Bio -based and A ppl ied Economics BAE Bio-based and Applied Economics 10(4): 305-323, 2021 | e-ISSN 2280-6172 | DOI: 10.36253/bae-10465 Copyright: © 2021 G. Layani, M. Bakhshoodeh, M. Zibaei, D. Viaggi. Open access, article published by Firenze University Press under CC-BY-4.0 License. Firenze University Press | www.fupress.com/bae Citation: G. Layani, M. Bakhshoodeh, M. Zibaei, D. Viaggi (2021). Sustainable water resources management under population growth and agricultural development in the Kheirabad river basin, Iran. Bio-based and Applied Economics 10(4): 305-323. doi: 10.36253/ bae-10465 Received: February 15, 2021 Accepted: October 14, 2021 Published: March 31, 2022 Data Availability Statement: All rel- evant data are within the paper and its Supporting Information files. Competing Interests: The Author(s) declare(s) no conflict of interest. Editor: Fabio Gaetano Santeramo. ORCID GL: 0000-0002-0110-0113 MB: 0000-0001-8217-3535 MZ: 0000-0003-4633-0593 DV: 0000-0001-9503-2977 Sustainable water resources management under population growth and agricultural development in the Kheirabad river basin, Iran Ghasem Layani1,*, Mohammad Bakhshoodeh2, Mansour Zibaei2, Davide Viaggi3 1 Shiraz University, Iran 2 College of Agriculture, Shiraz University, Iran 3 Department of Agricultural and Food Sciences, University of Bologna, Italy *Corresponding author. E-mail: Ghasem.layani.su@gmail.com Abstract. In this study, an integrated system dynamics model was developed for sce- nario analysis in sub-sectors of the Kheirabad River Basin in southwestern Iran where managing water resources is seriously challenging due to population growth and peri- odic drought. Afterward, the variability of water demand and supply under baseline scenario and different water demand management policies, including water conserva- tion and water pricing, was evaluated. Findings illustrated that with increasing popula- tion and cropland area if no further demand management policies were implement- ed, the total water demand and withdrawal of water resources increase by more than 0.75% annually. The annual surface water availability during 2018-2030 is expected to decrease by around -1.23%. Under these circumstances, the sustainability index of the water resources system is equal to 0.703, indicating that the water system would not be able to meet the total water demand in the near future. However, the water resource sustainability index increases significantly by improving irrigation efficiency and changing crop patterns at the basin. Also, the reduction in per capita water demand and domestic water pricing under the competition structure would help to improve the sustainability index to 0.963 and 0.749, respectively. Keywords: Sustainability Index, water system, system dynamics, agriculture, food security, Kheirabad River Basin. JEL codes: Q2, Q25. 1. INTRODUCTION Water is essential for people’s daily life, agricultural irrigation, fish farm- ing, and manufacturing (UNIDO, 2003). However, this vital resource is faced with several stresses in quantity and quality (Speelman & Veettil, 2013). Among the others, climate variability and increasing population growth have resulted in water scarcity in many countries especially in the arid regions (Hashemi et al., 2019; Mulwa et al., 2021). The water scarcity problem threat- ens nearly 80% of the world’s population (Vallino et al., 2020). Increasing http://creativecommons.org/licenses/by/4.0/legalcode 306 Bio-based and Applied Economics 10(4): 305-323, 2021 | e-ISSN 2280-6172 | DOI: 10.36253/bae-10465 Ghasem Layani et al. water demand in various economic and social sectors exacerbates the problem of water scarcity (Donati et al., 2013) and can make the water system more vulnerable (Cai et al., 2018). Therefore, the most challenging issue in water resources system in the world is to achieve a balance between supply and demand (Kotir et al., 2016; Xiong et al., 2020). The complexity of water systems is familiar to all those studying in the field because of fundamentally their large number of agents and interdependent sub- systems (Madani and Mariño, 2009; Balali and Viaggi, 2015). In a water system, there are dynamic feedback relationships among different factors on the supply and demand sides (Kotir et al., 2016). Furthermore, the changes in water resource have a dynamics behavior as it is affected by many socio-economic and climatic fac- tors over time (Sterman, 2001). In other words, popula- tion growth, climate change, agricultural development, changes in harvesting rate from ground and surface water are factors that affect the water system of a region over time with interaction (Brown et al., 2015). The use of water in one sector also affects other sectors, and the agents in the water system are contiguous. These inter- actions between different water users such as irrigation, drinking water, industrial production, and environmen- tal facilities lead to complexity in the water resources system (Berger et al., 2007). These complexities in the water resources system cause policymakers to face policy resistance in managing water resources. Policy resistance occurs when policy actions trigger feedback from the environment that undermines the policy and at times even exacerbates the original problem. Policy resistance is common in complex systems characterized by many feedback loops with long delays between policy action and result (Sterman, 2001). Besides, implementing dif- ferent policies to manage water resources, depending on the conflict of interest, may have different effects on dif- ferent stakeholder groups (Darbandsari et al., 2020). Addressing the complexities of water resources sys- tem, a holistic approach such as system dynamics (SD) can provide a sufficient water management framework based on conflict resolution approaches. System dynam- ics consider the interactions among different elements of different stockholders for simulating the behavior of the system and policy analysis (Frank, 2000). This helps decision-makers assess different management poli- cies considering various aspects (e.g., economic, social, environmental, etc.) for simultaneously reducing con- flicts and improving water resources conditions (Mirchi, 2013; Darbandsari et al., 2020). There are a large volume of published studies that have applied SD modeling to evaluate the effect of changes in some variables such as water demand, population control, water transfer as well as climate change on water availability (Gohari et al., 2017; Sun et al., 2017; Pluchinotta et al., 2018; Mah- davinia and Mokhtar, 2019; Keyhanpour et al., 2020). A great deal of previous research into water management has focused on mathematical programing, but they do not pay attention to the feedback processes in the water resources system (Donati et al., 2013; Archibald & Mar- shall, 2018; Zeng et al., 2019; Saif et al., 2020). Given the significant water consumption in the agricultural sector, these studies emphasize that local water management authorities, in addition to being aware of farmers’ pos- sible decisions to allocate farms, should also be able to provide an optimal cultivation pattern commensurate with the potential of each region (Donati et al., 2013). Although good progress has been made in the SD modeling of water resources system in different stud- ies, there are still limitations. Some important limita- tions of these studies are briefly as follows: (i) in gen- eral, less attention has been paid to theoretical foun- dations in modeling in the agricultural subsystem (Madani and Mariño, 2009; Gohari et al., 2017; Mah- davinia and Mokhtar, 2019); (ii) some studies (Kotir et al., 2016) considered the crop yields as a stock variable, which contradicts the definitions of the stock variable; (iii) in the population subsystem, few studies (Clifford Holmes et al., 2014; Goldani et al., 2011) have consid- ered the behavior of consumers to change in water prices; (iv) although most of the above-mentioned studies have focused on the interaction between ele- ments and feedback loops in the water system, a few of them (Madani and Mariño, 2009; Gohari et al., 2017) have been designed to analyze various water indica- tors, for instance, sustainability index that is defined as the ratio of water supply and demand and sum- marizes the performance of alternative scenarios and policies (Loucks, 1997). It should be noted that the above points are important in studying the behavior of the water system at the basin. Compared to previ- ous studies, to achieve a better result, we used a Ner- love (1956) partial adjustment framework to model the agricultural subsector and simulate cropland area and agricultural water demand. In more detail, farm- ers’ decisions to develop the cropland area were con- sidered in response to changes in crop prices in mod- eling. It can be an effective effort to more accurately simulate the agricultural water demand. Also in the population sub sector, consumers’ responses to water price changes were taken into account. Policies such as taxes  and  subsidies can  change the price of goods and correspondingly the quantity consumed. Thus, various indicators including sustainability (Loucks, 307 Bio-based and Applied Economics 10(4): 305-323, 2021 | e-ISSN 2280-6172 | DOI: 10.36253/bae-10465 Sustainable water resources management in the Kheirabad river basin, Iran 1997), reliability (McMahon et al., 2006), vulnerability (Hashimoto et al., 1982) and max deficit (Moy et al., 1986) indices, were considered to evaluate the effects of water resources management policies and to rank different policies base on their effects on water system behaviour. Because of increasing complexity and integration of environmental, social, and economic functions, the early water resource models still need to be developed and appropriate policies should be adopted based on the socio-economic and environmental characteristics of basin. Accordingly, this paper develops an integrated SD simulation model for exploring the water resource sustainable index in the Kheirabad river basin in south- western Iran where managing water resources is seri- ously challenging due to population growth and periodic drought. Put it simply, the present study aims to explore the water supply and demand dilemmas and calculate the water resource sustainability index at the basin. This paper is organized as follows. The case study and SD model features are presented in the next section. Then, the applied data are described. The simulation results of the model are presented in Section 4 and the conclusions are provided in Section 5. 2. THE STUDY CONTEXT AND SCOPE Iran is located in the mid-latitude belt of arid and semi-arid regions of the Earth. The arid and semi-arid regions cover more the 60% of the country Iran. The main source of water in Iran is precipitation in the form of 70% rainfall and 30% snow, which is estimated to be about 413 BCM (billion cubic meters). About 71.6% of the total rainfall (295 BCM) is directly evaporated. Considering 13 BCM of water entering from the bor- ders (joint border rivers), the total amount of the coun- try’s renewable water resources (long-term averages for 1977 to 2018) is annually estimated to be 124 BCM, of which about 73 BCM go to surface runoff. Groundwa- ter recharge is annually estimated to be about 51 BCM. Currently, total water consumption is approximately 88.5 BCM (Abbasi et al., 2015). Agricultural water con- sumption accounts for about 85% of total water resourc- es in Iran and 90% of them may be allocated in surface irrigation systems with low efficiency and full water supply (Lalehzari et al., 2020). According to the latest figures, the average population growth rate in Iran dur- ing 1999-2000 was 1.755 percent and lowered to 1.246 percent in 2010-2017. However, in all these periods, Iran’s population growth rate is above the global average (UNDATA, 2017). The annual water consumption in the urban areas of the country is about 5.4 BCM, of which 4.3 BCM is related to household consumption that implies to the per capita water consumption of 224 liters per person a day. As far as population growth is consid- ered, the increasing demand is not limited to fresh water use for drinking purposes. The growing population is results in increasing demand for agricultural products as well, especially for some strategic food stuffs such as wheat that are provided at subsidized prices and the Ira- nian government insists on their domestic supply (The Statistical Center of Iran, 2018). Considering the driving factors of water crisis, the water resources management issue is a national priority and the most important issues among policymakers in Iran (Madani, 2014). Kheirabad river basin is a part of the Zohre river basin in the Kogiluyeh and Boyerahmad province, south- western Iran (Fig. 1). The average annual rainfall of the basin, where the rainfall regime is Mediterranean (with dry and wet season), varies from less than 200 mm to more than 800 mm The average annual temperature also varies from 12°c to 25°c. The water consumption of the Kheirabad river basin in the drinking, industrial and agricultural sectors is provided of surface and ground- water resources. This basin is rich in surface water, but the un-normalized utilization of soil and water resources and also the increasing water resources withdrawal have reduced the basin’s water potential to meet increas- ing demands. Most of the surface water resource in the basin is provided by Kowsar reservoir dam located in Zohre river basin in the west of Gachsaran Coun- ty. Rainfall is extremely seasonal; about 50% of which occurs in winter (concurrently with the smallest water demand), 23% in spring, 23% in autumn, and 4% in summer (concurrently with the greatest water demand). Kheirabad river basin’s average annual precipitation is estimated to be 331 mm during 2012-2020 while evapo- ration amount is more than three times that. Not only the climate variability but also the population as an important factor affecting water demand, is continually increasing. While according to the report presented by the Regional Water Organization of Kogiluyeh and Boy- erahmad province (2017), the average per capita domes- tic water consumption of this province is more than 220 liters per day, which is about 20 percent higher than the national average. The combination of these factors led to the water stored in Kowsar dam has declined in recent years. Because one of the most important goals of the Kowsar dam construction is the supply of drinking water in the southern provinces of Iran and agricultural development in these areas, meeting the growing water demand in this basin is becoming a concern among pol- icymakers. 308 Bio-based and Applied Economics 10(4): 305-323, 2021 | e-ISSN 2280-6172 | DOI: 10.36253/bae-10465 Ghasem Layani et al. 3. SYSTEM DYNAMICS METHODOLOGY SD modeling is an iterative and feedback process to reach new understanding of how the problem arises and then design high leverage policies for improvement (Davies and Simonovic, 2011). A four-step SD modeling process introduced by Sterman (2001) and Ford and Ford (1999) is used in this study: (1) Problem articula- tion; (2) Model formulation; (3) Model testing; (4) Sce- nario design and simulation. The first step in SD mod- eling is to be specific about the dynamic problem and problem articulation (Ford and Ford, 1999). This step includes defining the problem, identifying the key vari- ables related to the problem, such as stocks, exogenous and endogenous variables, identifying the temporal and spatial scales to be considered (Zhuang, 2014). The aim of model formulation is representing the structure of the problem and formulating a SD simula- tion model of the causal theory (Sterman, 2001; Zhuang, 2014). There are several diagram tools to capture the structure of the system, including causal loop diagram (CLD) and stock and flow diagram. CLDs consist of var- iables connected by arrows for representing the feedback structure of the system (Sterman, 2001). In spite of the fact that stock and flow and feedback are the two central concepts of system dynamic theory, CLDs are not able to capture the stock and flow structure of a system (Ford and Ford, 1999; Sterman, 2001). This is an important reason for using stock and flow diagram to represent the structure of a system with more detailed information that is shown in a CLD. In general, the stock variable is an accumulator variable (Zhuang, 2014). A stock with a single inflow and single outflow can be mathematically formulated as: (1) Where s is any time between t0 and t. The stocks are the key variables in the model. They represent where accumulation or storage takes place in the system. Stocks tend to change less rapidly than other variables in the system, so they are responsible for the momentum or sluggishness in the system (Ford and Ford, 1999). Model testing begins as the first equation is writ- ten and it is a critical step in SD modeling (Sterman, 2001). Tests to rely on SD model can be divided into two groups, structure tests and behavior tests (Forrest- er, 1997). Structure tests compare the structure of the SD model with the available knowledge about the real system presented in historical data. Behavior test is to run the model and compare the results to the reference Figure 1. Kheirabad River Basin and Kowsar Dam. 309 Bio-based and Applied Economics 10(4): 305-323, 2021 | e-ISSN 2280-6172 | DOI: 10.36253/bae-10465 Sustainable water resources management in the Kheirabad river basin, Iran mode1 (Historical or observed data). When the simula- tion results match the reference mode, you have reached a major milestone in the modeling process (Ford and Ford, 1999). Following Kotir et al. (2016), mean relative errors (MRE) and coefficient of determination (R2) were applied to evaluate the performance of the model. MRE indicates the mean possible divergence between the observed and simulated data (Qin et al., 2011), the low- er values of MRE indicates that the model satisfactory fits the historical values. R2 describes the proportion of the variance in measured data explained by the model2 (Kotir et al., 2016). (2) (3) Where and are the observed and simulated val- ues of tested or variable and is the average of observed values of variable. After the validation of the model, we can use this model to evaluate the impact of different scenarios designed to solve the problem (Zhuang, 2014). 3.1. SD Modeling of Kheirabad River Basin 3.1.1. Water Supply Subsystem The water supply subsystem includes feedback rela- tionships between climate variables and water resources. This subsystem is constructed based on the surface and groundwater resources balance equation by taking in to consideration all inflows and outflows at the study area. This subsystem represents the measure of water resourc- es available at the basin (Hjorth and Bagheri, 2006). Surface water resources available are controlled by vari- ous factors such as measure of precipitation, runoff, water inflow and outflow of surface water, evaporation, transpiration and infrastructural conditions (Hjorth and Bagheri, 2006; Gohari et al., 2017). As shown in fig. 2, the water supply subsystem includes surface and groundwater resources. It is also worth mentioning that the surface and subsurface water inflows, return flow and precipitation are incoming inflows, and the surface and subsurface water outflows, evaporation, transpira- tion, water withdraw for kind of uses are outflows. Tem- perature and precipitation as climate variables affect 1. A reference mode is a pattern of behavior over time 2. The values of R2 range from 0 to 1, with values closer to 1 indicating that the model well simulates the system. the measure of available water. As a matter of fact, the increased precipitation can increase water availability. Strictly speaking, part of the precipitation is entered in to the water system as runoff (Eq. 4), taking into con- sideration of the runoff coefficient reported in the water balance studies of the study areas (Hjorth and Bagheri, 2006). Another part of the precipitation, joins to the groundwater resources considering the average perco- lation coefficient (Eq. 5). Also evaporation and transpi- ration was considered as a function of temperature in this study. Therefore, an increase of temperature in the future may affect the behavior of water resources system. Annual evaporation in water supply subsystem is meas- ured into available surface water multiplier in evapora- tion rate (Eq. 6). At each time step, the evaporation rate is taken from temperature at the basin which is repre- sented as a LOOKUP table3. Runoff=Runoff rate × Precipitation (4) Percolation=Percolation rate × Precipitation (5) Evaporation=Evaporation rate × Available surface water (6) Also, the return flow in water system, according to Eq. 7, is as a percentage of the water consumption in dif- ferent sectors that is added to the surface and ground- water resources. Total water withdrawal from the basin is measured into the sum of agricultural, domestic, envi- ronmental and industrial water demands. Following Davies and Simonovic (2011), domestic water demand is expressed as a function of population and per cap- ita water demand in the Kheirabad river basin model. Agricultural water demand is expressed as a function of cropland area and water requirement for each crop. Environmental water demand is assumed to be as an exogenous variable. For calculating industrial water demand, per capita industry water use is applied (Bala- li and Viaggi, 2015), in which industrial water demand equals population multiplier per capita industry water use. The amount of surface water withdraw is equal to the part of total water demand that is supplied from sur- face water sources. According to the report presented by the Regional Water Organization of Kogiluyeh and Boyerahmad province (2017), 49% of agricultural water demand, 66% of urban water demand and 51% of indus- 3. Lookup Tables are typically used in SD modeling to represent nonlin- ear relationships between two variables. A table function can be defined as a list of numbers whereby input values to a function are positioned relative to the x axis and output values are read from the y axis (Ford and Ford 1999; Vensim Reference Manual 2011). 310 Bio-based and Applied Economics 10(4): 305-323, 2021 | e-ISSN 2280-6172 | DOI: 10.36253/bae-10465 Ghasem Layani et al. trial water demand at the basin are supplied from sur- face water sources. Return flow = Return rate × Water demand of each sector (7) Total water demand = Agriculture D. + Domestic D. + Oil industry D. + Industrial D. + Environ- mental D. (8) Surface water withdraw = (water demandi × the share of surface water) (9) Ground water withdraw = (water demandi × the share of ground water) (10) 3.1.2. Population Subsystem Population is one of the factors that affect the water demand (Sušnik et al., 2012). Generally, population is the main driving factor in water demand. Popula- tion influence the domestic water demand directly and other sources of water demands indirectly (Davies and Simonovic, 2011). There are some towns and villages on the Kheirabad river basin. Most of the domestic water demand at the basin is provided by Kowsar dam. Also Kowsar dam supplied water to the Persian Gulf littoral cities and ports for nearly 20 years. Population sub-mod- el represents the population of the case study including one stock “Population” which is increasing by popula- tion growth rate. The population at time t is mathemati- cally represented by Eq. 11 as follows: population(t) = population(0) + (population growth rate)dt (11) In this study, the total population is divided into urban and rural population groups according to urbani- zation rates (Fig. 3). Therefore, the water demand in the urban sector equals urban population multiplier per cap- ita water consumption in the urban sector and similarly Available surface water Ground water availabilty Natural surface water inflow Run off Precipitation rate Evapotranspiration Surface water withdraw Surface water retern flow Ground water inflow Natural groundwater inflow + Groundwater return inflow + Ground water outflow Ground water withdrawal + total return flow + + Agricultural water demand Total water demand + Total domestic water demand Total industrial water demand + + Environmental water outflow Rate of percolatin Area of basin Volum of precipitation+ + + Run off rate + Percolation+ Minimum storage + Inflow + +