1 Introduction

Soil has formed the effect of many ecological factors and covers the earth’s surface. Soil meets the needs of humanity from past to present. Since humanity transitioned from nomadic life to permanent settlement, the soil has been an indispensable natural resource for many sectors, especially the agricultural sector. Changing requirements and technological advancements play a crucial role in the utilization of such invaluable lands. Soils, in addition to being a significant natural resource, genesis over many years under the influence of various factors. It is increasingly imperative for both academics and management to address the challenges posed by drought, which affects regions across various climates. According to current climate change projections, the frequency and severity of drought periods in Mediterranean and semiarid areas are expected to increase (Danandeh Mehr et al., 2020; Gaitán et al., 2020; Dursun & Babalık, 2021; Bayraklı et al., 2023; Uçar et al., 2023). It also shows that agricultural production must double to meet the projected global food demand for 2050 (Al Sayah et al., 2021). Globally needed food demand is met from arable land, 45% of the world’s food supply. Farming production is carried out on 11 million square kilometers, excluding fodder crops worldwide. Farming production areas constitute 7.38% of the world’s 149 million square kilometers of land surface (FAO, 2014). The sustainable management of soils, which is the most important source of agricultural production, is important for the reasons stated. It is necessary to determine the characteristics of agricultural soils to cover a very small area on the earth’s surface and respond to the increase in food demand. For this purpose, soil science researchers have been doing research for many years (Kellogg, 1937; Oakes, 1960; Dinç, 1975; FAO, 1976; Dinç, 1980; Dinç et al., 1988; USDA, 1993; Staff, 1999; Dengiz et al., 2003; Başayigit & Dinç, 2005; Dengiz, 2006; Özaytekin et al., 2012; Hartemik, 2015). Soil survey and mapping studies are carried out to develop and classify agricultural lands around the world with a specific methodology. As stated above, it is necessary to maintain the ecological balance sustainably in the face of the limited agricultural production areas and the demands of the rapidly increasing population. To reduce the increasing pressure on agricultural lands with the developments in technology, the necessity of determining the management models of the soils in a qualitative, quantitative, and sustainable way is an accepted idea that is necessary on a global scale with soil survey and mapping studies by researchers Şenol et al. (2014); Çoşkun and Dengiz, (2016); Hughes et al., (2017); Tuncay (2018); Dedeoğlu et al., (2020); Esetlili et al., (2020); Gerasimova et al., (2020); Şenol et al., (2020); Demir ve Başayiğit (2021); Yang et al., (2022); Regassa et al., (2023); Shafnas et al., (2023).

Soil composition and morphology are used as criteria in the classification of soils. The classification systems used today are grouped into two main groups, including pedogenic and morphological approaches, according to the criteria they are based on. The most widely known system is Soil Taxonomy, initially presented at the 1960 7th International Soil Science Society (ISSS) Congress, later refined and finalized in 1975 (Dinç, 1980; Dinç et al., 1987). It has undergone several updates since (Nactergaele et al., 2000). Soil Taxonomy serves two purposes: keeping users informed about changes and providing on-field classification keys. This system is recognized internationally, especially in Latin America and Asia, and is aligned with standards set by WRB (the World Reference Base for Soil Resources) and FAO (Nestory, 2007). Another notable initiative is the Corine program in the European Union, collecting land cover data since 1985. The LUCAS (Land Use/Cover Area Frame Survey) method ensures systematic data control, producing soil information for EU member states since 2009 (European Environment Agency, 2018). WRB was introduced in 1998, with the third edition published in 2014 (WRB, 2014). WRB aligns with modern soil classification concepts, including Soil Taxonomy, FAO Soil Map of the World 1988, and Russian Soil Classification Principles. It aims to facilitate global soil communication, supplementing national systems. The European Union utilizes WRB for its soil database, while researchers worldwide also adopt this system for classification studies (Costantini, 2023; Nedd et al., 2021; Panagos et al., 2012; Yadav et al., 2022).

Soil science in Türkiye, as a very young branch of science, started only in 1950. The first soil survey and mapping study in Türkiye was carried out by Kerim Ömer Çağlar on the basis of soil color with a scale of 1/1000000 at the subordinate level. Soil survey studies continued by updating these studies. “Türkiye Improved Soil Map” was made for the whole country at the subordinate level by using 1/25000 scale topographic maps between 1966 and 1971 (Dinç et al., 1988). Soil survey and mapping studies are carried out in Türkiye according to the new Soil Taxonomy classification system (Dinç, 1975; Dinç et al., 1988; Başayiğit & Dinç, 2005; Dengiz, 2006; Özaytekin et al., 2012; Çullu, 2012; Dingil et al., 2012; Şenol et al., 2014; Akça & Atatanır, 2020; Demir & Başayiğit, 2021; Tunçay et al., 2022). However, current studies are limited in number and do not cover large areas compared to Türkiye in general. Taking into account the soil survey and land evaluation studies in Türkiye and the developments in the world, Law No. 5403 on soil protection and land use was published on July 3, 2005. According to Article 7 of this law, the Ministry makes or has made the classifications and maps of the land and soil by using different systems for soil protection and use. It has been stated that the procedures and principles regarding the determination and classification of soil and land assets, setting standards for a survey, analysis, and classification, preparation of maps and databases, and making them available to users will be determined by a regulation to be prepared by the Ministry. The “Soil Survey Mapping Handbook” was published in 2015 by the General Directorate of Agricultural Reform for Modern Soil Survey and Land Evaluation, taking into account the law numbered 5403. Şenol et al. (2015) stated that LCT and land evaluation criteria can be combined. They stated that using this multi-criteria analysis approach, land evaluation can be made based on specific land use patterns and combinations of different criteria. Using this method, several small-scale studies on the suitability evaluation of agricultural areas have been conducted (Jankowski, 1995; Goovaerts, 1999; Yüksel & Dengiz, 2001; Cangir & Boyraz, 2002; Malczewski, 2004; Dengiz et al., 2005; Rossel et al., 2006; Dengiz et al., 2006; Dengiz & Sarıoğlu, 2013; Dengiz, 2013; Dengiz & Özyazıcı, 2018; Çakır & Dengiz, 2021; Yeo et al., 2022; Tercan ve Dengiz, 2022).

The urgent need for conducting comprehensive soil surveys and mapping across the entire country arises from the critical necessity to protect the proper utilization of suitable agricultural lands, thus ensuring the sustainable continuity of global initiatives aimed at addressing climate change, an increasing population, and food security. For this reason, at a scale of 1/25000 national soil database was used to determine suitable agricultural lands within the borders of Isparta province with a multi-criteria approach. The study aimed to determine according to the land suitability classification determined by FAO/UNESCO (1976) with the multi-criteria process. In the study, using the national soil database, the evaluation was carried out using 9 criteria (great soil groups, soil depth, land use capability classes, land use capability subclasses, other soil properties, soil depth, erosion degree, slope, aspect, and elevation) with a scale of 1/25000 belonging to the province of Isparta in 1994 (KHGM, 1994). Using the multi-criteria approach in Isparta province, suitable areas for agriculture were determined. According to the land use classification, it was compared with the use of arable land. This study determined the need for updating and using soil surveys and mapping studies in planning.

2 Material and method

2.1 Study field description

The study area, located within the borders of the Mediterranean Region, covers 13 district lands, including Atabey, Keçiborlu, Gönen, Eğirdir, Gelendost, Şakikaraağaç, Uluborlu, Yenişarbademli, Yalvaç, Senarkent, Sütçüler, and Aksu. It extends between coordinates 2,466,922–374,731 east and 4,242,849–4,250,019 north (Zone 36, UTM-m) (Fig. 1). Isparta province, encompassing the study area, has a surface area of 8946 km2 according to the General Directorate of Maps. Positioned on the western foot of the Taurus Mountains, the province experiences a transition between the Mediterranean and Central Anatolian climates. Mediterranean climate effects are prominent in the hollow areas south of the Taurus Mountains, transitioning to a cooler and rainy climate with increased altitude toward the mountains. The study area, situated at the center of the Lakes region, has experienced an average precipitation of 568.4 mm and an average temperature of 12.3 °C for many years (MGM, 2022). The land of the province of Isparta generally emerged in the III. Time formations and consists of white chalk stones. Basalt or trachyte deposits have emerged in the province area where volcanic activities are observed. Sulfur and gypsum dusts, which came to the surface with volcanic sprays, caused the formation of hardened large beds. The oldest building in the province belongs to the Paleozoic. I. Time formations of crystalline schist, phyllite, limestone, and marble cover the borders of the Yalvaç district. II. Time formations cover the borders of the Sütçüler district, east of Isparta and east and west of Eğirdir Lake as a result of limestone, marl, flysch, and serpentine tectonic activities. III. Zaman formations, Miocene limestone, marl, sandstone, and conglomerates intrude into the province area along the Aksu and Köprü Suyu valleys. These sediments are sometimes hardened as sandstones and sometimes as loose sand and gravel deposits. IV. Time formations, conglomerates, silt, clay, and pebbles have already covered the bottom of the depression areas in Isparta (KHGM, 1994). Soils in the study area are located in Cambisols, Lithosols, Fluvisols, and Xerosols soil units according to FAO (1981). According to the Soil map of the world, the mapping units of the study area soils are Be 122-2bc (289.9 km2), Bk45-2bc (3776.7 km2), Bk 49-2c (612.2 km2), I-Lc-E-2bc (3265.7 km2), jc49-1/3a (27.9 km2), and Xk56-2/3ab (8.6 km2) (FAO, 1981). Agricultural production is carried out in an area of 1939.98 km2 within the study area. 21.68% of the study area is used for agricultural production. Agricultural production in the study area is used as 20.38% fruit, beverage, and spice crops, 20% fallow land, 2.18% vegetable area, 0.09% ornamental plants, and 62.07% cereals and other ornamental plants. According to 1990 and 2022 census data, the overall population of the study area rose from 314,766 in 1990 to 445,325 in 2022 (TUİK, 2023).

Fig. 1
figure 1

Location map of the study area

Full size image

2.2 Parameters of the national soil database used in the land suitability analysis

Türkiye soils have been defined by “Soil Maps” classified according to the Old American Soil classification system since 1958 (Başayiğit & Dinç, 2001; Tümsavaş & Aksoy, 2008; Akıncı et al., 2013; Tanrıkulu, 2015; Şenol et al., 2015; Alaboz et al., 2019). The regions classified according to the new classification systems encompass a minor portion of Türkiye’s total land. Soil classification systems are multi-category, and the number of classes increases widely from upper to lower categories, forming a pyramid. Türkiye soils were formed from six categories according to the old classification system. These classes consist of categories such as ordo, suborder, large soil groups, families, series, and types. The top category is ordos, and the discriminating factors are soil formation factors. Zonal soils have distinctive characteristics of climate and vegetation. Its distinguishing characteristics are topography and the main material, intrazonal soils. Its distinguishing characteristics are azonal soils, in which lack of time, bedrock, rapid erosion, or re-deposition plays a role (Şenol et al., 2015).

The national soil database data was created using 1/25000 scaled data as a mapping unit in 1966–1971, taking into account the depth, slope, stony, erosion degree, and similar features of the 1938 American soil classification system, which are important phases determined by large soil groups and field observations. These maps were updated in 1982–1984 by adding differences in land use capability classes and land use situations. The report of Isparta province land assets and the suitability of the land for agricultural use was used (KHGM, 1994). In the study, large soil group (GSG), land use capability class (LUCC), land use capability subclass (LUCS), soil depth, slope, aspect, elevation and erosion degree, and other soil properties were used to determine the areas suitable for agriculture. In the study, these data were used for five reasons in determining the areas suitable for agriculture in Isparta province. The first reason is that it is sufficient to determine suitable areas for cultivation. The second reason is that there is no agricultural product restriction in areas suitable for agricultural production; therefore, other soil nutrients, climate, etc. values are not included. The third reason is that the parameters of these spatial data are created and ready for the use of the relevant institutions. The fourth reason is to investigate the extraction of non-agricultural areas determined in the national soil database data and to reveal the change in the use of suitable land according to Corine (2018) land use/land cover classification. The fifth reason is to protect the ecological richness of the region, which practices widespread agricultural and forestry activities, contributing significantly to the global food supply. The national soil database parameters used within the scope of the study are explained in Table 2.

2.3 Land suitability analysis datasets and purpose methodology

In this study, the methodology applied to determine suitable agricultural areas according to LCT and to determine the land use status according to Corine (2018) data is given in Fig. 2. In this study, the topographic parameters were determined from SRTM (Shuttle Radar Topography Mission) data in digital format. SRTM data, one of the most accurate digital height models in the world, has a spatial resolution of approximately 28*28 (≤ 30 m) meters. Elevation, slope, and aspect maps were made in ArcGIS software from these data. Digital soil maps at the scale of 1/25000 were used at the subordinate level regarding the characteristics of the soils in the study area. According to the raster data format of the DEM data of the study area, geographic and raster transformations of the 28*28 m cell size soil data were made. GSG, LUCC, LUCS, OSP, soil depth, and erosion degrees maps were produced in the soil database. After the digital mapping units of the study area were created, an agricultural land suitability map was created using LCT. The determination of the current use cases of suitable areas for agriculture was made according to CLC 2018 data.

Fig. 2
figure 2

Flowchart of purpose method for agricultural land use suitability map

Full size image

It is the calculation of the ratios of different land and soil parameters (criteria) for each defined mapping unit in the process of evaluating the lands in terms of suitability for agricultural purposes. It would be appropriate to approach land suitability analysis studies as a multi-criteria evaluation or multi-criteria decision-making problem. In this study, LCT was used as a multi-criteria land suitability assessment technique, although there are many techniques for a multi-criteria land suitability assessment technique (Patrona, 1998; Şenol et al., 2015).

In LCT, a weight value is assigned to each criterion that affects the agricultural land use pattern. The relative importance of the criteria determines this weight value. Afterward, these subcriteria are divided into subcriteria and subcriteria scores are determined by subjecting these subcriteria to a separate numerical evaluation. Then, these subcriteria scores are multiplied by the weight value of the criterion to which they belong. Thus, the criteria can be put together on the same scale, that is, they are combined. The equation for the conformity assessment approach of agricultural lands in this technique is given below.

$$S = \mathop \sum \limits_{i = 1}^{n} \left( {W_{i} \cdot X_{i} } \right)$$

where S is the total land suitability score, \({W}_{i}\) is the weight value of i parameter, \({X}_{i}\) is the subcriteria of i parameter, and \(n\) is the number of parameters. The values calculated by LCT and the FAO land evaluation classification (FAO, 1976) for each mapping unit are given in Table 1.

Table 1 Land suitability classes and values of classes
Full size table

A total of nine parameters that can be decisive in land suitability solutions were considered. These parameters are grouped as GSG, LUCC, LUCS, OSP, erosion degree, soil depth, elevation, slope, and aspect criteria. In addition, these criteria were divided into subfactors and weight scores between 0 and 10 were given. It takes either the value of 0 if the subfactors make it impossible to use the lands from an agricultural point of view or 10 if it provides optimum conditions for cultivating suitable soil tillage and cultivated plants. Values between 0 and 10 vary according to the degree of limitation of plant growth by the soil characteristic (Table 2). In the process of determining the weight scores of each of the parameters (criteria), the weight scores were determined using the analytical hierarchy process (AHP) technique developed by Saaty (1980), taking into account the importance of the evaluated criteria relative to each other (Table 3). This technique is a measurement theory based on the priority values obtained from the pairwise comparison of the considered parameters. It allows both quantitative and qualitative factors to be taken into account in the selection of the best decision alternative. The relative prioritization scale based on pairwise comparisons was collected to determine the lands suitable for tillage agriculture within the study area.

Table 2 Description class value parameters used in the model for agricultural land suitability classification and weight scores of subfactors
Full size table
Table 3 Pairwise comparisons scale used for preferences in the AHP technique
Full size table

While the weight scores of the criteria (parameters) evaluated in the study were determined by the AHP technique, considering the effect status of the criteria, the pairwise comparison matrix is created as given in the equation below.

$$A = \left[ {\begin{array}{*{20}l} {a_{11} } \hfill & {a_{12} } \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & {a_{1n} } \hfill \\ {a_{21} } \hfill & {a_{22} } \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & {a_{2n} } \hfill \\ \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill \\ \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill \\ \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill \\ \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill \\ {a_{n1} } \hfill & {a_{n2} } \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & \vdots \hfill & {a_{nn} } \hfill \\ \end{array} } \right]$$

where A is the pairwise comparison matrix, \({a}_{ij}\) is the importance value relative to criteria \(i\) under criterion \(j\) (\(i,j=\mathrm{1,2},\dots ,n)\).

The properties of the pairwise comparison matrix are explained below, respectively.

$$a_{ji} = {\raise0.7ex\hbox{$1$} \!\mathord{\left/ {\vphantom {1 {a_{ij} }}}\right.\kern-0pt} \!\lower0.7ex\hbox{${a_{ij} }$}}$$

\(a_{ij} > 0 \left( {i,j = 1,2, \ldots ,n} \right)\) It must satisfy the following property for the pairwise comparisons matrix to be consistent.

$$a_{ij} = a_{ji} a_{jk} \left( {i,j,k = 1,2, \ldots ,n} \right)$$

If the pairwise comparisons matrix is strictly consistent, that is, the above equation is satisfied, then the entries of the pairwise comparisons matrix A are error-free. \({a}_{ij}\) can be expressed by the following equation.

$$a_{ij} = \frac{{W_{i} }}{{W_{j} }}$$

where \({W}_{i}\) is the priority value for criteria \(i\) calculated by means of pairwise comparisons matrix A, \({W}_{j}\) is the priority value for criteria \(j\) calculated by means of pairwise comparisons. Using the above equation, the following equation can be written.

$$a_{ik} a_{kj} = \frac{{W_{i} }}{{W_{k} }}\frac{{W_{k} }}{{W_{j} }} = a_{ij} \left( {i,j,k = 1,2, \ldots ,n} \right)$$

The diagonal entries of the pairwise comparison matrix take the value 1. That is, as stated in the equation below.

$$a_{ii} = 1 \left( {i,j,k = 1,2, \ldots ,n} \right)$$

After the creation of the A matrix, the calculation of the priority of each of the compared parameters (largest eigenvalue vector or priority vector, or weight values) is done as given below.

The values in each column of the pairwise comparisons matrix are summed. Each element in the pairwise comparisons matrix is divided by the total value of the column it is in. The matrix obtained as a result of this process is called the normalized pairwise comparison matrix. The arithmetic mean values of the elements in the row of the normalized pairwise comparison matrix provide an estimate of the relative priorities of the elements being compared. In the last step of the method, the consistency check of the obtained eigenvector is done. A new vector is obtained by multiplying the pairwise comparisons matrix (A) with the resulting priority vector (W). A second new vector is obtained by dividing each element of this new vector by its corresponding value in the priority vector. By taking the arithmetic average of the values of this last vector, the maximum eigenvalue \({\lambda }_{{\text{max}}}\) is estimated. The closer \({\lambda }_{{\text{max}}}\) is to the number of elements (n) of the pairwise comparisons matrix, the more consistent the result (Dengiz, 2013; Şenol et al., 2015; Dedeoğlu & Dengiz, 2019; Şenol et al., 2020; Karaca et al., 2021). If the pairwise comparison matrix A is not perfectly consistent, the \({\lambda }_{{\text{max}}}\) value deviates from n and the other eigenvalues deviate from zero. These deviations are determined by the consistency ındex (CI).

$${\text{CI}} = \frac{{\lambda_{\max } - n}}{n - 1}$$

To calculate the consistency ratio (CR), the “random ındex (RI)” values must also be known. This value, suggested by Saaty (1980), is used by randomly filling 100 matrices in each dimension of 1–15-dimensional matrices and averaging the CI values calculated according to the above equation (Table 4).

$${\text{CR}} = {\raise0.7ex\hbox{${{\text{CI}}}$} \!\mathord{\left/ {\vphantom {{{\text{CI}}} {{\text{RI}}}}}\right.\kern-0pt} \!\lower0.7ex\hbox{${{\text{RI}}}$}}$$
Table 4 Random index values used for CR according to AHP for 15 criteria (Saaty, 1980)
Full size table

Consistency check measures the logical inconsistency of judgments and allows the identification of errors that may occur in judgments. The CI must be 0.10 (10%) or less for the method to be valid. If this ratio is greater than 0.10, pairwise comparison matrices need to be reconstructed (Saaty, 1980).

3 Results and discussion

3.1 Main criteria and subcriteria

As shown in Fig. 2, the spatial distributions of the main criteria and subgroup criteria of parameters inside the study area’s borders were derived from soil maps and DEM data. The results obtained for the study area are presented in Table 5. Detailed explanations for the results obtained for each parameter are provided below.

Table 5 Areal and percentile distributions of the main and subcriteria parameters in Isparta province
Full size table

3.1.1 Great soil groups

The size and spatial distribution of Isparta province according to large soil groups due to the effect of different soil formation factors are given in Table 5. According to the national soil database of Isparta province, it is in the zonal soil order of 46.89%, the intrazonal soil order of 26.89%, and the azonal soil order of 12.47%. Water bodies, urban fabric, bare rocks, and rubbles cover 25.44% of Isparta province (Table 5). The great soil groups map of Isparta province is given in Fig. 3.

Fig. 3
figure 3

GSG, LUCC, and LUCS maps used in the study

Full size image

3.1.2 Land use capability class (LUCC)

Land use capability class (LUCC), according to the system proposed by Klingebiel and Montgomery (1961), was established by the Soil Conservation Service of the US Department of Agriculture. LUCC has been widely used throughout the world with numerous adaptations. LUCC, in which the categorical system uses qualitative criteria, is stated in Law No. 5403 “Soil Protection and Land Use” enacted in 2005 in Türkiye (Şenol et al., 2015). Land use capability class (LUCC) refers to the classification made according to the degree of suitability of the lands for sowing and planting, pasture and forest without plowing under climate, soil, topography, and drainage conditions. The land use capability classification is defined as eight classes. There are two groups in this classification. Lands suitable for tillage agriculture and lands not suitable for tillage. Soils in the first four classes (I, II, III, and IV) are suitable for the agricultural culture and can grow locally adapted plants. Soils in V, VI, and VII classes are not suitable for cultivated agriculture culture. VIII. Class soils are not suitable for the cultivation of any culture plant. The areal and percentage distributions of the lands according to the land use capability classes of the province of Isparta are given in Table 5. The lands suitable for tillage agriculture cover an area of 22.38% with 201,450.30 ha. The lands unsuitable for tillage agriculture cover an area of 469,650.12 ha and 52.18%. VIII class agricultural production is not possible lands cover an area of 152,961.15 ha and 16.99%. Settlements and water bodies cover an area of 76,013.63 ha and 8.45% in the national soil database within the provincial borders of Isparta (Table 5). The land use capability class map of Isparta province is given in Fig. 3.

3.1.3 Land use capability subclass

The land use capability subclass is symbolized in lowercase letters according to the type of limiting factor to indicate land use problems and disability in LUCCs subcategories (Kellog, 1951; Kellog, 1963; Sitorus, 2010). The subclasses have the same kinds of limiting factors for agricultural uses as a result of soil and climate. They are groups of units within land use capability classes. Some soils may have agricultural production-limiting features such as erosion, groundwater height, depth, or other degraded soil characteristics when conservation practices are not implemented. The use of some other soils may be hindered by bad climatic conditions. Soil subclasses have been defined as four types of limiting factors. It is expressed as the combination of four factors to show their combined effects or the excess of the effect degree. The lower limits determined are erosion hazard (e), soil deficiency–root zone limitations (s), wetness, drainage problem and flood damage (w), and climate limitations (c). If these limiting factors are displayed together, the order of priority is e, w, s (Senol et al. 2014). According to Isparta province national soil database data, land use capability subclasses are expressed as e, es, s, se, w, and ws (Table 5). There are no climate limitations (c), which is a subclass of land use capability in Isparta. Other limiting factors cover an area of 621,818.99 ha and 69.09%. In the study area, it covers 49,281.43 ha and 5.48% of which does not contain any soil limiting factors. In the study area, no data (water bodies, urban fabric, and bare rocks and rubbles) covers the 49,281.43 ha and 25.44% classes of water bodies, urban fabric, and bare rocks and rubbles are unsuitable agricultural areas. The land use capability class map of Isparta province is given in Fig. 3.

3.1.4 Other soil properties

Other soil properties that limit agricultural production in the study area are given in Table 5. The areas that do not have other soil properties limiting agricultural production of the study area cover 505,967.86 hectares and 56.21%. Bare rocks and debris limiting soil formation factors and without agricultural production cover an area of 152,977.90 hectares, 17%. Among the other soil properties, stony (t) and rocky (r) cover an area of 138,501.46 hectares and 15.39%. The water body and settlement areas cover an area of 75,996.88 hectares and 8.44%. Factors limiting agricultural production in the study area including poor drainage (f), slightly saline (h), saline–alkaline (v), inadequate drainage (y), slightly salty-alkaline (k) and their combinations (e.g. hf, hy, kf, and ts) cover an area of 2.96% with 26,631.10 hectares (Table 5). The map of other soil properties limiting agricultural production in the study area is given in Fig. 4.

3.1.5 Erosion degree

Soil loss affects soil’s physical, chemical, and biological properties and prevents soil formation. Soil fertility decreases as erosion increases soil loss. The limitation of soil formation negatively affects the uptake of substances needed for physiological events in plant and root development. Inhibition of plant root growth restricts the uptake of water needed by plants. Preventing soil formation creates soils that are not suitable for agricultural production, as it causes a decrease in the amount of water, plant nutrient intake, and organic matter content (Alaboz et al., 2021; Başayiğit et al., 2016; Dengiz et al., 2013). Soil formation processes take place more slowly with the effect of erosion, resulting in bare rocks and debris without soil cover. No data class, which has no erosion data in the study area, covers 228,974.77 hectares and 25.44%. The bare rock and rubbles in the No data class cover 152,977.90 hectares, 17.00%, the waterbodies class cover 6915.2 hectares, 7.68%, and the urban fabric class covers 6838.68 hectares, 0.76%. 8.68% of the soils of the study area are under the influence of no or very little, 11.27% moderate, 23.01% severe, and 31.60% very severe water erosion. Water erosion is observed in 65.88% of the soils in the study area (Table 5). Due to the topographic factors of the study area, the source of erosion is precipitation at higher elevations. The source of erosion is streams and rivers until the slope of the bed decreases in the drainage lines where the precipitations converge in the lower parts. According to the national soil database, there is no wind erosion in the borders of Isparta province. The erosion degree map in the study area is given in Fig. 4.

Fig. 4
figure 4

Soil depth, erosion degrees, and OSP maps used in the study

Full size image

3.1.6 Soil depth

Soil depth is defined as the difference between the soil profile and the bedrock formed by the effect of soil formation factors. It is possible to define the soil depth in the profile with various phases, from hard layers that prevent root development to continuous groundwater. Depth phases are shown by dividing them into classes, ranging from very shallow to very deep, between 0 and 150 cm, according to the bedrock, parent material, hard layers in the profile, and other plant root growth limiting factors. Soil depth is important in determining the crop pattern in the cultivation of crop plants for agricultural production (Şenol et al., 2015). Soil depth refers to the depth at which cultivated plants can use their roots and available water and nutrients. Soil depth in the national soil database in the study area is lithosolic, very shallow (0–20 cm), shallow (50–20 cm), medium deep (90–50 cm), medium deep (90–50 cm), and deep (90+ cm) in the form of a combination of a great soil groups and slope, according to the old American soil taxonomy. According to soil depth classes, the study area covers 281,894.23 hectares with 31.2% lithosolic, 118,935.37 hectares with 13.21 percent very shallow, 99,827.11 hectares with 11.09% shallow, 49,267.02 medium deep, and 121,176.69 hectares with 13.46% deep. In the study area, soils without soil depth data cover 228,974.77 hectares and 25.44% (Table 5). The soil depth in Isparta province was determined as 18.94% arable land in a medium deep and deep class suitable for agriculture. The spatial distribution of the class limiting agricultural production in Isparta province was determined as 55.62 percent. The spatial distribution of water bodies, settlements, and bare rock and debris class without soil depth data in Isparta province was determined as 25.44% (Table 5). The soil depth map in the study area is given in Fig. 4.

3.1.7 Elevation

Topography and vegetation affect climate by helping determine how solar energy is used on Earth. Plant abundance and land cover type (soil, sand, concrete, or asphalt) affect evaporation and ambient temperature. Climatic features are wind, humidity, cloud cover, atmospheric pressure, and fogging. The elevation is a significant factor in determining climate. Land cover can assist in determining climate regional. A region’s elevation, proximity to the ocean or fresh water, and land use patterns can all affect climate. All climates are the product of many factors such as latitude, altitude, and topography (Türkeş et al., 2022). In general, climatic conditions become colder as altitude increases. However, it is stated that high regions warm faster than low-altitude areas and there is usually a greater increase in daily minimum temperatures than daily maximum temperatures. Elevation affects crop patterns and harvest dates in agricultural production. The air temperature decreases by 0.5 °C due to elevation increasing every 100 m in the first 4 km of the troposphere. This situation causes a change in the vegetation cover and vegetation period every 100 m. This situation is effective in product selection for agricultural production, sowing dates, harvest dates, first spring late frosts, and early autumn frosts (Akıncı et al., 2013; Başayiğit et al., 2008). It has been determined that it is below 700 m around the Karacaören dam located in the Sütçüler district in the south of the study area. In the study area, it has been determined that the height increasing during the Taurus Mountains range is over 700 m toward the inner parts. It was determined that the altitude of the study area covers 16,444.62 hectares with 1.83% between 248 and 700 m, 858,182.76 hectares with 95.35% between 700 and 2100 m, and 25,447.81 hectares with 2.83% above 2100+ meters (Table 5). The elevation map in the study area is given in Fig. 5.

Fig. 5
figure 5

Topographic maps of elevation, slope, and aspect generated from the digital elevation model used in the study

Full size image

3.1.8 Aspect

Aspect is considered an evaluation criterion in the selection of the land to be cultivated. Plants need the sun at certain intervals to maintain their physiological activities. Plants are classified according to the amount of sunshine they need daily. In general, most cultivars in the northern hemisphere continue to develop optimally in the south and west directions, which receive sunlight for most of the day (Özyazıcı et al., 2014; Karaca et al., 2021). Aspect classes determined according to the degree of aspect in the study area were flat, south, southwest, southeast, east, and west, and it was determined that 634,237.34 hectares and 70.46% of them had no sunbathing problem. Aspect classes determined according to the exposure degrees of the study area were determined as north, northwest, and northeast; 265,837.86 hectares and 29.54% of them had sunbathing problems (Table 5). The aspect map in the study area is given in Fig. 5.

3.1.9 Slope

The slope is closely related to topography, that is, geomorphological features, which is one of the soil formation factors. The slope is expressed as a percentage. As the slope increases, soil formation slows down, and as the slope decreases, soil formation increases. As the erosion effect decreases in the areas where the slope decreases, deep soils are formed. Erosion, which increases with precipitation in places where the slope increases, slows down soil formation. As soil formation is slow in the areas where the slope increases, the soil depth decreases. Since the soil depth decreases in agricultural areas where the slope is high, cultivating crops and tillage farming cannot be done. In addition, increasing slope accelerates the erosion of the surface soil, thus reducing soil fertility properties such as organic matter, plant nutrients, and water-holding capacity of the soil (Başayiğit et al., 2016). It was determined that 78.46% of Isparta province has a slope between 0 and 20%, and 15.46% has a slope over 20% (Table 5). The slope map in the study area is given in Fig. 5.

3.2 Corine (2018) land use-land cover (LULC)

CORINE (coordination of ınformation on the environment) consists of combining land cover/use maps and land data produced using computer-aided visual interpretation methods over satellite images according to the Land Cover/Use Classification determined by the European Environment Agency (EEA). In all EEA member countries, the same basic data has been collected and a standard database established for the purposes of identifying environmental changes in the land, rational management of natural resources, and establishing environmental policies. In this context, CLC 2018 codes were created at three levels according to land use situations, and these groups were described as Label 1, Label 2, and Label 3 (Corine, 2018; EEA, 2018; Dinç & Gül, 2021; Siyavuş, 2021). According to Corine’s nomenclature, the first group is defined as five types of land cover type Label 1. These are Agricultural areas, forest and semi-natural areas, artificial areas, wetlands, and water bodies. According to the Corine (2018) land use classification of the study area, it was determined that agricultural areas were 27.49%, forest and semi-natural areas were 63.17%, artificial areas were 1.39%, wetlands were 0.60%, and water bodies were 7.34%. A database was created by dividing the study area into subgroups according to these five land use classes. The maps created according to the Corine (2018) label codes of the study area are given in Fig. 6.

The area size and area percentage according to the CLC2018 data of the land cover and usage status of the study area are given in Table 6.

Fig. 6
figure 6

CLC2018 Maps produced from the Corine Land Cover (CLC2018) datasets referring to the Copernicus Land Monitoring Service used in the study

Full size image
Table 6 Distribution of land use classification of the study area according to Corine (2018)
Full size table

3.3 Suitable agricultural areas

Using the data of the national soil database and digital elevation model data determined nine mapping units within the administrative borders of Isparta province. GSG, LUCC, LUCS, soil depth, erosion degree, and OSP are the criteria chosen by considering the soil’s physical, chemical, and biological properties in determining the agricultural suitability of the lands. Elevation, slope, and aspect are the topography criteria in determining the agricultural suitability of the lands. The priority weight values obtained based on the pairwise comparisons matrix made are given in Table 7.

Table 7 Calculations of AHP technique to determine weighted values for parameters
Full size table

In Table 7, the GSG criterion, one of the agricultural suitability criteria, with a weight value of 0.284 is calculated. Major soil groups are classified according to the influence of soil formation processes as well as soil properties such as color, texture, lime, iron, drainage, organic matter, structure, and salt deposition. Large soil groups are the most important mapping unit in determining the suitability of soils for agricultural production (Bozdağ et al., 2016). A weight value of 0.226 was calculated for the LUCC agricultural suitability criterion. LUCC is a grouping of one or more individual soil mapping units. It demonstrates soil capability units with similar potentials and permanent hazards. The soils in a capability unit are sufficiently uniform. The soils in a capability unit produce similar kinds of cultivated crops and pasture plants with similar management practices. The soils in a capability unit require similar conservation treatment and management under the same kind and condition of vegetative cover. The soils in capability units have comparable potential productivity (Kellog, 1951, 1963; KHGM, 1994; Sitorus, 2010). A weight value of 0.165 was calculated for the LUCS agricultural suitability criterion. The capability subclass provides information as to the kind of conservation problem or limitations involved. The LUCS criterion contains information about the type and limitations of homogeneous preservation of soils for agricultural production. The capability subclass units indicate the place of soils in the main protection group. Soils’ main protection groups are factors that limit agricultural productivity. These limits are described as erosion and runoff, excess water, root zone limitations, and climatic limitations. The capability classification (unit, subclass, and class) is an interpretive classification by experts in soil science. These interpretations include based on the effects of combinations of climate and permanent soil characteristics on risks of soil damage, productive capacity, limitations in use, and soil management requirements. A weight value of 0.115 was calculated for the soil depth agricultural suitability criterion. Effective depth refers to the total depth of the soil profile suitable for root development. Since the soil depth criterion affects the agricultural productivity of the soils, it is effective in providing optimum conditions for the cultivation of cultivars. A weight value of 0.089 was calculated for the slope agricultural suitability criterion. The slope plays an important role in performing activities such as in-field mechanization or field traffic, as well as soil erosion (Dengiz et al., 2006). After the GSG, LUCC, and LUCS criteria, the slope comes in fourth place within the hierarchical relationship. It is closely related to many soil quality properties such as keeping enough water, water, and plant nutrients in the soil, ensuring the structural development of the soil, plant root development, air–water, and heat circulation, and microorganism activities (Dengiz & Sarıoğlu, 2013). High weight values were calculated because the agricultural suitability criteria of the study area were mapping units created by evaluating soil quality properties. Other agricultural suitability criteria are followed by aspect (0.089), elevation (0.031), erosion degree (0.027), and OSP (0.021). Consistency ındex and consistency ratio values for pairwise comparisons of these criteria were calculated as 0.648 and 0.045, respectively. Since the consistency check value of the binary comparison matrix results is calculated as 4% using the AHP technique, the determined weight values are consistent (Table 7). The weight values were calculated low because the agricultural suitability criteria, which were determined in the last row, can be eliminated with protection and human factors and the product pattern can be determined by taking into account the agricultural constraints.

Determinating using LCT, the agricultural suitability map of the study area is shown in Fig. 7. Conformity classes and proportional distributions of mapping units are shown in Table 8.

Fig. 7
figure 7

Land suitability map according to LCT

Full size image
Table 8 Distribution of land suitability classes based on LCT and CLC 2018’s Label 1
Full size table

In addition, the spatial and proportional distributions of the agricultural suitability classes determined for the study area are calculated according to the land use/land cover classes using Corine CLC 2018 data (Table 8). Distribution of current land use status according to Label 1 class according to five land cover types is calculated. Proportional distributions of other Label 2 and Label 3 classes are given in Appendix Tables 9 and 10.

The standard soil survey map shows the spatial conditions of soil units. These maps are intended to meet the needs of users with widely different problems. In addition, these maps contain considerable detail to show important basic soil differences. According to LCT model, 106,899.27 hectares (11.88%) of the soils of the study area are the S1 class according to the agricultural land suitability classification. According to CLC 2018 data of this class, 90.26% are agricultural areas, 5.06% are forest and semi-natural areas, and 3.73% are artificial surfaces. 90,037.86 hectares (10%) of the lands of the study area is the S2 class according to the agricultural land suitability classification. According to CLC 2018 data of this class, 79.13% of them are agricultural areas, 16.19% of them are forest and semi-natural areas, and 1.50% of them are artificial surfaces. 158,730.36 hectares (17.64%) of the lands of the study area is the S3 class according to the agricultural land suitability classification.

According to CLC 2018 data of this class, 28.20% are agricultural areas, 69.99% are forest and semi-natural areas, and 0.82% are artificial surfaces. 316,953.08 hectares (35.21%) of the lands of the study area constituted the N1 class according to the agricultural land suitability classification. According to the CLC 2018 data of this class, 8.17% are agricultural areas, 90.64% are forest and semi-natural areas, and 0.51% are artificial surfaces. 227,454.62 hectares (25.22%) of the lands of the study area is the N2 class according to the agricultural land suitability classification. According to the CLC 2018 data of this class, 4.18% are agricultural areas, 64.85% are forest and semi-natural areas, and 1.90% are artificial surfaces.

Using the national soil database data, the settlements separated in the Isparta province land asset map in 1991 were determined in the N2 class. However, according to CLC 2018 data, there are changes in the use of soils and it is distributed as artificial surfaces within suitable agricultural areas. This situation shows that agricultural lands are out of agricultural production according to the increasing population and usage needs. It also shows that most of the bare rocks and debris in the N2 class are in the forest and semi-natural areas class. Lake Eğirdir, Lake Burdur, Lake Beyşehir, Karacaören dam, and other water bodies are in the N2 class. However, according to CLC 2018 data, wetlands were also included in other classes. This shows that the reservoir areas of the irrigation structures built for use in agricultural, urban, and industrial irrigation are above areas of suitable agricultural areas. Furthermore, the forest and semi-natural areas fall under class N2 with the highest distribution. This is primarily due to the assignment of low land capability areas and bare rocks to class N2 according to the Corine (2018) data, resulting in a percentage of 65.85%. In class S3, it can be observed that agricultural production parcels are less prominent.

These results corroborate that forest areas are situated in regions more suitable for agricultural production (Table 8). In a study conducted by Büyükgebiz et al. (2009) in Isparta province Sütçüler district, it was determined that the Sütçüler district harbors various plant taxa, including Sütçüler thyme, utilized as a non-timber forest product in diverse industrial sectors such as food, medicine, pharmacy, chemistry, paint, and cosmetics. This insightful research sheds light on the multifunctionality of forested regions, extending beyond conventional perceptions. Our study results contributes to a comprehensive understanding of the diverse roles of forested areas, ranging from non-wood forest product utilization to their suitability for agriculture. This holistic perspective underscores the need for adaptive land use policies that consider the multifunctional nature of landscapes. Using the data from the old American soil classification system, it has been shown that the areas suitable for agriculture, determined by LCT, are open to different uses within the study area. It has been determined that the S1, S2, and S3 classes, which are the suitability classes for agricultural production, of the study area soils are in different land use situations. This situation shows that detailed soil survey and mapping studies should be done according to the new soil classification belonging to the study area and within the borders of Türkiye. The study revealed important results in revealing the change in land use status. It has been determined that the results of the study are compatible with both high-resolution satellite images and visual interpretation in the Google Earth Pro platform, and the results with CLC 2018 data.

According to 1990 and 2022 census data, the overall population of the study area rose from 314,766 in 1990 to 445,325 in 2022 (TÜİK, 2023). The increasing population supports the study’s results that indicate the improper use of S1 and other suitable agricultural areas as artificial surface areas. Furthermore, to assure food supply security for the world’s growing population in the face of climate change, agricultural productivity in recognized viable agricultural regions must be determined using new soil categorization methods. According to the findings of the study, S1 class regions, which are irrigable grounds, are used for fruit and vegetable cultivation. Furthermore, greenhouse farming is becoming increasingly common in S1 class locations of the study area. The results are significant on both a national and worldwide scale.

It was not possible to separate the study area’s identified suitable agricultural areas into categories for agriculture and forests because many food production activities, such as raising livestock and growing medicinal and aromatic plants, are currently carried out in forests to meet the demands of the expanding food supply (Büyükgebiz et al., 2009; Özkan ve Deniz, 2023). Legal backing has also been given in Türkiye for the agricultural use of lands that no longer qualify as forests. The studies have indicated that increasing urbanization due to population growth exerts pressure on natural areas (Öncel & Levend, 2023). The results of the study support the finding that the increase in artificial surfaces within areas suitable for agriculture is indicative of improper land use. Soil in forested regions was included in the evaluation of LCT approach employed in the study.

LCT method serves as a prevalent aggregation technique within GIS. It finds application in creating composite maps, particularly in scenarios like suitable agricultural areas determent, and mapping. Following the generation of relative weights through methods such as AHP, LCT combines these weights to yield a consolidated evaluation score. This approach can be viewed as a hybrid method that bridges the realms of qualitative and quantitative analysis. In the context of agricultural land suitability analysis, AHP method can help to determine weight by considering various factors such as soil quality, climate, topography, and land use. Suitable agricultural areas are determined by LCT method by using the features whose weights are determined (Şenol et al., 2015). Studies that have used LCT method to determine suitable agricultural areas have generally found it to be an effective tool for agricultural land evaluation. For example, a study conducted in Türkiye used LCT method to identify the most suitable areas and found that the method was able to accurately determine areas with high agricultural potential (Dengiz & Özyazıcı, 2018). Another study conducted in Türkiye used LCT method to assess the suitability of different regions for rice cultivation and found that the method was able to provide accurate results, taking into account various factors such as climate, soil, and land use (Tercan & Dengiz, 2022). These two investigations were conducted in local communities and in regions where extensive soil surveys were conducted. The S1, S2, S3, N1, and N2 land suitability classes identified by the research are said to be comparable with the mapping units determined by the soil survey study. For this reason, with the impact of climate change and increasing population, soil, which is the most important natural resource in agricultural production, needs to be used sustainably to meet food demand. Conducting soil surveys and mapping studies on a national and international scale shows the importance of preventing the misuse of agricultural lands. Overall, LCT method has been shown to be a useful tool for evaluating and identifying suitable agricultural areas, and it can be applied in different regions and contexts. However, it is important to note that the accuracy of the results depends on the quality and reliability of the input data used in the analysis.

It is also worth noting that while LCT method is a valuable tool for agricultural land suitability analysis, it should be used in conjunction with other approaches such as field assessments and expert knowledge to ensure comprehensive and accurate results. Therefore, it is crucial to conduct detailed soil survey and mapping studies using the appropriate soil classification system for the study area and to update the data regularly to reflect changes in land use and population growth.

4 Conclusions

This study presents an approach based on LCT and geographic information systems to determine suitable agricultural areas and current use within the borders of Isparta province. The identification of suitable farmlands includes data from the old American system of soil classification criteria and topographical factor criteria based on expert knowledge. It is determined that 39.52% of the criteria are for S1, S2, and S3 classes in the map of suitable agricultural areas using LCT. LCT used in this study and the agricultural suitability map has been great application example for users in Isparta province. Cartographic materials were produced for the institutions and organizations in the study area at the point of developing policies on the use of soils for agricultural and other purposes. According to CLC 2018 land use/land cover data, it is determined that areas with high agricultural production potential are in use as artificial surfaces, forest and semi-natural areas, wetlands, and water bodies. Türkiye’s forests, pastures, and wetlands are protected by law, and misuse is prevented. However, it is determined that 1.88% of suitable agricultural areas (S1, S2, and S3) were used as artificial surfaces. The need for detailed soil surveys and mapping is revealed by using new classification systems of soils at the country level for the food supply and security that agricultural lands are scarce resources and the growing population needs. This is critical to ensuring food supply security for an increasing population around the world. The use of S1 class regions for fruit and vegetable cultivation offers the opportunity to increase the diversity of food products. Therefore, it is important to grow different agricultural products and promote sustainable agricultural practices. Methods used for this study suggested a solution for the application of AHP in the weighting of the data for LCT to determine the areas with high agricultural suitability of increasing technology and land data. The results show that detailed soil surveys and mapping studies with high spatial resolution should be scientifically based to determine and use suitable agricultural areas in agricultural production on a country basis. It is suggested that the criteria determined for the study area can produce results by adding climate data in determining suitable crop-based agricultural areas in LCT. Making use of these techniques widespread is enable planners and decision-makers to reach results faster and more accurately and to use natural resources sustainably.