A multiple criteria decision analysis for agricultural planning of new crop alternatives in Turkey

Abstract

Agriculture is a crucial and strategic sector for developing countries. The agricultural sector in Turkey has been suffering from regression in recent years due to several reasons. In an attempt to reverse this process, we analyze the cultivation possibilities of high profit-margin crops in Turkish lands and develop a ranking among eight alternative crops. To perform a comprehensive analysis encompassing several dimensions, three MCDM methods are utilized; namely fuzzy AHP to determine the weights of evaluation criteria, and TOPSIS and PROMETHEE to develop a ranking among the crop alternatives. The crop alternatives are evaluated against several economic, technical, social and environmental criteria. The results favor the cultivation of soy bean, goji berry and buckwheat, while tamarind appears to be the least favored crop among the considered alternatives. The analysis results are enhanced with a sensitivity analysis.

Keywords

Multi criteria decision making fuzzy AHP TOPSIS PROMETHEE agricultural planning sensitivity analysis

1 Introduction

Agriculture has always been a crucial and strategic sector for developing countries [26]. Turkey, which was once a self-sufficient country in terms of agricultural products, has recently lost this peculiarity and became highly export-dependent to satisfy the domestic demand. The reasons for this occurrence are manifold, ranging from restricted agricultural subsidies after IMF-induced policies in the 2000 s to elevated input costs in agriculture, small farm sizes resulting in diseconomies of scale, the low education level among farmers, and their lack of technical know-how, as well as limited irrigation opportunities [15]. Many farmers have now left their land, migrated to bigger cities, so the agricultural sector of Turkey has been suffering from abandonment of fertile, arable land.

In this paper, we would like to inquire whether it is possible to revert this process by improving the producer welfare via the cultivation of high profit-margin crops. Towards this end, we have selected 8 crops that are not widely known by farmers in Turkey although the domestic demand for these plants (as well as the global demand) increases day by day. Moreover, these crops sell at higher prices compared to many other crops cultivated extensively in Turkish lands today, leaving a high profit margin for the farmers. We would like to analyze whether cultivating one or some of these crops could be a solution for reinvigorating the Turkish agriculture sector and help alleviate the poverty problem of Turkish farmers. In our analysis, we do not only take into account the demand and wealth potential of these crops, but we also consider the issue from several other perspectives such as the appropriateness of the Turkish lands to cultivate each one of these crops, the labor requirement and expertise, and environmental effects among others. A fuzzy Analytical Hierarchy Process (AHP) analysis is utilized to determine the weights of the selected criteria to evaluate the cultivation decision; and TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) and PROMETHEE II (Preference Ranking Organization Method for Enrichment Evaluations) methods are used to rank the crop alternatives. Fuzzy AHP method is selected for determining the criteria weights, since AHP combined with the fuzzy set theory is proven to achieve more flexibility in judgment and decision-making. Fuzzy AHP retains many of the advantages enjoyed by conventional AHP, in particular multiple criteria are considered smoothly and easily with this method, and a hierarchical structure is provided via pairwise comparison [49]. Furthermore, this method reflects the ambiguity in human cognition process, and takes into account the uncertainty in generating decisions. TOPSIS and PROMETHEE methods are selected mainly because they are well-known, well-renowned methods in the MCDM literature that prove to be successful in generating rankings of selected alternatives [43].

Furthermore, by the means of a sensitivity analysis conducted on the selected methods, the robustness of proposed solutions is discussed and the crop alternatives are further evaluated from various dimensions. The results could be illuminating not only for Turkish farmers, but also for the farmers in other countries who experience similar adversities as their Turkish counterparts. Moreover, our recommendations will be helpful for the policy-makers and experts in the agricultural sector in terms of developing more innovative, option-rich agricultural policies in the near future, hopefully leading to a stronger agricultural sector in Turkey.

The paper is structured as follows: In Section 2, a literature review regarding the applications of Multi-Criteria Decision Making (MCDM) methods in agriculture is presented; and the position of this work is stated. In Section 3, the details of our methodology are discussed. In Section 4, the background information regarding the relevant crop alternatives and the set of criteria are given, and the analysis is explained in detail. Finally, in Section 5, the results and potential policy implications for the policy-makers are discussed.

2 Literature review

MCDM models have been frequently applied in agricultural settings. They are used to solve various problems such as optimal site selection of the agricultural product warehouses [20], selection of post-harvest technology in cocoa production [31], developing climate decision support systems for tomato production [22], and the decision-making problem among different farming systems in a watershed [39]. Moreover, many researchers combine MCDM methods with Geographic Information Systems (GIS) techniques to identify the suitable zones for agriculture within a larger region [33 , 38].

Despite the wide range of MCDM applications in agriculture, the works which consider the selection problem among potential crop alternatives are particularly relevant for our research. Biswas and Pal [5] utilize fuzzy goal programming in an attempt to tackle the land-use planning problem for production of the principal crops of the Nadia District in West Bengal. They compute the priority levels and optimal production plans for these crops. Gupta et al. [21] formulate a multi-objective fuzzy linear programming (MOLFP) area allocation model. They consider various conflicting objectives involved in irrigation planning, and select crops based on the criteria listed as domestic needs of a particular crop commodity, productivity, market value, regional balance, and resource requirements and their availability. According to the multi-attribute value technique, they find that the pigeon-pea crop stood at first rank followed by a high yielding wheat variety while fodder and cotton occupied the last two positions. Qiu [39] assesses the economic and environmental impacts of farming decisions in an agricultural watershed, and identifies the selection criteria as increasing net returns (NR), reducing economic risk (ER), improving drinking water quality (DW), enhancing aquatic ecosystem health (AE), and reducing soil loss (SL). Yuan et al. [53] consider crop planning subject to value at risk constraint in a fuzzy environment. They develop a heuristic algorithm integrating approximation approach (AA), neural network (NN) and genetic algorithm (GA) to solve the fuzzy crop production planning VaR model.

Cobuloğlu and Büyüktahtakin [12] consider the problem of selecting the most sustainable biomass crop type for biofuel production and propose a stochastic AHP model towards its solution. They apply the proposed model to biomass alternatives including switchgrass, miscanthus, sugarcane, corn, and wheat in Kansas. As a result, the score of switchgrass increases if environmental criteria are emphasized, whereas wheat and corn become more favorable if priority is given to economic factors.

In an attempt to derive the most appropriate crop to cultivate in Turkish lands, we use similar ideas discussed in this line of research. As in the aforementioned papers, several economic, technical, social, and environmental criteria are determined to judge the crop alternatives. These criteria are stated in Section 4.1 and weighted by using fuzzy AHP method in Section 4.2. Next, the crop alternatives are evaluated according to the given set of criteria, and the results are triangulated by using more than one MCDM method, namely TOPSIS and PROMETHEE, in Sections 4.3 and 4.4 respectively.

Despite these similarities, our work differs from the above line of research in several ways. Most of the past studies evaluate the crop alternatives regarding their cultivation possibilities in a small agricultural region. In contrast, our main objective in this work is to propose a holistic agricultural strategy for Turkish agriculture. Secondly, the crop alternatives evaluated in other studies are often already being cultivated in the given regions; while we want to introduce and evaluate crops that are unknown (or narrowly known) by Turkish farmers. Furthermore, the set of evaluation criteria we consider is widely comprehensive, ranging from environmental to technical, social and economic dimensions.

The methodology implemented in this paper is also unique in many aspects. Although fuzzy AHP, TOPSIS and PROMETHEE are well-known and highly utilized methodologies in the literature, a combination of the three methods in a single study is not often observed. Furthermore, the sensitivity analysis presented at the end of the analysis provides further robustness on the proposed recommendations of this study. We believe that both the methodology and the results could be illuminating for the policy-makers both in Turkey and around the world.

3 Methodology

In this research, first fuzzy AHP, then TOPSIS and PROMETHEE methods are utilized to analyze the problem. Figure 1 summarizes the steps of the proposed algorithm.

Fig. 1

Procedure of the proposed methodology.

3.1 Fuzzy AHP

AHP is among the most prominently used multiple criteria decision-making techniques in the literature [41]. It enables building hierarchical relationships among different criteria to judge possible solution alternatives. In order to capture the ambiguity in human thinking style in a more accurate manner, fuzzy AHP, a fuzzy extension of AHP, was developed and several versions of this extension were tailored to address problems in many contexts [46]. In this work, we will utilize the same approach and compute the weights of the evaluation criteria using fuzzy AHP. The method is applied in accordance with the procedure described by Chang [10]. The steps of the application procedure are summarized below.

Step 1. Let X ={ x₁, x₂, x₃, …, x_n } be the criteria set and U ={ u₁, u₂, u₃, …, u_m } be the goal set. According to the method, an extent analysis is performed on each goal per each criterion, respectively. Therefore, one obtains m extent analysis values for each criterion, i.e. $M_{gi}^{1}, M_{gi}^{2}, M_{gi}^{3} \dots M_{gi}^{m}$ , for all i = 1, 2, … n. These values are triangular fuzzy numbers (i.e. expressed by three values $M_{gi}^{j} = (l, m, u)$ ) and $M_{gi}^{j}$ shows the triangular fuzzy number related to goal j according to criterion i.

Step 2. Next, the value of fuzzy synthetic extent with respect to the i-th criterion is defined according to the following equation: $S_{i} = \sum_{j = 1}^{m} M_{gi}^{j} ⊙ {[\sum_{i = 1}^{n} \sum_{j = 1}^{m} M_{gi}^{j}]}^{- 1}$ (1)

To obtain ${[\sum_{i = 1}^{n} \sum_{j = 1}^{m} M_{gi}^{j}]}^{- 1}$ , it is necessary to perform the fuzzy addition operation on m extent analysis values and compute the inverse of the resulting vector, i.e.: ${[\sum_{i = 1}^{n} \sum_{j = 1}^{m} M_{gi}^{j}]}^{- 1} = (\frac{1}{\sum_{j = 1}^{m} l_{j}}, \frac{1}{\sum_{j = 1}^{m} m_{j}}, \frac{1}{\sum_{j = 1}^{m} u_{j}})$ (2)

Step 3. In this phase, the significance vector is computed to measure the likelihood of one triangular fuzzy number being larger than another. To be more specific, the degree of possibility of M₁ ⩾ M₂ is defined as: $V (M_{1} ⩾ M_{2}) = {\begin{matrix} \begin{matrix} 1, & m_{1} > m_{2} \end{matrix} \\ \begin{matrix} 0, & l_{2} > u_{1} \end{matrix} \\ \begin{matrix} \frac{(l_{2} - u_{1})}{(m_{1} - u_{1}) - (m_{2} - l_{2})}, & otherwise \end{matrix} \end{matrix}$ (3) where M₁ = (l₁, m₁, u₁) and M₂ = (l₂, m₂, u₂).

The possibility of a fuzzy number being larger than the other n-1 fuzzy numbers can be stated by the significance vector W = (d (A₁) , d (A₂) , … d (A_n)) ^T. This vector is attained by normalizing W’ = (d’ (A₁) , d’ (A₂) , … d’ (A_n)) ^T, where d′ (A_i) = minV (S_i ⩾ S_k) , k = 1, 2, . . n, and k ≠ i.

Elements of W are calculated according to the formula: $d (A_{i}) = \frac{d^{'} (A_{i})}{[d^{'} (A_{1}) + d^{'} (A_{2}) + \dots + d^{'} (A_{n})]}, i = 1, 2, . . n$ (4)

3.2 TOPSIS and PROMETHEE

Among several MCDM solution methodologies, TOPSIS and PROMETHEE methods are among the most frequently used ones. TOPSIS was initially developed by Hwang and Yoon in 1981 [23] as an alternative to ELECTRE (ELimination Et Choix Traduisant la REalité) method. The main idea in this methodology is to calculate the proximity of decision alternatives to the ideal solution. Ideal solution is defined as the best performance on each criterion; however, in general, different alternatives produce the ideal solution under each criterion making this value impossible to reach by selecting a single decision alternative. Therefore, the decision-maker proceeds with selecting the alternative in closest proximity to the ideal solution. Towards that end, the vector of “Positive Ideal Solution (PIS)”, which maximizes the profits and the vector of “Negative Ideal Solution (NIS)”, which maximizes the costs, are developed. According to TOPSIS, the selected point should be closest to PIS, and farthest from the NIS [35]. The general procedure consists of the following steps:

Step 1. Given the decision matrix, D, the “normalized decision matrix”, ND, is computed. The ij^th entry (i.e. the value on the i^th row and j^th column) of ND is computed as: ${ND}_{ij} = \frac{D_{ij}}{{(\sum_{i = 1}^{I} D_{ij}^{2})}^{\frac{1}{2}}}, \forall i, j$ (5)

Step 2. Given the normalized decision matrix, ND, the “weighted normalized decision matrix”, V, is computed. The ij^th entry of V is calculated as: $V_{ij} = {ND}_{ij} * w_{j}, \forall i, j$ (6) where w_j is the weight of the evaluation criterion j (computed by AHP procedure), ∑_jw_j = 1.

Step 3. Given the weighted normalized decision matrix, V, the “positive ideal solution”, PIS, and “negative ideal solution”, NIS, are computed. $\begin{matrix} PIS = {v_{1}^{*}, v_{2}^{*}, \dots v_{n}^{*}} = {max_{j} v_{ij} | j ɛ J}, \\ NIS = {v_{1}^{-}, v_{2}^{-}, \dots} = {min_{j} v_{ij} | j ɛ J} \end{matrix}$ (7) where $v_{j}^{*}$ , j ɛ 1,2, ... n defines the value of the positive ideal solution with respect to j^th criterion, and $v_{j}^{-}$ , j ɛ 1,2, ... n defines the value of the negative ideal solution with respect to j^th criterion; both of which are computed from the respective columns of the matrix V.

Step 4. Next, separation measures are calculated. The separation of each alternative from the positive ideal solution is given as: ${PID}_{i} = \sqrt{\sum_{j} {(v_{ij} - v_{j}^{*})}^{2}}, \forall i$ (8) and the separation of each alternative from the negative ideal solution is given as: ${NID}_{i} = \sqrt{\sum_{j} {(v_{ij} - v_{j}^{-})}^{2}}, \forall i$ (9) where v_ij is the ij^th entry of matrix V, namely defining the value of alternative i with respect to criterion j.

Step 5. Finally, the relative closeness to the ideal solution is computed for each alternative. The closeness of the alternative i to the ideal solution set is defined as: ${CIS}_{i} = \frac{{NID}_{i}}{({PID}_{i} + {NID}_{i})}$ (10)

Clearly, the closer CIS value of an alternative is to 1, the relatively closer it will be to the ideal solution. Conversely, the closer CIS value is to 0, the closer it is to the negative solution. The alternatives can then be ranked according to their closeness to the ideal solution.

PROMETHEE is proposed by Jean Pierre Brans in 1982 [6], and has been updated as better versions later. We use the version developed in 1985 [7]. In this method, pairwise comparisons of alternatives are performed under each criterion, and common preference functions are computed. There are six different distributions to develop the preference functions. In our paper, we used type 5, namely “Criterion with Linear Preference and Indifference Area”, in order to create smoother comparisons rather than extreme “all or none” type of preferences. This preference function is defined as follows. $p (x) = {\begin{matrix} \begin{matrix} 0, & x ⩽ s \end{matrix} \\ \begin{matrix} \frac{x - s}{r}, & s ⩽ x ⩽ s + r \end{matrix} \\ \begin{matrix} 1, & x ⩾ s + r \end{matrix} \end{matrix}$ (11)

Next, a valued outranking graph is developed by using a preference index, which determines a total preorder on the set of possible actions [32]. Preference index of alternative a with respect to alternative b is computed according to the following formula. $π (a, b) = \sum_{j} w_{j} * P_{h} (a, b)$ (12) where w_j is the weight of the evaluation criterion j (computed by AHP procedure), ∑_jw_j = 1, and P_h (a, b) indicates the preference function computed for alternatives a, b. Finally, the outgoing flow and incoming flow values, and a net flow value are computed for each alternative using the following formulae, respectively: $\begin{matrix} φ^{+} (a) = \sum_{i} π (a, i), \\ φ^{-} (a) = \sum_{i} π (i, a), \\ φ (a) = φ^{+} (a) - φ^{-} (a) \end{matrix}$ (13)

The ranking of the alternatives can be then made according to the net flow values.

4 Analysis

4.1 Background information

In an aim to search for new opportunities for Turkish farmers in the agriculture sector, a total of eight crops, namely asparagus, goji berry, tamarind, chia, quinoa, buckwheat, soy bean and mung bean have been selected as prospective cultivation candidates. The reasons for focusing on these crops are that a) they have a high profit margin and an increasing demand pattern, b) they are commonly preferred by dieticians and nutritionists in designing healthy diets, c) although pilot studies of growing these crops have been performed by individual growers in the last years and considerable success rates have been achieved, these crops are not recognized by the majority of farmers in Turkey. Hence, according to the results of our analysis, it will be possible to recommend the policy-makers adopt programs that could make some of these crops to be more broadly known and grown by farmers in Turkey.

The general characteristics of each one of the selected crops are stated as follows.

Asparagus (Asparagus officinalis l.): Asparagus is a green vegetable that is a member of the lily family. It is highly preferred by the nutritionists in the design of low-calorie diets as it is packed with essential vitamins, minerals and antioxidants; and being a fiber-rich vegetable, it helps improve digestive health [11, 14]. It is produced extensively in China (7,843,162 tons/year as of 2017), which is followed by Peru (383,098 tons/year), Mexico (245,681 ton/year) and Germany (245,681 tons/year) [45]. The domestic demand for asparagus in Turkey is not clear. However, it has been increasing over the last years, and the production rates, despite still being limited in certain locations, are also growing [17]. The production amount reported to the Turkish Statistical Institute was 169 tons in 2018; and 174 tons in 2019 [48]. Both the climate and soil of Turkey, especially in Western Anatolia, are very appropriate for growing this crop [1].

Goji berry (Lycium barbarum): Belonging to the “lycium” genus, goji berry has long been used as both food and for pharmacological purposes in history. It is famous for its antioxidant potential, and is an effective supplement in diseases such as diabetes, cardiovascular diseases and cancer [29]. This plant is best grown in altitudes between 800 and 2000 m, loves sunlight, and can be grown in almost all types of soils. With irrigation, it reaches its maximum productivity in non-humid climates. Turkish climate and soil are typically appropriate to grow this plant, except perhaps for the coastal region, due to the risk of fungi infection and delay in fruit maturation [47]. The largest producer and exporter of this plant is China, with export values reaching 983.5 ton in 2012. There is unfortunately not much formal data about the production amounts in Turkey. Nevertheless, individual entrepreneurs are planting this product in different regions of the country. For instance, 6 decares of land in Afyon, 12 decares in Denizli, 18 in Kayseri, 50 decares in Adiyaman, 23 decares in Niğde, 8.5 decares in Antalya, and 10 decares in Sivas are reported to be allocated for goji berry production [47]. Considering that the yield is roughly 4–4.5 tons per hectare in the mature time of the plant, the total domestic production is less than 100 tons per year.

Tamarind (tamarindus indica): Tamarind originated from East Africa and grown extensively in India, Asia Pacific, and Egypt [50]. It is best suited to semi-dry tropical climates, but can also adapt to rainy and humid environments. Young plants are sensitive to cold weather and are seriously harmed below –2 C. Tamarind can be grown in all soils but reaches its best in drainage and in mildly acidic soil. Tamarind is an excellent source of vitamin C, and has antioxidant properties. Due to climatic conditions, tamarind would be best grown in the southern regions of Turkey. Nevertheless, no commercial cultivation of tamarind currently exists in Turkey. The domestic demand level is unclear; however, the export opportunities could be abundant due to the increased usage of tamarind in the production of cosmetics and personal care products all around the world, especially in Europe [19].

Chia (Salvia hispanica): Chia has recently become popular as a dietary food supplement due to its antioxidante and gluten-free properties, as well as healthy fatty oil profile [16]. Being native to Southern America, chia is mainly grown in Argentina, Mexico and Guatemala [37]. The expansion of chia is limited because it is a photoperiod-sensitive plant adapted to areas without cold. Chia grows naturally in tropical and subtropical environments; it is optimally established at altitude levels from 400 to 2500 m, and conditions below 200 m elevation are not adequate for its cultivation [36]. Coastal desert climate provides the best conditions for chia production in Chile, resulting in yields of around 2 tons per hectare. Although these climatic properties do not exactly exist in Turkey, higher altitude areas in the Mediterranean and Aegean coastal region could be suitable for chia production [18]. Although domestic demand increases in recent years, it is still not at a significant level. However, the European demand grows year by year, and as most of this demand is satisfied via imports from Southern America [8], there is a high potential for exports for the prospective Turkish chia growers.

Quinoa (Chenopodium quinoa): Being an alternative to crops like corn and wheat, quinoa is mainly famous for its high protein content and healthy nutrition properties. It is mainly produced in Southern America, Peru and Bolivia as the largest quinoa exporters for Europe [9]. Ecuador claims to be the largest global producer in 2017, accounting for about 1.3 thousand metric tons of the global quinoa production volume [44]. Quinoa has adapted to the harsh climatic conditions in the Andean region and can be cultivated effectively in the mountainous terrains of Anatolia as well [24]. Its productivity increases with irrigation. Its production is still limited in Turkey, but there are good opportunities. Several producer firms are assisting in the production and guaranteeing quinoa purchase from the farmers under certain conditions (e.g. cultivating this crop in fields that are at least 200 decares in size or using the seed of the producer firm [52]).

Buckwheat (Fagopyrum esculentum): Being gluten-free and having anti-oxidant properties, buckwheat has recently gained popularity as a dietary grain. In addition to health benefits, it is an appropriate grain for celiac patients. It can also be used as bee food, and in feeding small ruminants and poultry [13]. It is selective about the climate and best grows in humid and cool climates. However, it is sensitive to snow frost. If the climatic conditions are appropriate, it can grow in almost all types of soil except lime soil. Moreover, it is essentially an undemanding crop, and can be grown under various production systems [40]. It is extensively produced in China and Russia. Currently, it is grown in Middle Anatolia, in Aksaray and Bilecik, but at rather low intensity [48]. The domestic demand is growing, and there are increasing export opportunities mainly in Europe and America [28].

Soy bean (Glycine max): Soy bean is very rich in protein and is advocated by the health professionals due to its preventative features against heart and vascular diseases. Soy bean seeds are used heavily in the production of healthy oil acids and residues are the main ingredient for high-quality livestock feed. Soy bean production started in Turkey in the early 1930s and was initially grown in the Black Sea region due to its high irrigation requirement. However, today it is mainly grown in the East Mediterranean region, in Adana and Osmaniye. Unfortunately, the domestic production has been reduced drastically in the recent years, resulting in more than 90% of the domestic demand being satisfied by exports. Most of the Turkish lands are suitable for soy bean production [34]. Moreover, it can be grown in alternation with other crops enriching the nitrogen content of the soil [3]. It is mainly used as livestock feed; however, the domestic demand has been increasing over the last decade due to its health-promoting properties and as a main ingredient of the vegan/vegetarian diets.

Mung bean (vigna radiata): Mung bean was produced and exported extensively in Turkey until the 1980’s. However, its production has receded lately. Having proven its healthy dietary properties with the recent research, this crop has again started becoming popular in Turkey and all around the world [30]. Current production in Turkey is estimated to be around 2000 tons per year, half of this production is done in Gaziantep [48]. Its production is similar to that of regular beans, and the Turkish soil is generally appropriate to grow this crop. Mung bean requires hot and non-humid climate. Therefore, it is mainly suitable to be grown in the southern regions of Turkey. Domestic demand is growing and mainly satisfied via import.

In order to search for prospective opportunities each of these crops can bring to the agricultural sector of Turkey, we decided to evaluate the crops according to the following set of criteria:

Profit margin: This is the monetary benefit of growing the crop and calculated as “the selling price (in the domestic market)” – “the total production cost” per decare. We have not taken the possible export prices into consideration since they would be highly variable based on the export country and there has not been sufficient data in this aspect.

Domestic demand: This criterion refers to the current domestic market size for each candidate crop.

Export opportunities: This criterion refers to potential export opportunities of the crop. Mainly European markets are explored in this aspect since the distance of the supplier matters in terms of fresh food logistics.

Technological and know-how requirements: Within this criterion, we evaluate how much training the farmers would need and/or how technically competent they need to be to grow each crop. The lower these requirements are, the higher the crop scores in this criterion.

Time until obtaining yield: This criterion refers to the time between the initial plantation and the harvest of the mature product for each one of the crop alternatives. The longer this period is, the lower the candidate crop’s score will be according to this criterion.

Labor requirement: This criterion refers to the employment opportunities each crop can bring. Thus, the more labor intense growing a crop alternative is, the higher that alternative scores in this criterion.

Strategic importance: Some crops such as starch, corn or soybean can be used in the production of biofuels and can also bear an importance regarding an energy source. Some are critical in ensuring the self-sufficiency in feeding the entire population of a country, or the livestock if the domestic economy is heavily dependent on animal husbandry. This criterion encompasses these aspects.

Suitability of the crop to the climatic properties/land properties of Turkey: We also explore how compatible each crop alternative is with the climatic properties of Turkey and the Turkish land. The Turkish climate varies from region to region, and similarly, the soil types differ across regions. Therefore, we mainly developed the scores regarding how well each crop alternative can be grown in the main agricultural regions of Turkey.

Environmental effect: This criterion takes into account the waste amount that could occur in cultivation and harvesting of each crop, carbon emissions in the process, the effect of this process on the agricultural land, etc. Again, a higher score would mean that the crop in consideration would have lower waste rates, better effects on the soil or on the quality of other crops that can be grown in the same land via rotational planting practices.

The data for the above criteria have been gathered from various sources, such as the websites of online retailers (e.g. Migros, Carrefour, Macromarket among many others), reports of TÜİK and other European and American trade foundations (e.g. Center for Promotion of Imports (CBI), Global Agricultural Information Network (GAIN), German Federal Ministry of Food and Agriculture, U.S. Foreign Agriculture Service (FAS) ...), educational videos provided for farmers, newspaper articles, and last but not least, by semi-formal interviews of the experts. Having combined the data collected from all of these sources, a “scoring index” for the crop alternatives across all of the given criteria is formed as seen in Table 1. This table also forms the decision matrix for TOPSIS and PROMETHEE analysis.

Table 1
Scores of the Crop Alternatives across the Decision Criteria

Profit margin domestic demand export opportunities time to grow expertise requirement labor requirement Strategic importance Environmental effect land compatibility climate compatibility

Asparagus 70 40 90 60 60 60 40 40 80 90

Goji berry 90 60 90 50 50 60 30 60 80 80

Tamarind 70 50 40 60 60 80 20 50 60 50

Chia 40 70 90 90 70 50 40 70 70 70

Quinoa 60 70 90 100 60 60 40 50 70 70

Soy bean 50 90 70 100 60 70 100 80 80 70

buckwheat 40 70 60 100 80 50 60 60 90 80

mung bean 40 70 60 100 80 50 30 60 80 90

	Profit margin	domestic demand	export opportunities	time to grow	expertise requirement	labor requirement	Strategic importance	Environmental effect	land compatibility	climate compatibility
Asparagus	70	40	90	60	60	60	40	40	80	90
Goji berry	90	60	90	50	50	60	30	60	80	80
Tamarind	70	50	40	60	60	80	20	50	60	50
Chia	40	70	90	90	70	50	40	70	70	70
Quinoa	60	70	90	100	60	60	40	50	70	70
Soy bean	50	90	70	100	60	70	100	80	80	70
buckwheat	40	70	60	100	80	50	60	60	90	80
mung bean	40	70	60	100	80	50	30	60	80	90

4.2 AHP analysis

The objective of applying fuzzy AHP analysis on the different economic, social and environmental criteria is to understand how experts from the area perceive the importance of these features. To this end, we interviewed 4 experts from the faculty of agriculture in an emerging Turkish university and used fuzzy scale prepared by Chang [10] for them to compare each pair of criteria. According to this scale, linguistic variables “equally important”, “moderately important”, “important”, “very important”, “much more important” correspond to the fuzzy scales “(1,1,1)”, “(2/3,1,3/2)”, “(3/2,2,5/2)”, “(5/2,3,7/2)” and “(7/2,4,9/2)”, respectively.

By applying formula (1) on the fuzzied questionnaire responses, the fuzzy synthetic vectors were computed as in Table 2. Next, fuzzy significance vectors were calculated according to formula (3) as shown in Table 3. Finally, fuzzy significance vectors were translated into normalized W significance vectors. Based on the results of this analysis, Table 4 shows the perceived weights of the given eight features to determine the importance of each suggested crop.

Table 2
Fuzzy Synthetic Vectors of the Evaluation Criteria

S vectors Low medium high

Profit margin (PM) 0.079598 0.121257 0.181809

domestic demand (DD) 0.07809 0.119012 0.179054

export opportunity (EO) 0.061929 0.096557 0.151507

time to grow (TG) 0.057105 0.087575 0.135438

expertise requirement (ER) 0.058975 0.088698 0.134979

labor requirement (LR) 0.043331 0.065868 0.102238

strategic product (SP) 0.059035 0.087575 0.130847

environmental effect (EE) 0.069708 0.104416 0.157017

land compatibility (LC) 0.069648 0.11003 0.173545

climate compatibility (CC) 0.079296 0.119012 0.176299

S vectors	Low	medium	high
Profit margin (PM)	0.079598	0.121257	0.181809
domestic demand (DD)	0.07809	0.119012	0.179054
export opportunity (EO)	0.061929	0.096557	0.151507
time to grow (TG)	0.057105	0.087575	0.135438
expertise requirement (ER)	0.058975	0.088698	0.134979
labor requirement (LR)	0.043331	0.065868	0.102238
strategic product (SP)	0.059035	0.087575	0.130847
environmental effect (EE)	0.069708	0.104416	0.157017
land compatibility (LC)	0.069648	0.11003	0.173545
climate compatibility (CC)	0.079296	0.119012	0.176299

Table 3

Fuzzy Significance Vectors of the Evaluation Criteria

V vectors	PM	DD	EO	TG	ER	LR	SP	EE	LC	CC
profit margin (PM)	1.00	0.98	0.74	0.62	0.63	0.29	0.60	0.82	0.89	0.98
domestic demand (DD)	1.00	1.00	0.77	0.65	0.65	0.31	0.63	0.84	0.91	1.00
export opportunity (EO)	1.00	1.00	1.00	0.89	0.90	0.57	0.88	1.00	1.00	1.00
time to grow (TG)	1.00	1.00	1.00	1.00	1.00	0.68	1.00	1.00	1.00	1.00
expertise requirement (ER)	1.00	1.00	1.00	0.99	1.00	0.65	0.98	1.00	1.00	1.00
labor requirement (LR)	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
strategic product (SP)	1.00	1.00	1.00	1.00	1.00	0.67	1.00	1.00	1.00	1.00
environmental effect (EE)	1.00	1.00	0.91	0.80	0.81	0.46	0.78	1.00	1.00	1.00
land compatibility (LC)	1.00	1.00	0.86	0.75	0.75	0.42	0.73	0.94	1.00	1.00
climate compatibility (CC)	1.00	1.00	0.76	0.64	0.65	0.30	0.62	0.84	0.91	1.00

Table 4

Weights of the Evaluation Criteria

	PM	DD	EO	TG	ER	LR	SP	EE	LC	CC
Weights	0.132	0.129	0.098	0.082	0.083	0.038	0.080	0.109	0.118	0.129

Findings of the fuzzy AHP analysis indicate that profitability, domestic demand and climate and land compatibility are regarded as the most significant criteria by the experts for evaluating new crops, while labor requirement and strategic product characteristic happen to be the least favored criteria among all.

4.3 TOPSIS analysis

The TOPSIS analysis is performed starting with the “Decision Matrix”, which shows the scores of each alternative with respect to each criterion and was tabulated earlier in Table 1. This matrix is next converted to the “Normalized Decision Matrix” and “Weighted Normalized Decision Matrix” according to the procedure described in Section 3.2. Then, the positive ideal solution (PIS) and negative ideal solution (NIS) values of the evaluation criteria are computed as in Table 5. Finally, the PID, NID and CIS (closeness to ideal solution) values and the ranking of the crop alternatives are tabulated in Table 6.

Table 5
PIS and NIS values

Criteria PIS NIS

Profit Margin 0.070138 0.031173

Domestic Demand 0.061866 0.027496

Export Opportunity 0.041263 0.018339

Time to Grow 0.034313 0.017157

Expertise Requirement 0.035821 0.022388

Labor Requirement 0.017843 0.011152

Strategic Importance 0.055602 0.011120

Environmental Effect 0.051295 0.025648

Land Compatibility 0.048991 0.032660

Climate Compatibility 0.054120 0.030067

Criteria	PIS	NIS
Profit Margin	0.070138	0.031173
Domestic Demand	0.061866	0.027496
Export Opportunity	0.041263	0.018339
Time to Grow	0.034313	0.017157
Expertise Requirement	0.035821	0.022388
Labor Requirement	0.017843	0.011152
Strategic Importance	0.055602	0.011120
Environmental Effect	0.051295	0.025648
Land Compatibility	0.048991	0.032660
Climate Compatibility	0.054120	0.030067

Table 6

Final Computations and Ranking According to TOPSIS

Crop Alternatives	PID	NID	CIS	Ranking
Asparagus	0.059272	0.043928	0.425660	6
Goji Berry	0.051624	0.053639	0.509570	2
Tamarind	0.070618	0.026675	0.274171	8
Chia	0.056579	0.043443	0.434335	5
Quinoa	0.050812	0.043036	0.458574	4
Soy bean	0.036266	0.068302	0.653179	1
Buckwheat	0.051347	0.047280	0.479378	3
Mung bean	0.060417	0.043320	0.417591	7

4.4 PROMETHEE analysis

As a way to triangulate the results of TOPSIS, the PROMETHEE II methodology is applied next. The decision matrix presented in Table 1 is utilized in this approach, too, in order to create the pairwise comparisons. As argued before, the preference functions are of type 5 in Brans and Vincke [7]. The indifference area upper limit, s, is set to be 10 for all criteria while the linear preference parameter, r, is assumed to take a value of 30, leading to a linear preference upper limit of s + r = 40. These values amount to the following reasoning: if a crop scores at least 40 units higher than another crop according to a certain criterion, this is an indication that the former crop is superior to the latter for that criterion. If the difference of the scores is below 10, this is an indication that there is not enough evidence whether each of these crops is superior to the other per this criterion. Other steps of the approach are sustained according to the methodology explained in Section 3.3, and the final table showing the outgoing flow, incoming flow, the net flow and the ranking of the crop alternatives according to PROMETHEE II is presented as Table 7.

Table 7
Final Computations and Ranking According to PROMETHEE

Crop Alternatives Outgoing flow Incoming Flow Net Flow Ranking

Asparagus 0.895585 1.41078 –0.51519 7

Goji Berry 1.172728 0.819512 0.353215 2

Tamarind 0.423713 2.455779 –2.03207 8

Chia 0.826608 0.666115 0.160493 5

Quinoa 0.841263 0.540174 0.301089 4

Soy bean 1.871992 0.49337 1.378622 1

Buckwheat 1.12211 0.812682 0.309428 3

Mung bean 0.910297 0.865884 0.044412 6

Crop Alternatives	Outgoing flow	Incoming Flow	Net Flow	Ranking
Asparagus	0.895585	1.41078	–0.51519	7
Goji Berry	1.172728	0.819512	0.353215	2
Tamarind	0.423713	2.455779	–2.03207	8
Chia	0.826608	0.666115	0.160493	5
Quinoa	0.841263	0.540174	0.301089	4
Soy bean	1.871992	0.49337	1.378622	1
Buckwheat	1.12211	0.812682	0.309428	3
Mung bean	0.910297	0.865884	0.044412	6

4.5 Sensitivity analysis

Although two MCDM methodologies (TOPSIS and PROMETHEE) are jointly utilized to establish a triangulation on the results, there can still arouse some doubts on the rankings of the crop alternatives. In particular, extent analysis of fuzzy AHP method is often criticized in the literature, regarding the fact that the weights determined by this method may not represent the real values of relative importance [51]. To overcome these problems, we add a sensitivity analysis on the criteria weights. Following Jain et al. [25] and Kaya and Kahraman [27], we define a total of 6 different cases (the initial one being the original case, referred to as “base case”) that attribute different weights to particular criteria. Table 8 presents the composition of criteria weights in the considered cases.

Table 8
Criteria weights with respect to the considered cases

PM DD EO TG ER LR SP EE LC CC

Base Case 0.132 0.129 0.098 0.082 0.083 0.038 0.080 0.109 0.118 0.129

Case 1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1

Case 2 0.3 0.078 0.078 0.078 0.078 0.078 0.078 0.078 0.078 0.078

Case 3 0.078 0.078 0.078 0.078 0.078 0.078 0.078 0.3 0.078 0.078

Case 4 0.2 0.15 0.15 0.071 0.071 0.071 0.071 0.071 0.071 0.071

Case 5 0.064 0.064 0.064 0.064 0.064 0.064 0.064 0.15 0.2 0.2

	PM	DD	EO	TG	ER	LR	SP	EE	LC	CC
Base Case	0.132	0.129	0.098	0.082	0.083	0.038	0.080	0.109	0.118	0.129
Case 1	0.1	0.1	0.1	0.1	0.1	0.1	0.1	0.1	0.1	0.1
Case 2	0.3	0.078	0.078	0.078	0.078	0.078	0.078	0.078	0.078	0.078
Case 3	0.078	0.078	0.078	0.078	0.078	0.078	0.078	0.3	0.078	0.078
Case 4	0.2	0.15	0.15	0.071	0.071	0.071	0.071	0.071	0.071	0.071
Case 5	0.064	0.064	0.064	0.064	0.064	0.064	0.064	0.15	0.2	0.2

TOPSIS and PROMETHEE analysis are performed with the new weights of criteria in the new 5 cases. The rankings of the top 4 crop alternatives arising in each case are presented in Figs. 2 and 3, for TOPSIS and PROMETHEE, respectively.

Fig. 2

TOPSIS Rankings according to Sensitivity Analysis.

Fig. 3

PROMETHEE II Rankings according to Sensitivity Analysis.

5 Discussion and conclusion

This paper aims to analyze a set of unfamiliar crop alternatives for Turkish agriculture and develop recommendations regarding their cultivation considering several economic, technical, social and environmental dimensions. To this end, the performed analysis provides promising results. In particular, the results of TOPSIS and PROMETHEE analysis are mainly compatible in the sense that they both advocate growing soy bean and goji berry with a strong preference, and place tamarind as the least favorable crop among all. Moreover, both methods state buckwheat and quinoa also as preferable alternatives, ranking as 3rd and 4th crops. In fact, the two methods only differ in the ranking of mung bean and asparagus; where TOPSIS states that asparagus should be the 6th preferable crop alternative and mung bean ranks the 7th while PROMETHEE ranks the two crops oppositely. However, in general, these results indicate that soy bean, goji berry and buckwheat are the strongest alternatives to grow for Turkish farmers, while tamarind may not be considered at all.

When the sensitivity analysis results are considered, it is seen that soy bean stands out as one of the best alternatives under both TOPSIS and PROMETHEE in almost all cases. Goji berry is a favored option when high profit margins and demand considerations are considered to be more significant. If environmental concerns attract higher priority, besides soy bean, buckwheat and chia are also among the favored crops. Finally, soy bean and buckwheat are again the most favored crops if climate and land compatibility are concerned. Again, these results are in consistency with the findings of the initial round of analysis; where soy bean, goji berry and buckwheat are the most favored new crop alternatives among the selected set of crops. Considering the dietary benefits, increasing popularity, high profit margins and climate- and land-compatibility of these crops, increasing their cultivation opportunities in Turkish land appears to be highly recommendable.

With the results stated above, this work clearly has certain limitations. First, selecting an entire country as the potential area for agricultural analysis might not bring the most accurate results, as the crop alternatives could be compatible with the climatic conditions in one region while not so much in other regions. However, we chose to ignore this limitation as the aim of the research was to give Turkish farmers and policy-makers some insights regarding new crop alternatives to grow. The same study can be applied to a smaller region in the country with micro-climatic conditions and more accurate conclusions can be reached. Second, the quantitative data was not at the desired level of accuracy as these crops are not well-known in Turkey. That is, in computing the domestic demand and export opportunities of some products, we sometimes had to rely on subjective expert opinions rather than exact numbers. Finally, we had to confine our analysis to eight crops, even though there can be other crops to consider (e.g. stevia, jostaberry, guava fruit ...). Obtaining data on these crops were even more difficult than the ones discussed here. Although we are aware of some limited efforts of growing other agricultural products in some regions of Turkey, we had to eliminate them from discussion. Despite these shortcomings, we believe that the method and the analysis provided here could be illuminating for the Turkish farmers and policy-makers. Moreover, the analysis could be extended to other settings by changing the parameters appropriately, resulting in a valid analysis for evaluating potential agricultural policies in other parts of the world.

Footnotes

Acknowledgments

The author thanks Ms. Zeynep Begüm Aykut and Dr. Damla Güvercin for their help in obtaining the necessary data.

References

Akan Kuşkonmaz

Ö.

, (Asparagus officinalis L.) yetiştiriciliğinde ülkemizin ve dünyanin durumu, Tralleis 1 (2014), 24–30.

Akinci

, Özalp

A.Y.

and Turgut

, Agricultural land use suitability analysis using GIS and AHP technique, Computers and Electronics in Agriculture 97 (2013), 71–82.

Arslan

, Isler

and Caliskan

, Effects of cultivar maturity on growth and yield of double cropped soybean, Acta Agriculturae Scandinavica Section B-Soil and Plant Science 56 (2006), 39–46.

Baginsky

, Arenas

, Escobar

, Garrido

, Valero

, Tello

and Silva

, Growth and yield of chia (Salvia hispanica L.) in the Mediterranean and desert climates of Chile, Chilean Journal of Agricultural Research 76 (2016), 255–264.

Biswas

and Pal

B.B.

, Application of fuzzy goal programming technique to land use planning in agricultural system, Omega 33 (2005), 391–398.

Brans

J.P.

, Lingenierie de la decision la methode Promethee, Instruments et Perspectives Avenir. Presses del Universitete Lavel, (1982).

Brans

J.P.

and Vincke

, A Preference Ranking Organisation Method: (The PROMETHEE Method for Multiple Criteria Decision-Making), Management Science 31 (1985), 647–656.

CBI, The European market potential for chia seeds. https://www.cbi.eu/node/2088/pdf/. (2019), Accessed 17 March 2020.

CBI, Exporting quinoa to Europe. https://www.cbi.eu/node/2285/pdf/. (2019) Accessed 17 March (2020).

10.

Chang

D.Y.

, Applications of the extent analysis method on fuzzy AHP, European Journal of Operational Research 95 (1996), 649–655.

11.

Chin

C.K.

and Garrison

S.A.

, Functional elements from asparagus for human health, Acta Hortic 776 (2008), 219–226.

12.

Cobuloglu

H.I.

and Büyüktahtakin

I.E.

, A stochastic multi-criteria decision analysis for sustainable biomass crop selection, Expert Systems with Applications 42 (2015), 6065–6074.

13.

Dağdaş

, Uluslararasi Tarim Enstitüsü, Karabuğday. https://arastirma.tarimorman.gov.tr/bahridagdas/Belgeler/L%C4%B0FLETLER%20SON/Karabu%C4%9Fday.pdf. Accessed 17 March 2020.

14.

Demirkol

, Adams

and Ercal

, Biologically important thiols in various vegetables and fruits, Journal of Agricultural and Food Chemistry 52 (2004), 8151–8154.

15.

Doğan

, Arslan

and Berkman

A.N.

, Türkiye’de Tarim Sektörünün İktisadi Gelişimi ve Sorunlari: Tarihsel Bir Bakiş, Academic Review of Economics & Administrative Sciences 8 (2015).

16.

Erdoğdu

and Geçgel

Ü.

, Chia Tohumu (Salvia hispanica L.) ve Yağinin Fizikokimyasal Özellikleri ve Gida Sektöründe Değerlendirilmesi, Gida ve Yem Bilimi Teknolojisi Dergisi 21 (2019), 9–17.

17.

Eren

, Kuşkonmaz yetiştiriciliğinde (Asparagus officinalis L.) farkli saksi boylarinin ve harçlarin fide kalitesine etkisi, Master’s thesis, Adnan Menderes Üniversitesi, Fen Bilimleri Enstitüsü (2014).

18.

Erim

, Farkli ortam koşullarinda yetiştirilen salvia hispanica l.(chia) bitkisinin fitokompozisyon özelliklerinin belirlenmesi, Master’s thesis, Pamukkale Üniversitesi Fen Bilimleri Enstitüü, (2019).

19.

Fact.MR, Tamarind Extract Market Forecast, Trend Analysis, and Competition Tracking-Global Market Insights 2018 to 2028, https://www.factmr.com/report/2054/tamarind-extract-market. (2020) Accessed 17 March 2020.

20.

Garcia

J.L.

, Alvarado

, Blanco

, Jiménez

, Maldonado

A.A.

and Cortés

, Multi-attribute evaluation and selection of sites for agricultural product warehouses based on an analytic hierarchy process, Computers and Electronics in Agriculture 100 (2014), 60–69.

21.

Gupta

A.P.

, Harboe

and Tabucanon

M.T.

, Fuzzy multiple-criteria decision making for crop area planning in Narmada river basin, Agricultural Systems 63 (2000), 1–18.

22.

Habib

, Akram

and Ashraf

, Fuzzy climate decision support systems for tomatoes in high tunnels, International Journal of Fuzzy Systems 19 (2017), 751–775.

23.

Hwang

C.L.

and Yoon

, Methods for multiple attribute decision making, Springer, Berlin, Heidelberg, (1981).

24.

Jacobsen

S.E.

, The situation for quinoa and its production in southern Bolivia: from economic success to environmental disaster, Journal of Agronomy and Crop Science 197 (2011), 390–399.

25.

Jain

, Sangaiah

A.K.

, Sakhuja

, Thoduka

and Aggarwal

, Supplier selection using fuzzy AHP and TOPSIS: a case study in the Indian automotive industry, Neural Computing and Applications 29 (2018), 555–564.

26.

Johnston

B.F.

and Mellor

J.W.

, The role of agriculture in economic development, The American Economic Review 51 (1961), 566–593.

27.

Kaya

and Kahraman

, An integrated fuzzy AHP–ELECTRE methodology for environmental impact assessment, Expert Systems with Applications 38 (2011), 8553–8562.

28.

Kim

S.J.

, Sohn

H.B.

, Suh

J.T.

and Chang

D.C.

, Domestic and overseas status of buckwheat production and future trends, The Journal of the Korean Society of International Agriculture (2017).

29.

Kulczyński

and Gramza-Michałowska

, Goji berry (Lycium barbarum): composition and health effects–a review, Polish Journal of Food and Nutrition Sciences 66 (2016), 67–76.

30.

Myfikirler.org., Maş fasulyesi çiftçilerin yeni kazanç kapisi olmaya aday, https://www.myfikirler.org/mas-fasulyesi-ciftcilerin-yeni-kazanc-kapisi-olmaya-aday.html. (2014). Accessed 17 March 2020.

31.

Lenin

V.M.

, Baviera-Puig

and García-Álvarez-Coque

J.M.

, Multi-criteria methodology: AHP and fuzzy logic in the selection of post-harvest technology for smallholder cocoa production, International Food and Agribusiness Management Review 17 (2014), 107–124.

32.

Luo

S.Z.

and Xing

L.N.

, A Hybrid Decision Making Framework for Personnel Selection Using BWM, MABAC and PROMETHEE, International Journal of Fuzzy Systems 21 (2019), 2421–2434.

33.

Mishra

A.K.

, Deep

and Choudhary

, Identification of suitable sites for organic farming using AHP & GIS, The Egyptian Journal of Remote Sensing and Space Science 18 (2015), 181–193.

34.

Nazlican

A.N.

and Yetiştiriciliği

, https://arastirma.tarimorman.gov.tr/cukurovataem/Belgeler/Yeti%C5%9Ftiricilik/soya-yetistiriciligi_1.pdf. (2017). Accessed 17 March 2020.

35.

Opricovic

and Tzeng

G.H.

, Compromise solution by MCDM methods: A comparative analysis of VIKOR and TOPSIS, European Journal of Operational Research 156 (2004), 445–455.

36.

Orozco

G.N.

, Durán

, González

, Zarazúa

, Ramírez

and Mena

, Proyecciones de cambio climático y potencial productivo para Salvia hispanica L. en las zonas agrícolas de México,í, colas 10 (2014), 1831–1842.

37.

Peiretti

P.G.

and Gai

, Fatty acid and nutritive quality of chia (Salvia hispanica L.) seeds and plant during growth, Animal Feed Science and Technology 148 (2009), 267–275.

38.

Pourghasemi

H.R.

, Pradhan

and Gokceoglu

, Application of fuzzy logic and analytical hierarchy process (AHP) to landslide susceptibility mapping at Haraz watershed Iran, Natural hazards 63 (2012), 965–996.

39.

Qiu

, Using multi-criteria decision models to assess the economic and environmental impacts of farming decisions in an agricultural watershed, Review of Agricultural Economics 27 (2005), 229–244.

40.

Radics

and Mikóházi

, Principles of common buckwheat production, The European Journal of Plant Science and Biotechnology 4 (2010), 57–63.

41.

Riaz

and Hashmi

M.R.

, MAGDM for agribusiness in the environment of various cubic m-polar fuzzy averaging aggregation operators, Journal of Intelligent & Fuzzy Systems 37 (2019), 3671–3691.

42.

Saaty

T.L.

, The Analytic Hierarchy Process. New York: McGraw-Hill, (1980).

43.

Sałabun

, Wątróbski

and Shekhovtsov

, Are MCDA Methods Benchmarkable? A Comparative Study of TOPSIS, VIKOR, COPRAS, and PROMETHEE II Methods, Symmetry 12 (2020), 1549.

44.

Shahbandeh

, Global quinoa production 2010–2017, by country. https://www.statista.com/statistics/486460/global-quinoa-production-by-country/. (2020). Accessed 17 March 2020.

45.

Shahbandeh

, Major asparagus producing countries worldwide 2017. https://www.statista.com/statistics/279556/global-top-asparagus-producing-countries/. (2020). Accessed a7 March 2020.

46.

Sipahi

and Timor

, The analytic hierarchy process and analytic network process: an overview of applications, Management Decision (2010).

47.

Solak

, Erol

İ.

and Yildiz

, Goji berry –Kurt üzümü kitabi, http://www.gojiberrybahcesi.com/FileUpload/ks431142/File/kurt_uzumu.pdf. (2017). Accessed 17 March 2020.

48.

TÜİK: Statistical data for plant production. http://www.tuik.gov.tr/PreTablo.do?alt_id=1001 (2020). Accessed 17 March 2020.

49.

Vahidnia

M.H.

, Alesheikh

A.A.

and Alimohammadi

, Hospital site selection using fuzzy AHP and its derivatives, Journal of Environmental Management 90 (2009), 3048–3056.

50.

Van der Stege

, Prehsler

, Hartl

and Vogl

C.R.

, Tamarind (Tamarindus indica L.) in the traditional West African diet: not just a famine food, Fruits 66 (2011), 171–185.

51.

Wang

Y.M.

, Luo

and Hua

, On the extent analysis method for fuzzy AHP and its applications, European Journal of Operational Research 186 (2008), 735–747.

52.

Yerlitarim.com: Alim garantili uretim. http://www.yerlitarim.com/s3/sozlesmelitarim/alim-garantili-uretim (2020). Accessed 17 March 2020.

53.

Yuan

, Tian

and Wang

, A VaR-based optimization model for crop production planning under imprecise uncertainty, Journal of Intelligent & Fuzzy Systems 33 (2017), 1–14.