopia with organization of water users besides adequate huge (3.5 million ha) irrigable land, the area support services. under irrigation development is only five percent (195 thousand ha). This shows that The paradox of big expectation from water resources have made little irrigation development to alleviate food contribution towards the development of insecurity and rural poverty versus inability agricultural sector in particular and the to sustainably utilize developed schemes community in general to date. calls for detail scrutiny of the relative contribution of prevailing technical and In the last few years, heavy investments institutional problems of failed schemes. have been made to harness the water One such failed scheme selected for this resources of the country towards irrigation study is Geray Irrigation Scheme. This development. The ongoing Tendaho, scheme, with its source of irrigation is of Kessem, Gode, Koga irrigation projects over excellent quality spring water with virtually and above huge work on rainwater zero silt, was expected to operate with

343

minimum technical problem. However the were chosen among those that served on the scheme has not been able to live up to the same secondary canal at the head, middle expectations. Hence, this study was made to and tail reaches with the presumption that evaluate the contribution of the institution there will be differences in the availability related problems for its technical of water among these categories of underperformance. irrigators.

2. Data and Methodology Application efficiency (Ea) 2.1. Location of the Study Site Depth of water added to the root zone = Depth of water applied to the field Geray irrigation scheme is located 10’ 60’’ 2.2 latitude and 37’26’’ longitude in The water delivery performance Arbaetuensisa Keble located 5 km from indicator: This was calculated by Finoteselam in West Gojam zone. measuring the actually delivered volume of Finoteselam is located 380 km from Addis water to the intended (design) volume water Ababa on the main road to Bahir Dar. to be delivered. Construction of the solid masonry diversion started in 1971 and was completed in 1972. Construction of canals and other structures Water Delivery Performance continued slowly and reports indicate that Actually delivered volume of water the scheme was commissioned in 1984. = Intended volume of water to be delivered

2.2. Technical Performance Evaluation 2.3

Technical performance indicators monitored 2.3. Maintenance Indicators in this study included the measurement of Proper maintenance enables the keeping of conveyance efficiency of the main canal and water control infrastructure in good working tertiary canals, application efficiency and condition so that the design water level is water delivery performance. maintained. The change in head (level) Conveyance efficiency – This was over structures in irrigation canals is the estimated by measuring inflowing and out single most important factor disrupting the flowing water along the selected canal intended delivery of irrigation water. The lengths (Boss, 1997) maintenance indicators are evaluated by the following hydraulic performance indicators (Boss, 1997). Conveyance efficiency a. The relative change of water level water flowing in the canal (RCWL) - this was computed by = water flowing out of the canal taking the actual water level depth from the canal and comparing it with 2.1 design value at the same position in the main canal, i.e. changes of level to Application efficiency: The ratio of the the intended level. depth of water added to the root zone to the depth of water applied to the field was measured from three farmer fields that were Re lative Change of Water level growing potato. The criteria for selecting Change of Level farms were their location associated with the = reach of the canal, i.e. top, middle and lower Intended Level end of the canal. The three plots of farms

344

2.4 Kebele Council, and Cooperative Promotion b. Effectiveness of infrastructure – Staff of Agriculture and Rural Development this measures the ratio of the Office. Samples chosen include 13 WUC number of functioning structures to members, seven WUC executives, five the total number of structures Kebele Council representative, 10 irrigable initially installed. land owners who are not members of the WUC, three farmers owning irrigable land Effectivity of Infrastructure but not rainfed, three farmers that don’t possess but are renting irrigable land, four Number of functioning structures = elderly farmers who participated during the Total number of structures weir. 2.5 3. Major Findings and Discussion c. Dependability of duration – this is estimated as the ratio of the duration 3.1 Characterization of the scheme and the irrigation interval of water delivery compared to the plan. Reports indicate that Geray Irrigation scheme was officially commissioned in 1984 upon finalizing the construction of Dependability of duration tertiary canals for the four of the seven Actual duration of water delivery secondary canals by the then Irrigation = Intended duration of water delivery Development Department, while the head work was finished in 1974 by the Ministry of Agriculture. The project was government 2.6 initiated in an effort to build the capacity of d. Sustainability – is measured as the the then Arbaetuenissia Producers ratio of current area under irrigation Cooperative. to the initial total irrigable area. The net irrigable area of the scheme was 618 Sustainability of Irrigable Area ha. However, the actual area developed for irrigation was 454 ha. The main canal was Current irrigable area = 9.8 km long. The number of secondary Initial total irrigable Area canals was nine with a total length of 12.5 km while that of the tertiary canals was over 2.7 52 km long in total. The weir had 105 m crest length and 4 m height. 2.4. The Irrigation Institution Performance Assessment The numbers of beneficiary households that own irrigable land were 790 out of which By institution here refers to social 400 were members of the WUA and the arrangements that shape and regulate human remaining 390 were non-members. behavior and have some degree of However, during the study period (2006), permanency and transcending individual the area under irrigation was 215 ha only. human lives and intentions. In this survey, The number of potential beneficiaries of the the status of irrigation institution and scheme is estimated to be 3950 people. The availability of support services were lowest average river discharge was reported assessed through questionnaires to be 1.9 m3/s. The maximum main canal administered to the different stakeholders of design discharge with the gate fully opened the scheme. Focus group discussions were was 1.54 m3/s. During this study, the actual made with beneficiary farmers, elders, water flow measured when the gate was fully user cooperative (WUC) management, the opened was 1.1 m3/s.

345

The average main and tertiary canal During this survey, the majority of the drop conveyance efficiencies were measured as and other structures were dysfunctional. 92 and 81%. It was observed during the field There are a number of illegal water work that water was leaking on places where abstraction and canal breaching. Majority of the canal was breached, flow in canal installed tertiary canals were not operational. network was not uniform, canals were heavily vegetated, water flows over the Farmers practice furrow irrigation. The banks of the canal. furrow lengths range from as long as 54 m to as short as 8 m. There is no restriction for The application efficiencies measured on the the type of crop one grows, but 75% of the three farms selected for this study is farmers grow potato. presented in Table 1. The values obtained were close to each other and low, but not 3.2 Technical Efficiencies very far from that expected in surface irrigation system. Conveyance and application efficiencies

Table 1. Application Efficiency

Applied Depth Stored Depth Application Farm (mm) (mm) Efficiency (Ea %) Head end of the canal 215.5 123.7 57.39 Middle of the canal 188.4 83.7 44.41 Lower end of canal 123.5 60.9 49.33

Water Delivery Performance The water delivery performance calculated as the ratio of actually delivered (1.1 m3/s) Effectiveness of Infrastructure water to the intended volume of water (1.54 As per the design document, the total m3/s) was 71 %. A 29% reduction in the number of different structures constructed capacity of the system was large. The effect was 111, but only 74 of them were of this reduction in the carrying capacity of functional. As a result the value of the main canal was reflected in reduced area effectiveness of infrastructure was reckoned actually irrigated. to be 67%. Nearly one-third of the structures had been destroyed. Severe The Relative Change of Water Level mutilation of water control structures for As per the design document, the main canal their iron bars was in evidence. carries 1.54 m3/s discharge when the height of water table is 0.47 m; whereas the height Dependability of Duration of water table measured was 0.32 m when The intended duration of water delivery as the actual discharge was 1.1 m3/s, resulting per the design document was 10 hours in a 32% water level change. This showed irrigation time per day. However, farmers that the intended water level in the main were irrigating up to 18 hours a day. The canal was achieved; consequently less calculated dependability duration was 180% discharge was delivered to the farms forcing showing that it required more time than farmers either to increase the irrigation time anticipated to irrigate the fields. or decrease the irrigated area. From all the indicators monitored in this survey, the scheme’s technical performance

346

was sub-optimal in all standards, and organizing some of them under the requires urgent maintenance. cooperative umbrella while leaving others as non-members seriously challenge to the 3.2 Institutional and Support Services effectiveness of the cooperative in water management. This is against one of the 3.2.1 Scheme Management principles of enduring self governing institutions (Ostrom, 1992 as cited in Sarker The scheme management since it was and Itoh, 2001). Moreover, the cooperative commissioned 27 years ago has been manual lacks detailed operational rules changing. Earlier in the previous (such as entry, allocation, penalty, input government, it was managed by producer rules) and organizational structure specific cooperatives. But when the cooperatives to the irrigation scheme. disbanded, the local government through the Agriculture and Rural Development Office 3.2.3. Operation and Maintenance with all its inefficiencies tried to manage the Problems scheme. Later on, however, knowing that it can no longer effectively manage it, it As per the response of the farmers crafted farmers’ cooperative and passed the interviewed in the study area, some of the management without proper consultation major problems that the farmers face include with all stakeholders. weir and canal leakage, and siltation of canals. These were also in clear evidence For the past four years, Geray Irrigation during the survey. Although the Scheme has been administered by this water maintenance of leaking weir may be beyond user cooperatives (WUC) registered under the capacity of framers in which case they Amhara Regional Cooperatives Promotion may solicit the support of the local Agency. The cooperative adopted a generic government, the canal maintenance should bylaw drafted by the Agency. The bylaw be the duty of the cooperative. However, the indicated that the association is responsible status of the secondary canals and their for water distribution, system maintenance, water control structures showed that no collection of water fees, soliciting for input proper maintenance has been carried out for supplies, credit facilitation, planning and a long time and the cooperative was not monitoring, etc. But none of these effectively shouldering the scheme responsibilities have been executed as management. The beneficiary farmers also desired for various reasons. acknowledged that the scheme was poorly maintained, and they attributed the problem 3.2.2. Performance of the Water User to lack of fund for maintenance. It was, Cooperative however, noted that farmers are neither contributing money for operation and Institutional arrangement on irrigation is maintenance nor do they pay for the required to overcome problems relating to irrigation water they use. irrigation water as a common property resource, to provide incentive to members disincentive for free riding and shirking. As 3.2.4. Conflict and Conflict Resolution mentioned earlier, the WUA was provided Mechanisms with a generic bylaws binding on all cooperative members. The first challenge in Conflict in the scheme both between and the functioning of the cooperative was the among the beneficiaries and the downstream principle clash in which cooperatives were farmers who are not served by the canal organized – i.e., membership ought to be water was said to be rampant. Among the 45 fully on voluntary basis. When farmers interviewed beneficiary farmers, 44 share the same irrigation water resource, acknowledged the presence of conflict. They

347

attributed water shortage to be often times agreement between upstream and the cause of conflict. The survey also downstream users, and there was no equity showed that there was no water sharing in water distribution (Table 2)

Table 2. Beneficiary farmers’ responses to conflict related questions Frequency Percent 1. Have you ever come across a conflict between those living

around the canal tertiary units and the downstream users? i. Yes 44 97.8 ii. No 1 2.2 2. What are the probable causes of the conflict? i. Unequal water distribution 4 8.9 ii. Irrigation schedule disturbed 25 55.6 iii. Scarcity of water 11 24.4 iv. Both i and ii 5 11.1 3. Is there a water sharing agreement among each of the branches? i. Yes 3 6.7 ii. No 38 84.4 iii. Have no information 4 8.9 4. Is water equally available to all users in the scheme? i. Yes 2 4.4 ii. No 42 93.3 iii. No response 1 2.2 5. Who are the actors in conflict resolution? i. WUA 4 8.9 ii. Kebele council/social courts 11 24.4 iii. Both i and ii 30 66.7

The survey also showed that the majority of laws. The majority of the water management the interviewed farmers (73%) had no problems revolve around the inability of the confidence in the capacity of WUC association to sanction offenders. The management to resolve conflicts. In fact, it tortuous legal processes in the judicial was learned that the major actor in conflict system and the lack of recognition of the resolution was the Kebele Council and cooperatives by-laws were the most serious social courts. The problem with Kebele challenges that the cooperatives were facing leadership and social courts in handling currently. disputes was that the majority of them were not among irrigation beneficiaries. Hence, 3.3. Support Service when WUC lodge complaints to the social courts, verdicts took a long time and the 3.3.1. Extension and Training penalty was never proportionate to the offence. Given the potential benefit that the scheme could provide to the beneficiary farmers and In conclusion, the WUA is not empowered the local community, a qualified extension to take punitive action and enforce its by- agent would have been imperative. It was,

348

however, learned that farmers were not getting advice from an irrigation agronomist or from a qualified development agent. There were no organizations committed towards providing farmers with the needful training and extension services. Table 3. Response of farmers to the kind of support they get from the government

What kind of support the scheme gets from the Frequency Percent government now? i Input supply 3 6.7 ii Advice 13 28.9 iii I to ii 6 13.3 iv No support 10 22.2 v Have no information 12 26.7 Total 45

3.3.2. Access to Market and Input observed in the study area to take advantage Supplies of soaring demand for vegetables either in nearby Bahir Dar nor Addis Ababa markets. The main Addis Ababa – Bahir Dar highway The problem was identified to be the fact is only some 5 kms from the scheme. The that a truck is required to transport to the status of dry weather road to the farms markets at Bahir Dar and Addis Ababa. during the study period was fair. The zonal However, such facility is not available to the town, Finoteselam, is only some 7 km from farmers. Trucks do not come to collect one the scheme. Market factor does not play a farmers pick as it would be too small to fill a role in what and when to produce. When truck. In principle, the solution for this farmers were asked whether they have problem would be to organize farmers to problem to sell their produce, their jointly plan and produce for the market on a responses were that they don’t have market pre-negotiated price. If this happens, the problem. However, when farmer was asked produce would be large enough to hire a to explain why many beneficiary farmers truck, and transport the produce where grow only potatoes, the response he gave market is available. was: From the survey results (Table 4) shortage ‘We grow potatoes because if we can’t sell and/or unavailability of seeds was also cited them in Finoteselam market, we can dry as another problem for diversification of them, store and use them for home crops under irrigation. This again showed consumption’. that the cooperative was not discharging the Drying potato to extend its shelf life was a fundamental cause for its formation, that is useful technology farmers practice to keep providing inputs and facilitating markets. produce when it is in excess. Therefore, it is evident that farmers are not producing for a serious market. No effort of farmers were

349

Table 4. Causes for inability to diversify crop production What are the problems relating to diversifying Frequency Percent irrigated Agriculture? i. Scarcity of seeds 37 82.2 ii. Plant diseases 0 0 iii. Market availability 0 0 iv. Lack of sufficient irrigable land 0 0 v. Low market price for output 1 2.2 vi. Items i, ii and iii above 2 4.4 vii. Items i and ii above 2 4.4 viii. No response 3 6.7 Total 45

3.4. Other Key Findings of the Survey d. The scheme had been under management of the government From the focus group discussion held with authorities until some four years beneficiary farmers and the community, the ago. Hence farmers did not feel that following important issues were captured: they own the water and the a. The WUC was not empowered to irrigation infrastructure. Farmers take administrative actions and were uncertain of the land tenure enforce its by-laws. The social arrangements. They reflected that courts do not use the cooperative the land they own today may be by-laws as valid instrument to taken away by the government at sanction penalties. Farmers use local any time for lease to investors. They adage saying that the WUC tried to justify this worry by citing management is a ‘Lion without the 20 ha land leased to an investor teeth’. in the middle of the scheme. b. The penalties that the offenders pay e. There was no mutual understanding fail to be commensurate with the to forge equitable distribution of infraction to deter him or her from irrigation water among all committing similar offence in the beneficiaries. Those located in the future. Moreover, the penalties that upstream end of the secondary canal the offenders pay were deposited could take as much water as they into Kebele’s account. Therefore, wish; consequently, not enough the WUC would benefit by taking water reaches the downstream users. an offender into a lengthy legal Hence, to avoid crop failure due to battle. water supply interruption at critical growth stages, the farmers located in c. At the time of this survey, there was lower end of secondary canal prefer no development agent (DA) to rely on rainfed farming. assigned to the scheme. Farmers f. Most tertiary and field canals reported that there was a DA with constructed as raised canals have natural resource management become dysfunctional to convey background but he stayed only for water as intended in the design. As less than one season. Hence, all the per the design document, siphon indications were that farmers were was used to divert water from the not benefiting from any technical canals to the furrows. But none of backstopping. the irrigators was practicing the

350

same. Farmers were creating their holding of a family. Thus providing own conveyance system bypassing access to as many farmers as developed canal and water control possible. structures. Consequently, there was 3. Level of awareness of farmers tremendous water loss when water towards marketable and high value was made to flow on temporary crop production was low. Farmers conveyance system. tend to be complacent with what g. The responses to the question of they have and produce. They do not ‘who owns the scheme?’ – the seem to be motivated towards government or the community – market oriented production. were illusive and it was difficult to 4. The irrigation skill of farmers was obtain a mutually agreed response. below satisfactory. Farmers do not Although farmers felt that the head appreciate, protect and attempt to work, primary and secondary canals use water conveyance and control were the responsibilities of structures established during the government, they could not pin irrigation development. point any them. From the local 5. Farmers were not paying for the governments end, those implicated provision of irrigation water. in one way or the other are the Moreover, no cost recovery attempt Agriculture and Rural Development has been initiated. Office, Cooperatives Promotion Agency and Water Resource Office. 4. Conclusion h. There was no enough consultation with all the stakeholders before The irrigation potential for Geray transferring the scheme irrigation scheme to change the life of management to the farmers. Hence the community is tremendous. It is farmers were uncertain of the endowed with excellent quality and government’s motive. quantity of irrigation water. The soil is highly suitable for irrigated agriculture. 3.5. Key Reflections Captured during the However, all the technical performance Discussions with the Staff Members of indicators showed that the scheme’s Woreda Agriculture and Rural performance was far from satisfactory. Development Division. Many of the water control structures have become dysfunctional. The 1. The absence of specific roles that underperformance of the scheme was the offices of agriculture, water, attributed to extreme neglect and lack of cooperative promotion agency supervision. These it is concluded that should play has made none of them the overall performance of the scheme accountable for the poor was unsatisfactory. performance of the scheme. 2. Many beneficiary farmers own The neglect was mainly attributed to excess irrigable land than they can lack of ownership. Further, many years manage. Some farmers have even of inefficient state management in the rainfed land outside the command absence of committed development area. Consequently, they do not agent took the major blame for the have the capacity and the infrastructure damage. Moreover, the commitment to manage both recent transfer of management to service irrigated and rainfed farms. cooperatives did not confirm to the Simultaneously, one alternative idea principle of long-enduring irrigation suggestion to solve this problem institutions. The fact that not all was to limit maximum irrigable land irrigation beneficiary farmers are

351

members of the cooperative which again Bos, M. G., Murray-Rust, D. H., Merrey, D shows that institute does have clearly .J., Johnson, H. G. and Snellen, W. B., defined boundary condition – another 1993. Methodologies for assessing requirement for good irrigation performance of irrigation and drainage institution. management. Irrigation and drainage systems. Vol. 7 No.4. Successful traditional schemes not very Daniel Kassahun, 2007. Rainwater far from Geray Irrigation Scheme have Harvesting in Ethiopia: Capturing the been perfectly functioning over 70 years Realities Exploring Opportunities. FSS using rudimentary irrigation structures Research Report No. 1. Forum for and extensive annual maintenance Social Studies, Addis Ababa. requirements. Hence, there should not Dessalegn Rahmato, 1999. Water Resources be any reason why this scheme should Development in Ethiopia: Issue of not have been successful provided that Sustainability and Participation. FSS proper institution had been crafted right Discussion Paper. Irrigation Practices, from the very beginning. The major State Intervention and Farmers’ Life- problem identified during this study was Worlds in Drought-Prone, Tigray, the absence of functional institution to Ethiopia. properly manage the scheme. The Gashaye Chekol, 2007. Technical and cooperatives association recently crafted Institutional Evaluation of Geray has no power either to sanction penalties Irrigiation Scheme, West Gojam, or enforce them. Under this scenario, Amhara Region. MSc Thesis, School of efforts made by the government to Graduate Studies, Haramaya University rehabilitate the structure may not bring Kloezen W. H. and Garces-Restrepo C., the desired changes in the life of the 1998. Assessing Irrigation Performance beneficiaries. Rather, the WUC needs to with Comparative Indicators: The Case be reorganized and made effective. of the Alto Rio Lerma Irrigation Uncertainties relating to land tenure District, Mexico. Research Report 22. need to be addressed at the earliest. International Water Management Most importantly qualified development Institute. Colombo, Sri Lanka. agent with irrigation agronomy MoWR (Ministry of Water Resources). background should be assigned to the 2001. Irrigation Development Strategy. scheme. Moreover, the DA should be Ministry of Water Resources, Addis given the responsibility to mobilize the Ababa. community, following the major Mekonen Leulseged, 2003. Irrigation Sub- network rehabilitation skills. sector situation analysis. Ministry of Water Resources, Addis Ababa (unpublished) Sarker, A and Itoh, T., 2000. Design principles in long-enduring institutions References of Japanese irrigation common pool resources. Agricultural Mater Bitew Genet, 2005. Status of Small –Scale Management. 48(2001) 89-102. Irrigation Projects in Amahra Region: Woldeab Teschome, 2003. Irrigation The Case of Adrako Micro-earth Dam, Practices, State Intervention and MSc Thesis, School of Graduate Farmers Life-Worlds in Drought–Prone Studies, Haramaya University Tigray. PhD Disertation, Wageningen, Bos M. G., Performance indicators for the Netherlands irrigation and drainage. Irrigation and Drainage Systems, 11:119-137.

352

Plate 4.2 Water flowing through the gate (Intake) Plate 4.3 Main Canal partly lined

353

Discussion on Theme 3: Irrigation Institutions and Support Services

Chair: Dr Kim Geheb Rapporture: Dr. G. Makombe

The chairman for this session introduced the theme and the floor was opened for questions, comments and suggestions.

Questions and Discussions established for marketing and inputs supply. Did you notice conflict 3.1 Institutions are linked with between formally established coops performance and farmer and community based/informal participation is a big issue to organizations? Koga is planned to improve performance of all have a coop. There is a hot debate schemes. WUA farmers need to in Amhara about institutional set up. organize themselves. Need to take into account traditional practice vs Ans: In some schemes some culture. Need to build on the members of WUA are not members existing traditional and cultural of coop and vice versa. There is practices conflict but it is managed because coops have upper hand over WUA. 3.2 Cost recovery is an important issue. E.g. members of WUA are not There is indication of interest by allowed to sell produce through farmers for cost recovery. From coop. experience few schemes not There is conflict and there is bad functioning because of wrong design attitude regarding/towards coops. usually fail because of management Many people were not interested to issues. be coop members. Koga: If coops could be managed Ans: Agree that WUAs are the only like cooperates this is the best option. way out. For larger schemes can be Water should have an owner. Koga cooperate or the government but for will be large and there will be small scale schemes WUAs need to segments. Water needs an owner and be empowered. Traditional schemes there should be cost recovery are successfully managed by social mechanisms exclusion. For modern schemes need to go to court but courts are 3.4 Should there be a market first then inefficient. There is need to farmers produce for it? empower the WUAs. Sometimes there is a water master or 3.5 Who owns the traditional systems and a WUA but the challenge is when who empowers the farmers to mange one asks what the role of each is them. Some are function well and when they are all there on a scheme. some not so well. What do we now There is mention that responsibilities about the management and are there should be transferred but only one of lessons to be learned. Farmer to them should be empowered. WUAs farmer learning was used well in Asia are not supported by government. from schemes working well to those Institutional conflict and confusion not working se well. To improve the exists. performance of traditional schemes is it physical infrastructure or ids it 3.3 There is no legal status for WUAs. institutional issues that need to be This is why irrigation coops were addressed? Traditional systems came

354

about because farmers did something courts then it is reported to kebele and when we build a scheme for court, then to district. When there is farmers we take away this option that a serious case of theft and when there farmers do something for themselves. is serous case of water and or abuse of power reported to court. Two cases Ans: There is incentive is getting per day were handled by the curt. access to water. Some farmers in remote areas were not interested in 3.7 At Indris water scarcity is the issue. canal clearance because they do not At Giray, there is excess water and see the benefit of irrigation. Those excess command area which is not who have access to information and managed by the farmers. What markets have incentive to produce. In determines the interest of the farmer Tigray there are people called to mange certain land. We have to production cadres who went to recognize that the government does primary school and become farmers . not accept the WUA but accepts the They are trained by extension to help coop. Coop should deal with water information flow. Demonstration and water management issues fields are usually showing optimum conditions that are not accessible to farmers. Farmers always need to 3.8 Geray: In this area there was a land adjust to their situation. redistribution in west Gojam zone. We should not generalize that the There is a land certification program. traditional irrigation is the best, but Is the process impacting the upgrading traditional schemes is a development of the irrigation scheme. good idea. Institutional set up should always consider local knowledge and Ans: One of the reasons farmers are institutions. not developing irrigation scheme is One irrigation scheme was fascinating. because of uncertainty abut land Why did it lasted for 50 years. There ownership. 28 ha were given to an was a penalty practiced for offenders. investor. Land certification had not An offender irrigating at night was reached there. Farmers are willing to fined 50 birr and an offender invest in land if they are clear about irrigating during the day was fined 30 ownership.. birr. This is whey they are working well because if immediate penalty. 3.9 These are good research works Water father said you teach us to exploring the local situations. Would conserve the soil, but annually we like to see in-depth why despite exporting soil to neighboring positive effects and nice looking countries. Why are we not crops, what is wrong? Why do constructing permanent structures if people despite the incentives, not farmers are doing so year after year? make proper arrangements to manage Why do we tend to construct new schemes. schemes but give them low support? Ans: Farmers could not maintain the 3.6 How do local courts handle water water intensive horticultural rights conflicts? What percentage of production for long because water conflict goes to local courts? became unreliable and farmers were obliged to shift from water intensive Ans: A substantial amount is horticultural crops to perennial crops. reported to courts. The committee The income has declining because of operates at three levels…..When it the effect of management problems. moves beyond capacity of local

355

Theme four:

Environmental and Health Impact of Irrigation

Papers 1. Case study review of investigated irrigation projects in Ethiopia 2. Environmental impact analysis of large-scale irrigation in Ethiopia 3. Brief manual “Keys for successful irrigation in Ethiopia” 4. Minimizing the negative environmental and health impacts of agricultural water resources development in Sub-Saharan Africa 5. Health and Environmental Analysis Results (10 minutes each) 6. GIS and remote sensing integrated environmental impact assessment of irrigation project in Finchaa Valley area 7. Entomological studies on the impact of a small-scale irrigation scheme on malaria transmission around Ziway 8. Irrigation and socio-economic factors related to malaria transmission in ziway, Ethiopia

Case study review of investigated irrigation projects in Ethiopia

W. Loiskandl1, D. Ruffeis1, M. Schönerklee2, R. Spendlingwimmer2, S. B. Awulachew 3 and E. Boelee3

1 University of Natural Resources and Applied Life Sciences, Department of Water - Atmosphere - Environment, Institute of Hydraulics and Rural Water Management, Vienna, Austria 2 Austrian Research Centers GmbH –ARC, Business Area Water Seibersdorf, Austria 3 International Water Management Institute IWMI, Addis Ababa, Ethiopia [email protected]

Abstract generic environmental and health issues as related to irrigated agricultural development” Within the project “Impact of Irrigation are presented. The collected material is Development on Rural Poverty and the compiled and made accessible through a data Environment” emphasis was given to the rural base. The case studies results are critically development of communities performing reviewed and conclusions for future field irrigation. Main goal was to analyze the investigation are drawn. potential of irrigation development, which should be performed in an environmentally Key words: Irrigation development, impact sound way to ensure good living conditions for assessment, environmental factors, Ethiopia future generations. Environmentally sound means to maintain soil fertility, water quality, Introduction to be concerned about health impacts and maintaining biodiversity. It is commonly agreed that irrigation intensification contributes to poverty The starting point for this review is an alleviation. Access to water, poverty and irrigation database, a classification of irrigation people's livelihoods are interlinked (Figure 1). schemes in Ethiopia and the resulting selection These linkages are both direct and indirect. of case studies of irrigation schemes. The Direct linkages operate via localized and study aims to support the clarification of household-level effects, and indirect linkages environmental factors and processes related to operate via aggregate or national level impacts irrigation development. The socioeconomic (Hussain and Hanjra, 2004). At the same time, implications of environmental impacts of water is becoming a scarce resource in many irrigation investments such as links to poverty, countries, particularly in Sub-Saharan Africa. health and policies and institutions will also be This is also reflected in the Millennium goals treated. The methods used include field of the UNITED NATIONS (2005) especially measurements and observations, laboratory in Goal 7: ensure environmental sustainability. analyses, structured questionnaire surveys and In the project “Impact of Irrigation PRA. Development on Rural Poverty and the Environment” this was one of the working The investigated field studies performed under hypotheses, but set in a wider environmental the project’s specific task “Assessment of and socio-economic context.

357

Hydrology health & poverty

Environmental frame Socio/economic frame

Soil quality Water quality Society

Figure 1: Interdisciplinary frames for case study investigations.

The overall goal is the understanding of collaborative work with partners, based on impacts of past interventions and investments their individual roles and mandates. This is in irrigation. This will significantly contribute supported by an EU sponsored co-project to the planning of new investments and the “Dissemination of research results in semi-arid design of interventions for enhancing and arid ecosystems with a focus on agricultural production. Emphasis is given to sustainable water resource management in biophysical aspects such as ground and surface Ethiopia" (short name WATERMAN). This water use, soil quality parameters, land use Specific Support Action (SSA) focuses on patterns, nutrient recycling, agrochemical use analysis and dissemination of research results and its associated risks, and wildlife and agro- in sustainable, integrated water resource biodiversity. The findings of the investigated management at river-basin scale in Ethiopia. case studies are of relevance for an improved The SSA takes advantage from the existing planning and managing of irrigation and water collaborations linking together partners from resources in Ethiopia and other countries with different regions in Ethiopia and the creation similar climatic conditions, especially in Sub- of an inter-Ethiopian-network is supported. Saharan Africa. Case study review of irrigation To quantify the significant positive and development in Ethiopia negative impacts of irrigation development in Ethiopia the following five specific objectives Irrigated agriculture is a priority of the were defined (Awulachew, 2004): agricultural transformation and food security strategy of the Ethiopian Government. 1) to generate information that can be Increased availability of irrigation and less used to improve the performance of dependency on rainfed agriculture is taken as a irrigated agriculture and enhance its means to increase food production and self- positive benefits while minimizing its sufficiency of the rapidly increasing population negative externalities, of the country. In line with the development 2) to guide future irrigation investments policy regional states and NGOs are promoting and fill the gap in knowledge about the irrigation development so as to increase and total impact of irrigation development on stabilize food production in the country. economy, society and environment, 3) to address specific health and poverty Under the 15- year Water Sector Development alleviation issues, Program (WSDP), irrigation development 4) to develop methodological guidelines subprogram, a total of 1606 small-scale for assessing the impact of irrigation irrigation schemes planned to be implemented investment, mainly for the provision of food requirements 5) to strengthen Ethiopian capacity for (Ministry of Water Resources, 2001). Foreign interdisciplinary research and political governments and multi-lateral agencies are implementation. expected to co-operate with the government of Ethiopia and Non-government organizations The assessment of irrigation practices (AIP) (NGOs) to foster this program. Other non- has to trigger actions. A prerequisite is the governmental organizations and communities

358

are also undertaking water resource • State farms operated by state owned development activities with the same enterprises, such as the Upper Awash objective. For example a large number of Agro Industry Enterprise. Water is earthen micro-dams and river diversions have abstracted from rivers or lakes by been built in the Awash Basin. Besides the pumps or diversions. Most of the development of new schemes, some traditional public irrigation schemes can be found systems are also being rehabilitated. in the Awash River Valley.

The irrigation scheme classification used in For detailed studies irrigation schemes were this paper follows the work of Philippe selected and investigations were performed Lempérière who divided irrigation projects in according to the various assignments of the Ethiopia in four groups, (cited by Werfring, project. The following aims to synthesize the 2004): case studies, highlights the main findings and • Traditional irrigation schemes that reflects them in terms of the assigned have been practiced for centuries by deliverables of the project. Out of 26 project using perennial or seasonal streams. case study sites eight were selected for the These schemes usually were environmental impact assessment (Table 1, developed by the farmers themselves Figure 2). Included in Table 1 is also a case without any government involvement. study outside the project from Dominik Ruffeis (2006c): the Lomi Wuha irrigation • Modern communal irrigation scheme, for which the same approach was implemented by regional governments. used. The investigation of Wagnew Ayalneh Rivers and run-off water, lakes, (2004) was performed at a selected case study springs and groundwater are used. site, but not as part of the project. In total 12 Generally modern communal irrigation case studies, performed by ten investigators schemes are more sophisticated than from Ethiopia and Austria, are analyzed in this traditional ones. report. Additional information is provided • Modern private irrigation schemes when accessible and relevant. The fieldwork started in the 1950s initiated by Dutch took place at 13 irrigation schemes in three companies that implemented sugar mayor river basins of Ethiopia, with Awash estates. With the adoption of a market River basin, where most irrigation activities based economy the private schemes re- happen, being most prominently represented. emerged in the 1990s. Water is mainly The presented case studies are a good sample from rivers or lakes by pumps or in terms of different irrigation scheme sizes diversions, although some small farms and management practices. use water harvesting techniques.

359

Table 1. General case study site specification, numbers are given for each individual report, e.g. thesis works or unpublished documents. Nr Investigator Scheme Basin Type of irrigation 1 Wallner/Ruffeis Wonji/Shoa Sugar Upper Awash State Farm/Gravity Plantation 2 Wagnew Doni, Batu Degaga, Upper Awash Modern/Traditional Ayalneh Markos, Godino Communal/Gravity 3 Ruffeis Godino Upper Awash Modern Communal/Gravity 4 Damtew Goha Woriko Upper Awash Modern Communal/Gravity 5 Zewdie Amibara II Middle Awash State Farm 6 Ebissa Ziway Middle Awash Modern Holota Upper Awash Communal/Gravity 7 Wallner/Ruffeis Indris/Guder Blue Nile Modern Communal/Gravity 8 Ruffeis Lomi Wuha (MfM) Blue Nile Traditional/Gravity 9 Ruffeis Finchaa Valley Sugar Blue Nile State Farm/Sprinkler Estate 10 Ahmed A. Finchaa Valley Sugar Blue Nile State Farm/Sprinkler Estate 11 Judt Hare Rift Valley Modern Communal Lakes Traditional/Gravity 12 Ruffeis Hare Rift Valley Modern Communal Lakes Traditional/Gravity

360

Indr Mark

Doni

Figure 2: Investigation field site locations

Presentation of field sites and discussion of results For better comparison of the results obtained at different locations the most important analyzed The investigated field sites should be parameters and the main focus of the considered in the broader context of irrigation respective studies are presented in an in Ethiopia. The river basin development investigation matrix (Table 2). The summary master plan studies of the Ministry of Water of performed work shows a concentration on Resources (2001) provide an excellent water and soil analyses in a strong relation to overview of intervention potential in Ethiopia. environmental issues. For most of the Experiences and opportunities for promoting investigated sites the other aspects of poverty small-scale/micro irrigation and rainwater alleviation, health and socio/economic harvesting for food security in Ethiopia are considerations were treated as well or were presented by Seleshi B. Awulachew et al. available from other sources. Further detailed (2005). Another broad overview of irrigation information may be obtained through the cited development in Ethiopia covering all river reports, thesis works and other literature basins is provided by Seleshi B. Awulachew et sources. al. (2007).

361

Table 2. Investigation matrix Nr Environment/ Water Soil Poverty Socio- natural quality quality alleviation/ economic resources health 1 XX XXX XXX X Awash 2 XX X X XX XXX 3 X XXX XXX X 4 X XXX XXX X 5 XXX X XX X XXX 6 X XXX XXX XXX1) X 7 X XXX XXX X Blue Nile 8 X XXX XXX X 9 X XXX XXX XXX2) X 10 XXX XX XX XX X 11 X XXX XXX Rift Valley 12 XX XXX XXX XXX3) Lakes Investigation XXX high, XX medium, X low 1) Alemu (2007) & Kibret (2008), both for Ziway only; 2) Chala (2007); 3) Mabedo (2003)

As a general characterization of the Figures 3 and 4. For each irrigation scheme a investigated field sites various climatic data short description and main results are provided and the elevation, which are also important in in the subsections below. relation to health aspects, are presented in

Annual N Elevation 45

1400 2500 40

35 1200 2000 30 1000

1500 25 Tmin 800 20 Tmax 600 1000 Elevation (m) Temperature (°C)

15 (mm) N annual 400 10 500 200 5 0 0 0 i r a n s o o II e o ga o in ik ota ley ni ta D a r iway l oa aki o ey Akaki Z o al Hare h Do k l uha i/Sho God ib ara H A ara II Ziway o all Hare Deg Mark a V i/S H u m ris/Gud nj Degaga Markos Godino Woriko s/Guder i W A d ha m Wonj In Lomi Wuha ha Amib Bat Goha Wo inc Wo atu o Indri Lo F B G inchaa V F Figure 3: Temperature values of case study sites Figure 4: Elevation and annual precipitation of (NMA, 2006) case study sites (NMA, 2006)

is visibly well managed. Awash River is the Wonji/Shoa Sugar Plantation irrigation source of the scheme which is The first large public irrigation scheme located down stream of Koka dam. Koka dam Wonji/Shoa Sugar Plantation was constructed was originally constructed for hydropower in 1956 as a private farm by HVA (Handels production in 1960. Irrigation water is diverted Vereniging Amsterdam). The sugar production from the Awash River with electrical pumps is now subordinated under the authority of a and the plantation is managed as a gravity government agency (Ethiopian Sugar system. Excess water is stored in night storage Company). The scheme is located about 80 km reservoirs which cover a total surface area of south east of Addis Ababa, 15km south of the 60 ha. Wastewater from the sugar factory and town of Nazareth in the Upper Awash Valley. domestic supply is blended with freshwater The irrigation schema has good structures and and used for irrigation on some of the fields.

362

The whole irrigated area (6000 ha for the state The study of these small-scale irrigation farm, additionally 1000ha for the out-growers) schemes in the Awash Basin revealed some is drained by a constructed, open surface factors that are important for the successful drainage system. implementation of small-scale irrigation schemes. The system of furrow irrigation, Girma A. (2005) states that in most sugarcane which is practiced in most of the schemes, has fields of Wonji/Shoa slight to moderate soil high labor demands and hence some farmers infiltration problems are expected if practice a flooding system. This will aggravate groundwater occurs at shallow depths. Field erosion especially in sloping plots. High observations confirmed that the groundwater electricity, repair and maintenance cost of depth was at all the times too shallow (in the pumps in Batu Degaga showed that electric root zone) for nearly all sugarcane fields powered pumps might be too costly for considered in this study. Hence in Wonji/Shoa, smallholder farmers (Ruffeis, 2006a). Systems sugarcane fields are prone to water logging and that require less cost like gravity diversion physical degradation of topsoil (Wallner, 2006; systems should be looked at seriously. In the Ruffeis et al., 2007). analysis of the schemes it has come out clearly that small-scale irrigation projects built or upgraded by NGO’s and government are often Doni, Batu Degaga, Godino, Marko, handed over to the farmers without proper Goha Woriko - Oromyia Region completion of construction or technical The Godino and Goha Woriko irrigation training and without proper management schemes are on both sides of the Wedecha establishment. This creates problems at such River, located in East Shewa zone, in Oromia schemes as farmers are left with the region near the town of Debre Zeit about 70 understanding that the government or the NGO km from Addis Ababa. The schemes are part are still responsible. of a cascade system which includes two earthen dams and reservoirs. The upper dam which is located north of the two schemes Amibara II Godino and Goha Worko is called Wedecha Amibara Irrigation Project II is located in the dam and is the main water source of both south-eastern flood plain of the Awash River schemes. The dam was constructed in 1980 by 250 km away from Addis Ababa, in a typical Kuba Construction Brigade having the semi-arid agro ecological zone with an Agricultural Development Minster of Ethiopia extensive pastoral production system based on as a client. Wedecha dam lies at an altitude of with camels, cattle, goats and sheep. In 1980 2437 m.a.s.l. and the two schemes on an the Amibara Irrigation Project II (AIP II) was average altitude of 2010 m.a.s.l (Damtew, established and production is mainly focused 2006). on cotton for foreign markets. Socioeconomic and environmental data were collected through

informal discussions with the local community Wagnew Ayalneh (2004) gathered information elders who were relocated from their original at household level on crop production, market living area due to the irrigation development and irrigation management. Investigated are (Zewdie, 2005). Changes in the living four small-scale irrigation schemes, namely conditions (social and cultural settings) of the Doni scheme from Boset woreda, Batu Degaga rural Afar people were recorded. In addition to scheme from Adama woreda, Godino scheme the survey data, secondary data on surface and from Ada Liben woreda and Markos scheme ground water hydrology, production and land from Wolemera woreda. The selection was productivity, and primary data on soil physical based on accessibility, experience and type of and chemical properties were collected. scheme. 240 farmers were selected from the respective peasant association (30 irrigators In this system the irrigation development and 30 non-irrigators for each woreda). The considerably contributes to the local economy total population of irrigators and non-irrigator through establishing profitable enterprises and farmers of the four peasant associations are improving the livelihoods of thousands of 627 and 3207 respectively. people (mainly migrants who newly settled in

this area) through providing job opportunities

363

and good income. On the other hand land through the type of investment farmer diminishing livestock holding forced the rural make in land management. pastoral community to lead under subsistence lives and some of them to migrate to towns in The overall study indicates that community- search of work. So the total contribution of the based irrigation schemes are more feasible at irrigation development in improving the highlands (Holota) than at lowlands due to livelihoods (income) of the native community health impacts and water and soil deterioration, of the area is none, even negative. probably linked to the original water quality of the source. Improving the local supporting Ziway and Holota infrastructure such as health centers, transport, Ziway is located 170 km from Addis Ababa on input supply institutes and credit organizations the way to Awassa. At Ziway, the study was would greatly help to contribute to the carried out in the peasant association (PA) sustainability of the systems. An EIA with the called Edo-Gojola, which was established in active participation of the whole community 1992 by SEDA (Selam Environmental on the onset of the project could have removed Development Association). The total command or at least minimized the negative impacts of area of the scheme is 2440 ha whereby only 32 the irrigation scheme observed on health, ha of the farm are currently operational. Water environment and social conditions. for irrigation is diverted from Lake Ziway using motor pumps and distributed to the fields by gravity (Ebissa, 2005). Rainfall is not Indris satisfactory and as a result low yields have The Indris irrigation scheme is located in the been recorded that need to be supplemented western Shewa administrative region of Ambo with irrigation. Awraja, about 110 km west of Addis Ababa and about 10 km west of Ambo near the town Holota is situated approximately 45 km west of of Guder. The scheme was established in Addis Ababa. At Holota the study was carried 1985/86 to produce vegetables on out in the Misrak-Shola-Ber peasant approximately 400ha of land using irrigation association (PA), which was established before (Halcrow & partners, 1989). In 1991 further 1971. The total irrigated area is 90 ha (Ebissa, plans had been made to modernize the 2005). The water source of the scheme is irrigation scheme as the old structures like the Holota River. The irrigation water is diverted weir, flumes and the channels were in poor from the river and distributed to the fields condition and partly damaged (OIDA, 1991). using gravity. The total area of the site is 680 Since about 1995 the command area of the hectare of which irrigation occupied 90 system is approximately 1000ha. The whole hectare. Clinical data from the health center irrigation system is divided into several data and a questionnaire were used to generate smaller irrigation schemes situated along Indris additional information. River, which is its water source. The water of the investigated schemes is diverted by two Technical skills and innovative irrigation different dams situated about 10km and 2km technology are absent in these two schemes. south (upstream) from Guder, respectively. Both irrigators and non-irrigators use similar types of land management techniques Severe downstream impacts of irrigation have (traditional technology). Some irrigators were been noted under low flow regimes (Ruffeis, found to increase their production by 2006b; Wallner, 2006). In the dry periods expanding land size and changing to other land practically the whole discharge is diverted uses while they might have better improved from the river into the irrigation system. their poor land’s condition through intensified Directly after the diversions only water management. Provision of training to empower seeping through the diversion dam supplies the farmers with new land management techniques riverbed. Tributary rivers of Indris are is crucial. There is a need to have alternative providing the discharge for the second river water sources to alleviate the water shortage diversion. problem at Ziway and Holota. Land ownership and land use policies are very important as these determine the productive lifespan of the

364

Lomi Wuha people. It has also many socio-economic This case study site is located in Merhabete benefits for the valley and surrounding people. district in Northern Shoa zone of the Amhara In addition Finchaa Sugar factory plays a key Regional State about 180 km north of the role in addressing the current sugar demand on Addis Ababa. The topography of Merhabete is the local market. There are many efforts to predominated by two valleys formed by the exploit the by-products of the factory for Jema and Wenchit Rivers. The infrastructure additional purposes like using ethanol for fuel. of Lomi’s irrigation scheme is very basic. The estate has also an important role in the Three springs or spring areas are used as water growth of national GDP and GNP (Ahmed source for irrigating the farmland. The water is Amdihun, 2006). collected in natural basins and is diverted with two earth canals supplying different parts of The analysis of measured water parameters at the scheme. The size of the command area of different dates shows no significant differences the system is approximately 24 ha and on-field (Ruffeis et al., 2007). The soil degradation is in water distribution is done with furrows. its early to moderate stage and not difficult to be addressed by alleviation measures. No Main hydrological impacts result from reduced further adverse impacts are to be expected run-off water, but there are no downstream caused by the water source, though malaria users (Ruffeis, 2006c). A conflict appears transmission seems to have increased because between the destruction of natural vegetation of the irrigation (Chala, 2007). in the adjacent spring area versus food security in the region. Soil conservation and protection Hare against erosion and land degradation is Hare River irrigation schemes are located in provided through irrigation because of a the Gamo Gofa Zone, in the Southern Nations permanent vegetation cover. Soil physical and Nationalities Peoples Regional State of analysis results compared with parent material Ethiopia (SNNPR), about 495 km south of may explain the different soil types within the Addis Ababa. The irrigation systems are area. located between the shore of Lake Abaya and the escarpments of the highlands. Hare Finchaa Valley Sugar Estate Irrigation System comprises three different Finchaa Valley is located about 330 km west irrigation schemes with a total irrigable area of of Addis Ababa in the western part of Ethiopia, 2224 ha belonging to four kebele eastern Wollega zone in Oromiya region. In administrations, Kola Shara, Chano Mille, 1975 the state farm was established that mainly Chano Chalba and Chano Dorga. The three produced food and commercial crops until irrigation schemes use Hare River as water 1991. Starting from 1991 up to now more than source but have to be classified differently due 8064 ha of land has been cleared and irrigated to differences in abstraction and delivery for sugar production (Ahmed Amdihun, 2006). structures (Ruffeis, 2006d). Chano Dorga can The water source is the reservoir of Finchaa be classified as traditional. In the case of Kola Dam (Lake Chomen) initially built for Shara a traditional delivery system is used to hydropower production. 25 pumps at 5 pump allocate the irrigation water that is diverted stations divert the water serving 34 % of the from Hare River by a modern diversion command area. 66 % of the area is irrigated structure. Chano Chalba and Chano Mille can using gravity off-takes at 3 locations. The be classified as modern. Finchaa Valley Irrigation System is equipped with sprinkler irrigation devices. Farmers complain about insufficient amounts of water for irrigation. The cause is decreasing Observations made during the field visit rainfall over the last decade in the highlands, indicate that the obvious malfunction of the which minimized the discharge of Hare River waste water treatment plant of the sugar and therefore the availability of irrigation factory poses a threat to downstream water water. The mentioned problems related to the bodies, especially to Finchaa River and its high variability and low availability of water ecosystem. The project opened up large scale disproportionally affects farmers having their job opportunities for many thousands of farmland located at the tail end of the scheme. In order to solve these allocation and

365

availability problems one possible solution an easily comparative way. An example of a could be the construction of a reservoir to store checklist applied at the fields sites of Wonji the run-off water of Hare River (Judt, 2007). and Indris was provided by Wallner (2006) in On the other hand this might have severe her work. impacts on the hydrology of Lake Abaya, due to a reduction of annual inflow rates. The Most important general findings for locations options for design and water management to with similar conditions are summarized in control malaria in this region were investigated Table 3. In this table a picture of the project by Ashenafi Madebo (2003). investigations is provided and hence we do not claim it is complete. Still it may serve as a starting point for sharing knowledge and Summary of irrigation development experiences. The impacts shown are the impacts on the environment challenges for the future and need to be From the results a standardized sampling considered in irrigation planning and procedure such as checklists and guidelines for management activities. The more we know the field work could be developed for future more we can optimize our interventions for the investigations. Standardized does not mean benefit of poor people. that the investigations are exactly the same in each site, but that the results are presented in

Table 3: Impact matrix of irrigation development on the environment and health Nr Scheme Basin Hydrology/ Water Soil quality Poverty Others natural quality alleviation resources 1 Wonji Rising water Slight EC Water Heavy table, Seepage of increase logging machines reservoirs Infiltration 2,3 Oromyia Inefficient water Stagnant Salinity Marketing, Need for 4 use (except water poor access education, A Marcos) to health management W facilities Water A conflicts 5 Amibara S Flood hazard, Linkage with Salinity questioned livestock, II H degradation Lake Beseka Negative (very high Social sodium cont.) effects 6 Ziway Toxicity Salinity Malaria, New plant Permeability Market diseases, Holota increase access training 7 Indris Water diversion Risk of water detailed soil logging sampling B Low fertility needed 8 Lomi L Erosion risk Nutrients Wuha U deficit, E alkaline 9, Fincha Destruction of Low EC Low organic Employment Impact of 10 N Ecosystem matter, possibility vs. sugar I Infiltration destruction of factory on L ecosystem, downstream E Malaria water body 11, Hare Degradation, Water logging Low organic Malaria, 12 Rift erosion, matter questioned Valley deforestation Lakes

366

Conclusion care, improved communication or a rural credit system could also support the development of The assessment of irrigation practices (AIP) rural areas. claims not to be a new method, but tries to take advantage and adapt existing practices for irrigation projects. A step forward to quantify The role of universities is to connect research the significant positive and negative impacts of and education, to share and support partners in irrigation development in Ethiopia was the southern hemisphere, which is clearly achieved and the risk assessment of future visible by the “work force” named in this irrigation investments is supported. The report. The importance of education to development of methodological guidelines for improve livelihood and food security is not assessing the impact of irrigation investment questioned. Education in this context must has started. The knowledge of the total impact reach out to all stakeholders involved in of irrigation development on economy, society irrigation practice. Especially education and and environment could be substantially training in water management, marketing and improved. Health and poverty alleviation general crop production is of high important. issues have been addressed to some extent. Finally it is believed that this project provides Data sheets and concepts for field a contribution to strengthen Ethiopia’s capacity investigations are developed and applied at for interdisciplinary research and political field sites. Standardization of the assessment of implementation. irrigation practices and field investigations is especially useful for designing fieldworks and the comparison of results. The development of References guidelines for decision makers, planners and farmers depends on standardized methodologies and procedures. Ahmed, Amdihun, (2006). Geographic Information Systems and Remote Sensing Integrated multi-purpose water utilization for Integrated Environmental Impact irrigation has to be considered. This increases Assessment of Irrigation in Finchaa the demand for good management and the Valley, MSc. Thesis, Addis Ababa awareness of possible conflicts that may arise University, Ethiopia (10) due to competing interests. Efficient use of Alemu, Yihenew, (2007). Irrigation and socio- irrigation water is required to avoid water loss economic factors related to malaria and to control vector breeding and water- transmission in Ziway, eastern Oromiya related diseases. Well planned and maintained zone. MSc thesis Biomedical Sciences, small-scale irrigation schemes are a Addis Ababa University Ethiopia (6) contribution to economic development through Awulachew, Seleshi B., (2004). Work Plan for increased incomes, employment creation and the Research Project on Impact of food security. Negative environmental effects Irrigation Development on Rural Poverty could be minimized. Salinity and its mitigation and the Environment, International Water measures are a mayor concern. Large scale Management Institute (IWMI) and irrigation schemes show mostly a better Universitaet fuer Bodenkultur Wien. management and maintenance, but are at risk Awulachew, Seleshi B., Merrey, D.J., Kamara, of more negative environmental impacts. A.B., Van Koppen, B., Penning de Vries A major challenge concerns the application of F., Boelee, E., Makombe, G., (2005). hydrological research for surface and Experiences and opportunities for groundwater resources development and promoting small-scale/micro irrigation and management, surface water harvesting and its rainwater harvesting for food security in effect on groundwater recharge with Ethiopia. IWMI Working Paper 98, implication to the conjunctive use of water International Water Management Institute. resources. Infrastructure and market access of Awulachew, Seleshi B., Aster Denekew Y., is a key feature of the success of an Makonnen Loulseged, Loiskandl, W., intervention. Other issues like improves health Mekonnen Ayana and Tena Alamirew,

367

(2007). Water Resources and Irrigation Ministry of Water Resources (2001). Data Development in Ethiopia, International collected from different river basin Water Management Institute (IWMI), development master plan studies. Planning Addis Ababa. and Projects Department, Ministry of Chala, B., (2007). Assessment of malaria as a Water Resources, Addis Ababa, Ethiopia. public health problem in Finchaa Sugar National Meteorological Agency (2006). Factory. MSc. thesis, Biology Department, Meteorological Database, NMA, Addis Addis Ababa University. (9) Ababa, Ethiopia. Damtew, A., (2006). Evaluation of Gowa OIDA (1991). Project Proposal Report for Worko small-scale irrigation scheme at Indris Irrigation Project; RID-Office for Ada district, Oromia. A Thesis Submitted C.Z., Oromia Irrigation Development to the School of Graduate Studies through Agency. Department of Agricultural Engineering Ruffeis, D., (2006a). EIA report of Godina and Graduate Council of Haramaya University, Goha Woriko irrigation schemes, Ethiopia (14) unpublished project report, Austrian Ebissa, Gizaw, (2005). Health, Environment Research Centers, Seibersdorf, Austria. (3) and Socioeconomic Impact Assessment of Ruffeis, D., (2006b). EIA report of Indris Community-Based Irrigation at Ziway and irrigation system, unpublished project Holota. Master Thesis at the school of report, Austrian Research Centers, graduate studies-environmental science, Seibersdorf, Austria. (7) Addis Ababa University. (6) Ruffeis, D., (2006c). EIA report of MfM Girma A., (2005). Evaluation of Irrigation schemes, unpublished project report, Water Quality in Ethiopian Sugar Estates Austrian Research Centers, Seibersdorf, (Research Report). Research and Training Austria. (8) Services Division, Ethiopian Sugar Ruffeis, D., (2006d). EIA report of Hare Industry Support Centre Sh. Co., Wonji. irrigation system, unpublished project (1) report, Austrian Research Centers, Halcrow W. & Partners (1989). Master plan Seibersdorf, Austria. (12) for the development of Surface Water Ruffeis, D., Loiskandl, W., Spendlingwimmer, Resources in the Awash Basin, Final R., Schönerklee, M., Awulachew, Seleshi Report (Vol. 8, Annex J), Ministry of B., Boelee, E., Wallner, K., (2007). Water Resources of Ethiopia, Ethiopian Environmental impact analysis of two Valleys Development Authority, Irrigation large scale irrigation schemes in Ethiopia. and Drainage, Addis Ababa. Paper presented at Symposium on Impacts Hussain, I., Hanjra, M. A., (2004). Irrigation of Irrigation on Poverty and Environment, and Poverty Alleviation: Review Of The Addis Ababa, Ethiopia, 26-29 November empirical Evidence. Irrig. and Drain. 53: 2007. (1), (9) 1–15. DOI: 10.1002/ird.114. United Nations (2005). UN Millennium Judt, Ch., (2007). Impact of irrigation on Development Goals, Un Web Service livelihood and food security in the modern Section, department of Public Information, Hare river irrigation scheme in Southern http://www.un.org/millenniumgoals/. Ethiopia, Diplomarbeit Universität Wien, Wallner, K., (2006). Field Parameter Department für Ernährungswissenschaften, Evaluation to Support Environmental Fakultät für Lebenswissenschaften. (11) Impact Analysis of Irrigation in Ethiopia, Kibret, Solomon, (2008) Entomological MSc. Thesis, Institute of Hydraulics and studies on the impact of a small-scale Rural Water Management, Department of irrigation scheme on malaria transmission Water, Atmosphere and Environment, around Zeway, Central Ethiopia. MSc University of Natural Resources Applied thesis Biomedical Sciences, Addis Ababa Life Sciences, BOKU Vienna. (1), (7) University Ethiopia (6) Wagnew Ayalneh, M., (2004) Socio economic Madebo, Ashenafi, (2003). Design and water and environmental impact assessment of management to control malaria in Arba community based small-scale irrigation in Minch, Ethiopia. MSc Thesis Soil and the Upper Awash Basin. A case study of Water, Wageningen University, The four community based irrigation schemes. Netherlands. (12) Master Thesis at the school of graduate

368

studies-environmental science, Addis Zewdie, Moltot, (2005). Environmental and Ababa University. (2) Socio-economic Impact Assessment of Werfring, A., (2004). Typology of Irrigation in Irrigation Development at Amibara Ethiopia, Master Thesis at the Institute of Irrigation Project II, Master Thesis Post Hydraulics and Rural Water Management, Graduate Program Department Of Universität für Bodenkultur Wien. Irrigation Engineering, Arbaminch University. (5)

369

Environmental Impact Analysis of Two Large Scale Irrigation Schemes in Ethiopia

D. Ruffeis1, W. Loiskandl1, R. Spendlingwimmer2, M. Schönerklee2, S.B. Awulachew3, E. Boelee3, K. Wallner1

[email protected]

Abstract adjusted sodium ratio, which leads in some instances to soil crusting and has a negative This article presents the finding of a study impact on infiltration rate. In Wonji/Shoa the undertaken to assess the status-quo and groundwater table has risen due to improper significant environmental impacts of two irrigation management and seepage of selected large-scale irrigation on natural reservoirs. In Fincha a valuable ecosystem has resources in Ethiopia. Main focus is on the been destroyed due to the establishment of the environmental impacts of irrigation on natural scheme and increased migration. resources with special emphasis on soil quality, water quality and downstream impacts, Key words: hydrology and potential interference with Large-scale Irrigation / Environmental Impact / ecosystems. For this purpose two schemes Water / Soil / Ecosystem / Ethiopia / Finchaa / were selected. Wonji/Shoa Sugar Plantation is Wonji/Shoa located in the Upper Awash Basin and Finchaa Valley Sugar Estate located in the Blue Nile Introduction Basin. Under the title “Irrigation Infrastructures It is well known that irrigation projects can Development for Food Security and National have several adverse environmental impacts Economic growth“ the Ethiopian government that may threaten the sustainable production of started an irrigation development program in agricultural goods, which is of major order to meet the ambitious efforts of the 15 importance and interest in Ethiopia since it years Water Sector Development Program, contributes 44 percent to Ethiopia’s GDP, which is scheduled from 2002 until 2016. The employs 80 percent of the labor force, and investment in large-scale irrigation projects provides a livelihood to 85 percent of the should guarantee to meet the demands for nearly 80 million population (Awulachew, industrial raw materials for Agro-industries, 2006, Government of Ethiopia, 2006, UNDP, cash crops and food crops. 26 medium and 2006). large-scale irrigation projects are planned to be implemented. Irrigation infrastructures in the Irrigation projects inter alia can have potential country will more than double at the end of the impacts on the hydrological characteristics of program period, which is the year 2016 aquifers, quality of downstream water bodies, (currently only 200,000ha, this will grow by quality of soils and ecosystems. 275,000 ha). More recently, the irrigation sector development program is revised and the The most prominent results and environmental country has stepped up its efforts to bring impacts of the selected case study sites could additional irrigated area of 430,000 in five be summarized as follows. In general the years, see MOFED PASDEP (2006) irrigation water is of good quality, but the electric conductivity is unfavorable to the

370

In Ethiopia, irrigation projects with command This article presents the finding of a study areas larger than 3000 ha are classified as large undertaken to assess the status-quo and scale schemes according to MoWR (2002). significant environmental impacts of large- Werfring et al (2005) distinguish between four scale irrigation in Ethiopia. The main focus of different types of irrigation schemes in this report is put on the environmental impacts Ethiopia: traditional, modern communal, of irrigation on natural resources with special modern private and public. Most of the emphasis on soil quality, water quality and medium and large scale irrigation schemes are downstream impacts, hydrology and potential located in Oromia and Afar region interference with ecosystems. respectively. According to Tilahun and Paulos (2004) 31.981 ha and 21.000 ha of irrigated In several studies a number of different land are classified as large or medium scale environmental impacts have been identified schemes in these regions. In total 61.057 ha are which are directly caused by irrigation identified as large or medium scale irrigation projects. Sectoral guidelines to conduct schemes in Ethiopia. environmental impact assessments of irrigation projects (e.g. FAO, MoWR) use checklists Since the Ethiopian Government has started to which include the pertinent environmental focus its development strategies on the impacts. These potential impacts are grouped extension of irrigated agriculture especially of into impact categories such as economic, large scale projects during the last decade it socio-economic, natural resource and has become more important to explore the ecological impacts. This article puts its focus nexus between irrigation investments, on impacts on natural resources and sustainable agricultural development and ecosystems which are closely related to in-field potential environmental impacts in the impacts on soil, water quality, hydrological Ethiopian context. In Ethiopia very little issues and destruction of ecosystems due to information and baseline date are available irrigation development. In Tab.1 some of the regarding irrigation and its environmental potentially significant adverse impacts are implications. Hence research has to be listed. undertaken to fill this knowledge gap.

Table 1: Typical environmental impacts of irrigation on natural resources Impact category Potential adverse impact Cause Rise of local water table, Improper irrigation management, Hydrology Water logging, Changes to the low flow Poor water distribution system regime Low irrigation efficiency, Seepage losses Quality of Pollution of soil, Toxicity of products Inadequate assessment of the quality of irrigation source supplied water Downstream water Pollution of downstream water bodies Improper use of fertilizer and pesticides quality Alkalization Quality of irrigation source Long term leaching, Soil acidification Improper use of fertilizer Saline groundwater,Saline irrigation water, Soil degradation & Salinisation Saline soils, Improper irrigation management, damage of soil Insufficient drainage structure & change Intensive cultivation without additional of soil properties Reduction of fertility amendments

Improper irrigation management, Soil Water logging degradation, Raise of water table, Improper drainage Erosion Operation, Construction of the scheme Ecosystem, Creation of aquatic habitats Construction of reservoirs and canals Water bodies, Scheme construction Destruction of ecosystem within the

371

Wetlands project area Eutrophication and pollution of water Reduced quality and amount of return flows bodies, Change of return flows Destruction of ecosystem in the adjacent Migration tendencies, Improper land use area of the project strategy

Methodology grey cracking clays in the topographic depressions and semiarid brown soils. On the Study area basis of texture they are categorized into light (coarse textured) soils and heavy soils (clayey The study areas have been chosen for their black types) which are more common in diversity in terms of climatic conditions as Wonji/Shoa Sugarcane Plantation (Ambachew well as soil types. Two large scale irrigation et al, 2000). projects in Ethiopia were selected as case studies where primary and secondary were The soils in Finchaa valley are made of alluvial collected, Wonji/Shoa sugarcane plantation and colluvial materials from the surrounding and Finchaa Valley Sugar Estate. escarpments. Five major soil types can be The major predominant soil types in the area of found in the area of Finchaa Sugar Estate of Wonji/Shoa sugarcane plantation are described which Luvisols and Vertisols are predominant. as Fluvisols, Andosols and Laptosols These soils account for more than 95 percent according to FAO soil classification. Soils of of the cultivated and irrigated land. The Wonji/Shoa are of alluvial-colluvial origin irrigation scheme is divided by Finchaa River developed under hot, tropical conditions (Fig. into East and West Bank (fig. 2). Currently 1). In the region diverse soil types are observed only the West Bank is under cultivation, but which also vary in their production potential. the extension of the irrigated area to the East (Ambachew et al, 2000). In general, soils of bank is planned. Wonji/Shoa can be described as a complex of

Table 2: General description of the selected case study sites Finchaa Valley Wonji/Shoa Size 8064 ha 7000 ha Location 9°30’-9°60’N; 37°10’- 8.40°N; 39.25°E 37°30’E Altitude 1550 m 1540 m Date of 1995 1956 Establishment Type Public, large scale irrigation Public, large scale irrigation Management Government Agency Government Agency Basin Blue Nile Basin Awash Basin Water source Lake Chomen, Finchaa River Awash River, Reservoir (reused water from factory) Diversion Pump Pump Irrigation Pump, Gravity, Sprinkler Gravity, Furrow, N/ETo (mm/y) 1300/1500 747/2519 Agro-ecology Weyna Dega (1500-2300m) Weyna Dega (1500-2300m) Main crop Sugarcane (monoculture) Sugarcane (monoculture) Major Soil Luvisol Fluvisol Type

372

Figure 1: Scheme layout of Wonji/Shoa Sugar Plantation

Figure 2: Scheme layout of Finchaa Valley Sugar Estate

about the topology and geological Data Collection and Analysis characteristics of the area and about irrigation infrastructure. Research and data collection Both secondary and primary field level data have been conducted using the “with or were collected. Secondary data included without” and when ever possible the “before general description of the scheme, hydrological and after” approach comparing secondary and and meteorological relevant data, information primary data sets. The availability of

373

secondary soil and water quality data in the water bodies using 1 litre plastic bottles and two case studies allows, to some extent, impact immediately stored in a cooling box. As some analysis using the “before – after” approach. parameters may change rapidly after sampling Whenever possible, baseline and secondary (e.g. Nitrate) the pH was brought down to 2 by data of research documents have been adding acid (HNO3) to the samples to stop collected. chemical reactions. The samples were analyzed For data collection activities, specific data either using in-field measurements and tests or sheets have been designed for field level laboratory analysis. Water sampling locations research as well as for secondary data were chosen based on the spatial variations in collection. the water stream and irrigation system (irrigation sources, the distribution canals, Because only few data or baseline data were reservoirs, the main drain and from available to compare time series data and carry downstream water bodies) in order to obtain a out before and after type of analysis, soil representative sample. In table 3 chemical and physical parameters are listed which had been samples were taken from non irrigated and analysed either in the field or using laboratory irrigated fields but also from cultivated and facilities. uncultivated plots to assess potential impacts The standards and threshold values for water of irrigation on chemical and physical soil quality analysis were chose with respect to parameters. Disturbed soil samples were taken irrigation (FAO 1989). The widely accepted from the predominant soil types at different threshold values for classifying the suitability spots, depending on the accessibility and of water for irrigation are presented in table 4 influence of irrigation on the spots, using the (FAO, 1998). The standards and threshold comparative soil sampling method. From each values for soil chemical and physical were plot samples of different depth were taken, taken from the Ethiopian Ministry of Water depending on the structure of the soil and the Resources (2002). depth of soil horizons. Every soil horizon over 10cm was sampled, at least from 0-30, 30-60 and 60-90 cm depths (FAO 1986). Water samples were collected from the irrigation source and potentially influenced

Table 3: Measured chemical and physical soil and water parameters Samples In-field measurement Laboratory Analysis - chemical & physical parameters Electric Conductivity pH, Electric Conductivity Temperature Cations: Na, Ca, Mg, K Nitrate Anions: SO4, PO4 pH Toxic substances B, Cl Water Nitrogen: NH4-N, NO3 Alkalinity (CO3+HCO3) Trace elements Fe, Mn SAR, adj. RNa Profile Pit Particle size distribution of silt, clay, sand, soil investigation texture pH, Electric Conductivity Soil Nutrients: Na, K, Ca, Mg, P, tot N, Fe, Mn CEC, ESP, BSP Organic Matter, Nitrate

Results and Discussion Wonji/Shoa Sugar Plantation - Water In general the irrigation water used in Quality Anylsis Wonji/Shoa Sugar Plantation is of high quality.

374

Hardly any adverse impacts on the soil quality case it is reused for irrigation. These findings resulting from irrigation activities are to be are confirmed by the study conducted on expected on this scheme. However SAR/adj irrigation water quality by Girma A. (2005). RNa to EC ratio (fig. 3) indicates that a slight Irrigation water (i.e. Awash River water) and to moderate reduction in the infiltration rate factory used water were free of potential might occur. Low salinity water (< 500 µS/cm) hazards of salinity, sodicity and specific ion is corrosive and tends to leach surface soils toxicities. Moreover, drainage water was free of soluble minerals and salts (FAO 29, nearly of similar quality. As pointed out in the 1989). The EC value of Awash River, the previous section this could cause from slight to irrigation water source of Wonji/Shoa, is 293 moderate problems regarding infiltration. The µS/cm. The tested soils show an EC from 151 total salt content of groundwater for some to 475 µS/cm and can be rated as salt free sugarcane fields was relatively high (i.e., EC according to the Eth. Ministry of Water values exceeded 700 µS/cm); a critical limit Resources (2002). The absence of salt in the above which osmotic effects are known to soil indicates an ongoing dispersion process occur. SAR values (fig. 3) were low in all which leads to sealing of the soil and thus groundwater samples. The potentially toxic reduces the soils infiltration rate (FAO, 1989). ions Na+, Cl- and Boron were also found at low During the field study and the analysis in a soil concentrations. pit destruction of the natural structure of the The values for the pH of Wonji/Shoa irrigation soil in some places could be observed. To a system are in the normal/neutral range certain extent this destruction of the soil (pH = 6.5 to 8.5) for all water samples. Only structure might be due to unfavourable SAR the drainage water shows a lower pH of 6.3 values and SAR/adj RNa to EC ratio in the which indicates slightly acidic water. irrigation water of some of the water samples. Generally the pH in all the water samples taken Moreover soil tillage with heavy machines does not seem to be influenced by irrigation. might also have destructive effects on the soil Similar results are given by Girma A. (2005). structure, particularly where waterlogging and The measured pH values of both irrigation high water tables occur. water and Factory used water samples varied Girma A. (2005) states that in most sugarcane from 7.4 to 8.1 (from near neutrality to fields of Wonji/Shoa slight to moderate soil medium basic in reaction). pH values of infiltration problems are expected if drainage water samples varied from 7.0 to 7.8 groundwater occurs at shallow depths. In this (normal range). All groundwater samples regard, the groundwater depth was at all the varied in pH values from 7.1 to 8.3. times too shallow (in the root zone) for nearly The highest values for chloride and boron are all sugarcane fields considered in this study. 3.6 meq/l and 0.45 meq/l respectively. This indicates that in Wonji/Shoa, sugarcane Therefore the threshold values from no fields are prone to waterlogging and physical restriction on use to slight to moderate degradation of topsoil. Drainage water and restriction on use (4 meq/l for chloride, 7 mg/l factory used water are stored in reservoirs and for boron) were not obtained by any of the reused for irrigation, therefore it is of major water samples. importance to monitor the quality of this irrigation source. The drainage water and the reservoir water shows EC values increased values compared to the irrigation water, 357 and 388 µS/cm respectively. With regard to salinity the water is still of very good quality and should not pose any problems on soils in

375

Figure 3: Effect on Infiltration rate due to Sodicity of sampled water sources (Wonji/Shoa Sugar Plantation)

extremely sensitive crops (ESP = 2–10) but not Wonji/Shoa Sugar Plantation - Soil Quality for sugar cane. Analysis The content of total Nitrogen can be rated as high for all soils (ranging from 0.071 to As mentioned the tested soils show an EC from 0.114%). The rating for Potassium is high and 151 to 475 µS/cm and can be rated as salt free very high for all soil samples ranging from according to the Eth. Ministry of Water 1.08 to 2.63 for irrigated land and from 3.68 to Resources (2002). 4.05 for the virgin soil. The comparison of The pH of all soil samples ranges from 7.4 to irrigated soils and non-irrigated virgin soil 8.4 indicating moderately alkaline soils which shows that the content of Potassium is much is typical for soils of these climates (dry, arid higher in the virgin soil (4.05 meq/100g). The climate most of the year). This indicates low distribution of some parameters over the soil availability of phosphorus and other micro- depth does not seem to be natural anymore for nutrients (MoWR, 2002). the cultivated soil. In general there are hardly With a CEC from 46.2 to 58.4 meq/100g the any obvious differences between irrigated soils soil samples are all in the range of very high and the virgin reference soil as far as the CEC, according to the ratings of the Ethiopian chemical properties are concerned. The reason Ministry of Water Resources (2002). This might be that the irrigation water has a very indicates good agricultural soil. The BSP, good quality and therefore no severe impacts ranging from 62.6 to 96.65% can be rated as on the soil are to be expected. A possible high for almost all soil samples which decline or even lack in nutrients is indicates a high fertility of the soil. The ESP of compensated through fertilizer application or the soils lies below 6% (0.99 to 5.88%) which other soil fertility management practices. The is definitely below the threshold of 15% and virgin soil might also be influenced by therefore the soils can be rated as non sodic. irrigation through capillary rise of the Decreases in yield would only be expected for groundwater or surface runoff of the irrigated

376

fields and therefore not be perfectly virgin the root zone (in some places up to 10 days anymore. after irrigation). These drainage problems have even become one of the major factors in determining the composition of cane variety The content of Calcium is high and very high (Tariku, 2001). In fields located near in all soil samples, ranging from 18.32 to 40.72 reservoirs, irrigation canals and drains meq/100g soil. The same applies for suppressed cane growth due to seepage and Magnesium ranging from 4.94 to 12.36 ground water table rise could be observed. meq/100g soil. Kediru (1997) stated that cane loss due to this The content of organic carbon of all soil problem is economically significant. ranges from 0.76 to 1.88 % and can be rated as In 1998/99 a study was started in Wonji/Shoa very low in all the soil samples. This indicates Sugar Plantation to measure the groundwater that without application of fertilizers no level (GWL) fluctuation with piezometers adequate yields would be achieved. installed in different plantation fields. Piezometers were installed at different Wonji/Shoa Sugar Plantation - locations near reservoirs, distribution canals Groundwater Hydrology and drains. One piezometer was installed in a non-irrigated area as reference (WRS). In Wonji problems due to irrigation are In figure 4 groundwater tables of 4 different reported with regard to waterlogging. The fields (G1, G2, G3 and G4) and the reference ground water level within the sugar estate field (WRS) are displayed for 3 years (1999- shows a rising tendency (observation 1999- 2001). Measurements are given in depth in cm 2001, fig. 4). However, not only within the from soil surface. The measurement of sugar cane plantation but in the whole region a groundwater table of the reference area (WRS) rise of the water table has been reported shows the maximum water table depth values (Teshome, 1999). This might be due to in all years and months. irrigation (seepage losses out of reservoirs and The results indicate that fields which are channels, over watering, etc) but also due to located near reservoirs are affected by rising seepage losses in a great extent from the Lake ground water level more than fields nearby Koka reservoir. The water of the Awash River distribution canals and fields nearby drains. is retained by Koka dam. The dam was built to assure a constant discharge in the Awash River Even this short measurement time series of for two hydropower stations, but it also 1999 to 2001 shows significant increasing guarantees sufficient irrigation water supply. trends of rising groundwater table. Fields near What has been reported before and confirmed reservoirs (G1) are most affected. Habib in the field work are problems with a high (2002) concludes that seepage from reservoirs water table and waterlogged fields. Especially and unlined irrigation canals is the reason for on fields with heavy clayey soils problems rising groundwater level. Near drains (G3) with waterlogging have been discovered groundwater levels are significantly lower during the field study and the analysis in a soil compared to the measurements near the pit (Wallner, 2006). However, the lighter soils reservoirs. This result shows that drains if – light in relation to the heavy soils of the effectively utilized can lower the ground water sugar estate - seem to be of very good quality table. In the reference area GWL was rising the and not prone to negative impacts of irrigation first year but decreasing the second year. under the given conditions. Furthermore the GWL is generally at least In 2002 a study on groundwater level 50cm deeper compared to the GWL of the fluctuation at Wonji/Shoa sugarcane plantation irrigated fields. was conducted by Habib D. Investigations Three years of GWL measurement reveal that showed that on some fields the groundwater fields near reservoirs are highly affected by table is less than one meter deep (ARS Annual waterlogging followed by fields near unlined Report, 1994). The same report indicated that distribution canals, fields away from water in some fields downward percolation of bodies and fields near by drains. The water irrigation water below the root zone, especially level is increasing from year to year. in soils classified as heavy black soils is so Moreover, with the exception of fields near slow that it caused temporary storage within drains the GWL of other fields is less than 60

377

cm for more than 6 months per year which has a negative effect on sugar production. Figure 4: Trends in the ground water level at Wonji/Shoa sugar plantation over three years (Habib, 2002)

Trends GW-Fluctuation

9 0 1 9 99 0 00 1 0 01 99 00 y 01 ar ept 0 ay ept Jan 9 M May 99July Sept 9 Nov 99 Jan 0 Mar Ma July 00S Nov 00 Jan 0 Mar M July 01S Nov 01 0

50 WRS Mean G1 100 Mean G2 Mean G3 Mean G4 150 Linear (WRS)

Depth (cm) Depth Linear (Mean G3) 200 Linear (Mean G4) Linear (Mean G2)

250 Linear (Mean G1)

300

improvement or at least maintenance of the waste water treatment plant is required.

The values of the EC measurement in the field are all below the threshold value of non saline Finchaa Valley Sugar Estate - Water to slightly and moderate saline of 700µS/cm Quality Analysis given by FAO (1989). This indicates no problems with salinity caused by the available Water quality, regarding the physical as well as water source. A detailed analysis of the EC chemical parameters, of Finchaa Valley Sugar Values of different water bodies in the adjacent Estate can be rated as very good for irrigation area of the estate provides a deeper insight in purposes. However, the increase of the EC the hydrological system of the irrigation value of Finchaa River from 96 µS/cm scheme (fig. 5). The EC value of Lake Chomen measured in the upstream area of the scheme to (Finchaa Dam) is 96 µS/cm. The groundwater 121 µS/cm in the downstream area shows the samples show slightly increased EC values of impact of irrigation on the downstream water 352 and 477 µS/cm due to accumulation of bodies. This value does not include the indirect salts in the aquifer. Fig. 5 shows the EC values impact of irrigation by releasing pollutants by of different water bodies within the irrigation the sugar factory because the outlet of the scheme from upstream to downstream area of factory was closed while EC measurements Finchaa Sugar Estate. As mentioned before the were taken. The observation made during the EC value of the reservoir is 96µS/cm and field visit indicates that the obvious 88µS/cm for Korke River one of the tributaries malfunction of the waste water treatment plant of Finchaa River in the far upstream area of the of the factory poses a threat to downstream scheme. Finchaa River drains Lake Chomen to water bodies especially to Finchaa River and the Blue Nile Basin and therefore has the same its ecosystem. Further investigations are EC values. The assumption can be made that necessary to investigate into the potential the natural EC value of surface water bodies in impact of the factory and the scheme on Finchaa Valley Area is below 100 µS/cm. The downstream users. Additionally an EC values of the drainage water (164µS/cm) and of tributaries of Finchaa River, e.g.

378

Agemsa River (286µS/cm) which are located drainage system of the sugar farm. The further downstream of the estate are much increasing EC value of Finchaa River makes higher compared to the values in the upstream the impact of irrigation on downstream water area. This fact together with the increased EC bodies clearly visible. Drainage waters as well value of Finchaa River (121µS/cm) measured as surface run-off water possibly mixed with at a location downstream of the irrigated fields agrochemicals affect the water quality of the indicates that the tributaries of Finchaa River natural water resources in the adjacent area of crossing the irrigated area serve as the natural the scheme.

Figure 5: EC values of different water bodies from upstream to downstream areas at Finchaa Valley Sugar Estate

The ratio of adj RNa to EC (fig. 6) of the irrigation water source indicates a risk of severe reduction of the infiltration rate over time. This imposed sodicity has the potential to destroy the soil structure and lead to soil crusting of the affected soil especially when EC value is low (96 µS/cm). The dispersive effects adversely influence the physical properties, e.g. infiltration rate, aggregate stability of the soil (FAO, 1989). The extent of the impact highly depends on the texture and clay-sized particles of the soil.

Since the filling of smaller pore space, sealing the soil surface and therefore reduces infiltration rates, by dispersion of finer soil particles is more likely to happen in soils with high clay content impact on Luvisols which can be classified as sandy loam to sandy clay loam with sand content over 60 % might prove to be less severe compared to the impact on Vertisols with clay content up to 60 %.

379

Figure 6: Effect on Infiltration rate due to Sodicity of sampled water sources (Finchaa Valley Sugar Estate)

As salinity of both applied water and the soil reason could be found in mechanized soil solution is relatively low no further swelling tillage. The silt content of the primary data and dispersion of clay minerals are to be sample is twice as low as it is for the secondary expected caused by the water source. data sample. This ongoing process might lead to accumulation of clay particles in the topsoil The threshold values for chloride from no layers and therefore aggravate the gradual restriction on use to slight to moderate degradation of the soil structure by induced restriction on use of 4meq/l are not obtained by sodicity. any of the water samples. The SAR rating for The EC values of all analysed soils range from slight to moderate restriction on use ranging 10 to 60 µS/cm which is far below the from 3 to 9 neither is obtained using the SAR threshold value of 4 dS/m (= 4000 µS/cm). values nor the values for adj RNa. Therefore no Therefore they are not affected by salinisation. toxic effects on plants due to irrigation are to According to the Eth. Ministry of Water be expected. Resources (2002) the samples can be rated as non saline. The EC value of the reference soil Finchaa Valley Sugar Estate - Soil Quality of the upper 30 cm is double of the value of the Analysis irrigated plots. This could be an indication for an ongoing leaching process caused by the The comparison of the collected primary and application of water. secondary soil texture data for Luvisols show Secondary EC values of all analysed soil differences related to the proportion of sand, types are far below the threshold value of 4 silt and clay content. The primary data show a dS/m (= 4000 µS/cm). The EC value of during lower sand content for the upper soil and a the state farm cultivated Vertisols in the East higher sand content for the lower soil part and Bank area is significantly higher than the vice versa for the clay content which indicates uncultivated Vertisols of the West Bank. This a shifting and relocation of soil particles might indicate an accumulation of salts in the especially of the sand fraction. A possible rain fed cultivated soil with insufficient leaching by precipitation.

380

In general pH values of sampled irrigated effective. This makes it even more important to Luvisols are very acidic to moderately acidic plan agrochemical management for each soil with values ranging from 4.89 to 5.62 from type differently. The reference soil however deeper to shallower soil depths respectively. shows a contrariwise tendency with a CEC The pH value of non-irrigated reference soils value of 24.87 meq/100g in the upper 30 cm of ranges from 4.56 to 4.86. Significant changes soil depth and 15.52 meq/100g below 60 cm. in soil pH values comparing secondary and The primary data of the year 2006 confirm the primary data for Luvisols could not be findings of the feasibility study conducted in observed. All measured values show slightly the pre-design phase of the irrigation scheme. acidic tendencies, which is common for soils Proper management of agrochemicals is developed under these climatic conditions. indispensable to avoid contamination of Secondary data show that the pH of Vertisols Finchaa valley aquifers. in Finchaa Valley is higher compared to The BSP of the two sampling spots ranging Luvisols. The measurements however show from 55.0 to 74.0 % and from 51.0 to 67.0 % lower values for the uncultivated Vertisols (pH respectively can be rated as medium to high. 6.13 - 6.5) than for the during the state farm This fact indicates a medium to high fertility of period cultivated Vertisols (pH 6.9 - 7.37) with the soil. The reference soils BSP values are increasing trend for increasing soil depths. between 37.0 to 56.0 %. Therefore the soil can This difference between secondary and be rated as medium fertile with lower values primary data might be an indicator for ongoing compared to the other tested soils and decline of soil pH due to irrigation and increasing values with depth. agrochemical use within the sugar estate. None of the measured ESP values of the tested Washed-out-soils tend to acidification. Water soils exceeds the threshold value of 15 % and dominated soils (soils of humid regions) have can therefore be rated as none sodic soil (FAO, low values of pH, because their content of 1988). organic and carbonic acids is often subject to With regard to CEC and ESP only secondary replenishing and recharge by rainfall. Under data of Finchaa West Bank area were these conditions, the acids attack minerals, available. From the CEC values the difference producing more acidity (Mirsal, 2004). in soil fertility of Luvisols and Vertisols can be Therefore washed-out-soils tend to clearly seen (fig. 7). With a CEC value higher acidification. According to the rating of pH than 52 meq/100g the Vertisols are in the range values given by the Ethiopian Ministry of of very high CEC according to the ratings of Water Resources Al and Mn will be toxic if the Ethiopian Ministry of Water Resources present in very acidic soils. Ca, Mg and Mb (2002). The data for Luvisols however show may be deficient and the availability of medium to high CEC values ranging between phosphorus is low in the presence of free Al 21.55 and 28.99 meq/100g. The areas where and Fe. Nitrification of organic matter is taking the field work in 2006 was conducted show place. In moderately acidic soils P, Ca, Mg and CEC values ranging from approximately 25.0 Mb may be deficient. Fertilizers (ammonium meq/100g for the Luvisols cultivated and sulphate and triple super phosphate) which irrigated since the establishment of the Finchaa may increase the acidity should be avoided Sugar Estate, which is comparable to values of (MoWR, 2002). the uncultivated Luvisols, and below 15.0 The CEC values of the tested Luvisols (14.02 meq/100g for the Luvisols which have been to 14.56 meq/100g) indicate low to medium cultivated since 1975 and irrigated since 1998. response to fertilizer application which is This makes the long-term process of soil typical for soils with low clay content. Low degradation visible induced by agriculture and CEC values can be caused by losses resulting especially by irrigation. Minor to major from leaching-out, especially sodium (Na) is amendments might be required for the highly exposed to this process (MoWR, 2002). Luvisols, however the soil might show only The CEC of Vertisols (23.95 and 26.54 moderate to poor response to fertilizer meq/100g) however shows a different picture. application (MoWR, 2002). The CEC shows an increasing trend with None of the measured ESP values of the tested increasing soil depths. As the clay content of soils neither for Luvisols nor for Vertisols this soil type is significantly higher, the exceeds the threshold value of 15 % and can response to fertilizer application is more

381

therefore be rated as none-sodic soil (FAO The content of Calcium is medium (5.18 to 9.5 1988). meq/100g soil) for the irrigated soil and low The BSP ranging from 36.36 to 50.16 % (Lc (3.36 to 4.99 meq/100g soil) for the non West Bank), 40.67 to 57.38 % (Lc East Bank) irrigated plot. and from 42.98 to 52.55 % (Ve West Bank) The tested soils show a high content of can be rated as medium. Only the BSP from Magnesium. the Vertisols which can be found on the East The results show very clearly the naturally low Bank can be classified as high with values content of available phosphorus in the ranging from 59.47 to 67.02 %. This fact reference soil (fig. 8) which makes the indicates a medium fertility of the Luvisols in application of agrochemicals necessary in general and the Vertisols of the West Bank. order to achieve high crop yields. It is stated The Vertisols of the East Bank are however that in moderately acidic soils P may be characterised by high soil fertility related to the deficient which is true in this case, measured BSP values. nevertheless fertilizers (ammonium sulphate The content of total Nitrogen of the tested and triple super phosphate) which may soils can be rated as high for all tested soils increase the acidity should be avoided (ranging from 0.05 to 0.14%). The total (MoWR, 2002). The content of available P nitrogen content of the uncultivated reference (29.2 mg/kg) is only high within the first 30 soil is significantly lower compared to the cm of soil depth for sampling point F 1. irrigated plots. According to the statement of a staff member The rating for Potassium is low for all samples of the Sugar Estate DAP had been applied to ranging from 0.2 to 0.27 meq/100g soil except plot F 1 shortly before the field work was the value of the topsoil of the non-irrigated plot conducted. Obviously phosphorus is not which shows a medium content of Potassium leached to lower soil depth as its content in 30- (0.48 meq/100g soil). 90 cm is significantly lower.

Figure 7: Cation Exchange Capacity of soil samples of Finchaa Valley Sugar Estate – Secondary Data

(West and East Bank) is similar compared to The secondary data show that the content of the content of the irrigated soils. total Nitrogen of the Luvisols and Vertisols of The rating for Potassium of the Luvisols is West and East Bank can be rated as high for all medium (West Bank and East Bank; lower soil soils (ranging from 0.05 to 0.27%). The parts) to high (East Bank; upper soil parts) content of total nitrogen in the cultivated whereas the rating for K of irrigated Luvisols Vertisols is significantly higher (0.27%) is low. The tested Vertisols in both areas show compared to the other tested soils. The content higher contents of K and can be rated as high. of total nitrogen in the uncultivated Luvisols The formerly cultivated Vertisols of the East Bank have a significantly higher K content

382

compared to the uncultivated Vertisols of the and secondary data show similar Mg contents West Bank. for irrigated and non-irrigated Luvisols. The content of Calcium is medium for the The analysis of organic content and tested Luvisols and high to very high for the comparison of primary and secondary data Vertisols in both areas. The Ca values of the reveals an ongoing soil degradation process formerly cultivated Luvisols of the East bank caused by intensive sugar cane monoculture. are higher compared to the Luvisols of the The comparison of the organic carbon content West Bank which were uncultivated at that of uncultivated, cultivated and irrigated areas point of time. The Ca content of the tested shows a decreasing trend of naturally low Vertisols of the West and the East Bank are organic carbon content of the tested soil types. more or less similar ranging from 18.3 to 21.8 Two reasons can be spotted as potential causes meq/100g but are higher compared to the for this trend. Soil degradation which is caused highest Ca content of tested Luvisols (10.15 by intensive agricultural production of sugar meq/100g; Luvisols, East Bank; topsoil). cane monoculture might be one reason. The The Magnesium content ranges from 2.32 to low content of organic carbon can also be 4.36 meq/100g for Luvisols of East and West explained by the low natural pH value of the Bank and East Bank Vertisols. The Mg value soil. The nitrification process of organic for Vertisols of the East Bank is slightly higher carbon which is typical for these soils and with a content of 6.12 meq/100g in the topsoil. climatic conditions could be the direct cause The content of Mg in all soils can be rated as for the very low organic carbon content of the high, except the content in the topsoil of the tested soils. However, this makes fertilizers West Bank Vertisols which can be rated as application necessary to achieve adequate medium. The analysis results of the primary yields which additionally spells the risk of groundwater pollution by agrochemicals.

Figure 8: Content of available Phosphorus of soil samples of Finchaa Valley Sugar Estate Available P

25 high 20

15 mg/kg Soil mg/kg 10 medium 5 low 0 0-30cm 0-30cm 0-30cm 30-60cm 60-90cm 30-60cm 60-90cm 30-60cm 60-90cm

F 1 F 2 F 3, virgin sampling spot/Depth (cm)

Reference soil (non irrigated nor cultivated) cultivated and irrigated soil whole region and the Sugar Estate in Finchaa Valley Sugar Estate – Erosion particular. Deforestation of the highland regions caused by cutting trees for fire wood Two major types of soil erosion could be exposes the highland soils to erosive processes identified for the Finchaa Valley cases study. and contributes to high surface runoff. The effect of hinterland erosion and infield Population pressure aggravates this problem erosion jeopardizes the fertility status of the even more. The eroded material from the steep escarpments which is deposited on the irrigated areas of the valley bottom can lead to soil

383

degradation. Run off estimation of the valley scheme and the beginning of sugar cane bottom indicates loss of eroded fertile soils due production also sugar cane burning for to sheet erosion estimated in the range of 5 to harvesting can be identified as a major cause 10 mm/y or equivalent to approximately for forest fires. Furthermore migration 100ton/ha-y (Girma T., 1995). The eroded soil tendencies triggered by the attraction and is deposited to the tributaries of Finchaa River employment options of the sugar estate which finally drain to Abay (Blue Nile) River. increase the pressure on the ecosystem. Considering the high content of available In developing countries like Ethiopia the GDP phosphorous which tends to be fixed to soil of a country highly depends on agricultural particles in the topsoil layer, erosion of these production. Priorities have to be outweighed soil particles causes an additional threat to between conservation of valuable ecosystems downstream water courses. Assuming that and important contribution to a country’s these present trends continue the problem of economy. In order to justify clearance of large soil and land degradation induced by soil natural forest areas agricultural production erosion may threaten the sustainability of the needs to be sustainable to avoid large scale irrigation project. land degradation and further adverse environmental impacts. Finchaa Valley Sugar Estate - Impact on Ecosystem Conclusions

In 1975 the state farm was established which The main goal of this report was to assess the mainly produced food and commercial crops pertinent environmental implications of two until 1991. About 3500 ha of valuable selected large scale irrigation schemes in ecosystems have been cleared and destroyed Ethiopia. For this purpose two large scale for agricultural activities. Before that the schemes in different regions were selected valley was under natural vegetation cover and Wonji/Shoa and Finchaa Valley. The main very few agricultural plots could be observed. focus was put on impact on natural resources The valley area was a sanctuary for wild like water, soil and ecosystems. Based on the animals. Tall savanna grass mixed with trees pertinent environmental checklists of FAOs, which occupied most of the valley floor World Banks and MoWRs EIA procedures, created favorable conditions for a large data sheets and guidelines for primary and numbers of wild animals like carnivores, secondary data collection were designed browsers, grazers and other small animals. The (Wallner, 2006) and used during the natural vegetation in the area was dense and assessment. The water sources of the schemes with full canopy during the wet season. were sampled and analyzed against FAOs standards and threshold for water used in Starting from 1991 up to now more than 8064 agriculture. Additionally water samples of ha of land has been cleared and irrigated for downstream water bodies were tested to assess sugar production (Ahmed Amdihun, 2006). potential adverse impacts. Physical and From 1975 to 1991 these parts of Finchaa chemical soil parameters of primary and Valley area have changed from primary to secondary data were analyzed for possible tertiary economic production; from traditional changes. agricultural to industrial and commercial In general, the study conducted shows that the production respectively. The dominant land irrigation water used in the investigated case use classes are irrigation agriculture and agro- studies is of good quality with regard to FAOs pastoral within the valley area and rain fed standards for water used in agriculture and agriculture mainly in the high land area. In does not spell any risk for irrigation purposes. addition to the land clearing tree and grass The irrigation water sources used in the three species are exposed to extensive and severe case study sites have low EC values ranging bush fires (Ahmed, 2006). Two major reasons from 96 µS/cm (Finchaa Valley) to 293 µS/cm could be identified. In addition to natural (Wonji/Shoa) and therefore no primary factors inhabitants of the region earn their salinisation is to be expected. subsistence by collecting wild honey and crop With regard to impact of the used water source cultivation. To clear their land farmers burn the on the soil quality EC to SAR ratio however forest. Since the establishment of the irrigation indicates potential negative long-term effects

384

on infiltration rates due to damage of the soil sustainability is neglected. Due to the fact that structure and soil crusting through induced environmental sustainability of irrigation sodicity especially in case of Wonji/Shoa and projects is rather on the low end of the policy Finchaa Sugar Plantation. priority list, adverse and irreversible The most crucial environmental impacts of environmental impacts are bound to happen in large scale irrigation in Ethiopia which could the contrary nexus of profit maximisation, be identified are related to improper irrigation short-term benefits and environmental management and development of irrigation sustainability. project on saline and saline-sodic soils. In order to avoid possible negative impacts of Inefficient application of water and seepage of the expansion of irrigated agriculture in water from reservoirs and unlined distribution causing deterioration of land and soil quality, canals lead to rising of groundwater table at proper understanding of the quality of soil and Wonji/Shoa. Investigations showed that on irrigation water and implementation of some fields of the Wonji/Shoa Plantation the appropriate measures have paramount groundwater table is less than one meter below importance for sustainable development. There soil surface (Habib D., 2002). This tendency is no doubt that irrigation can increase has mainly two adverse effects. The rise of the intensification and productivity, can help to groundwater table up to the root zone limit the size of cultivated areas, can provide interferes with the proper development of the ample labour and agro-industrialization planted crop, leads to damage of the soil opportunities and other potentially positive structure and insufficient soil ventilation. benefits. On the other hand it can also cause Secondly it induces secondary salinisation due negative impacts such as deterioration of soil to capillary rising. Improper irrigation and water quality, impact on eco-system, management and attempts to leach the health and other negative externalities. It is accumulated salt by additional application of important to support such endeavours through water, which leads to rising groundwater table, proper study and continuous research, so that has even aggravated this effect (Tena, 2002). the positive roles could be enhanced with Besides installation of drainage systems to timely mitigation measures for the negative intercept deep percolation of the excess water, impacts. other in the long run more cost-effective measures need to be considered. Installation of drip systems could avoid excessively use of irrigation water and make the application of water more efficient and therefore increase the overall water productivity of the schemes. According to the Ethiopian Irrigation Development Program 26 medium and large- scale irrigation projects are planned to be implemented. Due to topographic reasons most of these already established or proposed large- scale irrigation schemes can be found in the lowlands of Ethiopians major river basins such as Awash, Blue Nile and Wabe Shebelle river basin. Over 11 million ha of land in the arid, semi-arid and desert parts of Ethiopia are known to be salt affected. Large areas of the Awash River Basin especially the middle and lower parts of the basin are saline or sodic or in saline or sodic phase and thus potentially exposed to salinisation and sodicity (EIAR 2006). Development of large scale irrigation, political decision making and investment strategies are often oriented on short-term profit maximisation whereby environmental

385

References EIAR (2006): Assessment of Salt Affected Soils in Ethiopia and Recommendations Abrol I.P., Yadav J.S.P., Massoud P.I. an Management Options for their (1988): Salt Affected Soils and their Sustainable Utilization, Taskforce Management, FAO Soils Bulletin No 39. Report, Addis Ababa, Ethiopia. FAO, Rome. Ethiopian National Meteorological Service Abdel-Deyem S. (1998): Waterlogging and Agency (2006): Meteorological Data. salinity. In: Biswas A.K. (ed), Water Ethiopian Ministry of Water Resources Resources: Environmental Planning, (2002): Study Guideline on Soil Survey Management and Development, Tata and Land Evaluation, Part B. Addis McGraw-Hill Publishing Company Ababa, Ethiopia. Limited. New Delhi, India. FAO (1965): Report on the Awash River Ahmed Amdihun (2006): Geographic Basin, Imperial Government of Ethiopia, Information Systems and Remote United Nation Special Fund FAO/SF: Sensing Integrated Environmental 10/ETH, 1995 Vol. IV, Addis Ababa, Impact Assessment of Irrigation in Ethiopia. Fincha Valley, MSc. Thesis, Addis FAO (N.d.): Key to the FAO Soil Units in Ababa University, Ethiopia. the FAO/Unesco Soil Map of the World Agricultural Services (1971): Annual Report http://www.fao.org/AG/agl/agll/key2soil. of 1970/71, Wonji, Ethiopian Sugar stm (accessed on 10/01/2007). Industry Support Center Sh. Co., Wonji. FAO Soils Bulletin No 42 (1986): Soil Agricultural Services (1974): Annual Report Survey Investigations for Irrigation, of 1973/74, Wonji, Ethiopian Sugar Rome. Industry Support Center Sh. Co., Wonji. FAO Soils Bulletin No 39 (1988): Salt Ambachew D., Girma A. (2000): Review of affected soils and their management, sugarcane research in Ethiopia: I. Soils, Rome. irrigation and mechanization (1964- FAO Irrigation and Drainage Paper No 29 1998), Research and Training Services (1989): Water Quality for Agriculture, Department, Ethiopian Sugar Industry Rome. Support Center Sh. Co., Wonji. FAO Irrigation and Drainage Paper No 48 Anonymous (1987): Annual Problem Fields (1992): The use of saline waters for crop Report of 1986/87 Campaign, production, Rome. Wonji/Shoa Sugar Estate. FAO Irrigation and Drainage Papers No 53 ARS Annual Report (1994): Wonji/Shoa (1995): Environmental impact Sugar Estate, Agricultural Research and assessment of irrigation and drainage Service Center, Wonji. projects, Rome. Awulachew, S.B. (2006): Improved Government of the Republic of Ethiopia Agricultural Water Management: (2001a): Water Sector Strategy, Ministry Assessment of of constraints and opportunities for agricultural Water Resources, Water Sector Development development in Ethiopia, in Awulachew, Program, 2002-2016, Irrigation S.B., Menker, M., Abesha, D., Atnafe, Development Program. T., and Wondimkun, Y. (Eds.) (2006): Government of the Republic of Ethiopia Best practices and technologies for small (2006): National Accounts Statistics of scale agricultural water management in Ethiopia, National Economic Accounts Ethiopia, Proceedings of a Department, Ministry of Finance and MoARD/MoWR/USAID/IWMI Economic Development, November, symposium and exhibition held at Ghion 2006. Hotel, Addis Ababa, Ethiopia 7-9 March, Girma A. (2005): Evaluation of Irrigation 2006, Colombo, Sri Lanka: International Water Quality in Ethiopian Sugar Estates Water Management Institute. 190pp. (Research Report), Research and Encyclopaedia Britannica: Training Services Division, Ethiopian http://www.britannica.com/eb/article- Sugar Industry Support Centre Sh. Co., 37681/Ethiopia (accessed 23/05/2007). Wonji.

386

Girma T. (1995): Finchaa Sugar Project Soil Szabolcs, I. (1989): Salt affected soils, CRS Survey/West Bank, Land Management Press. Inc., Boca Ratan, Florida. 274p. Unit Final Report Vol. I, Addis Ababa, Sahlemedhin Sertsu, Taye Bekele (2000): Ethiopia. Procedures for Soil and Plant Analysis, Girma T. (2001): Land Degradation: A National Soil Research Center, Ethiopian Challenge to Ethiopia, Environmental Agricultural Research Organization, Management Vol. 27, No. 6, pp. 815- Addis Abeba.e 824, Springer Verlag New York Inc. Tadele G.S. (1993): Degradation Problems Girma Teferi (N.d.): Infield Irrigation of Irrigated Agriculture, In Tekalign M. Assessment in the Finchaa Valley, Study and Mitiku H. (ed.) ESSS Proceeding report, Finchaa Sugar Factory Project, Soil the Resource Base for Survival, Finchaa Valley Sugar Estate Library. Addis Ababa, Ethiopia, pp. 200-203. Habib D. (2002): Groundwater level Tena A. (2002): Spatial and Temporal fluctuation study at Wonji/Shoa Variability of Awash River Water sugarcane plantation (Preliminary Salinity and the Contribution of Research Report), Research and Training Irrigation Water Management in the Services Division, Ethiopian Sugar Development of Soil Salinization Industry Support Centre Sh. Co., Wonji. Problem in the Awash Valley of Haider G. (1988): Irrigation Water Ethiopia, PhD. Thesis, Institute of Management Manual, Institute of Hydraulics and Rural Water Agricultural Research, Addis Ababa Management, Department of Water, Ethiopia Atmosphere and Environment, Heluf Gebrekidan (1985): Investigation on University of Natural Resources Applied Salt Affected Soils and Irrigation Water Life Sciences, BOKU Vienna. Quality in Melka Sedi-Amibara Plain, Teshome D. (1999): Water Balance and Rift Valley Zone of Ethiopia, MSc Effect of Irrigated Agriculture on Thesis, School of Graduate Studies, Groundwater Quality in the Wonji Area / Addis Ababa University. Addis Ababa, Ethiopian Rift Valley, MSc. Thesis, Ethiopia. 131p. Addis Ababa University. Kediru K. (1997): Annual Problem Fields Tilahun H. and Paulos D. (2004): Results to Report of 1996/97, Agriculture Division, date and future plan of research on Wonji/Shoa Sugar Estate, Wonji. irrigation and its impact, Workshop on Mirsal I.A. (2004): Soil Pollution, Origin, impact of irrigation on poverty and Monitoring & Remediation, Springer- environment, Workshop proceedings, (In Verlag, Berlin Heidelberg. press). McCartney, M. (2007): Decision support UNDP (2006): Human Development Report systems for dam planning and operation, 2006. Beyond scarcity: Power, poverty Draft IWMI Working Paper. and the global water crisis, New York. MOFED: PASDEP (2006). Ethiopia: USDA (2006): Keys to Soil Taxonomy, Building on Progress. A Plan for Tenth Edition, United States Department Accelerated and Sustained Development of Agriculture, Natural Resources to End Poverty (PASDEP), Volume I: Conservation Service, California. Main Text. Wallner K. (2006): Field Parameter Rahmeto A. (1998): Problems Related to Evaluation to Support Environmental Irrigation and Drainage in Wonji/Shoa Impact Analysis of Irrigation in Ethiopia, Sugar Cane Plantation, A Preliminary MSc. Thesis, Institute of Hydraulics and Survey, Wonji. Rural Water Management, Department Richards L.A. et al. (1954): Diagnosis and of Water, Atmosphere and Environment, Improvement of saline and alkali soils, University of Natural Resources Applied Agriculture Handbook 60, US Salinity Life Sciences, BOKU Vienna. Laboratory, US Department of Werfring A., Lempérière P. and Boelee E. Agriculture, Riverside, California. (2004): Typology of irrigation in Szabolcs, I. (1979): Review of research on Ethiopia, In: Proceedings of inception salt affected soils, Natural Resources workshop on IWMI–BOKU– Research XV, UNESCO, Paris. Seibersdorf–EARO–Arbaminch University collaborative study on the 387

impact of irrigation development on poverty and the environment, 26–30 April 2004, Addis Ababa, Ethiopia, (In press). Werfring A. (2005): Typology of Irrigation in Ethiopia, MSc. Thesis, Institute of Hydraulics and Rural Water Management, Department of Water, Atmosphere and Environment, University of Natural Resources Applied Life Sciences, BOKU Vienna. World Bank (2004): Ethiopia’s Path to Survival and Development: Investing in Water Infrastructure, Concept Paper (Background Note for FY04 CEM), Yeshimebet T. (2005): Summary of Weather Data for the Year 2004 and Average of the Last Ten Years 1995-2004, Ethiopian Sugar Industry Support Center Sh.Co., Wonji Research Station. Zeleke K. (2003): GIS and Remote Sensing in Land Use / Land Cover Change Detection in Finchaa Valley Area, MSc. Thesis, Addis Ababa University, Ethiopia.

388

Planning and management of irrigation in sub-Saharan Africa: reducing the environmental and health costs

Matthew McCartney1, Eline Boelee1, Olufunke Cofie2 and Clifford Mutero3 1IWMI-Addis Ababa, 2IWMI-Acra, 3Independant consultant [email protected]

impact assessment relevant to community Abstract led, NGO and small private projects.

Development of irrigation can result in 1. Introduction negative environmental and human health impacts. Irrigation undertaken without full It is widely acknowledged that irrigation can consideration of these impacts can have play a major role in improving food serious adverse repercussions, not only productivity, reducing poverty and undermining the investment but also sustaining rural livelihoods (Hussain and worsening poverty and contributing Hanjra, 2004; Smith, 2004). However, over considerably to peoples’ suffering. The the past two decades, investments in impacts are strongly inter-linked because it irrigation in sub-Saharan Africa have is changes in the environment that cause declined significantly (Kikuchi et al. 2005). changes in health. Furthermore, mitigation There are a number of reasons for the measures that reduce environmental damage decline, but the poor performance of often improve health outcomes. In the past, irrigation, especially with respect to capital- research into impacts, and the development intensive schemes, has undoubtedly of impact assessment methodologies, has contributed. Although not the sole reason, focused primarily on large scale, capital environmental factors and adverse health intensive, schemes. However, small scale impacts have been a prominent cause for the and less formal water management disappointing performance of many schemes interventions, which are increasingly (Oomen et al., 1990). Inadequate prevalent in sub-Saharan Africa, can also consideration of environmental and health have significant environmental and health issues in the planning and implementation of impacts. This paper summarizes the findings projects is widely perceived as a key cause of a study of environmental and health of project failure (Moradet et al., 2005). issues associated with all scales of irrigation in sub-Saharan Africa. It is not a The environmental and human health compendium of data, but rather provides an aspects of irrigation schemes need to be overview and framework for understanding considered in tandem, because they are policy and programming issues. It is strongly inter-linked. It is changes in the recommended that a pragmatic approach to environment, in conjunction with associated address current environmental and health socio-economic change, which results in planning includes three levels of changes in the health of local populations implementation: i) strategic planning at the (Figure 1). Environmental and health impacts national and regional level; ii) full of irrigation are generally site specific and are environmental and health assessment for multiple, varied and complex. They depend government and donor funded projects; and on a range of factors, including the scale of iii) development of simplified tools for development, bio-physical conditions, management and operation, as well as the extent to which safeguards are implemented.

389

Although investment in irrigation is supposed The potential negative environmental to bring health benefits, through improved impacts of large capital-intensive irrigation nutrition and income, it can have adverse schemes are extensively documented (e.g., impacts through the extension of water- Adams 1992; Dougherty and Hall 1995; related vector-borne diseases (e.g., malaria, Kay, 1999). Modification of river flow schistosomiasis, liver flukes, filariasis, regimes, depletion of groundwater, onchocerciasis, dengue fever, yellow fever, sedimentation effects, soil salinization, Rift Valley fever and encephalitis (Oomen et waterlogging, water contamination and al. 1990). For example, malaria has been biological effects24, have all been closely associated with the Gezira Irrigation responsible for undermining the System, in Sudan since it began in 1924. sustainability of schemes. Often farmers on Severe outbreaks in the mid-1970s were irrigation schemes are fully aware of many linked to changes in irrigation management environmental problems. However, because practices and the onset of pesticide resistance small incremental changes can take a long in malaria mosquitoes (Hunter et al. 1993). time to have a significant impact on productivity, often nothing is done until it is Contrary to widespread belief, community- too late. It is estimated that in the southern based and small-scale agricultural region of Ethiopia, approximately 50% of interventions also have environmental and irrigation scheme failures and below health impacts (Konradsen et al. 2000; capacity performance are due in part to Mutero et al. 2004). These impacts are often technical (as opposed to institutional and disregarded and in many instances, there is social) reasons, many of which are almost no knowledge of the cumulative environmental in nature (e.g., soil environmental and health impacts arising as salinization, sedimentation in headworks a consequence of up-scaling. For example, and channels, and drying up of rivers) small earth dams are being widely promoted (Robel, 2005). throughout much of sub-Saharan Africa for multiple uses of water including irrigation and livestock watering. In many places, Agricultural Water Development these dams have resulted in increased household income through improved agriculture. However, the potential Social Change Environmental Change environmental impacts and health consequences are rarely considered and the • food security • abiotic + biotic impacts of many thousands of dams are • poverty • local + distant • institutions unclear. In Cameroon, the development of • immediate + long-term • global trade hundreds of small agro-pastoral dams led to a rapid spread of schistosomiasis (Ripert and Raccurt 1987) Similarly in Ethiopia, the Health Impacts construction of small dams in Tigray has led to outbreaks malaria, where previously there positive & negative were none (Ghebreyesus et al. 1999).

2.Environmental and Health Planning Figure 1: Influence of agricultural water in Irrigation Projects

The need to take environmental and health considerations into account as part of 24 Examples include agricultural pests and weeds ensuring sustainable development is now and the establishment of aquatic vegetation in the widely recognized. Many countries in sub- water storage, distribution and drainage systems. 390

Saharan Africa have national policies, the environment, such as the health impacts of strategies (e.g., National Environment water pollution and many also consider a Action Plans) and legislation that stipulate range of physical/biological impacts on the need for appropriate environmental directly affected groups (e.g., displacement or planning and management of projects. Most adverse impacts on local communities). international financing institutions (e.g., The Nevertheless, current coverage of human World Bank, ADB and IFAD) as well as health aspects within environmental and many bilateral donors (e.g., CIDA, Danida, social assessments is widely regarded as DFID, GTZ and USAID) and international inadequate (Birley et al. 1997). Public health development agencies (e.g., FAO), have agencies are often excluded or only environmental policies that mainstream marginally involved and Environmental environmental issues at operational levels Health Impact Assessments (EHIA) are (Bos, 1999). Commercial organizations (e.g. generally underutilized as tools for health banks) are also increasingly environmentally protection (Fehr, 1999). aware and many have signed up to the Equator Principles, which provide a Practical approaches to EHIA have been common framework to manage advocated by the World Health Organization environmental and social issues and the Asian Development Bank. The (http://www.equator-principles.com/ WHO/FAO/UNEP/UNCHS Panel of Experts principles.shtml). on Environmental Management for Vector Control (PEEM), jointly with the Danish To support these policies a large number of Bilharziasis Laboratory, developed a training Environmental Assessment (EA) tools have course on rapid health impact assessments, been developed. These include project-level later further refined and disseminated by the Environmental Impact Assessment (EIA), Liverpool School of Tropical Medicine Strategic and Sectoral Environmental (Birley 1995; Furu et al. 1999; Bos et al. Assessments (SEA), Social Impact 2003). In addition, some good text books are Assessments (SIA), Health Impact now available (e.g., Kemm et al. 2004). Assessments (HIA) and Environmental However, for the most part EHIA Audits and Appraisals (EAA) (Table 1). development has occurred in parallel, but is However, currently there remains not integrated, with EIA methodologies. considerable diversity among donors and There is need for much better integration. A other institutions in their mandates and policy shift is required so that institutions approaches to dealing with social issues promote EHIA rather than EIA (Amerasinghe (including health). Most institutions routinely and Boelee 2004). consider social impacts that are mediated by

391

Table 1: Environmental, Health and Social Assessment Tools

Environmental Impact Assessment (EIA) A process for examining the environmental and human consequences, both beneficial and adverse, of a proposed development activity, and for incorporating appropriate measures to address them into project design and implementation. In many instances EIA is defined broadly to include social dimensions such as health. Health Impact Assessment (HIA) Similar to EIAs, these are intended to focus specifically on the health implications of a project, in situations where greater emphasis is required. Social Assessment (SA) Similar to EIAs, these are intended to analyze, manage and monitor both the intended and unintended social consequences of a development. They may be used to promote social goals such as social inclusion or poverty reduction. Strategic Environmental Assessment (SEA) A process to assess the environmental and social implications of strategic decision-making. SEA differs from EIA in that it is applied to policies, plans and programs rather than to projects. It addresses a number of shortcomings of EIA in that it is capable of addressing the cumulative impacts of projects (i.e., where one project stimulates other development), it can address synergistic impacts (i.e., where the impact of several projects exceeds the sum of the individual project impacts) and it can address global impacts such as biodiversity loss. Environmental Management Plans (EMP) Strategies developed for ongoing activities to avoid, mitigate or compensate for adverse impacts. They should include specific quantifiable aims and objectives and assign responsibilities and budgets for the environmental and social (including health) impact management measures. Environmental Audits and Appraisals (EAAs) Determine the effectiveness of mitigation measures conducted and, where appropriate, propose remedial measures.

392

3. Constraints to successful planning inadequate. For example, in Ghana a data and management checklist sent to 22 irrigation schemes, provided limited information for just three In common with other regions in the World, schemes. This lack of data was attributed to in sub-Saharan Africa, approaches to reduce poor data keeping and the fact that, for the the negative impacts of irrigation schemes majority of schemes, assessment of are successful in some circumstances but are environmental and/or health impacts had not effective in others. Constraints to never been undertaken (Kranjac- environmental and health management, and Berisavljevic and Cofie, 2004). In relation to the successful implementation of measures to health, baseline information on the health ameliorate negative impacts, arise for a range and socio-economic status of communities, of technical reasons as well as limitations in and hence their susceptibility to change, is human, financial and institutional capacity. often lacking (Fehr, 1999).

Many countries lack the resources to To a large extent the effectiveness of properly enforce policies and to ensure that environmental and health management recommended practices are followed. depends on the abilities of those people who Consequently, despite current national plan and manage mitigation measures. In policies, EIAs are most often restricted to many parts of sub-Saharan Africa, the large construction projects and are largely requisite professionals are unavailable or not donor-driven. For example, the effectiveness proficient in the interdisciplinary working of the Environmental Council of Zambia habits necessary for successful (ECZ), the lead institution for overseeing environmental and health planning and EIAs in Zambia, is severely curtailed by the management. Furthermore, there is often a inadequate budget and limited human lack of coordination between relevant resources (McCartney et al., 2004). government departments. In a review of World Bank projects in Africa, the most Failure of measures to mitigate negative frequently cited recommendation for impacts often stem from a lack of sufficient corrective action for environmentally information at the design stage in planning sensitive projects was improvement of (Morardet et al. 2005). More often than not, capacity in responsible institutions (Green baseline information is unavailable for and Raphael 2000). This lack of capacity is irrigation projects. Furthermore, adverse being be exacerbated by the HIV/AIDS environmental and health consequences often epidemic (Cohen, 2002). occur because schemes are planned and Another major limitation to formal managed in isolation from other things environmental and health procedures is that occurring within the catchment. In many often there are no mechanisms to ensure instances too little thought is given to the adaptation in the design of the project. dynamics of catchment change and there is Usually the people who demand the inadequate evaluation of the specific assessment are not the same as those who biophysical and socio-economic context in decide on changes in the project design or which the scheme is located. Available even whether or not the project will be descriptions of effective mitigation measures carried out. Most sub-Saharan African generally do not include the underlying countries have neither the necessary assumptions or specifications that were used framework to ensure legal compliance nor to design them. Without appropriate criteria organized civil society to ensure that and specifications for the design of recommended environmental and health measures, it is unusual for the measures to safeguards are implemented. In such achieve the desired goals. To develop the situations the contractual arrangement with necessary criteria and specifications, the donor may be the major means for sufficient information must be obtained. ensuring compliance. However, in the Very often the environmental, ecological and absence of a transparent accountable system socio-economic monitoring required, both this arrangement is rarely successful. Very for design prior to the implementation of a little is known about the proportion of scheme and afterwards to assess the assessments that lead to actual adaptations or effectiveness of protection measures, is implementation of mitigating measures. 393

However, a study of the follow-up to the EIA management, the recommendations are conducted for the Koga irrigation scheme in divided into three categories: Ethiopia found that only two of the twenty major recommendations made in the EIA • strategic planning at national and were being implemented satisfactorily. The regional level lack of follow-up was largely attributed to • agricultural water projects for which institutional failure, with no single authority full environmental and health being responsible for ensuring that the EIA assessment should be mandatory (i.e., recommendations were implemented (Abebe all government and donor funded et al. 2007). projects, whatever their size, plus all In a global review of the effectiveness of its other projects involved in commercial EA procedures the World Bank found that agricultural production and greater than key constraints to successful implentation in 20 ha in extent25) projects for which EA was not deemed to • agricultural water projects that by-pass have been performed effectively were: i) the current procedures and for which it is lack of a definitive Environmental unrealistic to expect full environmental Management Plan (EMP) with time-bound and health assessment to be conducted actions and responsibilities; ii) the absence of (i.e., private, community and NGO environmental monitoring indicators; iii) a organized projects smaller than 20 ha in lack of reporting requirements for project extent2). performance (including environmental and health indicators) and iv) the absence of legal 4.1 Strategic planning at national and commitments by borrowers to undertake regional level environmental actions (World Bank 1997). 4.1 Implement Strategic Environmental 4. Recommendations Assessment at regional and national level

Clearly, if irrigation is to make a significant Strategic environmental assessments (SEAs) contribution to realizing the potential of can be used to plan irrigation development at agriculture in sub-Saharan Africa, there is national level and for major international need for much improved and integrated river basins (e.g., Zambezi, Limpopo, Volta). planning. Measures that promote SEAs are most valuable if they integrate sustainability by, among other things, environmental, health and social concerns capitalizing on the opportunities for and attempt to reconcile development, enhancing human health, should be at the environmental protection, community rights core of agricultural water development. The and human health. Regional and national EA process, recommended by most donors development goals, as well as issues such as and governments, is widely recognized as a climate change and loss of biodiversity, and useful for identifying issues and developing commitments to international conventions plans to address them. However, within sub- (e.g., the Convention on Biological Saharan Africa there are, as outlined above, Diversity) should be considered. many constraints to the process and subsequent follow-up is often weak. 4.2 Improve and promote EHIA Furthermore, the process is inappropriate for many small-scale developments. Subjecting Currently health impact assessments are smallholder and community-led projects to often conducted in isolation from full environmental and health assessment, environmental assessments. Since much of and monitoring, although justifiable, is often the information to be collected on neither economically feasible nor practical. environmental receptivity and community Against this background, the following recommendations are a pragmatic attempt to address current limitations in environmental 25 The suggested value of 20 ha is arbitrary but and health planning and management intended to make recommendations pertaining to irrigation development. operationalizable. Governments could decide on Focused on what governments and donors a more appropriate figure, based on the specific agro-ecological conditions and development can do to improve planning and needs of their country. 394

vulnerability is the same it is mutually term cohort studies are required that are not beneficial if they are integrated. Where feasible within the context of individual necessary, EHIA should specifically include EHIA. the issues of migrants and livestock that hitherto have tended to receive very little Agricultural water projects for which full attention. environmental and health assessment should be mandatory 4.3 Improve regional capacity for Environmental and Health Assessment 4.6 Implement comprehensive options assessment All countries without compulsory environmental and health assessment Comprehensive options assessments, processes should consider enacting laws that undertaken during the scoping of irrigation make these mandatory for large infrastructure projects, provide a means, early in the projects, including large irrigation projects. In planning process, to eliminate unacceptable many countries strengthened institutional projects or project components. arrangements would assist in the Comprehensive environmental and social implementation of environmental and health audits can help determine the causality of assessment processes. For example, environmental and human impacts and the establishing environmental units within relative magnitude of impacts at a basin or Government ministries responsible for regional level, which can then be compared irrigation could be contemplated. The to alternative development scenarios. It is effectiveness of such units would be enhanced essential that environmental, health and if they work closely with national environment social criteria, as well as technical, economic agencies and appropriate health authorities. and financial factors, are considered when comparing alternatives. 4.4 Adopt harmonized environmental and health procedures 4.7 Identify and quantify intended livelihood and health benefits The ability of governments to implement sound environmental and health practices The environmental and health impacts of would be improved if donor agencies irrigation are diverse. As with any harmonized procedures and developed a development process, trade-offs between consistent framework for the evaluation, social, environmental and economic goals planning and management of environmental are inevitable. As far as possible these trade- and health aspects of irrigation. Procedural offs need to be identified and made explicit. requirements should conform to current Often it is assumed that, by improving food international best practice and be clearly laid security and/or peoples’ socio-economic down in regulations and operational manuals. status, water development will inevitably result in health benefits and improved 4.5 Conduct regionally specific research livelihoods. However, the intended livelihood and health benefits are rarely More research is needed on the benefits of made explicit, and in reality, neither costs incorporating environmental and health nor benefits are evenly distributed amongst safeguards in irrigation planning and stakeholders. Environmental and health operation versus the cost of not taking assessments, as well as management plans, potential negative impacts into account. need to take into account the socio-economic Another researchable issue stems from the diversity of communities and ensure that the lack of monitoring, both for water resources weakest and most vulnerable are not development projects and in the health adversely affected. Intended health and sector. With baseline data not available, livelihood benefits as well as means of proxies need to be developed to provide verification need to be identified and stated alternative ways to the same information. at the outset of any irrigation scheme. Specific tools need to be developed to facilitate assessment of long-term health and environmental impacts. For example, long 395

4.8 Plan and manage using a catchment- cover these, in many cases it is not. For this wide perspective reason donors and governments ought to investigate innovative ways of financing Given the inter-linkages between impacts recurring costs, such as trust funds26. and what occurs elsewhere in the catchment it is essential that projects are planned and 4.11 Develop innovative approaches managed within the specific socio-economic to ensure compliance with and biophysical context in which they are environmental and health located. Consideration must be given to requirements potential environmental impacts on, as well as impacts caused by, the development. Incorporating environmental protection and Assessments of impacts on the catchment health measures into irrigation projects is water balance and sediment fluxes, including made difficult by the failure of many project evaluation of possible future development operators to fulfill voluntary and mandatory (particularly relating to land-use change), are obligations. Innovative approaches to essential. The potential cumulative affect of encourage compliance ought to be small-scale interventions should be investigated. Options could include: a) the use specifically included in assessments. of performance bonds, supported by financial guarantees and expressed in wellbeing-related 4.9 Improve data generation and outcomes and not just agricultural yields and analysis related to environmental and health water use efficiency; b) implementation of a impacts sector-specific environmental management system, perhaps constructed around that A major constraint to the sustainability of developed by the International Standards agricultural water development is the lack of Organization (ISO); c) development of an site-specific data and long-term monitoring; ethical code for large-scale irrigation projects pre-requisites for informed decision-making. to ensure that environmental and health For this reason measures to significantly concerns are adequately addressed. improve data generation and analysis related to environmental and health impacts should Agricultural water projects for which it is be encouraged (e.g., coordination of existing impractical to conduct full environmental data collection efforts between sectors and/or and health assessment establishment of meta-databases). Ideally monitoring strategies would be mandatory in 4.12 Increase local-level awareness of all projects and governments environmental and health issues and donors must provide adequate funding to enable this. Governments and donors should support campaigns of health awareness carried out by 4.10 Develop innovative ways for community health teams and training financing environmental and health programs that, in collaboration with measures community groups (e.g., farmers associations, agricultural water user associations, water The cost of effective environmental and committees and women groups), increase health measures is often very high and must usually be borne by the organization 26 responsible for the irrigation development. Trust funds have been suggested as a possible mechanism for financing the mitigation of the The most common mechanism for financing environmental impacts caused by large dams these measures is to incorporate the costs (Bizer 2000). A project-specific trust could be into the capital financial package of the established at the outset of a project (by the project. The costs that are most readily project financier) with the condition that funds incorporated into the capital costs are those are used specifically for environmental and health that occur once (e.g., construction of fish management, including monitoring. The ladders in dams). Financing on-going approach could incorporate annual contributions obligations, such as environmental and from the scheme owners/beneficiaries as well as health monitoring, is more difficult. Whilst it other organizations (e.g., governments or donors), is sometimes appropriate for beneficiaries to with the environment and health program funded from the proceeds of the trust. 396

awareness of potential environmental hazards effort and long-term commitment from both and approaches to mitigation. Information on governments and donors. If implemented practical ways to maximize health benefits they will contribute to better awareness of should be provided, as well as outlining the linkages between environmental and potential hazards and approaches to mitigate health impacts and improve the sustainability negative impacts. of irrigation development in sub-Saharan Africa. 4.13 Develop “user-friendly” methods of rapid appraisal for evaluating small- References scale projects Abebe, W., Douven, W.J.A.M., McCartney, Donor and government funded programs that M.P. and Leentvaar, J. (2007) EIA promote small-scale development (e.g., the implementation and follow up: a case community driven development program of study of Koga irrigation and watershed the World Bank), should conduct program- management project. Paper presented at specific environmental and health IWMI workshop 4-5 June 2007, Addis assessments. These should assess the Ababa, Ethiopia. potential impacts of the micro-projects to be financed under the program and the possible AdamsW.M. 1992. Wasting the rain: rivers, cumulative impacts of scaling-up. They people and planning in West Africa. should set the context for lower-level Earthscan, London. assessments and, based on the priorities for attention, simple checklists, intended for use Amerasinghe, F. and Boelee, E. 2004. by small local organizations and Assessing the impact of irrigation communities, should be developed to development on the environment and evaluate the impacts for individual micro- human health. In: Proceedings, projects. Inception workshop on IWMI-BOKU- Siebersdorf-EARO-Arbaminch 4.14 Ensure programs that promote University collaborative study on the small-scale agricultural water impact of irrigation development on development are embedded within poverty and the environment, 26-30 rural development programs April 2004, Addis Ababa, Ethiopia.

Governments and donors should ensure that Birley, M.H. 1995. The health impact programs promoting small-scale irrigation assessment of development projects. are undertaken in conjunction with broader HSMO, London, UK. rural development programs that include water and sanitation, as well as health Birley, M.H., Gomes, M. and Davy, A. 1997. components. Care must be taken that these Health aspects of environmental projects are designed so that the main assessment. Environmental assessment beneficiaries are clearly identified and the sourcebook update 18. Washington D.C., objective of improving livelihoods through USA: World Bank irrigation remains the primary focus. Bizer, J.R. 2000. Avoiding, minimizing, Concluding remarks mitigating and compensating the environmental impacts of large dams. Addressing environmental and health Report to the World Commission on impacts are crucial for the sustainability of Dams. World Commission on Dams, future irrigation development in sub-Saharan Cape Town, South Africa Africa. The recommendations presented above focus on ways to improve the policies Bos, R. 1999. Water resources development and practices pertaining to impact and health: the policy perspective. Pp31- assessment and planning for both large and 46 in Kay, B.H. (ed) Water resources: small developments. To be effective the Health, environment and development. recommendations require a coordinated New York, NY, USA: E & FN Spon Ltd. 250 pages. 397

Hussain, I. and Hanjra, M.A. 2004. Irrigation Bos, R.; Birley, M.; Furu, P. and Engel, C. and poverty alleviation: review of the 2003. Health opportunities in empirical evidence. Irrigation and development. A course manual on Drainage 53: 1-15. developing intersectoral decision-making Kay, B.H. (ed) 1999. Water resources: skills in support of health impact Health, environment and development. assessment. Geneva: World Health New York, NY, USA: E & FN Spon Ltd. Organization. Kemm, J.; Parry, J. and Palmer, S. 2004. Cohen, D. 2002. Human capital and the HIV Health impact assessment: Concepts, epidemic in sub-Saharan Africa. Geneva, theory, techniques, and applications. Switzerland: ILO Programme on Oxford: Oxford University Press. HIV/AIDS and the World of Work. Kikuchi, M., Inocencio, A., Tonosaki, M., Dougherty, T.C. and Hall, A.W. 1995. Merrey, D., de Jong, I. and Sally, S. Environmental Impact Assessment of 2005. Cost of Irrigation Projects: A Irrigation and Drainage Projects Comparison of sub-Saharan Africa and Irrigation and Drainage Paper 53, FAO, Other Developing Regions and Finding Rome, Italy: FAO. Options to Reduce Costs. Report for the World Bank as part of the Investment in Fehr, R. 1999. Environmental health impact agricultural water management in Sub- assessment, evaluation of a ten-step Saharan Africa: Diagnosis of trends and model. Epidemiology 10: 618-625. opportunities. Pretoria, South Africa, IWMI. May 2005 Furu, P.; Birley, M.; Engel, C. and Bos, R. 1999. Health opportunities in water Konradsen, F.; Amerasinghe, F.P.; van der resources development: A course Hoek W. and Amerasinghe, P.H. 2000. promoting intersectoral collaboration. In Malaria in Sri Lanka: Current knowledge Water resources: Health, environment on transmission and control. Colombo, and development, ed. Kay, B.H. New Sri Lanka: International Water York, USA: E & FN Spon Ltd. pp. 86- Management Institute. 107. Kranjac-Berisavljevic, G. and Cofie, O. Ghebreyesus, T.A.; Haile, M.; Witten, K.H.; 2004. Comparison of formal and Getachew, A.; Yohannes, A.M.; informal agricultural water development Yohannes, M.; Teklehaimanot, H.D.; using selected environmental and social Lindsay, S.W. and Byass, P. 1999. indicators, in Ghana. Case Study Report Incidence of malaria among children for Collaborative Program on living near dams in northern Ethiopia: Investments in Agricultural Water Community based incidence survey. Management in Sub Saharan Africa: British Medical Journal 319: 663-666. Diagnosis of Trends and Opportunities. Accra, Ghana: International Water Green, K.M. and Raphael, A. 2000. Third Management Institute. environmental assessment review (FY 96-00). Environment McCartney, M.P., Daka A.E. and Ntambo, I. Department,Washington, D.C.: World 2004. Environmental and health impact Bank assessment of dambo utilization in Zambia. Case Study Report for Hunter, J.M., Rey, L., Chu, K.Y., Collaborative Program on Investments in Adekolu-John, E.O. and Mott, K.E. Agricultural Water Management in Sub 1993. Parasitic diseases in water Saharan Africa: Diagnosis of Trends and resources development. The need for Opportunities. International Water intersectoral negotiation. WHO World Management Institute Health Organization, Geneva. 152 pp Morardet, S., Seshoka, J., Sally, H. and Merrey, D. 2005. Review of irrigation 398

planning and implementation process. Report to the African Development Bank. Pretoria, South Africa: IWMI

Mutero, C.M.; Kabutha, C.; Kimani, V.; Kabuage, L.; Gitau, G.; Ssennyonga, J.; Githure, J.; Muthami, L.; Kaida, A.; Musyoka, L.; Kiarie, E. and Oganda, M. 2004. A transdisciplinary perspective on the links between malaria and agroecosystems in Kenya. Acta Tropica 89: 171-186.

Oomen, J.M.V., de Wolf, J. and Jobin, W.R. 1990. Health and irrigation: incorporation of disease-control measures in irrigation, a multi-faceted task in design, construction and operation. Wageningen, The Netherlands: ILRI.

Ripert, C.L. and Raccurt, C.P. 1987. The impact of small dams on parasitic diseases in Cameroon. Parasitology Today 3(9): 287-289.

Robel, L.W. 2005. Assessment of design practices and performance of small-scale irrigation structures in South Region. Unpublished MSc thesis, Arbaminch University, Ethiopia.

Smith, L.E.D. 2004. Assessment of the contribution of irrigation to poverty reduction and sustainable livelihoods. International Journal of Water Resources Development 20: 243-257.

World Bank, 1997. The Impact of Environmental Assessment: a review of World Bank Experience. Environment Department, Washington, D.C., USA: The World Bank.

399

GIS And Remote Sensing Integrated Environmental Impact Assessment of Irrigation Project In Finchaa Valley Area

Ahmed Amdihun Addis Ababa University, Ethiopia [email protected]

Abstract The post irrigation development years This research was conducted in order to witness that some water born diseases and assess the environmental impacts of Finchaa malaria case records have increased. It is irrigation project using GIS and remote also found out that expansion of irrigation sensing techniques. Because of the limited fields and the respective Malaria case resources only some environmental records are positively correlated. parameters were selected. These parameters Generally, despite the significance of the are natural vegetation, soil/land, water Finchaa irrigation project, it has negative quality, climate and health conditions. environmental repercussions. This is evident The normalized vegetation index (NDVI) from vegetation cover distraction, water analysis was used to detect the spatial and quality deterioration in the down stream temporal change of vegetation biomass in area, alteration of soil physical and chemical the study area. The result indicated that the components and increasing health threats. If natural vegetation biomass is declining. This the current condition continues the problems is mainly due to the expansion of may out weight the benefits of the irrigation agricultural land and escalating human made project. Thus in the project site and structures in the area. adjoining areas urgent environmental The water physico-chemical analysis conservation is necessary. This helps to demonstrated that the down stream water sustain the existing and revitalize the fading has more chemical substances and degraded resources. physical properties than the up stream counterparts. The direct leakage of industrial 1. Introduction liquid waste and the agro-chemicals from irrigation fields are supposed to contribute 1.1 Background Information for this result. The GIS analysis of a 100 The expansion of irrigation scheme in meter buffer around Major rivers and Ethiopia lend a hand to achieve food self tributary streams is found to be a necessary sufficiency and poverty reduction. Irrigation action to mitigate the problem of pollution. agriculture makes production more The soil chemical and physical property unwavering than the rain fed agriculture. analysis in the irrigated and non irrigated Proper planning and management aided fields reveals that the soil samples taken irrigation projects contribute for the growth from irrigated fields contain higher of national GDP and GNP. It also creates phosphorous, Nitrogen and organic carbon job opportunities for several thousands of compared to its counter parts. The use of people directly or indirectly. agro-chemicals in the irrigation fields Despite their significances, however, contribute for the result. irrigation practices have sometimes adverse From the twenty two years rain fall, impact on environmental conditions. It is temperature and humidity data no abrupt known that Human activities have a inclining or declining trend is observed that profound effect up on the natural could tell the possible impacts of the project.

400

environment and are becoming the main Geographic information system (GIS) and agent of environmental degradation. remote sensing can facilitate the study of Finchaa valley was one of the few areas in environmental impact assessment of Ethiopia to preserve its natural conditions irrigation projects for a better outcome. for years. The topographic set up made the area to be inaccessible. In 1975 the valley 1.2. Objectives of the study was selected as a suitable site by the state farm to produce food and commercial crops. 1.2.1. General Objectives After few years the area was again chosen to ¾ To Asses the impact of the establish a sugar factory. Following these irrigation project on the natural there were many activities carried out in the environment of Finchaa valley area. Some of these activities have an Area. enormous positive contribution while some others have negative effect on the 1.2.2. Specific Objectives environment. • To see the impacts of Finchaa In Finchaa valley, following the irrigation scheme on vegetation establishment of the sugar factory more cover using satellite images of pronounced land degradation is observed. different years. There is large scale land clearance • To explore the extent of soil quality (deforestation) by the factory for new degradation as a result of the irrigation field. irrigation scheme. In addition there are many people in and • To asses the impact of the irrigation outside the valley that earn their livelihood on water quality in the upstream from forest and forest products. and down steam water of Finchaa The factory uses agrochemicals like River. fertilizers, pesticides and herbicides in • To see the climatic change irrigation fields and the wash away collected (temperature and rain fall) over the in ditches and then join the nearby tributary past 22 years and interpret the stream. On the other hand some Part of the results on sustainability of the liquid waste from the factory spillover the irrigation project. treatment plant and joins Finchaa River. The • To investigate some health threats problems emanate from the little attention following the irrigation project given for environmental conservation as the spatially and temporally. main objective is to maximize production • To propose some valuable measures and productivity. The cumulative effects of to be taken to mitigate the negative these all problems can result in impacts of the irrigation project on environmental degradation.In Finchaa valley the environment in such away that there exists a continuous disturbance on assures sustainable development. vegetation cover, soil, and water. If the trend goes on, there will be an extreme effect on 1.3 Research Methodology the environment. Thus, it is indispensable to carry out environmental impact assessment In order to make out the positive and (EIA) of the irrigation project in the area. negative impacts of the irrigation project on Environmental Impact Assessment has been the environment of Finchaa valley area more recognized as an integral part of the early of primary and some secondary data are planning studies of irrigation projects in collected. Some of these data are integrated order to identify any expected negative with GIS and remote sensing techniques in a impacts and to suggest the necessary actions way that manifests the impacts of the to curb the problem. In addition, EIA can irrigation project on the environment. consider different designed alternatives for To best investigate the positive as well as the project as an essential step for better negative impacts some components of the decision making. The application of 401

environment are preferred for investigation. The unstructured interview to the concerned These are vegetation cover, soil/land, water, bodies and past research works furnish with climate and health cases records. valuable information with respect to the The first line consideration is given to the past-present natural and socio-economic direct environmental impacts of the setup of the area. These data are integrated irrigation scheme. In light of this the indirect with impact assessment and GIS/RS impacts are also inspected to the best of the techniques in such a way that shows the kind researchers’ knowledge and available and extent of changes that have been taking resources. place. Different years of satellite images are used for the vegetation cover change with the 2. Impacts of the Irrigation project on expansion of the irrigation in Finchaa valley. Vegetation and Soil Under this the scope and extent of variation in land cover, land use, reflectance 2.1. Impacts of the Irrigation project on properties, Image differencing, erosion vegetation cover estimation and the NDVI analysis are 2.1.1. General Conditions of Natural explored and quantified. Vegetation Cover To investigate the impacts of the irrigation In developing countries the attention given on the soil the physical and chemical soil for vegetation conservation is less compared analysis has been made. The soil samples in to the need for development. In realizing different sites were collected. These sites are their policies for food self sufficiency and the irrigated fields, ploughed but not yet agricultural productivity preeminent value is planted and non irrigated (vegetated) areas. given for irrigation developments even some The samples were taken in three layers and times at the expense of environmental totally nine samples were analyzed. The considerations. result is believed to show the soil Depending on the management system component anomalies in the irrigated and irrigation projects can have both positive as non irrigated areas and the possible causes. well as negative impacts on vegetation Visual presentation of the land with and cover. Undoubtedly the expansions of without irrigation also gives some idea about irrigation projects have many advantages. the level of land degradation. Some GIS However, in most cases there happens integrated slope analysis also provides slope change in the natural ecosystem following differences and the intensity of erosion. large scale irrigation developments. In order to investigate water quality Obviously in order to under take large scale problems water samples from upstream and irrigation projects the vegetation cover in the down stream areas were taken and these area needs to be cleared and different samples are supposed to show the spatial construction activities should be carried out. water quality changes. This intern helps to Natural Vegetation as one of the eminent examine the impacts of the project on water part of the ecosystem is negatively affected quality. Quantity wise the irrigation water with such development activities. Large use will be incorporated to asses the scale forest resource degradation can change problems emanating from under and/or over the natural environment. This in turn puts utilization of water. the sustainability of irrigation projects in Long year’s meteorological data are used in question. Conversely if appropriate order to evaluate the micro climate of consideration is given for vegetation Finchaa valley area for temporal anomalies. conservation the forest area can be In light of this panorama other delineated and effective afforestation and environmental components are examined reforestation can be carried out. For that and possible solutions recommended. This matter vegetation resource can be keep hold again helps to foretell the sustainability of around the hills, on vacant and marginally the irrigation scheme in relation to climatic suitable lands. The conservation of natural favorability. vegetation can fix the problem of soil 402

erosion, micro climatic disturbances, accessible and vulnerable for human biodiversity and it balances many of the interference. This opened up a new episode environmental systems. for the forest resource exploitation. Still to Well planned irrigation schemes have good the present Finchaa valley is considered as natural vegetation conservation and an ideal site for hard wood and bamboo management plans. Effective management forests used for fire wood and construction and proper balancing of these seemingly activities.The beginning of 1990s can be conflicting issues should be treated wisely. seen as the second turning point in the forest Finchaa valley in the pre 1975 years was history of the area. In these years Finchaa virtually under natural vegetation cover. The valley was selected as the most suitable site tall savanna grasses mixed with short and for sugar cane plantation and industrial medium trees predominate the elevation development. In the mean time the state below 1600m.The steep escarpments and the farm abandoned the farm and handed over far down stream areas experience dense the area to Finchaa sugar Factory. In 1991 vegetation growth. The gallery forests the Finchaa sugar project started extensive occupy the networks of major rivers and mechanized vegetation clearance. From the their tributaries. As it is evident from the three major Companies that carried out the MSS satellite image of 1972 there was no feasibility study any of them did not apparent human intrusion to the valley. recommend for any single area buffering for From unstructured interview made with natural vegetation conservation. Almost all local elders there were some individuals attention was on sugar cane production and who went to the area to collect wild honey strategies for expansion. Accordingly the from trees and hunters for valor. west bank of Finchaa River was considered The first intrusion to the valley was more suitable and fertile and at present successfully made by the state farm in 1975. about 8,064.88 hector is under sugar cane Since this time it is estimated that the state plantation. (See fig 4.3)The factory farm cleared about 3,500 hectors of land neglected the east bank until the recent (vegetation) .The 1986 TM image reveals years. This year Vegetation clearance and that some parts of the eastern and western land preparation has been taking place on banks of Finchaa River are occupied with the eastern bank. The total area of 7,108 some food and commercial crops. Even at hector is expected to be irrigated. Despite this moment most part of the valley was the fact that the expansion escalates the under the natural vegetation cover. industrial productivity, it further aggravates The construction of the road dawn the the problem of deforestation in the valley. escarpment made the forest resources

Figure 2.1. RGB/321 MSS satellite images of Figure 2.2. RGB/321 TM satellite Valley Area 1972 with 30 meter resolution. Valley Area 1986 with 30 meter resolution.

403

From the visual image interpretation it is evident that there is land cover change. Some features like the vegetation biomass are diminishing while some others like Finchaa Lake are increased in size. Even though there is large scale vegetation clearance there are efforts made by the factory to plant trees on unused areas. To the present Finchaa sugar factory under forestry department planted 600 hectors of land in the afforestation program. One of the tree species preferred is eucalyptus tree .This was so for the accessibility of the tree.

Figure 4.1. RGB/321 ETM satellite images of Finchaa Valley Area 2000

Currently the department has the plan to cover 2005/06 it increased by more than seven folds. 2,200 hectors of land under the reforestation By implication the vegetation cover is and afforestation programs.However, this retreating with the same or even more rate. amount is very less with respect to the Deforestation is the major problem in Finchaa vegetation clearance which has been taking valley area. The large scale vegetation place for long years in the area. In comparison clearances by the factory together with to the vegetation that has been cleared the individuals earning a livelihood from forest present afforestation program reclaim for not products are devastating the vegetation more than 7.5 percent of the land under resource. Forest fire is one of the critical irrigation. Even if the future goal of the causes for the vegetation degradation. In department is attained it reclaims only for addition to the naturally instigated fire the about 27.3 percent of the present land under factory and some individuals play a significant irrigation. At the time 13,000 hectors of land role in triggering the problem. The fire is irrigated if only 2,200 hectors of land is escaping from the frequent cane burning by under forest cover it means that less than 17 the factory and irresponsible action by percent of the cleared land is revitalized. individuals who are looking for timber, The irrigated field has increased a lot. In charcoal, fire wood, Wild honey collection, 1997/98 about 932.27 hectors of land was construction wood and others exacerbate the harvested and after eight years that is in obliteration.

2.1.2. The Normalized Difference Vegetation Index Figure4.6. The Normalized vegetation Index results of Sept.2000 ETM image Normalized difference vegetation index (NDVI) is a method used to analyze the vegetation cover of an area. NDVI is calculated from reflectance measured inthe visible and near infrared channels from satellite-based remote sensing. NDVI shows the temporal and spatial change of Vegetation cover. The difference between two images is calculated by finding the difference between each pixel in each image and 404

generating an image based on the result. The NDVI Analysis

Figure4.6. The Normalized vegetation Index results of Dec.1972 TM image of the 1972 Multi spectral scanner (MSS) compared to the 1972 image. The expansion image of Finchaa valley area reveals that there of cultivated areas, bare lands and built up is more vegetation biomass in the study area areas are apparent in the NDVI analysis. (NDVI>0) compared to the later years. These areas appeared as deep red and NDVI < 0.0. This means that many areas that were The Normalized vegetation index of 2000 formerly under vegetation cover are turned up image shows lesser vegetation biomass into Human made features.

Table 4.1: The Normalized Vegetation index result of the 1972, 1986 and 2000 satellite images.

Year Landsat MSS Landsat TM Landsat ETM Mean Standard Mean Standard Mean Standard deviation deviation deviation 1972 87.3427 66.2614 - -

1986 - - 80.4064 63.1612

2000 68.2834 58.5612

Table 4.1 Reveals that the mean and standard and large scale deforestation that has been deviation of the 1972, 1986 and 2000 images taking place for many years. Still the present has been decreasing. This could indicate the trend indicates that the deforestation will rate of vegetation cover destruction. continue to the virgin lands. By taking the Generally the Visual image interpretation and aspiration of the factory for expansion in to the Normalized vegetation index results consideration large effort should be made in confirm that the vegetation biomass of afforestation and reforestation projects. Strict Finchaa valley area has been diminishing. measures should be taken to stop illegal forest There are three major factors that can explain resource exploitation and the frequent fire. this circumstance. These are the expansion of Afforestation and reforestation activities agricultural lands, growing settlement areas should not be considered as a superfluous 405

activity. Beyond harmonizing many of the weather conditions, degradation of natural systems they can serve as a means to biodiversity, fire wood and wood product solve many problems like soil erosion, hot requirements and many others.

4.2. Impacts of the Irrigation project on Soil soils which include Salinization, Alkalization, Water logging, soil pollution and Soil Soil is one of the most decisive natural acidification.There are two dominant soil resources. It has been supporting the types in the project area; these are increasing number of life in our planet earth. the Luvisols and vertisols. Luvisols covers 75 Now a days the large number of population percent of the irrigated land. These soils are increased the demand for food, this in turn put partly made of alluvial and colluvial materials forth full-size pressure on land /soil resource. from the surrounding escarpments. Luvisols Areas formerly considered as marginal are has limited fertility and agricultural suitability. currently being cultivated. The demand for big Water logging is not a vital problem in the yield created enthusiasm to look for area as the factory is using over head sprinkler alternative means. One of these is getting irrigation system. This consecutively evades bigger yield through customary agricultural the problem of salinization. In order to practices like irrigation systems, use of maximize production the Agro-chemicals have fertilizers, pesticides, herbicides and many been used in the irrigation fields .The most other agricultural inputs. common ones are fertilizers, pesticides and Irrigation schemes beside their positive herbicides. The two commonly applied contributions have many shortcomings on the fertilizers are Urea and Dap. The brief physical and chemical properties of soil in summery of the total amount of agro- particular and the environment in general. The chemicals is presented in table 4.1. (See FAO repository document mentioned some of Appendix 4 for the details) the adverse impacts of irrigation schemes on

406

Table 4.2. Shows that the use of fertilizers, pesticides and herbicides has been increased in an alarming rate with the expansion of irrigation. The Use of large scale agro-chemicals alter the physical and chemical properties of the soil which can damage the soil quality and use full living organisms.

Year Fertilizers (Qunt.) Pesticides(Lts.) Herbicides(Lts.)

1994/95 219.4 480 5 1995/96 5,224 20,074.5 8 1996/97 5,806 10,403 267 1997/98 4,003 311.8 20 1998/99 15,952.68 5,278 716 1999/00 17,264.9 10,914.56 216.51 2000/01 23,097.01 25,794.31 2,943 2001/02 19,444.25 17,585.9 2,330 2002/03 23,274.7 15,899.51 1,019 2003/04 25,760.3 14,370.63 2,368 2004/05 36,538.69 17,837 2,712 Total 176,584.93 138,949.21 12604.51

In order to see whether there is change in the comparative results of samples from cultivated soil physico-chemical properties of the but not planted and non irrigated spots can dominant Luvisols in the irrigated fields reveal weather the change is due to human samples are collected and analyzed. Luvisols intervention or natural causes. The soil sample are preferred for analysis because 75 percent from vegetation cover area is supposed to of the irrigation is carried out in this soil. The reveals the natural properties of the Luvisols samples are taken from three sites in three in the area. Thus, the site selection for soil layers. The first site is the non irrigated field samples is intentional and made in such away where there is no human interference. The that shows the impacts of irrigation on the second site is ploughed but not yet planted physical and chemical properties of the soil. field. In this site none of the agricultural inputs are applied. The third site is the irrigated field 2.2.1. Irrigation and Physical properties of where the agricultural inputs have been used. the soil In each of the three sites soil samples from three spots are collected and mixed to form The Luvisols and vertisols occupy more than only one composite soil sample. The three 95 percent of the Finchaa valley area. Luvisols layers are the top layer (0-30 cm), the middle have reddish brown color and weakly layer (30-60) and the bottom layer (60- developed structure. They have also shallow 90cm).Totally nine samples were investigated profile and limited fertility. Luvisols are and the result will be presented under 4.2.1 composed of sand which decreases with and 4.2.2. The three spots are believed to show increasing depth. This soil is the most the possible positive and/or negative impacts exploited soil in the valley. About 75 percent of irrigation scheme on the physical and of the irrigation is carried out on Luvisols. The chemical properties of the soil. In addition the vertisols on the other hand have black color 407

and shallow profile. Vertisols contains more clay materials with increasing depth.

Table 4.3.The physical properties of the Luvisols from irrigated, Ploughed but not planted, and vegetated area in three layers. (0-30, 30-60, 60-90)

LUVISOLS FROM IRRIGATED FROM PLOUGHED FROM VEGETATED FIELD BUT NOT PLANTED AREA Depth(cm) 0-30 30-60 60-90 0-30 30-60 60-90 0-30 30-60 60-90 Total sand (percent) 63.55 57.33 53.34 51.16 41.87 41.78 51.56 35.78 41.35 Silt (percent) 4.17 2.08 7.26 7.60 4.39 4.39 12.37 25.27 11.52 Clay (percent) 32.28 40.59 39.40 41.25 53.74 53.83 36.07 38.95 47.13 Texture class SCL SC SC SC C C SC CL C Ph-H2o(1:2.5) 5.62 5.20 4.89 5.62 5.04 5.04 4.86 4.48 4.56 Ph-kcl(1:2.5) 5.16 4.79 4.76 5.08 4.57 4.58 4.42 4.00 3.87 Ec(ms/cm)(1:2.5) 0.03 0.04 0.02 0.03 0.01 0.01 0.06 0.02 0.01

Table 4.3 illustrates that the total sand content is more acidic than the soil in the irrigated of Luvisols decreases with increasing soil field. Theoretically the fertilizers, pesticides depth in all sampled layers. The silt content of and herbicides that have been applied to the the soil from vegetated area is higher. This can cane fields seem to increase the PH of the soil. be due to the lesser amount of erosion in But the result shows that the soil in the vegetated areas compared to the cultivated irrigation field is less acidic than in the areas. The clay content of the soil in ploughed vegetation area. Three main reasons can and vegetated areas increases with the explain this result. In the first place the surplus increasing depth. Unlike the non irrigated water use in the irrigated areas can wash the fields the clay content of the soil from chemicals vertically and laterally. Secondly irrigated site is higher in the 30-60 cm depth. cultivation by itself can alter the inherent Ph This can be due to the excess water that of the soil by exposing the soil. Finally the dissolves soluble minerals and percolates respective composition of the soil forming down. On its way it accumulates the insoluble parent material can be different in the sample clay in this horizon. Generally the texture sites. class of the Luvisols in all locations ranges from sand clay loam (SCL) to clay (C). Such 2.2.2 Irrigation and Chemical properties of soils are known to be suitable for irrigated the soil cane plantation with cautious soil management. The soils in all the three spots The Normal Soil chemical properties can be of the three layers are found to be acidic. The altered by natural and human made factors. PH is less than 5.6. It is investigated that there Industrial toxic wastes, hazardous chemicals, is perceptible PH difference between irrigated Agricultural malpractices and inputs, and and non irrigated soil. The average PH value many others constitute the human factors. of the soil in the irrigated area is 5.3 where as Alternatively due to some natural processes in in the vegetation area the value is 4.6. This the system there may be alteration of soil shows that the soil in the vegetation cover site chemical properties. In this respect the 408

physico-climatic conditions play a key role to irrigation fields. These higher amounts in the change the chemical properties of the soil.As irrigation field are due to the fertilizers (Urea it is illustrated in table 4.4 the average amount and DAP) that have been used in the irrigated of chemical elements in the three sample areas areas. The organic carbon is found in higher are different. The amount of exchangeable quantity in the vegetated area. The soil sample bases (Exch. Na, K, Ca and Mg) varies with from the vegetation area are found to be more increasing depth. Generally speaking the acidic than the irrigation Fields. This is mainly amount of Potassium, Calcium and due to high organic content in the vegetated Magnesium decreases with increasing depth areas. On the other hand the less acidic nature while sodium increases with of the soil in the irrigated area is related to the depth.Exchangeable calcium and sodium is exposure and excess water use in the irrigation higher in the irrigation and cultivated but not fields. planted fields than the vegetated areas. The total percentage of Nitrogen is higher in the irrigation field. The Available phosphorous is extremely high in the top layer of the Luvisols Irrigated field Cultivated but Vegetated area not planted Depth in cm 0-30 30- 60-90 0-30 30-60 60-90 0-30 30-60 60-90 60 Exch. Na(meq/100gm of soil 0.13 0.16 0.13 0.15 0.13 0.17 0.13 0.11 0.13 Exch. K(meq/100gm of soil) 0.20 0.15 0.16 0.27 0.20 0.21 0.48 0.18 0.16

Exch.Ca(meq/100gm of soil) 5.82 5.82 5.18 9.50 8.72 8.72 4.99 3.36 3.36 Exch. Mg (meq/100gm of soil 4.16 4.20 2.59 6.05 2.62 4.36 4.99 4.20 5.04 Sum of cations (meq/100gm of 10.48 10.54 8.19 16.13 11.90 13.69 11.39 9.66 11.34 soil) CEC(meq/100gm of soil) 14.02 14.48 14.56 23.95 25.12 26.54 24.87 21.46 15.52 Organic carbon (percent) 1.17 0.58 1.69 0.90 0.65 0.65 1.40 1.14 0.78 Nitrogen (percent) 0.14 0.10 0.14 0.13 0.09 0.09 0.09 0.07 0.05

Available P(mg p2o5/kg soil) 29.20 5.10 4.00 7.80 2.80 2.80 2.60 2.70 2.70 Exchangeable Acidity 0.17 0.21 0.13 0.13 0.22 0.22 0.38 1.79 2.65

409

In addition the acidic nature of the soils in the intensification of gully and considerable vegetation areas shows that the soil in the area deposition on the roads and cane fields. The is naturally acidic and the human intervention active erosion and expansion of gully in the minimizes the soil acidity. road side made road construction a year round Generally the analysis of the three soil samples activity. In addition beyond taking the fertile indicates that there is alteration of some of the top soil erosion has been expanding active soil physical and chemical properties as a gullies and turns the potentially irrigable lands result of the irrigation scheme. The level of in to bad land. alteration hardly results in full-sized soil Figure 4.9. Slope based Interpolation Map pollution at this level. However the showing General Conditions of erosion cumulative impact could grow in to soil quality degradation. Thus there are signs of soil There is a general elevation decline from south pollution in the Irrigated areas. There are to north and from the eastern and western several reasons that can explain this condition edges to Finchaa River. This indicates that the of which application of the agro-chemicals is general trend of erosion is to Finchaa River one. first and finally to the Abay gorge. The tributaries fed fertile soil to Finchaa River and 2.2.3. Soil Degradation the soil finally transported to Abay River. The digital elevation model based run off It is found out that one of the preeminent estimate indicates that there is high runoff problems of soil in Finchaa valley area is pattern in the areas that lies from the eastern erosion. There is active soil erosion in the and western escarpments to the banks of surrounding areas and irrigated fields. The Finchaa River. This is due to the steep slope surrounding steep escarpments with average down the escarpment to the valley floor. slope ranging from 5 to 65 percent create Obviously the high runoff in these areas favorable condition for erosion. There is also contributes for high rate of erosion. Thus the high rain fall intensity (90-120mm/hr) which is topographic set up and human activities make highly erosive. Rain fall intensity greater than soil erosion to be a critical problem in the 50mm/hr is believed to be erosive. In Finchaa study area. valley area human intervention exacerbates the In a nutshell one of the critical problems of soil problem of erosion, especially deforestation resource in the project area is erosion. The use and road construction. The large scale of agro-chemicals in the irrigation fields has deforestation exposed the soil for agents of also its own share to degrade the soil quality. erosion and contributes for high runoff. Due to Soil pollution emanating from chemical terrain inconvenience the roads have been pollutants is found to be moderate in the study constructed by dissecting hills and uplands area. This can be due to soluble nature of which facilitate the birth and intensification of chemicals that have been applied and the sheet, rill and gully erosions. Road quantity in proportion to the total sampled construction and the frequent maintenance also soils. But there is greater possibility of the play a vital role in aggravating the problem. agro-chemical use in the irrigation fields to The energetically operating sheet, rill and gully alter the soil quality in the long run. The Soil erosion mainly takes the fertile top soil and contributes for expansion of gullies and there by reduce the potentially irrigable lands. These two major soil problems needs proper follow up and management. If the present trend continues, in the long run the problem of soil/land degradation can put the sustainability of the project in question. Persistent and considerable efforts should be erosions around the escarpment donate the made to mitigate the impacts of erosion on the fertile top soil to the valley floor. This partly soil and potentially irrigable lands. fed fertile soil to the irrigation fields. On the other hand however, the high runoff from elevated ridges accelerates the formation and

410

2.2.4. Land use and Land cover In the TM image of 1986 some artificial developments have been observed (See Figure Finchaa valley area have transformed from 4.2).This is mainly because of the introduction primary to secondary and tertiary economic of the state farm to the valley and slight activities; from traditional agriculture to population growth in the surrounding high industrial and commercial activities. There is lands. From this time on wards agricultural land use land cover change in the area since lands have been expanding in the valley and 1975. the surrounding areas. it is evident that many of the area were under Savanna grasses, open wood land and dense forest. There were no significant land use classes in this year. The present Amerti Lake was in its swamp stage. In the upper right corner the area that appeared as lake is incorrect .Rather it is spike involved during Satellite image acquisition. In the land use classification of 1972 the built up area category is very small and insignificant and therefore not represented in the unsupervised classification. In this year very few agricultural plots are observed. In the ETM image of 2000, which is after 22 In the pre 1975 years there were no years, significant human made features are considerable land use classes in Finchaa valley evident in and around Finchaa valley. The vast area. Most of the areas were under the natural vegetation cover. From 1975-1991 the state farm used to produce some food and commercial crops on about 3,500 hectors of land in the valley floor. This incident attracted few people to the area to get jobs in the farm but still the number was not that much substantial. During this period there were no significant land use classes except for the state farm and few fragmented private holdings out side the valley. With the beginning of the sugar project in 1991 extensive land has been cleared irrigation fields and built up areas have and irrigated. Currently the irrigated land is increased. By implication the vegetation about 8,064.88 hectors and the built up areas biomass in these areas has diminished. (See fig occupy approximately 200 hectors of land in 4.3) the valley. The dominant land use classes are In the surrounding areas the rain fed irrigation agriculture; Rain fed agriculture, agricultural plots have intensified. Some built up areas, roads, artificial reservoir, lakes, smaller towns and villages are observed and others. including Finchaa town, Achane, Homi and The land cover of the study area can be Kombolcha villages. The size of Finchaa Lake categorized in to two classes. These are the is also increased compared to the pre 2000 natural and artificial land covers. The human years. However from the field observation it is made features in the area composed of towns, perceptible that in recent years the size of the roads, drainage canals, ponds, agricultural and lake is diminishing. In the irrigated fields, and artificial lakes. Only a few Figure 4.12. Supervised Land use/ land cover artificial structures are observed near Finchaa Map Of 2000 image. dam in the MSS image of 1972 following the unsupervised classification map agricultural construction of the dam. (See figure 4.1) In land, built up areas, bare lands and the size of these years approximately more than 95percent the lake have increased in size. The Amerti of the area was under natural environment. swamps grow in to a perennial lake. Conversely the total share of dense forest,

411

open wood land, savanna grass lands have This phenomenon affects the living organisms diminished. in water in particular and the environment in Generally there have been land use and land general. Again Finchaa River as one of the cover changes in the study area. This is mainly tributaries of Abay River crosses the due to favorable climate and environmental boundaries of Sudan and Egypt. Any water conditions which instigate agricultural and quality problem in this place arise dispute with industrial development activities in the area. these countries. Thus strict water quality This phenomenon was in turn followed by control works should be carried out in and population growth and intensification of around the industrial and agricultural sites. agriculture and industrial developments. The development of the irrigation scheme in the 2.2.3. The Physico-Chemical Properties of the project area facilitates the alteration of the Up Stream and Down Stream Water natural ecosystem and brought changes in the land use land cover of the study area. The physical and chemical properties of water characterize the water quality. These properties 2.2 Irrigation project and water Quality of water are susceptible for change. The and Health Conditions addition of toxic wastes to the surface or sub surface water alters the normal composition. 2.2.1. Impacts of Irrigation project on The PH for instance is sensitive and decisive Water Quality factor for the survival of living organisms in water.On experimental lakes in North West 2.2.2 General Conditions of water in Ontario Schindler (1988) find out that due to Finchaa Valley Area change of PH from 5.4 to 5.1 over all, the In the study area the Finchaa and Amerti- number of species in the lake at PH 5.1 was 30 Nashe rivers form the main drainage system. percent lower than in the pre acidification They both join the Abay River in the far down years. In order to asses the impacts of the stream area. The irrigation field and the agro-chemicals and industrial wastes, water Finchaa sugar factory lie with in the networks Samples from the up stream and down stream of Finchaa river system. They both relay on areas are taken and analyzed. (See Appendix this river to meet their water requirement. 5) Finchaa River is diverted to cane fields near The up stream area refers to the water near the the power house in the upstream area through power house where the water does not get in concrete canals. At present the west bank contact with water from irrigation and canals run for about 44 kilometers. Water from industrial wastes sites. The down stream the canal is pumped to irrigation fields and comprises the water after it mixes up with finally sprinklers shower the water to the water from streams in the irrigated and growing cane. The extra water from cane fields industrial waste sites. (See Figure 5.2 for flow to the near by ditches and join one of the sample sites) nearest tributary streams. In order to increase the accuracy of the results On the other hand the industrial waste water is water sample was not taken from the irrigation taken to the treatment plant which is situated to ditches and direct industrial waste water. the east of the factory. The factory uses a rock Rather the Mixed down stream water was filtration treatment method. However some of preferred so as to avoid inaccuracies and the instruments of the treatment are exaggerated results.The physical properties of nonfunctional. The waste water coming from water like pH, EC, Odor and color are found to the factory over flows due to these broken be different in the upstream and down stream parts and two stream-sized crude waste water areas. The PH and EC are lower in the flows to Finchaa River. These direct leakages upstream water compared to the down stream together with the agro-chemicals from water. This could be due to difference in the irrigation fields indisputably alter the physico- chemical constituents in these two sites. The chemical properties of the water. color differences are also discernable.

412

140 filtration and purification. As an alternative 120 approach the extra water from the irrigation 100 field can be collected in an artificial reservoir 80 60 Upstream and treated before it discharges to the main 40 rivers. 20 Dow n stream Generally, the water samples from Finchaa 0 River indicate that there is water quality -2 0 EC PH difference between the upstream and down Nitrite Nitrate Total stream area. This shows that, to a greater or Sodium C a lc iu m hardness Tem p(oc) Am m onia

P otassium lesser extent, there is water pollution in the Magnesium River. The discrepancy in the physico- The results of the water chemical analysis also chemical properties of water is supposed to be indicate that some elements are found in a from the industrial wastes, agro-chemicals higher quantity in the down stream than the from irrigation fields and to a lesser extent upstream water. (See appendix 5) metropolitan wastes forming the point and non As it is evident from Figure 5.3 almost all of point sources. the inspected elements are found to be higher Figure 5.4. 100m Buffer around the Finchaa in the down stream water. This could be due to and Nashe rivers and their major tributaries. two major reasons. In the first place the extra water washed the agro-chemicals from irrigated 2.2.4. Health Conditions in the Post Figure 5.3. The physical and chemical Development Years Of the Irrigation properties of upstream project and down Stream Water from Finchaa River. fields and joins the river. There is no recorded health status data before The Second main reason is the liquid waste the establishment of Finchaa sugar factory as from the industry and the urban areas that there was no settlement in the valley. directly or indirectly drains to surface or However, from unstructured interview made subsurface water. These two cases comprise with local inhabitants some people asserted the point and non point source for the that the area had been affected by epidemics pollution. The industrial waste water escaping even before the arrival of the state farm. The from the treatment plant forms the point source interviewee sited the problem as one of the pollution while the agro-chemicals from impeding factor for permanent settlement not irrigation field cover the non point source for to take place inside the valley in the former the pollution. years. In an effort to alleviate the problem of water Booker international agriculture Ltd (1977) in pollution the point and non point sources should the feasibility study for Finchaa sugar project be given priority. As a point source the affirmed that there exists Malaria and Tsetse contribution of waste Water from the industry fly in the valley. The company added that this can be addressed by continuous follow up and could be a challenge for the project workers maintenance of the treatment plant. It is again and residents of the valley. Currently there is advisable to replace the treatment plant with one health center in the valley and some modern and effective instruments and methods. recorded case information is available. And the problem of non point source can be Accordingly the intestinal and malaria cases mitigated by avoiding the direct contact of the have been increased from 1992 on wards. excess water from irrigation ditches and From the informal interview made with the metropolitan wastes with the river and tributary staff of the health center three possible streams. This can be possible by creating a rationales can explain this scenario. In the first buffer around Finchaa and Amerti-Nashe Rivers place the water used for drinking is pumped and major tributaries. (See figure 5.4) from the canal with diminutive treatment. The buffered zone needs to be covered with Secondly the expansion of irrigation can vegetation so as to enhance soil and plant litter facilitate the spread of malaria and access to unclean water. Finally the population explosion in a short period of time may inflate

413

the proportion of patients compared to the irrigation could be due to the time taken for early years. reproduction and stages of development in Malaria is the top health problem in the last 12 human body. years. The fluctuating Weather conditions In general there is no health data available on together with the expansion of irrigated fields the pre irrigation development years of the and ditches can be the factors behind the valley. Since 1992 malaria and intestinal problem. The fluctuation in malaria case parasite case records have increased. records arises from the inconsistent use of anti- Conversely the health facility given in the malaria chemical sprays and expansion of valley has improved a lot since 1992.It is irrigated lands. The Second and third top found out that there is a positive relation ship health threats are Guardia and Ascaries. These between malaria cases and expansion of intestinal problems in most cases are water irrigation fields. The intestinal health threats born disease which can be related with are also interlinked with unclean drinking unhygienic water use for drinking. The water. Well organized preventive and haphazard increment of intestinal parasite and controlling measures should be implemented malaria case records can be due to natural or as the health cases are interrelated with human made reasons. On one hand the natural workers productivity. set up of the valley and the climatic conditions can facilitate the birth and growth of 3. Conclusions pathogenic organisms in the area. On the other way round human interference have changed In order to see the possible environmental some of the existing natural systems. In other impacts of irrigation projects some parameters word the expansion of human made were selected. Some of the Geographic environments results in alteration and information system and remote sensing degradation of the natural ecosystem. techniques were also used. Accordingly, it is These environmental modifications create a observed that the natural environment in fertile ground for some insects and pathogenic Finchaa valley has been modified due to organisms which give birth to the spread of agricultural and industrial developments since diseases. The classical example here is the 1975. Following this modification the irrigation project have both positive as well as 4500 negative impacts on the environment. 4000 From the positive contributions the project 3500 3000 opened up large scale job opportunities for 2500 Maleria case many thousands of people. It has also many 2000 Irrigation field(ha) 1500 socio-economic benefits for the valley and 1000 500 surrounding people. In addition Finchaa Sugar 0 factory play a key role to address the current 1992 1993 1994 1995 1996 1997 1998 1999 2000 sugar demand in a local market. There are also many efforts to exploit the byproducts of the expansion of irrigation and increasing malaria factory for other extra purposes like using case records. ethanol for fuel. The project has also an Figure 5.9.Expansion of Irrigation fields Vs important role for the growth of national GDP Malaria case records. and GNP. As it is evident from figure 5.10 Malaria cases On the other hand the attention given for are increasing with the expansion of irrigation natural resource conservation is less and this fields. In the beginning few malaria cases were has been devastating some of the seen in the valley. For the 1992 and 1993 there environmental components. There has been were no recorded malaria cases data available large scale vegetation clearance taking place in in the health center. But for the consecutive the study area. The NDVI image analysis of two years fewer malaria cases were recorded. the 1972 MSS and the 2000 ETM images From 1995-1997 large number of malaria shows that the vegetation biomass is cases were observed. (See appendix 3) In these diminishing. The intensification of agricultural years extensive sugar cane plantation was and industrial developments together with carried out in the valley. The lag time between population explosion has the coin share for the highest malaria case and the expansion of decrement. The large scale deforestation has

414

been devastating the vegetation and wild life join the tributary streams. Thus, based on the resources in particular and the biodiversity in water samples inspected there is water general. There are efforts made by Finchaa pollution emanating from poor industrial waste sugar factory to rehabilitate the forest resource. water treatment and the leftovers of the agro- But the amount and rate of deforestation in one chemicals used in the irrigation fields.In most side and the reforestation and afforestation cases climate is the reflection of the natural projects on the other side are incomparable in environment. Any system disturbance on the any measure. environment can affect the climatic conditions. In addition deforestation is facilitating the The analysis of the 22 years rain fall, progress of runoff and accelerates erosion. temperature and humidity data can not Accordingly soil erosion is a critical problem meaningfully imply any climatic change as a in the project area. The active erosion, beyond result of the irrigation project. This is due to taking the fertile top soil, is changing some of the sluggish and unpredictable nature of the potentially irrigable lands in to Bad Lands. climatic anomalies. It is reasonable however to In some areas there are gullies that extend for say that there is imperceptible changes about 30 m. The topographic set up and following the environmental degradation. Still Human induced factors are responsible for the it is open for further specific and detailed active erosion in the area. The steep slope in works to see the impacts of irrigation on the the escarpments surrounding the valley local climate. Case records of Malaria and promotes greater runoff. The Road some water born disease have been increasing construction and the frequent maintenance following the opening of Finchaa irrigation down the valley made the soil ready for scheme. There is a positive correlation erosion. between malaria case records and expansion of The physical and chemical analysis of the soil irrigation fields. Although the health care samples taken from irrigated field, cultivated facility given has improved a lot, the number but not yet planted and Vegetation cover area of patients boost up by a large number. This are found to be different. The total sand shows that big attention is given on disease content of Luvisols decreases with increasing control than prevention. The environmental soil depth in all sampled layers. While the clay modifications and the diminutive prevention content increases with depth. The measures contribute for the large number of exchangeable bases are higher in the irrigated malaria and water born diseases case and cultivated area than the vegetation cover records.Generally despite of its positive area. Relatively Organic carbon, Nitrogen and consequences, the irrigation project in Finchaa phosphorous are found in large quantities in valley area has a negative impact on the the irrigation fields especially in the upper environmental components. Especially on the layer (0-30cm). The use of agro-chemicals in vegetation cover, soil quality, water quality the irrigation fields are supposed to contributes and partly on some health conditions. But this for this result. In general some of the physical does not, in any way, mean that the problems and chemical properties of the soil in the out weight the benefit of the factory and that irrigated and non irrigated sites are found to be the problems are out of control. The different. This shows that, to a lesser or greater degradation is in its early to moderate stage extent, there is soil contamination that could and even not difficult to address and alleviate lead to full-size soil pollution. The result of them all. The possible solutions are much water samples from up stream and down easier and cheaper in this moderate stage of the stream areas indicates that the physico- environmental degradation. But undoubtedly if chemical properties are different in these two the current trend keeps on the problem would areas. The down stream water contains more get more complex and difficult to reclaim. chemical substances than the up stream water. Thus, urgent attention should be given for the The point and non point sources contribute for environmental rehabilitation and conservation. the pollution. The point source comprises the industrial waste water that escapes from the 4. Recommendations treatment plant and join the river. The non point sources involve the use of agro- In line with the findings of this selective chemicals (fertilizers, pesticides and parameter based environmental impact herbicides) and the metropolitan wastes that

415

assessment the following recommendations are 7. In addition to disease controlling presented: strategies, research based preventive approaches should be adopted so as to 1. The environmental considerations mitigate the escalating malaria and should not be disregarded in any way water born disease case records and and with any justification seeing that their far reaching impact on production well organized environmental and productivity. management positively contribute for better productivity and sustainability. References

2. The rehabilitation of devastated African development fund. (2002). The vegetation biomass should be given Federal Democratic Republic of Ethiopia. first line attention as it helps to Agricultural review. Rural development maintain the soil, water, climate and department, Onar. biodiversity of the area. Finchaa sugar Agrawal.M.L. and A.K. Dikshit (2002). factory should tackle the problem of Significance of Spatial Data and GIS for frequent Forest fire and large scale Environmental Impact Assessment of deforestation that are observed inside highway projects. India. the valley. Ahmed A. (2002). A Geographical study on Finchaa sugar Factory. A senior essay 3. There should be well organized and .Haromaya University. effective afforestation and Alan Gilpin (1995).Environmental Impact reforestation programs to reestablish Assessment. Cutting Edge for the Twenty- the ecosystem. Some areas like the first century. Cambridge university press. surrounding escarpments, river sides, London. agriculturally non suitable and Anuctech private. Ltd (1994), Finchaa sugar marginal lands can be delineated and project proposal for the Finchaa Valley protected as the forest area. Forestry project, Finchaa sugar project office, 4. There should be continuous follow ups Addis Ababa: 61pp. and assessment of the physico- Atesmachew B., Yasin G., Girma T, Don chemical properties of the soil in the Peden. (2000). GIS Application for irrigation fields. This helps to see the Analysis of Land suitability and impacts of fertilizers, herbicides and Determination of grazing pressure in up pesticides on the soil quality and to land of the Awash river basin. ILRI, Addis take timely measures. Ababa Bandaragoda.D.J. (1998).Need for Institutional 5. Strict physical and biological measures Impact Assessment in planning Irrigation should be taken to impede the actively system Modernization. Delhi, India. operating erosion and growing gully Barrow.C.J. (2000). Environmental Impact problem in the irrigation fields and the assessment of developments. Longman surrounding areas. printing press. USA. Birhanu.M. (2005). Remote Sensing and GIS 6. Finchaa sugar factory should establish techniques in land use/land cover mapping, a modern and efficient waste water change detection and evaluation of treatment plant in order to stop the two environmental degradation in the main stream-sized industrial waste water Ethiopian rift (MER) between Koka and and irrigation field wash away Ziway. leakages to the river. Further Booker Agricultural International Ltd. inspections should be carried out for (1977).Finchaa valley Feasibility study the water quality problem in the down main Report.vol.1.London.12-15pp. stream area and appropriate measures should be taken. Chris Parker. (1998).GIS and Environmental Management in the Sultanate of Oman. Turkey.

416

Darout Gum’a. (2004). The socio-economic Schiondler D.W. (1990).Effects of acid rain on Aspects of irrigation management. Addis fresh water ecosystem Ontario. Canada Ababa University. Addis Ababa. 25pp. Dereje.A. (2005). Assessment of Socio- Shawinigan Company. (1981). Finchaa Sugar economic Impacts of Irrigation in Finchaa Project Design Report. Ethiopian Sugar valley, Addis Ababa University, Addis Corporation. Addis Ababa.9-11pp. Ababa. Smajstrla.A.G, Zazueat.F.S.and Haman D.Z. Donalds Ahrens (2005). Microsoft Encarta (2002). Potential Impacts of Improper Reference Library2005 .1993- Irrigation system Design. Florida 2004.Microsoft Corporation. University. Florida. FAO (2001). Environmental Impact Smol J.P. (2002). Pollution of lakes and rivers. Assessment report. Retrieved from A paleo environmental perspective. Arnold http//www. publishers. Network and Oxford Gideon. A. (2005).Environmental degradation. University press. London 126pp. Forum for social study. Addis Ababa. Sophie N. Laurence S. and Kai Sorensen Girma A. (2005). Evaluation of Irrigation (2003).Adaptive, participatory and Water Quality in Ethiopian sugar Estates. integrated assessment (APIA) of the Research report. 23-34pp. impacts of irrigation on Fisheries. Girma.T. (1995).Finchaa Sugar Project Soil Evaluation of the approach in Sri Lanka. Survey/west Bank. Land Management unit Final report.Vol. I. Addis Ababa. Stockle.C.O. (2002). Environmental Impact of Glen Paleto (2003). Environmental Impact Irrigation: A review water Research Assessment (EIA).Lecture Notes. Washington state Haklay.M, Feitelson E. and Doytsher.Y. university.Washington.121-129pp. (1998). The Potential of a GIS based Vicharn A. and Sirirat S. (2001). GIS Scoping Application on the Strategy for Sustainable System. An Israeli proposal and case study. Development in Phitsanulok Environmental Impact Assessment Province.Naresuan Unive4rsity. review.Israel.439-459pp. Phitsanulok. Thailand. Hartza Engineering Company (1965). Worku B. (1995).Agro ecological conditions Appraisal of Potential Agricultural and background information on Finchaa Development of Finchaa project. Final sugar project/Factory site. Finchaa report. Addis Ababa. valley.6-18pp. 43-48pp. Zeleke K. (2005).GIS and remote sensing in Mintesinot B., Mohammed A., Atinkut M.and Land use/Land cover change detection in Mustafa Y. (2004). Community Based Finchaa valley area. East wollega.Addis Irrigation Management in the Tekeze Ababa University. Addis Ababa.36-44p Basin: Impact Assessment. A case study on three small-scale irrigation schemes (micro dams) .Mekele University, ILRI, and EARO. Morgan M.K. (1998).Environmental impact Assessment. A Methodological Perspective.Kluwer Academic publisher. Dordrecht. Painnly J.P. (1994). Environmental impact Assessment: A case study on the irrigation project. Gujarat. India.37-39pp. Paul Engelking (2005). Micro soft Encarta Reference Library 2005. 1993-2004. Microsoft Corporation. Rosenqvst.I.T. (1978).Alternative sources of acidification of river water in Norway. The science of the total environment.Oslo.39- 49pp.

417

Entomological studies on the impact of a small-scale irrigation scheme on malaria transmission around Zeway, Ethiopia.

Solomon Kibret1, Beyene Petros1, Eline Boelee2 and Habte Tekie1

1 Department of Biology, Addis Ababa University, Ethiopia. 2 International Water Management Institute, Ethiopia. e-mail: [email protected]

Abstract malaria transmission might possibly extend from seasonal to year-round, involving the To evaluate the impact of a small-scale dry season. Proper water management irrigation scheme on the level of malaria coupled with environmental management transmission in a semi-arid area, such as source reduction could reduce vector entomological studies were conducted in abundance and hence malaria transmission Zeway area, Central Ethiopia. Larval and in the irrigation schemes. adult anophelines were sampled during the 6. dry and short-rainy seasons from irrigated 7. Key words: Anopheles, malaria and non-irrigated villages. Overall, transmission, small-scale irrigation scheme, significantly higher density of Anopheles Plasmodium falciparum sporozoite rate, larvae were found during the dry season in Zeway, Ethiopia. the irrigated village (Mean = 38.3 larvae/100 dips) than the non-irrigated village (7.4 1. Introduction larvae/100 dips). Canal leakage pools, irrigated fields and irrigation canals were the Development of irrigation schemes is widely major sources of Anopheles mosquitoes. recognized as a key for promoting economic Larval and adult Anopheles pharoensis and growth, ensuring food security and An. arabiensis, principal malaria vectors in alleviating poverty in most developing Ethiopia, were more abundant in the countries (Lipton et al., 2003). However, irrigated village than the non-irrigated past experience shows that inadequate village throughout the study period. Hourly consideration of both environmental and light trap catches revealed that peak indoor public health impacts can seriously and outdoor biting activities of An. undermine the sustainability of such arabiensis and An. pharoensis occurred schemes (Gratz, 1988; McCartney et al., during the early period of the night before 2007). Key among the potential negative the local inhabitants retire to bed. The impacts is the link between irrigation and majority of blood-engorged An. arabiensis malaria – a disease that affects between 300 (0.78) and An. pharoensis (0.69) had fed on and 500 million people each year globally humans, suggesting that their highly and claims the lives of 1.5 to 2.5 million anthropophilic nature in Zeway area. people annually (WHO, 2006). Plasmodium falciparum infection rates of 1.02% and 0.54% were determined for An. By increasing the availability of surface arabiensis and An. pharoensis, respectively, water for breeding, irrigation favors the in the irrigated village. This study development of large populations of disease demonstrated that due to poorly maintained vectors such as anopheline mosquitoes irrigation structures, the irrigation scheme responsible of transmission of malaria. created conducive breeding grounds for Hence, there is great concern that irrigation malaria vector species, particularly during can lead to increased malaria transmission the dry season. Consequently, the period of especially in sub-Saharan Africa where 90%

418

of the global malaria burden exists and the communities near irrigation microdams than prevailing climatic factors support those further away (Ghebreyesus et al., proliferation of malaria vector mosquitoes 1999). A recent entomological study in the and development of the parasite in the same area has reported 5.9-7.2 times more vector. However, the relationship between adult An. arabiensis (the main malaria irrigation and malaria is not straightforward vector in Ethiopia) in the dam villages than and varies according to endemicity and the controls, non-irrigated villages seasonality. In stable malaria endemic areas (Yohannes et al., 2005). The study also of sub-Saharan Africa, studies have shown indicated that seepage water at the base of that malaria transmission is equal or less in the dam, leaking irrigation canals and irrigated-rice growing areas compared with waterlogged fields were the main sources of neighboring areas without irrigated rice An. arabiensis throughout most of the year. cultivation (Josse et al., 1987; Lindsay et al., However, despite extensive development of 1991; Boudin et al., 1992; Faye et al., 1993; irrigation schemes in semiarid fertile areas Henry et al., 2003). The explanation for this of the country with unstable disease finding is yet unresolved, but in some cases transmission (MoWR, 2005), in-depth at least, could be attributed to displacement information on the link between irrigation of the most anthropophilic (human blood and malaria in such environmental settings seeking) malaria vector Anopheles funestus is lacking. The main objective of this study by An. arabiensis with lower vectorial was to assess the possible impact of capacity, as the later thrives more than the irrigation-based agricultural activities on former in irrigated fields (Ijumba and malaria transmission in a semi-arid area with Lindsay, 2001). It has also been suggested seasonal disease transmission. that many communities near irrigation schemes benefit from the greater wealth 2. Materials and Methods created by the schemes, often leading to better access to improved health care and The study area hence receive fewer infective bites compared to those outside such schemes. On The study was undertaken between February the other hand, in areas where malaria is and May 2006, in two rural farming villages, absent or unstable, introduction of irrigation Abene-Girmamo and Woshgulla, located in was found to place the non-immune Zeway area (8o00`N, 38o40`E), Central population at a high risk of acquiring the Ethiopia, 165 Km south of Addis Ababa, in disease, increasing malaria morbidity and the middle course of the Ethiopian Rift mortality (El Gaddal et al., 1985; Ijumba et Valley (Figure 1). Both study villages are at al., 1990). In such areas, irrigation, a distance of 5-6 Km from Zeway town, especially during the dry season, might alter which is situated alongside Lake Zeway. the malaria transmission pattern from The area receives between 700-800 mm of seasonal to annual, as observed in the annual rainfall, with the heavy rains during Sahelian region of Mali (Sissoko et al., the months of June to September and short 2004) and in sub-arid irrigated areas of rains in April and May (National Madagascar (Marrama et al., 2004). Meteorological Agency). The mean annual temperature is 20 0C, and February is the In Ethiopia, where three quarters of its land hottest month of the year. mass are potentially malarious, introduction or expansion of irrigation schemes can Malaria transmission in Zeway area is increase the burden of malaria in the generally unstable (seasonal), with peak country. A detailed epidemiological study in transmission occurring between the months the highlands of Tigray, northern Ethiopia, of September and November, immediately has reported that malaria incidence in young after the main rainy season, while the second children was sevenfold higher in less pronounced transmission period falls

419

between April and May in the short-rainy water, were commonly found at the season. Plasmodium falciparum is the most backyards of households that commonly prevalent malaria parasite in Zeway area, practice brick-making either for domestic responsible for 60-70% of malaria cases. use or for sale. These pits were mostly Vivax malaria is also common in the area, functional during the dry season but became particularly in the dry season, but generally non-functional in the wet seasons, because less prevalent. (Abose et al., 1998b; Zeway the rains could damage newly formed moist Malaria Control Unit, unpublished report). mud-bricks before they dry. Each village Anopheles arabiensis is the primary malaria had a water-harvesting pool, i.e., collection vector in Zeway area, and elsewhere in of rainwater in a wide and deep well Ethiopia, while An. pharoensis plays (volume ~ 2m width x 2m length x 6m secondary role (Rishikesh, 1966; Abose et depth) with corrugated iron-roofing. al., 1998a; Abose et al., 1998b; Ye-Ebiyo et al., 2000). The source of water for irrigation in Abene- Girmamo is Lake Zeway, located 5-6 Km Abene-Girmamo is an irrigated rural village, away from the scheme. Water is pumped situated at an altitude of 1647 m. The village from the lake by three long plastic pipes (0.4 is inhabited by 934 people, mainly m diameter and 4-5 km long) that run dependent on subsistence farming. The underneath the ground to reach the unlined community is mainly comprised of the surface canals at uplifted soil mass. The Oromo and Silte ethnic groups. Most surface irrigation canals feed smaller field families own livestock (mainly bovine, canals to cover the entire agricultural field. ovine and equine), with a human to cattle However, due to poor construction and lack ratio of 1: 0.4. Woshgulla is a non-irrigated of maintenance, there were many leaking agricultural village, situated at an altitude of canals, causing leakage pools at unwanted 1654 m, with a population size of 741. The places. These pools never dry because of village is located 8 Km away from the continuous water leakage from the irrigation irrigation schemes in Abene-Girmamo. The canals. Water logging also occurred in the inhabitants are dependent on subsistent rain- agricultural field as a result of over- fed agriculture during the months of the wet irrigation and water retaining characteristics seasons. They also keep livestock (mainly of the soil. Sometimes, poor drainage led to cattle, equine and ovine), with the mean water logging in the field. The uplifted soil human to cattle ratio of 1: 0.6. The domestic walls of the surface canals were also animals in both villages spend the night frequently perforated and formed leakages, either indoors in the same homesteads with mainly due to the action of domestic animals the owners or outdoors in open cattle while drinking water in the canals. Onion, enclosures. The main type of housing in cabbage and maize (Zea m. mays) were these villages was circular huts, made of commonly grown under irrigation mud-brick walls and thatched roof. Mud- throughout most of the year. brick-making pits, partly covered with

420

Figure 1. Location of the study area in Ethiopia and map of Zeway area (the two study villages are shown in bold rectangles). surveyed using standard dippers (350ml). Entomological surveys The surface area of each potential mosquito

Entomological surveys comprised larval and breeding site was estimated in square meter adult mosquito collections in the irrigated (m2) and sampling was made at a rate of 6 and non-irrigated study villages during the dips/m2. One ‘sample’ was defined as 30 dry (February/March) and short-rainy dips (or less, in smaller sites) taken over a (April/May) seasons of 2006. surface area of 5 m2. For sites in the range of 5-10 m2, one sample was taken, whereas two Anopheline larvae were sampled for eight samples were taken from sites in the range days between February and March in the dry of 11-20 m2 and so forth. An upper limit of season and also between April and May in six samples was set for all sites with water the short-rainy season of 2006. At each surface area exceeding 50 m2 (Amerasinghe survey, all available potential mosquito and Munasingha, 1988). Larval anophelines breeding habitats such as irrigation canals sampled from each type of breeding habitat (unlined surface canals with still water due were transferred to separate vials and killed to back-flow), canal leakage pools (pools by gently heating and preserved in 70% formed from leaking main canals), irrigated alcohol for later species identification. field paddies (water logging in the field due to over-irrigation and poor drainage canals), Adult anophelines were sampled from water-harvesting pools, mud-brick-making indoors and outdoors for ten consecutive pits and rain pools within one kilometer nights between February and March in the radius from each study village were dry season and between April and May in the short-rainy season of 2006. Three

421

techniques: i.e., CDC light traps, indoors and outdoors (Abose et al., 1998b). mouthpiece aspirators and space-spray, were Using white sheets of cloth and an aerosol of used for adult sampling. In order to obtain a pyrethroids (Mobil flit; Mobil Africa Sales representative sample of the mosquito Inc., Belgium; Composition [% weight]: population for the irrigated village, a total of Tetramethrin 0.12; Phenothrin 0.12; 6 houses were randomly selected (2 from the Allethrin 0.25; Solvents, Propellants and edge of the village at close proximity to the essential oils 99.43), indoor-resting irrigated field [~300 m], 2 from the middle anophelines were also collected from six [~600 m] and 2 from the far side of the selected houses in each village. Before village [~800m]). Similarly, the non- spraying the houses, all openings that could irrigated village was roughly divided into allow mosquito escaping (such as doors, three sampling zones based on proximity to windows, and holes on the walls) were the non-irrigated agricultural field and two closed, and the entire floor was covered with houses were randomly selected from each the white cloth. The houses were then zone. Untreated bed nets were distributed sprayed with Mobil flit for about 5 minutes for the households and the same houses and left closed for 10 minutes. Thereafter, were used throughout the study period. A the sheets were brought outside the rooms to total of six outdoor sites (~100 m away from inspect and collect the knock-down occupied houses) was also selected in each mosquitoes. Mosquitoes collected by the village for outdoor light trap collection. different techniques were counted and kept in separate paper cups for latter A total of twelve (six indoors and six identification and mosquito processing. outdoors) CDC light traps (Model 512; J. W. Hock Co., Atlanta, USA) was operated in Species identification and dissection each village from 1800 to 0700 hours throughout each sampling night. Each At Zeway Malaria Control Laboratory, indoor light trap was hung on a wall; with preserved anopheline larval samples were the bulb about 45 cm above the head of a counted and individually mounted on person sleeping under an untreated bed net microscope slides for species identification (Lines et al., 1991). Outdoor light traps were based on morphological characteristics hung on trees at close proximity (~50 to 100 (Verrone, 1962b). Only third and fourth m) to open cattle enclosures where some larval instars were used for species individuals spend the evening keeping their identification of anopheline larvae. Adult livestock from theft. To determine peak anophelines collected by the different activity of anophelines during the period of sampling methods were also sorted out into the night, hourly mosquito collections were species based on morphological also conducted indoors and outdoors by light characteristics (Verron, 1962a). One-third of traps. Using mouthpiece aspirators and unfed female Anopheles mosquito space-spray, female Anopheles mosquitoes collections obtained from light trap catches were collected from their daytime resting and all unfed female anophelines caught sites both indoors and outdoors (Service, resting indoors and outdoors were dissected 1993). While collecting light traps the to determine parity rates for each species following morning, a team of three during the dry and short-rainy seasons. collectors holding aspirators and torchlight Ovaries with coiled tracheal skeins were searched for resting mosquitoes from the six considered as nulliparous (did not lay eggs), light trap houses and from possible outdoor- while those with stretched out tracheoles resting sites (burrow pits, ground holes, tree were taken as parous (laid eggs) as holes, open cattle sheds and among described by Lewis (1958). All the vegetations) in each village. For the purpose remaining female anopheline samples and of comparison, the team of collectors spent the head-thorax region of dissected the same period of time (20 minutes) mosquitoes were stored in the silica-gel

422

dessicator and transported to Addis Ababa 3. Results University, Biomedical Science Laboratory, and kept at room temperature (19-22 oC) for Larval habitats and abundance later mosquito processing. Total number of positive larval habitats, Mosquito processing number of Anopheles larvae collected and larval density in the irrigated and non- The head-thorax region of each dried female irrigated study villages during the two anopheline was tested for the presence of P. sampling seasons are presented in Table 1. falciparum and P. vivax sporozoite antigens Four-times more positive Anopheles larval using Enzyme-Linked Immunosorbent sites were encountered in the irrigated Assay (ELISA) (Wirtz et al., 1987). The village (n = 51) compared to the non- direct ELISA procedure described by Beier irrigated village (n = 12) during the study and colleagues (Beier et al, 1988) was used period. Consequently, higher Anopheles to determine the sources of blood meals larval densities were found in the irrigated (human vs. bovine) of the blood-engorged village (mean no. larvae per 100 dips = 36.0; female anophelines. 95% CI = 25.4–48.5; z = -3.196, P < 0.001) than the non-irrigated village (mean no. Data analysis larvae per 100 dips = 14.9; 95%CI = 9.1– 20.8) throughout the study period. The Daily larval and adult mosquito collections difference in Anopheles larval abundance were entered into Microsoft Excel Database and positive larval sites between the dry and and log-transformed (log10 [n+1]), and tested short-rainy seasons was significant in the for normality before analysis. The non-irrigated village whereas there was abundance of larval and adult anophelines insignificant seasonal difference in the was compared between villages and seasons irrigated village. Overall, Anopheles larval using nonparametric Mann-Whitney U-test. production in the non-irrigated village was The same test was applied to compare the associated with the wet seasons while high indoor and outdoor density of adult larval production in the irrigated village was anophelines. The relative abundance of evident both in the dry and wet seasons. Anopheles species in the larval and adult collections was compared using Kruskal- Anopheles larval collections were composed Wallis Test, a non-parametric test for of five species, among which Anopheles ascertaining significance among more than arabiensis, An. pharoensis and An. coustani two variables. Larval density was expressed were the major species. The distribution of as the mean number of anopheline larvae per Anopheles species in different larval habitats 100 dips. Sporozoite infection rate of each in the irrigated and non-irrigated villages is Anopheles species was expressed as the shown in Table 2. Among the five types of proportion of mosquitoes containing malaria larval habitats in the irrigated village, canal sporozoite antigen from the total samples of leakage pools and irrigated field puddles a species tested by ELISA. The Human were the most important sources of An. Blood Index (HBI) for each Anopheles arabiensis, accounting for nearly 60% of the species was calculated as the proportion of larval collection during the study period. For samples positive for human blood from the An. pharoensis, canal leakage pools and total blood meals of a particular species irrigation canals were the major larval tested. The level of significance was habitats as more than 90% of larval determined at 0.05. All analyses were done collection of this species were obtained from using Microsoft Excel 2003 and statistical these habitats. In the non-irrigated village, software, SPSS version 13 (SPSS Inc, brick-making pits and rain pools were the Chicago, IL, USA). most important Anopheles larval habitats. Overall, around 80% of the total Anopheles

423

larval production in the irrigated village was Anopheles mosquitoes were generally very from three types larval habitats (irrigated low in the non-irrigated village during the field puddles, canal leakage pools and study period, similar indoor-outdoor trends irrigation canals) associated with the were noted for the three species. Overall, irrigation scheme. An. arabiensis was more endophagic while An. pharoensis and An. coustani were more Adult anopheline collections exophagic in the study area.

A total of 1271 adult anophelines was Peak hourly activity of Anopheles species collected from the two study villages during the study period, of which 94% (n = 1213) Peak indoor and outdoor activities of An. and 6% (n = 58) were from the irrigated and arabiensis were observed during the early non-irrigated villages, respectively (Table period of the night, between 18:00-19:00 3). Anopheles pharoensis was the major and 19:00-20:00 hours, respectively species predominantly sampled in the (Figure 2). Thereafter, its activity steadily irrigated village during the dry season decreased both indoors and outdoors (56.9%; n = 340; X2 = 52.294; df = 2; P < throughout the rest of the night. Peak 0.001) while An. arabiensis predominated in indoor and outdoor activities of An. short-rainy season (50.2%; n = 309; X2 pharoensis occurred between 20:00-21:00 =17.751, df = 2, P < 0.001). Of the few adult and 19:00-20:00 hours, respectively, anophelines collected in the non-irrigated which gently declined thereafter, but with village during the short-rainy season, the a remarkable increase between 22:00- majority (65.5%, n = 38) were An. 23:00 hours at outdoors (figure 3). For An. arabiensis. No mosquito was collected in coustani, its peak indoor and outdoor the non-irrigated village during the dry activities were recorded between 18:00- season. 19:00 hours, which sharply dropped thereafter but with a remarkable peak The density of An. pharoensis per light trap- between 22:00-23:00 and 05:00-06:00 night was higher during the dry season hours, indoors and outdoors, respectively (mean no. mosquito/trap/night = 2.24; 95% (Figure 4). Overall, about 75%, 66%, and CI = 1.21–3.17; Mann-Whitney U = 2422.0, 69% of the biting by An. arabiensis, An. z = -5.244, P < 0.001) than the short-rainy pharoensis and An. coustani occurred season (mean no. mosquito/trap/night = during the early period of the night (before 1.48). The difference between indoor and 22:00 hours), before the local people retire outdoor densities was significant, being to bed. higher outdoors (Mann-Whitney U = 4646.0, z = -4.257, P < 0.001) than indoors. Parous rate In contrast, the mean density of An. arabiensis was higher during the short-rainy Parous rate of Anopheles species in the season (mean no. mosquito/trap/night = irrigated and non-irrigated study villages 1.70; Mann-Whitney U = 1840.0, z = - during the dry and short-rainy seasons is 2.569, P = 0.01) than the dry season (mean presented in Table 5. In the irrigated no. mosquito/trap/night = 1.23; 95%). There village, the parous rate of An. arabiensis was also significant difference between the was higher during the dry season (58.7%, indoor and outdoor densities, being higher n = 46) than the short-rainy season indoors (Mann-Whitney U = 3849.5, z = - (22.2%, n = 81). In contrast, the parous 5.849, P < 0.001) than outdoors. The density rate of An. pharoensis did not vary An. coustani was higher outdoors (Mann- significantly between the two sampling Whitney U = 5772.5, z = -2.342, P = 0.001) seasons; 43.3% (n = 39) and 41.2% than indoors during the two sampling (21/51) in the dry and short-rainy season, seasons. Although the densities of respectively. For An. coustani, higher

424

parous rate was recorded during the dry result for bovine blood. The overall HBI season (18.2%, 4/22) than the short-rainy for An. coustani was 0.06. The ELISA season (3.4%, 1/29). In the control village, results showed that An. arabiensis and An. among few anophelines caught during the pharoensis are the most important short-rainy season, only two parous An. anthropophagic species in Zeway area. arabiensis females (5.3%; 2/38) were found. Overall, An. arabiensis and An. Sporozoite rate pharoensis had higher parous rate during the dry season, suggesting higher The P. falciparum sporozoite rates of longevity of these species in this season. Anopheles species in the irrigated village is presented in Table 7. None of the Host feeding preference samples tested were positive for P. vivax sporozoites. Among 424 female An. Table 6 shows the sources of mosquito arabiensis specimens collected from the blood meals in the irrigated village. irrigated village and tested for P. Among 120 blood-fed An. arabiensis falciparum sporozoites, 5 (1.18%) were specimens tested, 70.8% (n = 85) and found to be positive. None of the thirty- 14.2% (n = 17) were positive for only one An. arabiensis specimens caught in human and bovine bloods, respectively. the non-irrigated village were positive for Some (7.5%, n = 9) were mixed blood P. falciparum. Among the total of 509 An. meals originated from human and bovine, pharoensis collected from the irrigated and the remaining were unidentified village, three (0.59%) were tested positive (7.5%, n = 9) - possibly originated from for P. falciparum sporozoites. None of the other domestic hosts (e.g. equines and four An. pharoensis and sixteen An. ovines). Overall, the Human Blood Index coustani specimens collected in the non- (HBI) for An. arabiensis was found to be irrigated village was positive for malaria 0.78. Out of 142 blood-engorged female sporozoites. Seasonally, higher P. An. pharoensis specimens tested, 61.3% (n falciparum sporozoite rate of An. = 87) and 20.4% (n = 29) had human and arabiensis was recorded in the short-rainy bovine blood meals, respectively. Some season (1.47%; 4/272) than the dry season blood meals (7.7%, n = 11) were (0.66%; 1/152). The P. falciparum composed of both human and bovine sporozoite rate of An. pharoensis was bloods, and 10.6% (n = 15) of An. 0.92% (3/325) in dry season, while none pharoensis blood meals were not (0/184) were positive in the short-rainy identified. Overall, the HBI for An. season. Overall, the P. falciparum pharoensis was found to be 0.69. From a sporozoite rate of An. arabiensis and An. total of 16 blood-engorged An. coustani pharoensis suggests the potential of these specimens, only one specimen (6.2%) was species in malaria transmission in the positive for human blood while the irrigated study village during the dry and majority (75%, n = 12) gave positive the short-rainy seasons.

425

Table 1. Total number of positive larval habitats, number of Anopheles larvae collected and larval density (mean no. larvae/100 dips) in irrigated (Abene-Girmamo) and non-irrigated (Woshgulla) villages in Zeway area, Central Ethiopia, during the dry (February/March) and short-rainy (April/May) seasons of 2006.

Irrigated village Non-irrigated village Total no. positive No. of Anopheles Larval densityδ Total no. positive No. of Anopheles Larval density Season larval habitats larvae collected (%) (95%CI) larval habitats larvae collected (95%CI)

Dry 38 797 (46.0) 38.3 (26.2–50.5)* 5 69 (22.8) 7.4 (4.4 –10.5) Short-rainy 33 936 (54.0) 34.9 (24..9–45.9)* 11 233 (77.2) 15.2 (9.3–21.1) Overall 51a 1733 (100) 36.0 (25.4–48.5)* 12a 302 (100) 14.9 (9.1– 20.8)

δ Larval density refers to mean number of Anopheles larvae per 100 dips. 95% confidence interval is shown in brackets. * The difference in seasonal larval density between the two villages was significant (P < 0.001). a The overall total number of larval habitats in each village is not equal to the sum of positive larval habitats in the two sampling seasons since the same larval habitat can occur in both seasons.

426

Table 2. Distribution of Anopheles species* in different types of larval habitats in irrigated (Abene-Girmamo) and non-irrigated (Woshgulla) villages in Zeway area, Central Ethiopia, between February and May 2006. * Only third and fourth larval instars were sorted out into species

Village Larval habitats An. arabiensis (%) An. pharoensis (%) An. coustani (%) An. cinereu An. squamosus Total (%) (Total no. of positive sites)

Irrigated Brick-making pits (5) 70 (18.5) 2 (0.4) 72 (21.5) 0 0 144 (12.3) Canal leakage pools (12) 108 (28.5) 242 (52.8) 57 (17.1) 0 0 407 (34.6) Irrigated field puddles (23) 118 (31.1) 41 (9.0) 66 (19.8) 0 0 225 (19.1) Irrigation canals (4) 45 (11.9) 173 (37.8) 85 (25.4) 0 1 304 (25.9) Rain pools (7) 38 (10.0) 0 (0.0) 54 (16.2) 3 0 95 (8.1) Total (51) (%) 379 (32.2) 458 (39.0) 334 (28.4) 3 (0.3) 1 (0.1) 1175 (100) Non-irrigated Brick-making pits (7) 57 (52.8) 8 (88.9) 14 (45.2) - - 79 (53.4) Rain pools (4) 51 (47.2) 0 (0.0) 17 (54.8) - - 68 (45.9) Water harvesting pools (1) 0 (0.0) 1 (1.1) 0 (0.0) - - 1 (0.7) Total (12) (%) 108 (73.0) 9 (6.1) 31 (20.9) - - 148 (100)

427

Table 3. Number of adult Anopheles mosquitoes collected from irrigated (Abene-Girmamo) and non-irrigated (Woshgulla) villages in Zeway area, using different sampling methods, during the dry (February/March) and short-rainy (May/April) seasons of 2006.

Village Season An. arabiensis An. pharoensis An. coustani Total (%)

Irrigated Dry (%) 182 (30.4) 340 (56.9) 76 (12.7) 598 (49.3) Short-rainy (%) 309 (50.2) 212 (34.5) 94 (15.3) 615 (50.7) Total (%) 491 (40.5) 552 (45.5) 170 (14.0) 1213 (100) Non-irrigated Dry (%) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) Short-rainy (%) 38 (65.5) 4 (6.9) 16 (27.6) 58 (100) Total (%) 38 (65.5) 4 (6.9) 16 (27.6) 58 (100) Grand Total (%) 529 (41.6) 556 (43.8) 186 (14.6 1271 (100)

428

Table 4. Indoor and outdoor density of Anopheles species (mean no. mosquitoes /light trap/ night) in irrigated (Abene-Girmamo) and non-irrigated (Woshgulla) villages in Zeway area, during the dry (February/March) and short-rainy (April/May) seasons of 2006.

Village Species Mean no. mosquito/ light trap/ night Dry season Short-rainy season In Out Mean In Out Mean

Irrigated An. arabiensis 1.53 0.93 1.23 2.14 1.22 1.70 An. pharoensis 1.74 2.72 2.24 1.31 1.65 1.48 An. coustani 0.50 0.73 0.62 0.57 1.00 0.79 Any anopheline 3.77 4.38 4.09 4.02 3.87 3.97

Non-irrigated An. arabiensis 0.00 0.00 0.00 0.63 0.21 0.32 An. pharoensis 0.00 0.00 0.00 0.07 0.00 0.03 An. coustani 0.00 0.00 0.00 0.05 0.23 0.13 Any anopheline 0.00 0.00 0.00 0.75 0.44 0.59

429

50 Indoors (n = 24 light trap‐ 40 nights) Outdoors (n = 24 light trap‐ 30 nights)

10 collected each hour

Percentage of mosquitoes mosquitoes of Percentage 0

100 300 000 600 2200 0700 00-190000-2000 00-0200 00-040000-0500 18 19 2000-2 2100- 2200-2 2300-00000-010001 0200-030003 04 0500-0 0600- Hour

Figure 2. Hourly activity of Anopheles arabiensis indoors and outdoors from light trap catches (as percentage of mosquitoes collected each hour) in an irrigated village in Zeway area, during the study period (February to May 2006).

50.0 Indoors (n = 24 light trap‐ nights) 40.0 Outdoors (n = 24 light trap‐ nights) 30.0

20.0

10.0 collected each hour 0.0 Percentage of mosquitoesPercentage

0 0 00 0 00 100 0 19 00 10 - -2 -22 -23 -0 -0500 -0700 0 0-2000 0-0 0 0 00 00 00 0 0 100 2 3 0 18 19 2 2 2 2 0 0100-02000200-03000300-04000400 0500-06000600 Hour

Figure 3. Hourly activity of Anopheles pharoensis indoors and outdoors from light trap catches (as percentage of mosquitoes collected each hour) in an irrigated village in Zeway area, during the study period (February to May 2006).

430

Indoors (n = 24 light trap‐ 40.0 nights) Outdoors (n = 24 light trap‐ 30.0 nights)

20.0

10.0 collected each hour hour each collected

Percentage of mosquitoes mosquitoes of Percentage 0.0

0 0 0 00 90 000 100 20 30 000 100 2 300 400 700 1 2 2 2 2 0 0 -0 0 0 -0500 -0600 0 0- 0- 0- 0- 0- 0- 0- 0 0- 0- 0 0 0- 0 0 0 0 0 0 60 180 19 20 210 220 23 00 010 02 03 040 050 0 Hour

Figure 4. Hourly activity of Anopheles coustani indoors and outdoors from light trap catches (as percentage of mosquitoes collected each hour) in an irrigated village in Zeway area, during the study period (February to May 2006).

Table 5. Parous rate of Anopheles species in the irrigated (Abene-Girmamo) and non-irrigated (Woshgulla) villages in Zeway area, Central Ethiopia, in the dry (February/March) and short-rainy (April/May) seasons of 2006.

Village Season An. arabiensis An. pharoensis An. coustani N1 P (%)* N P (%) N P (%)

Irrigated Dry 46 27 (58.7) 90 39 (43.3) 22 4 (18.2) Short-rainy 81 18 (22.2) 51 21 (41.2) 29 1 (3.4) Non-irrigated Dry - # - - - - - Short-rainy 38 2 (5.3) 4 0 (0.0) 16 0 (0.0)

1 N = number of specimens dissected * P is the number of parous females, and the percentage of P is calculated from N. # - (minus sign) indicates that no mosquito was collected during that season and thus no dissection.

431

Table 6. Number (and percentage) of blood-fed Anopheles mosquitoes tested positive for human and/or bovine bloods by direct ELISA, from collections obtained from an irrigated village in Zeway area, Central Ethiopia, during the study period (February to May 2006).

Species No. of mosquito Human only (%) Bovine only (%) Mixed blood* (% Unidentified (%) tested

An. arabiensis 120 85 (70.8) 17 (14.2) 9 (7.5) 9 (7.5) An. pharoensis 142 87 (61.3) 29 (20.4) 11 (7.7) 15 (10.6) An. coustani 16 1 (6.2) 12 (75.0) 0 (0.0) 3 (18.8)

* Mixed blood meal refers to a blood meal containing both human and bovine bloods.

Table 7. Plasmodium falciparum sporozoite rate of Anopheles species collected from an irrigated village (Abene-Girmamo) in Zeway area, Central Ethiopia, during the dry (February/March) and short-rainy (April/May) seasons of 2006.

Season An. arabiensis An. pharoensis An. coustani N SR (%) N SR (%) N SR (%)

Dry 152 1 (0.66) 325 3 (0.92) 61 0

(0.00)

Short-rainy 272 4 (1.47) 184 0 (0.00) 70 0

(0.00)

Overall 424 5 (1.18) 509 3 (0.59) 131 (0.00)

N – number of mosquitoes tested by ELISA. SR – number of mosquitoes tested positive for P. falciparum sporozoites.

Note: none of the mosquitoes collected from the non-irrigated village were positive for malaria sporozoites.

432

4. Discussion different larval habitat requirements. Previous studies have shown that An. The present study revealed that the small- arabiensis prefers open, shallow and scale irrigation scheme in Zeway area has temporary breeding habitats while An. created breeding sites for the two malaria pharoensis thrives in shaded, permanent vector species, namely, An. arabiensis and water bodies with emergent vegetation An. pharoensis. The most important (Snow, 1983; Gillies and Coetzee, 1987). prolific Anopheles larval habitats were found to be poorly constructed irrigation Rains are known to have dual effect on the canals (that allow water to stand for a development of mosquito larvae. When it period of time), canal leakage pools rains, new mosquito-breeding sites are (formed due to perforated soil walls of the created; at the same time at other irrigation canals) and waterlogged fields previously existing sites, some individuals (field puddles formed due to over- will be washed away. We observed that irrigation). The same breeding habitats newly formed breeding sites were sooner have been shown to create conducive colonized by An. arabiensis and An. breeding grounds for An. arabiensis in the coustani (as these species prefer such dam villages of Tigray, where microdam- habitats) while older permanent larval based irrigation is practiced during the dry habitats of An. pharoensis diminished. season (Yohannes et al., 2005). In Similar observations were reported in agreement to our findings, in Mwea Mwea irrigation scheme in Kenya, where irrigation scheme, Kenya, it has been larval An. arabiensis were found reported that An. arabiensis, An. abundantly in newly flooded rice fields in pharoensis and An. coustani thrive well in the wet season but a few weeks later, irrigated fields where rice was commonly when the rice moderately grew, An. grown (Ijumba et al., 1990; Muturi et al., pharoensis was the one predominated 2006). In irrigation schemes of Faiyum (Mukiama and Mwangi, 1989). Governorate, Egypt, irrigation ditches, seepage water collections and irrigated This study generally confirmed that the fields with moderate crop growth were irrigation scheme in Zeway area has shown to be the major sources of An. created good Anopheles mosquito pharoensis during the dry season (Soliman breeding conditions by restoring the lost et al., 1967), in line with our finding for surface water during the dry season. Thus, the same species in the present study. Anopheles larval production in the irrigated villages of Zeway area is no We also observed that larvae of An. longer restricted to the wet seasons; rather arabiensis were predominantly abundant continuous breeding of Anopheles in newly formed canal leakage pools and mosquitoes throughout most of the year is field puddles, while larval An. pharoensis possible as the crucial linkage between the preferred canal leakage pools and rainy seasons is provided by the irrigation irrigation canals covered with vegetations. activities. Therefore, there is a potential Even in the same larval habitats where the for dry season malaria transmission in the two species coexisted (such as canal irrigated villages of Zeway area, as leakage pools), Anopheles arabiensis malaria vector mosquitoes (An. arabiensis mostly preferred the shallow, sunlit and and An. pharoensis) thrive well in disturbed (muddy) margins of the habitat breeding sites created by the irrigation while An. pharoensis was frequently scheme coupled with the prevailing sampled around the shaded and deeper climatic factors that could facilitate parts of the habitat with encroaching development of the aquatic stages of the vegetation. This indicated that An. vector as well as the malaria parasites arabiensis and An. pharoensis have inside the female anopheline.

433

Consistent with the observed seasonal We found that peak indoor and outdoor trend in larval abundance, variations in activities of An. arabiensis, An. seasonal adult densities were also evident pharoensis and An. coustani occurred during the study period. The density of during the early period of the night (before adult An. arabiensis was higher in the 22:00 hours), coinciding with the night short-rainy season than the dry season time behavior of the local people in the while the densities of An. pharoensis and study area. Similar early biting behavior An. coustani peaked in the dry season. was previously reported for An. arabiensis Similar seasonal trend was observed in and An. pharoensis in Zeway (Abose et villages at close proximity to Lake Zeway, al., 1998b), and An. arabiensis in Tigray where An. arabiensis outnumbered An. (Yohannes et al., 2005). In Sille, an pharoensis during the wet season while irrigated village in southern Ethiopia, the latter dominated the former in the dry Taye et al., (2006) reported that peak season. These species are common in biting activities of An. pharoensis and An. irrigated villages elsewhere in Africa coustani occurred between 18:00 and where they occur sympatrically (Snow, 20:00 hours, which is in agreement with 1983; Mukiama and Mwangi, 1989; the present findings for the two species. In Ijumba et al., 1990; Muturi et al., 2006). contrast to the observed early biting periodicity of An. arabiensis in the present Indoor and outdoor light trap catches study area, these authors reported a peak revealed that An. arabiensis was more biting activity between 23:00 and 3:00 endophagic while An. pharoensis and An. hours for the same species in Sille. coustani showed a more exophagic Interestingly, 40 years ago, in Zeway area behavior. We observed that the local most An. gambiae s.l. (presumably An. people in the study area spend the early arabiensis) fed readily after 23:00 hours part of the night (on average up to 10 pm) and little early evening biting activity was outdoors either working on their field or recorded (Rishikesh, 1966), suggesting taking care of their cattle. Such night time that the early biting behavior of this behavior of the local people might species has evolved since then. The early increase the chance of receiving more biting activity of An. arabiensis is likely to bites by the inherently exophagic be a consequence of long-term application populations of An. arabiensis and An. of residual insecticides, particularly DDT, pharoensis in the study area. Similar selecting for early biting behavior as it has suggestion for An. arabiensis was also been suggested recently in Tigray previously made by Ameneshewa (1995) (Yohannes et al., 2005). Moreover, such who worked in Gergedi (Awash valley, early biting activity of the malaria vector about 80 Km from Zeway) reported that populations in the current study area is the biting behavior of this species depends likely to compromise the efficacy of strongly on the availability of host either insecticide-treated bed nets as large indoors or outdoors during the period of proportion of bites occurred before the its biting activity in the evening. An. local people, including children, go to pharoensis and An. coustani are well sleep under their bed nets. known exophagic species in Ethiopia (Nigatu et al., 1994; Adugna and Petros, In the present study, An. arabiensis had a 1997; Abose et al., 1998b; Taye et al., higher parous rate during the dry season 2006) and elsewhere in Africa, such as than the short-rainy season while An. Kenya (Ijumba et al., 1990; Mukiama and pharoensis showed insignificant variation Mwangi, 1989), Sudan (El Gaddal et al., in its parous rate during the two seasons. 1985) and Cameroon (Antonio-Nkondjio This report is inconsistent with a previous et al., 2006). finding in Upper Awash that recorded a

434

higher parous rate for An. arabiensis (Habtewold et al., 2001) and Sille (Taye during the wet season than the dry season et al., 2006; Tirados et al., 2006), but (Ameneshewa, 1995). The explanation for lower than a 1.52% sporozoite rate in the this discrepancy is that following the adjacent, Wonji area (Ameneshewa, unusual heavy rains in April, high 1995). A 0.88% P. falciparum sporozoite recruitment of young ones into the rate of An. pharoensis in the dry season existing older population might have confirms the vectorial role of this species resulted in a higher proportion of in malaria transmission in the irrigated nulliparous females and hence lower villages of Zeway area particularly during parous rate. The parity rate of An. the dry season. On the other hand, arabiensis and An. pharoensis thus Anopheles arabiensis was found infected suggested higher longevity during the dry with P. falciparum sporozoites both in the season, hence likely to maintain malaria dry and short-rainy seasons, suggesting transmission during this season of the that this species play significant role in year. malaria transmission both in the dry and wet seasons of the year. These findings The reported Human Blood Index (HBI) could be the first report for the dry season for An. arabiensis (0.78) and An. and also for An. pharoensis. Hence, the pharoensis (0.69) in the present study role of An. pharoensis in transmitting P. reaffirmed the importance of these species falciparum should not be underestimated in malaria transmission in Zeway area. in areas where this species is abundant. Yohannes et al (2005) reported an HBI of 0.72 for indoor-resting An. arabiensis in The major short-coming of the current Tigray, northern Ethiopia, which is a study was that larval and adult collections comparable finding for the same species in were made merely on seasonal basis, only the present study. An HBI of 0.66 was focusing on the dry and short-rainy reported for An. arabiensis in Konso, seasons, due to which monthly variations southern Ethiopia, (Tirados et al., 2006), at different periods of the year were not which is lower than the present finding, as shown. Hence, further longitudinal studies the species population in Konso was in the same area are required to ascertain reported to be exclusively exophagic. the present findings. Adugna and Petros (1996) reported higher HBI for An. pharoensis (0.84) and An. In conclusion, although development of coustani (0.26) from samples collected in irrigation schemes is of paramount mixed dwellings. In our study, An. importance to increase crop yield and coustani had shown an exceptionally high hence to ensure food security and preference of An. coustani (75%) for economic growth in Ethiopia, its adverse bovine blood – hence less likely to play health problems may pose significant significant role in malaria transmission. public health concerns. The findings of the Hence, the present study confirmed that present study underscore the importance An. arabiensis and An. pharoensis are the of irrigation schemes in semi-arid areas two most important anthropophagic like Zeway in maintaining malaria species in Zeway area, which is in transmission particularly during the dry agreement with previous reports from the season, when mosquito abundance is same area (Rishikesh, 1966; Abose et al., normally presumed to be limited. Proper 1998b). water management and control measures such as source reduction through The P. falciparum sporozoite rate of environmental management could help to 1.18% for An. arabiensis in the present reduce mosquito-breeding sites and thus study is comparable to the 1.1% malaria transmission. sporozoite rate reported from Arbaminch

435

References comparative study on the acquired immunoprotection in native populations. Abose, T., Ameneshewa, B., Teklehaimanot, Acta Trop. 51:103–111. A., Tulu, A., Colluzi, M., Petrica., V. Bradley, D.J. 1988. The epidemiology of and Petrangeli, G. 1998a. Cytogenetic ricefield-associated diseases. studies on the Anopheles gambiae International Rice Research Institute in complex species in Ethiopia. Ethiop. J. collaboration with WHO/FAO/UNEP, Health Dev.. 12: 81-83. Panel of Experts on Environmental Abose, T., Ye-ebiyo, Y., Olana, D., Management for Vector Control. Alamirew, D., Beyene, Y., Regassa, L. International Rice Research Institute, and Mengesha, A. 1998b. Reorientation Philippines. pp 29-39. and definition of the role of malaria El Gaddal, A.A., Haridi, A.M., Hassan, F.T. vector control in Ethiopia. and Hussein, H. 1985. Malaria control in WHO/Mal/1998.1085. World Health the Gezira-Managil irrigated scheme of Organization. Geneva. Pp 38. the Sudan. J Trop Med Hyg. 88:153-9. Adugna, N. and Petros, B. 1996. Faye, O., Gaye, O., Herve, J.P., Diack, P.A. Determination of the human blood index and Diallo, S. 1993.Malaria in the of some anopheline mosquitoes by using sahelian zone of Senegal: Parasitic ELISA. Ethiop Med J. 34:1-10. indices. Ann Soc Belg Med Trop. 73: Ameneshewa, B. 1995. The behavior and 31–36. biology of Anopheles arabiensis in Ghebreyesus, T.A., Haile, M., Witten, K., relation to epidemiology and control of Getachew, A., Yohannes, A., Yohannes, malaria in Ethiopia. Ph.D. Thesis. M., Teklehaimanot, H. and Lindsay, University of Liverpool, UK. Pp 288. S.W. 1999. Incidence of malaria among Ameneshewa, B. and Service, M.W. 1996. children living near dams in northern Resting habits of Anopheles arabiensis Ethiopia: community-based incidence in the Awash Valley of Ethiopia. Annals survey. Brit Med J. 319:663-669. Trop Med Parasitol. 90: 515-521. Gillies, M.T. and Coetzee, M. 1987. A Amerasinghe, F.P. and Munasingha, N.B. supplement to the Anophelinae of 1988. A predevelopment mosquito Africa, South of Sahara (Afrotropical survey in the Mahaweli Development Region). Publications of the South Project area, Sri Lanka: immatures. J African Institute of Medical Research. Med Entomo.125: 286-294. Pp 66. Antonio-Nkondjio, C., Kerah, C.H., Simard, Gratz, N.G. 1988. The impact of rice F., Awono-Ambene, P., Chouaibou, M., production on vector-borne disease Tchuinkam, T. and Fontenille, D. 2006. problems in developing countries. Complexity of malaria vectorial system International Rice Research Institute in in Cameroon: contribution of secondary collaboration with WHO/FAO/UNEP, vectors to malaria transmission. J Med Panel of Experts on Environmental Entomo. 43(6): 1215-1221. Management for Vector Control. Beier, J.C., Perkins, P.V., Wirtz, R.A., International Rice Research Institute, Koros, J., Diggs, D., Gargian, T.P. and Philippines. pp7-12. Koech, D.K. 1988. Bloodmeal Habtewold, T., Walker, A.R., Curtis, C.F., identification by Enzyme-Linked Osir, E.O. and Thapa, N. 2001. The Immunosorbent Assay (ELISA), tested feeding behavior and Plasmodium on Anopheles (Diptera: Culicidae) in infection rates of Anopheles mosquitoes Kenya. J Med Entomo. 25: 9-16. in southern Ethiopia in relation to use of Boudin, C., Robert, V., Carnevale, P., insecticide-treated livestock for malaria Ambroise, T.P., 1992. Epidemiology of control. Trans Royal Soc Trop Med Plasmodium falciparum in a rice field Hyg. 95: 584-586. and a savanna area in Burkina Faso:

436

Henry, M.C., Rogier, C., Nzeyimana, I., works in sub-arid and humid regions. Assi, S.B., Dossou-Yovo, J., Audibert, Part 1. Entomological investigations. M., Mathonnat, J., Keundjian, A., Acta Trop. 89: 193–203. Akodo, E., Teuscher, T. and Carnevale, McCartney, M.P., Lankford, B.A. and P. 2003. Inland valley rice production Mahoo, H. 2007. Agricultural water systems and malaria infection and management in a water stressed disease in the savannah of Cote d’Ivoire. catchment: lessons from the Trop Med Int Health. 8: 449–458. RIPARWIN project. Research Report Ijumba, J.N. and Lindsay, S.W. 2001. 116. International Water Management Impact of irrigation on malaria in Institute. Colombo, Sri Lanka. Africa: paddies paradox. Med Vet Ministry of Water Resource (MWR). 2002. Ent.15: 1-11. Water sector development program Ijumba, N.J. Mwangi, R.W. and Beier, J.C. 2002-2016. Volume II. Main Report. 1990. Malaria transmission potential of Ministry of Water Resources, Federal Anopheles mosquitoes in the Mwea- Democratic Republic of Ethiopia. Addis Tebere irrigation schemes, Kenya. Med Ababa. October 2002. Vet Entomo. 4: 425-432. Mukiama, T. and Mwangi, R. 1989. Josse, R., Josseran, R., Audibert, M., 1987. Seasonal population changes and Malariometry and seasonal variations of malaria transmission potential of malaria indices in the rice fields of Anopheles pharoensis and the minor Maga (north Cameroon) and in the anophelines in Mwea Irrigation Scheme, adjacent region. Med Parasitol. 25: 63– Kenya. Acta Tropica. 46(3): 181-189. 71. Muturi, E., Shililu, J., Jacob, B., Gu, W., Lewis, D.J. 1958. The recognition of Githure, J. and Novak, R. 2006. nulliparous and parous Anopheles Mosquito species diversity and gambiae by examining the ovarioles. abundance in relation to land use in Trans Royal Soc Trop Med Hyg. 52: riceland agroecosystem in Mwea, 456-461. Kenya. J Vect Biol. 31(1): 129-137. Lindsay, S.W., Wilkins, H.A., Zieler, H.A., Nigatu, W., Petros, B., Lulu, M., Adugna, N. Daly, R.J., Petrarca, V. and Byass, P. and Wirtz, R. 1994. Species 1991. Ability of Anopheles gambiae composition, feeding and resting mosquitoes to transmit malaria during behavior of the common anthropophilic the dry and wet seasons in an area of anopheline mosquitoes in relation to irrigated rice cultivation in The Gambia. malaria transmission in Gambella, J Trop Med Hyg. 94: 313–324. Southwest Ethiopia. Ins Scie Appls. Lines, J.D., Curtis, C.F., Wilkes, T.J. and 15(3): 371-377. Njunwa, K.J. 1991. Monitoring human- Rishikesh, N. 1966. Observations on biting mosquitoes (Diptera:Culicidae) in Anopheline Vectors of Malaria in an Tanzania with light traps hung beside unsprayed Upland Valley in Ethiopia. mosquito nets. Bull Entomo Res. 81:77- WHO/MAL/66.554. Geneva. 26pp. 84. Service, M.W. 1993. Mosquito Ecology: Lipton, M., Litchfield, J. and Faures, J-M. Field Sampling Methods. 2nd Edition. 2003. The effects of irrigation on Chapman and Hall, London. UK. poverty: a framework for analysis. Sissoko, M.S., Dicko, A., Briët, O.J.T., Water Policy. 5: 413-427. Sissoko, M., Sagara, I., Keita, H.D., Marramaa, L., Jambou, R., Rakotoarivony, Sogoba, M., Rogier, C., Touré, Y.T. and I., Pock Tsi, J.M.L., Duchemine, J.B., S. Doumbo, O.K. 2004. Malaria incidence Laventure, S., J. Mouchet, J. and Rouxh, in relation to rice cultivation in the J. 2004. Malaria transmission in irrigated Sahel of Mali. Acta Trop. 89: Southern Madagascar: influence of the 161–170. environment and hydro-agricultural

437

Snow, F. 1983. Mosquito production and mosquitoes feeding on maize pollen. species secession from an area of Am J Trop Med Hyg. 63(1): 90-93. irrigated rice fields in The Gambia, Ye-ebiyo, Y., Pollack, R.J., Kiszewski, A. West Africa. J Trop Med Hyg. 86: 237- and Spielman, A. 2003. Enhancement of 245. development of larval Anopheles Soliman, A.A., Rifaat, M.A. and Ibrahim, arabiensis by proximity to flowering M.T.H. 1967. Biology of Egyptian maize (Zea mays) in turbid water and Anophelines. I. Biological activities of when crowded. Am J Trop Med Hyg. Anopheles pharoensis in nature. J Trop 68(6): 748-752. Med Hyg. 70(3): 63-67. Yohannes, M., Haile, M., Ghebreyesus, Taye, A., Haddis, M., Adugna, N., Tilahun, T.A., Witten, K., Getachew, A., Byass, D. and Wirtz, R.A. 2006. Biting P. Lindsay, S.W. 2005. Can source behavior and Plasmodium infection reduction of mosquito larval habitat rates of Anopheles arabiensis from Sille, reduce malaria transmission in Tigray, Ethiopia. Acta Trop. 97:50-54. Ethiopia? Trol Med Int Health. 10(12): Tirados, I., Constantini, C. Gibsson, G. and 1274-1285. Torr, S.J. 2006. Blood-feeding behavior of the malaria mosquito Anopheles arabiensis: implication for vector control. Med Vet Entomo. 20: 425-437. Verrone, G.A. 1962a. Outline for the determination of malaria mosquitoes in Ethiopia. Part I. Adult female anophelines. Mosq News. 22: 37-49. Verrone, G.A. 1962b. Outline for the determination of malaria mosquitoes in Ethiopia. Part II. Anopheline Larvae. Mosq News. 22:394-401. Wirtz, R.A., Zavala, F., Charoenvit, Y., Campbell, G.H., Burkot, T.R., Schneider, I., Esser, K.M., Beaudoin, R.L. and Andre, R.G. 1987. Comparative testing of monoclonal antibodies against Plasmodium falciparum sporozoites for ELISA development. Bull WHO. 65(1): 39-45. World Health Organization (WHO). 2006. Global Malaria Situation. WHO Technical Report Series. WHO/RBM/2006.3. Geneva. Pp34. Ye-Ebiyo, Y., Pollack, R. J. and Spielman, A. 2000. Enhanced development in nature of larval Anopheles arabiensis

438

Malaria Transmission In Ziway, Eastern Shewa Zone, As Related to Irrgation Development

Yihenew Alemu1, Eline Boelee2, Beyene Petros1, Dawit Alemu3, Solomon Kibret1

1 Department of Biology, Addis Ababa University (AAU), Addis Ababa, Ethiopia 2 International Water Management Institute (IWMI) Addis Ababa, Ethiopia 3 Ethiopian Institute of Agricultural Research (EIAR) Addis Ababa, Ethiopia

workers during the establishment of small- scale irrigation schemes are recommended. Abstract Malaria is the major public health problem in Ethiopia in general and in the Oromia Regional State in particular. Ziway is one of Key words: Agriculture, Anopheles Oromia regions with frequent malaria pharoensis, Ethiopia, Irrigation, Malaria, epidemics usually occurring at the end of the Plasmodium falciparum, P. vivax, Ziway. main rainy season, September to mid December. In addition it is one of the Introduction regions in Ethiopia where unreliable rainfall In Ethiopia, approximately 4-5 million cases frequently affects agricultural production. of malaria are reported annually (in a normal The government of Ethiopia supported the transmission year) (WHO, 2005). Malaria is development of small-scale irrigation found in about 75% of the total area of the schemes to ensure food self-sufficiency and country, and 40-50 million (>65%) of the sustain agricultural productivity. However, total population is at risk of infection (Tulu, without proper planning such schemes are 1993 and WHO, 2005). Transmission known to worsen vector-borne disease usually occurs at altitudes <2000 meters endemicity. The objective of the study was above sea level. The two main seasons for to assess the effect of small-scale irrigation transmissions of malaria in the country are schemes on malaria transmission around September–December, after the heavy Ziway. Blood smear samples were examined summer rains, and March–May, after the at the end of the main rainy season and the light rains. P. falciparum and P. vivax are dry season of 2005/2006. Overall irrigated the dominant human malaria parasites, areas had significantly higher (19.2%; which account for about 60% and 40% of p<0.05) malaria infection prevalence rate as cases, respectively (Tulu, 1993). compared to the non-irrigated study sites (16%). In irrigated areas all age categories Irrigation projects in Ethiopia are identified showed higher malaria infection prevalence as large scale irrigation (> 300 ha.), medium in the dry season as compared to the rainy scale irrigation (200-300 ha.) and small season. However, the difference was scales irrigation (< 200ha.) (Awulachew et significantly (P<0.05) higher only in the age al., 2005).Traditionally farmers have built category greater than 15 years old. Control small scale-schemes on their own initiatives, interventions through integrated malaria sometimes with government technical and control approaches that include education material support (MoWR, 2002). The farm about its importance, source reduction and size varies between 0.25 ha and 0.5 ha combined efforts by agricultural and health (Awulachew et al., 2005). The Federal or Regional government normally constructs

439

small scale modern schemes. Such schemes been identified for creative malaria were expanded after the catastrophic control if it is found increased with drought in 1973/74 to achieve food security irrigation. and better peasants’ livelihoods by producing cash crops. Such schemes involve Materials and Methods dams and diversions of streams and rivers (Awulachew et al., 2005). Study Area Irrigation schemes in some parts of the The study was conducted in Oromia country were found to increase malaria Regional state, East Showa Zone, Adamitulu transmission (Ghebreyesus et al., 1999). Judo Kombolcha Woreda (figure: 1). It is These irrigation schemes had been 163 Km south of Addis Ababa alongside introduced to reduce dependence of local Lake Ziway. Similar to other Ethiopian agriculture on irregular rainfall. However, regions the main rainy season of the area while the schemes have had a positive starts in June and extends up to impact on agriculture, the effect on health August/September while the short rainy has been worrying, with an increased season begins in March and extends to incidence of malaria. The findings in April/May. The area mean annual Ethiopia raise fears that, unless properly temperature of 20OC and annual maximum thought out schemes to improve the and minimum temperature of 27.5 OC and environment are in place, irrigation could do 13.9OC respectively. The mean annual rain as much harm as good (Ghebreyesus et al., fall of the area varies between 700 and 800 1999). Therefore, it is vital to gain a better mm (Figure: 2) .The area is characterized as understanding of the influence that irrigation lowland with sparsely distributed Acacia and agricultural activities have on the spread trees and thorny bushes. of malaria especially as developing At the beginning of the study, relevant countries extend their irrigated areas to feed socio-demographic information was sought rapidly growing populations. from the agricultural and health office experts of the area through structured The general objective of this study was interviews. This was followed by selection to generate information for an effective of specific sites/villages and two cross- and economically viable integrated sectional surveys at the end of the rainy and management strategy that describes during the dry season. The rainy season data health and irrigation factors, to create were colleted in September/October, which effective malaria prevention in irrigated is the period with the highest malaria area. Hence, malaria transmission has transmission almost every year in Ziway been assessed as related to irrigation (Abose et al., 1998; Seyoum et al., 2002). development at the study sites, by The dry season study was conducted in determining the prevalence of the February/March before the beginning of the short rainy season and before the second Plasmodium infections in irrigated and round indoor insecticide spray begins. non-irrigated areas at different villages around Ziway. Secondly, options have

440

160 35

140 30

120 25 an d )

% 100

y( 20 80 15 T em p erature 60R a in fa ll(m m ) 10 40 R e la tiv e H u m d iti

20 5

0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Months, 1995- 2006

Rain Fall (mm) Relative Humditiy (%) Mean Min Temp Mean Max Temp

Figure 2. Monthly rainfall (mm), mean relative humidity, mean maximum and minimum monthly air temperatures (oC) of Ziway area, 1995-2006. Figure 1: Map of Ethiopia showing location of the study site Source: National Meteorological Agency (NMA)

Selection of the irrigated area was done Abine Germamo is found at the northern based on pre-defined criteria 1) the irrigation outskirts of Ziway town by the side of the site should be far from the lake that is the road coming from Addis Ababa. It includes main source of irrigation water in order to previous kebeles Hizbawi Betele, Abine avoid the effect of vectors that breed in Gijota and Abine Germamo. It is located at natural habitats; 2) farmers should live in 1647 meters above sea level. According to close proximity to the irrigation fields but the information obtained from the away from the lake; 3) the site should be agricultural office, the number of accessible and close to the Woreda health households in this kebele reaches up to 934. center for easy of sample transport and However, the beneficiaries of irrigation in examination. Abine Germamo was found to the 2004/2005 production season were only be the area, which best fitted to the selection 75 households. Lake Ziway is the source of criteria. It is found more than 5 km away irrigation water. Both the eastern (the field from the source of irrigation water (the lake) between the lake and the road) and the and non- irrigated area. western side of the main road are irrigated

441

but almost all the farmers are living on the divided by the population of 2003/04 of western side of the road. Pumps are used in the area was used to estimate the order to lift the water, then it is delivered prevalence in the non-irrigated area. over a long distance through pipes and Using Win Episcope 2.0 (developed by surface canals. Finally the water is diverted facultad de veternaria Zara goza, down to all irrigation fields by surface Wagningen University and University of irrigation with the help of the force of gravity. The houses of the farmers are found Edinburgh) computer soft-ware the in close proximity to the irrigation fields. All sample size was determined. It was the inhabitants of the area, those benefited estimated by using the difference from irrigation (irrigators) and those without between proportions with 95% level of access to irrigation (non-irrigators) are confidence interval and value of 80% for almost all found in close proximity to the the power with the formula (for the irrigation fields. sample size) for a two-tailed test: 2 The main criteria for selecting the non- ⎡Z(a) + Z(b)⎤ irrigated area were: 1) it should be at a n = ⎢ ⎥ ∗[]( p1 * q1) + ( p2 * q2 ) p − p distance beyond the reach of mosquitoes ⎣ 1 2 ⎦ from the irrigated area and the lake (Ghebreyesus et al., 1999; Klinkenberg et al., 2005; Yohannes et al., 2005); 2) with the Where Z (a) and Z (b) are the value of same topographical position, comparable student’s t at specified confidence level agricultural production system and other and at the specified power, respectively, important features except irrigation as the and p1 and p2 are proportion in irrigated areas (Ghebreyesus et al., 1999; population one and two respectively and Klinkenberg et al., 2005; Yohannes et al., 2005); 3) accessibility. Three areas were p1.q1 and p2.q2 are the variance of the taken into consideration for the selection of proportion in population one and two. the non-irrigated area but Weshgule was Hence the sample size according to this found to be the area, which fitted to the method was calculated to be 700. selection criteria best. Weshgule is found at Meanwhile, based on prevalence rate the southern outskirts of Ziway town at 1654 obtained from parasitological survey meters above sea level. conducted at Ziway Eddo-Gojola irrigation area by Abose et al., (1998), the Sample selection sample size was calculated to be 280. Then by taking the average of these two We assessed malaria parasite prevalence, calculations the sample size was finally a direct indicator of the impact of determined to be approximately 500 malaria, in communities with and samples (individuals) each from the without irrigation (Klinkenberg et al., irrigated and non-irrigated areas. 2005). Six year of malaria case data in the study areas from the Health office at The applied sampling method was Ziway was taken for the determination probability sampling; where each member of of sample size. Average of positive cases the population has a known non-zero 1999/2000 - 2004/2005 of the irrigation probability of being selected. Systematic area divided by the population of sampling was used in the study. The list of 2003/04 of the area was taken as average households for the selected areas, knowing annual prevalence in irrigated area and that the list does not contain any hidden average of positive cases of 1999/2000 - order, was obtained from the kebele leaders and it was used as a sampling frame. After 2004/2005 of the non-irrigated areas

442

the required sample size had been calculated The quality of the Giemsa staining solution every nth record, calculated by dividing the was checked before using directly to the sample size by the average family size of the samples. A fresh 10% Giemsa solution in area, selected from a list of population tap water was prepared and the air-dried members. Every head of the household slides were immersed into a staining jar for found at home during sample collection was 10 minutes. The stained slides were rinsed informed about the study and the consent by tap water and put in upright position to form was read to them. When they agreed to dry (Schilchtherle et al., 2000). The stained participate in the study blood smear samples thick and thin films were observed under were collected from all family members. In 100x oil immersion objective light case a household did not want to participate microscope. First, the thick blood smear in the study or if the head of the household samples were observed in order to know not found at home, then the next house was whether the sample is positive or negative. taken as a replacement. When a sample was found positive then the thin blood smear was used to identify the Parasitological survey species. Ethical considerations Blood samples were colleted by skin pricking of the finger onto glass slides at the Ethical considerations were addressed by end of the rainy season and during the dry treating positive malaria cases according to season of the study. Specialized laboratory the standard drug regimen. Informed consent technicians of the Woreda malaria control was obtained from adult study subjects and laboratory did the blood sample collection from parents and guardians in case of from the selected households (Figure 6 & 7). children before sampling. The project The thick and thin blood smears were obtained ethical approval from the Ethical prepared on the same slide side by side, Committee of the School of Graduate properly labeled, air dried and then the thin Studies of Addis Ababa University. The blood smears were fixed with methanol on- project was also discussed with concerned site. The samples were carefully transported authorities in the study areas and their to the Ziway Woreda health office agreement was obtained. Those individuals laboratory where specialized laboratory with positive results were treated with the technicians did the staining and required treatment and dose free of charge. parasitological test by microscopic The treatment was given in collaboration examination. The samples were put into with the Woreda health office according to slide boxes and transported to Addis Ababa the guidelines of Ethiopian Ministry of and further examination of all the positive Health. In this study, no invasive procedure samples and 10% of the negative samples other than the finger pricking, which is used (Ghebreyesus et al.,. 1999) were also done for routine diagnosis, was used. Experienced in AAU Biomedical Science laboratory. The laboratory technicians performed finger- thick and the fixed thin blood film slides of pricking in order to avoid unnecessary pain the second season (dry season) samples were and bleeding. For finger pricking, a single properly stained and examined at Ziway disposable sterile lancet was used per person Woreda health office laboratory. to avoid possible transmission of infection. Afterwards, these were also put into slide boxes and then transported to the Ethiopian Results Health and Nutrition Research Institute (EHNRI) Laboratory, Addis Ababa, where Blood smear samples were examined from a further conformational examination was total of 2435 study participants, 1060 males done by technicians of the institute. and 1375 females. A total of 427 parasite positive slides were found in both study sites, 232 (19.2%) in irrigated villages and

443

195 (16%) in non-irrigated areas. Out of 198 significantly (P<0.05) higher than the (18.6%) infected males 10.3 % were in prevalence in non-irrigated areas. Moreover, irrigated areas and 8.3% were in non- the result revealed that there was no irrigated areas. Out of 229 (16.6%) infected significant (P>0.05) difference in malaria females 8.9% were in irrigated areas and infection prevalence between the two sexes 7.7% in non-irrigated areas. Overall malaria (Table 1). infection prevalence in irrigated areas was

Table 1. Overall malaria infection prevalence in the irrigated and non-irrigated areas, Ziway (2005/06).

Malaria positive cases (%)

No. of study participants Irrigated Non-irrigated

Male 1060 110 (10.3) 88 (8.3) Female 1375 122 (8.9) 107 (7.7)

Total 2435 232 (19.2)* 195(16.0) * Significant difference at (X2, P<0.05)

The rainy season malaria prevalence in the prevalence in the dry season (Table 2). irrigated area was (16.0%) and in the dry Moreover, additional analysis by season season it was (22.7%, Table 2). The finding revealed that malaria infection prevalence showed that malaria transmission in during the dry season was significantly irrigated areas during the dry season was (p<0.01) higher in irrigated areas (22.7%) as significantly (P<0.05) higher than during the compared to non-irrigated areas (11.5%). rainy season. For the non-irrigated areas, the Nevertheless, during the rainy season, rainy and the dry season malaria infection malaria infection prevalence in irrigated prevalence was (19.6%) and (11.5%), areas (16.0%) was significantly (p<0.05) respectively. The rainy season infection lower than in non-irrigated areas (19.6%, prevalence of non-irrigated areas was Table 2). significantly (P<0.05) higher than the

Table 2. Overall malaria infection prevalence in irrigated and non-irrigated areas during the rainy season and the dry season, Ziway (2005/06). Study areas Season Malaria positive cases (%)

Rainy (n=699) 112 (16%)

Irrigated areas Dry (n=528) 120 (22.7%)*

Rainy (n=692) 136 (19.6%)*

Non-irrigated areas Dry (n=516) 59 (11.5%) * Significant difference at (X2, P<0.05)

444

Comparison of infection prevalence of category greater than 15 years. In non- malaria in different age categories in irrigated areas higher, though not irrigated and non-irrigated areas during the significant, infection prevalence was rainy and the dry seasons of the study is depicted in all age categories during the shown in table 3. In irrigated areas all age rainy season as compared to the prevalence categories in the dry season showed higher during the dry season. Overall, there was a malaria infection prevalence as compared to decreasing but not significant trend in their infection prevalence in the rainy infection prevalence of malaria with season. However, the difference was increasing age categories within a season. significantly (P<0.05) higher only in the age

Table 3. Age-specific malaria infection prevalence in the irrigated and non-irrigated areas during the rainy and the dry season, Ziway (2005/06). Season Area Age Dry Rainy % No. Examined Positive No. Examined % Positive Irrigated <5 156 26.9 212 21.7 5 to 9 112 21.4 168 16.2 10 to 14 58 20.8 70 15.7 >15 202 20.3* 249 11.2 <5 153 17.5 189 24.7 5 to 9 125 13.2 165 20.4 10 to 14 46 10.1 82 17.3 Non- irrigated >15 192 7.1 256 14.4 * Significant at (X2, P<0.05)

Analysis of infection prevalence of the season was significantly (P<0.05) higher Plasmodium parasite species was done in than its prevalence in the dry season. For the irrigated and non-irrigated areas during the non-irrigated areas, rainy season infection rainy and the dry seasons of the study. In the prevalence of P. falciparum was irrigated areas infection prevalence of P. significantly higher (P<0.05) than the dry falciparum in the dry season was season infection prevalence and the same significantly (P<0.05) higher than its scenario was observed for P. vivax (P<0.05, prevalence in the rainy season, whereas Table 4). infection prevalence of P. vivax in the rainy

Table 4. Species-specific infection prevalence of malaria parasites in the study population during the rainy and dry seasons in the irrigated and non- irrigated areas, Ziway (2005/2006). Area Season No. Examined % Species % P. falciparum % P. vivax Irrigated Rainy 699 9 7* Dry 528 21* 1.7 Non-irrigated Rainy 692 12.6* 7.1* Dry 516 10.7 0.8 * Significant difference at (X2, P<0.05)

450

adult density of An. pharoensis in irrigated Discussion areas during the dry season (Abose et al., Irrigation structures often offer ideal habitats 1998). for the proliferation of anopheline An. pharoensis is the second important mosquitoes, including vectors of malaria. A vector, next to An. arabiensis, implicated in study conducted in Tigray (Northern malaria transmission in Ziway (Abose et al., Ethiopia) for instance, indicated an overall 1998; Seyoum et al., 2002). This species increase in incidence of malaria in the mostly rests outdoors on vegetation. The villages close to dams as compared with the irrigated farming in the area where control villages (Ghebreyesus et al., 1999). vegetables such as onion, cabbage, tomato, Another study in Southeastern Ethiopia had and cucumber, as well as cereal crops, shown malaria due to P. falciparum and P. mainly maize, were grown and other trees vivax to be the main health problems in planted for fencing irrigation fields would irrigation schemes along the Genale River provide ideal sites for resting and enhancing (Birrie et al., 1997). A study in Arba-minch longevity of this vector species. The also showed irrigation activity to have relevance of enhancement of longevity of created a year round breeding habitat for the vector mosquitoes to malaria transmission anopheline mosquitoes (Ashenafi, 2003). has been shown by the findings of Mukiama The present study looked into the situation in and Mwangi (1989) where higher adult Ziway where small-scale irrigation has been population of sporozoite-infected An. in practice for decades but the extent of its pharoensis was more abundant in irrigation linkage with malaria has not been studied. schemes in Kenya. Moreover, the higher The study findings are in agreement with the minimum temperature of the dry season in earlier reports that assessed the impact of the irrigated areas would contribute to construction of small irrigation dams on the completion of parasite life cycle inside incidence of malaria in Tigray (Ghebreyesus vectors. et al., 1999) and that of Klinkenberg et al. Furthermore, according to Carrara et al. (2005) in the irrigated areas in Ghana. (1990), An. pharoensis was the main malaria vector for intensive P. falciparum The difference in malaria prevalence transmission in Senegal River delta where all between irrigated and non-irrigated sites was examined An. gambiae Giles sensu lato were more pronounced during the dry season. P. falciparum sporozoite negative. Besides Probably, this is because the irrigation the report from Senegal, the role of An. structures in the area would provide suitable Pharoensis in P. falciparum-dominated breeding sites for malaria vector mosquitoes malaria transmission in Ziway study area during the dry season, while those in the during the dry season is supported by the non-irrigated area dry out. Since mosquito- entomological survey concomitantly breeding sites such as pools and puddles conducted in the area (Solomon Kibret et al., would be equally available both in the 2008 also in this proceedings). irrigated and non-irrigated areas during the The higher dry season malaria transmission, rainy season, such increase in malaria related to availability of water bodies, was prevalence was not observed with irrigation. similar to what was reported by Carlson et The humid environment during the small al. (2004) that water in pits created during rains (March, April and May) which was at brick-making supported the development of the same level of humidity following the big malaria vectors and thus enhanced dry rains (October) (National Meteorological season malaria transmission in western

Kenya. On the other hand, the higher Agency, NMA) and the humid environment, prevalence of malaria in the non-irrigated which also is created through irrigation areas during the rainy season might be due to during the dry season, would enhance vector “draw-down” phenomenon in the irrigated longevity. Moreover, Robert (2004) areas whereby rapid drying and over-flowing observed, the increase in the density of of the irrigation structures would disturb the foliage of plants will provide more shelter vector breeding sites and lead to decreased for adult mosquitoes, extending their transmission. Bziava and Kruashvili (2002) longevity. Likewise, when the residual soil had shown that periodic “draw-down” of moisture increases with irrigation, it will irrigation canals can be used as an ecological extend the range of sheltered habitats for measure for malaria control. mosquitoes. This would explain the higher

451

In irrigated areas farmers in older age groups could be altered in time through irrigation, will be at work early in the evenings when availability of surface water, humidity and the prominent vector of the dry season will temperature are factors which affect vector be active so these people will have a higher abundance, longevity and parasite risk of infection. Higher malaria prevalence development inside vectors (Woyessa et al., in older age groups in Ethiopia was also 2004; Teklehaimanot et al., 2004). The study reported by Yohannes and Petros (1996) by site (Ziway) has an average temperature of observing the absence of apparent decrease 20.70C with the higher temperature occurring in prevalence with increasing age in during dry season (Figure 4) (Teklehaimanot Nazareth, Ethiopia. Although Abose et al. et al., 2004). At 200C, 3 weeks are needed (1998) reported that men in the area were for P. falciparum development inside affected more than women, considering that mosquitoes while P .vivax can develop in 16 adult men are more likely to spend the days (Gilles and Warrell, 1993; CDC, 2004). evening hours outdoors working in their As the temperature increases, for instance at plantations and might receive more mosquito 220C, mosquito’s life cycle and P. bites than women, the present study revealed falciparum development inside mosquitoes no significant difference in malaria infection will be completed in 18 and 7.9 days, prevalence between the two sexes. This respectively (Teklehaimanot et al., 2004), suggests lack of difference in the outdoor and hence P. falciparum would be favored stay behavior of the two sexes and is and could suppress other human Plasmodium consistent with the findings of Himeidan et species (Smith et al., 2001). al. (2005) in an irrigated area in Eastern Persistence in P. falciparum malaria may Sudan. also be due to other more direct human activities (Singh et al., 2004). For example, It is generally known that P. falciparum is P. falciparum has developed resistance to dominant during the peak malaria sulfadoxine pyrimethamine in most parts of transmission season in September and Ethiopia (Worku et al., 2005) and hence October, while P. vivax tends to dominate Coartem (Artemisin-based Combination during the dry season in Ethiopia (Tulu, Therapy) is now recommended as substitute 1993). The present findings from the non- therapy. However, the population in Ziway, irrigated areas confirm the above during the study time, mainly uses generalization for P. falciparum. Decreases sulfadoxine pyrimethamine (personal in P. falciparum prevalence have been communication with health worker), as this shown to allow the potency of other parasite is available on the market. Such patients are species (Smith et al., 2001). The lower likely to serve as reservoirs of the parasite infection prevalence of P. vivax in irrigated and may lead to augmented P. falciparum areas during the dry season could be due to transmission in the dry season. the fact that P. vivax is most prevalent in unstable transmission conditions (Gilles and Warrell, 1993) and hence its rate of Conclusions and Recommendations transmission may not be influenced by irrigation. The following conclusions can be drawn Irrigation has created enhanced conditions from the present study on the impact of small for endemicity of falciparum malaria by scale irrigation on malaria transmission in changing its seasonal patterns. In irrigated Ziway. Significantly higher prevalence of areas, stable malaria transmission favors P. malaria was observed in irrigated areas as falciparum, whereas transmission of P. vivax compared to the non-irrigated areas in was much lower as its transmission would be Ziway. This suggests that the irrigated areas suppressed by P. falciparum (Smith et al., around Ziway are more favorable for 2001). An irrigation-associated increase of P. breeding and activity of malaria vector falciparum infection prevalence during the mosquitoes, which enhances their longevity dry season was also reported from India, and facilitates completion of parasite life where extensive irrigation had triggered P. cycle for most months of the year. Malaria falciparum dominated malaria in the virgin infection prevalence significantly increased levees of Thar Desert in India (Tyagi, 2004). at irrigated sites during the dry season when Although all factors that contribute to compared with the parallel scenario during malaria risk are not fully understood (Robert the rainy season. This was exhibited in all et al., 2003), seasonality features, which age groups but, older age groups (greater

452 than 15 years old) in irrigated areas had Konradsen, F., Van der Hoek W, (eds). significantly higher infection prevalence Malaria in irrigated agriculture: papers during the dry season as compared to the and abstracts for the SIMA Special same age group in the rainy season. Seminar at the ICID 18th International To counter the increased malaria risk near Congress on Irrigation and Drainage, developed water resources, there is a need Montreal, 23 July 2002. IWMI Working for special attention to health issues in the Paper 47 SIMA Document 2, Colombo, implementation of ecological and Sri Lanka. environmental development programs. Carlson, J.C., Byrd, B.D., Omlin, F.X. Simultaneously, awareness creation (2004) Field assessments in western educational interventions must be undertaken Kenya link malaria vectors to so that the use of ecological control environmentally disturbed habitats measures, mainly focusing on eliminating during the dry season. BMC Public breeding sites and personal protection Health 4: 33. through the use of ITN are effectively Carrara ,C., Petrarca, V., Niang, M. and practiced. In depth studies including Coluzzi, M. (1990). Anopheles comparisons between different agro- pharoensis and transmission of ecological zones are needed to assess the Plasmodium falciparum in the Senegal overall importance of irrigation-related River delta. West Africa. Med. Vet. epidemiology of malaria in Ethiopia. Entomol. 4:421-424. Ghebreyesus, T., Haile, M., Witten, K., Getachew, A., Yohannes, A., Yohannes, References M., Teklehaimanot, H. and Lindsay, S.W. (1999). Incidence of malaria Abose, T., Yeebiyo, Y., Olana, D., among children living near dams in Alamirew, D., Beyene, Y., Regassa, L. northern Ethiopia: Community based and Mengesha, (1998). Re-orientation incidence survey. BMJ. 319: 663-666. and Definition of the Role of Malaria Gilles, H.M. and Warrell, D.A. (1993). Vector Control in Ethiopia. Bruce- Chewatt’s Essential malariology. (WHO/MAL/ 98.1085) WHO, Geneva, 3rd edition. Arnold. London. pp.1-37. Himeidan, Y. E., Elbashir, M.I., Rayah, E.A. Ashenafi, Madebo. (2003). Design and water and Adam, I. (2005). Epidemiology of management of irrigation systems to malaria in New Halfa, an irrigated area control breeding of Anopheles in eastern Sudan. Eastern. Mediter. mosquitoes. Case study: Hara irrigation Health J. 11 : 499-504. project, Arba Minch, Ethiopia. M.Sc. Kibret, S., Petros, B., Boelee, E., Tekie, H. Thesis. Wageningen University, (2008) Entomological studies on the Wageningen. The Netherlands. impact of a small-scale irrigation scheme Awulachew, S.B., Merrey, D.J., Kamara, A. on malaria transmission around Zeway, B., Van koppen, B., Penning de Vries, Ethiopia. Paper presented at Symposium F., Boelee, E. and Makombe, G. on Impacts of Irrigation on Poverty and (2005). Experiences and opportunities Environment, Addis Ababa, Ethiopia, for promoting small-scale/micro 26-29 November 2007. irrigation and rainwater harvesting for Klinkenberg, E., McCall, P.J., Hastings, food security in Ethiopia. IWMI I.M., Wilson, M.D., Amerasinghe, F.P, Working Paper 98.Colombo, and Donnelly, M.J.. (2005). Malaria and International Water Management Irrigated Crops, Accra, Ghana. Emerg. Institute. Infect. Dis. 11:1290-1293. Birrie, H., Gemetchu, T., Balcha, F., Bero, Ministry of Water Resources (MoWR). G., Balkew, M. and Girmay, M. (1997). (2002). Water sector development Health risk assessment of planed programme 2002-2016, Volume II: Main irrigation scheme along the Genale Report. Ministry of Water Resource, River, South Ethiopia. Ethiop. J. Health Federal Democratic Republic of Dev. 11: 229-233. Ethiopia, October 2002. Addis Ababa. Bziava, J. and Kruashvili, L. (2002). Mukiama, T.K. and Mwangi, R.W. (1989). Ecological changes of malaria mosquito Seasonal population changes and malaria larvae at the rehabilitation of irrigation transmission potential of Anopheles and drainage systems. In: Boelee, E.,

453

pharoensis and the minor anopheline in Ethiopia? Tropical Medicine and Mwea Irrigation Scheme, Kenya. Acta International Health 10:1274-1285. Tropica 46: 181-189. Yohannes, M. and Petros, B. (1996). Urban Schlichtheerle, M., Wahlgren, M., Perlman, malaria in Nazareth, Ethiopia: H. and Scherf, A. (2000). Methods in parasitological studies. Ethiopian Malaria Research. Third edition Medical Journal 34: 83-91. MR4/ATCC, Manassas, Virginia. Seyoum, A., Balcha, F., Balkew, M., and Gebre-Michael, T. (2002). Impact of cattle keeping on human biting rate of anopheline mosquitoes and malaria transmission around Ziway, Ethiopia. East. Afr. Med. J. 79:485-490. Smith, T., Genton, B., Baea, K., Gibson, N., Narara, A. and Alpers, P. (2001). Prospective risk of morbidity in relation to malaria infection in an area of high endemicity of multiple species of Plasmodium. Am. J. Trop. Med. Hyg. 64 : 262–267. Teklehaimanot, H., Lipsitch, M., Teklehaimanot, A. and Schwartz, J. (2004). Weather-based prediction of Plasmodium falciparum malaria in epidemic-prone regions of Ethiopia I. Patterns of lagged weather effects reflects biological mechanisms. Malar J. 3: Published on line [http: //www.pubmedcentral.nih.gov/redirect3. cgi.]. Tulu, A. (1993). Malaria In: Kloos, H and Zein, Z.A (Eds).The Ecology of Health and Disease in Ethiopia, pp 341-352, .West view Press, Boulder. Tyagi, K. B. (2004). A review of the emergence of Plasmodium falciparum dominated malaria in irrigated areas of the Thar Desert, India. Acta Trop. 89: 227-239. WHO. (2005). Roll back malaria report. Ethiopia, country profile. 6 pp. Worku, S., Girma, T. and Shiferaw, Y. (2005). Therapeutic efficacy of Sulfadoxine/ Pyrimethamine in the treatment of uncomplicated malaria in children. Ethiop. J. Health Dev. 19:11- 15. Woyessa, A., Gebre- Michael, T., Ali, A. and Kebede, D. (2004). An indigenous malaria transmission in the outskirts of Addis Ababa, Akaki Town and its environs. Ethiop. J. Health Dev. 18: 2-7. Yohannes, M., Haile, M., Ghebreyesus, T. A., Witten, K.H., Getachew, A., Byass, P., Lindsay, S.W. (2005). Can source reduction of mosquito larval habitat reduce malaria transmission in Tigray,

454

Manuals KKKEEEYYYSSS FFFOOORRR SSSUUUCCCCCCEEESSSSSSFFFUUULLL How to enhance water efficiency? Minimise excess water use through implementation of efficient irrigation methods, effective IIIRRRRRRIIIGGGAAATTTIIIOOONNN IINNN EEETTTHHHIIIOOOPPPIIIAAA irrigation scheduling, and soil moisture determination and retention. The following practices are

designed to minimize water losses from evaporation, deep percolation and runoff: • Develop a crop demand-dependent irrigation schedule. Why irrigation? • Determine soil type characteristics to estimate effective irrigation application rates, durations Irrigation is one of many important initiatives that can offer and frequencies. For instance, sandy soils may require more frequent but shorter duration opportunities for change. Good access to a whole range of agricultural applications. technical skills and inputs is needed if the benefits are to be • Till the land along the topographic contours to reduce runoff. maximised. • Water early in the morning or evening to reduce evaporation losses. • Higher agricultural yields • Collect stormwater and irrigation runoff in a series of ditches or drains that return the excess water to a • Wider variety of crops in wet and dry seasons storage pond, if appropriate and necessary. • Longer growing seasons, multi-harvesting • Maintain all parts of the irrigation system, routinely inspect all water lines, tanks, valves and pumps for leaks. • Increase agricultural production and enhance food security Keep replacement and repair parts on hand. • Poverty reduction

The amount of water used by the plants depends on the crop type and How to prevent soil erosion? will rise with evapotranspiration as a result of: Soil erosion by water depends on: • High temperatures the slope: steep, sloping fields are more exposed to erosion; • High wind speed the soil structure: light soils are more sensitive to erosion; • Low humidity • High intensity of sunlight the volume or flow rate of surface runoff water: larger or rapid flows induce more erosion. the vegetation cover of the soil. Which water resources are available and can be used? Erosion is usually heaviest during the early part of irrigation, especially when • Rainwater harvesting (tanks or ponds) irrigating on slopes. The dry surface soil, sometimes loosened by cultivation, • Streams, rivers, lakes or pools is easily removed by flowing water. • Shallow or deep groundwater resources, springs

After the first irrigation, the soil is moist and settles down, so erosion is reduced. Newly irrigated areas are more sensitive to erosion, especially in their early stages. There are two main types of erosion caused by water: • Sheet erosion - even removal of a very thin layer or "sheet" of topsoil from sloping land. It occurs over large areas of land and causes most of the soil losses. • Gully erosion - removal of soil by a concentrated water flow, large enough to form channels or gullies. These gullies carry water during heavy rain or irrigation and gradually become wider and deeper. Therefore pay attention to use low flow velocities when applying water, especially at the beginning of the crop vegetation period and at higher terrain slopes.

How to prevent water related disease hazards? Irrigation alters the environment by creating conditions suitable for parasitic vectors, which then spread disease among producers and the wider population. Irrigation can in general increase the risk of water related diseases, like malaria etc. Especially stagnant irrigation water can serve as breeding site for mosquitoes.

Therefore reduce mosquito breeding sites by avoiding water stagnation on irrigation basins, cover open water surfaces of storage ponds or tanks and apply efficient drainage measures.

This leaflet was prepared in the framework of the collaborative project “Impact of Irrigation on Poverty and Environment in Ethiopia” (2004-2007).

Authors: M. Schönerklee, R. Spendlingwimmer, W. Loiskandl, D. Ruffeis

455

Consider the following key issues and natural conditions when using Which irrigation method is appropriate? irrigation for agriculture Each irrigation method has advantages and disadvantages that should be taken into consideration when choosing the most suitable irrigation method. The main factors influencing the selection of the type of irrigation are: Why is water quality important? • Natural conditions (soil type, slope, climte, water quality, water quantity) • Type of crops to be grown The suitability of water for irrigation depends on the amount and the type of salt the irrigation water contains. A high salt content of irrigation water leads to salinization of soils. • Salinity – amount of salt contained in the water (“salt concentrations”) Watering-Can/Bucket irrigation o Electric conductivity This water-efficient method is appropriate for small plots of land, such as vegetable o Total dissolved solids gardens, which are close to the water source. The salinity of water is easily measured by means of an electrical device, it is then expressed in terms of electrical conductivity. The higher the salt concentration of the irrigation water, the greater the risk of salinization.

Salt concentration Electric conductivity Soil salinization risk Restrictions on use Less than 0.5 g/l Less than 800 µS/cm No risk No restriction on its use Furrow irrigation Should be used with appropriate water 0.5 – 2 g/l 800 – 3100 µS/cm Slight to moderate risk management practices Furrows are narrow ditches dug on the field between the More than 3100 Not generally advised for use unless rows of crops. The water runs along them as it moves More than 2 g/l High risk µS/cm consulted with specialists down the slope of the field and is gradually absorbed into

If possible, a full water analysis should be carried out by an analytical laboratory to assess potential impacts on the bottom and the sides of the furrows wetting the soil. the infiltration rate and soil: Furrows are generally used on farms with large uniform fields where long furrows can be formed at regular intervals • Major mineral compounds (Calcium, Magnesium, Sodium, Potassium, Chloride, Sulfate, Nitrate, Carbonate, between the crop rows. Boron, Fluoride) • Calculation of Sodium Adsorption Ratio (SAR)

Basin irrigation How to conserve the soil fertility and avoid salinization? Basins are horizontal, flat plots of land, surrounded by small dykes or bunds which are The irrigation water will move much faster through a soil with large pores (sandy soil) than through a heavy clay totally flooded during irrigation. Flooding should be soil with smaller pores. Soil structure influences how easily water and air can move through the soil (permeability) carried out rapidly so that water spreads quickly and the penetration of roots. across the whole basin and an even water The soil structure can be improved by good practices such as crop rotation. distribution can be achieved. Cycles of wetting and drying improve soil structure whereas cultivation of very Water is normally brought to the basin in earth wet or very dry soils can destroy the structure. Adequate soil cultivation like channels using gravity. tilling is also required to prevent water-logging problems. Where water has to be pumped from the water source it should be pumped to the highest point and then be After irrigation, the water added to the soil is used by the crop or evaporates distributed in channels by gravity. directly from the moist soil. The salt, however, is left behind in the soil. If not removed, it accumulates in the soil; this process is called salinization. Soils that contain a high amount of salt are harmful to plants and therefore reduce soil fertility and agricultural yields. Drip irrigation Furthermore the presence of salts, especially in heavy clay soils, has the effect In drip irrigation the water is led to the field through a of breaking down the soil structure and therefore reducing the soil permeability pipe system. The distribution system within the field to air and water. consists of tubes perforated at regular intervals. In this way water is slowly and directly supplied drop by

drop to the plants. Why is drainage needed? Drip irrigation is most efficient for irrigating individual During irrigation the water infiltrates into the soil and is stored in its pores. plants such as trees or widely spaced row crops like When all the pores are filled with water, the soil is said to be saturated and no vegetables. more water can be absorbed. The water flowing from the saturated soil

downward to deeper layers, feeds the groundwater reservoir. As a result, the groundwater level rises. Following heavy rainfall or continuous over-irrigation, water-logging may occur which means that the groundwater table may even Sub-surface irrigation reach and saturate part of the root zone. If this situation lasts too long, the Water is supplied to the plants from below the surface by controlling the plants may suffer. level of naturally occurring shallow groundwater. The water is made to flow In order to prevent these negative effects, drainage measures are necessary. Excess surface water is removed into open channels where the level is controlled by earth banks. However, through shallow open drains. Excess groundwater is removed through deep open drains or underground pipes. a regular drawing down of the water level is essential to avoid water- Leaching of salts by sufficient percolation water can only be achieved if an efficient drainage system exists and logging of the plots and crop damage. water-logging can be avoided.

456

Discussion on Theme 4: Environmental and Health Impact of Irrigation

Chair: Dr. Akissa Bahri Rapporture: Gayathri Jayasinghe

The chairman for this session introduced the theme and the floor was opened for questions, comments and suggestions.

Questions and Discussions about irrigation as causing malaria spread it would be better to 4.1 Intervention in the development of associate this with improper small scale irrigation development management of irrigation water. will cause some environmental What is your comment on this? impact but their impact may not be significant. Is there a limit (lower) Ans. Yes, it is bad management of for which EIA study may not be water that causes leakages in the required in view of size? systems, but still the problem is due to the introduction of irrigation Ans: The impact is similar, thus there is no difference between 4.4 The papers from Solomon and schemes in terms of size. Focus Kibret are of the same supervisor should be made on strategic and the same area of investigation. approach and the assessment The facts were known like when should be made using simple there was incidence, type of questionnaires that could be used malaria occurance, so by the communities or NGOs. The i) what were the big cumulative effect of a number of differences between the two? small scale schemes should be ii) What was the significance of considered rather than isolated this study? individual schemes. iii) What idea are you forwarding to policy makers Comment: The issue of and the new findings drawn environmental impact should be from these studies? seen by balancing the positives and the negatives without affecting the 4.5 The comment about watershed resource base. Using inputs like management around Fincha where water will be positively they spend 2 million Birr per year; influencing the environment by - where is the soil erosion going producing more food per area! on? Is it in the sugar estate or outside? 4.2 In the two large scale irrigation - What would be the picture of schemes how come that NO3- Fincha if there is no leakage is not monitored? Notable intervention in the form of effect is expected on downstream sugar estate? users. Irrigation schemes are found in the Is it not possible that the soil Malaria belt of Ethiopia. It could organic matter accumulation could have been better to compare the be expected to be stable, given the before and after irrigation vegetation cover from the perennial development situation crop and good management (in the sugar cane schemes)? Comment: Put irrigation water management as one of the 4.3 Whether there is irrigation or non- preventive measures to be irrigation, there is malaria in the recommended in the research prone areas. Instead of talking undertaking.

457

security and economic status. Is it 4.6 What are the environmental possible to integrate the two impacts (negative) or concerns aspects? related to Wonji/Shoa sugar estate given the sugar estate is working 4.9 What is the optimum investment for the last 50 years or more? cost per hectare for SSI in Ethiopia? 4.7 What is the objective of the EIA What is the contribution of on Fincha and Wonji since this investment on flower industries to should have been done before the the national economy? (they also project is implemented? If, what use drip irrigation) is presented is environmental monitoring and evaluation how How are the riparian and can it be carried out in the absence appropriative laws being practiced of baseline data and initial EIA? in the context of the project How do you assess the impact of Koga before the construction is Comment: Although completed? There is integrated environmental issue is a common watershed management being concern that should be looked at done. The downstream impact for ensuring sustainability, one indicated is not based on concrete needs to be clear about whether or data and research. not we need to carry out EIA before and after interventions. Ans: What has been done is analysis (status quo) of existing 4.10 Irrigation and Malaria projects to see the impact of transmission in Ziway area: irrigation on environment. People living near the lake are The watershed management project anyway more exposed than those lags behind the dam. Downstream who are living away from the flooding of the dam is not lake, so how is it possible to considered. conclude that irrigation is responsible rather than the lake? Comment: Environment impact in Koga is an issue of sequencing. It Ans: This condition was taken into is a first measure which integrates consideration and the study watershed management, irrigation, covered villages away from the and storage. Some are delayed lake in 4-5km distance due to capacity, but those Brief discussion was made with the measures are yet to be presenter and satisfactory answers implemented and should not be were given. perceived as failures. 4.11 From the indicators of environmental resilience, in these 4.8 Paper on Wonji and Fincha: farms, Fincha and Wonji, lots of When we say there was no fertilizers are used. Can expect database, does it mean were/are no nitrogen leakage impact on project appraisal documents when downstream users. Has this been the estates were developed? investigated. Paper on “key success factors…..” Clarification: does this mean that Ans: Observed increase in Nitrogen such knowledge/practice(s) is not in the water. known here in Ethiopia? One slide shows a decline in The statement that health impact organic matter in the soil. I assessment should consider not thought in these intensive systems, only diseases, but also health because of the year round crop improvement from improved food cover one would expect to see

458

increased organic matter in the soil. 4.15 It is a known fact that in any intervention in irrigation there is Ans: Problem could be that it is a env impact. Must be some monoculture. But need to look relation to size of sheme and into this. delay of impact. Is there any conditions where EIA is not 4.12 Not clear about the objective of necessary? the EIA. My understanding is that EHIA is done prior to investment. Ans: In-field impact is the same Fincha and Wonji are very old. So regardless of size of scheme, but is what you are doing really env other kinds of impacts may have a monitoring and evaluation? If so, delayed effect, eg fertilizer use. are you doing this because pre For small scale irrigation, total investment EIA for these schemes amout of water abstracted from are not available? the water source can have an impact. 4.13 Not having baseline data is common. Is there no project Ans: Small schemes do have appraisal documents? If we don’t impacts. Yes, economically not have appraisal documents does it feasible to make EIA for every mean we cannot do an impact small scheme. For small assessment? schemens recommendations is to do strategic EIA to see what the Ans: True, but what we did was impact would be of upscaling. trying to assess the status quo of Also for a small scheme rapid impact of existing irrigation assessment methods through schemes. questionnaires…etc may be more appropriate rather than a full scale 4.14 Koga, is the first project of large EIA. scale scheme for smallholder irrigation. Scheme under mid Ans: has the experience that in construction. SO how did you the regions, consultants have to reach the conclusion that there are include EIA. They are there but environmental impacts? content is sometimes weak. One aspect of Koga is watershed management. And Environmental Ans: don’t ignore the cumulative authority is a member of the effects. steering committee. 4.16 Although I personally believe in Ans: EIA was conducted but this EIA. I have difficulty in didn’t consider downstream understanding if this EIA can be impacts of change in flow regime. done before implementation. Possible impact on fisheries. 2. Need basic understanding of EIA. EIA made recommendations about a lot of things that should Ans: EIA is a means of preventing be done before the dam is failure rather than costly repairs. constructed. Eg water shed Learning from past lessons rather conservation work that should than re-inventing the wheel. have been done before the dam is constructed. Although these have 4.17 Assessment of the health and commenced, not progressed as Env/: we tend to be biased by planned. The organization that is looking at impact in terms of supposed to make sure the diseases. How can we balance recommendations are carried out the negative of the diseases vis a is not pushing hard enough to vis the food security impact on make sure these are implemented. health.

459

Where does the soil erosion Ans: Absolutely true. Looking at actually occure? Is within the +ve and _ve is the best. What sugar estate or around the area happens now is that assumptions where there is deforestation? are made about +ve impacts but What would happen if the sugar these are not quantified. Useful estate was not there? Generally, as far as possible if impacts are unplanned interventions can have quantified in a way that they can devastating effect on ecosystems. be validated and trade-offs For example, in lake Tana where assessed. we thought there were large areas of wetlands and actually Ans: Koga is the first experience intensively cultivated. So, in the country to integrate whether there is a scheme or not, watershed management to reduce degradation is going to happen sedimentation in the dam. Not anyway. So it is better that complete ignorance but delay in studies compare impacts of implementation of certain planned vs unplanned use. components due to complex problems. In Ethiopia it is not Comment: Most of our irrigation water development that has schemes are in Malaria belt. contributed to degredation but the Better to compare before-after absence of it. Important to situation. reverse this. 4.20 Is it irrigation or type of crop that Ans: the koga study shows is related to increased Malaria? lessons on how to follow up on There is a recent discussion on EIA s that are done. maize contributing to the increase of Malaria. Can change in crop Comment: It seems there is a management mitigate malaria? North South conflict of priorities. What are the policy ? Advisors from WB “promote recommendations we can make? eco-agriculture” for example. Is this really appropriate for Comment: With or without Ethiopia? irrigation there is malaria. Try to find out the impact of quality of 4.18 Skeptical of the results because management of scheme on villages near the lake are more Malaria. It is preferable to say susceptible than villages farther that bad water management away. How did you attribute the causes fertile ground for malaria high malaria to irrigation rather breeding. Effective irrigation will than the lake? One alternative is not produce impounding, proper water management to seepage, etc so relating to prevent malaria. Advise management sounds better. inclusion of this point in the recommendations. Ans: Proximity of the study villages to lake zeway 4-5km a Ans: Both control and study little beyond a common flight villages were at the same distance range of arabiensis. Both villages from the lake and 4-5 km away are equi-distant from lake. There which is more than the average are also “buffer” villages between flight range of the arabiensis Ziway and the study villages that vector. deflect vectors from the study villages. 4.19 Watershed management is Canal leakage and over-irrigation practiced by Fincha with a has been observed in the study spending of about 2m per year. sites as ‘bad’ management.

460

Group discussion and findings

After the presentation of the four themes i.e. statues quo analysis, characterization and assessment of performance of irrigation in Ethiopia; irrigation impact on poverty and economy; irrigation institutions and support services; and environmental and health impacts of irrigation development. Participants were sub divided into four groups (working group I to IV) according to the above themes. The outcome of their discussions is shown in the group work presentations as indicated in the home page.

Way Forward and Synthesis Plan

Dr. Deboorah Bossio made a summary of the symposium and discussion points pertinent for the way forward. Her presentation is included in the paper presentations shown in the home page. Intensive discussions and recommendations were provided by the participants based on the following two leading questions

What impacts would we like to have? Who could make it happen?

Exhibition

Under the poster exhibition displays of posters related to the project and supplementing some of the papers presented in the symposium by the respective authors and the overall description of the project. A total of 11 posters were exhibited and demonstrated by authors as shown below.

Table: Posters presented Name Poster title Institution Abdu Beshir Analysis of irrigation systems using comparative Oromia Water Works performance indicators: A case study of two large Design Enterprise scale irrigation systems in the upper Awash River Basin

Abonesh Tesfaye The Impact of Small Scale Irrigation on Household MoWR Food security: The Case of Godino and Filtino Irrigation Schemes in Ada Liben District Ahmed Amdeyehun GIS and Remote sensing Integrated Environmental Impact Assessment of Irrigation Project in Finchaa Valley Aster Yilma Irrigation potential of Ethiopia IWMI-NBEA Belete Bantero Hare River Irrigation close to Arba Minch has World Vision become a way out for Many Households from Poverty to Prosperity Dr. Seleshi Bekele Impact of Irrigation on Poverty and Environment: IWMI-NBEA Research Project Yihenew Malaria Transmission In Ziway, Eastern Oromia Zone, As Related To Irrigation Development Wallner K. a Field Parameter Evaluation to Support a Department of Water, Environmental Impact Analysis of Irrigation in Atmosphere and Ethiopia Environment, Institute of Hydraulics and Rural Water Management, BOKU, Austria Judit Christine Impact of irrigation on livelihood and food security University of Natural

461

in the modern Hare river irrigation scheme in Resources and Applied Southern Ethiopia Life Sciences, Vienna- Austria

Closing

The symposium was officially closed by Ato Abara Mekonnen, Chief Engineer of the Ministry of Water Resources. In his closing remark he emphesized the importance and value that this research will add to irrigation development in Ethiopia.

462

List of Participants

Abdu Beshir Ademe Teferi (Head of Planning and Water Resource Engineer programming) Ministry of Water Resource Benishangul Gumuze BoARD P.O.Box 5744 P.O.Box; 30 A.A, Ethiopia Assossa, Ethiopia Tel: 0911-19 05 59 Tel; 0911-96 89 87 E-mail: [email protected] Fax 057-775 07 26

Abera Girma Abebe Adinew Abate Wonji-Shoa Sugar Factory Kobo Girana Valley Dev. Program office P.O.Box 51 P.O.Box; 182 Wonji, Ethiopia Woldia, Ethiopia Tel: 0911387216 Tel; 033-334 00 89, 0918-34 01 72 E-mail: [email protected] Fax 033-334 04 74 E-mail: [email protected]

Abera Mekonnen Afework G/Kidan(Irrigation Eng. Dept. Ministry of Water Resources Head) P.O.Box 5744 Fincha Sugar Factory A.A, Ethiopia P.O.Box; 5734 Tel: 0116-62 55 06 Fincha(Wolega), Ethiopia E-mail: [email protected] Tel: 057-664 10, 0911-39 70 63

Alemayehu Mengiste (Managing Abonesh Tesfaye (Economist) Director) Ministry of Water Resources Concert Engineering P.O.Box 11490 P.O.Box; 7553 A.A, Ethiopia A.A, Ethiopia Tel: 0911-19 15 58 Tel; 0116-63 92 44, 0116-63 92 45 E-mail: [email protected] E-mail: [email protected] Abraham Asmare (Manager) Alessandro Dinucci WVE WFP P.O.Box; 3330 A.A, Ethiopia A.A, Ethiopia Tel; 0911-75 43 62 Tel: 0116-29 33 50 E-mail: [email protected] E-mail: [email protected] Amhed Amdihun AAU Abraham Woldemichael A.A, Ethiopia Hawassa University Tel; 0911-48 47 21 P.O.Box; 5 E-mail: [email protected] Hawassa, Ethiopia

Tel; 0916-82 36 43

Fax 046-220 54 21

E-mail: [email protected]

Andegna Shinale (Office head)

Arba Mich District BoARD

P.O.Box:24

463

Arba Minch, Ethiopia Tel 0577-75 22 50, 0913-25 80 64, Tel; 046-881-01 22, 0911-95 56 28 0577-75 00 60 Fax 046-881-01 43 E-mail [email protected]

Asfaw Negassa (Dr.) Belay Demissie ILRI Agricultural Economist P.O.Box:5689 USAID A.A, Ethiopia P.O.Box 1014 Tel: 0911-74 06 99 A.A, Ethiopia E-mail:[email protected] Tel 0115-51 00 88 E-mail: [email protected] Asmelash Birhane MU Belete Bantero Mekelle,Ethiopia UN-Industrial Dev't Organization Tel 0914-72 65 02 0344-40 93 04 P.O.Box 23248 E-mail:[email protected] A.A, Ethiopia Tel 0115-54 45 87, 0911-86 16 95 Asnake Abera E-mail: [email protected] Food Security Expert EC Deligatgion to Ethiopia Benhard Meier A.A, Ethiopia Regional Director Tel 0116-61 25 11 German Agro Action E-mail [email protected] P.O.Box 1866 A.A, Ethiopia Asrat Melaku Tel 0116-62 47 65, 0116-62 47 31 ORDA 132 E-mail: [email protected] Bahir Dar, Ethiopia E-mail [email protected] Berhanu Adinew EEA Aster Denekew Ms P.O.Box 34282 P.O.Box IWMI 5689 A.A, Ethiopia A.A, Ethiopia Tel 0114-16 21 21, 0114-16 09 67 Tel 0116-17 21 94, 0911-60 19 94 E-mail: [email protected] E-mail: [email protected] Berhanu Gebremedhin Ayalew Abate ILRI, IPMS Water Bureau Engineering P.O.Box 5689 P.O.Box 5673 A.A, Ethiopia A.A, Ethiopia E-mail: [email protected] Tel 0116-63 70 34, 0911-46 54 80, 0116-61 07-10 Bob Yoder E-mail: [email protected] Technical Director IDE Bahri Akissa P.O.Box 7892 Director IWMI A.A, Ethiopia Accra, Ghana E-mail [email protected] Tel 0114-16 26 40 E-mail: [email protected] Belay Bancha Burke Vindevoghtel W.S.S study & design team leader Oxfam Canada, A.A, Ethiopia Benishangul Gumz Region Water, Mine Tel 0911-40 56 10 & Energy Resources Dev't Bureau E-mail: [email protected] P.O.Box 51, Asossa, Ethiopia

464

Dawit Kelemework Haramaya University P.O.Box 115, Dire Dawa, Ethiopia Dr. Dejene Aredo Tel 0911-61 78 70 Associate Professor E-mail: [email protected] AAU P.O.Box 150008, A.A, Ethiopia Deborah Bossio (Dr). Tel 0911-22 77 80 IWMI E-mail: [email protected] P.O.Box 2075, Sri Lanka Tel 0094-112-787-400 Dr. Solomon Seyoum E-mail: [email protected] Water Resources Specialist NBI-WRPM Degu Teressa P.O.Box 60173, A.A, Ethiopia Water Use Expert Tel 0116-46 70 11, 0912-05 50 08, Indris Scheme (tokke Kuttaye Irrigation 0116-46 70 14 Office) E-mail: [email protected] Guder, Ambo, Ethiopia Tel 0112- 82 04 05, 0911-38 06 90 Endeshaw Goshu Golgota Irrigation Scheem Dejene Abesha Merti, Ethiopia Dept. Head MoARD Tel 224530390, 0912-02 93 06 P.O.Box 21855, A.A, Ethiopia Tel 0115-15 68 04 , 0115-54 68 04 Enyew Adgo (Dr) E-mail:[email protected] ARARI P.O.Box 527, Bahir Dar, Ethiopia Dejene Kebede Tel 0918-76 56 21, 058-220 51 74 Program Manager E-mail: [email protected] Austrian Embassy Development Cooperation Etafa Emama P.O.Box 11553, A.A, Ethiopia MoWR Tel 0115-53 38 28, 0115-55 38 31 P.O.Box 5744, A.A, Ethiopia E-mail: [email protected] Tel 0116-61 11 11, 0911-12 17 25, 0116-61 08 85 Desta Bayena E-mail: [email protected] FC-Consultant Oromia KfW-SUN, A.A, Ethiopia Fekadu Fufa Tel 0114-42 10 89, 0911-90 19 65, Researcher (PhD) 0114-42 10 89 Institute of Biodiversity Conservation E-mail: [email protected] (IBC) A.A, Ethiopia Dominik Ruffeis Tel 0913-03 17 30 BOKU muthgasse E-mail: [email protected] P.O.Box 18, Austria Tel 0911-73 98 81 Fekahmed Negash E-mail:[email protected] Ministry of Water Resource P.O.Box 5744 A.A, Ethiopia Dr. Awad Ghanem Tel 0911-67 92 44, 0116-63 70 38 MoARD GTZ/CiM E-mail: [email protected] P.O.Box 91, A.A, Ethiopia Tel 0911-25 01 10, 011-61 30 71 Fentaw Abegaz Dr. E-mail: [email protected] EIAR

465

P.O.Box 2003 Tel 0116-17 21 92, 01116-17 20 01 A.A, Ethiopia E-mail: [email protected] Tel 0116-46 01 34, 0911-88 64 05, 0116- 48 94 12 Getachew Alem E-mail:[email protected] Private Consultant P.O.Box 30361 Feyera Abdi A.A, Ethiopia Executive Director Tel 0113-71 35 07, 0911-22 33 46 SOS Sahel - UK P.O.Box 3262 A.A, Ethiopia Getachew Alemayehu(Dr). Tel 0114-16 03 91/ 0114-16 02 78, 0911- ARARI 20 88 38, 0114-16 02 88 P.O.Box 527 Bahir Dar, Ethiopia Fitsum Hagos (Dr.) Tel 0918-76 26 64 0582-20 51 74 IWMI E-mail: [email protected] P.O.Box 5689 A.A, Ethiopia Getachew Haile Tel 0116-17 21 91, 0911-55 96 16, 0116- REST 17 20 01 P.O.Box 222 E-mail:[email protected] Mekelle,Ethiopia Tel 0344-41 08 20, 0914-70 65 62 Fitsum Tesfaye Lecturer Hawassa University Getachew Tikubet P.O.Box5 Director Hawassa, Ethiopia YINRM Tel 0911-71 34 78 P.O.Box 3893 E-mail: [email protected] A.A, Ethiopia Tel 0911-25 23 37 G/Hawaria G/Egziabher E-mail: [email protected] Noewegian University of life Sciences, IOR Gete Zeleke (Dr.) P.O.Box 5003, N1432 AS Global Mountain Program AS, Norway P.O.Box 5689 Tel 0047-952-302-27 A.A, Ethiopia E-mail: [email protected] E-mail: [email protected]

Gashaye Chekole Gezahegn Alemu Head, Water Resources Dev't Division JICA Ethiopian Orthodox Church Dev't and P.O.Box 5384 Inter church Aid Commission (EOC- A.A, Ethiopia DICAC) Tel 0115-50 44 65 A.A, Ethiopia E-mail:[email protected] Tel 0911-33 67 69, 0111-55 14 55 E-mail: [email protected] Gezahegne Ayele Senior Scientist Gayathree Jayasinghe EDRI Researcher (Biometrician) P.O.Box 2479 IWMI A.A, Ethiopia P.O.Box 5689 Tel 0911-21 69 35 0115-50 55 85 A.A, Ethiopia E-mail: [email protected]

466

Girma Tadesse(Dr) Judith Christie ILRI BOKU P.O.Box 5689 Vienna, Austria A.A, Ethiopia Tel 0043(0)650-8811223 Tel 0116-60 37 86 E-mail: [email protected] E-mail:[email protected]

Girma Yosef Kalkidan Assefa Head, Department of Agriculture IPMS/IWMI ESTC P.O.Box 101266 P.O.Box 2490 A.A, Ethiopia A.A, Ethiopia Tel 091171 77 43 Tel 0115-53 49 44, 0115-52 44 00 E-mail: [email protected] E-mail: [email protected] Katrien Descheemacker(Dr.) Godswill Makombe(Dr.) ILRI-IWMI IWMI P.O.Box 5689 P.O.Box 5689 A.A, Ethiopia A.A, Ethiopia Tel 0913-40 33 12 Tel 0116-46 32 15, 0911-68 40 82, 0116- E-mail: [email protected] 46 46 45 E-mail: [email protected] Kebede Thehayu GIS consultant FAO-Nile Project Habte Honeligne P.O.Box 521 Irrigation Agronomist Amhara Regional Kampala, Uganda Bureau of Agriculture P.O.Box 437 E-mail: [email protected] Bahir Dar, Ethiopia Tel 058-220 73 77, 0918-78 79 82,058- Kifle Abegaz 220 15 10 Program Officer II E-mail: [email protected] CRS-Ethiopia P.O.Box 6592 Habtemariam Kassa A.A, Ethiopia CIFOR Tel 0911-61 99 36, 0114-6544 50 P.O.Box5689, A.A, Ethiopia E-mail: [email protected] Tel 0116-46 32 15, 011-46 46 45 E-mail:[email protected] Hailemariam Hailemeskel Kifle Amene Agricultural Economist Irrigation Team Leader ADB Agrigulture & Rural Dev't P.O.Box 25543 code 1000 Bishoftu A.A, Ethiopia Tel 0911-77 05 94, 0114-33 17 21 Tel 0911-48 09 43 E-mail: [email protected] Kim Geheb IWMI Ibrahim Ahmed P.O.Box 5689 MoARD A.A, Ethiopia P.O.Box 62347 E-mail: [email protected] A.A, Ethiopia Tel 0115-15 78 64 Lakech Haile E-mail:[email protected] Department Head

467

Ministry of Water Resource E-mail: [email protected]; P.O.Box 335243, A.A, Ethiopia [email protected] Tel 0911-19 40 94 E-mail: [email protected] Menkir Girma Irriggation Engineering (Expert) Lemessa Mekonta Upper Awash Agro Industrhy Enterprise Head, Water Resources Study & Design Dep't P.O.Box 12624 Oromia Water Resources Bureau Merti Jeju, Ethiopia P.O.Box 21685 CODE 1000 Tel 0221-12 17 05, 0221-12 27 03 A.A, Ethiopia Tel 0115-52 70 68, 0911-44 25 31 Mesfin Shiferaw E-mail: [email protected] Water Management Expert Ade'a Irrigation Dev't Office Leonhard Moll P.O.Box318, Akaki Beseka Austrian Embassy Development Debre Zeite, Ethiopia Cooperation, A.A, Ethiopia Tel 0911-47 18 00, 0114-33 17 21 Tel 0115-53 38 28 E-mail: [email protected] E-mail: [email protected] Michael Abebe Luaskandl Willibald Head, Dams & Hydropower Des. University of Bodenhultus Wien Department Vienna, Austria Ministry of Water Resource Tel 0043(1)36006-5499 P.O.Box 17598 A.A, Ethiopia E-mail: [email protected] Tel 0911-60 45 45 E-mail: [email protected] Makonnen Loulseged IWMI Michael Menkir Mr. P.O.Box 170121, A.A, Ethiopia IWMI Tel 0116-17 21 48, 0116-17 20 01 P.O.Box 5689 E-mail: [email protected] A.A, Ethiopia Tel 0116-17 22 49, 0911-11 19 99 Fax Matthew McCartney (Dr.) 0116-17 20 01 IWMI 5689 E-mail: [email protected] A.A, Ethiopia Tel 0116-17 22 46, 0116-12 20 01 E-mail:[email protected] Mekonnen Ayana Million Alemayehu AMU Head, Liason Office-ORDA P.O.Box 21 ORDA Arba Minch, Ethiopia P.O.Box8122 Tel 046-8810453, 0916-83 10 51, 0468- A.A, Ethiopia 81 02 79 Tel 0115-5044 55/54 E-mail:[email protected] Fax 0115-517244 E-mail: [email protected] Melaku Mekonnen Metaferia Consulting Engineers Mitiku Bedru P.O.Box 3192 Deputy bureau head for Irrigation Dev't A.A, Ethiopia Sector, SNNPRS Tel 0115-51 71 93, 0911547808, 0115- SNNPR Water Res. Dev't Bureau 51 44 66, 0911-48 72 84 Irrigation Dev't Sector

468

P.O.Box 153 Rebeka Amha Hawassa, Ethiopia Monitoring & Evaluation Assistant Tel 046-220 91 70 046-221-00 20 ILRI, IPMS Fax 046-220-20 37 P.O.Box 5689 E-mail: [email protected] A.A, Ethiopia Tel 0911-63 43 60 Mohamed Badel Mohamed E-mail: [email protected] SORPARi P.O.Box 398 Rehel Deribe Jigjiga, Ethiopia MSc. Student Tel 0915-74 48 22 AAU Fax 0257-75 31 27 P.O.Box 4991 E-mail: [email protected] A.A, Ethiopia Tel 0911-15 39 25 Monika Jaeber E-mail: [email protected] University of Innsbruck Kenwelscheiben Rieckh Helene P.O.Box 21, 6173 Oberperpuss University of Bodenkuetur Innsbruck, Austria Vienna, Austria E-mail: [email protected] E-mail: [email protected] Mulatu Daba Golgota Irrigation Scheem Robel Lambisso P.O.Box 3334 Program Coordinator Merti, Ethiopia WVE Tel 022-119-08 43, 0912-03 76 30 P.O.Box 3330 A.A, Ethiopia Mulugeta Dejene Tel 0911-18 66 51 Plant production & protection desk head Fax 0116-29 33 46 E-mail: [email protected] Arba Mich District BoARD P.O.Box 24 Schneider Jean Arba Minch, Ethiopia Prof. Dr. ChairIAG-BOKU VIENN Tel 046-881-01 22 0916-83 22 57 Vienna, Austria Fax 046-881 01 43 0043147654/5401, 5409 E-mail: [email protected]

Nadia Manning Schonerklee Monika DI, MSc Researcher / outreach coordinator Austrian Research Centers Seibersdoaf IWMI Seibersdoaf, Austria, Vienna P.O.Box 5689 Tel 0043-664-62 07 637 A.A, Ethiopia Fax 0043-50 550-3452 Tel 0116-46 32 15 E-mail: [email protected] E-mail: [email protected]

Phillippe Lemperiere Shimelis Dejene BRLi/ENTRO Development Studies Lecturer (MA) A.A, Ethiopia Jimma University, Ambo College Tel 0911-76 05 36 Ambo, Ethiopia E-mail: [email protected] Tel 0911-30 48 50 E-mail: [email protected]

469

Fax 01 607 255 40 80 Sihin Tekle E-mail: [email protected] Researcher Debre Zeit Agricultural research Center Taye Duressa P.O.Box32, Debre Zeite, Ethiopia Irrigation & Drainage Enginer Tel 0114-67 19 00 (Res.), 0913-10 74 00 WWDSE E-mail: [email protected] A.A, Ethiopia Tel 0116-63 18 96, 0911-31 82 28 Sintayehu Asfaw PR Officer (Public Relation) MoWR Tegegne Sishaw P.O.Box 40955 Haramaya University A.A, Ethiopia P.O.Box 189(138) Tel 0116-62 63 24 0911-13 97 32 Dire Dawa, Ethiopia Fax 0116-61 08 85 Tel 0915-74 20 60 E-mail: [email protected] E-mail: [email protected]

Solomon G/Selassie Temesgen Birhanu ILRI MoFED P.O.Box5690 A.A, Ethiopia A.A, Ethiopia Tel 0911-07 23 01 Tel 0116-46 32 15 E-mail: [email protected] E-mail: [email protected] Tena Alamirew(Dr.) Solomon Kibret Haramaya University Independent P.O.Box 138 P.O.Box 14001 Dire Dawa, Ethiopia A.A, Ethiopia Tel 0915-33 04 71 Tel 0911-65 29 78 Fax 0255-53 03 31 E-mail: [email protected] E-mail: [email protected]

Spendling-Wimmer Robert (Dr.) Teressa Alemayehu Austrian Research Centers Seiberdou Water Use Expert Seibersdoaf, Austria, Vienna Indris Irrigation Scheeme Tel 0043-664-6207654 Guder, Ambo, Ethiopia E-mail: Tel 0112-82 04 05, 0911-97 67 60 [email protected]

Tesfaye Zeleke Lecturer Svat Matula Prof. Axum University CzechUniversity of Life Science Axum, Ethiopia Prague, Czech Republic Tel 0911-30 21 70 Tel 00420-23438 4636 E-mail: [email protected] Fax 00420-23438-1835 E-mail: [email protected] Teshome Asmare Expert Amhara Regional Bureau of Tammo Steenhuns Prof. Agriculture Cornell University Bahir Dar, Ethiopia Iuhaca NY Tel 058-220 46 16 USA Fax 058-220 29 69 Tel 01607 255 4080

470

Teshome Atnafe Mekelle, Ethiopia Head, Irrigation and Drainage Dev't Tel 0912-04 86 69 Studies Department E-mail: [email protected] Ministry of Water Resource P.O.Box 100894, A.A, Ethiopia Woldeab Teshome (Dr.) Tel 0911-61 12 38 AAU Fax 0116-63 70 48 P.O.Box 150197, A.A, Ethiopia E-mail: [email protected] Tel 011-167 34 13 E-mail: [email protected] Teshome Lemma Program Manager Wubishet Yilma Millinium Water Program D/G/Manager P.O.Box4710 Middle Awash Agri. Dev't Ent. A.A, Ethiopia M/Sedi, Ethiopia Tel 0911-41 93 64 Tel 022-456-01 45, 0911-20 18 95 Fax 0116-18 32 95 Fax 022-459-01 39 E-mail: [email protected] Yihenew Alemu Tilahun Amede AAU ILRI-IWMI P.O.Box 40762 P.O.Box5689 Jimma & A.A, Ethiopia A.A, Ethiopia Tel 0911-30 43 89 Tel 0116-46 32 15 E-mail: [email protected] E-mail: [email protected] Yodit Balcha Tilahun Hordofa(Dr.) IWMI EIAR P.O.Box 5689, A.A, Ethiopia P.O.Box 2003 Tel 0911-88 01 41 Nazareth, Ethiopia E-mail: [email protected] Tel 022-111 14 49, 0911-84 24 92 Fax 022-111 46 22/23 Yohannes Belay E-mail: [email protected] German Agro Action P.O.Box 1866, A.A, Ethiopia Tirufat Hailemariam Tel 0116-62 47 31 BOKU E-mail: [email protected] Vienna, Austria Tel 06991-01-61 660 Yohannes Geleta E-mail: [email protected] Design Enginer OromiaIrrigation Dev't Wagnew Ayalineh A.A, Ethiopia Research Assistant Tel 0114-67 15 03, 0911-98 16 65 ILRI E-mail: [email protected] P.O.Box 5689, A.A, Ethiopia Tel 0116-613215 Zachar, Easton (Dr.) E-mail: [email protected] Cornell University USA Wiebke Foerch Tel 607-255-2463 University of Arizona GTZ-SUN E-mail: zme2@cornell

471

Partial View of Symposium Participants

472