Turkey - Eastern Anatolia Watershed Rehabilitation Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No: 24181 IMPLEMENTATION COMPLETION REPORT (CPL-35670; SCL-3567A; SCPD-3567S; TF-25800) ONA LOAN IN THE AMOUNT OF US$ 77 MILLION TO THE GOVERNMENT OF TURKEY FOR THE EASTERN ANATOLIA WATERSHED REHABILITATION PROJECT March 31, 2002 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS (Exchange Rate Effective September 30, 2001) Currency Unit = Turkish Lira (TL) TL 1,00 = US$ 0,00000064 US$ 1 = 1,518,666 TL FISCAL YEAR January 1 December 31 ABBREVIATIONS AND ACRONYMS AGM Department of Reforestation & Erosion Control CAS Country Assistance Strategy EAWRP Eastem Anatolia Watershed Rehabilitation Project ERR Economic Rate of Retum FCPCPS Farmer-Centered, Problem-Census, Problem Solving GEF Global Environmental Fund GIS Global Information System GOT Govemment of Turkey ICR Implementation Completion Report KHGM General Directorate of Rural Services M&E Monitoring & Evaluation MARA Ministry of Agriculture & Rural Affairs MC Micro Catchment MOF Ministry of Forestry ORKOY General Directorate of Forest Village Relationship MTR Mid-Term Review NGO Non-Governmental Organization OGM General Directorate of Forestry PCSU Project Coordination & Support Unit PPF Project Preparation Facility QAG Quality Assurance Group SAR Staff Appraisal Report SPO State Planning Organization TKV Turkish Development Foundation TAGEM General Directorate for Agricultural Research TUBITAK Turkish Scientific & Technology Research Organization TUGEM General Directorate of Production and Development Vice President: Johannes Linn Country Manager/Director: Ajay Chhibber Sector Manager/Director: Laura Tuck/Marjory-Anne Bromhead, ECSSD Task Team Leader/Task Manager: Nedret Durutan  FOR OFFICIAL USE ONLY TURKEY TURKEY EAST ANATOLIA WATERSHED PROJECT CONTENTS Page No. 1. Project Data 1 2. Principal Performance Ratings 1 3. Assessment of Development Objective and Design, and of Quality at Entry 2 4. Achievement of Objective and Outputs 3 5. Major Factors Affecting Implementation and Outcome 13 6. Sustainability 14 7. Bank and Borrower Performance 15 8. Lessons Leamed 17 9. Partner Comments 19 10. Additional Information 23 Annex 1. Key Performance Indicators/Log Frame Matrix 24 Annex 2. Project Costs and Financing 26 Annex 3. Economic Costs and Benefits 29 Annex 4. Bank Inputs 32 Annex 5. Ratings for Achievement of Objectives/Outputs of Components 34 Annex 6. Ratings of Bank and Borrower Performance 35 Annex 7. List of Supporting Docmtnents 36 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.   j [toject ID: P009023 Project Name: TURKEY EASTERN ANATOLIA WATERSHED PROJECT ream Leader: S. Nedret Durutan TL Unit: ECSSD -CR Type: Core ICR Report Date: May 20, 2002 1. Project Data Name: TURKEY EASTERN ANATOLIA WATERSHED L/C/TF Number: CPL-35670; PROJECT SCL-3567A; SCPD-3567S; TF-25800 Country/Department: TURKEY Region: Europe and Central Asia Region Sector/Isubsector: AG - Agency Reform; AY - Other Agriculture KEY DATES Original Revised/Actual PCD: 03/05/1990 Effective: 07/26/1993 Appraisal: 06/15/1992' MTR: 11/01/1995 11/30/1995 Approval: 03/11/1993 Closing: 10/20/2000 10/30/2001 Borrower/lmplempienting Agency: REPUBLIC OF TURKEY/MOF/MARA Other Partners: iTAFF Current At Appraisal 'ice President: Johannes Linn Wilfred Thalwitz Countrv Manager: Ajay Chhibber Michael Wiehen .ector Manager: Marjory-Anne Bromhead Jamnes Goering 'ream Leader at ICR: S. Nedret Durutan Marjory-Anne Bromhead .'CR Primaiy Author: R. Suppa (FAO/CP) ;;. Principal Performance Ratings (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HL=Highly Likely, L=Likely, UN=Unlikely, HUN=H-ighly lJnlikely, HU=Highly Unsatisfactory, H=High, SU=Substantial, M=Modest, N=Negligible) Outcome: S Sustainability: L institutional Development Impact: SU Bank Performance: HS Bozrower Performance: S QAG (if available) ICR Quality at Entry: HS S Project at Risk at Any Time: Yes  3. Assessment of Development Objective and Design, and of Quality at Entry 3.1 Original Objective: The specific objectives of the project were to help restore sustainable range, forest and farming activities in three provinces in the upper Euphrates watershed, reducing soil degradation, erosion and sedimentation in reservoirs as well as increasing productiveness and incomes in this impoverished region of Turkey. By using a participatory approach, the project was designed to strengthen farmers' planning and implementing capacity while improving the responsiveness of rural services agencies to farmers' needs. A key underlying objective of the project was the environmental rehabilitation of degraded land for the conservation, in their natural habitat, of the genetic resources of globally significant herbaceous and woody species indigenous to Turkey. The objectives of the project were clear, realistic and in line with Government's priorities and Bank's country assistance and sector strategies that supported poverty alleviation through productivity increases, participatory approaches to development and institutional building to enhance sustainability. Active village participation was an innovative and essential feature of this project. The project therefore conformed to the Bank lending strategy for Turkey which included strong attention on environmental issues, and increased assistance to the poorer provinces of Eastern Turkey. By improving sustainability of forestry, livestock and agriculture, the project was designed to adequately contribute to increased productivity and higher rural income in these areas. Because of the focus on poverty and natural resources protection, the Government of Turkey (GOT) was fuldly committed to the project. Although institutions were considered able to expand their activities, financial resources and equipment were limited. The Bank's support-for the Eastern Anatolia Watershed Rehabilitation Project (EAWRP) was therefore justified as a response to the GOT's commitment and efforts to reduce ecological degradation, protect the natural resources and thus contribute to rural growth, employment and poverty alleviation. The project, at entry, was not considered exempt from some technical and institutional risks. Like all watershed rehabilitation efforts, it required the integration of conservation and development, coordination of interventions in crop, livestock and forestry production and revisions in the way land would be managed. The interventions for improved range management were perhaps the most risky as they required full participation by villagers and agricultural extension agents who did not possess such experience in this field. This problem was adequately addressed at appraisal by: (i) providing technical assistance and training; (ii) focussing on low cost participatory approaches to range management; and (iii) allowing for a review of progress within the first two years of project implementation and modifying the approach if considered necessary, possibly by contracting with Non-Governmental Organizations (NGOs) with experience in community organization. Intensive Bank supervision was foreseen at appraisal. 3.2 Revised Objective: The project objectives were not revised during irnplementation and components were maintained as foreseen at appraisal. The Mid-Term Review (MTR) Mission carried out by the Bank in October/November 1995 decided that "no changes in the treatment measures for forest and agriculture and income generating activities were justified". However, as a result of the strong project ownership and enthusiasm by local and provincial officials and the village communities, the GOT decided to increase the project area and in mid-1997, based on the experience of the initial three project provinces and together with considerable project savings, requested the Bank to initiate preparation for additional participatory micro-catchment (MC) plans in major watershed areas. Finally, eight more provinces, namely Adana, Kabramnaras, Sivas, Antalya, Isparta, Icel, Gaziantep and Sanliurfa were added to the project. The inclusion of the new provinces under the project required amendments to the original Loan Agreement, which were approved in October 1999. Expansion enabled the Borrower to: (i) test the "Participatory Watershed Management" - 2 -  ap proach in different socio-economic settings; and (ii) expose more provincial agencies to the approach. 3.3 Original Components: The EAWRP was designed to include the following components: (i) rehabilitation of an estimated 54 MCs through a series of treatment on cultivated, range and forest lands with participation of the local pzpulation; (ii) supporting activities including small-scale irrigation, horticulture and agriculture; (iii) support to project planning and management; (iv) adaptive research and pilot work to supplement and inprove the menu of treatment options; and (v) Global Environmental Fund (GEF) sub-project activities including survey and inventory, management of selected sites, monitoring, data management, institutional sirengthening and the preparation of a national plan for gene conservation. The original components fitted well with the objectives of the project and were within the administrative and financial management capacities of the main implementing agencies, i.e. the Department of Reforestation and Erosion Control (AGM) within the Ministry of Forestry (MOF), and the General Directorate of Production and [levelopment (TUGEM), the General Directorate for Agricultural Research (TAGEM) and the General E lirectorate of Rural Services (KHGM) within the Ministry of Agriculture and Rural Affairs (MARA). In aidition, the project design incorporated the main lessons from three previous Bank projects in lurkey namely the Second Agricultural Credit Project, the Erzurum Rural Development Project and the Northern Forestry Project and Agricultural Extension and Applied Research Project. 3.4 Revised Components: NWithout changing the project objectives, project components and physical targets were revised and funds were maintained throughout the project as also indicated in the MTR mission. However, as indicated above, during the course of implementation in mid-1997, the scope of the project was expanded to cover a total of 11 provinces, 8 more than the original number envisaged at appraisal. 3.5 Quality at Entrv: IThe Quality Assurance Group (QAG) did not exist at the time of the project's approval. Peer reviewers and Bank decision-makers endorsed the project's design and objectives which were consistent with the Country Assistance Strategy (CAS) and considered its quality at entry as satisfactory. The Project piloted key elements of a then evolving participatory watershed rehabilitation approach that focussed oni restoring sustainable range, forest and farming activities in the upper Euphrates watershed. The project was subject t) a QAG analysis in April/May 1998, and the Panel rated the EAWRP in respect to both the quality at entry and the quality of supervision as "Best Practice" and an example of the World Bank at its best. The I.roject made a sophisticated multi-sectoral and sequential project concept operational in remote, resource-poor mountainous rural areas of Eastern Turkey. It introduced a variety of small physical ilfrastructural and agricultural interventions demanded by village communities aiming at increasing their incomes in private lands. The Bank nominated the EAWRP a Project Excellence Award in 1999. As foreseen at appraisal, the project faced technical and institutional risks. These problems were addressed during implementation by provision of adequate technical assistance and training and effective liaison committees at national and provincial levels through which the implementing team could operate satisfactorily under continuous monitoring. However, some difficulties were experienced in relation to GOT budget allocations. The concem about the financial contribution from the GOT to suppolt project activities did not fully materialize at the beginning of the project and this was the main cause for some d elays of project implementation up to 1995. 4. Achievement of Objective and Outputs 4.1 Outcome/achievement of objective: After a short and frustrating start in the first years, and despite slow disbursement because of late -3 -  availability of counterpart funds and immature project management, the project finally had considerable impact on the degraded natural resources and households it was targeting and, indirectly, on policy, institutions and behavioral standards. The EAWRP"may well be considered as a flagship for the Bank in the important area of watershed management. In pursuing its objectives, the project commendably adopted a demand-driven approach, implemented through participatory MC planning, as distinct from a target-driven approach. At appraisal, the scope, scale and cost of activities could not be determined in advance with accuracy since communities would select from a menu of interventions. Therefore, based on the field work during preparation, best estimates were used for the purpose of cost and benefit calculations. During implementation it was observed that MCs were diverse in terms of both ecological and socio-economic conditions and attitudes of communities which in turn affected the scope and scale of interventions. This inevitably led to departure from the basic assumptions made at appraisal regarding the areas to be achieved under the various components. The project is noteworthy for having focused on quality rather than aiming at simple area targets and showing considerable flexibility and resulting in learning by doing. The project has achieved unprecedented coordination among the implementing agencies, MOF, MARA and KHGM. In conclusion, the outcome of the project as a whole, both in respect to its development objectives and its physical implementation, is rated satisfactory. The success of the implementation created an interest in many countries and Turkish implementing agencies hosted visitors from Morocco, Ethiopia, Georgia, Azerbaijan, and Albania. The achievement of each development objective is summarized below. This report takes no account of the GEF financed In-situ Conservation of Genetic Diversity sub-project which was evaluated separately in the Implementation Completion Report (ICR) dated April 30, 1999. 4.1.1 Restore Sustainable Range, Forest and Farming Activities, and Increasing Productivity and Incomes: The project initiated innovations in the protection and rehabilitation of the environment while helping rural people to increase incomes. A comparative baseline evaluation study, based on sample data, indicates that poverty in the project villages was reduced and that average incomes have more than doubled over three years. The increased income is elaborated in para 4.4 and in Annex 3. It is reasonable to assume that the project has made a significant contribution to the increase in farm income, which, in turn, has facilitated the communities' interest in the improvement of the management of the communal and public lands, on which they rely for grazing and forest products. The MOF, in turn, actually introduced the innovations for the management of these lands. In terms of area treated, the project has over-achieved on forest land (including forest rangeland) and under-achieved on non-forest land particularly on rangelands compared to the Staff Appraisal Report (SAR) estimates. The appraisal estimates depended on provincial annual statistics. However, statistics regarding rangelands did not reflect the reality due to obscure perception of rangeland and lack of legal definition. Generally, most non-crop land was classified as rangeland and this resulted in over-estimates. It was only after the issuance of the "Rangeland Act" in 2000 that efforts have started to re-define the "rangeland" with the support of the ground work to delineate the rangelands in the country and bring legal clarification to the issue. Therefore, during the MC planning process and also in implementation the actual situation of the rangelands were understood. The project activities regarding forest rehabilitation are expected to make significant contributions to combating erosion and will, over time, also produce wood, other forest products and feed for the livestock. In particular, villagers were very happy with rangeland closure which provided feed for the livestock and also for the bees during drought periods. The productivity of marginal farm lands has been successfully addressed by assisting farmers to adopt improved practices. Soil fertility management and the use of vegetative means of soil and moisture conservation have been successfully promoted through both agronomic and horticultural practices and the significant advances made in forage production have been associated with an increase in the stall-feeding of livestock. The supporting activities, designed to reduce farmers' dependence on forest and rangelands which led to reduced pressure on the natural resources have been popular, small-scale irrigation especially -4 -  so. The small-scale irrigation component of the project was very popular and provided an increase of the net value of production of about 380% in MC villages. 4.1.2 Strengthening Farmers' Planning and Implementing Capacity: The Farmer-Centered, Problem-Census, Problem-Solving (FCPCPS): Approach has been successfully implemented and ensured the participation of the villagers' in decision making, MC planning and implementation. The achievement of a high level of community participation in the planning process is a considerable strength of the project. This meant that the scale of various interventions and the implementation pace in a MC was detennined largely by community preference. In the early years of the project, some communities were sceptical about the project approach and decided not to participate in the project at all. However, after observing the successful implementation in neighboring villages, they applied to the agencies to be included in the project. Villagers' volunteering ensured commitment and contributions for rehabilitation and use of their natural resources. As a result of this, the project implementation teams organized field trips regularly for the communities of the newly included MCs to those in which the rehabilitation activities were more advanced and the results were visible. This enabled the villagers observe the developments and also interact with other communities where such practices were proven to be successful. In strengthening the communities' planning and implementing capacity, coordination among the implementing agencies contributed a lot and proved to be a major project achievement. From a pre-project situation of wholly independent entities, the three implementing agencies involved now coordinate their activities within the project. This collaboration took time to mature and is evident both in Ankara, where the Project Coordination & Support Unit (PCSU) resides in the AGM and at provincial level which is generally adequate to good. However, in those provinces most recently included in the project it needs more time to mature. 4.2 Outputs by components: 4.2.1 Watershed Rehabilitation: Physical achievements are summarized at Annex 1. The total area treated including forest, range and arable land was 116,521 ha and was thus was less than the total of 216,849 ha envisaged at appraisal (Annex 1: Key Performance Indicators). The shortfall was on agricultural land rather than on forest land, where the SAR figure of 62,370 ha was exceeded in implementation by 53%. The major sub-components are reviewed separately below. Overall performance in this component is considered satislractory. Improvements in farming practices: This sub-component consisted of a series of further and inter-related sub-components and steps. The area treated under this was 29,810 ha, against an SAR estimate of 38,167 ha. The comparative baseline study shows that yields increased over the three year (as per the survey) period of comparison i.e. the average yield increase in irrigated wheat in Adiyaman, Malatya, and Elazig (original project provinces); 40% (1800 kg/ha in 1998 versus 2520 kg/ha in 2001). In these provinces, the average increase was less pronounced under rainfed conditions; only 13% due to severe drought prevailing since 2000 (1490 kg/ha versus 1690 kg/ha). However, despite the severe drought in some MCs, farmers managed to be affected less as a result of better variety introduced through the project and agronomic practices i.e. average annual rainfall dropped 60% in Isparta in 2001, but the yield decrease was only about 18% (30 kg/ha). On the average, in project provinces, the yield of irrigated barley increased by 54% and that of rainfed barley by 22% due to the introduction of a winter hardy variety. For maize, tomato and pepper under irrigated and for lentil and chickpea under rainfed conditions, the figures show increases of 44%, 46 %, 53 %, 33 % and 18 % respectively. These are all substantial increases over a three year period. -5 -  The comparative figures for agronomic packages indicated an evolution from subsistence farming with almost no inputs and poor tending to semi-conmmercial fanning with some inputs and better agronomic practices i.e. the number of farmers using phosphorus fertilizers for major crops almost doubled, there was a shift from broadcast seeding to seeding with a crop drill; 12% increase in the number farmers adopted drill use in wheat and barley production, and the number. of farmers weeding their fields almost tripled. The project promoted forage crops and in turn provided strong support to livestock production which has been a declining sub-sector. Farmers who used to be familiar with only alfalfa as a forage crop produced under irrigation benefited from Hungarian vetch produced in rotation with wheat/barley under rainfed conditions, i.e. in Adiyaman, there was no Hungarian vetch in the first two MCs before the project, as of December 2000, there was an area of 191 ha under Hungarian vetch produced with the seeds provided by the project, and an additional 43 ha, gained as a result of adoption of the practice by the farmers who financed the seed themselves. In seven years, farmers of these two MCs adopted this forage crop and the area under Hungarian vetch reached 234 ha. The project also introduced sainfoin, a perennial forage crop produced on marginal agricultural land on which wheat production caused severe erosion due to annual cultivations, i.e. in Sivas, the marginal agricultural land under wheat production was converted to sainfoin on an area of 2,000 ha. Another important project achievement which was introduced in later years was the successful promotion of silage, made from maize, previously unknown in the majority of the project area. As a result of efforts in this area, in 11 project provinces, forage crop production is expected to increase to about 3,100 tons/year of dry matter as a result of project efforts. The project was not successful in expanding the contour tillage. A particular constraint to contour tillage on slopes lies in the narrowness of the parcels into which much of the land is divided. Parcels lie length wise up the slope rather than across it, the result of dividing the land up equally between family members. Farmers are reluctant to use tractors in narrow parcels and contour tillage is therefore practiced only by farmers who still have horses. The project achieved increased awareness among the farmers in the MC communities on replacing the fallow with food legumes (chickpea and lentil) and particularly with feed legumes (H. vetch and sainfoin) produced in rotation with wheat under rainfed conditions. The project also introduced chickpea varieties resistant to a disease which is currently causing important harvest losses throughout the country. Rangeland: Against an SAR estimate of 134,112 ha of rangeland to be treated, the project achieved 21,664 ha, mostly on forest rangeland, under both MARA and MOF responsibility. The area of agricultural rangeland treated under MARA was 2,382 ha against an SAR estimate of 116,312 ha. The main reason is the misleading agricultural statistics as explained in 4.1.1. It should also be noted that in some areas, it was not worth investing in rangeland rehabilitation because the degradation was irreversible. Progress on rangeland activities which were the responsibility of MARA was impeded by uncertainties as to ownership (especially where rangeland typically adjoins private lands and infrastructure rights), the absence of cadastral maps and related surveys, a history of encroachment and an inadequate legal framework for MARA's participation and authority. These issues together meant that, in practice, there was considerably less rangeland available to the project than had been anticipated. These issues were addressed during the course of, and outside of, the project by new Rangeland Act approved by Government in 1998 which had reportedly been under consideration by Parliament for 37 years. The Act defines the respective roles and responsibilities of government (MOF and MARA) and local people in rangeland management. It thus gives MARA institutional authority and responsibility with regard to rangeland, which was formerly lacking. The law also requires cadastral mapping to be undertaken of all rangeland which is likely to take many years to complete. - 6 -  There were technical constraints within MARA associated with the lack of previous rangeland experience md, as noted at appraisal, a lack of adaptive research into rangeland issues. MARA's focus had formerly teen on providing assistance to the private owners of cultivated land, rather than to the communal mnanagement of poorly understood common property resources. The SAR's prescription for the piloting of aerial seeding was not pursued due to the cost involved and unsuitable ecological conditions. T rials regarding enrichrnent seeding and broadcast application of fertilizers did not work due to drought and/or :old. The project staff were mislead by the over emphasis on physical interventions (seeding, fertilizing) recommended by international and national consultants and academicians and it took a while for them to realize that the most cost effective and practical intervention was the closure of land to grazing and letting the flora regenerate. It is still noteworthy that rangeland efforts have had some definite success. Fencing off of the area and rotational closure to grazing was proven to be very successful in some provinces, i.e. Sivas, Malatya, Adana and Icel. Local people accepted the closure since: (i) they were aware of the fact that the existing capacity of the rangelands was far from supporting a profitable livestock production; (ii) there has been a tendency to shift from small ruminants (rangeland based) to big ruminants (mostly stall fed); (iii) the project introduced and /or supported cultivated forage which could easily compensate the feed loss due to closure; (iv) they recognized that productive rangeland and cultivated fodder are complementary for a profitable livestock production; (v) they realized that livestock production could not be improved only with upgraded genetic material, feed is a must; (vi) the extension staff created awareness about the impact of overgrazing on the productivity of the rangelands; and vii) the communities developed confidence in the implementing agencies after observing the achievements realized by other interventions in the same MC and/or in other MCs. Forest Land: Physical achievement on forest land exceeded the estimates made at appraisal in all except one sub-component, namely oak coppice rehabilitation. The project understandably decided to amalgamate three of the original sub-components in the SAR and to re-define soil conservation afforestation to include also conifer plantations and fuelwood coppice plantations (oak planting and acom seeding). The SAR estimates for the original constituents of this newly defined category amount to 26,700 ha, against which the project achieved 71,991 ha. This considerable over-achievement reflects the fact that these activities were taking place on land which was: (i) relatively distant from villages; and (ii) under the control of the MOF. Local people may have previously grazed animals on the land used for afforestation. But they did not rely heavily on it and were content to have it included in MC plans and to benefit from the employment afforded by the afforestation activities. The situation was thus in contrast to that of oak coppice, discussed below. Although survival rates for newly planted seedlings are reportedly 70% or higher, there is considerable variation in the quality of the planting stock used, reflecting often low nursery standards. AGM has made limited progress towards using improved seed. It has established a seed orchard of red pine. But much of the planting, including all that of oak, is from seed collected indiscriminately by villagers. Against an SAR estimate of 17,800 ha of oak coppice rehabilitation (cutting of degraded oak stands to encourage coppicing), the project achieved finally completed an area of 2,240 ha. The explanation for this lies partly in a relative unavailability of oak coppice in the particular MCs selected by the project. The selection of MCs partly on the basis of the amount of eroded land they contained inevitably tended to exclude oak coppice, which is typically associated with less eroded areas. There was alsc a degree of social resistance. Local people depend on oak coppice for fuelwood and regard it in much the same light as they do rangeland, as a common property resource. In this regard, they proved generally unwilling to have the closure of oak coppice included in MC plans. A relatively minor institutional factor was that, because oak coppice rehabilitation involves cutting trees, it is the responsibility of the MOF's General Directorate - 7 -  of Forestry (OGM) as it was the case for cedar rehabilitation, rather than of the AGM which was responsible for other forestry activities within the project and indeed for overall project coordination. The area of forest rangeland treated, under the responsibility of MOF, was 19,282 ha and thus exceeded the SAR estimate of 17,800 ha. Forest rangeland is generally remote from villages and at higher altitude than the category of rangeland for which MARA was responsible. It is not subject to the same level of uncertainty regarding ownership, boundaries and encroachment and there is little or no ambiguity about the MOF's authority over it. Against this background and partly because of the employment afforded, the local people were more willing to accept and actually to support the rehabilitation activities undertaken by the MOF than those attempted by MARA. But there have been instances, notably in Sanliurfa and in the Toros Mountains of Icel and Antalya, where the project has had to address issues associated with transhumants' use of the high altitude rangeland for summer grazing. The project concluded that, in both categories of rangeland, the most effective means of rehabilitation lay in simply closing the area to animals for a period prior to the introduction of rotational grazing. The MOF had the power (which MARA did not) to fence off areas of rangeland in order temporarily to prevent grazing. It is finally to be'reported that farmers in Sanliurfa indicated that, despite the severe drought in some years, they did not migrate from the area because, thanks to the project, the rangeland gave considerable support to their livestock in terms of sufficient feed availability also in these years. The forest rangeland treatments have included not only fencing but also erosion control measures such as check dams and some limited infrastructure (mainly water troughs and dips) for animals. There appears to have been some variation in the costliness of the techniques and materials used and, in some instances, a degree of over-engineering. The project planted 81 ha for the purpose of riverbank protection, against an SAR estimate of 70 ha. This planting, on Treasury land, was done using poplar, Eleagnus and oak. The project also created two new sub-components, namely cedar rehabilitation and participative plantations, for which there are therefore no SAR estimates The rationale for the inclusion of cedar rehabilitation came when the project expanded from its original three provinces, where cedar forest was rare or absent, into provinces where it was important, notably Adana, Icel (Mersin) and Antalya. The project achieved 1,687 ha of cedar rehabilitation and 35 ha of participative plantations. Strengtheningfield services: This was achieved through training, some technical assistance, and the upgrading of infrastructure at the field and central levels. There was a substantial overseas training program for the project staff, administered by external consultants, under which 282 trainees, participated in 34 study tours between 1995 and 2000. Eleven study tours were in the United States, eight were in Israel and the rest were distributed between ten different countries. There were also three short term overseas training courses in the United States and were attended by 20 project staff. Subject areas covered mainly the technical aspects of watershed development and erosion control. Twelve staff received English language training. Those who attended confirmed a high degree of satisfaction with the training provided. Based on the identified needs, the project agencies provided training for their staff through programs organized by TUGEM of MARA. The program built up slowly, from 2-6 truning events in each of the years 1993 to 1997 to 16 in 1998, the year in which 30% of a total of 53 courses and workshops took place. By the end of the project a total of 1,533 project staff had undergone local training. Thirty eight per cent of the trainees were from the MOF, 47% from MARA and 15% from KHGM. The expansion of the project from 3 to 11 provinces, combined with a considerable staff turnover and arising problems during implementation, meant that there was a continuing and growing need for local training. The project developed training notes that can be useful for both national and international stakeholders. Overall the - 8-  quality of the training was quite good, although the program could have included more training on extension methodology, and financial analysis. Deficiencies in procurement experienced in the early years of the project were overcome starting from 1996, when procurement responsibility was decentralized away from the Central Procurement Committee in the MOF and became the responsibility of each individual implementing agency. There were some delays especially in the procurement of vehicles, adversely effecting performance; in some instances cases tCe project had available to it some vehicles which existed already. The project decided against buying same of the heavier earthmoving and ripping equipment budgeted for in the SAR, to be used by KHGM and AGM. This resulted in significant savings in funds. These agencies instead relied upon subcontracting the works. 4.2.2 Income-supportingActivities: T'hese included measures to improve rainfed productivity for field and horticultural crops, assist with small-scale irrigation, promote bee keeping and grafting of wild fruit species. At project completion, the raajority of the agricultural activities have been completed as expected and in some cases have exceeded SAR targets (Annex 1). Overall performance is considered satisfactory. Income supporting activities were conditionally linked to the villagers adopting forestry treatments and improved range management p ractices. MC planning emphasized the linkages between the income supporting activities supporting the a doption of desirable resource management activities. In the majority of the MCs, costs of these activities sid not exceed about one third of the total investment cost for the MC. The project staff focused on three raajor groups in a community as identified through the FCPCPS process: (i) owners of large livestock f locks who are using the rangelands; (ii) groups who are benefiting from oak coppice as fuelwood and forage; and (iii) groups who are using marginal agricultural lands. In order to make behavioral change in the use of these natural resources, the groups who are using these resources were given the opportunity to shift to alternative income supporting activities. This being the principle, favoritism was avoided in granting the project inputs. laken as a whole, the rainfed horticultural treatments achieved an area of about 14,350 ha and so exceeded the estimate made at appraisal of 11,862 ha. The project has combined two SAR sub-components into one. 'hese are rainfed horticulture and rainfed terrace agriculture. Against an SAR estimate for these of 5,616 ha, the project achieved 5,554 ha. Rainfed horticulture in gullies achieved 1,642 ha, against an SAR estimate of 3,246 ha and suffered frorn farmers' understandable preference to concentrate on more familiar types of land. In all three sub-components there were some difficulties in supplying the required quantity End type of planting material. A popular crop combination was almond and grape, these being planted in combination on terraced land, the almond (or other fruit tree such as cherry) on the terrace and the grape on the terrace edge. There was also a small proportion of forage crops and cereals. The two sub-components involving pistachio both performed well. About 1.2 million wild trees were grafted and this represents an area of about 6,000 ha, or three times the SAR figure. For planted pistachio, actual and SAR figures were comparable at approximately 1,000 ha. A total of 240,000 trees have been distributed by the project under I he tree planting on field boundaries covering an area of about 200 ha or 60% of the SAR target. In the earlier years of the project, there were problems regarding apiculture activities mainly due to the overlapping and uncoordinated responsibilities of the related agencies, namely General Directorate of I roduction & Development (TUGEM), General Directorate of Forest Village Relationship (ORKOY) and I he Turkish Development Foundation (TKV), a private supplier. Colony losses occurred as a result of late (delivery, insufficient basic and on-the-job training, mis-selection of beneficiaries. In 1997 a pnrtocol was signed clearly defining the role and responsibilities of each agency, TUGEM, ORKOY and TKV and the performance increased significantly. Against an SAR estimate of 1,620 beehives, the project has supplied -9-  1,728. This sub-component has proved especially popular in the dry and relatively treeless province of Sanliurfa, where crop production significantly failed as a result of severe drought and beekeeping accounted for a very high proportion of the cash income of those engaged in it and its promotion by the project represents real success in poverty alleviation. The comparative baseline study shows significant advances in upgrading livestock in the project MCs. Over the three year period, the percentage of cross-bred cows increased by 17%. Milk yields increased by about 20% in both local and cross-bred animals. The project's direct contribution to the genetic improvement of cattle was limited and its effect was more through the promotion of feed base rather than from more direct intervention. A serious constraint to the intended artificial insemination of cows was the relative remoteness of many of the project farms. The provincial animal health services were already fully employed by the farmers in the lower-lying and more accessible parts of their provinces. A decision to employ private veterinary expertise on a pilot scale was implemented in 1997 in one province, Malatya. The remoteness of many of the project farms made it difficult for the private veterinary specialists to reach them with the necessary punctuality. A low success rate of about 50% led to the discontinuation of artificial insemination efforts. SAR proposals to use breeding bulls between groups of farmers (to obviate the need for timely visits by veterinary specialists) were not implemented, owing to the difficulty of managing such animals communally. Small-scale irrigation was understandably the most popular single sub-component with farmers. The total area treated was 12,368 ha, against an SAR estimate of 10,530 ha. This sub-component improved the existing water holding capacity by building concrete ponds versus the existing earthen ponds by reducing the water losses while improving the efficiency of conveyance by the provision of concrete canals as opposed to earth canals. Furthermore, the employed technology allowed collection of water from sources as small as 1 lit/second enhanced the capacity of the farmer to irrigate from sources which were previously unused. The project supported only a few schemes with pumping, however, the pumping equipment was financed totally by the farmers, i.e. in Kahramanmaras. The irrigation provided has permitted the expanded cultivation of forage species, especially alfalfa, silage maize and a range of fruit trees, including apricot (especially in Malatya), apple, peach and cherry. 4.2.3 Planning and Management: The amount and nature of technical assistance used in the project differed from that anticipated at appraisal. There was a substantial shift away from intemational towards local technical assistance. International technical assistance was drawn upon only in 1995 and 1996 and amounted to 392 days, or 16% of the total anticipated. Total local technical assistance amounted to about six times the 39 person-months anticipated at appraisal. From 1997, each of the main implementing agencies were allowed to select their own local consultants, the previous arrangement whereby AGM selected all consultants having been found unsatisfactory for disciplines other than forestry. Of the 6,000 person-days of local technical assistance, half was accounted for by two long-term appointments related to monitoring and evaluation. Much of the rest was accounted for by the local consultants in forestry, agriculture and irrigation who were employed up to 1997 to assist in MC planning, related soil and water surveys. From its original design embracing 54 MCs in 3 provinces, the project expanded in 1998 to include a further 18 MCs in 3 new provinces. In 1999, it expanded further, bringing its overall scale to a total of 88 MCs in 11 provinces. During appraisal it was foreseen to select three MCs per year in any one province, there were some instances which this was not achieved due to insufficient and/or immature capacity of the provincial agencies and/or attitude of the MC communities, i.e. in Antalya the community withdrew their agreement to project activities due to their perception of forestry activities. In the initial years of the project, the one year period allowed for selection and planning is barely sufficient. - 10-  7he Mid-Term Review in 1995 commented on the tight scheduling of planning and implementation and iecommended a longer time frame. The project did not select any new MCs the following year (1996) to catch up with the work. Despite the coming project completion date, agencies wanted to wante(d to continue to select new MCs by The rapid expansion towards the end of the project placed some strain on the capacity of the implementing agencies and left 37 MCs with work remaining to be completed from GOT funding (regarding which the mission received assurances from the implementing agencies, as discussed below). The project provided farmer training for both men and women and its public awareness and community 1raining extended also to children. The project also promoted farmer to farmer contact between new and old MCs. The effectiveness of the latter is partly demonstrated by the figures for the sowing of fodder crops. The total pre-project area sown to fodder in the project MCs was about 1,100 ha. With the project, Ihis rose to approximately 7,500 ha. But a further 3,000 ha was established by farmers not themselves directly assisted by project staff. Their contribution to the total incremental area under fodder wvas nearly 30% over the life of the project. In total, the project organized 279 field days and 457 training courses for 'arming family members and through them directly reached 12,242 participants. This amounts to less than the estimate at appraisal of 30,000 person-days of farmer training, a discrepancy which partly reflects the slow start. I ionitoring and evaluation was addressed by the employment, from 1997, of a specialist consultant who first helped to design and then managed the PCSU's monitoring system. Earlier international technical assistance in this area had proved ine'fective. The system functions extremely well in monitoring physical and financial inputs to the project The system in its present form does not attempt to monitor outputs or r esults, for example changes in crop yield, or variations in factors such as climate and streamflow. It iherefore performs a more limited function than was envisaged at appraisal. MARA's Monitoring and ]-valuation Section, which is outside of the project, undertook the comparative baseline evaluation with advice from TUGEM. This was based on interviews with farmers and compared information from approximately 600 farms in 63 MCs from all provinces. The years compared were 1998 and 2001. While it would have been preferable to have had an earlier initial baseline for those provinces covered by the project prior to 1998, the exercise was a valuable one and presented much useful data. 'rhere is additional information on project results which has been collected by the participating government bodies, especially TUGEM. Such inibormation is not generally available, nor is it available in E.nglish, and could very usefully be made so. It should ideally be incorporated into the monitoring and evaluation :;ystem, as was envisaged at appraisal. Although the SAR allowed for the creation of a Geographic Information System (GIS) and mapping i:apability (in Year 3), the project did not proceed with this. Instead it contracted in 1999 the Space 'rechnologies Department of the Turkish Scientific & Technology Research Organization (TUBITAK) Marmara Research Centre, to undertake two comparative studies of two MCs of Kumludere and Hancayi in Malatya. The years for which satellite imagery was compared were 1992 and 1998. The study showed that, while the project activities initially reduced vegetation cover slightly, there is subsequently an increase n the proportion of denser vegetation in these MCs. The comparison noted increases in the area under field crops, orchards and vineyards. On the basis of the above, the performance under this component is considered satisfactory. 4.2.4 Applied Research: 'Performance under this component is considered unsatisfactory. Except for the GIS work described above (which was research into, rather than the routine application of, a particular methodology) and some - 11 -  research just starting into fertilizers on rangeland (which will be too late to fall within the project), there was little activity in this component. Certain of the research identified in the SAR, for example that on aerial seeding and fertiliser application, was considered unsuitable in the light of the disappointing results from comparable non-aerial treatments and of the particular local conditions. Other research, the need for which was identified either at appraisal or from project experience, did not materialize. There was a suggestion, for example, of a need for research into the root fungi (Mycorrhiza) associated with cedar, as it was assumed that failures in cedar planting might be due to the absence of the fungi. Supervision missions repeatedly drew attention to the lack of research in, for example rangeland treatments and sedimentation rates. One limited justification for the lack of research under the project is that relevant research was already being done elsewhere. The World Bank, for example, was supporting agricultural research and extension in eastern Turkey through a separate project, approved in 1992. Also, the need for research was questioned. But one major reason appears to be the difficulty of incorporating into the project activities which would have to be done by organizations not already in the project's mainstream. There was reportedly reluctance on the part of some research bodies to enter into contracts with AGM. A number of potential research organizations are named in the SAR. The difficulty appears to have obtained even when the appropriate research organization, for example in forestry, was a sister organization to one of the implementing agencies. 4.3 Net Present Value/Econonmic rate of return: The project initiated innovative ways in the protection of the environrnent while helping rural people to make a living. An exercise carried out with the provincial staff by questioning them about their perception of project benefits included: (i) institutional benefits; (ii) social benefits; (iii) economic benefits; and (iv) environmental benefits. The economic re-evaluation of the project is presented in details in Annex 7, Appendix A. The Monitoring and Evaluation Unit (M&E) of AGM was able to collect data on the performance of different activities as implemented during the project including: (i) soil conservation and afforestation; (ii) oak coppice rehabilitation; (iii) rangeland rehabilitation; (iv) fallow reduction; (v) agronomic package; (vi) rainfed horticulture; (vii) rainfed gully horticulture; (viii) pistachio grafting; (ix) pistachio establishment; (x) irrigated horticulture; (xi) irrigated forage; (xii) small-scale irrigation; and (xiii) beekeeping. Basic data regarding these activities are based on a "Baseline Survey" carried out by both MOF and MARA during 1998-2001. Although it may still be argued that these activities are somehow complementary each other (SAR approach), the availability of good related data as provided by M&E particularly on area yields, production costs and investment costs, gave the ICR mission the possibility to estimate the Economic Rate of Return (ERR) for the project as a whole and, separately, for each of the above activities. On the basis of the assumptions, summarized in Annex 3 and given in detail in Annex 7, Appendix A, the Economic Rate of Return (ERR) for the project as a whole has been reassessed atl6.8% in line with the estimate made at appraisal. The project activities, taken separately, are expected to generate ERRs ranging from 18.4% to 32.9% for the agronomic package and gully horticulture, respectively. 4.4 Financial rate of retirn: At appraisal, financial benefits were only expressed in terms of average incremental income per family in the project area. At completion, it was possible to estimate the impact of the project on family income that would directly deriving from the activities supported by the project In general, farmers in the project area have contributed to the investment programs with their own labor. The net annual incremental income is estimated, on average, at US$270,000 per MC taking into account the expected aggregated benefits from the various activities supported by the project. This is equivalent to about US$590/family, which is in line with SAR estimates. While the increases in the net value of production are generally expected to range from 28% (agronomic package) to 390% (rainfed horticulture), the gully horticulture and the irrigated horticulture activities are expected to generate, at full development, values that are tenfold the related - 12 -  without project levels. The beekeeping is to be considered as a new activity in the region. A kit of 20 be ehives, representing the average package given to farmers, is currently generating about US$5'20/year. Details on expected net value of production at full development are given in Annex 7, Appendix A, Tables 8-15 for the activity under MARA responsibility and in Tables 16-18 for the soil conservation alforestation, oak coppice rehabilitation and rangeland rehabilitation respectively. 4.5 Institutional development impact: This has been manifested in two principle ways. The first concerns the institutional development which has taken place within the three implementing agencies, which were exposed to new technologies and good pr-actices in various fields including environmental survey and monitoring cultural heritage awareness and community mobilization. They have developed the capacity to coordinate their activities and so deliver a mrore effective service. They have also gained experience in communicating and collaborating with firmers, and this marks an important shift in the relationship between government and the rural population b)th in the project MCs and more widely. The second main impact on institutional development has been al. village level. The capacity of farmers to collaborate with each other has been strengthened by their awvareness of the advantages afforded by the project in the form of technology, infrastructure and the ixaproved management and conservation of community land resources. The project thus successfully intermediated knowledge and technology. I he M&E Unit, well staffed, is now working satisfactoiily and will remain a sustainable asset of AGM. It has gained the necessary capability to adequately follow, in the future, the implementation experience and p -rformance of any similar project without the assistance of consultancy expertise. All project staff have had access to updated information on costs and as well as on physical progress of field activities. The quality of their progress reports during the last years of project implementation was very satisfactory. 5. Major Factors Affecting Implementation and Outcome 5.1 Factors outside the control of government or implenmenting agency: E:asically, most factors were within the control of both the Government and implementing agencies. In n trospect, the main factor that affected negatively project implementation was the unexpected nmacro-economic difficulties combined with restricted budget allocation during the early years of-the project. However, the implementing agencies were able to cope efficiently with the greater than expected rite of inflation experienced during the life of the project. Delays and unpredictable reductions in cperational budgets were experiences in 1999 because of the provisional limitation of funds due to the Marmara Earthquake Recovery Program. The latter was one of the main reasons to justify the extension of tie closing date to 30 September 2001. A temporary factor in the early years of the project was the uncertain security situation, which prevailed in 1994 which made more experienced contractors reluctant to wY fork in the project area between 1993-94. Soon after, implementation took place in a stable social environment 5.2 Factors generally subject to governnent control: Complex GOT budgetary and disbursement approval procedures caused delays in World Bank disbursement and GOT expenditure and thus in project implementation. Until 1998, disbursements were slowed done due to the difficulties in using the Special Account for local expenditures. The Turkish Lira (1L) payments had to be recorded in Government budget and transferred to the TL budget account in the Central Bank before the payments were made. Therefore, it used to be taking almost 2 months until the funds were reaching to the budget of the implementing agency. However, both the World Bankc Projects I)epartment and the Public Finance Department of the Treasury made serious efforts to shorten this process and developed a new procedure which opened the way for TL expenditures after 1998. In addition, there was throughout much of the project life a shortage of GOT counterpart funding, the result of government - 13 -  austerity measures. In some cases, even the amount of funds was satisfactory, the timing of the release of funds hampered the implementation due to the seasonal nature of the interventions. Toward the end of the project, there has been some improvements in timing. 5.3 Factors generally subject to implementing agency control: The project has suffered from staff turnover at provincial level and this has often meant the loss of trained staff and the need to train replacements. Lack of capacity in one province (Sivas) were overcome by placing the project activities there under the coordination of the project staff in a neighbouring one (Malatya). The delays in procurement which were a feature of the early years of the project were overcome by transferring responsibility for procurement to the Project Coordination & Support Unit (PCSU). Initial inexperience in collaboration between the three main implementing agencies was successfully overcome and the coordination that ensued was a key factor in project success. It was strong at headquarters and in most but not all of the provinces. Where coordination at field level was weak, there were some instances in which the intended sequence of activities was not adhered to. Within MARA, institutional arrangements at provincial level, which do not mirror those at headquarters, were in some instances not clear and this led to procedural delays. 5.4 Costs andfinancing: The project's original closing date was extended from 09/30/2000 to 09/30/2001. It is estimated that the total project cost, excluding the GEF component, will reach US$78.3 million or 71% of appraisal estimate. This could be attributed mainly to the impact of high inflation on project implementation was intimately linked to exchange rate policies. In Turkey, where the exchange rate floats freely and fully adjusts, inflation has reduced project costs in US$, as the cost of civil works rises more slowly than the general price level, and the change from expensive mechanical to cheaper manual works as experienced in some project components. The project, at appraisal, was to be financed 70.1% by the Bank and 29.9% by the GOT; the final figures are 61.2% and 38.8% respectively. More detailed figures on project costs and financing are presented in Annex 2. 6. Sustainability 6.1 Rationalefor sustainability rating: The project sustainability is considered likely. The key to the project's sustainability lies in the fact that its activities have been undertaken with the agreement and participation of villagers with a strong measure of ownership by the primary stakeholders. The project has built on technology already familiar to villagers and reacted positively to villagers' preferences and suggestions. The project has had a significant impact on agriculture, horticulture and fodder production and, through these, on people's income. The comparative baseline study shows an increase in average annual net income per farmer of 136% over the three year period, from $1,197 to $2,827. A noteworthy feature of the change in income is that, while the increase in maximum net income was 53%, that for minimum income was 86%. While increases in income cannot be attributed exclusively to the project (and are partly related to off-farm activities) it is reasonable to conclude that the project has reached the poorest and addressed rural poverty, the root cause of environmental degradation. Notwithstanding that much of any increase in income to the poorest can be attributed to wage labor, including that provided by project-sponsored forestry activities, the figures provide strong evidence for sustainability. A related indicator of sustainability is a reduction in the rate of outward seasonal migration from the MCs in most of the project provinces. Overall the average reduction over three years was 7%, although there was considerable variation between provinces. The original three provinces showed an average reduction of 25%. Again, it is reasonable to assume that the project is partly responsible for a decline in outward migration. - 14 -  There was a cost sharing arrangement for the income supporting activities. The project provided the seed and seedlings for newly introduced varieties but only for the first year. Farmers were responsible for financing the tending and also provision of all inputs for the following years. The project's sustainability will be strongest on private land and especially in those activities involving irrigation for which the cost sharing was about 3% to 5%. Farmers have generally provided the cement for the irrigation ponds constructed under the project. They were also responsible for the operation and maintenance which was handled through the village administration. For bee keeping, villagers paid the value-added tax (VAT) upfront which was about US$130. Sustainability is also likely on the non-forest rangeland which, in the event, accounted for relatively little of the project effort. The combination of villagers' support and MOF presence on forestland, ensures protection, from any set back of related activities initiated by the project. Further, the sustainability of project activities under the responsibility of the MOF, is also assured by the fact that the Ministry is able to cover some of the budgetary needs out of intemally generated funds. 6.2 Transition arrangement to regular operations: The implementing agencies presented the mission with itemized budgets for 2002 relating to further work in the project MCs. Together they amount to the equivalent of $3.6 million for the year. KHGM's proposed program is for only four provinces (Gaziantep, Icel, Sanliurfa and Sivas), where there is engineering work remaining, the cost of which is estimated at US$1.8 million. AGM's budget for all eleven project provinces is $1.2 million and MARA's is $552,000. While these budgets have yet to be approved by GOT, the mission views them as a strong commitment to the completion of work outstanding in the project. The prospect of a follow-up project is under consideration by both the GOT and the World Banlc. Such a project would greatly benefit from the implementation experience and lessons to be learned of the present project in addressing land degradation in Anatolia. 7. Bank and Borrower Performance Bank 7.1 Lending: The Bank's performance in the identification, preparation assistance and appraisal/post appraisal of the project was satisfactory. The project was consistent with GOT and Bank development priorities and the Bank's country assistance strategy. Relationships with govermment were excellent overll during identification/preparation/appraisal and discussions were lively and professional centering on the most appropriate coordinating mechanism among the concerned agencies involved with the project implementation. As requested by the GOT the Bank was able to successfully arrange for Project Preparation Facility (PPF) funds necessary to prepare the six 1993 MC plans with assistance of TA and to finance study tours and the procurement of computers and office equipment. The technical approach proposed for watershed management and conservation was sound and a good deal of flexibility as well as a strong supervision program were worked into the project design. A post-appraisal mission was necessary to review, with MOF, MARA and KHGM, the implementation program for the first year of the project in order to avoid delays in project start-up and to agree on issues/actions to be discussed at negotiations including the organizational set-up at provincial and central level, sufficient budget allocations to meet GOT contribution and to agree on procurement arrangements. 7.2 Supervision: The Bank performance during supervision is considered to have been highly satisfactory. A total of 20 Bank supervision missions, including a MTR in October/November 1995, were carried out during the course of the project to review project progress at intervals of six months. Appropriate skill mix and staff continuity were always assured. Bank and project staff of MOF and MARA started to carr out a joint work for the first time and maintained throughout the life of the project a constructive role forming, at the - 15 -  same time a very harmonious team capable of taking advantage of the project flexibility when the project scope was extended to cover a total of eleven provinces. In general, Bank staff have been persistent, firm and clear in dialogue with the Government working to find constructive solutions to problems identified. The QAG noted that supervision was highly attentive to development issues and kept the project focus in the right places. Issues were well identified and their significance for the overall operation were presented in supervision reports, rating the project "Best Practice" both in terms of quality at entry and quality of supervision. The EAWRP was considered an example of the World Bank at its best. 7.3 Overall Bank peyfonnance: The overall performance of the Bank is considered highly satisfactory. As indicated above, from the design phase to implementation, the Bank staff developed a pioneer project Watershed Management using a participatory approach, which proved to be the keynote of the successful implementation of the project The increased role of the resident mission in supervision in Turkey is clearly a help in cushioning projects from changes at the headquarters level. Borrower 7.4 Preparation: The Borrower performance during preparation is rated satisfactory. All the concerned entities closely participated in and effectively contributed to Bank's work from identification to appraisal. The original intention at design stage was to include 17 provinces in the project. The GOT responded flexibly as, for logistical and other reasons, the number of provinces was reduced, first to ten and finally to three. The Govermnent's decision to initiate first project activities in six out of an intended estimated total of 54 MCs so as to avoid delay in the start of implementation, was very effective. 7.S Government implementation performance: The Government's performance during project implementation was mixed. While the GOT in many ways displayed strong commitment to the project, its complex financial procedures, combined with austerity measures, served to delay and reduce expenditure and so to retard implementation. All three implementing agencies were affected, KHGM most severely. Supervision missions also commented adversely on the standard of auditing of the project accounts. The Govermment was able to provide the necessary support to expand the project scope from the original three provinces as envisaged at appraisal to the final target of eleven. 7.6 Implementing Agency: The performance of the principal implementing agencies, MOF, MARA and KIHGM during the various stages of the project is rated satisfactory. The expansion of the project from 3 to 6 provinces in 1998 and from 6 to 11 provinces in 1999 was weUl taken care by the implementing agencies and, in particular, their effective capacity to coordinate their activities during the life of the project was highly appreciated by the WB. However, the number of women professional staff remained inadequate throughout the life of the project, a deficiency repeatedly noted at supervision. MARA especially would have benefited from having women staff in its extension activities. AGM performed well both as the lead agency and as the project coordinating body, once it assumed the overall responsibility for the PCSU. It also performed well at field level in the implementation of forestry activities. In these it had the advantage of firstly employing good dedicated technical staff, secondly operating on land free of serious pressure from viUlagers. MARA performed most effectively in its activities with private farmers and its capability in extension grew in the light of project experience. MARA's performance was less successful on rangeland, for which it lacked the technical and social expertise and an appropriate legal framework within which to operate. KHGM started one year later than - 16 -  the other two implementing agencies and this in turn adversely affected MARA's activities at the start of the project. Thereafter its performance was satisfactory although, with the expansion towards the end of the project, supervision mission missions noted the increasing cost of its activities and instances of inadequate awareness within KHGM of the supporting nature of its role. KHGM suffered from being more dependent than the other two implementing agencies on GOT finance. 7.7 Overall Borrowerperformance: In view of the above, the overall performance of the borrower is rated satisfactory. 8. Lessons Learned * A participatory project cannot be target-driven: The lesson for project design is that the design should focus on process rather than on physical targets, although, clearly, informed assumptions have to be made for the purpose of estimating costs and returns. The design should allow for an annual cycle of work starling from participatory MC planning. Problems should be solved with clients, not for them. One specific issue arising from this approach which will need to be addressed in any further project concerns the ratio between conservation and supporting income-generating expenditure. Early involvement of key stakeholders in project design is essential in order to ensure ownership and build commitment. * Major government ministries can collaborate effectively in delivering services asfield level: Although not unique, the experience of this project provides an infrequent and invaluable lesson in successful collaboration between three implementing agencies who had no previous history of working together. But the project showed also that key elements in establishing such cooperation (and that of villagers) are time and training and that flexibility is needed (for example in procurement procedures) to allow each agency to contribute to its fullest extent. Experience towards the end of the project, when the numbers of MCs and provinces were substantially increased, tends to support Mid-Term Review concerns regarding the risks associated with too fast a pace. * The project should operate in unambiguous legal conditions: There were legal uncertainties surrounding the rangeland, especially the non-forest rangeland, and an inadequate legal framework for MARA's participation in activities related to it. It was noted at appraisal that new rangeland legislation was in process but this legislation was approved only in 1998. The associated cadastral mapping will take many years. For any given piece of rangeland, the difficulties will remain until all uncertainties are resolved. The World Bank should therefore request the GOT to give priority in mapping to those provinces and areas where it is most needed for the implementation of World Bank-assisted activities, specifically the anticipated second watershed rehabilitation project. The design of this project should be realistic in any objectives relating to rangeland. * Design and implementation should build on existing local technology and capacity: The project's principal successes came from promoting and adapting existing local practice and the project found relatively little need for technical assistance, especially that from overseas. The project was implemented directly by local administrations, from the village up, and the project's sensitive employment of its farmer-centred problem-census problem-solving (FCPGPS) approach is, at its best, exemplary For any follow-up project. * A project of this kind needs social and extension skUls: In both design and implementation, the - 17 -  project exhibited technical strength but would have benefited from the provision of social and extension skills. In particular the need for extension training should be considered when designing further assistance to what is in large measure an extension project. * AU stakeholders need to be included: This an important lesson for any future project whose MCs include high altitude rangeland used by trans-humants. Although these grazers are likely to have their home bases distant from the project villages, it is essential that they be included in the FCPCPS. * Training should be timely and appropriate: In the present project training gathered momentum relatively slowly and the Mid-Term Review noted that, three years into the project there were staff who had not been trained and did not fully understand the project. Any follow-up project would benefit from more timely training. Subject areas in which training for project staff will be needed include procurement, disbursement, supervision and financial management. * The project design must ensure that the time allowedfor participatory planning and implementation is (a) sufficient and (b) likely to be efficiently utilised: Partly related to the question of social and extension expertise is the fact that the time (one year) allowed for MC selection and planning in some instances proved inadequate for the purpose of ensuring the necessary level of commitrnent by villagers. This problem was also related to coordination at field level. It is noteworthy that the Mid-Term Review recommended increasing the time allowed for implementation in any one MC from 3 years to 5 years and that the project desisted from selecting any new MCs in 1996, the year after the review. The pace accelerated considerably thereafter, with 28 MCs selected in a single year (1998). The design of the second project should address the issue of scale and timing. * The project design should be such as tofacilitate the inclusion of all necessary sources of expertise: The project design assumed, explicitly or otherwise, that technical or other support would be forthcoming not only from the identified implementing agencies but also from certain of their sister departments or from as yet unidentified sources such as research bodies. It proved difficult or impossible to bring into the project expertise not specifically identified and budgeted for at appraisal. The limited success on rangeland is attributed in part to a lack of research information. The lesson is therefore is that more explicit and/or more flexible budgeting is required. * Monitoring and evaluations should (a) be sustainable and (b) include data on outcomes: While monitoring and evaluation were ably addressed by the employment of consultancy expertise, its sustainability was not, and this should be an issue for the next project. Also important is the nature of the monitoring and exercise, which currently focuses only on inputs. It is important that it should also incorporate data and, ideally, specific and simple indicators, relevant in assessing project outcomes. Such data as exist are currently kept by individual implementing agencies rather than by the PCSU. * There were various technical lessons: Noteworthy among these are: (a) the technical and cost-saving innovations developed (e.g. in Malatya) in soil conservation and irrigation technology; (b) the difficulty of improving rangeland by means other than simple closure and protection; (c) the merit of upgrading (grafting) naturally occurring fruit trees (where available) on rainfed land, rather than planting new ones; and (d) the need in forestry to ensure good seedling quality: the follow-up project could usefully include a nursery component and address the issue of genetic - 18 -  quality also. 9. Partner Comments (a) Borr-ower/implementing agency: Comments from Undersecretariat of Treasury: * Publicity among beneficiaries and general public: An important activity that was missing until late years of the implementation has been the less than required level publicity of the project objectives, coverage and accomplishments among both the current and potential beneficiaries and also outsiders, including potential sponsors and other public and private agencies that are willing to plan, implement or fund similar projects. This is important for both being able to secure funds if required from such organizations and for demo effect utilization and sustainability purposes. * Project costing: Costs should be computed realistically so that actual disbursements and physical perfornance that could in fact be at reasonable rates shall not be perceived by an outsider as too low as compared to the loan disbursement ratios envisaged in PAD. * Projects of this kind should take into consideration the investment program. budgetarv constraints. if any. and internal procedures of the borrower: Since the Bank-financed project iunds are not considered as additional resources to be treated outside the budget according to the Turkish state accounting system, matching funds for the loan proceeds need to be provided besides the counterpart fundings and internal taxes. This requirement needs to be taken into account unless there is a change to the said system. This is not to be regarded as a problem or a temporary practice but as a long established part of the government procedures. At times of scarce funds, this can cause some difficulties but the reason we are carrying out projects financed partly by the World Bank is that the intemal funds are not sufficient to do them through internal budgets only, whereas the need for such projects will be there as long as there is a need to improve the life standarts of especially the people living under harsh geographic and financial conditions. There also is a need to improve project design, implementation and coordination capacity of the government agencies responsible of leading the activities in related sectors together with their ability to cooperate with inhabitants and non-governmental bodies both during and post implementation periods. Also the Bank-financed projects need to be designed such that they form a good blend with the government's investment program and procedures in terms of objectives, priorities, locations, means and timing of implementation. Such projects should not be seen as fancy and isolated projects created out of scratch in a country with no prior setting of rules and regulations and no procedural frameworks, but instead these all need to be studied and utilized during all phases of the project for the best outcomes. On the other hand, we should take into consideration the fact that Turkey suffered from an economic crisis in 2001 which had far-reaching effects on all sectors of the economy. During those times it becomes necessary to re-adjust the budgetary allocations for the implementing agencies and this was a difficult process since measures had to be taken for the whole economy. Nevertheless, the Turkish side achieved not to be taken over by the negative factors but instead reached successfully the objectives set at the beginning. Therefore, the performance of the - 19 -  implementing agencies should not be under-estimated. The best outcome of the project has been that it has improved greatly the vision and business approaches of both the beneficiaries and the agencies involved. This is an indicator of the fact that such activities designed and coordinated with a project concept have to be replicated wherever feasible in Turkey. A continuation of this effort, therefore, would be very useful if based on the lessons learned. The evaluation jointly made by MOF, MARA, KHGM and Treasury is given below: I . The objectives of the EAWRP are the rehabilitation of the degraded watersheds and raising of the incomes of the rural communities. Until the preparation of the project in 1992, a number of projects have been implemented separately and/or jointly by the agencies which are providing services to the rural areas. However, none of those projects approached the natural resource degradation and rural poverty holistically. Natural resource conservation and rehabilitation, and poverty reduction has become two indispensable elements of rural development projects in recent years. Today, the international donors are seeking this integrated approach in the proposed rural development and natural resource conservation projects. These raising values have been successfully integrated into the EAWP about a decade ago. 2. The EAWP, also adopted two concepts; participation of rual conmmunities and sustainability. These concepts were not only envisaged by the project but also achieved. 3. The Project started with an approach that was in harmony both with the Bank's and the Turkish Govermment's rural development policy and strategies. At the preparation and implementation period, it was a participatory project. At the implementation stage, it exhibited a flexible approach, open to revisions by maintaining the development objectives. It was also a project continuously supervised, monitored and periodically evaluated. Performance of the Bank 4. The Bank was successful in pursuing its strategies from the first contact and subsequently guided the client accordingly. The implementing agencies were the central and provincial agencies of the three Ministries namely; Ministry of Forestry, Ministry of Agriculture and Rural Affairs, and General Directorate of Rural Services. The project had a substantial risk by being implemented with a number of agencies which had no previous experience in working together in the same area and had a complex administrative structure. The Bank was successful in mitigating this risk by acting as a catalyst and motivating the project staff to work around the common goals. 5. The Bank helped the client to create a momentum and all through the project life made substantial efforts to maintain the gained momentum by emphasizing the importance of the approach of the Project for rural Turkey, and its potential as a model. 6. The Bank, provided a comfortable environment for the implementing agencies in terms of flow of funds during both preparation and implementation stages. 7. The Bank was also keen in supervising and monitoring the Project. After the task management responsibility was transferred to the Bank's field office, continuous supervision was provided. In addition, at least two times a year, formal supervision missions were conducted . Field supervision - 20 -  tours were made together in the sense that all agencies participated reviewing the others progress. Frequent joint meetings organized and hosted by the Country Office between formal supervision missions to discuss technical and as well as administrative problems further strengthened the team spirit. At the field superv/isions, half of the time was allocated to listening to and instigating dialogue between the MC community and the project staff about the Project approach and interventions. 8. The Bank gave special emphasis to the capacity building both for the technical staff and rural communities. The Bank encouraged the agencies to conduct technical training programs, field study tours, farner study tours (exchange of infornation between communities of the new and the old project MCs), lessons learned workshops. 9. Despite all the positive aspects underlined above, there were also some negative sides in working with the Bank. Particularly in the first years of the Project, frequent changes of task managers and team members, and insufficient understanding of the project approach on the mission members' side created some communication problem between the client and the Bank. The fBank's lengthy procedures and bureauciacy was another difficulty that was faced during the project life. 10. The field supervision of the project was found time consuming and physically exhaustive due to the: (i) wide geographical distribution of the project sites; (ii) MCs being in the remote areas; (iii) rough terrain; and (iv) a number of multi-disciplinary activities. Therefore, two-week field supervision missions need to be extended and more frequent. 11. The Bank's performance was rated satisfactory in terms of flow of funds and supervision. The agencies' confidence in the Bank was increased due to the Bank's overall approach, commitment, dynamism in supervising the project and determinism. Performance of the Governmenat 12. Preparation stage: The contribution of the agencies to the preparation of the project was limited mainly due to the following reasons: (i) integration of natural resource rehabilitation and rural poverty was a new approach to the agencies; (ii) the agencies involvement in the preparation was limited since the preparation was contracted to a consulting company; (iii) MOF had limited experience in working with local communities despite its technical expertise; (iv) MARA used to concentrate on low lands and resource rich areas and paid little attention to MCs/high altitudes and resource poor farmers; (v) KHGM had not been developing water resources less than 5 It/second and working in rough terrains and carrying out limited terrace construction and the agency's experience was limited in participation. Therefore, we rated the performance of the central and field level agencies as marginally satisfactory during the preparation period. 13. Implementation stage: State Planning Organization (SPO) has not been involved in supervision/monitoring of the project activities and therefore, failed to provide necessary budget allocations particularly for KHGM resulting in implementation delays in some years leading to weakened the confidence of the local communities in government agencies. 14. The World Bank Project's Department of Treasury, although paid particular attention to the project progress and visited several MCs during implementation, could not join the supervision teams for systematic field monitoring. - 21 -  15. Despite the difficulties and bottlenecks which affected the project implemnentation from time to time, the project has achieved its objectives. The monitoring data indicates this. Project Impact The following evaluation has been made jointly by the provincial agencies. 16. Social impacts: * Reduced tendency in rural out migration as a result of increased rural income. * Increased awareness about better living conditions and nutrition and willingness to apply. * Increased awareness about natural resource degradation and adoption of environmentally friendly production techniques. * Improved community-state relationship and community's voice in works performed. * Increased confidence in state. * Improved ownership of the investments and renewed attachment to the land. * Voluntary involvement in project activities. * Increased awareness in collaboration. * Demanding services from agencies without an intermediary. * Weakening of the opponent groups in rehabilitation and conservation of the natural resources and strengthening of the supporters. * Positive overall change in the traditional community. 17. Institutional impacts: * Increased tendency in providing better services as a result of strengthened institutional infrastructure. * Improved skills in using office equipment and improved performance in reporting, presentation and monitoring. * Improved inter-agency collaboration and improved understanding about services provided. * Improved vertical and horizontal communication within an agency. * Created awareness about monitoring of the performance of the other project agencies. * Better provision of services as a result of interactive, integrated and planned work. * Increased ownership of work done and created competition among agencies and project staff in providing quality services. * Increased professional satisfaction and willingness for self development. * Increased professional success and self confidence. * Reduced institutional chauvinism. * Change in approaching to local problems, increased tendency for decentralization. * Change in the way of managing tasks, adoption of cost efficient methods. * Putting more emphasis to the work done rather than the individuals and/or agencies. * Improved skills in maldng decisions at the local level. * Taking actions based on the technical and social priorities instead of political priorities. * Demanding budget allocations based on the planning of the work needed rather than creating work to get more allocations. * Adoption of holistic approach to the problem identification and solution finding. * Improved understanding of natural resource degradation, rehabilitation and rural poverty. * Better understanding of collaboration among agencies in addressing natural resource rehabilitation and rural poverty. - 22 -  * Change in approach to erosion control, better understanding of the soil degradation on agricultural land. * Conceptual agreement in addressing natural resource degradation, rehabilitation and conservation among agencies. 18. Enviromnental Impact: * Reduced risk in floods and land slides. * Reduced soil degradation and biodiversity loss. * Reduced pressure on forests and rangelands. * Increased landscape value. * Improved enviromnent for wildlife. * Improved access to clean and continuous water. * Increased natural resource productivity. * Increased fertility and productivity of farmland. 19. Economic Imrpact: * Increased employment. * Increased productivity of natural resources. * Increased crop and livestock productivity. * Increased income as a result of decreased production costs. * Increased land value. (b) Cofinanciers: (c) Other partners (NGOs/private sector): 10. Additional Information N.A. - 23 -  Annex 1. Key Performance Indicators/Log Frame Matrix Output Indicators Output Description Projected in Actual SAR (1993) Latest Estimate Unit 1. FOREST LAND (MOF) (i) Soil conservation afforestation (a) Soil conservation afforestation ha 10,000) (b) Conifer plantations ha 4,900) 71,991 (c) Fuelwood coppice plantations ha 11,800) Sub-total ha 26,700 71,991 (ii) Oak coppice rehabilitation ha 17,800 2,240 (iii) Rangeland rehabilitation ha 17,800 19,282 (iv) Riverbank protection (gallery plantations) ha 70 81 (v) Cedar rehabilitation ha Not in SAR 1,687 (vi) Participative plantations ha Not in SAR 35 Total ha 62,370 95,316 2. RANGELAND (MARA) (i) Rangeland management ha 58,650 1,320 (ii) Range management and fertiliser ha 30,500 234 (iii) Range management, fertiliser and seed ha 20,000 364 (iv) Pilot aerial fertiliser application ha 5,000 0 (v) Pilot aerial fertiliser and seed application ha 2,000 0 (vi) Demonstrations ha 162 464 Total ha 116,312 2,382 3. ARABLE LAND (MARA) , _ (i) Agronomic packages ha 11,667 7,886 (ii) Fallow reduction/forage production ha 25,960 20,800 (iii) Demonstrations ha 540 1,124 Total ha 38,167 29,810 4. SUPPORTING ACTIVITIES (MARA) (i) Rainfed horticulture and conservation ha 1.124) 5,554 (ii) Rainfed terrace agriculture ha 4,492) (iii) Rainfed gully horgiculture ha 3,246 1,642 (iv) Rainfed pistachio grufting ha 2;,000 6,000 (v) Rainfed pistachio establishmnent ha 1,000 1,149 Total rainfed ha 11,762 14,345 (vi) Irrigated horficultur and conservation ha 2,574 3,310 (vii) Irrigated forage ha 7,898 1,903 Total irrigated ha 10,472 5,213 Total (MARA) ha 22,334 19,558 (ix) Beekeeping (I kit= 20 beehives) kit 1,620 1,728 (x) Tree planting on field boundaries no. of trees 1,200,000 240,000 ha 341 200 6. SUPPORTING ACTIVITIES (KHGM) (xi) Small-scale irrigation with terrace construction ha) 10,530 1,203 without terrace construction ha) 11,165 Total small-scale irrigation 10,530 12,368 (xii) Rainfed terraces ha 5,616 I1,440 Total (KHGM) ha 16,146 13,808 (xiii) Riverbank protection Ian 0 40 - 24 -  Outcome/Impact Indicators Outcome/Impact Indicators Description Projected in SAR Actual 1) Capacity of local communities to Participating approach to strengthen Project activities have been undertaken manage their own natural resources. farmers' planning and implementing and implemented with the agreement capacity and participation of village people and _______________________________ _________________project beneficiaries. 2) Annual incremental income per US$ 525/family US$587/family. Based on the total family in each of the participating project incremental value of productive villages in the MC. (US$23,500,000) for 40,000 families. 3) Cost sharing arrangement for Individual and community share of the Farmers in the project area have watershed rehabilitation. rehabilitation efforts included labor for actively participated in the project establishment but mainly contributed activities and have mainly contributed in the form of O&M of facilities and with their own labor and cement for improvement created. irrigation facilities. In the case of Kuzucak, farmers formed a cooperative to hire a private company to seek technical and economic advice and set up an efficient cost sharing arrangement. 4) Income generating activities. GOT and beneficiaries to share the Completed as expected with strong investment cost while O&M cost measure of ownership by the primary would be covered fully by the stakeholders. beneficiaries. 5) Restoration of sustainable natural The reforestation component expected The project initiated innovating in the resources. to focus on indigenous species, in protection and rehabilitation of the particular native oak. The GEF environment while helping rural in-situgene conservation sub-project people to increase incomes. The GEIE expected to promote preservation of in situ conversation of genetic diversity wild forest and crop species. subproject was evaluated separately in the ICR dated April 30, 1999. 6) Microcatchments (no.) 54 87 7) Microcatchments area (ha) 400,000 520,000 8) Beneficiary families (no.) 38,333 40,000 9) Beekeeping farmers (no.) 80 86 10) Beehives (no.) 1,620 1,728 11) International TA (p/in) 97 15 12) Local TA (p/m) 39 234 13) Ponds (no.) .270 1,260 = 14) Irrigation canals (km) 594 1,069 - 25 -  Annex 2. Project Costs and Financing Project Cost by Component (in US$ milion equivalent) Appraisal Actual/Latest Percentage of Estimate Estimate Appraisal Project Cost By Component US$ million US$ million 1. Strengthening Agency Capacity 5.88 1.54 26 2. Watershed Rehabilitation' 58.66 43.74 75 3. Supporting Activities 22.82 32.84 144 4. Applied Research 0.80 0.21 26 Total Baseline Cost 88.16 78.33 Physical Contingencies 8.58 0.00 Price Contingencies 13.05 0.00 Total Project Costs 109.79 78.33 Total Financing Required 109.79 7833 Project Costs by Procurement Arran menents (Apprai al Estimate) (US$ million equiv lent) Expenditure Category Procurement Total Method 1/ IBC NCB Other NBF Cost 1. Civil Works 22.5 34.82/ 57.3 (13.4) (20.8) (34.2) 2. Plant and Equipment 8.3 0.9 9.2 (7.3) (0.8) (8.1) 3. Apiculture Kist 4.8 4.8 (2.4) (2.4) 4. Materials 19.9 19.9 (17.5) (17.5) 5. Vehicles 5.0 5.0 X (4.) (4.4) 6. Technical Assistance and Training 6.0 6.0 (6.0) (6.0) 7. PPF 0.8 0.8 .___________ ___________ (0.8) (0.8) 8. Incremental Operating Cost 27.3 6.9 6.9 .____________ (15.9) (3.4) (3.4) Sub-total 13.3 69.2 109.8 (11.7) (49.3) (77.0)+ GEF Sub-project 1.9 27.3 3.7 5.7 1.7 (15.9) (3.4) (5.1) Grand Total 15.2 72.9 115.5 (13.4) (52.7) (82.1) - 26 -  1/ Figures in parenthesis indicate amounts financed by IBRD and GET. 2/ Force account. Project Costs by Procurement Arran ements (Actua/Latest Estimate (US$ million e uivalent) Expendittire Category Procurement Total Method I/ IBC NCB Other NBF Cost ]. Civil Works 27.72 30.62/ 58.32 (14.34) (15.98) (30.32) 2. Plant and Equipment _ 3. Apiculture Kist 2.55 _ 2.55 (2.55) (2.55) 4. Materials 0.18 7.67 . 7.85 _______ _ (0.14) (6.76) (6.90) 5. Vehicles 3.78 3.78 (3.36) . (3.36) 6. Technical Assistance and Training 2.2 0.86 0.27 3.33 _ (2.2) (0.96) (0.27) (3.33) 7. PPF 0.44 0.08 0.52 (0.44) (0.08) (0.52) 8. Incremental Operating Cost 1.98 1.98 __________________________ _ _(0.99) (0.99) Sub-total 6.42 31.31 40.60 78.33 (6.00) (17.80) (24.08) (47.97) GEF Sub-project - - - Grand Total 6.42 31.31 40.60 78.33 (6.00) (17.89) (24.08) (47.97) Project Financing by Disbursement Categories (US$ Million EquIvalent) : - . .' rwritim.ril --> F*zlq itist __.we"a,tihl .__ _ _ __ _ _ _ _t F Civil Works 34.2 23.1 0.0 57.3 30.32 28.0 0.0 58.32 89 121.0 00 101.0 Goods 30.1 4.0 0.0 34.1 10.26 1.36 0.0 11.63 34 34.0 0 0 34.0 Agriculture Klts 2.4 2.4 0.0 4 8 2.55 0.00 0.0 2.55 106 0 0 0.0 53 0 echnical Assistan 6.0 0.0 0.0 6.0 3.33 0.00 0.0 3.33 55 0.0 0.0 55.0 Training I PPF 0 8 0 0 0 0 0.8 0.52 0.00 0.0 0.52 65 0.0 0.0 65.0 _~~~~~~~~~~~~~~~~~~~ _ Incremental 3.4 3.4 0 0 6.8 0.99 0.99 0 0 1.98 29 29.0 0 0 29.0 Operating Cost _ Sub-Total 76.9 32.9 0.0 109.8 48 0 30 4 0.0 78.3 378 184.0 0.0 337. - 27 -  I t
 Annex 3. Economic Costs and Benefits Quantifiable Benefits Annual incremental production benefits attributable to the project at full development are expected to include, as far as the activities implemented under MARA responsibility are concerned, some :3,500 tons pulses, 34,500 tons of mixed fruits, about 6,000 tons of pistachio, 9,200 tons of grapes, 3,500 tons of vegetables, 25,000 tons of vetch/sainfoin, 17,000 tons of fodder (dry matter) and about 700 tons of honey (Annex 7, Appendix A, Tables 8 to J15). Similarly, other project benefits that would derive at full development from the activities under MOF responsibility include 220,000 m3 of wood/year, 780 tons/year of leaves/branches, 2,600 m3/year of firewood and 10,800 tons/year of fodder (Annex 7, Appendix A, Tables 16 - 18). Farm Income The net annual incremental income is estimated, on average, at US$270,000 per MC taking into account the expected aggregated benefits from the various activities supported by the project. This is equivalent to about US$590/family, which is in line with SAR estimates. While the increases in the net value of production are generally expected to range from 28% (agronomic package) to 390% (rainfed horticulture), the gully horticulture and the irrigated horticulture activities are expected to generate, at full development, values that are tenfold the related without project levels. The beekeeping is to be considered as a new activity in the region. A kit of 20 beehives, representing the average package given to farmers, is currently generating about US$520/year. Details on expected net value of production at full development are given in Annex 7, Appendix A, Tables 8-15 for the activity under MARA responsibility and in Tables 16-18 for the soil conservation afforestation, oak coppice rehabilitation and rangeland rehabilitation respectively. Economic Re-Evaluation The economic analysis has been carried out for the project as a whole and for each of the activities indicated above. Actual investment costs, expressed in constant December 2001 prices, are given in Table 3. As the analysis has been carried out in US$, the official exchange rate of US$1 = TL1,330,000 has been used. Total investment costs have been included in the analysis and in estimating the specific ERR for each project activity, the cost of the supporting activities and demonstrations, estimated at US$10.2 million, has been allocated proportionally in relation to the respective area actually developed. In view of the impact of price reforms in Turkey, key distortions in the foreign exchange rate, wage rates and product prices are considered minimal. The Turkish Lira is freely convertible and there are no major trade restrictions on agricultural goods to deflect market prices widely from border economic values. The financial prices, being very close to the economic prices, were therefore considered appropriate for the economic analysis (Annex 7, Appendix A). The value of the incremental production of the project has been estimated taking into account the following output prices expressed in economic terms. - 29 -  Economic Prices Wheat 150.4 US$/ton Firewood 16.5 US$/m3 Commercial Timber 27.1 US$/m3 Leaves/Branches 23.0 US$/ton Fodder (dry matter) 75.0 US$/ton Fruits: mixed 0.37 US$/kg pistachio 1.35 US$/kg grapes 0.30 US$/kg Honey 2.25 US$/kg On the basis of the above, the ERR for the project as a whole has been reassessed at 16.8%. The ERRs for the various project activities under MARA range from 18.4 % for the agronomic package subcomponent to as high as 32.9% for gully horticulture. Similarly, activities under the implementation responsibility of the MOF, would generate ERR ranging from 9.5% for the oak coppice rehabilitation to 20.3% for the rangeland rehabilitation. Detailed calculations on the ERRs are given in Annex 7, Appendix A and are summarized in the table below. Results of the Economic Analysis Project Activities Economic Rate of Return (%) SAR ICR MARA Fallow Reduction n.a. 21.7 Agronomic Package n.a. 18.4 Rainfed Agriculture n.a. 23.5 Gully Horticulture n.a. 32.9 Pistachio Grafting n.a. 23.4 Pistachio Establishment n.a. 28.9 Irrigated Horticulture n.a. 25.0 Irrigated Forage n.a. 22.2 Small-Scale Irrigation n.a. 18.9 Apiculture n.a. 31.4 MOF Soil Conservation & Afforestation n.a. 11.4 Oak Coppice Rehabilitation n.a. 9.5 Rangeland Rehabilitation n.a. 20.3 Total Project 17 16.8 The above ERRs estimates indicate that the project as a whole, as well as the individual project activities, remain, at project completion, satisfactory, although the ERR for the soils conservation and afforestation and oak coppice rehabilitation may be considered marginal. However, as at appraisal, the analysis is still considered conservative since it does not take into account other benefits due to reduced run-off or resource - 30 -  conservation such as the likely increase in the economic life of dams in the project area through reduced ;edimentation. Furthermore, the analysis does not take into account likely falls in productivity -due to ,ieclining soils fertility in the absence of the project. .Sensitivity Analysis: At project completion, a meaningfil sensitivity test is carried out on expected future benefits only. If the project incremental benefits are reduced by 10% and 20%, the ERR drops to 15.1% and 13.40/, respectively, indicating the robustness of the project to this possible risk. - 31 -  Annex 4. Bank Inputs (a) Missions: Stage of Project Cycle No. of Persons and Specialty Performance Rating (e.g. 2 Economists, I FMS, etc.) Implementation Development Month/Year Count Specialty Progress Objective Identification/Preparatdon 03-04/1991 6 E, A, RN, WM, EC, F 11/1991 5 E, A, NA, NA, NA Appraisal/Negotiation 01-02/1992 6 E, A, NA, NA, NA NA 06-07/1992 7 E, A 11/1992 6 E, A 01/25-28/1993 Supervision 05/1993 3 E, A, F S S 11-12/1993 3 E, A, IE S S 05-06/1994 2 IE, A u s 10-11/1994 4 A, IE, E, BS U S 05/1995 4 A, E, F, BS S S 10-11/1995 (MTR) 4 F, A, IE, BS S S 05/1996 4 TM, A, BS, OP S s 10/1996 4 TM, A, IE, OP S S 05-06/1997 5 A, TM, IE, F, OP S S 07/1997 2 A, OP S S 11-12/1997 5 A, IE, F, OP, E S S 04/1998 4 A, IE, F, OP S S 07/1998 2 A, OP S S 11/1998 3 A, IE, F S S 05/1999 5 A, OP, IE, F, L S S 10/1999 4 A, OP, IE, F S S 05/2000 S S 11/2000 S S 04/2001 2 A, IE S S 09/2001 5 A, IE, IE, F, OP S S ICR 12/2001 2 E, NR S S (b) Staff: Stage of Project Cycle Actual/Latest Estimate No. Staff weeks US$ (000) Identification/Preparation Appraisal/Negotiation na 243,301 * Supervision na 581,000 ICR na 36,000 Total - 32 -  *From inception to Board presentation. - 33 -  Annex 5. Ratings for Achievement of Objectives/Outputs of Components (H=High, SU=Substantial, M=Modest, N=Negligible, NA=Not Applicable) Rating Oi Macro policies O H OSUOM ON * NA Oi Sector Policies O H *SUOM ON O NA Ol Physical O H *SUOM ON O NA OI Financial O H OSUOM ON O NA O Institutional Development 0 H * SU 0 M 0 N 0 NA D Environmental O H *SUOM O N O NA Social Oi Poverty Reduction O H *SUOM O N O NA Li Gender OH OSUOM ON ONA O Other (Please specify) O H OSUOM O N * NA El Private sector development 0 H O SU O M 0 N * NA L Public sector management 0 H O SU *M 0 N 0 NA L Other (Please specify) O H OSUOM O N * NA - 34 -  Annex 6. Ratings of Bank and Borrower Performance (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HU=Highly Unsatisfactory) 6.1 Bankperformance Rating O Lending OHS OS OU OHU U Supervision *HS OS O U O HU O Overall OHS Os OU OHU 6.2 Borrowerperformance Rating O Preparation OHS OS 0 U O HU O Government implementation performance O HS O S 0 U 0 HU I Implementation agency performance OHS OS OU O HU O] Overall OHS OS O u O HU - 35 -  Annex 7. List of Supporting Documents * Staff Appraisal Report, Report No. I 1294-TU, dated Februar 9, 1993 * Loan Agreement dated March 25, 1993 * Regional and Departmental Files * Mid-term Review * Information submitted by the Borrowers - 36 -     IMAG8NG Report No.: 24181 Type: ICR