overcoming limited information through participatory...

Author's personal copy

Overcoming limited information through participatorywatershed management: Case study in Amhara, Ethiopia

Benjamin M. Liu a, Yitayew Abebe b,c, Oloro V. McHugh a, Amy S. Collick a,Brhane Gebrekidan c, Tammo S. Steenhuis a,c,*

a Department of Biological and Environmental Engineering, Cornell University, Ithaca, NY, United Statesb Swedish International Development Cooperation Agency, Bahir Dar, Ethiopia

c Amhara Micro-enterprise development, Agricultural Research, Extension, and Watershed Management (AMAREW) Project, Bahir Dar, Ethiopia

Received 24 May 2006Available online 13 May 2007

Abstract

This study highlights two highly degraded watersheds in the semi-arid Amhara region of Ethiopia where integrated water resourcemanagement activities were carried out to decrease dependence on food aid through improved management of ‘green’ water. Whiletop-down approaches require precise and centrally available knowledge to deal with the uncertainty in engineering design of watershedmanagement projects, bottom-up approaches can succeed without such information by making extensive use of stakeholder knowledge.This approach works best in conjunction with the development of leadership confidence within local communities. These communitiestypically face a number of problems, most notably poverty, that prevent them from fully investing in the protection of their naturalresources, so an integrated management system is needed to suitably address the interrelated problems.

Many different implementing agencies were brought together in the two study watersheds to address water scarcity, crop production,and soil erosion, but the cornerstone was enabling local potential through the creation and strengthening of community watershed man-agement organizations. Leadership training and the reinforcement of stakeholder feedback as a fundamental activity led to increasedownership and willingness to take on new responsibilities. A series of small short term successes ranging from micro-enterprise cooper-atives to gully rehabilitation have resulted in the pilot communities becoming confident of their own capabilities and proud to share theirsuccesses and knowledge with other communities struggling with natural resource degradation.� 2007 Elsevier Ltd. All rights reserved.

Keywords: Integrated community watershed management; Food aid; Ethiopia; Water scarcity; Erosion

1. Introduction

Ethiopia’s population grows by 2 million per year. In acountry dependent on subsistence farming, this increasingpressure has resulted in average land holdings falling from0.5 ha/person in 1960 to 0.11 ha/person in 1999 (SpecialReport: FAO/WFP Crop and Food Supply AssessmentMission to Ethiopia, 2006). The Ethiopian government

has officially recognized that 5–6 million of its people havelost the capacity to procure enough food to meet annualneeds under normal conditions, with another 10 millionsusceptible to any shock, so projects that protect andimprove the natural resource base are necessary for long-term survival. Since both arable land and surface watersources are severely limited, improving the soil’s capacityto store rainwater was investigated. Over 90% of Ethiopia’sfood production comes from rain-fed agriculture soimprovements in the management of this ‘green’ waterhave enormous significance.

Ongoing land degradation in Ethiopia requires urgentaction at different levels of society (Nyssen et al., 2004).

1474-7065/$ - see front matter � 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.pce.2007.04.017

* Corresponding author. Address: Department of Biological and Envi-ronmental Engineering, Cornell University, Ithaca, NY, United States.Tel.: +1 607 255 2489.

E-mail address: [email protected] (T.S. Steenhuis).

www.elsevier.com/locate/pce

Available online at www.sciencedirect.com

Physics and Chemistry of the Earth 33 (2008) 13–21


Nyssen et al. (2004) indicated that the stagnation of agricul-tural technology and lack of agricultural intensification inthe Ethiopian highlands is the origin of present land andresource degradation. This degradation in turn becomesthe underlying root of poverty. Thus, the challenge ofbreaking the poverty-environment trap and initiating sus-tainable intensification requires policy incentives and tech-nologies that confer short-term benefits to the poor whileconserving the resource base. Only by improving the natu-ral resource base can one increase food production and les-sen the need for external food supplies (Shiferaw andHolden, 1998). Therefore, to explore sustainable methodstowards increased food production and its connected eco-nomic development, the goal of this study was to revitalizewatersheds by keeping more rainwater (‘green’ water) onthe land and increasing fertility. This revitalization is multi-dimensional and complex, requiring interdisciplinary effortto carefully design development activities. Since ‘green’water development encompasses so many different activitiesat various scales, with each location having unique needs,an integrated approach is especially appropriate.

1.1. Integrated watershed approaches

One integrated approach for watershed management isthrough the use of computer models – various attemptshave been made in Ethiopia to apply such methods. TheAgricultural Non-Point Source Model (AGNPS) was testedon the highlands Augucho catchment by Haregeweyn andYohannes (2003) but could not reproduce runoff patterns.The Precipitation-Runoff Modeling System (PRMS) wassimilarly tested by Legesse et al. (2003) for South CentralEthiopia, and needed extensive calibration to predict themonthly runoff. Ayenew and Gebreegziabher (2006) fitteda spreadsheet type water balance to predict water levels inLake Awassa of the Rift Valley, but found that the modeldid not perform well in more recent years, possibly due tochanging land use and neotectonism. Finally, Hengsdijket al. (2005) applied a suite of crop growth, nutrient bal-ance, and water erosion models to conclude that commonconservation practices such as bunds, crop mulching, andreforestation may actually result in lower overall crop pro-ductivity in the highlands of northern Ethiopia.

In response to Hengsdijk’s conclusions, Nyssen et al.(2006) compared Hengsdijk’s predictions with field obser-vations from the same region and found that the modelsover predicted crop yields while under predicting soillosses. Although such models can be applied for policyanalysis, they typically need extensive calibration and can-not simulate the intricacies that farmers have to deal withon a day by day basis (Nyssen et al., 2006). When datasetsare incomplete or of poor quality, other integratedapproaches will likely be more effective management tools.

The modeling techniques described above were typicallydeveloped for conditions in the United States or Europewhere availability of datasets for both input and calibra-tion was not a great limitation. In contrast, the required

datasets in developing countries are often available onlyat very limited locations, and data collection standardscan be inconsistent. Integrated water managementapproaches that either require less information or makeuse of indigenous knowledge within the watershed aretherefore needed. One such approach is the SmallholderSystem Innovations (SSI) programme in Tanzania andSouth Africa, which concentrates on ways to increase foodproduction, improve rural livelihoods, and safeguard criti-cal ecological functions through participatory developmentand interdisciplinary research (Rockstrom et al., 2004).Similarly, using the joint Vertisols project in Ethiopia asa case study, Jabbar et al. (2001) described the need to tran-sition away from the traditional single discipline manner ofresearch if complex interrelationships between environ-ment and human are to be addressed.

Another reality of development work in Ethiopia is adependency on food aid. It has become a contentious issuefor many developing countries, with some officials evenadvocating the cessation of foreign aid. After more than25 years of Food for Work programs, productivity of majorcrops (barley and wheat) remained stagnant, and local live-lihoods in Ethiopia were not improved (Herweg, 1993; Shi-feraw and Holden, 1998; Tekle, 1999). Conservationpractices proposed under the Food for Work program werenot seen by households as valuable means towards increas-ing food production. Farmers took part in the Food forWork programs because either they needed the grain orwere forced to participate (Tekle, 1999). Tekle furthershowed in a study of South Wollo, Ethiopia that the mainproblem with Food for Work programs was the lack ofattention given to attitudes of local people towards conser-vation programs and what their priorities were. In mostprograms local people were not consulted at all, making itimpossible for these communities to accept any kind ofresponsibility. Public opposition to projects most oftenarises from either a lack of accurate knowledge or inade-quate involvement in the decision-making process (Plan-ning and Decision Making, 1999), so the obvious solutionis to have farmers become involved in both the informationgathering and decision processes. Therefore, our integratedwatershed management approach was to give local farmingcommunities control of both the planning and distributionof Food for Work aid, thus meeting short-term food needswhile sustainably improving the resource base.

1.2. Pilot watersheds

With this in mind, two pilot watersheds from food inse-cure areas of the Amhara region, Yeku and Lenche Dima,were selected in conjunction with the Amhara Micro-enter-prise development, Agricultural Research, Extension, andWatershed Management Project (AMAREW) to test inte-grated watershed management techniques in Ethiopia.Locations of these sites are shown in Fig. 1. Both water-sheds are almost entirely comprised of subsistence farming.Rain-fed crops, mostly cereals, are cultivated in the flatter

14 B.M. Liu et al. / Physics and Chemistry of the Earth 33 (2008) 13–21


valley bottoms while the hillsides are used for free livestockgrazing. Yeku watershed is more at risk of chronic foodshortage than Lenche Dima and was recently involved inthe government’s resettlement scheme. As shown in Table1, Yeku is higher in altitude, receives less rain, and is eco-nomically poorer as illustrated by smaller average land-holdings. While the only developments within eachwatershed are small traditional villages and paths, LencheDima is near the larger town of Woldiya so has easieraccess to outside resources and markets.

Feasibility studies to select the pilot watersheds werecarried out in the late 1990s, with the project officially start-ing in September 2002. The first community meetings to setwatershed priorities were held in January and February of2003. Leadership workshops were held then CommunityWatershed Management Organizations (CWMOs) createdin June. With these formal community structures in placeconservation projects (both biological and physical) beganand continue through the present. Late in each calendaryear, project progress is reviewed by the communities andthe following year’s activities planned. Our involvementwith AMAREW stopped at the end of 2005.

1.3. Watershed management design with limited data

When trying a grassroots approach to watershed man-agement, local farmers’ knowledge forms the foundation

that all other efforts build upon. This approach is especiallyapplicable to places like the pilot watersheds where knowninformation is limited and disparate. Without accuratedata, technical experts have no advantage over politiciansor stakeholders (Steenhuis and Pfeffer, 2000). Yet withoutthese experts, new ideas and management practices willnever be introduced. Farmers’ knowledge must be supple-mented with scientific data collection to properly identifythe needs of each watershed.

1.4. Gathering information

Obtaining data and making it accessible to both plan-ners and communities is a crucial first step in appropriatewatershed development. In the initial watershed selectionfor the 1999 USAID request for applications, comprehen-sive feasibility studies were undertaken at the Yeku andLenche Dima watersheds. A traveling technical teamformed a planning group with 30 representative farmersat each watershed then carried out exercises on describingthe watersheds, identifying problems, formulating solu-tions, and developing community action plans for eachsubsystem of the watershed (Gizaw et al., 1999a,b). Withassistance the farmers also mapped available resources,land use, and soils of their watersheds. These in turn wereused by the technical team to create scientific maps of thesame watershed characteristics. Results from the planninggroup were presented to the communities, local govern-ments, and cooperating agencies.

Unfortunately, after such a promising start it took fouryears to begin the actual work. During this break the com-munities were not kept informed and lost confidence in theproject. Other lapses in communication during this time ledto important parts of the feasibility study, such as soilmaps, being lost. Cases like this underlined the recurringproblems associated with participating agencies and groupsnot sharing the same vision. In a system accustomed todirectives coming from above, grassroots initiatives requirea daunting amount of inter-agency communication to havea chance at succeeding.

Due to difficult site access, satellite imagery with GPSregistration was used to create detailed land use and eleva-tion maps (Heatwole, 2003). These maps are very useful forpresenting information to technical colleagues, but thetechnology has not been successfully made accessible tothe stakeholders. Further work needs to be done to let allparticipating parties benefit from this expensive knowledge.While remote sensing may one day provide the basic infor-mation needed for watershed management planning in

Fig. 1. Ethiopia site map.

Table 1Pilot watershed characteristics

Altitude(masl)

Yearly rainfall(mm)

Rainpattern

Number ofhouseholds

Average landholdingsize (ha)

Total area(ha)

Cultivated Area(ha)

Yeku 2050–2360 400–800 Unimodal 285 0.75 582 230Lenche

Dima1520–1820 700–1000 Bimodal 865 1.56 1546 979

B.M. Liu et al. / Physics and Chemistry of the Earth 33 (2008) 13–21 15


remote locations, the lack of detailed data available forAfrica and extreme rainfall patterns mean that onsite datacollection is the only real option for now.

The climate data used in the feasibility studies was fromthe nearest government weather stations, but Ethiopia’shighland climate varies greatly both in space and time. Col-laborating with the AMAREW project, the governmentresearch institution ARARI and graduate students sta-tioned at each pilot watershed researched the effectivenessof interventions and collected basic scientific data such asrainfall, stream flow, temperature, and evaporation. Enlist-ing local children allowed for the gathering of a moreaccurate spatial distribution of data and helped to ensurelong-term interest in data collection. Continual monitoringwill be essential to successfully adapting managementstrategies to rapidly changing conditions.

2. Watershed management through leadership development

Even if all the outputs from initial efforts had been flaw-lessly preserved, there still was not sufficient informationfor a top-down watershed development approach thatrelied on precise hard data for planning. The only practicalway to continue was to make full use of community knowl-edge. To do this the communities needed to be cooperative,informed, and organized such that data and feedback couldbe relayed with ease. Since time was limited a two-prongedapproach was followed: community problems were identi-fied and prioritized in workshops and, at the same time,the community was helped to develop leadership confi-dence.

2.1. Workshops

Once the project started, community problem identifica-tion and prioritization workshops were again held in eachwatershed. Yeku determined that soil erosion and watershortage were their most important priorities, while LencheDima was most concerned with water shortage and cropproductivity. Given less priority were fodder shortage,pests, livestock health, deforestation, erratic rainfall pat-terns, and lack of access roads. The communities then pro-posed solutions for each problem and, based on theirfarming timetables, determined how much time could bespent each month on project tasks. These estimates werecombined with field visits by technical staff to formulateannual work plans and budget requirements.

2.2. Organizing communities

Farming communities have a high interest in improvingand protecting their natural resource base but do notalways have the capacity to effectively address these issues.Developing leadership skills is essential to these communi-ties wishing to assume more responsibilities. As an initialstep various leaders from each community were enrolledin the Community Organizing and Leadership Training

for Action (COLTA) program. Upon returning, these lead-ers were able to hold elections and form the communitywatershed management organizations (CWMOs). Furthertrainings were conducted in conflict management, consen-sus building and community organization. Without suchincremental steps, the watershed management organiza-tions would not have been able to organize themselves,coordinate food aid programs, or plan watershed interven-tions.

Any person receiving food aid automatically became amember of the CWMO in their respective watershed, buteach of the four village clusters selected four males andfour females to form the organization’s leadership. Genderequity was a major concern in strengthening the commu-nity participation process. Social and cultural norms com-bined with a lack of land ownership discourage womenfrom participating in leadership positions. After some ini-tial opposition it has become acknowledged that the 50%rule resulted in a stronger organization. This planninggroup split into four focus committees: natural resources,crop production, livestock and social development. Greatlyenhancing extension possibilities, the influential watershedmanagement organizations also serves as a conduit for fur-ther information dissemination such as soil and water con-servation, post-harvest storage, and animal health.

2.3. Promoting confidence

The project also targeted households with small or nolandholdings for income generation through more informalself-help groups. An example of the success achievedthrough these programs is the Yeku women’s energygroup. Increasing scarcity of fuel sources and the necessityof alternative sources of income brought these womentogether to produce and market improved wood stoves,which reduce fuel use and curb smoke pollution. They cur-rently produce the stoves in the watershed and sell themthroughout the Sekota area with the continued supportof the Sekota Agricultural and Rural Development Office.Involving women in new income generating activities andleadership opportunities are important emerging compo-nents of integrated watershed management projects.

It must be noted that while stakeholder feedback is cen-tral to the participatory method, new processes still requirethe support of outside institutions to seem credible andattainable. Comparing the early progress of the two pilotwatersheds made this apparent. The Bureau of Agriculture(BoARD) was the primary partner agency responsible forimplementation of the AMAREW project, acting as theinterchange point between the communities and outsideagencies. BoARD is understaffed and stretched thin, yetat Yeku the supportive representative played a critical rolein helping the watershed produce real gains from projectworks within months. At Lenche Dima, at least three dif-ferent Development Agents were assigned to the area dur-ing the first two years of the project, and sometimes therewere months when no one was responsible. Despite the



community’s interest, very little was accomplished. Theyneeded outside support to help jumpstart the process andbuild confidence in sustainable development practices.

3. Specific achievements

There are encouraging attitudinal changes within thecommunity. In the last three years, through the facilitatingrole of the CWMOs and the incremental successes of smallindividual achievements, the communities now contribute30–40% free labor for watershed management tasks. Main-tenance of soil and water conservation structures, treeplanting, and access road construction works are carriedout through free community labor. This was not the caseprior to the AMAREW project, and is a welcome changefrom the much lamented dependency trap that food aid isoften associated with. Increased community responsibilitywas accomplished by promoting from the start that com-munities should own the watershed and its activities,whereas outside projects contribute the financial, technicaland legal support necessary to create an enabling environ-ment and encourage local initiative.

The vital monitoring of the pilot watershed developmentis also primarily the responsibility of the CWMO. Eachyear’s activities are evaluated by the communities prior tothe next work plan preparation. Farmers identify technol-ogies that have performed well and should be scaled upalong with those that should be dropped because of poorperformance. Increased confidence and responsibilitywithin the communities has the added benefit of strength-ening partnerships between government institutions andthe CWMOs.

To illustrate what can be achieved through communityparticipation, specific examples are given on water short-ages, land degradation, improved agricultural productionand gully rehabilitation.

3.1. Potable water shortages

Communities in Yeku identified water shortage as theirprimary constraint and their first priority for an integratedwatershed development effort in their watershed. House-holds in the watershed are forced to access on average morethan three different water sources during the course of theyear to fulfill their domestic and livestock water needs.Water sources include gully sand bed holes, communityearthen ponds, rivers, springs and hand-dug shallow wells.Each source has its own constraints: poor water qualityand health problems are associated with the ponds andgully beds, springs are especially susceptible to fluctuationsin rainfall, and wells can be prohibitively expensive to con-struct. Most of these sources dry up during the dry periods,so long distances have to be traversed in search of reliablewater.

It is women and children of school age, particularlygirls, who collect drinking water while boys are responsiblefor watering livestock by collecting leech-free water from

distant rivers. Water scarcity here has significantlyimpacted enrollment of school-aged children. One farmerexplained the situation as follows: ‘‘Since the opening ofelementary school at Woleh (5 km away), the incidenceof leeches on our cattle has increased as we started sendingour children to school and our livestock are left unat-tended. I know sending children to school is a good thingbut I should give priority to my livestock as they are theones on which my family’s livelihood depends. That iswhy I made my children drop out of school and tend thelivestock to protect them from getting leeches’’.

In line with their prioritization exercises, the communityat Yeku developed potential water sources includingsprings and shallow hand-dug wells with support fromAMAREW. Labor and locally available constructionmaterials such as sand, stone, gravel and water were pro-vided by the community while the project provided materi-als that were not locally available, including cement,reinforcing iron rods, pipes, fittings, and skilled labor.The major new source is the Bambaw spring, comprisinga spring box with sand filtering system, separate waterdelivery point for humans and livestock along with a pro-tected washing stand. This spring provides clean potablewater for over 200 households and meets the water needsof more than 600 livestock per day. Water use is managedand operated by a water committee established by the com-munity, with moderate usage fees collected to cover routinemaintenance and security costs. The newly developedspring has reduced workload for women and childrenand improved human and livestock health. The farmerquoted above commented the following: ‘‘Now that I haveaccess to filtered water from a spring which is leech free,time has come for me to send my children back to school’’.

3.2. Land degradation

In addition to water shortage, land degradation andyield reduction were some of the primary concerns of thepilot watersheds. Extensive physical and biological conser-vation works were carried out in Yeku and Lenche Dimathrough Food for Work and free community labor mobili-zation. Physical interventions included extensive hillsideterracing, check dams using stone/gabion/sand bags, stoneand soil bunds, eyebrow and micro-basins, trenches, sedi-ment storage dams, and rock-fill dams. Biological interven-tions mainly focused on area closure, gully rehabilitation,hillside planting, and individual homestead plantations.Stabilization of farmland soil by direct sowing on bundsof forage species such as Sesbania, Leucenea, and pigeonpea has shown very encouraging results. Increased forageproduction by beneficiaries allowed households to meetlivestock feed requirements and enabled some to earn addi-tional money from the sale of forage seeds to the localBureau of Agriculture (BoARD) and NGO offices.

The closed area management system in Yeku is one ofthe finest examples of the cooperative process. Closing offgrazing lands when livestock herds were already underfed



seemed like a preposterous solution to the communities, sothe project arranged for farmers, other community leaders,and local development agents to visit another watershedwhere closed area management was practiced. After dis-cussing with the more experienced community the difficul-ties and benefits associated with closed areas, Yeku decidedto try closing 25 ha of their land. Management of these hill-side areas was given to self-help groups comprised of per-sons with limited or no land holdings, often youths orwomen-headed households. The hillsides showed veryrapid regeneration of natural vegetation, resulting in cutand carry profits for the user groups, decreased soil loss,and increased water infiltration and groundwater recharge.This led to seasonal stream flows at Yeku being extended.Due to the success of the trial, the Yeku community hasvoluntarily closed other grazing areas and become a show-piece watershed that holds Farmers’ Days to educate oth-ers about closed area management.

3.3. Agricultural productivity

To increase crop yields, experimental farms were estab-lished within farmers’ plots to test improved varieties oftraditional crops and new crops such as cotton, groundnut,and garlic. Some crops, such as beets, grew well but it wasquickly discovered that they did not have any marketpotential in the communities. This kind of instant feedbackis not possible at farm trials in specialized outside researchsites. Improved plowing techniques aimed at increasedwater infiltration and storage were also tested in a similarfashion. Subsistence farmers not sure of meeting theirrequired needs are unwilling to try new methods that couldleave them with less food, so aid incentives were used todecrease the risk of technology transfers such as seed stor-age improvement, fruit tree propagation, private nurseries,beekeeping, and poultry production. Working with localknowledge to select the best locations, small-scale waterharvesting structures have also been constructed to enableorganic vegetable gardens. Progressive farmers such as AtoWossen in Yeku now produce a variety of vegetables andfruits such as garlic, pepper, onion, pumpkin, coffee,papaya, avocado and mango using this water. Last year,he was able to get 400 birr (US$ 47) from the sale of hisgarlic crop alone, so other farmers are becoming greatlyinterested in such improvements.

3.4. Gully rehabilitation

Deforestation exacerbates excess run-off and has causedsevere gully erosion in productive farmlands at the foot ofhillsides of both pilot watersheds. Increasing amounts ofproductive farmland are lost through gully erosion everyyear. While gullies are not the only places in the landscapewhere erosion takes place, they were the logical candidateto address first since flow is concentrated in relatively smallareas, meaning less labor to construct and maintain conser-vation structures. Another important point is that before

the project gullies had no value other than transportationpathways – constructing structures in other areas wouldalmost certainly involve a reduction in the amount of cropor grazing lands.

Therefore, in 2004, a gully in the Kolo Kobo sub-catch-ment of Lenche Dima was selected for a demonstration ofrehabilitation measures. Community level discussions wereheld on how to rehabilitate gully areas within farm plots,assessing traditional practices and new ideas. It wasdecided to use simple sand bag check dams since the stonesfor gabion construction were scarce. Sand bag check damsare not permanent structures but help trap silt washed fromsurrounding hillsides. Once adequate silt is accumulated,usually after the first three or four rains, forage specieswere directly sown on the silt layer. Fig. 2 shows the signif-icant stabilization that occurred due to these physical andbiological interventions.

The CWMO divided the entire gully length into smallsections, usually the area between two sand bag checkdams, and allocated it to a farmer who had landholdingsadjacent to the gully. The user has the right to utilize thevegetation, employing the cut and carry system, in returnfor maintaining the physical structures before and afterthe rains and planting trees and other plants as appropri-ate. During the initial stage of construction, frequent loot-ing of bags, eucalyptus poles, and stone was common. Itwas not possible to assign guards along the 600 m of gullylength, and knowing that government decrees would havelittle effect, the CWMO turned to Shelegas, religious lead-ers in this Muslim area who use their authority to resolvedisputes in the community. The project invited the Shelegas

to visit the development efforts under way, then a tradi-tional chat chewing ceremony was arranged and the issueof damages to the physical conservation works wasbrought to the attention of the Shelegas. The Shelegas

uttered a warning statement locally recognized as Ergeman

stating that ‘‘whoever inflicts any damage on the commu-nity’s or project’s structures or materials and supplies,may invite Allah’s severe punishment on him’’. Since thewarning, there has not been any damage reported or mate-rials stolen.

Gully rehabilitation has been associated with moderat-ing the extremely violent flow events that are associatedwith intense rainfall patterns, increasing the duration ofstream flow and lowering peak rates. We investigated if thiswas true for the Lenche Dima watershed. Due to the rela-tively short duration of the study and remoteness of thelocation, examples are only illustrative. The stream flowpatterns in the Kolo Kobo sub-catchment were comparedwith a control sub-catchment lacking gully stabilizationimprovements. Although these two micro-catchmentswithin the Lenche Dima watershed were not identical, theresults still showed the benefits of interventions. The con-trol sub-catchment, Hartibo, was almost twice as largebut had 120 m less of an elevation change as the KoloKobo sub-catchment, so standard logic would predict thatit have smaller peak discharges and longer retention times.



Two runoff events are plotted in Fig. 3, the first runoffevent of the season (a and b) and an event during the mid-dle of the rainy season (c and d). Both storms show theKolo Kobo gully flow lasting 2–3 times longer than Har-tibo, and the first storm (a) shows Hartibo having a muchhigher peak flow rate. The increased retention time at KoloKobo was due to the sandbag check dams holding backwater and silt. The many peaks during the 8/14/04 event(b), most notably the last one seen at 5 h, were likely dueto one or more of the check dams being overtopped or giv-ing way, but even with such failures the constructed worksdid a remarkable job of slowing down erosion. The reten-tion of sediment and moisture in the gully bed can also

clearly be seen in Fig. 2(a and c). Most years see less than10 runoff producing storms so it is worth the effort to repairthe check dams or construct new ones as the bed fills withsoil after each large event.

4. Discussion and conclusions

Bottom-up participatory watershed development pro-cess can only be achieved with strong communitywatershed management organizations (CWMOs), whichfacilitate the introduction of appropriate managementstrategies and, even more importantly, continue this effortafter outside help leaves. Therefore, our single most

Fig. 2. Kolo Kobo gully rehabilitation photos. Sandbag check dams constructed in main gullies of Kolo Kobo catchment: (a and c) shortly afterinstallation of check dams and (b and d) three months later showing good vegetation cover and gully stabilization.



important result is that the community organizations havebecome strong enough to manage watershed developmentthemselves. Lack of information at the beginning wastransformed into a strength by allowing the project toinclude farmers at a very early point. Involvement in eachphase of planning has resulted in increased trust and con-fidence. This has in turn allowed introduced technologiessuch as area closures and water harvesting basins to bemore readily accepted, and such changes have becomedesirable with or without outside funding. Empowermentof local organizations allows them to take control of theirown watershed problems and work towards meeting theirown livelihood needs.

Specifically we have learned that (1) communities havehigh interest in development initiatives, including sustain-ing natural resources; (2) only when people make real con-tributions of their own resources will they strive to ensurethe implementation of the planned activities; (3) emphasismust be given to effective organization of communitiesrather than only focusing on technology development; (4)including non-technical leaders in the information loopprovides great benefits at the community level.

Introduced ideas, such as the fair representation ofwomen and poor in the management process, women’smicro-enterprise groups, and closing off areas to livestock,faced strong cultural opposition at the beginning but haveproven very successful thanks to trial tests and continuedcollaboration efforts by farmers, non-technical leaders,

and project personnel. We also found that communitiesmost at risk were also the most open to new ideas. LencheDima was close to a relatively well-off town that offeredalternative opportunities for income, so the establishmentof a committed CWMO in Lenche Dima was initially diffi-cult. Only after a project manager stayed for a two-monthperiod in the watershed demonstrating what could beachieved by cooperation did the CWMO take off. In con-trast, farming was the sole option for people in the moreremote Yeku watershed, so it was easy for them to seethe importance of improving their natural resource base.Motivation was thus much stronger in Yeku, and theirCWMO immediately started making progress. A goodmeasure of the strength of this community organizationis that three years after its inception approximately 40%of the conservation works labor was being implementedthrough volunteer work instead of foreign aid. We havenot yet been back to Lenche Dima.

Without ignoring the encouraging improvementsdescribed above, many challenges still exist for these pilotwatersheds. In a society that is used to top-down adminis-tration, having a strong community base is not by itselfenough. Since the government currently owns all land,communities and individuals will not be able to take fullresponsibility for their natural resources until furtherreform of land rights is pursued. In addition, stakeholderscientists and engineers are still required to draw upon theirwider ranges of experience to ensure the sound design, solid

HB8/14/04

0

25

50

75

100

125

0 2 4 6 8 10 12

0

5

10

15

20

KK8/14/04

0

5

10

15

20

25

Time (h)

Time (h) Time (h)

Time (h)

0

5

10

15

20

Rai

nfal

l (m

m)

Rai

nfal

l (m

m)

Rai

nfal

l (m

m)

Rainfall = 30 mm

Rainfall =15 mm

Rainfall = 30 mmI30 = 27 mm h-1 I30 = 23 mm h-1

0

25

50

75

100

125

10 12 0 2 4 6 8 10 1210 12

0

5

10

15

20 0

5

10

15

20

25

Dis

char

ge (

mm

d-1

)D

i

Rai

nfal

l (m

m)

Dis

char

ge (

mm

d-1

)

Di

Dis

char

ge (

mm

d-1

)D

isch

arge

(m

m d

-1)

0

5

10

15

20

I30 = 27 mm h-1 I30 = 23 mm h-1

Hartibo discharge 8/14/04 Kolo Kobo discharge 8/14/04

0

10

20

30

40

50

0 2 4 6 0 2 4 6

0

5

10

15

20

I30 = 22 mm h-1

KK9/13/04

0

10

20

30

40

50 0

5

10

15

20

Rainfall = 21 mmI30 = 22 mm h-1

HB9/13/04

0

10

20

30

40

50 0

5

10

15

20

I30 = 22 mm -1

0

10

20

30

40

50 0

5

10

15

20

I30 = 22 mm h-1

Hartibo discharge 9/13/04 Kolo Kobo discharge 9/13/04

a b

dc

Fig. 3. Stream discharge and 15-min rainfall (bar graph) for two sub-catchments in Lenche Dima: ‘‘control’’ Hartibo (HB) and ‘‘conserved’’ Kolo Kobo(KK). Measured values shown by a solid line with open circles and recession fits shown by a broken line.



construction, proper use and implementation, and appro-priate maintenance of new technologies. The projectinvolves numerous government agencies and foreign uni-versities, but all the partners do not have the same vision.Gains will quickly falter unless the project continues tobuild organizational relationships and recognize changingneeds.

Luckily, even localized successes, such as the progressseen in the pilot watersheds, work like a positive feedbackloop, leaving those involved more capable of tackling lar-ger problems. Examples discussed from both watershedshave illustrated the effectiveness and influence of small-scale projects. Projects at the local level are less likely tooverwhelm the resources of an organization or excludethem, and both the local communities and stakeholderorganizations that work together in these projects improvetheir decision-making powers and sense of ownership. Thepilot watersheds of Yeku and Lenche Dima have becomeshowpieces for watershed conservation, motivating andteaching other communities that similar progress is withintheir reach.

Acknowledgements

The key ingredients for success in this project were themany hard working people in the watershed and the expe-rience and knowledge of Yitayew Abebe and Tesfaye Habt-amu in promoting participatory watershed managementunder supervision of Kent Reid. In addition the help ofthe personnel of BoARD, ARARI, and other partner agen-cies were instrumental in facilitating the project. USAIDprovided the funding for the project to Virginia Tech (Con-tract No. 663-C-00-02-00340-00) which subcontracted thewatershed component to Cornell University.

References

Ayenew, T., Gebreegziabher, Y., 2006. Application of a spreadsheethydrological model for computing the long-term water balance ofLake Awassa, Ethiopia. Hydrolog. Sci. J.-J. Sci. Hydrolog. 5, 418–431.

Gizaw, S., Mekonnen, K., Desta, L., 1999a. Lenche Dima IntegratedWatershed Development Project Feasibility Study Report. AmharaNational Regional State, Bahir Dar (19 May).

Gizaw, S., Mekonnen, K., Desta, L., 1999b. Yeku Integrated WatershedDevelopment Project Feasibility Study Report. Amhara NationalRegional State, Bahir Dar (24 May).

Haregeweyn, N., Yohannes, F., 2003. Testing and evaluation of theagricultural non-point source pollution model (AGNPS) on Auguchocatchment, western Hararghe, Ethiopia. Agr. Ecosyst. Environ. 99 (1–3), 201–212.

Heatwole, Conrad. 2003. QuickBird Satellite Image B 6 November 2003.AMAREW Project internal map.

Hengsdijk, H., Meijerink, G.W., Mosugu, M.E., 2005. Modeling the effectof three soil and water conservation practices in Tigray, Ethiopia. Agr.Ecosyst. Environ. 29–40 (1–2).

Herweg, K., 1993. Problems of acceptance and adoption of soil conser-vation in Ethiopia. In: Baum, E., Wolff, P., Zobisch, M.A. (Eds.),Acceptance of soil and water conservation. Strategies and technolo-gies. In: A Series on Soil and Water Conservation. Strategies andTechnologies, vol. 3. Deutsches Institut fur Tropische und Subtrop-ische Landwirtschaft, pp. 391–411.

Jabbar, M.A., Mohamed Saleem, M.A., Li-Pun, Hugo., 2001. Towardstransdisciplinarity in technology and resource management research: aproject in Ethiopia. Outlook Agr. 30 (4), 257–260.

Legesse, D., Vallet-Coulomb, C., Gasse, F., 2003. Hydrological responseof a catchment to climate and land use changes in Tropical Africa: casestudy South Central Ethiopia. J. Hydrol. 275, 67–85.

Nyssen, J., Poesen, J., Moeyersons, J., Deckers, J., Haile, M., Lang, A.,2004. Human impact on the environment in the Ethiopian andEritrean highlands – a state of the art. Earth-Sci. Rev. 64 (3–4), 273–320.

Nyssen, J., Haregeweyn, N., Descheemaeker, K., Gebremichael, D.,Vancampenhout, K., Poesen, J., Haile, M., Moeyersons, J., Buytaert,W., Naudts, J., Deckers, J., Govers, G., 2006. Comment on ‘‘Mod-elling the effect of soil and water conservation practices in Ethiopia’’[Agr. Ecosyst. Environ. 105, 29–40]. Agric. Ecosyst. Environ. 114, 407–411.

Planning and Decision Making. 1999. New Strategies for America’sWatersheds. Committee on Watershed Management, NationalResearch Council.

Rockstrom, J., Folke, C., Gordon, L., Hatibu, N., Jewitt, G., de Vries,F.P., Rwehumbiza, F., Sally, H., Savenije, H., Schulze, R., 2004. Awatershed approach to upgrade rainfed agriculture in water scarceregions through Water System Innovations: an integrated researchinitiative on water for food and rural livelihoods in balance withecosystem functions. Phys. Chem. Earth 29 (15–18), 1109–1118.

Shiferaw, B., Holden, S.T., 1998. Resource degradation and adoption ofland conservation technologies in the Ethiopian highlands: a casestudy in Andit Tid, North Shewa. Agric. Econ. 18, 233–247.

Special Report: FAO/WFP Crop and Food Supply Assessment Mission toEthiopia, 2006. Food and Agriculture Organization of the UnitedNations, World Food Programme, Rome (24 February).

Steenhuis, T.S., Pfeffer, M.J., 2000. Sustainable and integrated manage-ment of water, land and bio resources. In: Wu, W.-S., Chang, S.-T.,Guan, B.T. (Eds.), Proc. Agriculture of the New Century: ManagingBio-Resources and Bio-Diversity. Nat. Taiwan Univ., Republic ofChina, pp. 127–135.

Tekle, K., 1999. Land degradation problems and their implications forfood shortage in South Wello, Ethiopia. Environ. Manage. 23, 419–427.


overcoming limited information through participatory...

Documents