Integrating agricultural policies and water policies under water supply and climate uncertainty



[1] Understanding the interactions of water and agricultural policies is crucial for achieving an efficient management of water resources. In the EU, agricultural and environmental policies are seeking to converge progressively toward mutually compatible objectives and, in this context, the recently reformed Common Agricultural Policy (CAP) and the EU Water Framework Directive constitute the policy framework in which irrigated agriculture and hence water use will evolve. In fact, one of the measures of the European Water Directive is to establish a water pricing policy for improving water use and attaining a more efficient water allocation. The aim of this research is to investigate the irrigators' responses to these changing policy developments in a self-managed irrigation district in southern Spain. A stochastic programming model has been developed to estimate farmers' response to the application of water pricing policies in different agricultural policy scenarios when water availability is subject to varying climate conditions and water storage capacity in the district's reservoir. Results show that irrigators are price-responsive, but a similar water-pricing policy in different agricultural policy options could have distinct effects on water use, farmers' income, and collected revenue by the water authority. Water availability is a critical factor, and pricing policies are less effective for reducing water consumption in drought years. Thus there is a need to integrate the objectives of water policies within the objectives of the CAP programs to avoid distortion effects and to seek synergy between these two policies.

1. Introduction

[2] Water is a scarce resource in the Mediterranean basin and the agricultural sector is the largest user of water. Spain's irrigated agriculture consumes close to 80% of the total available water resources in the country but it extends over 3.6 million ha that, being the largest irrigated surface in the EU, it constitutes a mere 18% of the total cultivated lands. However, irrigated crops account for 60% of all agricultural production and 80% of farm exports. Thus irrigation water has become an essential input to sustain agricultural activity and therefore a more efficient water management is progressively required.

[3] Spain is divided into eight different publicly owned and independently managed river basin authorities that are responsible for water management in the basins and for delivering water to the irrigation districts. The irrigation districts are well established Water User Associations that allocate water among its members. Water in Spain is of public ownership, including underground waters, and irrigators have the right to use water under nontradable concessions, granted by the public authority, that determines their water allotments.

[4] In Spain as in other Mediterranean countries increasing water shortages in many regions have called for a progressive implementation of water policies that will ameliorate water management and increase water use efficiency. In fact, several water policies have been proposed for this purpose in Spain and elsewhere such as pricing policies [Varela Ortega et al., 1998; Tsur and Dinar, 1997; Dinar and Subramanian, 1997; Rosegrant et al., 1995], water markets [Garrido, 1998; Rosegrant and Binswanger, 1994], water rights adjustments [Sumpsi et al., 1998] and financial incentives [Spencer and Subramanian, 1997]. In this connection, the EU Water Framework Directive (WFD) 2000/60/EC enacted in 2000 has established a Framework for Community action in the field of Water Policy and represents an important step toward sustainable use of water resources in the EU. The WFD proclaims an integrated management of all water resources by establishing river basins as the basic unit for all water planning and management actions. In addition, the WFD requires Member States (Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Portugal, Spain, Sweden, Netherlands, and United Kingdom and, from 1st May 2004, Czech Republic, Estonia, Cyprus, Latvia, Lithuania, Hungary, Malta, Poland, Slovenia, and Slovakia) to undertake an economic analysis of water use (article 5) and stipulates that all Member States must take into account the principle of recovery of the costs of water services, including environmental and resource costs, in accordance with the polluter pays principle (article 9). For this purpose, Member States, among other things, shall ensure by 2010 that water-pricing policies provide adequate incentives to use water resources efficiently. Unquestionably, the implementation of the WFD will have important consequences in the irrigated agricultural sector of the EU.

[5] On the other hand, EU agricultural sector is strongly determined by agricultural policy programs aimed to secure farmers' income and protect the rural environment. The Agenda 2000 constitutes the current CAP framework and aims to procure a multifunctional, sustainable and competitive agriculture throughout the EU territory. The Agenda 2000 is based on the establishment of production-related direct aid payments and gives a prominent role to agrienvironmental instruments to support a sustainable development of rural areas and to respond to society's increasing demand for environmental services. Alongside, Rural Development measures seek to stabilize and support rural communities by further integrating environmental and socioeconomic aspects as Member States are given the option to condition the access to the CAP aid payments to meeting certain environmental requisites (cross-compliance option). This tendency is being further reinforced in the strongly debated new CAP reform, recently approved on June 2003. It continues the reform process of the Agenda 2000 to enable a more competitive and less trade-distorting EU agricultural production. For this purpose, one of the fundamental measures of the new reform is the replacement of most of the current direct aid payments by a single farm payment scheme. This single payment will not be linked to production and beneficiaries will be obliged to accomplish certain environmental and food safety requirements. The single aid payment will come into operation on January 2005, although Member States may decide to delay implementation up to 2007.

[6] Thus the irrigated agricultural sector in Spain as in other EU countries is influenced by two types of policies: Agricultural Policies and Water Policies. The new CAP reform of 2003 and the European Water Framework Directive (WFD) will constitute the future policy context for the development of the irrigated agricultural sector and hence for water use in the agricultural sector. Therefore, in the near future, the analysis of the interactions of the agricultural and water policies in the irrigated agriculture will be essential to avoid conflicts in the implementation of both policies and to promote a real integration of agricultural and water policies. Unfortunately, published research in this area is still limited. For Spain, an example is given by Gómez-Limón et al. [2002], who analyze the effects of different CAP and WFD scenarios in an irrigation district situated in north central Spain. The results of this study show that the joint implementation of the CAP and the WFD instruments would raise conflicts, given that the WFD implementation will impose additional costs on irrigated farming negatively affecting EU agriculture competitiveness. The present article tries to further contribute to this area of research, bringing some understanding to the difficult task of integrating agricultural policies and water resources policies.

2. Objectives of the Research

[7] The aim of this research is to predict the irrigators' response confronted to the combined application of water policies and agricultural polices. We analyze comparatively the effects of the implementation of a water pricing policy, defined by the application of volumetric water charges, in different agricultural policy alternatives. In this combined policy context, we have also taken into account the different water availability situations that farmers face that constitute one of the major sources of uncertainty for attaining the farmers' programmed production plan. Water availability is determined by varying climate conditions and storage capacity in the district's reservoir. Specifically, the analysis focuses on the evaluation of the effects of these policies on water use, farmers' income and on the revenue collected by the water management agency. In particular this paper will address the changing strategies that farmers will follow when water-pricing policies are applied in conjunction with agricultural policies as envisaged in the new EU policy context. That is, changes in the cropping pattern, land allocation between irrigated and rain fed farming, cropping intensification and management of water in the farm. All these issues are intended to throw some light into the difficult task that the EU, as other countries elsewhere, is facing to integrate water conservation policies and agricultural policies.

3. Methodology

3.1. Model

[8] To analyze the impacts of the integrated application of agricultural and water policies, we have built a Discrete Stochastic Programming Model (DSPM) that simulates the farmer's behavior in different policy scenarios. The model takes into account the uncertainty that farmers face on water availability along their decision making process in a given cropping season [Apland and Hauer, 1993; Taylor and Young, 1995; Torkamani and Hardaker, 1996; Jacquet and Pluvinage, 1997; Keplinger et al., 1998; Blanco, 1999]. In fact, risk analysis in farmer economic behavior models has become an essential element for models to be useful. We find in the literature several mathematical programming techniques that take into account different sources of risk and uncertainty [Hazell and Norton, 1986; Rae, 1994; Hardaker et al., 1997]. We also know that most of the decisions in agricultural production are taken in a sequential way [Rae, 1971a, 1971b; Antle, 1983; Adesina and Sanders, 1991; Taylor and Young, 1995; Torkamani and Hardaker, 1996]. We consider that dynamic models can be more valid than static models to understand risk effects in agricultural production [Antle, 1983]. The Discrete Stochastic Programming Model (DSPM), developed by Cocks [1968] and Rae [1971a, 1971b], allows us to solve a multistage decision process in which the decision maker's knowledge about random events changes through time as economic choices are made.

3.1.1. Uncertainty in Water Availability

[9] Sumpsi et al. [1998] and Blanco [1999] show that the main source of risk in the southern Spanish irrigated agriculture is water availability. Water availability for irrigation in these areas depends on climatic conditions along the crops' growing period and on the amount of water collected in dams and reservoirs. Following this line, the main assumption in our analysis is that farmers make their production choices over time based on their knowledge about the amount of water that will be delivered from the dam and on existing climatic conditions.

3.1.2. Model Structure

[10] The model (represented in Figure 1) is defined by three decision stages along an agricultural year. It assumes full knowledge of the past and considers three “states of nature” for stage two and nine for stage three. The different “states of nature” are defined by the combination of two random variables, namely the volume of water available from the reservoir and the climate conditions. We have considered three levels of water volumes delivered from the district's reservoir (S1, S2, S3) and three different climatic situations (A1: humid year, A2: average year, A3: dry year) according to statistical information. Spring rainfall has been taken as the reference climatic condition. The combination of these variables along the decision making process generate nine final “states of nature” (A1S1, A2S1, A3S1, …).

Figure 1.

Decision tree. Water demand functions according to the availability of water.

[11] The farmers' decision process is represented as follows: (1) In the first stage, farmers take decisions concerning the crop surface allocation for winter season. The model allows certain crops to be irrigated with different techniques, so that farmers also take a decision on which irrigation technique would be applied to these specific crops. These decisions are taken prior to the farmers' knowledge of the yearlong climate conditions and of the water that will be actually delivered from the district's reservoir (which is decided in the spring). Thus, in this stage farmers decide what to crop in the wintertime considering their expectations on water availability. (2) In the second step, decisions concerning the surface assigned to spring crops and as explained above, for certain crops the irrigation technique applied, are taken according to the farmer's knowledge of the water that will be released from the reservoir. (3) Finally in stage three farmers decide on the irrigation intensity (intensive or extensive) once they have precise information on the actual climatic conditions, that is the spring rainfall, that will determine crop yields.

[12] Additionally, farmers can decide to invest in new irrigation equipment (drip irrigation). Given that the model represents an agricultural year, the investment cost is represented by the annual repayment of the new irrigation equipment.

[13] The model maximizes the expected farmer's income under different constraints (technical constraints, economic constraints and policy constraints). The objective function is

equation image

where Zsa is the farmer's income in the state of nature sa, and pspa are the probabilities of the different states of nature. Farmers' income is defined by

equation image

where subindex j accounts for crop type, k for soil quality, t for the irrigation intensity, r for the irrigation technique (gravity, drip irrigation) and sa for the different states of nature; X3 is the crop surface in the last step of the decision making; Yj,k,r,t,a is crop yield; pj is crop prices; subj are product subsidies; CIRRs,a are water application costs; CMOs,a are labor costs. Total water application costs CIRRs,a are composed of four elements:

equation image

where Cx is fees paid by the irrigators to the water authority in the river basin (equation image/ha), Cq is a variable volume charge (equation image/m3), Cr is irrigation equipment maintenance costs (equation image/ha), Car is repayment of the investment in irrigation equipment (equation image/ha).

[14] The objective function is subjected to the following constrains (omitting labor, production possibilities, and policy constraints):Availability of land

equation image

where Sk represents the surface by soil quality.Irrigation equipment

equation image

where Sr is the surface by irrigation technique (drip and surface) and Ir accounts for the surface in which new investment in drip irrigation technique is undertaken.Irrigation restrictions

equation image
equation image

where bej,k,r,t are the crop water requirements, Qs,a is water availability in the state of nature sa, h is a water distribution efficiency parameter and Ds is water allotment in the state of nature s. The model has been validated comparing the observed values with the results predicted by the model.

3.2. Zone of Study

[15] For the empirical application of this study, we have selected the irrigation district El Viar, a medium-size Water Users Association (WUA) that extends over 11,958 ha in the Guadalquivir river basin in the region of Andalusia in southern Spain. The district is an old self-managed irrigation district established in 1958 constituted by 3978 farms. Water supply depends on the storage capacity of a reservoir directly managed by the own district. Main crops grown are cereals, cotton, sunflower, vegetables and fruit trees. Irrigation techniques used are mainly gravity and drip irrigation and the water allotment right for the farmers in the district is around 8000 m3/ha, which may vary considerably depending on water availability conditions. Irrigators pay a fixed fee of 102.17 equation image per ha to cover partially the costs of distribution, maintenance of infrastructure and administration. This fixed fee is composed by different elements: the basin agency fees that cover the capital, operation and maintenance cost of the state financed infrastructure and the irrigation district fee intended for the operation and maintenance cost of the irrigation district. Historically, this irrigation district has suffered from the effects of water scarcity in drought periods. Therefore public water authorities are considering the establishment of water conservation practices in the irrigated agriculture as a major priority in this area [Iglesias et al., 2000].

[16] For the purpose of the analysis, the irrigation district has been characterized by a typology of four representative farms (F1, F2, F3, F4) (Table 1), which represents the irrigation agriculture in the area. The representative farms were selected according to farm size, soil quality, water availability, crops grown and irrigation technique. The technical and economic parameters of the representative farms have been obtained from a survey conducted in the study area [Sumpsi et al., 1998] and further revised to adjust to the new policy conditions.

Table 1. Farm Typologya
Farm TypeArea, haSoil Quality, %Irrigation Technique, haCropping Pattern, %
F15K1: 60% K2: 40%gravity: 4 drip: 1cotton 20%, vegetables 29%, sugar beet 20%, others 31%
F240K1: 100%gravity: 30 drip: 10cotton 34%, vegetables 4.5%, others 40%, sugar beet 20%, fruit trees 1.5%
F340K1: 40% K2: 60%gravity: 30 drip: 10cotton 34%, vegetables 7%, others 40%, sugar beet 19%, fruit trees 4.5%
F4100K1: 50% K2: 50%gravity: 70 drip: 30cotton 15%, corn 15%, fruit trees 10%, others 30%, wheat/sunflower 30%

3.3. Policy Options

[17] We have defined three policy options within the framework of the CAP: (1) CAP reform of 1992, (2) Agenda 2000 measures currently applied in the area and (3) the establishment of equal direct payments for all crops decoupled from crop yields (calculated to maintain the same level of aid payments as in Agenda 2000). This scenario represents the trend followed by EU policies toward production-neutral payments (to comply with the WTO agreements) and it can be considered a first step to the complete decoupled payments envisaged in the new reform of the CAP. For all agricultural policy options considered we have simulated jointly the application of a water pricing policy defined by administered volumetric water charges. Consequently, we will be able to study the interactions between the implementation of agricultural and water policies. The policy scenarios are summarized in Table 2.

Table 2. Policy Scenarios
Agricultural Policy OptionsaPolicy Instruments
Price SupportDirect Payments
  • a

    For all policy options a water policy has been simulated defined by volumetric water prices (a charge per volume applied, t equation image/m3).

CAP Reform 1992 (E1)high price supportlow direct payments tied to crop yields
Agenda 2000 (E2)low price supporthigh direct payments tied to crop yields
Equal Aid Payment (E3)low price supportequal direct payments for all crops independent of crop yields

4. Results

4.1. Effects of Water Availability and Climate Uncertainty

[18] In all agricultural policy scenarios, water demand in water scarcity situations is very inelastic (Figures 2, 3, and 4). Thus water consumption in dry periods (low water supply from the reservoir and drought year) is not significantly reduced until prices reach high levels but farm income is negatively affected. Farmers' rigidity to adapt to water scarcity situations prevails across all policy options. Agenda 2000 shows as like more inelastic demand as prices have to mount 0.13 equation image/m3 to reduce water consumption (as compared to 0.11 equation image/m3 in the 1992 CAP scenario and 0.09 equation image/m3 in the Equal Aid Payment scenario).

Figure 2.

Water demand functions according to the availability of water: 1992 CAP scenario.

Figure 3.

Water demand functions according to the availability of water: Agenda 2000 CAP scenario.

Figure 4.

Water demand functions according to the availability of water: Equal Aid Payment scenario.

[19] With regard to the effects of water availability and climate uncertainty in the different CAP scenarios, results show that farm income is more severely affected by water availability reductions in the Agenda 2000 scenario (Tables 3 and 4). Thus farmers' capacity to adapt to water scarcity situations is based on their possibilities to change the cropping pattern, which are substantially reduced in the Agenda 2000 scenario. In fact, low-water demanding crops such as oil seeds are less profitable in Agenda 2000. In general, it should be pointed out that according to the results obtained, farmers do not react to water scarcity situations by changing the irrigation technology applied to the different crops.

Table 3. Water Consumption by Policy Option and Water Availability
CAP ScenarioWater Consumption
Dry Year, ×103 m3/haDry Year, %Normal Year, ×103 m3/haNormal Year, %Humid Year, ×103 m3/haHumid Year, %
CAP 922.0850.14.151004.96119.5
Agenda 20002.0850.14.151005.03121.2
Equal Aid Payment2.0850.14.0697.84.27102.3
Table 4. Farm Income by Policy Option and Water Availability
CAP ScenarioFarm Income
Dry Year, equation image/haDry Year, %Normal Year, equation image/haNormal Year, %Humid Year, equation image/haHumid Year, %
CAP 92937.0458.01614.141001970.48122.1
Agenda 2000879.5254.51572.9797.41939.71120.2
Equal Aid Payment967.0759.91615.68100.11619.39100.3

4.2. Combined Effects of Agricultural Policies and Water Policies

[20] Pricing policies have been simulated to cover an ample range of price rates (ranging from 0 equation image/m3 to 2 equation image/m3 when water use equals zero, in 48 simulation levels). For the purpose of clarity we have summarized the results selecting three levels of water prices, including the price levels that will recover the O&M costs. Cost recovery is considered a crucial issue as the WFD considers the introduction of a water pricing policy that will ensure an efficient use of water taking into account the principle of cost recovery for water services, including environmental and resource costs. The recovery of the O&M costs can be considered a first step toward the realization of this principle. The price levels selected are as follows: (1) P1 is no water pricing policy. As explained above, farmers pay a fixed fee of 102.17 equation image/ha. (2) P2 is water price to recover the O&M costs of the irrigation district. Following statistical information contained in the River Basin Hydrologic Plan 237.27 equation image/ha has been considered the minimum amount to cover the O&M cost of the irrigation district. (3) P3 is water price that maximizes the district's revenue.

[21] Table 5 shows the aggregate results of the effects on water consumption, farm income, and the revenue collected in the irrigation district in the average state of nature. Table 6 shows the results of the cropping patterns chosen by the farmers in the simulated scenarios in the average state of nature. Figure 5 depicts the average revenue collected and the cost recovery level for all policy options simulated.

Figure 5.

Average irrigation district revenue in the agricultural policy scenarios.

Table 5. Results of Policy Options (Average State of Nature)
TariffsWater consumptionFarm IncomeIrrigation District Revenue
000 m3/ha%equation image/ha%equation image/ha%
CAP 92
P1 = 0 equation image/m33.811001474.64100101.03100
P2 = 0.04 equation image/m33.7598.41339.5490.8238.32233.91
P3 = 0.09 equation image/m32.8875.61160.9178.7360.31356.64
Agenda 2000
P1 = 0 equation image/m33.811001431.3797.1101.03100
P2 = 0.05 equation image/m33.5492.911232.8083.60281.57278.69
P3 = 0.09 equation image/m33.0880.81117.1075.7364.93361.20
Equal Aid Payment
P1 = 0 equation image/m33.2284.51481.55100.5101.03100
P2 = 0.07 equation image/m31.7846.721283.8487.06249.83247.28
P3 = 0.14 equation image/m31.3341.51164.4278.9324.55258.77
Table 6. Cropping Pattern Selection by Policy Option (Average State of Nature)
TariffsWheat, %Corn, %Sunflower, %Cotton, %Irrigated Surface, %
CAP 92
P1 = 0 equation image/m31019162595
P2 = 0.04 equation image/m31020162495
P3 = 0.09 equation image/m329200580
Agenda 2000
P1 = 0 equation image/m3101403582
P2 = 0.05 equation image/m3221303593
P3 = 0.09 equation image/m338003596
Equal Aid Payment
P1 = 0 equation image/m338003596
P2 = 0.07 equation image/m359001293
P3 = 0.14 equation image/m34500493

[22] In the 1992 CAP reform scenario (E1) (reference scenario), water use is high (almost 4000 m3/ha) and farmers grow a high proportion of water-demanding crops such as cotton, corn and vegetables. When water charges are applied in this scenario, water use is only slightly reduced by 2% but farmers' income will decrease by 10%. O&M costs will be recovered only if water prices increase beyond 0.04 equation image/m3, well below the water price that will maximize the district's revenue (0.09 equation image/m3). In addition, as water prices rise, farmers grow less water-demanding crops such as wheat, sunflower and the irrigated surface will be progressively reduced.

[23] Compared to the reference scenario, Agenda 2000 (E2) will not induce water use reductions. However, total irrigated surface will decrease by 13% as traditional crops such as sunflower and corn will diminish and farm income will decline slightly (3%). When water prices are charged in this scenario, farmers will tend to adopt water saving strategies by cropping less water demanding crops. In addition, recovering O&M costs will require prices to mount to 0.05 equation image/m3 as water consumption will be further reduced by 7% and farmers' income by 13%. Thus cost recovery policies will have greater effects in this scenario.

[24] The Equal Aid Payment Policy (E3) will achieve a substantial reduction in water consumption even without applying any water policies, as opposed to the other policy options. Comparatively to the reference scenario, water demand will decline by 15.5% but the irrigated surface will remain constant. This means that crops with low water requirements will be cultivated in the irrigated lands. However, this policy alternative will induce a crop specialization toward wheat production and other crops such as corn and sunflower, traditionally grown in the area, will disappear. Wheat production will be profitable and hence farm income will increase slightly (1%). Water policies are very effective in this scenario for attaining water saving objectives. When water charges are applied, price rates have to mount to 0.07 equation image/m3 to recover O&M costs and this will in turn decrease water demand by 38% but farm income will decrease proportionally less (13%). The revenue collected in the district will continue to rise and will reach a maximum at 0.14 equation image/m3. Additionally, in all the agricultural policy scenarios, results show that the application of water charges does not induce farmers to invest in new irrigation equipment.

5. Conclusions

[25] 1. Results show that a water pricing policy based on volumetric charges could have a low impact on water consumption in dry years. In fact, the analysis reflects that when water availability is low, water demand is very inelastic for low price ranges. Water consumption will decrease only if prices mount considerably. Thus it indicates that this water policy will not produce the desired effects of water conservation in the case of drought.

[26] 2. The recovery of the O&M cost will be attained at different water charges in each agricultural policy scenario. Thus the effects of these water charges will be different across agricultural policies, as strategies followed by the farmers will vary accordingly. That is, cropping patterns will change in each agricultural policy context when the recovery of the O&M costs is attained. Within the water price range selected (recovery of O&M costs up to the maximum revenue collected by the water district), we can observe that the price interval is wider in the Equal Aid Payment scenario. This will permit a more flexible integration of Water Policies and Agricultural Policies when payments received by the farmers are decoupled from production levels. This can be considered as a positive effect of the widely discussed decoupled policies within the EU context. However it cannot be generally argued that this policy is unquestionably beneficial. We have observed from the discussion of the results that the application of an equal direct payment for all the crops may produce also negative effects such as the reduction of crop diversity.

[27] 3. The implementation of an Equal Aid Payment Policy, neutral from production, will attain the objective of water conservation and farmer's income maintenance more efficiently than the other policy alternatives. In this policy scenario water is reduced without inflicting any income loss to the farmers who in fact attain a slightly higher income. In the other policy options water conservation objectives are only met at the cost of incurring in farm income reductions.

[28] 4. With respect to the introduction of a water pricing policy, in the agricultural policy option currently applied in the area studied, Agenda 2000, it is expected that the introduction of water pricing policies will produce certain detrimental effects on farmers' income and water will not be substantially reduced. Thus, to avoid this distortion, the application of the European Water Directive will need to be done in accordance with the CAP programs and specifically with the CAP agrienvironmental measures.

[29] 5. As we have already argued, a similar water-pricing policy could have distinct effects in different agricultural policy options. Thus it can be concluded that there is a need to integrate the objectives of the EU WFD with the objectives of the CAP programs to avoid distortion effects and to seek a synergy between these two policies. In this way, the use of cross compliance introduced in the “Agenda 2000” and further reinforced in the new CAP reform of 2003, that enables Member States to condition CAP subsidies to meet certain environmental requirements, could be a good opportunity to integrate the WFD objectives into agricultural policies.

[30] 6. It should be noted that these conclusions refer to a specific area in the South of Spain characterized by a highly productive and intensive agriculture. In this area traditional subsidized crops such as cotton, wheat and sunflower are grown together with competitive nonsubsidized horticultural production. Results would be different in other agricultural systems affected by the same policy scenarios.

[31] 7. In this analysis we have tried to bring some understanding to the difficult task of integrating agricultural policies and water resources policies, to attain the dual objective of maintaining farmers' income and conserving water resources and environmental sustainability. This is a crucial undertaking in many areas in the world where water availability is a major problem such as the Mediterranean region.


[32] This paper has been based on the research project carried out at the CIHEAM - IAMM (Mediterranean Agronomic Institute of Montpellier, France): Etude des impacts socio-économiques des Politiques de Gestion de l'eau et des Politiques Agricoles: Modélisation de la production agricole d'un périmètre irrigué situé dans le Bassin du Guadalquivir (Espagne). This paper expresses the views of the author and does not necessarily reflect the policy of FAO and its Member Countries.