E C O L O G I C A L E C O N O M I C S 6 8 ( 2 0 0 9 ) 1 4 9 6 – 1 5 0 6

ava i l ab l e a t www.sc i enced i rec t . com

www.e l sev i e r. com/ l oca te /eco l econ

ANALYSIS

The projected costs and benefits of water diversion from and tothe Sultan Marshes (Turkey)

Filiz Dadaser-Celika,⁎, Jay S. Cogginsb, Patrick L. Brezonikc, Heinz G. Stefanc

aWater Resources Science Program, University of Minnesota, St. Paul, MN 55108, USAbDepartment of Applied Economics, University of Minnesota, St. Paul, MN 55108, USAcDepartment of Civil Engineering, University of Minnesota, Minneapolis, MN 55455, USA

A R T I C L E D A T A

⁎ Corresponding author. Erciyes University, D+90352 437 57 84.

E-mail addresses: dada0004@umn.edu, fdad

0921-8009/$ – see front matter © 2008 Elsevidoi:10.1016/j.ecolecon.2008.10.012

A B S T R A C T

Article history:Received 13 February 2008Received in revised form17 October 2008Accepted 20 October 2008Available online 25 November 2008

The Sultan Marshes in the Develi Basin, Anatolia, one of twelve internationally importantwetlands of Turkey, have been severely affected by the construction of an irrigation projectin 1988. Intensive use of surface and ground water in irrigation has caused more than a 1 mdecline in water levels and has affected the wetlands' ecological characteristics. Previousstudies indicate that Sultan Marshes will need more water to restore viable ecologicalconditions. In this study, we analyze how economic benefits from agriculture and wetlandswould be affected if moderate amounts of water were diverted from agriculture back towetlands in the Develi Basin. By estimating total andmarginal costs and benefits associatedwith water diversions, we determined the optimum or economically-efficient amount ofwater diversion. When only direct-use values of the wetland (animal grazing, plantharvesting, and ecotourism) were included in the analysis, the optimum amount of waterdiversion to the wetlands was found to be 5.2 million m3 year−1 (165 L sec−1), whichcompares to about 62 million m3 year−1 (1,957 L sec−1) used in irrigation. When wastewatertreatment benefits (an indirect-use value) were added, the optimum amount rose to7millionm3 year−1. Overall, the analysis showed that water diversion fromagriculture to theSultan Marshes is economically preferable.

© 2008 Elsevier B.V. All rights reserved.

Keywords:AgricultureEconomicsSultan MarshesTurkeyWater diversionWetlands

1. Introduction

Wetlands are ecosystems with many hydrologic, biological,and ecological functions (Heimlich et al., 1998). Wetlandsrecharge/discharge ground water, maintain or improve waterquality by sediment and nutrient retention, reduce erosionand floods by water storage, and provide habitat for numerousplant and animal species. They help regulate local climate andplay important roles in carbon sequestration. Wetlands alsooffer products for human use and opportunities for natureobservation, hunting, and fishing.

ept. of Environmental En

aser@erciyes.edu.tr (F. D

er B.V. All rights reserved

Intensification of irrigated agriculture has negativelyaffected wetlands worldwide in recent decades (Gerakis andKalburtji, 1998; Lemly et al., 2000; Kingsford and Thomas,2004). Water available for wetlands has diminished as morewater is diverted for irrigation. Residues of fertilizers andpesticides used on agricultural lands have been transported aspollutants to wetlands. As a result, wetlands have becomesmaller or even completely dried out, and their chemical andbiological characteristics have been altered. The conflictbetween agricultural intensification and wetland conserva-tion has become more apparent as the value of wetlands has

gineering, 38039 Kayseri/Turkey. Tel.: +90352 437 49 01/32804; fax:

adaser-Celik).

.

1497E C O L O G I C A L E C O N O M I C S 6 8 ( 2 0 0 9 ) 1 4 9 6 – 1 5 0 6

been recognized by the public. In this study, we provide aneconomic analysis of water use by irrigated agriculture andthe Sultan Marshes wetlands in the Develi Basin, Anatolia,Turkey.

Irrigated agriculture has always existed in the semi-aridDeveli Basin but became more intense after the completion ofthe first phase of the “Develi Irrigation Project”. Since 1988,

Fig. 1 –Geographic setting of the Sultan Marshes (Örtülüakar MarBasin of Turkey (modified after Paşaoğlu (2004)). Components of tbasin surrounded by mountains.

almost all the surface water flows that formerly fed the SultanMarshes have been collected in irrigation reservoirs. Groundwater and flows from springs also have been used at anincreasing rate for irrigation.

The Sultan Marshes are located at the center of the DeveliBasin (Fig. 1) and consist of two salt-water lakes (Yay and Çöl)and two freshwatermarshes (Örtülüakar and Kepir) covering a

shes, Kepir Marshes, Yay Lake, and Çöl Lake) in the Develihe irrigation system are also shown. Develi Basin is a closed

1498 E C O L O G I C A L E C O N O M I C S 6 8 ( 2 0 0 9 ) 1 4 9 6 – 1 5 0 6

total of 17,200 ha. Themarshes were designated a Ramsar Site(according to the “Ramsar Convention on Wetlands of Inter-national Importance especially as Waterfowl Habitat”), a“Nature Conservation Area”, and recently a “National Park”by the Turkish Government. Despite their strong conservationstatus, the marshes have undergone rapid degradation inrecent years. Water levels have dropped significantly, and themarshes had several dry periods (Dadaser-Celik et al., 2006).Decreases in water levels, in turn, have affected the biologicalcharacteristics of the wetlands, particularly bird habitats(Özesmi et al., 1993). Previous studies have resulted in therecommendation that irrigation water use be reduced torestore the conditions in the Sultan Marshes. In this study,we evaluate possible economic outcomes of allowing waterfrom agriculture to flow into the wetlands of the Develi Basin.In this analysis, we compare the changes in economic benefitsfrom irrigation of agricultural lands to the changes ineconomic benefits from the wetlands.

There is a growing body of literature that evaluates theeconomic impacts of reductions in agricultural water use tosustain in-stream and wetland flow and prevent water qualitydegradation. Lee et al. (1997) estimated the economic impact ofa hypothetical 25% cut in surface irrigationwater supply to theSacramento Valley in California. Effects on agricultural pro-duction and local economies were included. They concludedthat a 25%water cutwould cause a 32millionUSD revenue lossfrom agriculture. Sundling et al. (1997) presented a methodo-logical framework to analyze the impact of water supplyreductions on agriculture. They developed three complemen-tary impact models and estimated the economic effects ofwater diversions to improve the water quality in the SanFrancisco Bay/Delta region of California. Barbier and Thomp-son (1998) and Barbier (2003) investigated the economicimpacts of water diversions and ground-water extraction foragriculture on wetland goods and services (floodplain agricul-ture, forestry, and fishing) in the Hadejia-Jama'are wetlands,Nigeria. Their analysis showed that wetlands can provideoverall economic benefits as large as those from agriculture.

This study addressed three questions. (1) Howwould watercuts affect agricultural income in the Develi Basin? (2) Howwould diversion of water to the Sultan Marshes affect theeconomic benefits obtained from the wetlands? (3) What

Table 1 – Crops cultivated in the region irrigated from Kovalı an

Crop Area(ha)

Average yield(kg ha-1)

Average inc(USD kg-

Cereals 3555 3750 0.25Legumes 104 1500 0.93Sugar beets 1481 47,500 0.08Sunflower 1595 1900 1.00Maize 388 42,500 0.06Plantations 365 0 0.00Vineyards 9 6500 0.37Orchards 2074 30,000 0.24Vegetables 77 32,500 0.32Potato 3 22,000 0.33Fodder 167 10,000 0.19Poplar 2 0 53.07Total 9819

would be the optimum (or economically-efficient) amount ofwater restoration from agriculture to the wetlands? Byanswering these questions, we intend to provide an economicperspective to the water allocation problem in the DeveliBasin. By comparing the value of water for different uses, thisanalysis can provide a foundation for better water allocationdecisions in the Develi Basin.

2. Agricultural production in the Develi Basin

The Develi Basin is near the city of Kayseri and covers an areaof about 100,000 ha (Özaslan and Şeftalici, 2002). Agriculture isone of the main economic activities in Kayseri and in theDeveli Basin. According to the 2000 census, 47% of the workingpopulation in Kayseri was employed in the agricultural sector(DPT, 2003). The area allocated to agriculture is approximately78,000 ha, and almost half of this land is irrigated. A significantportion of the irrigated land was put in production aftercompletion of the first phase of the Develi Irrigation Project in1988. The three irrigation reservoirs in the Develi Basin(Akköy, Ağcaaşar and Kovalı, with annual storage capacitiesof 7.5, 62, and 25 million m3, respectively) irrigate 28,000 ha offarmland. Another 9700 ha are irrigated by ground water(Özaslan and Şeftalici, 2002). A second phase of the DeveliIrrigation Project, incomplete as of date, includes plans totransfer 111–150 million m3 year−1 of Zamantı River waterthrough a 12-km channel from the Zamantı Basin to the southto irrigate an additional 34,000 ha in the Develi Basin. Theirrigation systems in the Develi Basin were designed as opencanal and gravity distribution systems using the floodirrigation method; therefore, irrigation efficiency is low.About 50% of the water used in irrigation is lost duringconveyance and application.

Areas allocated to different crops and the economic benefitsfrom agriculture (in 2005 USD) in the region irrigated by Kovalıand Ağcaaşar Reservoirs are listed in Table 1. For the totalcultivated area (approximately 10,000 ha), 36% was allocated tocereals, 25% to orchards, 16% to sunflower, and 15% to sugarbeets. The highest income per unit area (ha) was obtained fromvegetables, followed by potatoes, orchards, and sugar beets.Total agricultural income was about 29 million USD in 2005.

d Ağcaaşar reservoirs (data obtained from Ünlü (2005))

ome1)

Average income(million USD)

Average income per ha(USD ha-1)

3 9460.2 13926 38783 19041 25500 00.02 241215 72371 10,2520.02 73480.3 18560 1630

Table 2 – Use and non-use values of the Sultan Marshes

Use values Non-use values

Direct use Indirect use Option/quasi option

Plant harvesting Waste elimination/nutrient retention Potential use for additional ecotourism BiodiversityAnimal grazing Ground-water discharge Future value of information Culture/heritageRecreation/ecotourism Micro-climate regulation Bequest value

Carbon sequestration

1499E C O L O G I C A L E C O N O M I C S 6 8 ( 2 0 0 9 ) 1 4 9 6 – 1 5 0 6

Both surface and ground water are owned by the state inthe Develi Basin, but most of the irrigation schemes aremanaged by farmers through irrigation associations or irriga-tion cooperatives. Historically, water requirements of theSultan Marshes have been overlooked, and almost all surfacewater resources are used in irrigation. An analysis of irrigationmanagement in the Develi Basin showed that a basin-widewater planning system that considers water requirements ofthe Sultan Marshes is necessary for the sustainability ofirrigation and the wetlands in the Develi Basin (Dadaser-Celiket al., 2008).

3. Economic value of the Sultan Marshes

Wetlands provide both use (direct, indirect, option, and quasi-option) values and non-use (existence) values (Barbier et al.,1997). In Table 2, direct and indirect use values, and non-usevalues of the Sultan Marshes are listed.

The direct use value of the Sultan Marshes is derived fromplant harvesting, animal grazing, and ecotourism/recreationaluses. Reeds (Phragmites australis (Cav.) Trin. ex Steud.) andother plants are harvested in the Sultan Marshes by localpeople for commercial and domestic uses. Reed harvesting is aparticularly important activity in the villages and municipa-lities adjacent to the Sultan Marshes (Dadaser and Özesmi,2002). Harvested reeds are exported as thatch or insulationmaterial and used locally for thatching houses and animalbarns, as well as for fodder production. Local people alsoharvest cattails (Typha angustifolia) for weaving mats orbaskets, bulrushes (Scirpus validus) for making traditionalcushions, and salt cedar (Tamarix sp.) as an energy source(Karabaşa, 2001). A reed harvesting management plan wasprepared to regulate reed harvesting (e.g., timing and location)for ecological and economic sustainability (Özesmi, 2002).

Animal grazing takes place on pasture areas surroundingthe water bodies in the wetland. In 2002, around 9,000 cattle(cows and water buffalo) used the pastures around the SultanMarshes (Şarkışla, 2002).

With a wide range of habitats, the Sultan Marshes are animportant breeding and wintering site for 301 bird species.Moreover, they are located 50 km away from Cappadocia (inTurkish, Kapadokya), and close to Erciyes Mountain, whichhas become a popular winter sport center. Because of thesecharacteristics, the Sultan Marshes are potentially importantfor national and international tourism.

The indirect use value of the Sultan Marshes is derivedfrom current wetland services, including waste eliminationand nutrient retention, micro-climate regulation, and carbon

sequestration, and potential uses in the future. The SultanMarshes have received and assimilated untreated (raw)domestic and industrial wastewater (from municipalitiesDeveli, Yahyalı and Yeşilhisar and many villages within theirjurisdiction) and agricultural drainage for many years. Waste-water treatment facilities do not exist in the Develi Basin fortreatment of domestic wastewater. Although the efficiency oftreatment processes taking place in the marshes has not yetbeen explored, it is known that wetlands can remove organicmatter, suspended solids, metals and nutrients from waste-water through physical (sedimentation, filtration, adsorption),chemical (precipitation, adsorption, decomposition) and bio-logical processes (microbial and plant metabolism, plantadsorption, natural dieoff) (Hammer and Bastian, 1989;Watson et al., 1989). Emergent (e.g., cattail, bulrush, reeds),floating (e.g., water hyacinth), and submerged (e.g., water-milfoil) wetland vegetation, which are present in the SultanMarshes, can play important roles in the treatment processes(Reed et al., 1995). Through waste storage and treatmentfunctions, the Sultan Marshes contribute to a variety of otherservices in the Develi Basin. The Sultan Marshes reduce thecontamination of the groundwater, which is used for drinkingand irrigation purposes. The marshes reduce the cost ofwastewater treatment and disposal by providing a sink forwastewater and preventing the distribution of many diseasesthat could develop due to improper disposal of wastewater.

The SultanMarshes also transfer large amounts of water tothe atmosphere through evapotranspiration, which may havea significant impact on the local climate. Wetlands are sinksfor carbon andmay have a role in global carbon sequestration,but this value of the Sultan Marshes is yet unexplored.

The Sultan Marshes also have option and quasi-optionvalues because of their potential uses in the future. Someservices that the wetlands could provide (e.g., additionalecotourism) are not used at present. Information for theassessment of future uses of the Sultan Marshes is currentlyinsufficient for accurate measurement.

Non-use or existence values of the Sultan Marshes arederived from biodiversity, cultural and heritage, and bequestvalues. The Marshes sustain a rich biodiversity consisting of25 species of mammals, 25 species of mollusks, 40 species ofhymnoptera, five species of fish, 301 species of birds, 125species of algae and 401 species of plants (Karadeniz, 2000).Many civilizations have lived around and used the SultanMarshes since the early ages of mankind; the Sultan Marshestherefore occupy an important place in the region's historyand culture. Finally, the Sultan Marshes have a bequest valueas important ecosystems because of their use and non-usevalues that should be preserved for the future generations.

1500 E C O L O G I C A L E C O N O M I C S 6 8 ( 2 0 0 9 ) 1 4 9 6 – 1 5 0 6

4. Methods

This study was conducted in two stages. First a hydrologicanalysis was performed to identify the amount of waterrequired to restore water levels in the Sultan Marshes. Then,an economic model was constructed to analyze the economicimpact of water diversions on agricultural production andwetland uses.

4.1. Hydrologic analysis

A deterministic, process-based hydrologic simulation modelof the Örtülüakar Marsh (a subsystem of the Sultan Marshes)(Dadaser-Celik et al., 2006) was used for the hydrologicanalysis. We assumed that Örtülüakar Marsh can representthe whole system because all subsystems of the SultanMarshes are interconnected and Örtülüakar Marsh is locatedon the upstream end of the Sultan Marshes. The model used amonthly timescale and covered a timeframe of 7 years (1997–2003). It simulated the changes in water levels of theÖrtülüakar Marshes based on monthly water inputs (i.e.,precipitation, overland flow, flows from springs, net ground-water flow, and irrigation return flow) and outputs (i.e.,evapotranspiration and surface flow to the Yay Lake). Theanalysis showed that ground water constituted about 50% ofthe water inflows to the Örtülüakar Marsh during the 1997–2003 period, 22–37% of the water inflow was provided bysprings, and 10–27% by precipitation. There was almost nosurface water inflow to the Sultan Marshes during this periodexcept for irrigation return, which was only 2–3% of waterinflows. Almost all water losses from the Örtülüakar Marshwere through evapotranspiration. More information about thehydrologic characteristics of the Sultan Marshes and theÖrtülüakar Marsh hydrologic model is provided by Dadaser-Celik et al. (2007) and Dadaser-Celik et al. (2006).

Fig. 2 –Water levels in the Örtülüakar Marsh with variousamounts of water diversion (m3 day-1). “Base simulation"refers to the model simulated water levels under existingconditions. “Minimum levels required” are the levels recom-mended by Ministry of Environment (2002).

Simulations weremadewithwater diversions to the SultanMarshes of 10,000–90,000 m3 day−1 for 5 months (May toSeptember) every year. These amounts represent 1.5–13.5 mil-lion m3 year−1, respectively, compared to 62 million m3 year−1

of water used for irrigation. Because water diversions con-stitute only a small portion of irrigation water uses in theDeveli Basin and the contribution to the marsh from irrigationreturn flow is not significant, we ignored the changes in returnflows under water diversion scenarios. The May–Septemberperiod was selected because it coincides with the irrigationseason in the Develi Basin. Minimum water levels in theÖrtülüakar Marsh recommended by the Ministry of Environ-ment of Turkey (2002) were used as targets.

Fig. 2 shows the base water levels observed in theÖrtülüakar Marsh and minimum water levels required forthe 1997–2003 period. Base water levels were obtained usingthe hydrologic model and they represent the water levelsunder current conditions (i.e., the conditions without waterdiversion from agriculture to the wetlands). Minimum waterlevels required are target water levels that have to beprotected for the sustainability of the marshes. Intra-annualfluctuations observed in base and target water levels are dueto the seasonal variations inwater inputs and outputs, and aretypical for wetlands. In Fig. 2, a multi-year decline (about 1 m)in base water levels is also observed. This decline is due to thedisruption of natural water regime of themarshes by diversionof surface and ground-water resources to agriculture.

Fig. 2 also shows the water levels simulated for differentwater diversion options (i.e., addition of 1.5 to 13.5 million m3

year−1 water to the marshes). Simulation results showed thataddition of 1.5 million m3 year−1 to the wetlands would beenough to obtain minimum water levels in the Sultan Marsh.For the complete restoration of water levels (i.e., to obtainhistorical (unaltered) water levels), 7.5 million m3 year−1 ofwater would be sufficient.

4.2. Economic model description

The economic model was developed using the softwarepackage STELLA. In STELLA, the model is constructed byvarious blocks categorized as stocks, flows and converters.Stock variables show quantities that accumulate in thesystem. Flow variables describe processes that change stockvariables. Converter variables are used for transformationsand catalytic effects. The connectors represent the cause andeffects within the model structure. For example, in a hydro-logic system, the volume of a water body is modeled as a“stock”, while precipitation is modeled as a “flow”.

The economic model had two sub-models (Dadaser-Celik,2008). A wetland sub-model simulated the increase in wetlandbenefits with increase in water diversion; an agriculture sub-model simulated the increase in agricultural costs with anincrease in water diversion. Economic values were expressedin 2005 USDs; the exchange rate used was 1 USD=1.35 YTL (inTurkish, Yeni Türk Lirası–New Turkish Lira).

4.2.1. Wetland sub-modelOnly direct-use values of the Sultan Marshes were consideredin the wetland sub-model because of insufficiency or unavail-ability of data regarding other wetland values. Direct-use

Table 3 – Costs and benefits of reed harvesting in theDeveli Basin

Capital requirements (USD year-1) 14,285Gross value (USD year-1) 285,690Variable costs (USD year-1) 0Gross margin (USD year-1) 271,406Fixed overhead costs (USD year-1) 129,275Depreciation (USD year-1) 714Annual net income (USD year-1) 141,417Number of bundles harvested 413,870Net benefit per bundle (USD bundle-1) 0.34

1501E C O L O G I C A L E C O N O M I C S 6 8 ( 2 0 0 9 ) 1 4 9 6 – 1 5 0 6

values are important because they represent the amount ofdirect cash flow. Includedwere changes in (a) grazing benefits,(b) reed extraction benefits, and (c) ecotourism income. Weassumed that water supply to the marsh causes independentchanges in the benefits obtained from the marshes.

4.2.1.1. Change in grazing benefits. The substitute costapproach was used to estimate the change in grazing benefitsfrom the Sultan Marshes. It was assumed that water restora-tion to the wetland would affect pasture productivity, but notpasture area. In Central Anatolia, where the Sultan Marshesare located, pastures on average produce 450 kg dry matterha−1 year−1 (DPT, 2001). The productivity of pastures aroundthe Sultan Marshes was assumed to be lower than this valuebecause of drought and overgrazing (300 kg dry matter ha−1

year−1). A direct relationship can be established betweenground-water levels and plant density on pastures (Liu et al.,2005). We assumed that pasture productivity would increaselinearly up to 500 kg dry matter ha−1 year−1 as water diversionto the wetlands (and hence the groundwater level) increases.The animal carrying capacity of the pastures was calculatedbased on the average dry matter requirement of animals(Eqs. (1) and (2)).

CCC =PA⁎CPPAFR

ð1Þ

where CCC is current pasture carrying capacity (animals), PA ispasture area (ha), CPP is current pasture productivity (300 kgdry fodder ha−1 year−1), and AFR is animal feed requirement(12.5 kg dry matter day−1 (Anonymous, 1998) multiplied by365 days year−1).

ECC =PA⁎EPPAFR

ð2Þ

where ECC is estimated pasture carrying capacity (animals)after a given level of water diversion and EPP is estimatedpasture productivity (kg dry matter ha−1 year−1).

The difference between current and estimated carryingcapacity was multiplied by the amount of alternative feedrequired per animal (AFR; kg dry matter ha−1 year−1), and thecost of the alternative feed (CAF; USD kg−1) to find the changein grazing benefits (GB) (Eq. (3)). During non-grazing periods,animals are fed with a variety of other feed such as cultivatedforages (e.g., clover, maize silage, trefoil, fig, hay), concen-trates, cereal and grain legume straw, agro-industry by-products, and residues in the Develi Basin. In this study, weassumed that alternative feed cost 0.15 USD kg−1 in 2005. Thisnumber was reported during interviews with farmers in theDeveli Basin and corresponds to the cost of “hay” (Dadaser-Celik, unpublished field data, 2006).

GB = ECC� CRCð Þ⁎AFR⁎CAF ð3Þ

4.2.1.2. Change in reed extraction benefits. The market pricemethod was used to estimate the changes in reed extractionbenefits (Mmopelwa, 2006). We assumed that reed area wouldexpand in relation to the amount of water diverted to thesystem, but average reed density (128 bundles ha−1) wouldremain the same. Because the economic benefits fromharvesting plants other than exported reeds are not signifi-

cant (Karabaşa, 2001), we considered only the benefits fromreed harvesting.

Data on harvested reed quantities and priceswere obtainedfrom a 2001 study conducted by the General Directorate ofNational Parks and Game Wildlife in Turkey (Özesmi, 2002;Şarkışla, 2002). No information was available, however, aboutcosts associated with reed harvesting; therefore, we madeseveral assumptions regarding costs. First, we assumed thatthe capital requirement for the harvesting equipment is 5% ofthe gross income. This assumption is reasonable because reedis harvested with sickles and carried to the shore with sharedwooden boats (locally produced). Althoughwe do not have theexact prices for this equipment, it is most likely that they costless than 5% of gross income. Second, we assumed that theopportunity cost of labor (the income that could have beengenerated if the labor had been employed elsewhere) was86 YTLmonth−1 (71 USDmonth−1) in 2001. This is the statutoryminimum wage payable to laborers in small businesses(Anonymous, 2004). Third, we assumed that one personharvests 75 bundles per day. The number of bundles thatcan be harvested in a day depends on the density of the reedsin an area and ranges between 25 and 150 bundles (Karabaşa,2001). Finally, we assumed that cost of maintenance is 5% ofthe capital requirement (Mmopelwa, 2006). Results of thecalculations are in Table 3. Variable cost in Table 3 refers to thetransportation cost, but the transportation cost is zero becauselocal businesses buy reeds at the shore and provide their owntransportation. Fixed overhead costs include labor and main-tenance costs and a fee of 0.025 YTL (0.02 USD) per bundle paidby reed harvesters to the General Directorate of National Parksand Game Wildlife. All values were converted to 2005 YTLs(USDs) using consumer price indices in 2001 and 2005. Thechange in reed extraction benefits was calculated with Eqs. (4)and (5).

CBE = ERA� CRAð Þ⁎RD ð4Þ

where CBE is the change in the number of bundles extracted,ERA is the estimated reed area (ha), CRA the current reed area(ha), and RD is density of reeds (bundles ha−1).

REB = CBE⁎RI ð5Þ

where REB is the change in reed extraction benefits and RI isthe income obtained per bundle of reed (USD bundle−1).

4.2.1.3. Change in ecotourism benefits. Change in ecotour-ism benefits was projected by estimating the change innumber of tourists visiting the site. Bird watching is the

Table 4 – Construction and operation costs of preliminary, primary and secondary treatment plants

Treatment option Construction cost(million USD)

Annual O&M costs(million USD year-1)

Annual investment requirement(million USD year-1)

1. Preliminary treatment 2.97 0.11 0.522. Primary treatment 5.47 0.24 1.023. Secondary treatment 11.01 0.45 2.01

Table 5 – Crop yield response factor (ky) values for cropscultivated in the Develi Basin

Crop Ky Location Reference

Wheat 0.32 Erzurum, Turkey Kırda et al. (1999)Legumes 1.04 Trakya, Turkey Erdem et al. (2006)Maize 1.02 Eskişehir, Turkey Öğretir (1993)Sunflower 0.91 Kırklareli, Turkey Karaata (1991)Sugar beets 1.27 Kırklareli, Turkey Yakan and Sabuncuoğlu (1991)

1502 E C O L O G I C A L E C O N O M I C S 6 8 ( 2 0 0 9 ) 1 4 9 6 – 1 5 0 6

most important activity for the tourists. Therefore, weestablished a relationship between number of tourists visitingthe site and number of birds present. We defined an index,called “habitat suitability index or HSI”, to explain the degreeof suitability of the marsh for birds. This index was calculatedby taking the ratio of the amount of water diversion in a year(WD in terms of million m3 year−1) to optimum amount ofwater diversion (OWD or 7.5 million m3 year−1)(Eq. (6)).

HSI =WD=OWD ð6Þ

In our analysis, the greater flamingo (Phoenicopterus rubber)was takenas thekeyspecies, because greater flamingonumberswere found to be correlated with water levels in other locations(Cezilly et al., 1995) and in the Sultan Marshes (Gürpınar, 1994).We assumed a direct relationship between habitat suitabilityindex and number of flamingos observed in the site. Maximumand minimum number of flamingos observed at the site were70,000 and 20, respectively (Gürpınar, 1994).

We also assumed that the number of tourists visiting thesite was related to the number of flamingos observed. Wecould not find any data related to the number of touristsvisiting the SultanMarshes in a year. Therefore, wemade veryconservative assumptions about tourist numbers. The num-ber of tourists visiting the site currently (CNT) was assumed tobe 1000 per year. This is about 3% of the international touristswho visited Kayseri in a year between 1996 and 2001 (Özaslanand Şeftalici, 2002). We also assumed that this number wouldrise up to a maximum of 5000 with water diversion (estimatednumber of tourists or ENT) (Eq. (7)). No data were availablerelated to the economic benefits obtained from ecotourism (TIin terms of USD person−1 year−1) in the Sultan Marshes. Aliterature review showed that the value of bird watchingactivity in the U.S. was 29.6 (±8.35) USD cap−1 day−1 with arange of 5.8–78.5 USD person−1 day−1 (Loomis, 2005). Anotherstudy conducted in Turkey (Pak and Türker, 2006) estimatedthe value of a forest recreation site in 22.1 USD person−1

visit−1. In this study, we assumed that the economic incomeper tourist is 20 USD.

TB = ENT � CNTð Þ⁎TI ð7Þ

4.2.1.4. Indirect-use benefits: waste water treatment. Weadded wastewater treatment benefits to the direct-use benefitmodel. Without data related to the degree of wastewatertreatment taking place in the marshes, we considered threetreatment levels corresponding to preliminary (i.e., screening),primary, and secondary treatment in wastewater treatmentplants. We assumed that as water diversion to the marshesincreases, the treatment level in the marsh will also increase.Thisassumption isbasedon the fact that vegetationdensity andarea in the marsh is directly related to amount of water

diversion. Vegetation density and area improves treatmentefficiency by increasing water residence times and potentialsurfaces for nutrient intake (Kjellin et al., 2007). As shown inmanystudies, reedsare responsible for the treatmentprocesses.

The substitute cost approach was used to estimate waste-water treatment benefits for various treatment levels. Waste-water treatment costs were obtained using the relationshipsestablished by Öztürk et al. (2005) among wastewater quan-tities, wastewater treatment levels, and construction/opera-tion costs of wastewater treatment plants for Turkey. Annualinvestment requirement was estimated by adding annualizedconstruction costs (calculated assuming that a wastewatertreatment plant operates for 30 years) and operation/main-tenance costs (Table 4).

4.2.2. Agriculture sub-modelThe agriculture sub-model estimates the losses in agriculturalproduction based on the amount of water allocated to thewetlands instead of agriculture. We assumed that farmers firstdivert water from low value crops. Based on this assumption,decreases in water for irrigation are projected to bemade in thefollowing order: wheat, legumes, sunflower, maize, and sugarbeets. For example, all wheat is abandoned before any produc-tion of legumes is abandoned. To represent the decrease in cropyield with respect to water diversion, we used the crop yieldresponse factor (ky) defined by Stewart et al. (1977).

YYm

= 1� ky 1� ETETm

� �ð8Þ

In Eq. (8),Ym ismaximumcropyield,Y is cropyieldwithdeficitirrigation, ETm is maximum evapotranspiration, and ET is theevapotranspiration under deficit irrigation. According to Eq. (8),relative crop yield is a function of relative evapotranspirationdeficiency.Thecropyield response factor (ky) variesdependingoncrop species, variety, irrigation method, and growth stage whendeficit evapotranspiration occurs (Kırda, 2000). We obtained kyvalues from the literature (Table 5) and assumed that theevapotranspiration deficit occurred all season.

Potential evapotranspiration (ETm) values and evapotran-spiration values that have to be met by irrigation wereprovided by DSI (DSI, 2007) for the Develi Basin. DSI (2007)

Table 6 – Total and marginal costs and benefits associated with water diversions

Waterdiversion(million m3)

Ecotourism(million

USD year-1)

Grazing(million

USD year-1)

Reedharvesting(million

USD year-1)

Agricultural costs(million

USD year-1)

Wetland benefits(million

USD year-1)

Marginal costs(million

USD year-1)

Marginal benefits(million

USD year-1)

0 0.000 0.000 0.000 0.000 0.000 0.000 0.0001 0.011 0.043 0.010 0.041 0.064 0.041 0.0642 0.021 0.086 0.020 0.082 0.128 0.041 0.0643 0.032 0.129 0.030 0.124 0.191 0.042 0.0644 0.043 0.172 0.041 0.171 0.255 0.046 0.0645 0.053 0.215 0.051 0.217 0.319 0.046 0.0646 0.064 0.258 0.061 0.354 0.383 0.137 0.0647 0.075 0.301 0.071 0.609 0.447 0.255 0.064

The benefits from ecotourism, grazing, and reed harvesting which were used to calculate total benefits were also presented.

1503E C O L O G I C A L E C O N O M I C S 6 8 ( 2 0 0 9 ) 1 4 9 6 – 1 5 0 6

estimated ETm values using the Braney-Criddle method andcalculated irrigation water requirements based on averageeffective rainfall values. Using cereals as an example, thewater savings and economic losses due to deficit irrigationwere calculated as follows.

Water saving (WS) denotes the difference between theamount of irrigation water used by cereals at present (IWU)and the amount that would be used under deficit irrigation(IWUD) divided by irrigation efficiency (IE) (Kırda, 2000):

WS = IWU� IWUDð Þ=IE ð9Þ

In Turkey field application efficiency for flood irrigation isaround 50% (Van Tuijl, 1993).

With the assumption that changes in water use would notaffect production costs, yield loss (YL) and economic loss (EL)were calculated from Eqs. (10) and (11).

YL = Y � Y⁎1� ky⁎ 1� IWUD + PPTð Þ½ �

IWU + PPT

� �� �⁎A ð10Þ

where Y refers to current yield (kg ha−1) and A refers to theareal coverage of cereals (ha).

CEL =YL⁎CI ð11Þ

where CI refers to crop income (USD kg−1).Total economic losses from agriculture and total economic

benefits fromwetlands for various amounts of water diversionwere obtained using the economic models. Marginal (orincremental) costs andmarginal benefits thenwere calculated

Fig. 3 –Total costs and benefits associated with waterdiversions.

by using total costs and benefits. Marginal costs (benefits) canbe defined as the cost of (benefit from) diverting 1 unit (m3) ofwater from agriculture (to the wetlands). We accepted thatoptimum (or economically-efficient) amount of water diver-sion occurred when marginal costs became equal to themarginal benefits.

5. Results

5.1. Total and marginal benefits and costs

Total costs and benefits and marginal costs and benefits ofwater diversions are given in Table 6 and plots of total andmarginal costs and benefits for various amounts of waterdiversions are presented in Figs. 3 and 4. Total benefits werehigher than total costs until 6 million m3of water year−1 werediverted. Above 6 million m3 of water year−1, total costsbecame higher than total benefits.

Total benefits include the benefits from ecotourism,grazing and reed harvesting. The increases in grazing benefitswere always the largest, followed by ecotourism and reedharvesting (Table 6). Grazing benefits constituted 57% of thetotal benefits, while ecotourism and reed harvesting provided17% and 16% of the total benefits, respectively.

Fig. 4 –Marginal costs and benefits associated with waterdiversions.

Fig. 5 –Sensitivity of the model to changes in (a) grazing benefits (b) ecotourism benefits (c) reed cutting benefits (d) total costs.

Fig. 6 –Simulation results with inclusion of wastewatertreatment benefits.

1504 E C O L O G I C A L E C O N O M I C S 6 8 ( 2 0 0 9 ) 1 4 9 6 – 1 5 0 6

Marginal costs were constant until 5 million m3year−1 ofwater were diverted (Table 6 and Fig. 4). After 6 million m3

year−1 of water diversion, marginal costs showed a rapidincrease. Marginal benefits, however, were constant for allamounts of water diversion.

Marginal benefits were equal to the marginal costs at5.19 million m3year−1 of water diversion. This amount isalmost equal to the amount of irrigation water used for cerealand legume production. This result suggests that water cutsfrom other crops (maize, sunflower, sugar beets) are noteconomically feasible. Here we have to emphasize that thebenefits from wetlands were estimated based on veryconservative assumptions and included only direct-use ben-efits. The real benefits from the wetlands are most probablymuch higher than values calculated here.

5.2. Sensitivity analysis

A sensitivity analysis was conducted by sequentially changingecotourism, grazing and reed harvesting benefits and totalcost by 20%. Fig. 5 presents how total benefits and total costswould change for various amounts of water diversions. Theresults showed that the system was not sensitive to theincreases and decreases in ecotourism and reed harvestingbenefits. Increasing (or decreasing) ecotourism and reedharvesting benefits by 20%, increased (or decreased) total

benefits by only 3%. Optimum water diversion (wheremarginal benefits equal to the marginal costs) moved from5.19 million to 5.22 (5.17) million m3 with 20% increase (20%decrease) in both of them. Animal grazing benefits had moreeffect on the system compared to the other components.Changing animal grazing benefits by 20% caused a 14% changein total benefits. Optimum water diversion was 5.29 (5.10)

1505E C O L O G I C A L E C O N O M I C S 6 8 ( 2 0 0 9 ) 1 4 9 6 – 1 5 0 6

million m3 with 20% increase (decrease). Increasing (decreas-ing) total costs by 20%moved optimumwater diversion to 5.08(5.37) million m3.

We analyzed the effect of uncertainty in the inputparameters of the grazing benefits sub-model using MonteCarlo simulation. The change in grazing benefits was deter-mined as the difference between current and estimatedgrazing capacities, which were defined by the grazing produc-tivity function. In the Monte Carlo simulation we randomlyselected values of two parameters from a feasible range andcalculated the change in grazing benefits for a constantamount of water diversion (5 million m3 year−1). This processwas repeated 10,000 times. The parameters were 1) pastureproductivity at current conditions (or before water diversion)and 2) rate of increase in pasture productivity with waterdiversion. Because specific ranges for these parameters werenot available from the literature, we accepted that theparameters varied by ±20%. Results showed that grazingbenefits ranged from 0.17 to 0.26 USD year−1 for 5 million m3

year−1 ofwater diversion. Themeanvaluewas 0.22USDyear−1.The probability of obtaining a value for grazing benefitssmaller than or equal to 0.22 USD year−1 (model estimatedvalue) was 54%.

5.3. Wastewater treatment benefits

Costs associated with preliminary, primary and secondarywastewater treatment are provided in Table 4. Fig. 6 shows theresults of the economic analysis with different levels ofwastewater treatment benefits included (i.e., a plot ofmarginal costs and benefits vs. various amounts of waterdiversions). Results show that if the lowest (preliminary)treatment benefits are included in the analysis, the optimumwater diversion would be 5.96millionm3 year−1. With primarytreatment, optimumwater diversionwould rise to 6.53millionm3 year−1. With secondary treatment, optimum water diver-sion would be higher than 7 million m3 year−1. These resultsshow that inclusion of indirect and non-use benefits from theSultan Marshes change the results significantly and make thediversion of more water to the wetlands not only feasible, butdesirable.

6. Summary and conclusions

In this study, we calculated the total and marginal costs andbenefits of allocating a fixed flow of water to the wetlands inthe Develi Basin. This water is currently used for irrigation–butnot very efficiently. The economic losses in agriculturalproduction due to the reduced water allocation have beendetermined and included in the cost/benefit analysis. Ouranalysis shows that an economically-efficient amount ofwater allocation is 5.2 million m3 year−1 when direct benefitsfrom the wetlands are considered. Inclusion of indirectbenefits, particularly that raw wastewater discharged to thewetlands receives “natural assimilative” treatment and non-use values of the wetlands increase this amount even further.These results suggest that diversion of water from agricultureto the wetlands is economically feasible and preferable to thedemise of the Sultan Marshes in the Develi Basin.

R E F E R E N C E S

Anonymous, 1998. Mera Yönetmeliği. (Pasture ManagementRegulation). July 31, 1998.

Anonymous, 2004. Yıllar itibariyle günlük ve aylık asgari ücretler(Daily and monthly statutory minimum wages). Çalışma veSosyal Güvenlik Bakanlığı (Ministry of Labour and SocialSecurity). Available at: http://www.csgb.gov.tr/istatistik/cgm/yillar_brut_asgari_ucret.htm (accessed on November 16,2006).

Barbier, E.B., 2003. Upstream dams and downstream waterallocation: the case of the Hadejia-Jama'are floodplain,northern Nigeria. Water Resources Research 39, 1311–1319.

Barbier, E.B., Thompson, J.R., 1998. The value of water: floodplainversus large-scale irrigation benefits in northern Nigeria.Ambio 27, 434–440.

Barbier, E.B., Acreman, M., Knowler, D., 1997. Economic Valuationof Wetlands: A Guide for Policy Makers and Planners. RamsarConvention Bureau, Gland, Switzerland.

Cezilly, F., Boy, V., Green, R.E., 1995. Interannual variation ingreater flamingo breeding success in relation to water levels.Ecology 76, 20–26.

Dadaser-Celik, F., 2006. Unpublıshed field data.Dadaser, F., Özesmi, U., 2002. Stakeholder analysis for Sultan

Marshes ecosystem: fuzzy cognitive mapping approach forconservation of ecosystems. Environmental Problems of theMediterranean Region Conference, Nicosia, North Cyprus.

Dadaser-Celik, F., 2008. Impact of large-scale irrigation on a closedbasin wetland: water flow alterations and participatoryirrigation management effects on the Sultan Marshesecosystem in Turkey. Ph.D. Dissertation, University ofMinnesota, Twin Cities, MN, USA.

Dadaser-Celik, F., Stefan, H.G., Brezonik, P.L., 2006. Dynamichydrologic model of the Örtülüakar Marsh in Turkey. Wetlands26, 1089–1102.

Dadaser-Celik, F., Brezonik, P.L., Stefan, H.G., 2007. Hydrologicsustainability of the Sultan Marshes in Turkey. WaterInternational 32.

Dadaser-Celik, F., Brezonik, P.L., Stefan, H.G., 2008. Agriculturaland environmental changes after irrigation managementtransfer in the Develi Basin, Turkey. Irrigation and DrainageSystems 22, 47–66.

DPT, 2001. 8. Bes Yıllık Kalkınma Planı-Hayvancılık Özel İhtisasKomisyon Raporu (8. Five-Year Development Plan - AnimalHusbandrySpecial ExpertiseCommissionReport).Devlet PlanlamaTeşkilatı (State Planning Organization), Ankara, Turkey.

DPT, 2003. Kayseri İli Gelişmişlik Performansı. (KayseriDevelopment Performance). Devlet Planlama Teşkilatı (StatePlanning Organization). Available at: http://www.dsi.gov.tr/duyuru/bitki_su_tuketimi.xls.

DSI, 2007. Bitki su tüketimleri ve sulama suyu ihtiyaçları. (Plantwater consumptions and irrigation water requirements). DevletSu İşleri (State Hydraulic Works), Ankara, Turkey. Available athttp://www.dsi.gov.tr/duyuru/bitki_su_tuketimi.xls.

Erdem, T., Erdem, Y., Orta, A.H., Okursoy, H., 2006. Response ofagronomic crops to use of deficit irrigation in the TrakyaRegion-Turkey. International Symposium on Water and LandManagement for Sustainable Irrigated Agriculture, Adana,Turkey.

Gerakis, A., Kalburtji, K., 1998. Agricultural activities affecting thefunctions and values of Ramsar wetland sites of Greece.Agriculture, Ecosystems and Environment 70, 119–128.

Gürpınar, T., 1994. Sultan Sazlığı'nın ornitolojik önemi ve dogal sudüzeni (Ornitological importance of Sultan Marshes and itsnatural water regime). Kayseri Sultan Sazlığı - Erciyes ÇevreSorunları Sempozyumu (Kayseri SultanMarshes -Mount ErciyesEnvironmental Problems Symposium). Kayseri Valiliği (KayseriGoverner's Office), Kayseri, Turkey, pp. 64–78.

1506 E C O L O G I C A L E C O N O M I C S 6 8 ( 2 0 0 9 ) 1 4 9 6 – 1 5 0 6

Hammer, D.A., Bastian, R.K., 1989. Wetlands ecosystems: naturalwater purifiers. In: Hammer, D.A. (Ed.), Constructed Wetlandsfor Wastewater Treatment. Lewis publishers, Chelsea,Michigan, USA, pp. 5–20.

Heimlich, R.E., Wiebe, K.D., Claassen, R., Gadsby, D., House, R.M.,1998. Wetlands and Agriculture: Private Interests and PublicBenefits. Resource Economics Division, Economic ResearchService, U.S. Department of Agriculture, Washington, DC, USA.

Karaata, H., 1991. Kırklareli Koşullarında Ayçiçeği BitkisininSu-Üretim Fonksiyonları (Water-Yield Relationships ofSunflower at Kırklareli (Turkey) Conditions). Köy HizmetleriGenel Müdürlüğü Atatürk Araştırma Enstitüsü (Directorate ofRural Affairs, Atatürk Research Institute), Kırklareli, Turkey.

Karabaşa, S., 2001. Saz Kesimi Reed Harvesting. Çevre ve OrmanBakanlığı (Ministry of Environment and Foresty), Ankara,Turkey.

Karadeniz, N., 2000. Sultan Sazlığı, ramsar site in Turkey.Humedales Mediterráneos 1, 107–114.

Kingsford, R.T., Thomas, D.H.L., 2004. Destruction of wetlands andwaterbird populations by dams and irrigation on theMurrumbidgee River in arid Australia. EnvironmentalManagement 34, 383–396.

Kırda, C., 2000. Deficit Irrigation Scheduling Based on Plant GrowthStages Showing Water Streets Tolerance Deficit IrrigationPractices. Food and Agriculture Organization of the UnitedNations (FAO), Rome, Italy.

Kırda, C., Kanber, R., Tulucu, K., 1999. Yield response of cotton,maize, soybean, sugar beet, sunflower and wheat to deficitirrigation. In: Nielsen, D.R. (Ed.), Crop Yield Response to DeficitIrrigation. Kluwer Academic Publishers, Dordrecht, TheNetherlands.

Kjellin, J., Wörmana, A., Johansson, H., Lindahla, A., 2007.Controlling factors for water residence time and flow patternsin Ekeby treatment wetland, Sweden. Advances in WaterResources 30, 838–850.

Lee, H., Sumner, D.A., Howitt, R.E., 1997. Economic Impacts ofIrrigation Water Cuts in the Sacramento Valley. University ofCalifornia, Agricultural Issues Center, Davis, CA, USA.

Lemly, A.D., Kingsford, R.T., Thompson, J.R., 2000. Irrigatedagriculture and wildlife conservation: conflict on a global scale.Environmental Management 25, 485–512.

Liu, H., Cai, X., Geng, L., Zhong, H., 2005. Restoration of pasturelandecosystems: case study of Western Inner Mongolia. Journal ofWater Resources Planning and Management 131, 420–430.

Loomis, J., 2005. Updated Outdoor Recreation Use Values onNational Forests and Other Public Lands. Department ofAgriculture, Forest Service, Pacific Northwest Research Station,Portland, OR, USA.

Ministry of Environment (Çevre Bakanlığı), 2002. Sultansazlığı SuBilançosu ve Alınması Gereken Tedbirler (Sultan MarshesWater Balance). Çevre Bakanlığı (Ministry of Environment),Ankara, Turkey.

Mmopelwa, G., 2006. Economic and financial analysis of harvestingand utilization of river reed in the Okavango Delta, Botswana.Journal of Environmental Management 79, 329–335.

Öğretir, K., 1993. Eskişehir Koşullarında Mısırın Su-Verim Ilişkileri(Water-Yield Relationships of Corn at Eskişehir Conditions).Köy Hizmetleri Genel Müdürlüğü Eskişehir Araştırma Enstitüsü(Directorate of Rural Affairs, Atatürk Research Institute),Eskişehir, Turkey.

Özaslan, M. and Şeftalici, H., 2002. Kayseri Il Gelisme Raporu(Kayseri Development Report). Devlet Planlama Teşkilatı (StatePlanning Organization), Bölgesel Gelişme ve Yapısal Uyum

Genel Müdürlüğü (General Directorate of RegionalDevelopment and Structural Adjustment), Ankara, Turkey.

Özesmi, U., 2002. GEF II, Sultan Sazlığı Doğal Kaynak YönetimProjesi, Sultan Sazlığı Saz Kesim Planı (GEF II, Sultan MarshesNatural Resource Management Project, Reed Harvesting Plan).Erciyes Üniversitesi (Erciyes University), Kayseri, Turkey.

Özesmi, U., Somuncu, M., Tuncel, H., 1993. Sultan Sazlığıekosistemi (Sultan Marshes ecosystem). Ankara ÜniversitesiTürkiye Coğrafyası Araştırma ve Uygulama Merkezi Dergisi(Journal of Ankara University Turkish Geography Research andApplication Center) 2, 275–288.

Öztürk, I., Timur, H., Koşkan, U., 2005. Atıksu Arıtımının Esasları:Evsel, Endüstriyel Atıksu Arıtımı ve Arıtma ÇamurlarınınKontrolü Principals of Wastewater Treatment: Municipal,Industrial Wastewater Treatment and Sludge Control). Çevreve Orman Bakanlığı (Ministry of Environment and Foresty),Ankara, Turkey.

Pak, M., Türker, M.F., 2006. Estimation of recreational use value offorest resources by using individual travel cost and contingentvaluation methods (Kayabaşı Forest Recreation site sample).Journal of Applied Sciences 6, 1–5.

Paşaoğlu, S., 1994. DSI Develi Havzası Ekoloji Koruma Projesi (DSIDeveli Basin Ecology Conservation Project). Kayseri SultanSazlığı - Erciyes Çevre Sorunları Sempozyumu (Kayseri SultanMarshes - Mount Erciyes Environmental Problems Sympo-sium). Kayseri Valiliği(Kayseri Governer's Office), Kayseri,Turkey, pp. 86–122.

Reed, S.C., Crites, R.W., Middlebrooks, E.J., 1995. Natural Systemsfor Waste Management and Treatment. McGraw-Hill, NewYork, NY, USA.

Şarkışla, A., 2002. GEF II - Biyolojik Çeşitlilik VeDoğal KaynakYönetimiProjesi, Sultansazlığı Tabiatı Koruma Alanı, Sosyal DeğerlendirmeRaporu (GEF II - Biological Diversity and Natural ResourceManagement Project - Sultansazlığı Nature ConservationArea - Social Environment Evaluation Project). Çevre ve OrmanBakanlığı (Ministry of Environment and Foresty), Ankara, Turkey.

Stewart, J.I., Cuenca, R.H., Pruitt, W.O., Hagan, R.M., Tosso, J., 1977.Determination and Utilization of Water Production Functions forPrincipalCaliforniaCrops.UniversityofCalifornia,Davis,CA,USA.

Sundling, D., Zilberman, D., MacDougall, N., Howitt, R., Dinar, A.,1997. Modeling the impacts of reducing agricultural watersupplies: lessons from California's Bay/Delta problem. In: Parker,D., and Tsur, Y. (Eds.), Decentralization and Coordination ofWater Resource Management. Kluwer Academic Publishers,Boston, MA, USA.

Ünlü, F., 2005. DSİ'ce Inşa Edilerek Işletmeye Açılan Sulama veKurutma Tesisleri 2004 Yılı Mahsul Sayım Sonuçları (CropProduction in 2004 in Irrigation and Drainage FacilitiesConstructed by DSI). Devlet Su İşleri (State Hydraulic Works),Ankara, Turkey.

Van Tuijl, W., 1993. Improving Water Use In Agriculture:Experiences in the Middle East and North Africa. World Bank,Washington, D. C., USA.

Watson, J.T., Reed, S.C., Kadlec, R.H., Knight, R.L., Whitehouse, A.E.,1989. Performance expectations and loading rates forconstructed wetlands. In: Hammer, D.A. (Ed.), ConstructedWetlands forWastewater Treatment. Lewis publishers, Chelsea,Michigan, pp. 319–351.

Yakan, H. and Sabuncuoğlu, S., 1991. Kırklareli KoşullarındaŞekerpancarı Su Tüketimi (Water Consumption of Sugar Beetsat Kırklareli (Turkey) Conditions). Köy Hizmetleri GenelMüdürlüğü, Kırklareli Araştırma Enstitüsü (Directorate of RuralAffairs, Kırklareli Research Institute), Kırklareli, Turkey.