Land suitability evaluation for brackish water aquaculture development in coastal area of Hormozgan, Iran

DOI 10.1007/s10499-014-9818-y Land suitability evaluation for brackish water aquaculture development in coastal area of Hormozgan, Iran Abouzar Hadipour Freydoon Vafaie Vahid Hadipour Received: 18 July 2013 / Accepted: 5 August 2014 Ó Springer International Publishing Switzerland 2014 Abstract Land suitability analysis is a prerequisite for successful aquaculture, and site selection affects both the success and sustainability of any aquaculture development. There is an urgent need for appropriate methodology to assist planners for site selection in aquaculture development. Site selection can be viewed as a multi-criteria decision-making (MCDM) problem. The analytical hierarchy process (AHP) is a proven, effective method used to solve problem of site selection. This paper applied geographic information systems (GIS), the AHP method, and MCDM to identify areas that are suitable for shrimp aquaculture development in coastal area of Hormozgan, Iran. To create models, combination of layers was carried out through Boolean operators and weighted linear combination (WLC) method. After performing the combination models, the results are presented and compared. Evaluation of the results shows that the most of the areas classified suitable in WLC model coincide with the existing shrimp farms and this indicates the validity of the GIS-based WLC model. The areas with the highest priorities are situated in eastern part of the study area. Since existing shrimp farms cover a small extent in the study area, further expansion of shrimp farming to other areas is possible. Keywords GIS Site selection AHP Aquaculture WLC Introduction Aquaculture activities have been improved significantly in recent years aiming at the increased production target. The socioeconomic benefits derived from aquaculture expansion provide nutritious foods and improve lifestyle of the poor, income generation and employment opportunity, diversification of fish production, and create scope for foreign exchange earning through export of high-valued products (Hossain and Das 2010). A. Hadipour (&) F. Vafaie V. Hadipour Civil and Environmental Engineering Department, K.N.Toosi University of Technology, Tehran, Iran e-mail: abha571@yahoo.com

Despite its rapid growth, aquaculture development continues to be hindered by a number of constraints. These include limited suitable sites and multi-use conflicts (Radiarta et al. 2008). The first step for scientific and sustainable aquaculture development is appropriate site selection, and the success or failure of any aquaculture development largely depends on the right selection of the site. To ensure sustainable aquaculture development, there is, therefore, a great need to allocate aquaculture to suitable locations. Appropriate location of aquaculture development will minimize the risk of environmental impact, maximize the overall economic return and minimize conflict between aquaculture and other resource uses (GESAMP 2001). In recent years, the southern coasts of Iran have become the main focus for government investment in shrimp culture, and this industry is playing an important role in the economic and social welfare of coastal communities by creating job opportunities and improving revenues accrued from export. Due to favorable climatic conditions, suitable water quality, soil high natural productivity and space availability, the long southern coastline of Iran is regarded as highly suitable for shrimp culture development. Besides the above, this region possesses suitable environmental parameters in terms of pond water salinity and temperature, to have two harvests for most of the year annually. Despite the need noted above, no studies have addressed the selection of suitable sites through the use of geographic information systems (GIS) to develop shrimp farming in Iran, and traditional methods have continued to be used for this purpose. The traditional method is followed by individual site assessment and is time consuming. Geographic information system is a digital database management system designed to manage large volumes of spatially distributed data collected from a variety of sources. GIS are ideal for site selection studies because they efficiently store, retrieve, analyze and display information according to user-defined specifications (Guiqin et al. 2009). GIS is a particularly useful tool for coastal planners and managers to facilitate spatial decisionmaking process regarding aquaculture in order to optimize the use of natural resources. Applications of GIS in site selection are reported in the literature for various different aquaculture industries such as hard clam culture in Florida (Arnold et al. 2000), site selection for land-based shrimp farming in the Australian Coastal Zone (McLeod et al. 2002), marine fish cages within the tourism industry in Tenerife (Perez et al. 2003), shrimp and crab farming in Bangladesh (Salam et al. 2003), shrimp farming in Vietnam (Giap et al. 2005), Brackish water aquaculture site selection in India (Karthik et al. 2005), assessing suitable carp farming areas in Bangladesh (Salam et al. 2005), marine fish cage culture in Tenerife (Perez et al. 2005), single-use site selection for oyster culture in Venezuela (Buitrago et al. 2005), tilapia farming areas in Bangladesh (Hossain et al. 2007), Japanese scallop site selection in Japan (Radiarta et al. 2008; Radiarta and Saitoh 2009), urban aquaculture development in Bangladesh (Hossain et al. 2009) and land suitability modeling for giant prawn in Bangladesh (Hossain and Das 2010). This paper presents a GIS-based multi-criteria evaluation (MCE) modeling that uses the AHP method to identify the most suitable sites for shrimp farming development in the coastal areas of the Hormozgan province, Iran. The paper organized as follows. First, explanation of theoretical background including the AHP method and Combination models are introduced in Theorical background section. In Materials and methods section, the methodology including criteria identification, database development and weighting procedure presented. In Results and discussion section, results of models implementation and discussion presented. Finally, conclusions stated in Conclusion section.

Theorical background The multi-criteria decision-making (MCDM) methods deal with the process of making decisions in the presence of multiple criteria. An MCDM approach is often used to solve various decision-making and/or selection problems. This approach often requires the decision makers to provide qualitative and/or quantitative assessments for determining the performance of each alternative with respect to each criterion, and the relative importance of evaluation criteria with respect to the overall objective (Kuo et al. 2006). Site selection can be viewed as spatial MCDM problem. A spatial MCDM aims to achieve solutions for spatial decision problems, derived from multiple criteria. In the most general term, GIS-based MCDM involves the utilization of geographical data, the decision maker s preferences, and the combination of data and preferences according to specified decision rules (Malczewski 2006). Spatial MCDM is more complex and difficult in contrast to conventional MCDM, as large numbers of factors need to be identified and considered, with high correlated relationships among the factors (Malczewski 2004). The AHP method is widely used for examining MCDM problems in real situations. The AHP method has some advantages. One of the most important advantages of the AHP is based on pairwise comparison. Besides, the AHP calculates the inconsistency index, which is the ratio of the decision-makers inconsistency. Analytical hierarchy process (AHP) The analytical hierarchy process (AHP), first proposed by Saaty, is a popular method for solving multi-criteria analysis problems involving qualitative data (Deng 1999). Since its introduction, the AHP has become one of the most widely used MCDM methods (Lee et al. 2008). The AHP is a flexible and yet structured methodology for analyzing and solving complex decision-making problems by structuring them into a simple and comprehensible hierarchical framework (Boroushaki and Malczewski 2008). Pairwise comparison is the basic measurement procedure employed in the AHP method. This comparison is used in the decision-making process to form a reciprocal decision matrix, thus transforming qualitative data to crisp ratios, and making the process simple and easy to handle (Deng 1999). By making pairwise comparisons at each level of the hierarchy, participants can also develop relative weights to differentiate the importance of the criteria (Boroushaki and Malczewski 2008; Hossain et al. 2009). Saaty recommended a suitable measurement scale ranging from 1 to 9 for pairwise comparisons in which 1 means no difference in the importance of one criterion in relation to another, and 9 means one criterion is much more important than another (see Table 1). Reciprocals of these numbers are used to express the inverse relationship (Saaty 1980). An eigenvector method is used to solve the reciprocal matrix to determine the criteria importance and performance of each alternative (Saaty 1980). In order to measure the degree of inconsistency associated with the pairwise comparison matrix, the consistency index (CI) is calculated as follows: CI ¼ k max n n 1 ; ð1þ where k max is the biggest eigenvalue that can be obtained once its associated eigenvector is known and n is the number of columns of matrix A. Further, the consistency ratio (CR), which is defined as follows, can be calculated so:

Table 1 Scales for pairwise comparisons Intensity of importance Verbal judgment of preference 1 Equal importance 3 Moderate importance 5 Strong importance 7 Very strong importance 9 Extreme importance 2,4,6,8 Intermediate values between adjacent scale values CR = CI RI ; ð2þ where RI is the random index, i.e., the consistency index of a randomly generated pairwise comparison matrix. RI depends on the number of elements being compared (Saaty 1980). Saaty suggests that if the CR value is smaller than 0.10, it indicates a reasonable level of consistency in the pairwise comparison, and if it is larger than 0.10, there are inconsistencies and the AHP method may not yield meaningful results. The procedures of the AHP involve six essential steps (Lee et al. 2008), namely: 1. Define the unstructured problem and state clearly the objectives and outcomes. 2. Decompose the complex problem into a hierarchical structure with decision elements (criteria, detailed criteria, and alternatives). 3. Employ pairwise comparisons among decision elements and form comparison matrices. 4. Use the eigenvalue method to estimate the relative weights of decision elements. 5. Check the consistency property of the matrices to ensure that the judgments of decision makers are consistent. 6. Aggregate the relative weights of the decision elements to obtain an overall rating for the alternatives. Combination models In the context of GIS applications, a decision rule specifies how to combine a set of criterion maps so that alternative decisions (locations) can be ordered according to some preferences with respect to evaluation criteria (Malczewski and Rinner 2005). MCE is perhaps the most fundamental of decision support operations in GIS (Jiang et al. 2000). There are two fundamental classes of MCE methods in GIS: The Boolean overlay operations and the weighted linear combination (WLC) methods. They have been the most often used approaches for land use suitability analysis (Malczewski 2006). The primary reason for the popularity of these methods is that they are easy to implement within the GIS environment using map algebra operations. The methods are also easy to understand and intuitively appealing to decision makers (Malczewski 2004). In Boolean overlay, all criteria are assessed by thresholds of suitability to produce Boolean maps, which are then combined by logical operators such as intersection (AND) and union (OR). Given a set of suitability maps and corresponding threshold values, the Boolean intersection (AND operation) results in classifying areas as suitable for a

particular land use if each suitability map meets its threshold. Conversely, the Boolean union (OR) identifies suitable areas as those that meet at least one suitability threshold value (Malczewski 2004). The WLC approach involves standardization of the suitability maps, assigning the weights of relative importance to the suitability s maps, and then combining the weights and standardized suitability maps to obtain an overall suitability score (Malczewski 2004). The result is a continuous suitability mapping. The higher the score, the more suitable the location is for the intended land use. In this combination method, the weights derived from the AHP method were combined with the suitability score and both were applied to spatial modeling created using the GIS (Malczewski 2004). The weight of the relative important assigns to each parameter and a total score (A i ) is obtained for each parameter by multiplying the weight assigned by the scale value for that parameter, and summing the product over all parameters as follows: A i ¼ X j w j x ij ð3þ where x ij is the score of the ith alternative with respect to the jth attribute, and the w j is the normalized weight, so that P w j ¼ 1. Materials and methods Study area This study was conducted in the Hormozgan province situated in the southern part of Iran and covers approximately 258000 ha. The scale of maps is 1:25,000, and pixel size is 50 9 50 m. The geographical location of the study area is shown in Fig. 1. The Hormozgan province has favorable environmental conditions for aquaculture and is the most important center of shrimp farming in the southern part of Iran. Regarding the long coastline and climatic conditions, this province has high potentials for shrimp aquaculture development. Software used The software used in this study included MATLAB version 7.0 and ArcGIS 9.2. To compute the weight by the AHP method; programming was done in MATLAB 7.0. Also data processing, preparation of the layers, and modeling were performed with ArcGIS 9.2 Identification of criteria and suitability rating The selection of a suitable site for aquaculture development is not based on one criterion but on multiple criteria. The key factors to be considered for selecting the optimal location in aquaculture are the availability of good quality water, soil quality, water salinity, water temperature, distance from pollution source, exposure to flood, infrastructural facilities, and access to essential inputs and markets (Nath et al. 2000; Mahalakshmi and Ganesan 2009). Desirable water quality and quantity are perhaps the most important requirements of aquaculture. The water supply must be of adequate quantity to fill the pond and maintain

Aquacult Int Fig. 1 Geographical location of the study area in Hormozgan, Iran the water level and of sufficient quality to provide an environment that is suitable for aquaculture (Hajek and Boyd 1994). The water source should be as close to the farm site as possible, so that provision of supplies is easily done, and costs are reduced. It is also

important that there should not be any industrial and pesticide pollution in the area surrounding the shrimp site. Properties of soils should be considered in selecting a site, designing earthworks, and specifying construction methods to provide a water-tight pond with stable levees, and bottom slopes (Hajek and Boyd 1994). A satisfactory pond bottom soil, apart from being impervious to water, permits rapid mineralization of organic matter, absorbs nutrients, loosely binds and releases them over a long period. Social and infrastructural factors also affect shrimp farming operations (Nath et al. 2000). A good network of roads is a prerequisite for shrimp farms, transport of shrimps, food, and necessary equipment. Hence, shrimp farming sites should be as close to the road system as possible to enable fast access. The shrimp farm should be located where the product is in high demand. After harvesting, shrimps require prompt marketing, otherwise, their quality deteriorates. Thus, shrimp farm proximity to markets is important. To establish a shrimp farm, the source of fries must be ensured. This source should be close to the farm, because fries are vulnerable to adverse environmental conditions and mortality usually increases with transportation distance. In this study, the most important criteria for site selection of shrimp farming (12 criteria) were selected and scored through reviewing the literatures and holding consultations with aquaculture experts. Thematic maps were prepared for each of the criteria. Then, the prepared thematic maps (base layers) were classified into three categories: (1) engineering parameters (slope, land use, soil texture, and elevation), (2) water quality and quantity (distance to water source, water temperature, water salinity, and distance from pollution source), and (3) infrastructure parameters (distance to roads, to markets, to processing plants, and to hatcheries). These maps were classified according to suitability ratings. Suitability is a measure of how well the qualities of a land unit match the requirements of a particular form of land use (FAO 1976). Suitability ratings were established according to FAO classification on the appropriateness of land for defined uses. These ratings are as follows: highly suitable, suitable, moderately suitable, and unsuitable (Kapetsky 1994; FAO 1976). In this paper, the same framework has been incorporated with addition of the more number of parameters like socioeconomic and infrastructure. The suitability ratings for these criteria were determined from different literatures and were modified, based on experts opinions to suit the Iranian environmental condition. Since the criterion maps contain the ordinal values (high, medium, and low) that indicate the degree of land suitability with respect to a particular criterion, the maps were standardized. Making the scores of the criteria compatible is often called standardization. There are many different standardization methods that can be used in GIS-based multi attribute analysis. The method to use depends on the feature of the problem and the feature of the criteria. Pairwise comparison technique can be used for the purpose of standardizing these ordinal values (Malczewski 2004). In this paper, the criteria are standardized using pairwise comparison technique. The standardization of criteria resulted in rating between 0 and 1. Suitability ratings and scores of criteria are presented in Table 2. Data source and database generation The data were collected from a variety of sources. The primary data sources used in this study include a Landsat7 satellite image, topographic maps on the scale of 1:25,000, and

Table 2 Suitability rating and scores of engineering, water, and infrastructure parameters for shrimp farming site selection Criteria Highly suitable Suitable Moderately suitable Unsuitable Land use type Aquaculture pond Poor range, salt farm and bare land Agricultural land Mangrove forest Slope (%) \2 2 5 5 10 [10 Elevation (m) 2 2.5 1 2 or 2.5 4 4 5 [5 or\1 Soil texture (% clay) [35 18 35 \18 Distance to water source (km) \1 1 2 2 4 [4 Water temperature ( C) 28 32 32 34 or 22 28 34 36 or 15 22 [36 or \15 Water salinity (ppt) 30 40 40 45 or 20 30 45 50 or 10 20 [50 or \10 Distance to pollutant source (km) [4 3 4 2 3 \2 Distance to market (km) \3 3 7 7 12 [12 Distance to road (km) \2 2 3 3 5 [5 Distance to processing plant (km) \3 3 7 7 12 [12 Distance to hatcheries (km) \3 3 7 7 12 [12

statistical data obtained from various reports, soil types data, and the results of socioeconomic surveys. All data included in the database needed modification and reclassification to create thematic layers. Land use pattern was derived from the Landsat7 satellite image. Elevation and slope data were generated from digital elevation model (DEM). A DEM is a digital map or 3D representation of a terrain s surface created from terrain elevation data. Layers for land use type, road networks, water source, hatcheries, processing plant, and market location were extracted from topographical maps and satellite image. Preparation of the soil type map, water temperature, and salinity maps was carried out by transformation of their data in the GIS environment. The layer of pollution source was created through the determination of the potential polluting industries, and integration of their coordinates into the GIS database. All vector-based data layers were converted into grids with 50 m 9 50 m cell size. The ArcGis raster analysis module was used to overlay and classify the layers, on which spatial logic and algebraic computations were performed. The grid layers were then reclassified according to their suitability scores. The reclassified grid layer of each criterion was multiplied by the respective weight calculated in the pairwise comparison matrix by the AHP method. Finally, these criteria layers were combined to develop suitability maps. Construction of hierarchy and weighting procedure The model structure for site selection of shrimp aquaculture in Hormozgan was built on the basis of hierarchical structures. Hierarchical structures break down all criteria into smaller groups (or submodels). To break down a hierarchy into clusters, first it was decided which elements to group together in each cluster. This was done according to the similarity of the elements with respect to the function they perform or the properties they share (Saaty 1980). The weight for each factor was determined by pairwise comparisons in the context of the AHP method. To assign weights by pairwise comparison, questionnaires were used. Eighteen experts were asked to evaluate the suitability rating and the weight of each criterion according to the check list. These experts were selected from universities, research institutions, government agencies, and private companies at national or local levels who have useful information in fields thought to be related in some way with aquaculture development. They were specialized in some related sciences such as civil engineering, fisheries science, biology, water resource management, environmental science, economic, and natural resource management. The top or first level in the hierarchy represents the ultimate goal of the MCDM analysis process. The intermediate or second hierarchy level lists the relevant evaluation criteria that were compared pairwise to assess their relative weights. Each of these clusters was regarded as a submodel. The lowest level in the hierarchy contains the evaluation objects. All these criteria are identified as affecting the goal of the study and may represent primary data or be the results of secondary data. Figure 2 represents the suitability analysis for the selection of shrimp farming sites in Hormozgan as a hierarchical structure. Results and discussion For models preparation, the criteria were weighted and scored in terms of their significance for shrimp farming. The pairwise comparison matrix and calculated weights of criteria are shown in Tables 3, 4, 5 and 6. The consistency ratios (CR) of 0 0.03 for the tables were

Fig. 2 Hierarchical Schematic diagram for modeling shrimp farming site selection in Hormozgan, Iran well within the ratio less than 0.10 signify that consistency is acceptable. Suitability maps of engineering, water, and infrastructure submodels generated for shrimp aquaculture development in the Hormozgan province, Iran are shown in Figs. 3, 4 and 5, respectively. There are two commonly used classes of map combination (overlay) operations in GIS: Boolean overlay and weighted linear combination (WLC) (Malczewski 2004). In this study, both of these two combination models were used and after performing the combination models, results of them were presented and compared. In Boolean model, combination of layers through Boolean intersection (AND) and union (OR) operators were carried out. The map resulting from the combination of layers with Boolean intersection is presented in Fig. 6. The Boolean method of characterizing the criteria is too black and white. Boolean intersection results in a very strict assessment, i.e., a region with only one criterion falling short of the threshold of the intersection is regarded as unsuitable. Conversely, the Boolean union operator employs a very liberal attitude, i.e., a region will be regarded as unsuitable with only one criterion meeting the threshold. To implement WLC model, the weight for each criterion is determined by AHP method. The model computes the CR associated with the pairwise comparison matrix. The resulting weights can then be used as input for the WLC model. The suitability ratings were established according to the FAO classification. The map prepared by applying WLC model to select sites for shrimp aquaculture in the Hormozgan province is shown in Fig. 7. The final suitability maps produced from models were verified to ensure that the models were suitable with the actual conditions in the field. Verification of the model was carried out by drawing comparisons between predicted suitable sites and locations of existing

Table 3 A pairwise comparison matrix for assessing relative importance of engineering factors for shrimp farming site selection in Hormozgan, Iran (numbers show the rating of the row factors relative to the column factor) Land use type Slope Elevation Soil texture Weight Land use type 1 2 5/2 2 0.41 Slope 1/2 1 2 1/2 0.19 Elevation 2/5 1/2 1 1/2 0.13 Soil texture 1/2 2 2 1 0.27 Consistency ratio (CR) = 0.03 Table 4 A pairwise comparison matrix for assessing relative importance of water quality and quantity factors for shrimp farming site selection in Hormozgan, Iran (numbers show the rating of the row factors relative to the column factor) Distance to water source Water temperature Water salinity Distance to pollutant source Weight Distance to water 1 1/2 2/3 2 0.20 source Water temperature 2 1 3/2 3 0.39 Water salinity 3/2 2/3 1 3 0.30 Distance to pollutant source 1/2 1/3 1/3 1 0.11 Consistency ratio (CR) = 0.01 Table 5 A pairwise comparison matrix for assessing relative importance of infrastructure factors for shrimp farming site selection in Hormozgan, Iran (numbers show the rating of the row factors relative to the column factor) Distance to market Distance to road Distance to processing plant Distance to hatcheries Weight Distance to market 1 2/5 2 2 0.25 Distance to road 5/2 1 5/2 3 0.46 Distance to 1/2 2/5 1 3/2 0.16 processing plant Distance to hatcheries 1/2 1/3 2/3 1 0.13 Consistency ratio (CR) = 0.02 farms. There are some farms in the eastern part of the study area such as Tiyab farm. These sites established many years ago and served as reference to evaluate models efficiency. An important assumption was that local people had a preference for more suitable locations based on their indigenous knowledge (Hossain et al. 2007). Comparison of methods is carried out according to existing shrimp farms. It is apparent from Figs. 6 and 7 that the results from the Boolean and the weighted linear combination methods are very different. The results show that the Boolean model is not appropriate to select shrimp farming sites, because the model is not capable to prioritize suitable areas. The WLC model is more effective and flexible than the Boolean method. It allows for

Table 6 A pairwise comparison matrix for assessing relative importance of land use requirements for shrimp farming site selection in Hormozgan, Iran (numbers show the rating of the row factors relative to the column factor) Engineering Water quality and quantity Infrastructure Weight Engineering 1 2/3 2 0.34 Water quality and quantity 3/2 1 5/2 0.48 Infrastructure 1/2 2/5 1 0.18 Consistency ratio (CR) = 0.00 Fig. 3 Suitability maps of engineering submodel for shrimp aquaculture development in the Hormozgan province, Iran Fig. 4 Suitability maps of water quality and quantity submodel for shrimp aquaculture development in the Hormozgan province, Iran Fig. 5 Suitability maps of infrastructure submodel for shrimp aquaculture development in the Hormozgan province, Iran

Fig. 6 Overall site selection map prepared by application of Boolean intersection model for selecting shrimp aquaculture sites in the Hormozgan province, Iran Fig. 7 Overall site selection map prepared by the application of WLC model for selecting shrimp aquaculture sites in the Hormozgan province, Iran criteria to be standardized in a continuous fashion, differentially weighted, and to tradeoff with each other. Conclusion This paper used GIS-based processing and analysis, along with the AHP method to identify areas that are suitable for aquaculture activities in a study area located in Iran. Combination of layers was achieved through Boolean operators and WLC method. Comparison of Boolean and WLC model which was carried out by model verification shows that the latter is more accurate and flexible to evaluate suitable sites for shrimp farming development. Unlike the Boolean operations, WLC is a compensatory method in the sense that a low score of one criterion can be compensated for a high score of another. The WLC combination model is a straightforward application and can easily be integrated spatially into GIS using raster-based map algebra. It allows the criteria to be differentially weighted and to tradeoff with each other. Most of the areas classified suitable in WLC model coincide with the existing shrimp culture farms, and this shows the validity of the GIS-based WLC model for shrimp farming site selection. The applied method including the AHP is effective approach for aquaculture site selection and can be applied to other region. The areas with the highest priorities are situated in the eastern part of the study area in which most of the criteria are favorable for shrimp farming and coincide with each other. Since existing shrimp farms cover a small extent in the study area, further expansion of shrimp farming to other areas is possible.

Results can be useful for policy makers and decision makers in Hormozgan province. It must be noted that in making final decisions, socioeconomic and cultural factors which allow coastal planners to make better decisions must be taken into account. On the ground, verification is the most reliable and also the most time consuming. Such an approach is appropriate to verify individual sites after the GIS have been employed to identify the most promising areas. Acknowledgments The authors would like to acknowledge the support of K.N.Toosi University of Technology in Iran (Tehran) and Department of Civil and Environmental Engineering. Conflict of interest The authors have declared no conflict of interest. References Arnold WS, White MW, Norris HA, Berrigan ME (2000) Hard clam (Mercenaria spp) aquaculture in Florida, USA: geographic information system applications to lease site selection. Aquac Eng 23:203 231 Boroushaki S, Malczewski J (2008) Implementing an extension of the analytical hierarchy process using ordered weighted averaging operators with fuzzy quantifiers in ArcGIS. Comput Geosci 34:399 410 Buitrago J, Rada M, Hernandez H, Buitrago E (2005) A single-use site selection technique, using GIS, for aquaculture planning, choosing locations for mangrove oyster raft culture in Margarita Island, Venezuela. Environ Manage 35:544 556 Deng H (1999) Multi-criteria analysis with fuzzy pairwise comparisons. Int J Approx Reason 21:215 231 FAO (1976) A framework for land evaluation. Food and Agriculture Organization of the United Nations, Rome GESAMP (IMO/FAO/UNESCO-IOC/WMO/WHO/IAEA/UN/UNEP) (2001) Joint group of experts on the scientific aspects of marine environmental Protection. Planning and management for sustainable coastal aquaculture development. FAO Rep. Stud. GESAMP No. 68, 90 pp Giap DH, Yi Y, Yakupitiyage A (2005) GIS for land evaluation for shrimp farming in Haiphong of Vietnam. Ocean Coast Manag 48:51 63 Guiqin W, Li Q, Guoxue LM (2009) Landfill site selection using spatial information technologies and AHP: A case study in Beijing, China. J Environ Manage 90:2414 2421 Hajek BF, Boyd CE (1994) Rating soil and water information for aquaculture. Aquac Eng 13:115 128 Hossain MS, Das NG (2010) GIS-based multi-criteria evaluation to land suitability modelling for giant prawn (Macrobrachium rosenbergii) farming in Companigonj Upazila of Noakhali, Bangladesh. Comput Electron Agric 70:172 186 Hossain MS, Chowdhury SR, Das NG, Rahaman MM (2007) Multi-criteria evaluation approach to GISbased land suitability classification for tilapia farming in Bangladesh. Aquac Int 15:425 443 Hossain MS, Chowdhury SR, Das NG, Sharifuzzaman SM, Sultana A (2009) Integration of GIS and multicriteria decision analysis for urban aquaculture development in Bangladesh. Landsc Urban Plan 90:119 133 Jiang H, Eastman JR (2000) Application of fuzzy measures in multi-criteria evaluation in GIS. Int J Geogr Inf Sci 14(2):173 184 Kapetsky JM (1994) A strategic assessment of warm-water fish farming potential in Africa, CIFA Technical Paper, No. 27 edn. FAO, Rome Karthik M, Suri J, Saharan N, Biradar RS (2005) Brackish water aquaculture site selection in Palghar Taluk, Thane district of Maharashtra, India, using the techniques of remote sensing and geographical information system. Aquac Eng 32:285 302 Kuo RJ, Liang GS, Huang WC (2006) Extensions of the multicriteria analysis with pairwise comparison under a fuzzy environment. Int J Approx Reason 43:268 285 Lee AHI, Chen WC, Chang CJ (2008) A fuzzy AHP and BSC approach for evaluating performance of IT department in the manufacturing industry in Taiwan. Expert Syst Appl 34:96 107 Mahalakshmi P, Ganesan K (2009) Mahalanobis Taguchi System based criteria selection for shrimp aquaculture development. Comput Electron Agric 65(2):192 197 Malczewski J (2004) GIS-based land-use suitability analysis: a critical overview. Prog Plann 62:3 65

Malczewski J (2006) GIS-based multi-criteria decision analysis: a survey of the literature. Int J Geogr Inf Sci 20:703 726 Malczewski J, Rinner C (2005) Exploring multi-criteria decision strategies in GIS with linguistic quantifiers: a case study of residential quality evaluation. J Geogr Syst 7:249 268 Mcleod I, Pantus F, Preston N (2002) The use of a geographical information system for land-based aquaculture planning. Aquac Res 33:241 250 Nath SS, Bolte JP, Ross LG, Aguilar-Manjarrez J (2000) Applications of geographical information systems (GIS) for spatial decision support in aquaculture. Aquac Eng 23:233 278 Perez OM, Telfer TC, Ross LG (2003) Use of GIS-based models for integrating and developing marine fish cages within the tourism industry in Tenerife (Canary Islands). Coast Manag 31:355 366 Perez OM, Telfer TC, Ross LG (2005) Geographical information system-based models for offshore floating marine fish cage aquaculture site selection in Tenerife, Canary Islands. Aquac Res 36:946 961 Radiarta IN, Saitoh SI (2009) Biophysical models for Japanese scallop, Mizuhopecten yessoensis, aquaculture site selection in Funka Bay, Hokkaido, Japan, using remotely sensed data and geographic information system. Aquac Int 17:403 419 Radiarta IN, Saitoh SI, Miyazono A (2008) GIS-based multi-criteria evaluation models for identifying suitable sites for Japanese scallop (Mizuhopecten yessoensis) aquaculture in Funka Bay, southwestern Hokkaido, Japan. Aquaculture 284:127 135 Saaty TL (1980) The analytic hierarchy process: planning, priority setting, resource allocation. McGraw- Hill, New York Salam MA, Lindsay GR, Beveridge MCM (2003) A comparison of development opportunities for crab and shrimp aquaculture in south-western Bangladesh, using GIS modeling. Aquaculture 220:477 494 Salam MA, Khatun NA, Ali MM (2005) Carp farming potential in Barthatta Upazilla, Bangladesh: a GIS methodological perspective. Aquaculture 245:75 87