Modelling organic farming suitability by spatial indicators of GIS integrated MCDA in Golestan Province, Iran

ABSTRACT

Organic farming suitability can improve the health of environment, agroecosystems and humans, quality of products, and local economy. Organic agricultural system is not very much evolved in Iran. In this paper a model is proposed to identify the suitable zones in 14 counties of Golestan Province, northeast of Iran, for the development of organic farming using spatial indicators, spatial analysis and Multi-Criteria Decision Analysis (MCDA). In this model, some important criteria such as climatic variables, topographic factors, soil characteristics, ecological variables, environmental variables and developmental variables were evaluated and considered as spatial indicators. The thematic layers were classified based on agronomical requirements tables of organic farming and were overlaid based on Weighted Overlay Analysis (WOA) in ArcMap software. Final maps were separately generated and classified to five classes of suitability degree for spring and winter crops. According to the results of model, development of organic farming is possible for up to 14.72 and 17.76 percent of the current lands of Golestan Province in Iran for organic spring and winter cropping, respectively. In this research, we developed a land suitability model for organic farming based on the evaluation of spatial variables in Geographic Information System (GIS) and MCDA. Results provided useful information that can be used as decision support tools in the development of organic agriculture in Golestan Province, other similar regions in Iran and other countries in the world.

KEYWORD:

• AHP
• GIS
• land suitability model
• MCDA
• organic farming

1. Introduction

A key challenge for ecological intensification in agroecosystems is to produce increasing amounts of food and feed with minimal greenhouse gas (GHG) emissions, soil erosion, biodiversity loss, and leaching (Schrama et al., 2018). Adverse ecological effects of conventional agroecosystems have increased the demand for more sustainable production systems. Organic farming is accepted as a possible way forward to improve sustainability in these systems (Tuomisto et al., 2012). Organic farming represents a production system without agrochemicals, including insecticides or pesticides, using plants that are not genetically modified. According to Pimentel et al. (2005) and Franz et al. (2010) organic agriculture can augment ecological processes that foster plant nutrition, yet conserve soil and water resources. According to the International Federation of Organic Agriculture Movements (IFOAM International Federation of Organic Agriculture Movements, 2004), organic agriculture is an agricultural production system that promotes environmentally, socially and economic sound production of food and fibres, and excludes the use of synthetically compounded fertilisers, pesticides, growth regulators, livestock feed and additives and genetically modified organisms (Rattanasuteerakul & Thapa, 2010). IFOAM defines the four principles of organic agriculture as ecology, health, fairness and care. These principles should be integrated and considered as a whole when applied to the development of organic farming (Luttikholt, 2007).

The objectives, rules and principles of organic agriculture are clearly stated by the European Commission (EU European Commission, 2015). Based on European Commission, there are three rules as (1) prohibition of the use of synthetic fertilisers and synthetic herbicides and pesticides, (2) the requirement to use only seeding material and propagating material produced organically, and (3) the requirement to apply wide crop rotations (Berentsena et al., 2016).

Organic agriculture originated as a response to a growing awareness that the health of the land is linked to the health and future of the people. This system is today practiced in almost every country in the world, and the amount of certified organic land is growing as well (Carolan, 2016). Organic farming systems showed lower surface runoff and higher water infiltration capacity than the conventional agriculture systems. Therefore, the conversion from conventional to organic farming resulted in higher infiltration and lower soil surface runoff which is a benefit of long-term organic farming systems to reduce floods and erosion hazards (Zeiger & Fohrer, 2009). Cranfield et al. (2010) revealed that profit, economic and financial issues, environmental, health and safety concerns, and ideological and philosophical motives were important to conversion from conventional systems to organic systems. Several studies have investigated the potential of organic farming as a tool to enhance farmland biodiversity, with results varying mainly because of the moderating effect of landscape complexity (Goded et al., 2018; Tuck et al., 2014).

Identification of suitable sites for organic farming requires consideration of different criteria such as climatic, soil, socio-economical factors, water resources, topography and topological and geophysical variables (Roig-Tierno et al., 2013). Multi-Criteria Decision Analysis (MCDA) analysis by Geographic Information System (GIS) and Remote Sensing (RS) techniques is used as a decision-making tool for the identification of suitable organic farming sites. As a MCDA method, the analytic hierarchy process (AHP) has been applied for solving different problems that involve complex criteria across different levels, where the interaction among criteria is common (Feizizadeh et al., 2014; Tiwari et al., 1999). GIS-based MCDA is usually used in land suitability analyses. For example, Mishra et al. (2015) concluded that remote sensing and GIS can play an important role in the identification of the suitable zones for the development of organic farming in Uttarakhand region, India. In another study, Kazemi and Akinci (2018) developed a land evaluation model for rainfed farming using GIS and MCDA in Golestan Province, Iran. In similar research in China, a comprehensive evaluation of tobacco land suitability was carried out by principals of fuzzy and hierarchy analysis mathematics, and the technique of GIS in the Henan region (Chen et al., 2010). A web-based framework was developed by Yalew et al. (2016). They were integrated various data from different sources for multi-criteria-based agricultural land suitability assessment by the Google Earth Engine platform. In Iran, Kazemi et al. (2016) suggested a land suitability model for faba bean cropping using spatial multi-criteria analysis.

Sarkar et al. (2021) investigated the soil fertility for Tulaipanji rice cultivation in Kaliyaganj of India, using AHP and machine learning algorithms. Results showed that 18.01% of the land was in excellent health condition to support Tulaipanji cultivation. In another study, Saha et al. (2021) developed a GIS-based AHP model to identify suitable sites for agricultural practices in the Anabranching site of Sooin river, India. Final map was classified into five categories as very high suitable zone (4.11%), high suitable zone (28.65%), moderate suitable zone (49.67%), low suitable zone (14.33%) and unsuitable zone (3.24%).

Organic farming system covered 50.9 million hectares worldwide in 2017, about 1.1% of total agricultural land in 179 countries (Willer & Lernoud, 2017). Organic agricultural system is not very much evolved in Iran. According to FAO statistical report, the organic farming areas in Iran covered 18,000 ha in 2016. In this paper a model is proposed to identify the suitable zones in Golestan Province, northeast of Iran, for the development of the organic agriculture. One of the most important areas for crop production in Iran is Golestan Province. The cultivated areas of this province are about 700,000 ha under conventional cropping systems (Agricultural Organization of Golestan Province, 2017). The main economic activity in Golestan Province is agriculture and the most important crops grown in this province are wheat, barley, soybean, rape seed, cotton, rice, faba bean, sunflower, and potato. Some reports showed the high dependency of agroecosystems in Golestan on purchasing resources, especially chemical fertilisers, pesticides, fossil fuel and pretty high economic costs. From an environmental viewpoint, high chemical fertiliser and pesticide consumption, agrobiodiversity loss, soil erosion, low quality of produced crops are the main factors that influence the health of agroecosystems and humans. Mostly, production methods are unsustainable, economically and ecologically risky, contributing to soil degradation, water depletion, and poor health in Golestan Province. Organic agriculture reflects one option towards more sustainable development. Organic agriculture represents a production system without agrochemicals, including pesticides, using plants that are not genetically modified (Rattanasuteerakul & Thapa, 2010). Therefore, the objective of the present study is a land suitability model development for organic farming based on the evaluation of spatial variables using MCDA in the agricultural lands of 14 counties in Golestan Province. Accordingly, farmers can develop their production under organic agriculture systems in this region. In addition, this study provides raw information at a local level that can be used by provincial managers to grow organic farming system in the developing counties.

2. Materials and methods

2.1. Study area

This study concentrated on the croplands of 14 counties in Golestan Province, the northeast of Iran (Figure 1). The regions coordinates range from 36° 44′ and 38° 5′ N latitudes and 53° 51′ and 56° 14′ E longitudes. According to De-Martonne’s advanced climate classification system, Golestan Province has the five different climate classes: semi-arid, mediterranean, humid, semi-humid, and arid desert (Kazemi et al., 2016). The province covers approximately 20,033 km². The eastern part of the Alborz Mountains Range surrounds the coastal plains of the Caspian Sea as a long, high wall, thus, all over the Golestan Province, the land slope decreases from the southern and eastern mountains towards sea (Golestan Province Government, 2009).

Figure 1. The location of the Golestan Province and its counties in Iran (reference:www.iranmap.com) and location of meteorological stations in Golestan Province.

Figure 2. (Continued).

2.2. Spatial indicators and data used

2.2.1. Soil

The soil properties data were obtained from 485 sampling sites distributed in croplands of Golestan Province, including electrical conductivity (EC), organic carbon (OC), pH, and texture. Sampling depth was 0–30 cm. First, we separated agricultural land use from other land uses and all soil sampling was done exactly in this land use. We had to use random distribution for soil sampling points. Because, agricultural lands in Golestan Province have a non-uniform distribution. Therefore, we identified the rural districts and then, in each village we sampled in four main geographical directions. The spatial distribution of these characteristics on the agricultural lands was evaluated using different geostatistical and interpolation methods such as, Ordinary Kriging and Inverse Distance Weighted and Local Polynomial Interpolation in ArcMap var.10.3 software. The layers of these soil indicators were produced by Geostatistical Analyst tools in the ArcGIS software. Soil-related maps of the study area are shown in Figure 2.

Figure 2. Soil maps in croplands of Golestan Province, Iran.

2.2.2. Climate

Climatic data were obtained from 95 meteorology stations located within the study area (Figure 1). The data were averaged from 1995 to 2017. Relative humidity, average, minimum, and maximum temperatures, radiation, and annual rainfall, were considered as climatic factors affecting organic farming land suitability model. Some interpolation methods, e.g. Local Polynomial Interpolation, Ordinary Kriging, and Inverse Distance Weighted were selected to provide a raster-based spatial distribution of climatic parameters. Final interpolated layers of these indicators are presented in Figure 3.

Figure 3. Climatic maps in croplands of Golestan Province, Iran.

Figure 3. (Continued).

Figure 3. (Continued).

2.2.3. Topography

The digital elevation model (DEM) dataset with a 90 × 90 m resolution was obtained from the Natural Resources and Watershed Management Organization of Golestan Province. Also, topographic indicators such as slope and elevation were obtained from the DEM by surface analysis function in ArcGIS var. 10.3 software (Figure 4).

Figure 4. Elevation and slope maps in croplands of Golestan Province, Iran.

2.2.4. Ecological indicators

In order to determine the land suitability degree for organic farming, some ecological indicators such as distances to water resources, agrobiodiversity, distances to natural ecosystems and Normalised Difference Vegetation Index (NDVI) as health vegetation index were used. The natural ecosystems and water resources data were obtained from the Natural Resources and Watershed Management Organization of Golestan Province and Regional Water Company of Golestan, respectively. Also, data of agrobiodiversity (species richness) were obtained from Ministry of Jihad-e-Agriculture, in 2017. The layers of these indicators were produced by Geoprocessing and Multiple Ring Buffer tool in the ArcGIS software.

Satellite data consisting of multi-spectral data acquired by OLI/TIRS sensor of Landsat satellite 8 for 7 and 14 April and 12 and 19 September 2017, that provided by The United States Geological Survey (USGS) Earth Explorer (https://earthexplorer.usgs.gov/) were used for detecting NDVI. It is one of the most commonly used vegetation indices in environmental studies. NDVI is calculated based on near-infrared (NIR) and red (RED) light reflectance assessments as:

NDVI = (NIR−RED)/(NIR+RED) (1)

Where NIR and RED are the amounts of light reflected by growing vegetation and registered by satellite sensor (Borowik et al., 2013; Jackson & Huete, 1991). This layer was produced by Erdas Imagine var. 2014 software. All produced layers of ecological indicators are presented in Figure 5.

Figure 5. Ecological variables maps in croplands of Golestan Province, Iran.

Figure 5. (Continued).

Figure 5. (Continued).

2.2.5. Development parameters

In order to determine the land suitability for organic farming, two important development variables include distances to local markets and distance to local roads were used. Farmers can sell their organic products to local consumers, restaurants and processors. This short supply chain markets represent profitable opportunities for organic farmers. These direct markets can bring farmers gross returns higher than sales to whole sellers (Dimitri, 2012). Also, local roads are required for easy transportation and access to local markets. Therefore, if access to roads is improved, products will sell and transport easier and faster. In this research, row data of these parameters was obtained from Golestan Province Government and Golestan Roads and Transports Organization, respectively. The layers of these indicators were produced by buffer functions in the ArcGIS software. Development-related maps of the study area are shown in Figure 6.

Figure 6. Development and environmental variables maps in croplands of Golestan Province, Iran.

Figure 6. (Continued).

2.2.6. Environmental indicators

One of the polluting factors of the environment is the excessive consumption of chemical fertilisers and pesticides. In most areas of Golestan Province, agriculture is performance as conventional system with high amounts of pesticides and chemical inputs. So, in the transition phase and in order to convert conventional agricultural lands to organic farming based on organic farming standards, if the soil of the fields is less contaminated with pesticides and chemicals, this phase will be short and the fields will convert to organic farms sooner and easier. In this study, two indicators from environmental variables include rates of pesticides and chemical fertilisers were evaluated. The required data was obtained from Agricultural Organization of Golestan Province (2017), in the regional scale. These maps of these indicators are presented in Figure 6.

2.3. Standardization of layers

After data gathering and production of thematic layers, all the spatial indicators were converted into raster layers with 50 m resolution and georeferred to UTM (WGS-84) coordinate system in ArcGIS environment. The land use map of province was used to extract layers of non-planting regions. Then, all the raster layers from soil, topography, ecology, environment, development and climatic factors were standardised and divided into five classes based on requirements of the spring and winter crops in an organic farming system, which is tabulated in Tables 1 & 2. These five classes included highly suitable (S1), moderately suitable (S2), marginally suitable (S3), currently unsuitable (N1) and permanently unsuitable (N2) (FAO, 1976; Sys et al., 1991).

Table 1. Criteria for delineating land suitability of spring organic farming in Golestan Province, Iran.

Variables	High suitable(S1)	Moderately suitable (S2)	Marginally suitable(S3)	Currently unsuitable (N1)	Permanently unsuitable (N2)
Annual precipitation (mm)	550<	450-550	350-450	250-350	250>
Annual average temperature (◦c)	20-23	23-25 18-20	25-28, 15-18	28-33	33< 15>
Maximum temperature (◦c)	20-30	23-25	25-30, 15-20	30-35	35<
Minimum temperature (◦c)	Dec-14	10-Dec	09-Oct	08-Sep	8>
Relative humidity (%)	45-55	55-65, 45>	65-75	75-85	85<
Radiation (MJ/m^2)	12<	10-Dec	09-Oct	08-Sep	8>
Slope (%)	0-3	03-May	05-Aug	08-Dec	12<
Elevation (m)	1000>	1000-1500	1500-2000	2000-3000	3000<
Organic matter (%)	3<	02-Mar	1.5-2	1-1.5	1>
Soil texture	Loam, Sandy	Clay loam, Sandy clay loam, Silty clay loam	Sandy loam, Silty clay, Sandy	Clay, Silty, Silty clay	Sandy and other classes
	loam, Silty
	loam
EC (dS/m)	0-4	04-Jun	06-Aug	08-Dec	12<
pH	6.5-7.5	7.5-8 6-6.5	5.5-6	8-8.5 5-5.5	8.5< 5>
Distance to natural ecosystems (Km)	1>	01-Mar	03-Jun	06-Oct	10<
Distance to water resources (Km)	0.5>	0.5-1	1-1.5	1.5-2	2<
Agrobiodiversity	31<	25-31	20-25	16-20	16>
Normalized difference vegetation index (NDVI)	0.3<	0.2-0.3	0.1-0.2	0.1-0	0
Distance to roads (Km)	1>	01-Mar	03-Jun	06-Oct	10<
Distance to local markets (Km)	1>	01-Mar	03-Jun	06-Oct	10<
Consumption of chemical fertilizers (Kg/ha)	174>	174-209	209-250	250-300	300<
Consumption of pesticides (Kg.ha)	2.18>	2.18-2.62	2.62-3.15	3.15-3.78	3.78<

Table 2. Criteria for delineating land suitability of winter organic farming in Golestan Province, Iran.

Variables	High suitable(S1)	Moderately suitable (S2)	Marginally suitable(S3)	Currently unsuitable (N1)	Permanently unsuitable (N2)
Annual precipitation (mm)	500<	400-500	300-400	200-300	200>
Annual average temperature (◦c)	18-25	25-28, 15-18	28-30, Dec-15	30-33, 10-12	33<, 10>
Maximum temperature (◦c)	18-25	25-28	28-31, 15-20	33-33	33<
Minimum temperature (◦c)	Oct-16	08-Oct	06-Aug	06-Apr	4>
Relative humidity (%)	45-55	55-65, 45>	65-75	75-85	85<
Radiation (MJ/m^2)	12<	10-Dec	09-Oct	08-Sep	8>
Slope (%)	0-3	03-May	05-Aug	08-Dec	12<
Elevation (m)	1000>	1000-1500	1500-2000	2000-3000	3000<
Organic matter (%)	3<	02-Mar	1.5-2	1-1.5	1>
Soil texture	Clay loam, Clay	Loam, Silty clay loam, Sandy loam, Sandy clay loam	Sandy loam, Silty loam, Clay Sandy	Silty, Silty clay	Sandy and other classes
EC (dS/m)	0-4	04-Jun	06-Aug	08-Dec	12<
pH	6.5-7.5	7.5-8 6-6.5	5.5-6	8-8.5 5-5.5	8.5< 5>
Distance to natural ecosystems (Km)	1>	01-Mar	03-Jun	06-Oct	10<
Distance to water resources (Km)	0.5>	0.5-1	1-1.5	1.5-2	2<
Agrobiodiversity	31<	25-31	20-25	16-20	16>
Normalized difference vegetation index (NDVI)	0.3<	0.2-0.3	0.1-0.2	0.1-0	0
Distance to roads (Km)	1>	01-Mar	03-Jun	06-Oct	10<
Distance to local markets (Km)	1>	01-Mar	03-Jun	06-Oct	10<
Consumption of Chemical fertilizers (Kg/ha)	174>	174-209	209-250	250-300	300<
Consumption of pesticides (Kg/ha)	2.18>	2.18-2.62	2.62-3.15	3.15-3.78	3.78<

2.4. Model description

In order to determine organic agriculture land suitability, a spatial model based on GIS was created to match the requirements of organic farming system for spring and winter crops with the land characteristic of Golestan Province. Schematic diagram of the steps for performance of this model is shown in Figure 7. After mapping of essential layers, the agronomical requirements for organic farming system were identified from a comprehensive literature review and asking local experts’ opinions (Tables 1& 2)

Figure 7. Flowchart of land suitability model for organic farming performance. a) Spring farming, b) Winter farming.

In the next step, AHP acquired weights were assigned to the criteria. Basically, AHP is as an important and universal method from MCDA. It was developed by Saaty (1980), for setting-up a hierarchical model based on criteria and alternatives for representing the complex problems (Roig-Tierno et al., 2013). In AHP model, the objective was set as the land use suitability analysis for organic farming in Golestan Province, and the criteria including topography, development, environment, ecology, climate, and soil. A nine-point scale based on Saaty (1980) approach is used to assign the relative importance of criteria. The weights of factors for land suitability were obtained from local experts, through a pairwise comparisons statistical analysis in Expert Choice software. Weights are within the range of 0–1, and their sum is equal to 1. To measure the consistency of pairwise comparison judgements, the Inconsistency Ratios (IR) proposed by Saaty (1980) was used. The Inconsistency Ratios (IR) for this pair-wise comparison matrix was equal to 0.1. This shows that the comparisons of land characteristics were perfectly consistent, and the relative weights were appropriate for applying in land suitability analysis using AHP. In this research, 25 agricultural specialists from Golestan Province were invited to fill pairwise comparison table to estimate the relative importance of the criteria. Then, parameter weights and sub-parameter scores were appointed to the related layers, and thematic and classified maps of 20 factors were overlaid using Weighted Overlay Analysis (WOA) in the ArcGIS software by raster calculator tools. Finally, the organic farming suitability maps were separately generated for spring and winter crops and divided into five classes of equal ranges including highly suitable (S1), moderately suitable (S2), marginally suitable (S3), currently unsuitable (NS1), and permanently unsuitable (NS2) areas.

3. Results and discussion

3.1. Results of AHP analysis

The results indicated that the most important criteria were climate, soil, ecology, environment, development and topography according to their specific weighting. Also, among the climate criteria, annual rainfall and maximum temperature had the highest and lowest weights, respectively. It should be noted that, among of the soil criteria, organic matter and electrical conductivity (EC) played a major role in the delineation of suitable areas for organic farming (Table 3). Lai et al. (2002) applied AHP in group decision making, which has proved to be more beneficial than the conventional techniques such as the Delphi techniques. In general, combining the potential of spatial analysis with AHP analysis enables researchers to understand the potential value of the organic crops production in this region. Mishra et al. (2015) reported that the AHP technique can help to identify and prioritise the potential sites for the organic farming in Uttarakhand province (India). Based on their results, weighted overlay method along with the AHP had very favourable result for the site suitability analysis of organic farming in this region. The results of Sajadian et al. (2017) showed that management of pest and disease, yield, soil nutrient management, water consumption rate, chemical fertiliser consumption rate and the use of transgenic materials had the highest weights for development and assessment of organic farming in Iran, respectively.

Table 3. The weights of criteria and sub-criteria in land suitability analysis of organic farming in Golestan Province, Iran.

Criteria /Sub criteria	Weight
1. Climate	0.348
Annual precipitation	0.434
Annual average temperature	0.181
Minimum temperature	0.092
Maximum temperature	0.053
Relative humidity	0.134
Radiation	0.106
2. Topography	0.074
Slope	0.778
Elevation	0.222
3. Soil	0.203
Organic matter	0.474
Soil texture	0.128
EC	0.205
pH	0.193
4. Ecology	0.175
Distance to natural ecosystems	0.221
Distance to water resources	0.257
Agrobiodiversity	0.338
NDVI	0.184
5. Environment	0.126
Consumption of chemical fertilizers	0.537
Consumption of pesticides	0.463
6. Development	0.073
Distance to roads	0.242
Distance to local markets	0.758
Inconsistency Ratio (IR)	0.01

3.2. Spring crops

3.2.1. Highly suitable zone (S1)

The highly suitable zone for organic agriculture involved all the southern areas of Golestan Province (Figure 8). This area can have a high potential production for organic production and it is suitable for the growth of current spring crops. The organic farming suitability model performance indicated that 14.72% of the surveyed area was located in highly suitable class (116,944.88 ha) (Table 4). These areas had the suitable properties such as low elevation, sufficient rates of annual rainfall and optimum temperatures, low EC, high organic matter in soil, high agrobiodiversity, suitable pH, low consumption of pesticides, near distance to local markets, roads and natural ecosystems.

Figure 8. Land suitability map for spring and winter organic farming in Golestan Province, Iran.

Table 4. The distribution of land suitability degree for organic farming in Golestan Province, Iran.

Suitability degree	Area (ha)		Area (%)
	Spring	Winter	Spring	Winter
Highly suitable (S1)	116944.88	140901.23	14.72	17.76
Moderately suitable (S2)	199405.29	193396.86	25.11	24.39
Marginally suitable (S3)	225548.20	207751.66	28.40	26.19
Currently unsuitable (NS1)	205727.32	189715.95	25.91	23.92
Permanently unsuitable (NS2)	46580.59	61410.70	5.86	7.74

Our results based on long-term climate data showed that organic farming of spring crops is not faced by limiting cardinal temperatures and radiation during the growing season. According to Levin (2007), the positive influence of organic farming on landscape composition can be the result of several factors.

3.2.2. Moderately suitable (S2) zone

Moderately suitable (S2) areas were identified as sites that are moderately advantageous for the performance of organic farming. These areas covered an area of 199,405.29 ha, which represented 25.11% of the total surveyed area (Table 4). Basically, moderately suitable zone is defined by annual rainfall of 450–550 mm, elevation levels between 1000 and 1500 m, slope 3–5%, soil pH range of 7.5–8 and 6–6.5, EC 4–6 dS m⁻¹, and organic matter range of 2–3%. It was found that the moderately suitable class (S2) was located from west to east in Golestan Province (Figure 8). Our results, based on long-term climate data, showed that organic farming of spring crops is faced by limiting rainfall, distance to water resources and relative humidity during the growing season.

3.2.3. Marginally suitable (S3)

The marginal class was identified as the boundary between suitable zones and unsuitable zones (Figure 8). This class consisted of as 225,548.20 ha area that covered about 28.40% of the studied area (Table 4). These regions were characterised by slope 5–8%, elevation 1500–2000 m, insufficient annual rainfall, EC6–8 dSm⁻¹, more distance to local markets and natural ecosystems, low agrobiodiversity and more consumption of chemical pesticides and fertilisers and higher relative humidity in west areas. In this class, the cultivation of spring crops will face high risk due to drought and salinity in north and northeast. From a climatic viewpoint, rainfall and temperature are two main factors that influence crop productivity in arid and semi-arid regions, and these two variables are the principal weather variables that determine the variability of crop yield in Iran (Kazemi & Akinci, 2018). By applying some management methods, crop production can play an important role in improving this class to a higher class.

3.2.4. Currently unsuitable (NS1) and permanently unsuitable (NS2) zones

The north and northeast areas, as well as parts of the west of Golestan Province, were identified as a currently unsuitable (NS1) class for spring crop production as organic farming system (Figure 8). In these areas, development of organic farming is not possible with the current condition of the ecological resources. Spring crop production in the northern regions is constrained by low and variable rainfall. Also, the results showed that some variables such as low agrobiodiversity and organic matter, high relative humidity and consumption of pesticides in west, high EC and low rainfall in north were limiting factors in these regions. Globally, agricultural expansion and intensification led to biodiversity loss in agroecosystems (Tscharntke et al., 2012) and reduction in the types and levels of ecosystem services (Barral et al., 2015). Agrobiodiversity is base for provision of ecosystem services needed to sustain agriculture per se and the environment as a whole (Overmars et al., 2014).

The permanently unsuitable (NS2) zone involved the northern and southern areas of Golestan Province (46,580.59 ha) (Figure 8). In this zone, variables such as slope, salinity, organic matter content in soil, soil texture classes, rainfalls, and distance to roads were identified as limiting factors for organic farming performance in Golestan Province. From a geomorphological viewpoint, slope and elevation in this province varies, decreasing from the south to the north. The results revealed that in the southeast parts of Golestan, changing land use from forest ecosystem to cropland acts as a limiting factor for the cultivation of organic crops because of high slope, hazards of runoff and soil erosion and land degradation. We recommend that this area should be converted to other land use near to natural ecosystems such as pasture, grassland, forest planting, ecotourism and wildlife. Generally, the conversion from conventional system to organic farming system resulted in lower surface runoff and higher infiltration which is a benefit of long-term organic farming systems to reduce erosion hazards and floods (Zeiger & Fohrer, 2009). In this study, the distance to roads, local markets and natural ecosystems were recognised as limiting factors for organic farming system development.

3.3. Winter crops

3.3.1. Highly suitable zone (S1)

This zone is most suitable for production of some winter crops such as wheat, barley, canola, faba bean, field pea and potato. Highly suitable regions were identified as sites that are highly advantageous for the performance of organic farming system (Figure 8). These areas covered an area of 140,901.23 ha, which represented 17.76% of total surveyed area (Table 4). Therefore, the highly suitable (S1) areas have a high potential production and sustainability of yield from year to year (Ghafari et al., 2000). Highly suitable areas were characterised by organic matter amounts range of 2–3%, EC < 4 dS m⁻¹, slope<4%, near distance to roads and markets, NDVI>0.4, sufficient rainfall, low consumption of chemical fertilisers and pesticides, and elevation<1000 m. Also, this zone had the high agrobiodiversity (species richness) than other zones. Some researchers showed that the intensification of modern agroecosystems is amongst the chief current pressures to biodiversity (Baudron & Giller, 2014; Hole et al., 2005; Kazemi et al., 2018; Shoyama & Yamagata, 2014).

3.3.2. Moderately suitable zone (S2)

Moderately suitable areas covered an area of 199,396.86 ha, which represented 24.39% of the total surveyed area (Table 4). Basically, moderately suitable zone is defined by annual rainfall of 400–500 mm, elevation levels<1500 m, slope<5%, soil pH range of 6–8, EC < 6 dS m⁻¹, suitable temperatures and relative humidity, and organic matter range>2%. It was found that the moderately suitable class (S2) was located from west to east in Golestan Province (Figure 8). It was identified that relative humidity, accessibility of water resources and roads were the limiting factors in this class than highly suitable class (S1).

3.3.3. Marginally suitable zone (S3)

Marginally suitable zone (S3) covered an area of 207,751.66 ha, which represented 26.19% of the total surveyed area (Table 4). There is marginally suitable zone in the central and northern parts of the investigated region (Figure 8). These areas show the variable potential of production, with considerable associated risks of low yields and high economic costs, which are due to the climate interacting with soil properties, diseases and pest problems (Ghafari et al., 2000) or developing variables such as distance to roads and markets. In Golestan Province, with the removal of some restrictions such as soil salinity, low organic matter, availability of water suitability, high consumption of chemicals, the suitability degree of these areas can be enhanced. There are some main variables, such as salinity and low rainfall, which limit both suitable land area and actual yield of winter crops in organic farming. For more accurate and beneficial results, the study needs to be focused on some specific plants like cover crops, which have great ecological value.

3.3.4. Currently unsuitable (NS1) and permanently unsuitable (NS2) zones

In currently unsuitable (NS1) and permanently unsuitable (NS2) zones, performance of organic farming system is not possible with the current state of the environmental resources. Winter crop production in the northern regions is constrained by low and variable rainfall, high soil salinity, low organic matter and low agrobiodiversity. The north and northeast areas, as well as parts of the west and south of Golestan Province, were identified as a currently unsuitable (NS1) class (Figure 8). These areas covered an area of 189,715.95 ha, which represented 23.92% of total evaluating area (Table 4). The results revealed that in the west parts of Golestan, high relative humidity and high consumption of chemical fertilisers and pesticides acts as limiting factors for the performance of organic farming system. These zones had the low organic matter percent in comparison to other zones. Principally, the use of green manures and organic fertilisation can be beneficial for soil fertility in modern agricultural systems (Baldivieso-Freitas et al., 2018). Also, green manuring can maintain soil organic carbon and increase total nitrogen in soil, only if introduced for a sufficient number of years during crop rotation (Sacco et al., 2015). Generally, in many farming systems, N fertiliser management poses a major challenge due to its high mobility and propensity for loss from the soil-plant system into the ecosystem. Efficient nitrogen management techniques are required to improve nitrogen delivery and retention in soils in order to increase N-use efficiency, improve economic yields and improve the economic sustainability of most farming systems (Garnett et al., 2009; Musyoka et al., 2017; Van Eerd, 2005).

In this study, the permanently unsuitable (NS2) regions were characterised by slope>12 percent, insufficient annual rainfall, EC > 12 dS m⁻¹, more distance to local markets and roads (>10 km), low agrobiodiversity, higher consumption of chemical pesticides and fertilisers and higher relative humidity and NDVI<0.1. Results of Borowik et al. (2013) indicated that in the forage fields of eastern Poland, NDVI (>0.3) and ground vegetation biomass were positively related, with a stronger correlation between the two variables occurring in summer. The permanently unsuitable zone had the lowest area among land suitability classes. This zone covered an area of 61,410.70 ha, which represented 7.74% of the total studied area (Table 4).

In Golestan Province, use of green manures, cover crops and organic fertilisation can improve the fertility of soil in organic system and reduce the environmental effects of chemical fertilisation. Several studies have investigated the potential of organic farming as a tool to enhance soil fertility, farmland biodiversity, higher infiltration and lower soil surface runoff (Goded et al., 2018; Zeiger & Fohrer, 2009). Sacco et al. (2015) demonstrated that organic fertilisation is known to improve soil chemical properties and to increase its soil microbial biomass activity. Some researchers showed that the conversion from conventional to organic farming system resulted in lower soil surface runoff, erosion, floods and higher infiltration (Zeiger & Fohrer, 2009). Also, organic farming benefits biodiversity mainly because of the restricted use of chemicals, choice of crop type and crop rotations (Goded et al., 2018; Smith et al., 2010). Over the last quarter of the 20th century, dramatic declines in both range and abundance of many species associated with farmland have been reported in Europe, leading to growing concern over the sustainability of current intensive farming practices. Therefore, organic agriculture systems are now seen by many as a potential solution to this continued loss of biodiversity (Hole et al., 2005). In this study, some developmental variables were recognised as limiting factors for organic farming system development. Bjørkhaug and Blekesaune (2013) showed that there were especially strong connections between the level of organic farming system with the farm processing of organic products, the population level in the municipalities, neighbourhoods effects and access to consumers, in particular regions of Norway

4. Conclusion

This study provides useful information at provincial level that could be used by farmers to select organic cropping patterns in accordance with land use suitability results. According to the results of model, development of organic farming is possible for up to 14.72 and 17.76 % of the current croplands of Golestan Province for organic spring and winter cropping, respectively. Adverse ecological effects of conventional agriculture have increased the demand for more sustainable systems such as organic farming. Today, organic farming is accepted as a possible way forward to improve sustainability level in agroecosystems. Achieving the development goals of organic agriculture requires dedicated compliance with the diversified principles of organic farming. Also, our results confirmed that GIS, spatial indicators and MCDA play important roles in this study as a platform for spatial analysis. This study could also serve as a reference of organic farming suitability model in similar regions especially in developing countries.

In this study, some factors include precipitation, relative humidity, access to water resources and roads, EC, organic matter and consumption of chemical inputs in some regions of Golestan were identified as limiting factors for development of organic agriculture. In general, we can conclude that the development of organic farming in the areas with highly suitable of Golestan Province can lead to healthy and higher quality agricultural products and also prevent environmental problems such as water and soil pollution and incidence of some diseases. Farmers can sell their organic products in local markets. This short supply chain markets represent profitable opportunities for organic farmers. These direct markets can bring farmers gross returns higher than conventional agriculture systems. These results were useful for decision makers to determine the quality of agricultural lands for this system as a decision and planning support tool. We recommend that for more clarification, the same studies using other environmental criteria such as nitrate leaching, soil phosphate contents, greenhouse gas (GHG) emissions, situation of drainage and irrigation systems and other social-economical parameters also should be planned.

f this article.

Acknowledgments

We thank the Gorgan University of Agricultural Sciences and Natural Resources (GUASNR) and Jihad-e-Agriculture Management of Golestan Province (Iran) that supported this research.

Disclosure statement

The authors report there are no competing interests to declare.

Additional information

Funding

No funding was received to assist with the preparation of this manuscript.