Agricultural Water Management Delivers Returns on

Agricultural Water Management Delivers Returns on Investment in Africa A Compendium of 18 Case Studies from Six Countries in Eastern and Southern Africa

Coordinating Lead Author Bancy M. Mati Contributing Authors: Enyew Adgo, Firew Akalu, Geophrey Kajiru, Jackson Nkuba, Wellington Mulinge, Andriantahina Rakotondralambo, Davie Kadyampakeni, Desire Kagabo, Jean Nsabimana, Alexandra Ravelombonjy, Samson Kazombo-Phiri, Josephine Thome

September, 2010

Preface This publication is the culmination of three years of research conducted concurrently across six countries, namely; Ethiopia, Kenya, Madagascar, Malawi, Rwanda and Tanzania, in Eastern and Southern Africa (ESA), between 2007 and 2009. The study sought to identify agricultural water management (AWM) technologies and practices at project-scale in each country, which have scope for out-scaling regionally. The technical viability and economic returns of the interventions at farmer level were also quantified. Conducted by paired country-research teams, 18 project sites (three study sites per country) were assessed and some 12 AWM technologies and practices identified. These are presented here as 19 scientific papers co-authored by 16 researchers. This publication targets senior decision makers who include; researchers, development partners, policy makers, managers of programmes and projects, extension workers, public and private practitioners of AWM. It is meant to inform, educate and enhance the reader’s knowledge on the positive aspects of AWM development and investments in ESA, to stimulate supportive action and investment in the sector. This publication has 19 Papers, of which Paper 1 is the Regional Synthesis, while the other 18 Papers are case studies from each of the field sites assessed. Acknowledgements The authors wish to thank the International Fund for Agricultural Development (IFAD) which provided financial support, the Improved Management of Agricultural Water in Eastern and Southern Africa (IMAWESA), International Crops Research Institute for Semi-Arid Tropics (ICRISAT) and the Association for Strengthening Agricultural Research in Eastern and Central Africa (ASARECA) and all the institutions and individuals who provided information logistical and material support to the research work. The views expressed in this publication do not necessarily reflect those of IFAD nor any other institution and are the sole responsibility of the authors.

Table of Contents 1. Agricultural Water Management Delivers Returns on Investment in Eastern and Southern Africa: A Regional Synthesis................................ 1 2. Effects of Long-Term Soil and Water Conservation on Agricultural Productivity: A Case Study of Anjenie Watershed, Ethiopia............. 42 3. Water Harvesting with Geo-membrane Lined Ponds: Impacts on Household Incomes and Rural Livelihoods at Minjar Shenkora, Ethiopia ........................................................................................................ 59 4. Comparing the Productivity of Field Crops under Irrigated and Rainfed Systems at Sewur Irrigation Scheme, Ethiopia........................ 74 5. Assessment of Runoff Harvesting with ‘Majaluba’ System for Improved Productivity of Smallholder Rice in Shinyanga, Tanzania . 90 6. Impacts of Mulching on Productivity of Maize and Beans in Smallholder Farming Systems of Kagera Region, Tanzania...............104 7. Effects of Ridging on Crop Productivity in Shinyanga Tanzania.....118 8. Improving Crop Productivity with the System of Rice Intensification (SRI) in Tsivory, Madagascar ..................................................................132 9. Impacts of Permanent Soil Cover on the productivity of rainfed rice farming systems in Alaotra Lake Region, Madagascar ........................147 10. Assessment of the Productivity of Rice in the Cascade System of Andranomanelatra Highlands, Madagascar...........................................163 11. Water Harvesting with Small Storage Ponds: Impacts on Crop Productivity and Rural Livelihoods in Lare, Kenya.............................173 12. Improving Crop Productivity and Farmer Incomes with Sprinkler Irrigation at Sagana-Maganjo, Kenya .....................................................189 13. Impacts of Long-term Soil and Water Conservation on Agricultural Productivity in Kaiti Catchment, Kenya ...............................................203 14. Effects of Off-Season ‘Dambo’ Irrigation on Crop Production and Farmer Incomes at Kamalambo, Malawi ..............................................217 15. Impacts of Treadle Pump Irrigation on Crop Productivity in Mchinji, Malawi .........................................................................................226 16. Assessment of the Productivity of Field Crops with Regulated Surface Irrigation in Domasi, Malawi ....................................................236 18. Radical Terracing and Hedgerow Practices for Improved Potato Production in Buberuka, Rwanda...........................................................247 18. Improving Rice Productivity with Water Control at RusuriRwamuginga Marshland, Rwanda...........................................................261 19. Performance of Farmer-Managed Valley Bottom Irrigation for Rice Production in Cyabayaga, Rwanda .........................................................267 i

List of Acronyms AWM ANAE ASARECA CA CBA ESA FGP ICRISAT IFAD IMAWESA IRR KARI NPV PSC ROI RWH SPSS SRI SSA SWC

Agricultural water management Association Nationale d’Actions Environnementales Association for Strengthening Agricultural Research in Eastern and Southern Africa Conservation Agriculture Cost-benefit Analysis Eastern and Southern Africa Farm Gate Prices International Crops Research Institute for Semi Arid Tropics International Fund for Agricultural Development Improved Management of Agricultural Water in Eastern and Southern Africa Internal Rate of Return Kenya Agricultural Research Institute Net Present Value Permanent Soil Cover Return on Investment Rain Water Harvesting Statistical Package for Social Scientists System of Rice Intensification Sub Saharan Africa Soil and Water Conservation

ii

1. Agricultural Water Management Delivers Returns on Investment in Eastern and Southern Africa: A Regional Synthesis Mati, B. M. Challenges facing agricultural water management in ESA Poverty reduction efforts face the greatest challenge in the rural areas of Sub-Saharan Africa (SSA), where over 70% of the populations live. This challenge is more pertinent in the Eastern and Southern Africa (ESA) region, which has some 85% of its rural population being poor people engaged in smallholder agriculture. Three core assets for these people include land, water, and human capital, albeit farmlands are small and fragmented. At continental level, 83% of the arable land in Africa (NEPAD, 2003) is prone to soil and/or water related limitations and negligible investment in agricultural water management (AWM). The World Development Report (2008) states that only 4% of Africa's annual renewable water resources have been developed for agriculture, water supply and hydropower use, compared to 70-90% in developed countries. Moreover, the average water storage capacity in Africa is about 200 m3/person/year compared to 5,961 m3/person/year in North America (Global Risk Network, 2008). Thus, the much hyped statement that 70% of water in the region is used for irrigation is a misnomer, because it considers only the amount of water mobilized from developed infrastructure. The real problem is that tthere has been minimal development of inftrastructure. Currently, water withdrawals for agriculture account for 3.6% of total renewable water resources in Africa and if all land suitable for AWM were to be developed, it would consume only 12% of the “available water” (FAO, 2003; Darghouth, 2007). Thus, the region experiences severe economic water scarcity, caused by lack of human and infrastructure capacity to satisfy water demand (Molden et al., 2006). Therefore, both economic and physical water scarcity pose special problems especially to smallholder agriculture in ESA.

1

The management of available water and land resources in the ESA region poses another constraint. The combination of high evaporation losses compounded by the rainfall unpredictability and tropical storms results in substantially lower percentage of precipitation contributing to renewable water resources. This means that most of the rainfall is lost as surface evaporation and runs as flash floods into swamps, rivers, lakes and saline sinks before it is used for agricultural production (Hatibu et al., 1997; van Koppen, 2002). These losses result in only about 19.7% of rainfall becoming renewable water, which is equivalent to about 150 mm per year (ECA, 1995; FAO, 1995). These effects are manifest in crop failures, low agricultural productivity and rural poverty. Moreover, the subsistence-based agronomic practices of smallholder farmers have resulted in the “mining” of the natural resource base, due to the need to produce more from the same land. Since land is inelastic, innovative ways that enable higher productivity need to be adopted to meet the growing food gap. Otherwise, left to natural forces, smallholder agriculture will continue to under-perform, impacting on livelihoods. However, it is an established fact that investments in water for agriculture have made a positive contribution to rural livelihoods, food security, and poverty reduction (Molden et al, 2007). Another dimension is climate change, which is expected to have the greatest impact on the rural poor reliant on rainfed agriculture in SSA. Over the last three decades, farmers have faced weatherrelated events that seem to get increasingly unreliable. These result in drought, prolonged dry spells, erratic rainfall and sometimes floods. The average incidence of serious drought has increased from around seven serious droughts during the period 1980-1990 to 10 during the period 1991 to 2003 and drought-induced crop failures are common in the region (FAO, 2005). Africa as a whole is warmer by 0.5°C than it was 100 years ago, putting extra strain on water resources. The six warmest years recorded in Africa have all occurred since 1987 (Global Risk Network, 2008). There is broad agreement that one of the biggest climate change impacts will be on 2

rainfall, making it more variable and less reliable (Lenton and Muller, 2009). Thus, adaptive strategies will be needed to improve the availability of water through storage, and better water management leading to increased demand for AWM for disaster response and management. Adaptation and mitigation measures will require more investment in terms of financial, technological, human and social capital, as there is an element of ‘overcoming nature’ in upgrading smallholder agriculture in ESA. Natural resource base of the ESA Region The Eastern and Southern Africa (ESA) region is a geographic as well as economic block, having diverse yet certain common features. In the context of this study, ESA refers to a block of 23 countries which include; Angola, Botswana, Burundi, Comoros, Democratic Republic of Congo (DRC), Eritrea, Ethiopia, Kenya, Lesotho, Madagascar, Malawi, Mauritius, Mozambique, Namibia, Rwanda, Seychelles, South Africa, Sudan, Swaziland, Tanzania, Uganda, Zambia and Zimbabwe (excludes Somalia and Djibouti). The natural resource base for agriculture is relatively marginal in most of ESA, in contrast to mineral resources, which are vast. Mean annual rainfall ranges from less than 100 mm in the arid zones to humid mountainous areas that receive over 1500 mm (Figure 1.1). At medium altitudes of about 1000 m.a.s.l., the savannahs provide relatively cool temperatures (for the tropics) allowing maize-based mixed farming systems. These take large tracts of land in Angola, Kenya, Lesotho, Malawi, Mozambique, Swaziland, Tanzania, Zambia and Zimbabwe. About 32% of the region’s poor people live in these maize-based systems. The high-rainfall and potentially highly productive areas cover large parts of Uganda, Burundi, DRC and Rwanda, and the highlands of Ethiopia, Kenya and Madagascar, amounting to about 31% of the total land area in the region. However, even with the high production potential, poverty prevalence in these zones is quite high, and about 54% of the region’s poor live in high rainfall areas. The ESA region has one of the highest poverty rates in the world, averaging 56% in the rural areas (IFAD, 2002). 3

Figure 1.1: Eastern and Southern Africa showing agroclimatic zones In general, low agricultural productivity and rural poverty are most evident in arid, semi-arid and sub-humid areas, which occupy 69% of the land area in the ESA region (IFAD, 2000; FAO, 1999). Rainfed cropping represents the main type of agriculture in the region and will continue to do so in the near future. While the potential for improving production and incomes from rainfed systems is considerable (Bossio et al., 2007), dependence on rainfall often limits the possibility of intensifying production and inevitably adds to the risk of crop failure due to drought (FAO, 2000). This vulnerability persists even where the risks of adverse weather are well known or are a regular occurrence. Therefore, AWM provides an important entry point for reducing rural poverty and mitigating 4

the impacts of climate change, by upgrading agricultural productivity. Defining Agricultural Water Management Until recently, water management for agriculture was synonymous with irrigation. Nevertheless, formal irrigation is only one dimension of agricultural water. There is strong evidence to suggest that improving management of water on rainfed lands, complemented by supplementary irrigation during dry spells, can go a long way towards increasing food security in much of semi-arid SSA (Falkenmark and Rockström, 2008). Furthermore, 95% of cultivated land in the region is rainfed, and this is unlikely to change soon. Hence, there is growing focus to improve water productivity in rainfed and quasi-rain-fed agricultural systems. In this context, professionals and practitioners have adopted a more holistic approach to water for agriculture, referred to as AWM. The original definition by FAO (1995) defined water management as “any kind of human action that influences the natural flow of water to farmers’ crops, or any form of agriculture that takes advantage of naturally rising or falling water levels for crop production.” The land on which water management is practiced is known as water managed area. ESA stakeholders (IMAWESA, 2007) modified this definition to one more inclusive of all sources of water and all types of agriculture, in which AWM is defined as; “all deliberate human actions designed to optimize the availability and utilization of water for agricultural purposes. The source of water could include direct rain as well as water supplied from surface and underground sources. AWM is therefore the management of all the water put into agriculture (crops, tree crops and livestock) in the continuum from rainfed systems to irrigated agriculture. It includes agronomy, soil and water conservation, rainwater harvesting, irrigation and drainage, interventions such as integrated watershed management and all relevant aspects of management of water and land”.

5

Rationale for focusing on AWM The Africa Water Vision for 2025 visualizes “An Africa where there is an equitable and sustainable use and management of water resources for poverty alleviation, socioeconomic development, regional cooperation and the environment” (UN Water/Africa, 2003). Meanwhile, agricultural water development in SSA is fairly invisible in comparison to its potential or compared to other regions of the world. For instance, out of the 39.4 million hectares potential, only 7.1 million hectares (18% of the potential) and just 3% of the total farmed area has been equipped for AWM in SSA (World Bank, 2006). The distribution of the area in ESA equipped for agricultural water is extremely inequitable because three countries (Sudan, South Africa and Madagascar) account for two thirds of the developed area. The water-managed area in the remaining countries does not exceed 300,000 ha per country. Meanwhile, over the last 40 years, only 4 million ha of land has been put under new AWM in SSA. Over the same period, China added 25 million ha, and India 32 million ha (World Bank, 2007; Molden, 2007). Therefore, progress in enabling agriculture to benefit from water management has been very slow in SSA, equivalent to a growth rate of just 1%, and targeted development is needed. In the past, sectorized approaches between what constitutes rainfed as opposed to irrigated agriculture led to competition rather than complementarities for resources and services. Projects were sectorized as either soil and water conservation or irrigation, and implementation usually maintained this divide (Lundgren and Taylor, 1993; Hurni and Tato, 1992). Examples of these are scattered throughout Africa and have formed the foundation of many development projects with agriculture and land management on their agendas (Reij et al., 1996; Negassi et al., 2000; Cleveringa et al., 2009). The integrated approaches adopted these days require rural development initiatives to embrace holistic AWM. There is adequate knowledge and experiences on a wide range of AWM technologies and practices adoptable in smallholder agriculture in 6

ESA, which have been tested and applied on the continent (Oweis et al., 2001; Mati, 2005; IWMI, 2006; Hatibu, 2000; SIWI, 2001). Thus, farmers can make technological choices to suit different biophysical and socio-economic conditions. Attempts by smallholder farmers to fully engage AWM best practices are constrained by the disconnection between available knowledge, its applicability and adoption, as well as lack of capital and support from value chains. The private sector, government, NGOs and other development agencies have provided sporadic support, which at times is inadequate. Consequently, AWM in most countries within ESA over the past decades has performed poorly. In particular, several communal irrigation schemes have performed below their potential (Kauffman et al., 2003; Inocencio et al., 2007). In addressing these constraints, farmers apply a vast array of lowinput, ecologically friendly agricultural technologies. Recent reviews have cataloged hundreds of technologies and approaches on AWM (Reij et al., 1996; Hatibu et al., 2006; Mati, 2007). Most assessments of these technologies register substantial increase in farmer yields and incomes, often exceeding double those achieved by conventional methods (Reij and Waters-Bayer, 2001; Penning de Fries et al., 2005). However, in ESA, most of the reporting on AWM gains has been negative, yet so much has changed over the last few decades. Positive data that combines both the technical and socio-economic components is relatively scarce. There is need for data to quantify AWM interventions which have worked at project scale, in order to guide policy and investment decisions in the region. Objectives and Scope The main objective of this study was to: contribute to guiding future investments in management of agricultural water by development partners and other organizations in the region. The specific objectives were:

7

(i)

To undertake studies on key water management issues of operational significance for investment on best-bet options for management of agricultural water. (ii) To establish conclusive evidence of the benefits for rural poverty reduction of best-bet options for water management under rainfed and small-scale irrigation systems. (iii) To develop communication products and activities targeted at investment decision makers, its consultants and partners involved in the identification and design of interventions. The regional study was designed as a series of country-based studies to establish and quantify the beneficial effects of specific AWM practices on improving productivity and curbing rural poverty. Given the overwhelming importance of rainfed farming in ESA, a priority issue covered was the identification of options for field level innovations in rainfed farming and quasi-irrigated systems, focusing particularly on mitigating the effects of dry spells. Prior to this, a baseline desk study on AWM technologies and practices across ESA was conducted, which identified over 100 technologies and practices adaptable to smallholder agriculture in the region (Mati, 2007). This provided a baseline showing the wealth of information on AWM in the region. However, available information was insufficient and there was need to find quantified data on what is working across the region, based on project interventions at field level. These types of data were collected and helped identify real AWM interventions with proven returns on investment on the ground that has projectscale applicability. It was this knowledge gap that this study was designed to address. The study Area The study was conducted concurrently across six countries; Ethiopia, Kenya, Madagascar, Malawi, Rwanda and Tanzania in ESA. These countries were selected based on their relative representation of AWM issues impacting on smallholder agriculture in the region. They represent East Africa, Horn of Africa, southern Africa, the Great Lakes region and an island. They all have large 8

numbers of smallholder farmers with high poverty prevalence, and where AWM interventions could have greatest impacts on agricultural productivity and poverty reduction. They also have diverse types of AWM successes, albeit not necessarily widespread, from which promising AWM interventions can be identified. Implementation of the regional study The study to identify “AWM interventions bearing returns on investment in ESA” was a multi-country, participatory applied research project designed to identify AWM technologies and practices, which have scope for up-scaling regionally, and which are being practiced at project-scale in each country, then to quantify the technical viability and economic returns of the interventions at farmer level. It was conducted by paired country-research teams, starting with desk studies followed by field work and participatory synthesis of results as follows: 1. Identifying main issues: Main issues were identified based on thorough background check, which included a scoping study (Chisenga and Teeluck, 2006). In addition, a desk study of AWM technologies and practices in ESA was conducted (Mati, 2007). Workshops and stakeholder forums were also conducted to gather ideas and insights on key AWM issues and knowledge gaps which could be addressed through a short-term regional study (Mati, 2006; Mahoo et al., 2007). All these materials and instruments were used to design the way the study would be conducted, and what it would seek. 2. Selection of pilot countries: This was done in January 2007. The criteria used meant ensuring inclusivity of eastern, southern, island and Franco-phone African countries in the ESA region where smallholder agriculture has had AWM project level interventions for quite some time, water management problems still persist and rural poverty is still prevalent. The selection of research teams was made from among the target countries. Teams of two researchers per country were identified each comprising a water management specialist (soil and water 9

engineer or economist or done at the participatory formats.

related field) and an economist (agricultural related field). The training of research teams was outset in a workshop which combined with planning and development of identical data

3. Criteria for identifying adaptable AWM technologies and practices: The criteria were developed to include AWM interventions which: o Can bring direct income to the farmer, o Have scope for wide applicability, o Can form a basis for large-scale investment (bilateral). The criteria also included those that were: o Technically feasible for adoption by smallholder farmers o Environmentally acceptable o Gender sensitive/equity considerations o Sustainable and resilient o Affordable/cost effective o Socially acceptable 4.

Collection of baseline data from literature and records: Each country team sought quantified literature and data on AWM from published material, for example, books, journal papers and from reports kept by institutions, projects and other sources. Discussions were also held with key stakeholders including researchers, project implementers and extension workers. These data were logged in identical spreadsheets to show both technical and socio-economic components. The types of data sought included background information (location, climate, agriculture), technical information on water management for agriculture and socio-economic data.

5. Selection of best AWM technologies and sample field sites: This utilized the findings of the desk studies to identify in each country, three technologies visible at three project-scale sites in each country, for further field observations and verification of the interventions. Thus, a total of 18 field sites were identified and used for this study (Table 1.1) across the six countries. The 10

criteria for selecting the site included areas where a project on AWM technology had been implemented for at least four years (in the past or on-going) and where there was availability of good records/data. In addition, the technology should have been practiced by a relatively large number of farmers, with visibility on the ground and have field characteristics which are replicable. Table 1.1: Technologies and sampled sites for field data Country

Ethiopia

Kenya

Madagascar

Malawi

Technologies/practices Research Site (i) Soil and water (i) Anjenie Watershed, West conservation on Gojam Zone, Amhara steep slopes Region (ii) Gravity fed (ii) Sewur Irrigation Scheme, smallholder irrigation North Shewa, Amhara (iii) Runoff harvesting Region (iii) Minjar Shenkora District, and storage in underground tanks Amhara Region (i) Soil and water (i) Kaiti Division, Makueni conservation with fanya District juu terraces (ii) Runoff harvesting and (ii) Lare Division, Nakuru storage in earthen pans District at household level (iii) Gravity-fed sprinkler (iii) Sagana-Maganjo Irrigation Irrigation scheme, Nyeri District (i) Conservation (i). Alaotra Lake region, South Agriculture East Vakinankaratra (permanent soil cover) (ii). Andranomanelatra, (ii) In–field soil and water Antsirabe II District management (iii). Tsivory, Southern (iii) System of Rice Amboasary District Intensification (SRI) (i) Smallholder surface 11

(i) Domasi Irrigation

Country

Rwanda

Tanzania

Technologies/practices Research Site irrigation with Scheme,Machinga community District, Southern management Region (ii) Improved wetland (ii) Kamalambo Dam, Mzimba (dambo) utilization and District, Northern management Region (iii) Small holder irrigation (iii) Mchinji District, Central Region. using treadle pumps (i) Buberuka Region, (i) Radical Terracing Northern Province (ii) Umutara Region in (ii) Irrigated paddy Eastern Province (iii) Rusuri-Rwamuginga (iii) Marshland marshland, Southern rehabilitation Province (i) Bukoba, Misenye District, (i) Conservation Kagera Region Agriculture (ii) Shinyanga Disttrict, (ii) Runoff harvesting into Shinyanga Region “majaluba” for rainfed (iii) Kwalei Lushoto and rice Bukoba Districts, (iii) Soil conservation and Shinyanga Region desert control

6. Data collection from field sites: At each selected site, this involved; (i) interviews conducted through questionnaires and focused group discussions, (ii) general field observations to provide ground truthing for the data, and (iii) stratified samples with random sampling within the strata and simple measurements where possible. It was not possible to do controlled experimentation given shortage of time and resources, data were either obtained from records kept by farmers and/or projects and extension staff. In other cases, estimates were made by farmers during interviews.

12

7. Data analysis: This involved calculations of yields, cost-benefit analysis, gross margins, net incomes, return on family labour, marginal return on family labour, water productivity and viability to investment. Excel spreadsheets and SPSS computer package were used in data management and analysis. Hydrological and economic modeling was not done due to the scatter and the low resolution of the available data. There were no predictive models developed through this study and this could be the focus of future work. The data analysis sought to shed more light, with quantification of: (i) AWM interventions that show viability for improving crop productivity and return on investment through:
Increase in productivity with, versus without technology (also before and after), under different conditions (biophysical, agronomic, water management)
Cost of achieving productivity increase
Estimate of total investment cost/benefit. (ii) Scope for improving water productivity (kg or US$ per ha per mm of water)
Amount of water applied at farm level (system efficiency)
Amount of yield per unit of water (iii) Rural poverty reduction and equity
Return on labour (income vs labour)- with and without the technology
Costs of inputs, labour and discounted returns
Socio-economic implications including gender equity. 8.

Peer reviews and writeshops: These were implemented every six months to assess progress, conduct joint analysis of the data, to plan for gap-filling of data and corrections, and to identify the regional key messages in a participatory manner.

Determining economic parameters The economic performance parameters were determined as shown in Table 1.2. The gross value of product or revenue was 13

determined as product of crop yield and farm gate price. Net value of produce (profit) excluding family labour was computed by deducting total expenses excluding the costs of family labour from the total costs that includes annualized cost of AWM technologies, maintenance costs of the technology, all costs related to agronomic inputs, husbandry practices and handling and marketing activities. The main idea of such cost-benefit analysis was to quantify if the investment on AWM technologies including yearly maintenance costs, since this is justifiable in terms of higher agricultural production and incomes (profits). The average prices of inputs and outputs for the year 2007 at each study site were the basis for the calculation. Costs of family labour were not considered in the costbenefit analysis because it was assumed that it has relatively less opportunity costs and would be idle if not be involved in such production practices. However, the calculated “profit” was related to the amount of family labour invested to earn the return on family labour of that particular AWM technology. The gross return on family labour was therefore determined by dividing the net income or profit excluding the costs of family labour inputs with the quantified family labour, expressed in adultdays. Marginal rate of return was calculated to quantify the rate of return generated by every currency unit of investing on AWM technology as compared to not investing. The marginal rate of return was thus computed by expressing the difference between the net benefit (profit) of ‘with’ and ‘without’ AWM technology as a percentage of the difference of total costs of labour. The computed marginal rate of return gives an indication of what a producer can expect to receive, on average, by switching technologies. Hence, a 150% marginal rate of return in switching from having no AWM practices to implementing AWM implies that for each dollar invested in AWM, the producer can expect to recover the US$1.00 invested plus an additional return of US$1.50.

14

Table 1.2: Economic performance parameters and their determination Performance Parameters

Source of information

Productivity and profitability A

Yields (kg ha-1yr-) (grain)

Household Survey Opinion of farmers, consumer

B

Farm gate prices (US$ kg-1)

C

Gross value of produce or revenue (US$ ha-1)

Calculated (A*B)

Costs excluding family labour(US$ ha-1) D E F G H I J

Annualized cost of the AWM system (establishment) Annual maintenance of the AWM system All agronomic inputs (fertilizer, pesticide, seeds) Husbandry practices (plowing, sowing, weeding) Harvesting, handling and marketing costs Total costs excluding family labour (US$ ha-1) Net profit excluding family labour (US$ ha-1)

Recorded/survey (annualized) Household Survey Household Survey Household Survey Household Survey Sum (D:H) Calculated (C-I)

Scope in improving water productivity (US$ mm-1) K L

Total water used during production period (mm season-1) Water productivity (US$ mm-1)

Rainfall data /irrigation water Calculated (C/K)

Contribution to poverty reduction M N O P Q

Family labour inputs (Adult-days ha-1yr-) Maintenance and improvement of the AWM system Husbandry practices Harvesting, handling and marketing Total family labour (days ha-1yr-) Gross return on family labour (US$ 15

Household survey Household survey Household survey Household survey Sum (M: O) Calculated (J/P)

R S T

Performance Parameters adult-day-1) Opportunity cost of labour (US$ adult-day-1) Total cost of family labour (US$ ha-1) Marginal return on family labour (US$/adult-day)

Source of information Household survey (local wage rate) Calculated (P*R) Calculated (∆J/∆S*100)**

Return on Investment U V W X Y Z

Discount factor Discounted costs Discounted benefits NPV IRR ROI

Interest rate (bank) Calculated Calculated Calculated (W-V) Calculated Calculated

The Water Productivity of the crop was calculated as crop yield or its monetary value divided by the amount of water utilized during the vegetative crop growth period. The calculation was done based on the total and effective rainfall amount in the area and expressed in terms of either kilograms (kg) of grain or US$ per millimetre of water consumed. The foregoing data analysis were complemented with further financial analysis, such as net present value (NPV), internal rate of return (IRR), investment on return (IOR) and marginal rate of return (MRR) (Leiber, 1984). NPV compares the value of a dollar today to the value of that same dollar in the future (discounting), taking inflation and returns into account. The difference between the sum of all discounted benefits and costs represents the NPV. This difference reflects how much the investment has brought benefits. If the NPV is positive, it should mean the investment on an AWM technology was profitable. However, if NPV is negative, clearly the costs outweigh the benefits and the investment was not economical. While NPV is expressed in monetary units (dollars in this case) the IRR is the true interest yield 16

expected from an investment on AWM technology expressed as a percentage. It shows the discount rate below which an investment results in positive NPV and above which an investment results in a negative NPV. It is thus the break-even discount rate, the rate at which the value of costs equals the value of benefits. A simple decision-making criterion can be stated to accept the profitability of the investment if its IRR exceeds the cost of capital and rejected if this IRR is less than the cost of capital. ROI is a performance measure used to evaluate the efficiency of an investment. It is the ratio of money gained or lost on an investment relative to the amount of money invested. To calculate ROI, the benefit (return) of an investment is divided by the cost of the investment; the result is expressed as a percentage or a ratio. In other words, if an investment on terracing does not have a positive ROI, then the investment should not be undertaken. AWM technologies and practices identified Major findings This study found that the ESA region has many community scale successes, where AWM interventions are improving agricultural productivity, supporting environmental recovery and contributing to poverty reduction. Interventions were considered successful if they had; (i) visibility of the AWM intervention on the ground, (ii) adaptability of the intervention to smallholder agriculture, (iii) relatively large spatial coverage of the intervention by local standards, (iv) technological and socio-economic viability whereby some drawbacks had been overcome to improve livelihoods, (iv) widespread adoption by local communities, and (v) quantifiable and/or inherently visible impacts on poverty reduction. Broadly classified, some ten AWM technologies and practices identified. They include: (i) Construction of small individual storages for runoff harvesting (blue water), with the water used for

17

supplemental irrigation of high-value crops, as in Ethiopia and Kenya, (ii) Harvesting runoff flows and storage in bunded basins within the soil profile (green water), in order to augment natural rainfall for production of field crops, especially rice, as in Tanzania and Madagascar. (iii) Terracing steep slopes to reduce soil erosion and runoff losses and thereby conserve water and nutrients for the crops, as in Ethiopia, Kenya, Rwanda and Tanzania (iv) Piped streamflow diversion and utilization for gravity-fed, sprinkler irrigation of high value commercial crops as in Kenya, (v) In-field water management, including conservation tillage techniques, mulching and permaculture as in Madagascar and Tanzania, (vi) Improving water control and management in existing irrigated systems as in Rwanda, Malawi and Madagascar, (vii) Use of small pumps and manually operated pumps to lift water from shallow water tables for irrigation as in Malawi (viii) Utilization of high water tables and valley bottoms for crop production as in Malawi and Rwanda. (ix) Stream diversion with surface irrigation and improved marketing of produce as in Malawi, Ethiopia and Rwanda (x) Soil fertility management, including use of manures and fertilizers alongside the water management technologies to improve crop productivity was found across all countries. Investment in Agricultural Water Management pays This study confirmed that “investment in the AWM pays”, and this was true across all the technologies and practices identified. In nearly all the cases, water management “with” technology increased crop yields by factors ranging from 20% to over 500%, while net returns on investment increased by up to ten-fold On average, AWM interventions increased crop productivity, income returns on investment, improved returns on labour and better water productivity. In addition, AWM was linked to poverty reduction, 18

employment creation and environmental conservation. Subsequent sections summarize this evidence. Evidence of increased crop yields Across all the countries and sampled sites assessed, crop yields had increased with adoption of targeted AWM technologies and practices. Terracing is necessary on many hilly and highland areas of the region, contributing to the conservation of both water and soil, along with other benefits. For instance, in the Anjenie watershed of Ethiopia, long-term terracing increased average yields of teff achieving 0.95 t ha-1 (control 0.49), barley 1.86 t ha-1 (control 0.61) and maize 1.73 t ha-1 (control 1.69), respectively. These results are in line with the findings of Vancampenhout et al. (2005) on the positive effects of soil conservation on the yields of field crops in the highlands of Ethiopia. Another example from Buberuka highlands of Rwanda showed that potato production increased on bench terraces by about 38% compared to cultivation on slopes. Similarly, in Kenya’s Kaiti catchment of Makueni District, fanya-juu terraces increased yields of maize by more than 100% compared to cultivation on slopes. Water harvesting with supplemental irrigation has multiple benefits for farmers in areas with droughts and prolonged dry spells (Hatibu et al., 2000). In Shinyanga region of Tanzania, farmers make excavated bunded basins locally known as majaluba. These are made to hold water harvested from surface runoff for crop production. About 35% of the rice in Tanzania smallholder individual farming is produced this way. The study found that runoff diversions with majaluba had increased paddy rice yields significantly from 2.17 t ha-1 for ordinary rainfed rice to 3.43 t ha-1 with majaluba, which was equivalent to 58% yield increment. Conservation agriculture (CA) is an AWM intervention with positive impact on crop production. In Tanzania’s Bukoba and Missenyi districts of Kagera region, mulching practice by smallholder farmers increased the average maize yields from 2.50 t ha-1 to 3.40 t ha-1. However, this increase varied by zones, being highest in the low 19

rainfall zone (170%) and lowest in medium rainfall zone (13%). Tanzania has been fostering the adoption of CA because of its potential to address three areas of crucial importance to smallholder farmers, that is, demand on household labour, food security through increased and sustainable crop yields, and household income (Lofstrand, 2005; Mariki, 2004). For best results, CA practices such as mulching must be accompanied by requisite agronomic practices such as fertilizers, manures, pesticides, and high quality seed, as well as proper water application and management. There is evidence that all forms of irrigation have contributed to improved crop yields. In Madagascar’s Tsivory area of Southern Amboasary District, rice yields doubled as a result of improving water management through the system of rice intensification (SRI). The smallholder farmers with SRI practice in Tsivory obtained rice yields averaging 3.63 t ha-1 from small fields and 4.10 t ha-1 from medium sized fields, compared to 1.85 t ha-1 normally obtained from conventional flooded paddies. In Kenya’s Sagana-Maganjo Irrigation Scheme, sprinkler irrigation increased the yield of cabbage from 530 t ha-1 without irrigation to 4,000 t ha-1. In Malawi, dambo irrigation at Kamalambo wetland recorded higher yields than upland rainfed systems, achieving 12.75 t ha-1 of tomato and 8.26 t ha-1 for onion, which was equivalent to an increment of 37% and 33% on both crops, respectively. These results agree with the findings of other researchers in the region (Wiyo and Kasomekera, 1994; Mloza-Banda et al., 2001). Moreover, improvements in water control and management can substantially increase crop yields. In another example of Cyabayaga valley bottom of Eastern Province of Rwanda, water control for paddy rice production increased the average yield of rice from about 3.5 t ha-1 to over 6.5 t ha-1. This compares well with the value of 6.0 t ha-1 reported by GoR (2005) and Jagwe et al. (2003). Another example from Rwanda’s Rusuli-Rwamuginga marshland in Southern Province obtained that development of the marshland and irrigation scheduling practices improved rice yield on average from 5.84 t ha-1 to 9.52 t ha-1 comparing before and after marshland development respectively. Another case from Domasi Irrigation 20

scheme of Malawi demonstrated that regulated water control in irrigation had significantly higher rice yields of 3.6 t ha-1 as compared to 1.2 t ha-1 under unregulated streambed irrigation. All these findings quantified evidence in support of AWM as the engine for improving smallholder crop productivity, and thus, improving food security. Evidence of increased returns on investment Increased returns on investment accruing from AWM interventions were recorded across all crop enterprises assessed. For instance, in Rwanda, the net benefit was found to vary between US$490 ha-1 and US$1,180 ha-1 from rice grown in valley bottoms in Umutara region. The corresponding values were 500 to US$1,425 ha-1 for improved water control with lined canals at Rusuli-Rwamuginga marshland of Rwanda. This is equivalent to an average 119% increment in net returns comparing with and without technology. In Kenya, the net present value for maize in Kaiti catchment ranged from US$ 800 ha1 for the upper slopes to US$1,500 ha-1 for the lower slopes. Another scenario from Machinga in Malawi showed that regulated water control in irrigation obtained significantly higher rice yields of 3.6 t ha-1 as compared to 1.2 t ha-1 under unregulated streambed irrigation. Net farm incomes were also higher under regulated irrigation, increasing from an average of US$248 ha-1, without regulation to US$1,184 ha-1 with regulated irrigation. Also, in Kenya’s Lare Division, supplemental irrigation from rainwater harvesting ponds resulted in increased yield of kales from 6.4 t ha-1 to 15.8 t ha-1, and onions from 1.6 t ha-1 to 11.9 t ha-1. Similar results were observed (Mati and Penning de Vries, 2005; Malesu et al., 2006) in Lare showing that crop yields increased due to improved intensity of cropping and water utilization to fill the rainfall deficits. The incomes benefits of soil and water conservation cannot be understated. In Anjenie catchment of Ethiopia, net value of product was significantly higher on terraced fields, recording 20.9 (US$-112 control) for teff, 185 (US$-41 control) for barley and -34.5 (US$-101 control) US$ha-1yr-1 for maize, respectively. Using a discount rate of 10%, the average net product value of barley with terracing was 21

about US$1,542 over a time period of 50 years. In addition, the average financial internal rate of return was 301%. Also in Ethiopia, water harvesting with small ponds for supplemental irrigation of vegetables and seedlings at Minjar Shenkora obtained average net incomes of U$155 per 100 m2 plot from onion seedlings, while incomes from bulb onions grown in the field provided equivalent of US$1,848 ha-1, adding up to US$2,003 ha-1, from onion crop alone. Comparatively, net incomes from teff and wheat were 523 and US$ 525 ha-1, respectively. Thus, water harvesting with small storage ponds could make major contributions to household incomes and rural poverty reduction. These results are consistent with the findings of Gezahegn et al. (2006) and Nega and Kimeu (2002), who assessed small scale water harvesting technologies in Ethiopia and found that returns on investment were high. Meanwhile, in Shinyanga region of Tanzania where farmers grow rice with water harvesting using bunded basins (majaluba), household incomes increased from 430 US$ ha-1 without runoff harvesting to 720 US$ ha-1 with the technology. The study found that runoff harvesting should be encouraged and accompanied by use of inputs such as fertilizers and manures, improved seeds, good agronomic practices, value addition and access to markets. The main constraint was that, with or without runoff harvesting, the majaluba system is predominantly rainfed with water storage in the soil profile (green water). Consequently, climatic uncertainties and prolonged dry spells adversely affect the system. It was noted that improvements could be made to allow water harvesting with storage infrastructure so as to drought prove the system. All forms of irrigation also recorded incremental returns on investment with AWM. For instance, at Malawi’s Domasi Irrigation Scheme, the net value of produce (NPV) for maize, wheat and rice averaged 322, 162 and 1,184 $ ha-1 respectively under regulated irrigation, which was equivalent to an improvement on incomes by factors of 636%, 128% and 377% respectively. It should be noted that maize grown outside the irrigation scheme attained a negative 22

NPV (US$-60 ha-1) excluding family labour. This could be ascribed to the limited access and use of streambanks of Domasi River and wetland gardens (Kambewa, 2004; Ferguson and Mulwafu, 2004). Similar results were obtained from the Sagana-Maganjo Irrigation Scheme of Kenya, where net income returns for beans, cabbage and French beans increased from US$-178 ha-1, -382 and US$919 ha-1 without irrigation to US$247 ha-1, US$2,991 and US$6,032 ha-1 respectively with irrigation. Generally, negative value of incomes for certain crops underlies the fallacy of smallholder rainfed systems in semi-arid zones of ESA. These results show that irrigation, particularly supplemental irrigation of field crops in marginal zones has an important contribution to income benefits of smallholder farmers in the region. AWM impact on returns on labour The adoption of AWM technologies and practices improved the returns on labour for most crop enterprises, for certain interventions, total labour input increased. Gross return on family labour varied with the type of water management technologies and crops grown in Ethiopia, Tanzania, Kenya as well as in Rwanda. In Kenya’s Lare Division, the average return on labour for marketable crops under supplemental irrigation with water harvesting was US$9.8 per adult-day compared to US$0.9 per adult-day without water harvesting. Similarly, in the Anjenie catchment of Ethiopia, returns on family labour were US$2.33, US$1.01 and US$0.74 per adult-day, respectively for barley, teff and maize grown on terraced plots, as compared to US$0.44, 0.27 and US$0.16 per adult-day, respectively for crops grown on sloping land. This difference was slight because although the opportunity cost of family labour was low at US$ 1.7 day-1, the share of family labour costs from the total production costs in terraced plots was significantly high. Assuming an income of US$1.0 per day as the threshold for poverty line, terracing can help the rural poor to move out of poverty. Also in Ethiopia at Sewur Irrigation Scheme, gross return on family labour was affected more by crop choice than by the type of AWM technology. This is because the return from rainfed onion 23

production was US$23 while that from the irrigated crop was US$28 per adult-day. Also, returns on labour were 800% higher from onions, than from teff and munge bean production. Contrary to onion, returns on labour for teff and munge bean were higher under rainfed production system than under irrigation. This is because the humid climate of the rainy season in the Ethiopian highlands favors teff and munge bean yields and disease incidences are relatively low. Moreover, teff and munge bean are broad casted and not suitable for irrigated agriculture. This shows that families constrained by labour can choose to invest it in the most opportune crop. In Tanzania’s Shinyanga region, rice production in majaluba with runoff harvesting improved return on labour by 62%. Marginal return on family labour was higher in lowland fields with runoff harvesting, averaging U$1.65 per adult-day while equivalent upland fields achieved US$ 0.48 per adult-day. The better performance of lowland fields was attributed to the benefits of seepage inflow and relatively more fertile alluvial soils. The high clay content in the lower fields hold more water for long time compared to both middle and upper fields. This is supported by data by Meertens et al. (2003), Kajiru (2006) and Ngailo et al. (2007). On average, the return on labour from majaluba was higher than the opportunity cost of labour, being about US$0.7 per adult-day. This implies that farmers can make better use of their family labour by deploying it on majaluba with runoff harvesting. It was also observed that male farmers obtained higher return on labour than female farmers. Further, AWM technologies engage more labour for productive purposes even during the dry spells, thereby contributing to increased employment in rural areas. Returns on labour can also be quantified as improved efficiency from adopting a technology. In Mchinji District of Malawi, it was observed that adoption of treadle pumps enabled farmers to pump more water per unit time (about 1.6l s-1) as opposed to bucket irrigation. The bucket may take the farmer about twice to thrice the time required to irrigate an area as compared to a treadle pump. 24

Normally, the average discharge of a treadle pump is between 0.4 to 1.7 l s-1, depending on the person operating it and the suction lift required (Van Leeuwen, 2003; Wiyo, 2002; Mangisoni, 2004; Maweru, 2004). Generally, as depth of pumping increases, the discharge rate that can be achieved decreases. Since the water tables at Mchinji are high, a treadle pump requires less labour to irrigate a unit area compared to bucket irrigation. Therefore, the benefit of using treadle pumps lies in improving the efficiency of irrigation by saving time and labour. Typically, the treadle pump is appropriate where there is a water source close to the surface (total pumping head less than 7 meters) and close to the field to be irrigated (distance from water source less than 200 meters). Impediments to their rapid adoption include the fact that they utilize manual labour and can be tiresome, require regular maintenance and have limited capacity in areas requiring high delivery heads. It was also observed that generally, the costs of farm inputs, e.g. fertilizers and manures, pesticides, far outweighed the costs of labour. For instance, in Kenya’s Kaiti catchment, the total costs of labour to grow maize averaged US$35 ha-1 as compared to the costs of inputs at US$282 ha-1. Another example from Lare Division of Kenya obtained that the total costs of labour for cabbage averaged US$139 ha-1as compared with total costs of inputs at US$302 ha-1. This means that the major constraint to improving productivity lies with inputs. As poorer smallholders are most vulnerable in terms of accessing inputs, strategies are needed to support them and to cushion them from unfair market forces. Impacts of agronomic practices Striking results were obtained from a new method of water and agronomic management in rice paddies called SRI in Rwanda’s Bugesera District. By adopting SRI, farmers’ rice yields increased on average from 3 t ha-1 to 6.5 t ha-1, with some farmers achieving 8 t ha-1. The main features of SRI are: transplanting of young seedlings singly in a square pattern with wide spacing, using organic fertilizers and hand weeding, and keeping the paddy soil moist during the 25

vegetative growth phase (Henri de Laulanie, 1983; Vermeulen, 2009). SRI was invented in Madagascar and has many environmental benefits including breaking the pest cycle of water borne diseases, as well as reducing water logging and water application. SRI was adopted in Rwanda from Madagascar, through a learning exchange. In Madagascar’s Tsivory area, SRI increased rice yields from 1.8 t ha-1 with conventional flooding, to 4 t ha-1 in this study. It is not common for rice to be grown in permanent crop cover systems. Permanent crop covers involve growing a crop preferably under a mulch, and planting the next crop while the old one is still in the field thereby ensuring that there is a crop cover throughout the year. The system works well with areas having high and well distributed rainfall or under irrigated agriculture. But in Madagascar’s, Alaotra Lake Region, adoption of permanent soil cover for rice with various levels of inputs yielded 1.46 t ha-1 for sole manure (control) as compared to 3.11 t ha-1 for manure with cover, 3.58 t ha-1 for manure and fertilizers and 4.81 t ha-1 for manure, fertilizers and permanent cover system. This translates into a doubling and tripling of the yields. In Anjenie catchment of Ethiopia, one observation was that before terracing, maize was not grown in the area due to moisture stress, but farmers with terraced fields could grow maize due to improved water conservation. Also, farmers with terraced barley fields adopted two crop seasons per year in place of a single crop on farms without terraces. Also in Ethiopia, double cropping was achieved through availability of irrigation water, at Sewur Irrigation Scheme thereby increasing profitability of crop production by between 40 and 942%. In Tanzania, mulching is a conservation agriculture practice widely adopted by farmers in Bukoba and Missenyi districts of Kagera Region. In these areas, intercropping of coffee with bananas, beans and other field crops is typical of these farming systems. Mulching is used for soil and water conservation and for improving crop productivity. The types of mulches applied by farmers under maize 26

crop include: cut grasses, crop residues and green (crop) covers. Different legume species are used as cover crops including beans and forage species; Mucuna (Mucuna pruriens), Lab-lab (Lablab purpureus) and Tropical Kudzu (Pueraria phaseoloides). Generally, CA is a system of farming which conserves, improves and makes more efficient use of natural resources through integrated management of available soil-water and biological resources (Lofstrand, 2005; Bwalya, 2005). Impacts on water productivity Water productivity depicts the overall efficiency of water applied and its conversion to a product (yield) or value of product (earnings), regardless of whether water was from rainfall, irrigation or supplemental irrigation. In Minjar Shenkora District of Ethiopia, the water productivity of onion seedlings irrigated from storage ponds was estimated to be about US$38.7 mm-1, while that of rainfed bulb onions was US$2.4 mm-1 per season, teff and wheat were US$0.69 mm-1 per season. Accordingly, water productivity of onions was 3.5 times higher than that of wheat and teff. Whereas the water productivity of onion seedlings was the highest, it should be noted that seedlings are not really an end product, but their sale has far reaching impacts on spreading the net benefits of water harvesting to other farmers in the community. Production of high value crops such as onion seedlings with water harvesting, utilized scarce water and contributed to improved economic returns of onion grown under rainfed conditions by the community. In Tanzania’s Shinyanga region, runoff harvesting with majaluba improved water utilization and made the crop resilient to droughts and prolonged dry spells. Generally, farmers who had adopted channeling water to their majaluba achieved higher yields and profits thereby improving their livelihoods and well-being as compared to farmers relying on rainfed rice production. The biophysical, climatic and management conditions to encourage ease of adoption by smallholder resource-poor farmers exists since the land terrain is usually suitable. Generally, the adoption of improved AWM technologies enabled improvements in water productivity of most 27

enterprises. AWM also optimizes crop water use and enhances the productivity of other agronomic practices. Social and Environmental impacts Generally, AWM interventions enhance agricultural productivity, providing rural employment, cash incomes and food security, thus improving the livelihoods of thousands of people in the region. Positive AWM impacts on the environment were easily identifiable from interventions that target soil and water conservation, rainwater harvesting, CA and irrigation with water control infrastructure. On average, irrigation had positive social and environmental impacts, especially overhead irrigation and supplemental irrigation with water harvesting. First, farmers could grow a variety of crops and thus maintain a kind of permanent cover on land. Secondly, farmers could irrigate tree seedlings and thus facilitate tree planting in areas beyond the irrigation schemes. Among the irrigation schemes assessed, salinity was not a problem and this was linked to good site identification, layout of the scheme and water management by farmers. Moreover, nearly all AWM interventions enhanced soil moisture retention thereby facilitating the utilization of soil nutrients and significantly contributing to improved crop growth and yields. In this study, mulching in Bukoba and Missenyi districts of Kagera region of Tanzania proved to have positive returns on labour and capital invested, and was compatible with the banana based farming systems. Farmers interviewed indicated that mulching practice had positive impacts by improving soil fertility, reducing soil erosion and improving soil cover and thus significantly reducing evaporation losses. Similar sentiments were obtained by Sambrook et al. (2004) in Babati and Karatu districts of Manyara Region, showing that mulching and cover crops increased maize production and household incomes, while enhancing soil fertility and conserving the environment. Generally, permanent crop cover practiced with rice crop in Alaotra Lake Region of Madagascar was found to contribute to 28

improvement in crop productivity, household food and income security and reduction in labour costs. The types of cover applied by farmers on rice crop include different legume species like beans, forage and agroforestry species: Stylosanthes guianensis, Mucuna spp, Dolichos lablab, Pueraria phaseoloides, Cajanus cajan, Vigna umbellate (GSDM, 2004; Séguy et al., 2009). The environmental advantages of permanent soil cover have been identified as soil and water conservation, improved infiltration, enhanced soil fertility and efficient management of water in the catchment (Séguy et al, 2009). However, permanent soil cover (PSC) is not easy to adopt in areas with low rainfall or where crop residues are needed to feed livestock. Such social-economic implications need to be assessed in adopting these technologies. Meanwhile, experiences from Kamalambo wetland of Malawi indicate that management of wetland ecosystem presents a challenge to farmers using traditional methods. It is difficult for them to manage water for crop production efficiently, and still maintain the environmental conservation of wetlands and riparian lands. However, farmers view these lands as areas with infinite water supplies, and as such, water use is rarely subjected to any strict irrigation scheduling and/or water conservation techniques. Extension workers reported that as the water table recedes in the dry season, drinking water for humans and livestock tends to be low and people and livestock have to travel further from the homesteads for water. Poor water management and use are becoming key concerns in wetlands of Malawi. Thus, farmers should be trained on how to manage the fragile dambo ecosystems and outlying catchment areas sustainably, including methods to improve water control, drainage and irrigation. These types of cases notwithstanding, AWM technologies/ practices generally resulted in improved food security and nutrition, increased household incomes and thus poverty reduction (Mati et al., 2008). Furthermore, higher returns on labour than the prevailing opportunity cost of labour without technology were obtained across 29

most enterprises, thereby improving the economic use of rural labour. This increases farmers’ choices and reduces rural unemployment. However, the benefits from AWM varies with households, with resource-rich households (by local standards) obtaining relatively higher productivity compared to the resourcepoor households. This is because households with limited resources could be constrained to apply the required input levels and crop husbandry leading to low productivity. There is therefore need to support poor households by making inputs affordable and thus AWM profitable, while safe-guarding the environment. Other supportive initiatives A major threat to enhancing the AWM in ESA includes negative perceptions about the returns on investment from AWM, including from irrigation (Inocencio et al., 2007). However, evidence is emerging that these sentiments may not hold true for smallholder individual investment in irrigation and other forms of AWM. This study has shown that smallholder water management, especially where the farmer has some level of individual autonomy in decision making, is highly profitable as well as sustainable. Another constraint is the high initial investment required. Sometimes, supporting infrastructure such as roads, stores, processing facilities and markets, may have to be constructed to support the value chain needed to make AWM profitable. If these facilities are lacking, then it is very difficult for AWM to pay. Moreover, the poorest and most vulnerable communities tend to be located in the driest and remotest (far from roads, towns) part of the country where transaction costs of any activity tends to be high. This poses further challenges as to where to allocate resources, especially given the slim chances of payback from such vulnerable groups as the poorest. In addition, many communities do not come from a tradition in which irrigation, water harvesting, soil and water conservation or drainage were practiced, having been used to shifting cultivation. There is a lot of learning and adaptation needed because this takes time. Moreover, the trial and error tendencies of farmers exposed to irrigation, water harvesting or other new technology for the first 30

time can be challenging (Mati, 2008). This sometimes leads to mistakes which could discourage both the farmers and investors. However, even with these limitations, the benefits of managing water for agriculture optimally far outweigh the threats, especially as there is increased food security, wealth creation, poverty reduction and improved livelihoods for beneficiaries. It was also noted that some of the technologies with very high returns on investment, e.g. surface irrigation of paddy rice, require the adoption of good agronomic practices, e.g. SRI to optimize water utilization. The costs of inputs tend to exceed those of labour, and in many cases, production costs were beyond the farmer’s control. There has been very little metrics of water used in agriculture by smallholder farmers, especially under surface irrigation. Although adequate knowledge on AWM exists, the best of it is not documented and is held in the heads of farmers, managers and implementers of projects, extension workers, researchers and other practitioners. Therefore, rural farmers, donors, non-governmental organizations (NGOs), private investors and governments should give due attention in promoting AWM and more investment as a strategy for poverty reduction, food security and wealth creation. Other supportive aspects include implementation of AWM interventions as part of a more inclusive package that accommodates the full value chain and water for multiple use systems (van Koppen et al., 2006). Since most development is targeted at groups, the institutionalization of management structures such as water users associations (WUAs) is crucial. Whenever possible, AWM interventions should aim to provide water for multiple purposes and enhance cost effectiveness. Capacity building for all cadres of stakeholders and local ownership are necessary for success. Conclusions and Recommendations The study to “identify AWM interventions bearing returns on investment in ESA” was conducted on 18 project-scale sites in six countries; Ethiopia, Kenya, Madagascar, Malawi, Rwanda and 31

Tanzania. It revealed that “Investment in AWM pays”. It has been observed that there are many examples of successful AWM interventions in the region. Examples of these include; construction of small individual storages for water harvesting, runoff harvesting and storage in the soil profile for crop production, soil and water conservation on steep slopes, stream diversion with gravity-fed irrigation, in-field water management including mulching and permaculture, improved water control in existing irrigation schemes, use of small pumps and manually operated pumps with shallow water tables for irrigation, use of valley bottoms and wetlands for crop production and soil fertility management. This study did not assess livestock interventions and benefits. Detailed reports of the findings are in succeeding chapters and they contain fresh data and a wealth of knowledge on AWM technologies and practices in the sampled countries. There is no doubt that AWM constitutes a key ingredient to the success of the agricultural productivity in ESA. Unlike the Asian green revolution which found in place a relatively well developed irrigation infrastructure, the African agricultural sector is predominantly rainfed, even in ecological zones which should of necessity be fully or partially irrigated. Again, unlike fertilizers and improved seeds which can be imported, purchased and distributed in small packages to farmers, water has to be sourced locally, is niche relevant and requires higher technological input. Such services as surveying, design, layout and construction of water supply and application equipment requires some level of specialized expertise and cost. This is not to say that the farmer is completely helpless. There are many technologies which the farmer can survey, layout and construct water management infrastructure using own labour at farm level with minimum training and facilitation. The interventions identified in this study are by no means exhaustive but constitute broadly, AWM interventions that can form the basis for extensive project-scale interventions in most countries. Within each category, there are several technologies to allow local 32

adaptability and cost-effectiveness. Another aspect of these interventions is their applicability for smallholder agriculture and drylands. The success of these interventions will depend to a large extent on the operational framework against which they are implemented, especially the inclusion of the farmers in planning, implementation and management of the systems. Other supportive aspects include implementation of AWM interventions as part of a more inclusive integrated watershed management, and thus the institutionalization of management structures such as WUAs. Whenever possible, AWM interventions should target to provide water for multiple purposes and enhance cost effectiveness. Capacity building for all cadres of stakeholders and local ownership is necessary for success. The major threats to enhancing the AWM for the improved productivity of drylands in ESA include; negative perceptions about the returns on investment by investors and development partners, as well as high poverty levels among land users. Other constraints include; weak linkages between research, science and smallholder farmers, lack of or ineffective policies to support the needed investment, technologies mis-matched to local conditions and failure to learn from each other, within and across countries in Africa. Sometimes, the high initial investments pose a limitation to adoption, while supporting infrastructure such as roads, stores, and markets may have to be constructed. Moreover, the poorest and most vulnerable communities tend to be located in the driest and remotest (far from roads, towns) parts of the country where transaction costs of any activity tend to be high. This therefore poses a challenge as to where to allocate resources, especially given the slim chances of payback from such resource-poor vulnerable groups. The trial and error tendencies of farmers exposed to irrigation and/or water harvesting for the first time can lead to many mistakes which could discourage both the farmers and investors. However, even with these limitations, AWM is necessary for agricultural productivity to be achieved along the full value

33

chain, for wealth creation, improved livelihoods and sustainable environments. References Bossio, D. Critchley, W., Geheb, K., van Lynden, G. and Mati, B. (2007). Conserving land-protecting water. In: Water for Food, Water for Life: A Comprehensive Assessment of Water Management in Agriculture, London: Earthscan and Colombo: International Water Management Institute, 551-583. Bwalya, M., 2005. Soil and water Conservation to Conservation Agriculture Practices (experiences and lessons from the efforts Eotulelo Farmer Field School – a community based organization) January 2005. Eotulelo Farmer Field School Group, Likamba Village, Arumeru. Arusha Region, Tanzania. Chisenga, J. K. and Teeluck, J.P. 2006. Initial Scoping of Status of Knowledge Management in Kenya, Madagascar, Malawi, Mauritius and Tanzania. Mission Report. SWMnet Working Paper 10. Nairobi. Cleveringa, R., M. Kay and A. Cohen. 2009. Synthesis of Strategic Approaches: Enhancing Pro-poor Investments in Water and Livelihoods. Rome: IFAD. Darghouth, S. 2007. Reengaging in agricultural water management in Sub-Saharan Africa: Edition GSDM/ CIRAD, 32 pp. Falkenmark, M. and J. Rockström. 2008. Building Resilience to Drought in Desertification-Prone Savannas in Sub-Saharan Africa: The Water Perspective. Natural Res. Forum 32:93-102. FAO 2000. Agriculture Towards 2015/30 (AT 2015/30): FAO Perspective study on agriculture in 2015 and 2030. FAO, Rome. FAO 2003. AQUASTAT. Global Information system of water and agriculture. Land & water Dev Division of FAO. Rome. FAO 2005. Food security and agricultural development in subSaharan Africa: Building a case for more public support. Main Report. NEPAD and FAO, Rome 34

FAO, 1995 Water Report 7. Irrigation in Africa in figures. FAO Rome. Ferguson, A.E. and W.O. Mulwafu, 2004. Irrigation Reform on Malawi’s Domasi and Likangala Smallholder Irrigation Schemes: Exploring Land-Water Intersections. Final Research Report to BASIS CRSP, October 2004, Madison, WI. Gezahegn Ayele, Gemechu Ayana, Kiflu Gedefe, Mekonnen Bekele, Tilahun Hordofa and Kdane Georgis.2006. Water Harvesting practices and impacts on livelihood outcomes in Ethiopia. EDRI,Addis Ababa,Ethipia. Global Risk Network (2008). Africa@Risk. World Economic Forum, Geneva, Switzerland. GoR, 2005 Programme national pour la production du riz période: 2006-2016. MINAGRI, NEPAD (PDDAA/FAO), Kigali Rwanda. Groupement Semis Direct de Madagascar (GSDM). 2004. Stratégie du GSDM pour la mise au point, la formation et la diffusion des techniques agro-écologiques à Madagascar. Ministère de l'Agriculture, de l'Elevage et de la Pêche, Antananarivo, Madagascar, Edition de 2007. GSDM, 38pp. Hatibu N., Mutabazi K. , Senkondo E.M, Msangi A.S.K. 2006. Economics of rainwater harvesting for crop enterprises in semiarid areas of East Africa. Agricultural Water Management 80, 74–86 Hatibu, N., 2000. Introduction. In: Rainwater harvesting for natural resource management: A planning guide for Tanzania. Hatibu, N. and Mahoo, F. (Eds). Technical Handbook No. 22. RELMA, Nairobi. pp. 1-5. Hatibu, N., E. Lazaro, Mahoo, H.F., 1997. Farming systems assessment of rainwater harvesting for crop production in Tanzania: Case of Bahi-Sokoni and Uhalela villages in Dodoma district. SWMRG; FAO-AGSP Project. pp 82. Hatibu, N.; Mahoo, H. F.; Kajiru, G. J. 2000. The role of RWH in agriculture and natural resources management: From mitigating 35

droughts to preventing floods. In Rainwater Harvesting for Natural Resources Management, eds. N. Hatibu; H.F. Mahoo. A planning guide for Tanzania. Technical Handbook No. 22. RELMA, Nairobi. 58-83. Henri de Laulanie. 1983. Le riz à Madagascar. Un développement en dialogue avec les paysans. Association Tefy Saina. Ambondrona. Madagascar. IFAD 2000. Knowledge Management . A thematic Review. IFAD support for water management and irrigation in east and Southern Africa. IFAD, Rome. IFAD, 2002. Assessment of Rural Poverty in Eastern and Southern Africa. International Fund for Agricultural Development. IMAWESA (2007). Policy Formulation & Harmonization with respect to Water Management for Smallholder Agriculture: Experiences from 15 Countries in Eastern and Southern Africa. Proceedings of the IMAWESA Joint Learning Workshop; Mauritius, January29th to February 2nd 2007. Mati, B.M., Hatibu, N.H., Mulinge, W., Balloo M. and Phiri, S.K. (eds). SWMnet Proceedings 10. Nairobi. Inocencio, A., D. Merrey, M. Tonosaki,, A. Maruyama, I. de Jong, and M. Kikuchi. 2007. Costs and Performance of Irrigation Projects: A Comparison of Sub-Saharan Africa and Other Developing Regions. IWMI Research Report 109. Colombo, Sri Lanka: IWMI. IWMI (2006) Agricultural water management technologies for small scale farmers in Southern Africa: An inventory and assessment of experiences, good practices and costs. Final report produced by the International Water Management Institute (IWMI) Southern Africa Regional Office Pretoria, South Africa. For Office of Foreign Disaster Assistance, Southern Africa Regional Office, United States Agency for International Development. Jagwe, J.N., Okoboi, G., Hakizimana, P., Tuyisinge, J., Rucibigango, M., 2003. Marketing Survey of the Rice sub-sector in Rwanda. ATDT-CIAT/ISAR/IITA-FOODNET Monograph. Kajiru, G.J., 2006. Soil fertility status and response of rice to Nitrogen, Phosphorus and Farmyard manure under rainwater harvesting systems in semi arid areas of Maswa, Tanzania. PhD. 36

Thesis, Soil Science Department. Sokoine University of Agriculture. Morogoro, Tanzania. Kambewa, D. 2004. Patterns of access and use in wetlands: Lake Chilwa Basin. Research Report to BASIS CRSP, October 2004, Madison, WI. Kauffman, J.H., Mantel, S., Ringersma, J., Dijkshoom, J.A., van Lynden, G.W.J., Dent, D.L., 2003. Making better use of green water under rain-fed agriculture in sub-Saharan Africa. In: Proceedings of the symposium and workshop on water conservation technologies for sustainable dryland agriculture in Sub-Saharan Africa (WCT). Held at Bloem Spa Lodge and Conference Centre, Bloemfontein, South Africa 8-11 April 2003. pp 103-108. Lenton, R., and M. Muller, eds. 2009. Integrated Water Resources Management in Practice: Better Water Management for Development. Global Water Partnership. London: Earthscan. Lofstrand F., 2005. Conservation Agriculture in Babati district, Tanzania. Impacts of conservation Agriculture for small-scale farmers and methods for increasing soil fertility. Master of Science Tesis. Swedish University of Agricultural Science. Dep. of Soil Science. Uppsala, Sweden. 84pp. Lundgren, L. and Taylor, G. (1993) From Soil Conservation to Land Husbandry. Guidelines Based on SIDA’s Experiences. Natural Resources Management Division, SIDA. Stockholm. Mahoo, H.F., Admassu, H. Rao, KPC, Mugugu, A.T. and Mati, B. M., 2007. Agricultural Water Management, A critical Factor in the Reduction of Poverty and Hunger: Principles and Recommendations for Action to Guide Policy in eastern and southern Africa. IMAWESA Policy Report 1. Nairobi. Malesu, M. M J. K. Sang, A. R. Oduor, O. J. Odhiambo and Nyabenge, M, 2006. Hydrologic impacts of ponds on land cover change: Runoff water harvesting in Lare. Kenya. 2006. Technical Report No. 32 Nairobi, Kenya: Regional Land Management Unit (RELMA-in-ICRAF) 37

Mangisoni, J. 2008. Impact of Treadle Pump Irrigation Technology on Smallholder Poverty and Food Security in Malawi: A Case Study of Blantyre and Mchinji Districts. International Journal of Agricultural Sustainability 6 (4): 1-19. Mangisoni, J.H. 2008. Impact of treadle pump irrigation technology on smallholder poverty and food security in Malawi: a case study of Blantyre and Mchinji districts. International Journal of Agricultural Sustainability, 6 (4):248-266. Mariki, W.L., 2004. The impact of Conservation Tillage and Cover Crops on Soil fertility and crop production in Karatu and Hanang Districts in Northern Tanzania. TFSC/GTZ Technical Report, Arusha Tanzania 1999 – 2003. Mati, B.M, Mwepa, G. and Temu, R. 2008. Farmer Initiatives in Managing Water for Agriculture in Eastern and Southern Africa. A booklet of farmer innovations in agricultural water management in Eastern and Southern Africa. IMAWESA Working Paper 15. Nairobi. Mati, B.M. 2006. IMAWESA Stakeholder Partnerships, Learning Needs, and Baselines for Project M& E. Report of a Baseline Assessment. SWMnet Working Paper 12. Nairobi. Mati, B.M. 2007. 100 Ways to Manage Water for Smallholder Agriculture in Eastern and Southern Africa. A Compendium of Technologies and Practices. SWMnet Working Paper 13. Nairobi, Kenya. Mati, B.M. 2008. Capacity development for smallholder irrigation in Kenya. Irrigation and Drainage. IRD 57(3): 332-340. Wiley InterScience Mati, B.M. and Penning de Vries, F.W.T. 2005. Bright Spots on Technology-Driven Change in Smallholder Irrigation. Case Studies from Kenya. In: Bright Spots Demonstrate Community Successes in African Agriculture. International Water Management Institute. Working Paper 102: 27-47. Maweru, S.C.Y. (2004). Current initiatives on microirrigation and smallholder markets in Malawi. pages 19-24. In: Mangisoni, J.H. (Ed) (2004). Proceedings of Stakeholders’ Experience and Lesson Sharing Workshop on Small Scale Irrigation and Market 38

Development held at Malawi Institute of Management (MIM), Lilongwe, 27th – 29th January, 2004. Meertens, H. C. C., Kajiru, G. J., Ndege, L. J., Enserink, H. J. and Brouwer, J., 2003. Evaluation of on farm soil fertility research in the rainfed lowland rice of Sukumaland, Tanzania. Experimental Agriculture. 39: 65-79pp. Messages for action. Paper presented at the ICID 2nd African Regional Conference, SANCID, Johannesburg (Nov. 6, 2007). The World Bank & INPIM. Mloza-Banda, H.R., J.W. Banda, R. Lunduka and Y.M. Mohamoud. 2001b. The dambo ecosystems of the Diamphwe and Lilongwe River catchments.II. Appraisal of potential indicators of limits to resource availability and use. UNISWA Research Journal of Agriculture, Science and Technology, 5(2):148-158. Molden, D. (ed). 2007. Water for Food, Water for Life: The Comprehensive Assessment of Water Management in Agriculture. Earthscan, UK. Molden, D., Frenken, K., Barker, R., de Fraiture, C., Mati, B., Svendsen, M., Sadoff, C and Finlayson C. M. 2007. Trends in water and agricultural development. In: Water for Food, Water for Life: A Comprehensive Assessment of Water Management in Agriculture. London: Earthscan and Colombo: IWMI, 57-89. Mulwafu, W., C. Chipeta, G. Chavula, A. Ferguson, B.G. Nkhoma, G. Chilima. 2003. Water demand management in Malawi: problems and prospects for its promotion. Physics and Chemistry of the Earth, 28:787–796 Nega, H and Kimeu, PM.2002. Low -cost method of rainwater storage: Result from field trials in Ethiopia and Kenya. RELMA Technical Report Series 28. Negassi, A.; Tengnas, B.; Bein, E.; Ghebru, K. 2000. Soil conservation in Eritrea. Some case studies. Technical Report No. 23. Regional Land Management Unit, Nairobi NEPAD (New Partnership for Africa’s Development). 2003. Comprehensive Africa Agriculture Development Program (CAADP). Midrand, South Africa: NEPAD. 39

Ngailo, J.A., Kaswamilla, A.L. and Senkoro, C.J., 2007. Rice production in Maswa District Tanzania and its contribution to poverty alleviation. RAPOA Report. Nissen-Petersen, E., 1982. Rain Catchment and Water Supply in Rural Africa: A Manual. Hodder and Stoughton, Ltd., London. Oweis, T.; Prinz, P.; Hachum, A. 2001. Water harvesting. Indigenous knowledge for the future of the drier environments, ICARDA, Aleppo, Syria. Penning de Vries, F.W.T., Mati, B., Khisa, G., Omar, S. And Yonis, M. 2005. Lessons Learned from Community Successes: A case for Optimism. In: Bright Spots Demonstrate Community Successes in African Agriculture. Colombo, Sri Lanka: International Water Management Institute (IWMI). Working Paper 102: 1-6. Reij, C.; Scoones, I. and Toulmin, C. (eds). 1996. Sustaining the Soil: Indigenous Soil and Water Conservation in Africa. London: Earthscan Publications. Reij, C.; Waters-Bayer, A., eds. 2001. Farmer innovation in Africa. A source of inspiration for agricultural development. London, U.K.: Earthscan. Sambrook, Clare Bishop-, Kienzle, J., Mariki, W., Owenya, M. and Ribeiro, F., 2004. Conservation agriculture as a labour saving practice for vulnerable households. A study of the suitability of reduced tillage and cover crops for households under labour stress in Babati and Karatu districts, Northern Tanzania. IFAD and FAO. Rome Italy. Séguy, L., Husson, O., Charpentier, H., Bouzinac S., Michellon R., Chabanne A., Boulakia S., Tivet F., Naudin, K., Enjalric F., Chabierski S., Rakotondralambo P., Rakotondramanana. 2009. La gestion des écosystèmes cultivés en semis direct sur couverture végétale permanente. SIWI, 2001. Water harvesting for upgrading of rain-fed agriculture. Problem analysis and research needs. Report II. Stockholm International Water Institute. Stockholm, Sweden: Stockholm International Water Institute. 40

Van Koppen, B., 2002. Water reform in Sub-Saharan Africa: What is the difference? A paper presented at 3rd WaterNet/Warfsa Symposium 'Water Demand Management for Sustainable Development', Dar es Salaam, 30-31 October 2002 pp. 8. Van Koppen, B., P. Moriarty, and E. Boelee. 2006. Multiple-use Water Services to Advance the Millennium Development Goals. IWMI Research Report No. 98. Colombo, Sri Lanka: IWMI, IRC, CPWF. Van Leeuwen, N. 2003. Agriculture and Water Management for Crops. In: Gopal, B., Pathak, P.S., Raman, A. and Lee, S.Y. (Eds) Special Issue: Wetlands and Agriculture. International Journal of Ecology and Environmental Sciences. 29 (1):11-14. Vancampenhout, K., Nyssen, J., Gebremichael, D. Deckers, J., Poesen, J. Mitiku Haile, Moeyersons, J. 2005. Stone bunds for soil conservation in the northern Ethiopian highlands: Impacts on soil fertility and crop yield, Soil & Tillage Research (2005). Vermeulen, M. 2009. More from less, from less to more: Dissemination of a rice cultivation technique. In Farming Matters, LEISA 25(4):22-24. UN (United Nations) Water/Africa. 2003. Africa Water Vision 2025. Economic Commission for Africa. Addis Ababa. Wiyo, K.A. and Z.M. Kasomekera. 1994. Study on Dambo Farming Communities in Malawi (Cash or subsistence farming?). Bunda College of Agriculture, University of Malawi, Lilongwe, Malawi. 157 pp. Wiyo, K.A., R.W. Lunduka and P.C. Nalivata. 2002. Malawi Treadle Pump Impact Assessment Study. Department of Irrigation, Ministry of Agriculture and Irrigation. Lilongwe, Malawi. World Bank. 2006. Re-engaging in Agricultural Water Management: Challenges, Opportunities and Trade-offs. Directions in Development Report. Washington, DC: World Bank.

41

2. Effects of Long-Term Soil and Water Conservation on Agricultural Productivity: A Case Study of Anjenie Watershed, Ethiopia Adgo, E.T. and Akalu, T.F. Introduction Agricultural development in Ethiopia is hampered by many factors, key of which is land degradation, which threatens to a large extent, the sustainability of agricultural production in the country (HansJoachim et al., 1996). Ethiopia is a mountainous country with about 45% of its land in highland zones at altitudes exceeding 1,500 m.a.s.l. The mountainous and rugged terrain makes the country prone to land degradation. Among the different forms of land degradation processes in Ethiopia, soil erosion by water is the most serious, threatening food security, environmental sustainability and prospects for rural development in the country. Soil erosion is not a new phenomenon in Ethiopia, being a process as old as the history of agriculture in the country (Hurni, 1988). Ethiopia has been identified as containing some of the most seriously eroded areas in the world (de Graaff, 1996), with an estimated annual soil loss of about 42 t ha-1 yr-1 from croplands, resulting in an annual loss in crop production equivalent to 1 to 2% (Hurni, 1993). The problem of land degradation is particularly critical in the highlands. The recurrent incidents of famine and starvation in Ethiopia have been blamed to a certain extent, on the soil erosion (Hurni, 1988). Despite extensive soil erosion, the importance of soil and water conservation was largely neglected in Ethiopia prior to 1974. The problem attracted the attention of policy makers only after the disastrous drought and famine of 1974 following which the Ethiopian government initiated massive soil conservation programs based on the 1975 land reforms. Investments in soil and water conservation made by the government and nongovernmental 42

organizations over the last three decades show viable benefits (Nyssen, 1998; Yohannis, 2005). Meanwhile, the adoption rates of soil and water conservation technologies show mixed results. Terrace technologies are well adopted by farmers and doing well in specific environments and socio-economic contexts. In other areas, the adoption rate has been very low due to diverse perceptions of farmers regarding the threat of soil erosion, household size, land and farm characteristics, technology-specific attributes, land quality differentials and tenure insecurity (Bekele and Holden, 1998). Investment in soil and water conservation contributes to intensification of agricultural systems, enhancing food production and alleviating poverty. In particular, terrace technologies control soil erosion by reducing the slope of the cultivated land and this facilitates the conservation of moisture for crop use, which in turn leads to increased crop yields. Cognizant of these potentials, huge investment have been made in different parts of the country. Although a number of biophysical research activities in soil and water issues have been implemented in the Anjenie watershed, impacts of soil and water conservation initiatives on improving yields and incomes of the beneficiary households remained largely un-quantified. This study was initiated to bridge this knowledge gap. The main objective therefore, was to quantify the contribution of long-term soil and water conservation initiatives on crop productivity, profitability, returns on investment and socioeconomic implications, towards improving rural livelihoods in Ajenie watershed of Ethiopia. The study area Anjenie watershed is located in the central highlands of Gojjam, Amhara Region of Ethiopia, about 370 3’ E, 1004’ N and about 260 km south east of Bahir Dar town. The watershed lies at altitudes about 2,400 m.a.s.l. The watershed covers an area of 108 ha but the size of the study area is about 113.4 ha. It is home to 95 households having a total population of about 512. Soil and water conservation measures include fanya juu terraces and grass strips constructed in 43

1984 by the then Soil and Water Conservation Project (SCRP), which was initiated by Bern University of Switzerland in collaboration with the Ethiopian Ministry of Agriculture. Terrace construction was done by local communities without payment for individual participating farmers. As an incentive, a health clinic was constructed by the project to compensate for labour and material contributions by the community. Agricultural systems in the study area are typical of both the upland cereal-based system and the smallholder crop-livestock mixed system of agriculture, growing barely, wheat, teff, maize and legumes as major crops. Teff and barley are the predominantly cultivated crops followed by maize and wheat. Over 80 % of the cultivated area is occupied by cereals. Oil crops and legumes are cultivated on smaller areas (Kohler, 2005). The average land holding is significantly lower than 1 ha. Before the introduction of terraces, farmers faced serious soil erosion problems which adversely affected soil moisture availability and soil nutrient status. This led to low land productivity and poor crop yields. As a result, food insecurity and poverty were common problems. Methods and data collection Anjene watershed was selected for this study because nearly all the households in the watershed had adopted the terracing. The area is a model site for soil and water conservation activities for the Amhara region as well as the nation. Moreover, the terraces at Anjenie are well maintained and there has been expansion of newly terraced farm lands in the surrounding villages since the project closed. For data collection, Anjenie watershed was divided into three sampling strata; the upper, middle and lower part of the watershed. This is because the upper zones are steeply sloping, but the slope decreases as one moves down the watershed profile. From each stratum, farmers growing at least one or more of the test crops; teff, barely, and maize, were selected using the stratified purposive random sampling techniques. Questionnaire surveys were administered to 20 farmers randomly selected from each stratum of the watershed and 44

from adjacent un-terraced farmlands for comparison. In total, 120 farmers were interviewed across the two treatments (with and without terracing). Other data were collected from records kept by the Regional and District Agricultural offices, as well as from published and unpublished reports on terracing, farming systems and other socio-economic data. Further, group discussions were undertaken with extension staff and leaders in the watershed. Cost-Benefit Analysis Cost benefit analysis were done using the average prices of inputs and outputs. The average yields across the topo-sequence were obtained for the upper, middle and lower catchment of the watershed. Costs of family labour were calculated using the opportunity cost of labour in the area which US$0.93 adult-day-1. The benefits from crop residues and grass were considered in this analysis because straw and grass have market value in the catchments and used for animal feed. The main cost components were fertilizer, seed, labour and animal power. Farmers used their own planting material by putting aside viable seeds from previous year’s harvest and they rarely bought certified seed. Therefore, the price of the seed was the average purchasing price between farmers. Sometimes herbicides were used by the interviewed farmers and hence were included in the cost-benefit analysis. Labour costs were considered for the preparation of land, planting, weeding, harvesting, threshing and marketing activities. Results and Discussion Viability for improving productivity The mean yields of teff, barely and maize on terraced fields were 0.95, 1.86 and 1.73 t ha-1 respectively (Figure 2.1). The corresponding values for un-terraced farms were 0.49, 0.61 and 1.69 t ha-1. The respective variations in yields across the slope strata are presented in Figure 2.1. Terraced farms were obviously more productive than un-terraced ones showing an average yield 45

increment of 93, 203 and 125% respectively for teff, barely and maize, as compared to without terraces. These results are in line with the findings of Vancampenhout et al. (2005), who has also found similarly positive effect of soil conservation on the yield of the crop in the highlands of Ethiopia. On the other hand, Herweg and Ludi (1999) and Menale et al. (2008) found that fanya juu, soil/stone bund, grass strips did not increase crop yield and biomass production in the highlands of Ethiopia and Eritrea. It was found that unless productivity was increased, for example, by increasing fodder grass production on bunds, soil and water conservation measures could not be characterized as a “win-win” measure to reduce soil erosion. This is contrary to the win-win findings of this study. There were variations of yield data along the top sequence of the terraces, but this variability did not conclusively show a marked pattern of the different crops and treatments. The percentage yield increases showed that barely benefited the most from terracing as the yields were doubled. This might be associated with the positive effects of terracing in improving moisture availability and conservation of soils. Another observation was that farmers having terraces could produce two crops of barley per year, in the main season and using residual moisture. This is possible because farmers plant barely early in the rainy season in the beginning of May and harvest at the end of August/ beginning of September. After the first crop is harvested, farmers immediately prepare their land and sow again the second barley crop until the third week of September. The rain ceases usually end of September/first week of October and the second barely crop is then grown using residual moisture and harvested in December/January. Double cropping of barley was rare on un-terraced plots.

46

Crop yield (t/ha)

2.50 2.00 1.50 1.00

Upper catchment Middle catchm ent

0.50

Lower catchment Average yield

0.00 Terrace Teff

Unterraced

Terrace

Unterraced

Terrace

Barely

Unterraced

Yield advantage

Maize

Crop type and slope treatment

Figure 2.1: Average yields of teff, barley and maize in Anjenie watershed with and without terraces Another new development associated with terracing in Anjene was the introduction of maize, a crop that was never grown in the watershed before the project was introduced in the 1980s. Maize is also found outside the catchment area but usually around homesteads where soil fertility is much better than in fields far from the village. The increased presence of maize crop in the watershed was associated with improvements in soil moisture and nutrient retention within the terraced fields. Grasses grown on terraces and increased crop residues from terraces were also additional incentives for farmers, as these were utilized as livestock fodder. Generally, these results show that terracing has long-term positive effects on crop productivity through improved soil and moisture conservation. Terracing practices are disseminated and widely accepted in the area. This negates claims that small holder-farmers are unlikely to adopt high cost technologies unless substantial subsidy is provided (Bekele and Holden, 1998, 2001). Investment in soil and water conservation has greatly contributed to land productivity.

47

Viability for increasing farm incomes The gross values, aggregated expenses and net incomes from terraced and un-terraced farms, for each of the three crops are presented in Table 2.1. It shows a clear advantage in terracing, especially in gross revenues. However, there were slightly higher expenses for terraced fields compared to un-terraced, which was attributed to the extra costs of maintenance of the structures. Table 2.1: Cost- benefit analysis of teff, barely and maize with and without terraces in Anjenie watershed Crop /treatment Teff Terraced Un-terraced Barely Terraced Un-terraced Maize Terraced Un-terraced

Gross Revenue (US$ ha-1)

Total Expenses (US$ ha-1)

Net profit* (US$ ha-1)

292.6 144.1

271.7 256.3

20.9 -112.2

382.3 98.5

197.1 139.6

185.2 -41.1

245.7 102.2

280.2 203.0

-34.5 -100.8

* Net profit excluding family labour The net income (profit) showed a clear advantage for terracing, reflecting positive incomes for barley and wheat. This result is in agreement with the findings of Bekele (2005) who estimated positive contributions of level bunds compared to ‘without’ in Hundi-Lafto catchment of Ethiopia. However, negative incomes from all unterraced and terraced maize fields are a reality facing farmers in the region. The costs of inputs sometimes outweigh the income benefits, even though the added value of terracing is reflected in the yields. In addition, the values were depressed by the inclusion of family labour. These findings agree with similar ones from Kenya’s Kaiti catchment (Mulinge et al. 2010 of this book), where positive impacts on yields were obtained but the net incomes were negative, 48

a factor associated with high costs of inputs, and poor commodity prices.

Profitability of terracing excluding family labour The profitability analysis excluding family labour was calculated using the average prices of inputs and outputs around Anjenie. The respective values of net profit are shown in Figure 2.2. Teff, barely and maize grown on terraces earned incremental income advantage amounting to US$159, 275 and US$141 ha-1yr-1 over the un-terraced farms. This indicates that though the opportunity cost of family labour was accepted to be low (US$1.7 day-1), the share of family labour costs from the total production costs in terraced plots was significantly high. Investment in soil water conservation can thus greatly contribute to increased income and household food security. These results are therefore consistent with the findings of Ludi (2004) who claimed that the profitability of SWC investments depends on cropping intensity.

P r o fita b ilit y (U S $ /h a )

400 300

Upper

200

Middle Lower

100

Average profit Income advantage

0 -100

Terrace

Un-terraced Teff

Terrace

Un-terraced

Barley Crop type and slope treatment

Terrace

Un-terraced Maize

Figure 2.2: Profitability of terraced and un-terraced farms excluding family labour Economic return on family labour It was found that family labour requirements per hectare for crop production were higher under terraces than without (Figure 2.3). This is because terracing requires more labour for construction and 49

maintenance and it needs additional labour for husbandry, harvesting and marketing activity for marginal output. Production of teff in terraced plots absorbed the highest amounts of family labour at 203 adult-days ha-1, followed by maize (178 adult-days) and barely (135.5 adult-days). This was because teff requires a fine seed bed, necessitating multiple ploughing and trampling before sowing, and frequent weeding more than barley. Moreover, teff is the major cash crop in the area and it requires additional labour to transport to the market. There is higher family labour demand on terraced fields as compared to un-terraced ones since the terraces have to be rebuilt regularly.

Labour inputs (man-day ha -1)

250 200 150

Upper catchment Middle catchment Lower catchment Average Additional labor

100 50 0 Terraced

Un-terraced Teff

Terraced

Un-terraced

Terraced

Barley Types of crops and treatm ent

Un-terraced

Maize

Figure 2.3: Family labour inputs for teff, barley and maize in Anjenie watershed

One of the parameters considered in this study was the return on family labour1. Gross returns on family labour are presented in Table 2.2. Family labour was the main source of labour for crop Return on family labour is determined by subtracting all costs related to crop production from the total revenue excluding family labour inputs. Divided this net profit with the number of family labour inputs in adult-days then it gives gross return on family labour. 1

50

production in both terraced and un-terraced fields. Increased production and cropping intensity need additional labour per unit area and year. Barley, teff and maize production on terraced fields had an average gross return on family labour of US$2.3, 1.0 and 0.74 per adult-day, respectively while the respective values without terracing were US$0.44, 0.27 and 0.16 per adult-day. This means that terraces provided higher gross return on family labour and thus, increased the incomes of rural farm households. Surprisingly, barley on terraced fields had higher return on labour, that is, US$2.3 per adult-day. The gross return on family labour from terraced fields was higher than the opportunity cost of labour in the study area. Assuming an income of US$1.0 per day as the threshold for poverty line, terracing can enable the rural poor to move out of poverty. Generally, terraces brought higher return on family labour. They create job opportunities and improve rural incomes and ultimately rural livelihoods. However, terracing takes up more family labour which is a limitation to vulnerable and elderly farmers. The average marginal returns on family labour (Table 2), had positive values for teff and barley, being US$5.60 and 5.56 per adult-day respectively. Meanwhile, marginal returns on family labour was negative for maize (US$-28.3 per adult-day). This implies that an additional labour requirement for teff and barely production on terraced fields had generated additional revenue. In the case of maize, additional labour was retrogressive. The rate of marginal return on family labour, meaning shifting from non-terracing practices to terraced farming practices was accompanied by positive returns for many of the studied crops. Generally, gross and marginal returns on family labour were higher for terraced fields as compared to that from un-terraced ones, indicating that investment in terracing improved incomes to family labour.

51

Table 2.2: Gross and marginal return on family labour Gross return (US$ per adult-day) Crop/ Catchment

Teff Terrace

Barley

Unterraced

Maize

Terrace

Unterraced

Terrace

UnTerraced

2.66 2.28 2.03 2.32

0.85 0.46 0.03 0.44

0.77 0.73 0.73 0.74

0.59 0.02 -0.13 0.16

Upper Middle Lower Average

1.09 0.96 1.00 1.01

0.16 0.34 0.33 0.27

Upper Middle Lower Average

Marginal return (US$ per adult-day) Teff Barley 7.95 7.3 3.7 4.5 5.3 4.9 5.6 5.6

Maize 1.2 1.5 - 87.6 -28.3

Scope in improving water productivity Water productivity2 was calculated based on the total and effective rainfall amount in the area and expressed in terms of kg grain or its monetary value per mm water consumed. The average long-term amount of total rainfall in the area, 1,690 mm yr-1 and estimating a run-off coefficient of 25% were the basis for the calculation. The length of growing period of teff, barley and maize was also considered in calculating the values presented in Table 2.3. It was observed that terraces improved water productivity of the three crops by at least 100% against un-terraced plots, which clearly shows the advantage of terracing in terms of efficient use of rainwater. Terraced barley had the highest water productivity in 2 Water productivity value was calculated as crop yield or its monetary values divided by the amount of water utilized during the vegetation period.

52

terms of grain yield per mm of water consumed (1.35 kg mm-1) followed by maize (1.21 kgmm-1) and teff (1.0 kg mm-1) respectively. On the contrary, the calculated monetary values (economic water productivity) showed that barely and teff were almost similar (0.28 and US$0.31 mm-1 respectively), a factor associated with the fact that both teff and wheat take just three months in the field while maize (US$0.17 mm-1) had the lowest economic productivity of water. Table 2.3: Water productivity of major crops at Anjenie Watershed Crops/ Systems

Teff Terraced Unterraced

Crop water Water Productivity requirement (mm season-1) kg mm-1 US$ mm-1

Water productivity at 25% run-off kg US$ mm-1 mm-1

933 933

1.01 0.52

0.31 0.15

1.35 0.70

0.42 0.20

Barley Terraced Un-terraced

1358 731

1.35 0.86

0.28 0.13

1.80 1.11

0.37 0.18

Maize Terraced Un-terraced

1445 1445

1.21 0.56

0.17 0.07

1.61 0.74

0.22 0.10

Returns on investment The parameters considered in this analysis were net present value (NPV) and financial internal rate of return (IRR) on investment (Table 3.4). Although terraces remain for a long time after construction, a lifespan of 50 years was chosen for this calculation. Initial investment cost of terracing was equivalent to US$46.51 ha-1 in the first year of construction, requiring about US$48.8 ha-1 to 53

maintain in the first three years. Using a discount rate of 10%, the average NPV of investment in terraces for barely production over a period of 50 years was US$1,542. Table 3.4: Net present value, financial internal rate of return on investment for barley Performance Parameters

Upper

Middle

Lower

Average

Discount Factor

10%

10%

10%

10%

Discounted costs(US$ ha-1)

1616

2107

2011

1911

Discounted Benefits(US$ ha-1)

3114

3881

3366

3454

NPV (50)

1498

1774

1355

1542

IRR

291%

356%

258%

302%

ROI

108%

95.1%

77.7%

93.6%

Across the catchment, the NPV were found to be positive, indicating that terraces were financially viable. Investment in terracing is feasible and financially promising (positive net present value). Moreover, the IRR was calculated and found to be 302% (average of the three catchments). This figure is higher than the discounted factor at 10%, indicating again the financial viability of terracing. The findings of this study are different from those by Shiferaw and Holden (1998) and Kappel (1996) who showed that, except for low-cost technologies like grass strips, returns on soil conservation investments were low, especially when the rate of discount may be high and subsidy was suggested as an incentive for adoption of SWC practices. Posthumus and de Graaf (2005) found negative NPV values for bench terraces in Peru when crop yield data were actually measured and profitability was lower than farmers' estimation. 54

Social, environmental and economic sustainability The introduction of soil and water conservation activities in Anjenie watershed in 1984 did not adopt the food-for-work compensation system popular at the time. Instead, farmers worked voluntarily and were rewarded with the construction of a health centre by the project. This approach was successful in instilling a sense of ownership and responsibility by the community. Therefore, the soil conservation structures and the clinic still exist and are functional after 25 years. The terraces which are well maintained, have since stabilized, and in many places, developed into level benches. The maintenance and management of the terraces has been the full responsibility of the farmers and these activities are carried out well and on time. Two decades ago, Anjene catchment was highly eroded and farmers could not produce enough food and thus the community was highly food insecure, a situation that caused regular out-migration of inhabitants to other areas in search of food in times of even minor droughts. The introduction terraces improved crop production thereby reducing seasonal migration of farmers. Fodder became available and improved animal production. Moreover, farmers around Anjenie have started to grow maize crop in the area. Maize was not grown in the watershed previously due to water stress. But terracing has improved moisture conservation and hence crop diversity in the area. Farmers pointed out that during the first five years, there were problems of oxen ploughing due to the narrow spacing between the terraces. As a result, modifications of bunds (removal of alternate bunds) ensured wider terrace spacings suitable for oxen to plough. Soil and water conservation practices have been widely adopted over the years, with many farmers constructing terraces on their own initiatives beyond the Anjenie catchment. Terracing also reduces the risk of crop failure during dry spells in the rainy season, due to the water conserving effect (Posthumus and de Graaf, 2005). The economic sustainability of terraces could vary across agro-ecological zones as crop productivity; prices and opportunity cost of labour vary and thus, are site-specific. Better 55

security in land tenure could help improve commitment by farmers to conserve the soil in the area (Tenge et al., 2004). Conclusions This study confirmed that soil and water conservation has had longterm benefits to smallholder farmers especially in the Anjenie catchment for teff, barley and maize. Increased crop yield was attributed to terracing, with increases of 93%, 203% and 125%, respectively, as compared to crops grown in the same area but without terraces. This implies that investment in soil and water conservation increases crop production and land productivity. The results of the study also indicated that teff, barley and maize grown on terraced plots enabled net values of US$20.9, 185.2 and -34.5 ha1 , respectively. On the other hand, crop grown without terrace had negative net returns of US$-112, -41, and -100.8 ha-1 for teff, barley and maize, respectively, a factor associated with high costs of inputs and labour demands. It was also found that barley, teff and maize with terrace had an average gross return on family labour of US$2.33, 1.01 and 0.74 adult-day-1, respectively, fivefold higher than the un-terraced fields. Water productivity of the different crops was also 100% higher than without terrace intervention. With the average NPV of US$1,542 and financial internal rate of return (IRR) of 302%, investment in terracing was financially realized. Generally, it was found that terracing had contributed to food security, and household income, thereby impacting on poverty reduction. Construction of soil conservation structures often has high initial costs and long payback periods. This could reduce household incomes in the short-term, but in the long term, the benefits far outweigh the costs.

56

References Graaff J de. 1996. The price of soil erosion: An economic evaluation of soil conservation and watershed development. Doctoral dissertation, Wageningen University Herweg, K. and E. Ludi. 1999. The performance of selected soil and water conservation measures – Case studies from Ethiopia and Eritrea. Catena 36: pp. 99-114. Hurni H 1988. Degradation and conservation of the resources in the Ethiopian Highlands. Mountain Research and Development 8(2/3), 123–13 Hurni, H. 1993. Land degradation, famine, and land resource scenarios in Ethiopia. In: Pimentel D., [ed.]. World Soil Erosion and Conservation. Cambridge: Cambridge Studies in Applied Ecology and Resource Management. pp. 27-61. Hurni, H. and Tato, K. (eds). 1992. Erosion, Conservation and small-scale farming. Geographisca Bernesia, Walsworth Publishing Company, Missouri, USA. Hurni, H., 1990. Degradation and conservation of the resources in the Ethiopian highlands. Mountain Res. Dev. 8 (2/3),123±130. Kappel, Rolf. 1996. Economic analysis of soil conservation in Ethiopia: issues and research perspective Ludi, E. 2004. Economic Analysis of Soil Conservation: Case Studies from the Highlands of Amhara Region, Ethiopia. Geographica Bernensia, African Studies Series A 18. Bern: Institute of Geography: 416 pp. Menale Kassie, S. Holden, G. Köhlin, and R. Bluffstone. 2008. Title: Economics of Soil Conservation Adoption in High-Rainfall Areas of the Ethiopian Highlands Nyssen, J., 1998. Soil and water conservation in the Tembien Highlands (Tigray, Ethiopia). Bull. Soc. GeÂogr. LieÁge 35, 517.

57

Posthumus, H. and de Graaff, J. Cost-Benefit analysis of bench terraces. A case study in Peru. Land Degrad. Develop. 16: 1– 11 René Kohler. 2005. The Status and Dynamics of Agricultural Production and Productivity in a Small Catchment Area in Anjeni, Gojam, Ethiopia. A Master’s Thesis. University of Berne, Switzerland. Shiferaw B, Holden, S.T. 1998. Resource degradation and adoption of land conservation technologies in Ethiopian highlands: a case study in Andit Tid, North Shewa. Agricultural Economics 18: 233–248. Shiferaw B. and S.T. Holden. 2001. Farm-level benefits to investments for mitigating land degradation: Empirical evidence from Ethiopia. Environment and Development Economics (2001), 6 : 335-358 Cambridge. Yohanes Afework .2005. The status of soil and water conservation measures in Amhara Regional State. Environmental Protection Study, Policy and Regulation Department, Bahir Dar Vancampenhout, K., Nyssen, J., Gebremichael, D. Deckers, J., Poesen, J. Mitiku Haile, Moeyersons, J. 2005. Stone bunds for soil conservation in the northern Ethiopian highlands: Impacts on soil fertility and crop yield, Soil & Tillage Research (2005). Wagayehu Bekele 2005: Stochastic dominance analysis of SWC in subsistence crop production in the Eastern Ethiopian Highlands the Case of the Hunde-Lafto area. Environment & Resource Economics (2005) 32: 533-550 Tenge, A. J., de Graaff, J. and Hell J. P. 2004. Social and economic factors affecting the adoption of soil and water conservation in West Usambara Highlands, Tanzania. Land Degrad. Develop. 15: 99–114.

58

3. Water Harvesting with Geo-membrane Lined Ponds: Impacts on Household Incomes and Rural Livelihoods at Minjar Shenkora, Ethiopia Akalu, T.F. and Adgo, E.T. Introduction Agriculture is the backbone of Ethiopia’s economy, contributing almost 50% of the GDP and employing 80% of the population. Because agriculture plays such a significant role, any fluctuation in the sector easily destabilizes the national economy. The structural problem with Ethiopian agriculture is its extreme dependence on rainfall (Devereux, 2000). Rainfall is so crucial that it is has become the single most important factor explaining the performance of Ethiopian economy from year to year. The challenge lies in the high variability and unpredictability of the rainfall patterns. This variability poses major limitations to agricultural production and productivity at household level, as much as to the national economy. It has been stated that a single drought occurrence in a 12-year period will lower GDP by 7% to 10% and increase poverty by 12% to 14% (UNDP, 2006). The rainfall variability is associated with frequent droughts and in many cases, crops suffer moisture stress even during normal rainfall seasons. Farm productivity has declined substantially and farmers find themselves sliding into poverty (Regassa et al., 2008). Ethiopia has 122 billion cubic metres of surface water and 2.6 billion cubic metres of ground water resources, but only 3 to 5% of the nearly 1 million ha of arable land is currently irrigated, while the irrigation potential has been estimated to be about 3.7 million ha (WCD, 2000). Similarly, the Amhara region is endowed with a potential irrigated land area of 600,000 ha within the four major river basins (BoWRD, 2005). Meanwhile, the potential for water harvesting using small dam storages and river diversions is considerable. However, the total area under irrigation to date amounts to only about 76,000 ha, which is less than 2% of the total 59

cultivated land in the region (BoWRD, 2005), but the potential water resources are not fully tapped. Increasing food supplies and economic prosperity for the rural poor in Ethiopia is expected to come from among others, the adoption of modern AWM, which is an entry point to ensure food security. Access to water will allow farmers to intensify agricultural production systems and thereby increase production and productivity. The vast majority of the rural poor rely on rainfed agriculture. The dominant smallholder agricultural systems of Ethiopia are not significantly benefiting from improved AWM technologies that could be used to alleviate food shortage and break the vicious cycle of poverty. Improved AWM innovations such as water harvesting could bring tremendous impacts to enhancing agricultural productivity and poverty alleviation. Since around 2003, the government of Ethiopia has invested heavily in small storage ponds in different parts of the country. However, the returns on investment and socio-economic impacts remain largely unquantified and thus unknown. The objective of this study, therefore, was to determine the impact of water harvesting with geo-membrane lined small storage ponds, on agricultural productivity, incomes and socio-economic impact to rural livelihoods in Minjar Shenkora District of Ethiopia. The study area The study was undertaken in Minjar Shenkora District of Amhara Region of Ethiopia. The district lies within the Rift Valley some 130 km southeast of Addis Ababa. The total area is 2,295 km2, having a population of 140,000 grouped into 29 peasants association Kebeles. The area traverses three major agro-ecological zones: highland (Dega) agro-climatical zone, which lies above 2,300 m.a.s.l., intermediate (Woinadega) agro-Climatic zone lying between 1,500 to 2,300 m.a.s.l. and lowlands (Kolla) agro climatic zone at below 1,500 m.a.s.l. The average land holding per household is about one hectare. The major crops grown in the area are teff, barley, wheat, sorghum, chickpeas, maize, field peas, onions, potatoes, pepper and 60

fenugreek. Livestock, has a major role in the farming system, especially the oxen which provide the draught power used during land preparation. Uneven and erratic rainfall is a common problem in Minjar Shenkora district. A majority of the farmers in the district have faced food insecurity due to prolonged dry spells. Lack of potable drinking water within realistic distance from homesteads is another problem in the district impacting on the workload of women and children who fetch domestic water since there are no permanent rivers in the area (MSWoARD, 2008). Rainwater harvesting therefore offers adaptable technology to provide water at home for various purposes, including agriculture. Design features of the water harvesting systems Rainwater harvesting initiatives were introduced in Minjar Shenkora in 2004, through a government-supported project on a pilot basis with 308 households. By 2008, the technology had expanded at exponential rates such that 7,618 households in the area had adopted water harvesting ponds by August 2008. The typical design of each pond is trapezoidal in shape, measuring 8 by 8 metres topwidth, 5 by 5 metres bottom width and 3 metres depth and a total storage volume about 102 m3. The plastic lining or geo-membranes are factory–manufactured with standard shape and size to fit these dimensions. The ponds are made with this standard design to enable bulk purchase and supply of geo-membranes, to make use of economies of scale. The cost of the geo-membranes was subsidized by up to 85% by the government, thus making them affordable to the poor. Participatory approaches were used to identify farmers, site selection and layout of ponds, with the assistance of technical staff. However, the excavation was done by the farmers and checked by technical staff for compliance. The farmer would then purchase the geo-membrane, install it and ensure safe use and management the runoff water and the pond. The volume of water harvested and stored in the ponds varies with agro-ecological zones, averaging 95 m3, 90 m3 and 80 m3 in the highland, midland and lowland areas respectively. However, water 61

retention within the ponds over time differs with individuals as well as with agro-ecology as affected by usage, evaporation and seepage losses. Treadle pumps are sometimes used to lift water by some of the farmers. Among most households, the water from the pond is used for domestic, livestock watering and supplemental irrigation, especially of horticultural crops. About 45% of the water is used for seedling and fruit production, 50% for livestock watering and 5% for domestic use. Thus, quantifying the yield and productivity of the total harvested water from each pond was difficult. On average, an area of about 100 m2 is cultivated with supplemental irrigation from each pond. Onion seedlings are grown under supplemental irrigation using the harvested water and transplanted in the field for bulb production under rain-fed conditions. Extra seedlings are sold to other farmers who have no ponds. Moreover, farmers produce several types of seedlings and fruits with the harvested water. This complicates quantification of yields and revenue. Sampling procedure, data collection and analysis Minjar Shenkora district was selected for this study due to its relative wealth of experience with water harvesting among smallholder farmers, as well as the relatively large number of households that had adopted the technology within a short time. A two-stage stratified random sampling method was used to select farmers for the study. At the outset, the district was stratified into three categories, that is, highland, midland and lowland. One Peasant Association (Kebele) from each strata was selected out of which 30 farmers were randomly selected for questionnaire interviews. In total, 90 farmers were interviewed and their ponds and farms assessed. Baseline data were collected from publications and records kept by the district, zonal and regional office of agriculture. These data covered issues such as water harvesting, farming systems and socioeconomic data. Also, supporting data were collected from the regional and district offices of the Ministry of Agriculture and Rural Development. Field data were then collected through farm visits 62

and interviewing farmers with questionnaires. In addition, group discussions were undertaken with agricultural experts and leaders of the respective Peasant Associations. The profitability analysis was done using the average costs of inputs and farm-gate prices produce, especially with onions as a test crop. Data analysis included partial budget analysis, gross margin analysis and water productivity. The discounted measures used to determine whether the technology was profitable or not also included the Internal Rate of Return (IRR) and the Net Present Value (NPV). Results and Discussion Viability for improving productivity The results are presented comparing a ‘before’ and ‘after’ scenario, in which farmers were originally reliant mostly on the field crops; teff and wheat, but ‘after’ the introduction of water harvesting, they could grow vegetables in small gardens in addition to the field crops. Of special interest were onions, which are used in the analysis. Onions were important because the availability of water in storages enabled farmers to grow onion seedlings during the dry season, making them available for planting at the onset of the rains. This in turn made it possible for onions to be grown as a rainfed field crop by more farmers, including those without storage ponds. The area became a source of onions as a marketable crop. As a ‘before’ scenario, the average yields of teff and wheat were 1.85 and 2.84 t ha-1 respectively. Generally, teff is considered a cash crop, earning on average farm gate price of US$0.48 ha-1 compared to wheat at US$0.28 ha-1. Consequently, although wheat has higher yields, the gross incomes for teff and wheat are US$871 ha-1 and US$801 ha-1 respectively. On the other hand (the ‘after’ scenario), onions are a bulky cash crop yielding on average, 13.36 t ha-1, and with an average farm gate price of US$0.17 kg-1, which translates to a gross income of US$2,258 ha-1. Thus, onions are a more profitable crop per unit area, with incomes far exceeding the total from teff and wheat combined. In addition, onion seedlings were also sold as cash crop, produced on plots measuring 100 m2 and utilizing about 40 m3 of the water from the 63

pond. Onion seedlings earned a gross income of US$229 (Table 3.1) when farmers sold their extra seedlings. As a result, the gross incremental income due to water harvesting by onion crop alone, adds up to U$2,487ha1. Thus, the added value of water harvesting is visible before and after the introduction of water harvesting. Table 3.1: Gross and net incomes from major crops at Minjar Shenkora Crops enterprise Teff rainfed(US$/ha) Wheat rainfed(US$/ha) Onion seedlings (US$/100m2) Field onions rainfed (US$/ha) Onions (seedlings +field crop) (US$/ha)

Gross Value 871 801 229 2,258

Costs of Inputs 348 276 74 410

Net Income 523 525 155 1,848

2,487

484

2,003

Net profitability excluding family labour The average net present value (NPV) of produce excluding family labour was US$1,848 ha-1 for bulb onions grown rainfed in the field, while teff and wheat, also rainfed, earned 523 and US$525 ha-1 respectively (Figure 1). But the average net income from onion seedlings alone was U$155 per 100 m2 plot, so the combined income from onions was US$2,003 ha-1. However, most farmers cannot plant a full hectare of onions as they need to grow other food crops on the same land. Normally, most farmers grow about 0.2 ha of onions, making the average take-home incomes from onions alone about US$400. This exceeds the dollar a day poverty threshold, and does not include incomes from other vegetables grown using the harvested water. Thus, water harvesting with small storage ponds can make major contributions to household incomes and rural poverty reduction. These results are consistent with the findings of Gezahegn and et al. (2006) and Nega and Kimeu (2002), who assessed smallscale water harvesting technologies in Ethiopia. 64

Net income(US$)

The costs of production are generally low since farmers utilize family labour and low levels of inputs. Family labour constituted about 72% of total cost of seedling production with water harvesting. Since labour is available in most households, adoption of water harvesting is affordable to the poor. 2400 2250 2100 1950 1800 1650 1500 1350 1200 1050 900 750 600 450 300 150 0

Teff rainfed ($/ha)) Wheat rainfed ($/ha) Seedling production with WH ($/100m2) Onion rainfed ($/ha) Incremental income due to WH intervention (onion crop only)

Crop system

Figure 3.1: Net incomes from major crops at Minjar Shenkora excluding family labour Scope for improving water productivity Water productivity depicts the overall efficiency of water applied and its conversion to a product (yield) or value of product (earnings), regardless of whether water was from rainfall, irrigation or supplemental irrigation. The water productivity of onion seedlings was calculated by dividing the net income with the amount of water applied during the cropping season water (40 m3) to irrigate the seedlings per unit area (100 m2). For the field crops, the water input was quantified as the average seasonal rainfall, which was 750 mm per season. Thus, the water productivity of onion seedlings using storage ponds was obtained to be US$38.7 mm-1, while that of rainfed bulb onions, teff and wheat were US$2.4, 0.69 and 0.69 mm1 per season, respectively. Accordingly, water productivity of onions was 3.5 times higher than that of wheat and teff. Whereas the water 65

productivity of onion seedlings was the highest, it should be noted that seedlings are not a final product. Bulb onions and seedlings have different market prices and thus influence values for unit of water consumption. This shows that production of high value crops such as onion seedlings with water harvesting made comparatively better use of scarce water and contributed to improved economic returns of onions grown under rainfed conditions. These results are consistent with the findings of Tadesse and Bekelle (2006) in Tigray region of Ethiopia. Returns on family labour Total family labour and gross economic returns on family labour3 are presented in Table 3.2. The family labour used in the production of bulb onions, teff and wheat per hectare was 150, 120 and 90 adult-days respectively. Meanwhile, onion seedling production utilized 13 adult-days meaning that production of onion crop utilized in total 163 adult-days. The higher labour demand for onions was because of the need for daily watering of the seedbeds which is usually done manually, while bulb production also utilizes more labour for agronomic management than cereal production. Moreover, the increased cropping intensity requires additional labour each growing season. Thus, water harvesting interventions involve additional labour, which could discourage some farmers and thus fit well in areas where labour is abundant. In Minjar Shenkora, water harvesting activities also helped increase off-farm employment opportunities for rural poor particularly during the construction of ponds. Consequently, water harvesting was used as a way of transforming “idle” human capital into financial capital.

3

Return on family labour is determined by subtracting all costs related to crop production from the total revenue excluding family labour inputs. Divided this net profit with the number of family labour inputs in adult-days then it gives gross return on family labour. 66

Table 3.2: Total family labour inputs and gross return on family labour Cropping Systems Teff rainfed Wheat rainfed Seedling production with WH Onion rainfed Incremental labour due to WH (onions only)

Total family labour 120 (adult-day ha-1) 90 (adult-day ha-1)

Return on family labour (US$ adult-day-1) 6.0 7.6

13 (adult-day/100m2)

13.6

150 (adult-day ha-1)

14.0

163 (adult-day yr-1)

15.0

It was found that the gross return on family labour from onion seedlings with water harvesting was US$13.6 manday-1, while incremental return on labour was US$15 manday-1. But the returns on family labour for rainfed wheat and teff were 7.6 and US$6 adult-day-1, respectively. This indicates that the returns on labour with water harvesting are higher than in rainfed systems. It should be noted that financial returns on family labour of high value crops like onions are produced either in rainfed or irrigated system. In terms of engaging more labour per unit area, making use of storage ponds demand obviously more family labour than under rainfed system. The return on family labour for storage ponds was higher than the opportunity cost (US$1.7 day-1) of labour for the study area. Moreover, considering incomes of US$1.0 day-1, as the threshold for poverty line, meaning that water harvesting ponds can help the rural poor move above the poverty line. Generally, results show that rainwater harvesting for production of onions had returns on labour around US$15 per adult-day invested. These benefits are very high due to the fact that without rainwater harvesting it is not possible to produce onion in the study area and rainfed wheat and teff crop realize returns on labour of around US$7 per adult-day. Thus, farmers with water harvesting have a bigger chance to move out 67

from poverty as compared to non-adopters. These results are consistent with the findings of Gezahegn and et al., (2006) with the role of water harvesting technologies in poverty alleviation in Ethiopia. Therefore, storage ponds are crucial to poverty reduction and improving the livelihoods of rural society by creating job opportunities (Hatibu and et al., 2006; Rebeka, 2008). In addition, storage ponds reduce wasteful labour engagement, such as fetching domestic water from distant streams or taking out the cattle for watering, because water is stored in the pond within or near the homestead. Returns on investment The financial analyses based on agricultural enterprises alongside water harvesting with storage ponds are presented in Table 3.3. The parameters considered were net present value (NPV) and financial internal rate of return (IRR). Initial investment costs of water harvesting were 154, 175 and US$187 per pond in highland, midland and lowland areas, respectively. The maintenance and production costs were US$48.8, 35.8 and 27.1 per pond in highland, midland and lowland areas, respectively. Farmers get benefit from water harvesting in many ways such as from seedling production, sales of water and animal watering. The gross incomes from seedling production were 301, 212 and US$174 in high, middle and lowland areas, respectively. However, these values do not include the water used for livestock and household consumption. If these other uses of harvested water were considered based on the opportunity cost of fetching water from distant areas, then the opportunity cost of labour would be US$0.87 day-1 or US$157 for the whole dry season. In general, the benefits of storage ponds with a discount rate of 10%, the average NPV of investment in storage pond over a period of seven years was about US$1,223. Across the different agro-ecological zones, the NPV values were positive, indicating that the water harvesting technology is financially viable. Moreover, the average IRR for the three agro-ecologies was 203%. This figure is higher than the discounted factor 10% again indicating

68

the financial viability of water harvesting. Generally, investment in storage ponds could improve the investor’s net worth. Table 3.3: Net Present Value, Financial Internal Rate of Return and Rate on Investment from onion seedlings Performance Highland Middle Lowland Average Parameters Discount Factor 10% 10% 10% 10% Discounted costs 333 301 277 304 (US$/100 m2) Discounted 1811 1459 1310 1527 2 Benefits(US$/100m ) NPV (7) 1477 1158 1033 1223 IRR 256% 189% 163% 202% ROI 514% 467% 468% 483%

Adoption of water harvesting ponds An important indicator of the success of water harvesting in Minjar Shenkora was the success rate at which farmers adopted the practice. The project started with 308 households in 2004 and by August 2008 the number of adopters increased to 7,618 households having water harvesting ponds, an adoption rate that can be described as exponential. It was not difficult to convince the farmers on the necessity of water harvesting since water scarcity is a major problem in the area. After the introduction of the ponds, distances travelled to fetch water for livestock and household consumption were drastically reduced releasing the time and labour for more socio-economic activities. Moreover, policy support like subsidy of the cost of the geo-membranes, agricultural extension and training initiatives contributed to the dissemination of the technology. For instance, the government provides subsidy by up to 85% of the cost of geo-membrane and 50% of the cost of treadle pumps used by farmers. Moreover, water harvesting has contributed to livestock

69

production by delivering water in the area and generally improved the standards of living of the farmers. Nearly all the ponds were in good condition and functional at the time of the study in 2007. The maintenance and management of the ponds were done by the farmers themselves. Surprisingly, several farmers had constructed more than one pond on their own initiative. Thus, Minjar- Shenkora district was a model site for water harvesting activities in the country. Water harvesting and storage in ponds also reduces runoff and soil erosion damage. It also improves incomes per unit area and reduces the pressure on land. Due to possibility to grow other types of seedlings using the harvested water, farmers started to grow tree crops like avocado, papaya and coffee, which make important economic use of land and water. However, some health hazards associated with water harvesting and storage include pests, especially mosquitoes. There were safety concerns since the ponds are open, while contamination of the water reduces its value for domestic use. Equitability across gender and wealth ranks Water harvesting in Minjar Shekora facilitated equitable availability of water as both men and women had benefited from the technology. Both undertook various activities in crop and livestock production activities. Storage ponds had brought tremendous advantages for women and children because they could produce vegetables which improved household nutrition, as well as incomes from the sale of some crops. Women also benefited from the ponds through reduced workload required to fetch water from distant locations. For instance, on average, water is fetched twice a day from a dam situated about 5 km from home, using a 20-litre container. This takes up about four hours per day. Assuming that the time spent could be used for income generating activities US$1.74 per day), the opportunity cost of fetching water is about US$0.87 per day or US$157 for the whole dry season. Farmers do not need a large amount of external inputs to construct the ponds. The major cots is that of the geo-membrane and

70

sometimes, the cost of a manually operated pump. Therefore, storage ponds support all income categories including the poor. Conclusions This study sought to quantify the incremental benefits on household incomes and rural livelihoods as a result of the introduction of rainwater harvesting utilizing small, plastic-lined ponds in Minjar Shenkora District of Ethiopia. It was found that the water harvesting intervention offers positive multiplier effects on improving the productivity of smallholder agriculture and household incomes in the study area. One of the most outstanding findings was the high adoption rates of the technology, starting with 308 ponds in 2004 and achieving 7,618 adopters by 2008. Poor adoption of water harvesting technologies is a major limitation to agricultural productivity in dry areas of Eastern and Southern Africa region. Thus, there are lessons to learn from the Minjar Shekora case study. They include:
The technology allowed household-level rainwater harvesting which is broadly suitable to almost any site where runoff water can be collected.
Dealing with the problem of seepage of ponds in an easy and adaptable manner.
Importance of making the technology affordable to the poor, through subsidy to the cost of the geo-membrane used for lining the pond.
This technology makes use of family labour as the main input cost, especially during the construction of the ponds, which is a once-off activity, since the ponds last several years.
It meets the felt need by the community who suffered effects of prolonged dry spells, food insecurity and shortage of water
Extension workers and researchers support in the design, site selection, layout, construction and use of the pond.
There is very low failure rates (if any) among the early adopters
There are comparatively higher returns on investment and use of harvested water for crops with high water productivity and high financial internal rates of return. 71







Its use provides the opportunity to diversify crop productivity, especially vegetables and improve family nutrition. Making use of “idle’ labour especially during the dry season. The enterprises from water harvesting did not really affect the production of other field crops so farmers did not feel the burdens of ‘technology crush’. There was reduction of the workloads of women and children spending a lot of time fetching water from remote areas, thus improving the labour productivity of rural households.

The major limitations of the technology included safety concerns, water losses due to evaporation and heath issues as the water storage was a stagnant pool. Also, water harvesting involves additional labour and this could discourage some farmers. The water productivity of onion seedlings estimated to be 38.7 $ mm-1 was higher than that of field crops ,that is, bulb onions, teff and wheat. However, further research is required to compare the water productivities of crops produced using the harvested water so as to determine the most opportunistic enterprise for the technology, and this could form the basis of a future study. References Amhara Bureau of Water Resources Development (BoWRS) 2005. Volume I.1. Water Harvesting. Regional Irrigation, Land and Water Inventory Book Devereux ,S, 2000.Food Insecurity in Ethiopia. A discussion Paper for DFID, IDS Sussex UNDP,2000. Human Development Book 2006.New York, USA. Gezahegn Ayele, Gemechu Ayana, Kiflu Gedefe, Mekonnen Bekele, Tilahun Hordofa and Kdane Georgis.2006. Water Harvesting practices and impacts on livelihood outcomes in Ethiopia. EDRI,Addis Ababa,Ethipia. Hatibu N., Mutabazi K. , Senkondo E.M, Msangi A.S.K. 2006. Economics of rainwater harvesting for crop enterprises in semi-arid areas of East Africa. AWM 80 , 74–86

72

Menjar Shenkura Woreda office of Agriculture and Rural Development (MSWoARD) 2008. Annual Book. Ministry of Water Resources (MoWR) 2002b: Water sector development progamme 2002-2016, Volume II: Main Book. Ministry of Water Resources, Federal Democratic Republic of Ethiopia, Addis Ababa, October 2002. Pg. 142 Nega H and Kimen PM.2002. Low -cost method of rainwater storage: Result from field trials in Ethiopia and Kenya. RELMA Technical Book Series 28. Rebeka Amha. 2008. Impact Assessment of Rainwater Harvesting Ponds: The Case of Alaba Woreda, Ethiopia. Awlachew, S.B.; M. Loulseged.; Yilma, A.D(eds). In impact of irrigation on poverty and environment in Ethiopia. IWMI Regassa E. Namara, Godswill Makombe, Fitsum Hagos, Seleshi B. Awulachew 2008. Rural poverty and inequality in Ethiopia: does access to small-scale irrigation make a difference? Awlachew, S.B.; M. Loulseged.; Yilma, A.D(eds). In impact of irrigation on poverty and environment in Ethiopia. IWMI. Tadesse N. and Bekelle E. 2006. "The Productivity of Shallow Wells Groundwater in Agriculture and Interacting Systems: A Case Study at Debre Kidane Watershed, Eastern Tigray, Northern Ethiopia”. Agricultural Engineering International: CIGAR Ejournal. Manuscript LW 06 017. Vol. IX. July, 2006. WCD (World Commission on Dams), 2000. Assessment of irrigation options. Technical review IV.2. Cape town, South Africa.

73

4. Comparing the Productivity of Field Crops under Irrigated and Rainfed Systems at Sewur Irrigation Scheme, Ethiopia Akalu, T.F. and Adgo, E.T. Introduction Agriculture is the mainstay of Ethiopia’s economy, contributing almost 50%of the GDP and employing 80% of the population (Befakadu and Berhanu, 2000). As agriculture plays a significant role in the economy, any fluctuations in agriculture can easily shake the GDP. Much of Ethiopia’s agriculture is rainfed, thus food deficits and famines usually occur as result of erratic rainfall and drought (IPCC, 2007). Ethiopia has 122.0 billion m3 surface and 2.6 billion m3 ground water resources, but only 3 to 5% of the nearly 10 million ha of arable land is currently irrigated, while the irrigation potential has been estimated to be about 3.7 million ha (WCD, 2000). Currently, this huge water resource has not been fully utilized. With conservative estimates, the country has the lowest irrigated agriculture compared with other developing countries. The vast majority of the rural poor rely on rain-fed agriculture. Smallholder dominates and farmers have not significantly benefited from improved AWM technologies which could help eliminate food shortage and alleviate poverty. The history of drought and famine in Ethiopia is as old as the history of the country itself. Over 19 periods of widespread and severe food shortages have been recorded in the past 100 years alone. More seriously, the problem has gradually expanded from the North to the rest of the country and its intensity has deepened over time (Christopher et al., 2004). According to Ethiopia’s Plan for Accelerated and Sustainable Development to End Poverty (PASDEP), 38 irrigation projects with an estimated budget of US$2.52 billion will have been implemented by the year 2010. Accordingly, nearly 530,000 ha of land will be irrigated which covers 14% of the irrigable land (Teshome, 2007). Increasing food supplies 74

and economic prosperity for the rural poor in Ethiopia is believed to come from mainly modern AWM technologies, which is among the best entry points to ensure food-security. Access to water will allow farmers to intensify agricultural production and thereby increase crop yields and water productivity. The Amhara Region with a population of about 18 million is the second most populous region in Ethiopia. Some drought- prone areas of the region are food insecure due to a combination of factors such as erratic and unreliable rainfall, degraded natural resource base, high population density and low productivity caused by poor agricultural management practices (BoRD, 2003). An estimated 18-20 % of the population is chronically food insecure. The Region is endowed with a potential irrigable area of 0.6 million ha within the four major river basins (BoWRD, 2005). In addition, the potential for water harvesting by small-scale dams and river diversions is considerable. However, the total area under irrigation to date amounts only to about 76 thousand ha, which is less than 2% of the total cultivated land in the region (BoWRD, 2005). Cognizant of the fact that Amhara region is food insecure and given the huge potential of water resources, the Regional as well as Federal Governments have given high priority for establishing modern irrigation facilities. This is aimed at mitigating the impacts of drought and associated crop failures and to increase agricultural productivity to improve household food insecurity (BoWRD, 2005). However, the impacts of small-scale irrigation on productivity and farmer incomes had remained largely un-quantified. This creates a gap in available information needed for better and targeted strategic planning. The objective of this study, therefore, was to determine the benefits such as profitability, water productivity, returns on labour, returns on investment and contributions to poverty reduction of irrigation of common crop enterprises grown under irrigated and rainfed conditions in Ethiopia.

75

The Study area The study was undertaken in Sewur Irrigation Scheme of Ethiopia, which is situated in Amhara Region, North Shoa Administrative Zone, Kewot district, Yele and Washo Peasant Association. It is located 23 km northwest of Shoa Robit town, about 240 km from Addis Ababa city. The area lies at an elevation of 1200 to 1300 m.a.s.l. in moist-lowland agro climatic zone. The water for the schemes is diverted from the Sewur River which has a diversion weir but there is no night storage tank or dam. The scheme was upgraded in 2003 through financial assistance of the International Fund for Agricultural Development (IFAD). The Sewur command area is almost flat to gently undulating valley plain flanked by the high hill ranges on both the banks. The area has a bimodal rainfall pattern, a long rainy season from July to September, locally called “meher” and short rainy season from February to April, locally called “beleg”. The latter is not dependable and leads often to crop failure. The annual rainfall of the study area is estimated to be 1,075 mm. All the command areas have a soil texture of clay with neutral soil pH within the plowing zone of the soil profile. The primary canal that supplies water to the scheme is only partly lined while all the secondary and tertiary canals including part of the primary canals are unlined. Therefore, seepage losses are a major problem. Flooding and furrow irrigation are the major methods of water application, depending on the types of crops grown. Both full irrigation and supplemental irrigation are practiced. The scheme has a command area of 191 ha with 764 beneficiary households. The average irrigated land holding size is thus approximately 0.25 hectares. The main crops grown in rainfed system are mainly sorghum, teff, munge bean, maize and vegetables. In irrigated systems, the dominant crops are onion, teff and munge bean. Both pure and supplemental irrigation are practiced in the area. Estimated planted areas for each crop enterprise during the wet season include; 10% onion, 55% teff, 25% munge bean and 10% other crops. Similarly, estimated relative area coverage during dry season is 75% onion, 15% teff, 5% munge bean and 5% other crops. Onion is the dominant commercial crop 76

during the dry season while it has the lowest planted area during the wet season. Sampling Procedure and Data Collection Sewur Irrigation Scheme was selected for this study due to its relative functional success and easy access to markets. At the outset, the scheme was stratified into three categories as upper, middle and lower to capture water availability and slope variation impacts across the whole scheme. Thirty farmers were randomly selected from each stratum. In total, 90 farmers were interviewed and their farms assessed. Other background data were collected from published and unpublished information about irrigation scheme, farming system and other socio-economic information. This information was collected from the regional and district office of the Bureau of Agriculture and Rural Development. Primary data were collected from sample farmers using a semi-structured questionnaire. Moreover, group discussions were undertaken with irrigators and leaders of the scheme. Data was analyzed using partial budget analysis, the gross margin analysis and water productivity analysis. Some discounted measures were also considered including an analysis of the Internal Rate of Return (IRR) and the Net Present Value (NPV). Result and Discussions Viability for improving crop productivity Productivity of onion with irrigated was on average about 13.39 t ha1 while yield under rainfed production system was about 9.89 t ha-1. The average yields of teff and munge bean with irrigation were 1.12 and 0.900 t ha-1 respectively. But the average yield of teff and munge bean under rainfed system was 1.19 and 0.96 t ha-1, respectively (Table 4.1). Irrigation at Sewur showed positive impacts on yields of irrigated onion as compared to rainfed. However, there was no significant difference between the yields obtained from irrigated teff and munge beans compared to those from rainfed systems. This implies that crop choice matters as to what extent a given water 77

management system can bring crop yield benefits. However, irrigation contributed to produce two times per year on the same plot of land. Therefore, irrigation has great roles to increase productivity and improving food security of the farmers. This result is in line with the findings of Tilahun (2004) that maize yield increased around 20-70% as the result of using irrigation in Ethiopia. Similarly, Biswas et al. (2003) verified that drip irrigation increased onion yield by 175% in Bangladesh. Moreover, Qiuqiong et al. (2006) found that irrigation had increased yield of crop more than 80 % in China. Table 4.1: Crop yields for rainfed and irrigated systems across the toposequence in Sewur Irrigation scheme Crop Yields (t ha-1) Upper fields

Middle fields

Lower fields

Mean

Onion Rainfed Irrigated

9690 12970

11350 14740

8620 12490

9890 13390

Teff Rainfed Irrigated

1300 1150

1220 1200

1020 920

1190 1120

900 860

950 880

750 910

960 900

Crops/systems

Munge beans Rainfed Irrigated

Diseases and pest attacks on onion during the rainy season were the major causes for the lower yields. The most prevalent were crops pests like cut worms (Agrotis septum) and blister beetle (Packnuda interval) not only for onion but also for other crops. Contrary to onion, productivity of teff and munge bean was higher under rainfed production system than irrigation. This is because of the humid climate of the rainy season which favors high yields of teff and munge beans, as disease incidences are relatively low. Moreover, 78

teff and munge bean are broad casted and not suitable for irrigated agriculture. For all crop types, crop productivity was relatively and consistently better in the middle of the irrigation scheme while at the lower end of the scheme crop yields were lower than both at the upper and middle of the scheme. The sloppy topography and relatively sandy soil nature at the upper, and water shortages and salinity problems at the lower end of the scheme could be mentioned as reasons for such yield variations across the command area. Viability for improving profitability The partial budget analysis on Table 4.2 clearly shows that rainfed systems had some net returns, albeit the values were lower than those obtained for irrigated systems. The profitability of the rainfed and irrigated systems was calculated for each crop separately for the season. In addition, the annual net value per unit command area using the frequently practiced crop rotation system of the area was determined. Usually irrigated onion fields were occupied by rainfed teff and irrigated teff plots replaced by munge bean during the rainy season. Rainfed onion comes after munge bean was harvested. Two harvests per year (double cropping) were common in Sewur irrigation scheme, albeit there were cases where farmers tried to produce three crops in a year. Onion was the most profitable crop in the Sewur irrigation scheme, under both irrigated (US$1716 ha-1) and rainfed (US$1,129 ha-1) systems. Production of onion under irrigation gave about 50% more additional profits in comparison to its rainfed system. According to the study result, the highest annual net value of US$1804 ha-1 was obtained when onion cultivation during the dry season was followed by rainfed teff production. Production of teff and munge bean under irrigated farming was not profitable. On the other hand, onion production with irrigation was more profitable than the rainfed system. As to farmers opinion irrigated production of onion was relatively free from disease and pests that favors higher productivity and profitability during the dry season. 79

Table 4.2: Partial budget analysis with and without irrigation Crops

Treatment

Revenue

Expenses

Net Profit

(US$ ha-1)

(US$ ha-1)

(US$ ha-1)

Onion Irrigation + teff Rainfed Irrigation Teff Irrg.+ munge bean Rainfed Irrigation Munge Irrigation + onion Bean Rainfed Irrigation

3272.3 2020.3 2736.1 891.6 536.2 503.4 2397.3 387.8 395.0

1478.3 891 1020.1 875.7 458.2 560.6 1311.1 315.1 420.1

1804 1129.3 1716.0 15.9 78 -56.8 1086.2 72.7 -43.1

Returns on investment The financial analyses of irrigation are presented in Table 4.3. The parameters considered in the analysis were net present value, financial internal rate of return of investment and rate on investment. An initial investment cost of irrigation of the whole scheme including 10% of community participation costs was US$247,321 gross or US$1,295 ha-1. Production onion and maintenance cost of onion, which was the most profitable crop in the scheme, was US$955 ha-1. The gross income of onion production was US$2,738 ha-1. Table 4.3: Cost-benefit analysis of irrigated onion production Performance Parameters Discount Factor Discounted costs(US$ ha-1) Discounted Benefits(US$ ha-1) NPV (20) IRR ROI

Onion 10% 8,571 21,190 12,618 138% 168%

80

Generally, onion could earn more than ten times the net value earned by teff or munge bean. Farmers responded to this by planting more irrigated onions, taking up about 75% of the scheme area during the dry season. Profitability of teff appeared to be better than munge bean. Profitability of teff and munge bean production under irrigation was below half of that of the rainfed system. Obviously, irrigation is a big investment venture and thus it requires careful selection of the type of crops to be grown. Market access and value of produce are also important factors to be considered. Farmers around Sewur irrigation scheme complained about the low prices they were being offered for their onion produce though it is near an asphalted road and not far from Shewa Robit town. A service cooperative and a union exist in the scheme, but were nonfunctional. The crop production strategy of Sewur irrigation farmers were occupying as much land as possible with teff and to some extent munge bean during the wet season and covering 75% of the command area with onion during the dry season was justifiable and profitable. It is also beneficial from maintaining the soil health conditions point of view. This might be the reason why the two crops are dominating the command area only during the wet season. Though benefits are very limited, munge bean and teff are important break crops in the area. Farmers can achieve better profitability if at least two crops are grown annually. This is a common practice in the irrigation scheme. The benefit of irrigation with a discount rate of 10%, the average NPV of investment in irrigation over a period of 20 years was US$12,619 ha-1. This positive NPV is indicating that the irrigation is financially viable. Generally, investment in irrigation is promising activity (positive net present value). This indicates that the profitability of irrigation for production of onion. Moreover, the IRR was calculated and found to be 138%. This figure is higher than the discounted factor 10% again indicating the financial viability of irrigation. The financial analysis shows that small-scale irrigation is 81

very low investment ventures. The financial viability performance is in line with the observations made by Inocencio et al (2007) in recent irrigation projects in Sub-Saharan Africa. The net benefits of irrigation practices would have been very high if farmers have produced onion both during the rainy season and dry periods with irrigation as farmers in Sewur irrigation scheme have practiced during the early times. However, farmers have noticed that onion yields have been decreasing dramatically and thus crop rotation became indispensable. Though there are variations from plot to plot, most farmers practiced the following crop rotation systems: onion-teff-munge bean. The above figures imply that small-scale irrigation has a great contribution for household income and substantially contributing to poverty reduction. This result is consistent with findings of Fitsum et al. (2008) which revealed that the net gross margin from irrigation is more than twice higher than the gross margin from rainfed agriculture in Ethiopia. Even though onion production through irrigation needed extra labor for making furrows, transplanting of seedlings, irrigating crops every five days interval etc, the family labour inputs for harvesting, transportation and marketing was negligible. Farmers around Sewur irrigation scheme were selling their onion produce directly on the field and traders were harvesting, transporting and marketing the bulb through hired labour. This led to lower labour demand for onion than was expected. Teff production was more labor intensive than munge bean. This is because teff requires a fine seedbed, frequent weeding and higher threshing costs. Returns on labour The study obtained the total family labor required, gross return on labour and marginal return on family labor for the three crops with and without irrigation. These are summarized in Tables 4.4 and 4.5. As depicted from the table, the family labour requirement per hectare for irrigated crop production was higher than that of rainfed system. This is because irrigated farming demands more labor for 82

making furrows, maintaining canals, irrigating crops at certain intervals etc. Irrigated onion and teff production require equal labor inputs (about 80 adult-days/ha) which was higher than that of munge bean (59 adult-days/ha). Obviously, increasing cropping intensity increases labor demand per hectare and year. For example, if onion production in the dry season is followed by rainfed teff cultivation during the wet season, the family labour demand was increased by 58%. Therefore, increasing cropping intensity as a result of irrigation infrastructure could create year round employment opportunities for family labor which could be otherwise idle during the dry season of the year.

Table 4.4: Total family labor inputs and gross return on family labour Crops/systems

Total family labour (adult-day/ha)

Average return on family labour (US$/adult-day)

Onion Rainfed Irrigated Irrigated + teff rainfed

43.3 80.0 138.6

27.9 23.1 14.7

Teff Rainfed Irrigated Irrigated + M. bean rainfed

58.5 80.2 116.1

3.0 1.0 1.8

Munge bean Rainfed Irrigated Irrigated + onion rainfed

35.9 59.1 102.1

3.7 2.7 12.1

83

Table 4.5: Marginal return on family labour

Crops/systems

Marginal return on family labour (US$ per man day) Upper Middle Lower Mea fields fields fields n

Onion Irrigated vs. rainfed (Irrigated + teff rainfed) vs.rainfed

16.71 8.29

16.91 8.65

19.15 9.17

17.59 8.7

Teff Irrigated vs. rainfed (Irrigated + bean rainfed) vs.rainfed

-5.96 0.46

-3.45 1.72

-4.01 -0.19

-4.47 0.66

Munge bean Irrigated vs. rainfed (Irrigated + onion rainfed) vs.rainfed

-3.83 16.35

-4.05 20.76

-1.92 13.93

-3.26 17.01

The contribution of irrigation in creating employment opportunities throughout the year and thereby improving income of farming households is significant and can help to achieve food security and reduce rural poverty. High value crops like onion can play a significant role in improving return on family labour. The return on family labour could be as high as US$25 per adult-day for onion while it was only US$-3.26 per adult-day for munge bean. Thus, irrigation significantly contributes in reducing poverty. It can create year-round employment opportunities for family and hired labor, increase incomes up to $US10 per adult-day which is way above the poverty line. The rate of marginal return on family labor means shifting from rainfed to irrigated agriculture was accompanied by positive return for many of the studied crops. Farmers who were using irrigation were able to realize positive net profits across a number of crop enterprises compared to those farmers without the technologies. It is therefore, more likely for a farmer applying a 84

water management practice to win the poverty war compared to non-adopters of the practice. Scope in improving water productivity The water productivity of crops was estimated based on the crop water requirements determined using climatic data. Cropwat software was used to determine the required water amount of individual crops. More than 15 years of climatic data of Shewa Robit town were used to calculate the crop water requirement. The water productivity of crops at Sewur irrigation scheme is summarized in Table 4.6. Water productivity was calculated based on the water amount determined based on crop water requirement and 40% of scheme efficiency, which is widely accepted irrigation efficiency figure in the Amhara region. Accordingly, onion appeared to bring more output per volume of water consumed (on average US$5.3 mm-1 water) as compared to teff (average US$0.78 mm-1) and munge bean (average US$0.6 mm-1). Irrigated onion was therefore more efficient water user (US$5.26 mm-1) than rainfed production (US$3.95 mm-1). However, irrigated munge bean had slightly higher water productivity than rainfed system, while rainfed teff recorded lower water productivity than under irrigation. Frequently practiced water application intervals in Sewur irrigation scheme are 5, 15, and 13 days for onion, teff and munge bean, respectively. A combination of furrow and flooding irrigation method is used for onion while only flooding is practiced for teff and munge bean which might have influences on water productivity. Direct relationship between duration of the vegetation period and productive uses of water might not exist because all crop types have about 90 days vegetation period but vary in their water use efficiency. Irrigated and rainfed farming did not vary much in terms of water use efficiency for the low yielder crops teff and munge bean while irrigation led to better water use efficiency for onion.

85

Table 4.6: Water productivity of crops for the different production systems Croping Systems

Crop water requirement

Water productivity

mm/season

Kg m-1 US$ m-1

Water productivity at 40% scheme efficiency Kg mm-1 US$ mm-1

Onion Rainfed Irrigated Irrigated + rainfed teff

319 325 713

31 41

6.32 8.41 4.59

19.3 25.7

3.95 5.26 2.87

Teff Rainfed Irrigated Irrigated + munge bean

388 401 806

3.06 2.79

1.38 1.25 1.10

1.92 1.72

0.86 0.78 0.69

Munge bean Rainfed Irrigated Irrigated + rainfed onion

405 395 714

2.37 2.27

0.95 0.95 3.35

1.48 1.42

0.59 0.61 2.09

Social, environmental and economic sustainability Sewur irrigation scheme has changed the lives of hundreds of farm households in the area. Recorded improvements include; increased number of animals (oxen, camels), households improving their housing from grass thatched to corrugated iron roofed houses, construction of commercial buildings in urban areas, even purchasing of vehicles has been documented. All the farmers interviewed expressed that their lives have been drastically improved by the irrigation facilities. The construction of the scheme in 2004 was initially requested by the community. Therefore, irrigation is seen as a way out from the erratic nature of the rainfall and poverty and hence irrigation is socially acceptable. Farmers observed some problems in the scheme such as pest and disease attack of onion crops and to a lesser extent yield reduction. The highest disease 86

incidence on onion may have been due to mono cropping. Proper crop rotation and introduction of other high value new crops like vegetables could help to mitigate the problem without compromising benefits. Malaria incidences were also booked in the area. However, due to frequent spraying with DDT, it was not a serious problem. The main canal was only lined up to 400 meters from the source of the water diversion point. It was seriously endangered by a very big gully near to the command area. Water distribution points had been constructed with cement but they were not functional and needed maintenance. Though the main, secondary and tertiary canals were unlined, water shortages in the downstream part of the scheme were not considered a problem. The clay nature of the soil limits excessive seepage but it is still necessary to line the canal to stop any seepage losses. Accessible roads within the scheme were not available which makes transportation of produces during the wet season very difficult. A majority of farmers did not take the responsibility to handle water conveyance problems, and this has implications for scheme sustainability. There is a traditional water user association in the scheme but its responsibility was mainly in water allocation, canal cleaning at the beginning of the irrigation season and to some extent, managing emerging conflicts among water users. Conclusion and recommendation The study revealed that irrigation increased yields of onion bulb by 26.1% as compared to rainfed system. Irrigation contributed to production of two crop-seasons in a year unlike a single crop under rainfed systems. The study also identified that production of onion under irrigation gave about 50% more additional profits in comparison to rainfed system. The study showed that the highest annual net value of US$1804 ha-1 was obtained if onion cultivation during the dry season was followed by rainfed teff production. Production of teff and munge bean under irrigated farming was not profitable. This indicates that irrigation has great potential for increasing productivity, profitability and improving the livelihoods of farmers. Marketing facilities should be improved in order to 87

maximize the profit of small holder farmers. Moreover, the study showed that US$14.7 per adult-day return on family labor was achieved if onion irrigated and rainfed teff were produced in a given year. The study also identified increasing cropping intensity under irrigated agriculture could create year-round employment opportunities for family labor. The water productivity of irrigated onion was US$5.26 mm-1, while that under rainfed conditions was US$3.95mm-1. But teff and munge bean had low water productivity for both irrigation and rainfed system. This implies that irrigation has a great contribution for improving water productivity resulting in improving productivity and reduction of poverty. The NPV of investment in irrigation over a period of 20 years with a discount rate of 10% was US$2,619 ha-1. Moreover, the IRR was found to be 138%. This indicates that the financial viability of small-scale irrigation. The financial analysis shows that small-scale irrigation is a very low investment venture. This makes investment in small-scale irrigation a potentially viable low investment to poverty reduction tool. Generally, irrigation is able to increase crop productivity, income, reduce poverty, improve efficient use of water, create employment opportunities and bring about gender equity. Therefore, rural farmers, donors, NGO, private investors and the government itself should give due attention in promoting it and further investment as a strategy for poverty reduction and achieving food security. References Befakadu D. and Berhanu N. .2000. The Ethiopian Economic Association Annual Book on the Ethiopian Economic Vol. 1. Biswas, S.K., P.K. Sarker, A.K.M. Mazharul Islam, M.A. Bhuiyan and B.C. Kundu 2003. Effect of irrigation on Onion Production. Pakistan Journal of Biological Science 6(20): 1725-1728 Bureau of Rural Development (BoRD). 2003. Rural Households Socio-economic Baseline Survey of 56 Districts in the Amhara Region, Bahir Dar, Ethiopia. 88

Bureau of Water Resources Development of Amhara (BoWRD). 2005. Volume I1. Water Harvesting. Regional Irrigation, Land and Water Inventory Book Fitsum H., Makombe, G. Namara R. E., and Awulachew, E.B..2008. Importance of Irrigated Agriculture to the Ethiopian Economy: Capturing the direct net benefits of irrigation. Awlachew, S.B.; M. Loulseged.; Yilma, A.D (eds). In impact of irrigation on poverty and environment in Ethiopia. IWMI. Inocencio, A., Kikuchi, M., Tonosaki, M., Maruyama, A., Merrey, D., Sally, H., de Jong, I., 2007. Costs and performance of irrigation projects: A comparison of SSA and other developing regions. Colombo, Sri Lanka: International Water Management Institute. 81 pp. (IWMI Research Book 109). IPCC. 2007. Working Group II, Fourth Assessment Book (www.ipcc.ch) Plan for Accelerated and Sustained Development to End Poverty (PASDEP). 2006. Ministry of Finance and Economic Development (MoFED). Volume I, Addis Ababa Qiuqiong H. Rozellea, S. Lohmarb, B. Huang, J. and Wang, J. 2006. Food policy. Volume 31,Issue 1 Feburary 2006.pp30-52. Teshome A. (2007). Current and Future plans of irrigation and drainage development in Ethiopia. Symposium on Impact of Irrigation on Poverty and Environment in Ethiopia organized by IWMI/Ethioipia, Addis Ababa, Gihon Hotel 27-29 November. Tilahun A. 2004, IFAD interim evaluation mission on small scale irrigation schemes- special country programme-SCPIIiEthiopia, working paper and contribution to the main book. UNDP.2006. Human Development Book. New York, USA. World Commission on Dams (WCD) ( 2000). Assessment of irrigation options. Technical review IV.2. Cape town.

89

5. Assessment of Runoff Harvesting with ‘Majaluba’ System for Improved Productivity of Smallholder Rice in Shinyanga, Tanzania Kajiru, G.J. and Nkuba J.M. Introduction The development of effective strategies to improve agricultural productivity and adaptation to climate change faces greatest challenges under rainfed agriculture, especially in the arid and semiarid regions. The livelihoods of millions of the world’s poorest people in these zones are dependent on agriculture and are thus vulnerable to variable rainfall patterns and water scarcity (UNDP, 2006). A case in point is Tanzania, where 74 percent of the total area under paddy rice comprises rainfed, semi-arid lowlands, utilizing bunded basins locally known as “majaluba” (Kanyeka et al., 1994; ECI, 2003). The majaluba rice production system accounts for about 35 percent of the total rice produced in Tanzania (Meertens et al., 1999; Hatibu et al., 2006). The term semi-arid is used in this paper to refer to conditions where rainfall amounts range from 400 to 800 mm per year, the potential evapotranspiration exceeds rainfall most of the time and the rainfall regime is highly variable in quantity, timing and distribution (Hatibu et al., 2006). In the rainfed rice production systems, the amount and timing of water supply are considered the most severe constraints to rice productivity (Zeigler and Puckridge, 1995). Wade et al. (1999) reported that within a small area, like for example 10 ha, the hydrology of rainfed lowland rice differs greatly depending on the surrounding landscape. Previously, the production of rainfed lowland rice was not recommended in Tanzania in the semi-arid areas of the Lake Zone, due to lack of suitable soil profiles and inadequate soil moisture storage. However, the introduction and adoption of rainwater harvesting systems in these areas has significantly reduced soil moisture constraints (SWMRG, 2001). Based on 2002-03 figures, total paddy production in Tanzania’s Lake zone was 399 t yr-1 which 90

was able to feed about 6 million people (based on FAO rice intake of 40 kg per person and per year). This left about 3.1 million people deficient of rice (URT, 2005). The rice demand in Tanzania and many parts of sub-Sahara Africa has been increasing at an average of more than 6% per annum due to increase in rice consumption contributed by the change of dietary preferences. This has been driven by the rapid urbanization, women’s growing participation in the formal labour force and population growth resulting in a shift away from traditional food such as maize, millet, sorghum and cassava to rice, as it is easier to prepare for consumption. Thus, annual rice consumption in Tanzania has, increased from 7 kg per person in the 1970s to 40 kg per person in 2003 (FAO, 2005). However, the demand for rice in most African countries outstrips the domestic supply, placing a heavy burden on governments to import rice from outside (WARDA, 2007). As a result, there is need to improve production and enhance quality. It is for this reason that rice is ranked high priority among crops research in Tanzania’s Ministry of Agriculture, Food Security and Cooperatives. Concerted efforts have gone towards reducing the adverse effects of the intermittent droughts on paddy rice production in semi-arid areas of Tanzania. A major intervention includes runoff harvesting from roads, home compounds, open areas and grasslands, and its channeling into the majaluba for paddy rice production. This has boosted the production of rice under rainfed conditions in Shinyanga, Mwanza, Mara and parts of Kagera regions. The majaluba system, which is also widespread in the Lake Zone, is credited with production of one-third of all rice grown in Tanzania (ECI, 2003). Support to the majaluba system was boosted in the 1990s by a programme co-funded by IFAD, World Food Programme, Irish Development Aid, the Government of Tanzania and the local people, as part of the Smallholder Development Project for Marginal Areas (SDPMA). Thereafter, the Participatory Irrigation Development Program (PIDP) was introduced in 2000. Coordinated from Dodoma, PIDP covered six regions across 13 districts located in semi-arid areas of Tanzania (IFAD, 2007). The programme supported irrigation 91

development, agricultural development, strengthening farmers’ organizations and local institutions. It was planned to increase crop productivity through expansion and improvement of farmer-initiated and well-managed small-scale irrigation schemes for paddy rice (Mahoo, et al., 1999). However, there have been limited studies to assess the benefits of the majaluba system to smallholder agriculture and household incomes. This study aimed at filling that gap. Characteristics of the study area Location This study was conducted in the rural district of the Shinyanga region of Tanzania. Shinyanga is located in the Eastern part of the Lake Zone, bordering the Mara, Mwanza, Kagera and Tabora regions on the north, west, south-west and south respectively. On the east, it borders with Maswa Game Reserve and Serengeti National Park, and on the southeast borders with the Singida Region. Temperatures in Shinyanga are moderately high while the annual rainfall, which is bimodal, ranges between 600 and 900 mm per annum being highly variable in distribution and amounts (FSR, 1996). Shinyanga Rural district, where this study was conducted lies at the centre of the region. Farming system zones (FSZ) found in the Shinyanga region Five farming system zones (FSZs) bearing local names are found in the Lake zone Tanzania in which Shinyanga is located (Enserink and Kaitaba, 1996). These include Luseni (FSZ 1), Itogolo (FSZ 2), Mbuga (FSZ 3), Ibushi (FSZ 4), Kikungu and Nduha (FSZ 5/6). According to FAO (2005), the local names correspond to soils classifications equivalent to Arenosols (luseni), Planosols (Itogolo), Vertisols (Mbuga) and Ferrarisols (Kikungu and Nduha). However, most parts of the study area comprise flat plains which are dominated by Planosols (FSZ 2). These plains (Itogolo) are characterized by considerable seasonable water logging and in the past, were mainly used for grazing. More recently, these plains have been used for paddy rice cultivation under rainfed conditions with the adoption of majaluba systems. Rice has 92

become the most important food as well as cash crop in this area (Meertens, et al., 2003; Meertens, et al., 1999; and Kajiru, et al., 1998). The major constraints faced by of the Itogolo farming system are (Magoggo, 1991): (i) The long-term sustainability of Itogolo rice production is at risk due to continuous rice cropping and the constant removal of all crop residues. The variable rainfall patterns lead to major risks in the rice production enterprise since, about once in every three years, rainfall is insufficient for a good crop. (ii) Poor timing of planting due to low mechanization of farm operations. Lack of oxen and ox-ploughs leads to frequency food insecurity particularly in resource-poor households. (iii) Low use of agricultural water management technologies. Data sources and acquisition Data collection was organized into two phases; baseline data from literature/reports and field data from farmer surveys. The first phase was devoted to collection of recorded data from selected stakeholders who included departments of District Agricultural and Livestock Development Offices (DALDOs) in Shinyanga Rural, non-governmental organizations (NGOs), farmer groups and Farmer Field Schools. The next step involved field surveys aimed at verification, gap filling and collection of field data. The fieldwork used rapid rural appraisal (RRA) techniques involving key informants and face-to-face interviews using structured questionnaire. The sample size was 60 respondents randomly selected to represent approximately 5 percent of the respondents in the study area. Among the 60 respondents, 30 were farmers who combined their majaluba with runoff harvesting (with technology) and the other 30 relied only on natural rainfall, without runoff harvesting with their majaluba systems (without technology). The respondents were also selected to represent the upper, middle and lower lying areas of the study area. Selection of villages and respondents also considered geographical and demographic factors such as agro-ecological zones, access to markets, gender of household heads and social status of households. Household 93

interviews were then conducted using semi-structured questionnaires. The data collected included types of crops cultivated, types and costs of farm inputs such as organic and inorganic fertilizers, and insecticides/herbicides, labour requirements (adult-days) and costs of farm operations, prices of farm produces and products and relevant climatic and soil factors for agricultural production. Data analysis The data collected were organized and analyzed with spreadsheets and SPSS statistical computer software. Descriptive and qualitative approaches in the form of simple means, ratios and percentages were used for analysis. The performance of paddy under runoff harvesting and without runoff harvesting was assessed in term of economic yield in t ha-1, net value of produce per hectare, return on labour as gross margin per adult-days. An adult-day refers to number of adult-equivalents working for 8 hours per day. The analyses focused on benefits of majaluba systems “with” and “without” runoff harvesting. Results and discussions The results of this study are discussed under the following (i) impacts of runoff harvesting on viability of majaluba for improving productivity and profitability (ii) contribution to reduction of rural poverty and (iii) social environmental and economic sustainability. Viability for improving productivity and profitability The performance of majaluba with and without runoff harvesting showed that paddy rice yields were significantly different between positions along the toposequence. The average rice yields level was higher, being 3.43 t ha1 with runoff harvesting as compared to 2.17 t ha-1 without, an increment of about 58% which was statistically significant at P