A Land Sustainability Model for Identifying Strategic Agricultural Areas: Application to Novo Mesto Municipality, Slovenia

Abstract

The conversion of agricultural land for urban and infrastructure uses poses a significant threat to food security and agricultural sustainability across Europe. Slovenia has experienced particularly severe agricultural land loss, with approximately 70,000 hectares converted to non-agricultural purposes since 1991. This reduction is considerable, as only 24% (481,000 ha) of the country’s land is suitable for agricultural production. In response, the Government of the Republic of Slovenia has enacted regulations to define and protect strategically important agricultural areas, which require robust methodologies for their identification and classification. This study develops and implements a transparent land sustainability model to classify agricultural land into three strategic categories: permanently protected agricultural land of the highest quality requiring strict protection (PPAL), other agricultural land suitable for continued agricultural use (OAL), and potentially suitable areas for agriculture (PSAs). The methodology utilises a multicriteria evaluation of soil quality, agricultural infrastructure, topography, current land use, and the cultural characteristics of agricultural production or landscape, drawing on readily available datasets to ensure replicability. In Novo mesto Municipality, southeastern Slovenia, the classification identified 6490 ha (28%) as PPAL, 1020 ha (4%) as OAL, and 5078 ha (22%) as PSA. This research provides a transparent and replicable multicriteria evaluation framework for strategic agricultural area classification, addresses methodological gaps in existing approaches, and demonstrates practical application in a post-transition European context.

1. Introduction

The loss of agricultural land to urbanisation and infrastructure development poses a critical threat to food security and agricultural sustainability across Europe. Agricultural land systems play a fundamental role in supporting food security and multiple United Nations (UN) Sustainable Development Goals (SDGs), yet face mounting pressures from competing land uses [1]. Slovenia has experienced particularly severe agricultural land loss, with approximately 70,000 hectares converted to non-agricultural uses since independence in 1991, representing a significant reduction in a country where only 24% (481,000 ha) of land is suitable for agriculture [2]. This loss occurs against a backdrop of global food supply disruptions, climate change impacts, and growing concerns about regional food self-sufficiency [3]. In response, the Government of the Republic of Slovenia introduced regulations to define and protect agricultural areas of strategic importance, necessitating robust methodologies for identifying and classifying such areas.

Recent European research demonstrates the critical role of spatial planning instruments and land evaluation frameworks in protecting agricultural land. Scenario modelling in Switzerland found that the revised spatial planning act could reduce potential prime cropland consumption for new urban areas by a factor of six compared with extrapolated past trends. However, strict implementation remains essential to meet protection targets and maintain food self-sufficiency levels [3]. Similarly, coastal regions across the Mediterranean face intense urbanisation pressures, with land consumption growing even where populations stabilise or decline, underscoring the need for strategic agricultural area designation in spatial planning frameworks [4]. These studies demonstrate that effective farmland protection requires both robust evaluation methodologies and strong regulatory implementation.

Multicriteria decision analysis (MCDA) has emerged as a powerful tool for assessing agricultural sustainability and supporting land-use planning decisions. MCDA methods enable systematic integration of multiple environmental, economic, and social criteria to evaluate land suitability and prioritise protection efforts [5,6]. When applied to agri-environmental assessment across European Union countries using MCDA methods (Simple Additive Weighting (SAW), TOPSIS, and EDAS), Finland, Ireland, and Sweden ranked best overall, while the Netherlands, Denmark, and Germany ranked worst, demonstrating MCDA’s capacity to reveal cross-national patterns in agricultural sustainability [5]. Mathematically sophisticated methods are valuable for ranking alternatives; however, in the participatory and politically sensitive stages of the planning process, user-friendly, transparent methods are more appropriate and recommended. The integration of MCDA with geographic information systems (GISs) has proven particularly valuable for spatial suitability analysis, with fuzzy PROMETHEE and other approaches enabling robust land suitability rankings under uncertainty in criteria weighting [7]. These methodological advances support more transparent and defensible land classification decisions in planning contexts.

Peri-urban agricultural landscapes require comprehensive evaluation frameworks that integrate multiple sustainability dimensions to ensure effective protection. Integrated approaches combining ecosystem services evaluation with spatial planning instruments have proven valuable for systematically identifying highly valued agricultural landscapes based on their multifunctional contributions [8]. Such frameworks reveal that effective agricultural land protection depends on both biophysical suitability criteria and explicit recognition of diverse ecosystem services provided by peri-urban agricultural areas [8]. Integration of agricultural land evaluation into binding spatial planning regulations is critical for balancing development with preservation. Multiple evaluation approaches systematically assessing peri-urban agricultural areas across environmental, economic, and social dimensions provide essential foundations for effective regulatory instruments [9]. European case studies demonstrate that successful integration requires explicit evaluation criteria, transparent processes, and clear linkages between classification categories and planning regulations [9]. These frameworks must function as policy instruments embedded within statutory planning processes to ensure legally binding protection.

Strategic area designation and farmland zoning approaches increasingly integrate multifunctional perspectives that consider supply, demand, and coupling relationships. In China’s Hangzhou Metropolitan Area, a farmland zoning approach combining multifunctional supply, demand, and coupling analysis yielded six distinct farmland-use zones, demonstrating how multifunctional metrics yield differentiated protection zones tailored to urban-region contexts [10]. Chinese spatial regulation frameworks for farmland protection demonstrate the importance of explicit zoning based on systematic multicriteria evaluation integrating soil quality, topography, infrastructure, and landscape fragmentation [11]. Stakeholder-driven spatial targeting for cultivated land consolidation in Jiangsu Province integrated spatial multicriteria analysis with stakeholder co-production, categorising townships into seven consolidation priority types and identifying that 41.82% of the area suited resource improvement consolidation, while 33.83% suited pattern optimisation [12]. These approaches demonstrate the value of integrating stakeholder perspectives and multifunctional assessment into strategic agricultural area classification.

The deliberative potential of MCDA in spatial planning has received increasing attention, with critical reviews highlighting both opportunities and challenges. MCDA methods can support deliberative spatial planning by making criteria and trade-offs explicit, yet face challenges in transparency, stakeholder engagement, and method selection [6]. Sensitivity analyses comparing various methods (AHP, TOPSIS, VIKOR, and SAW) with different weighting schemes for land consolidation project ranking found relatively small variance and potential equal rankings, suggesting possible use of valuation instead of strict ranking to allocate funds and highlighting the importance of method selection and weighting in MCDA applications [13]. Addressing uncertainties in MCDA for sustainable systems planning requires techniques to make outputs more robust and decision-relevant in the face of data uncertainty [14]. These methodological considerations underscore the need for transparent, well-justified MCDA frameworks in strategic agricultural area classification.

This research develops and implements a comprehensive model for land sustainability, aiming to identify strategic agricultural areas at the municipal level and establish a replicable framework for protecting agricultural land within spatial planning. The model classifies agricultural land into three categories: permanently protected agricultural land (PPAL), representing the highest-quality land requiring strict protection; other agricultural land (OAL), which remains suitable for agriculture; and potentially suitable areas (PSA), where agriculture could expand or be restored. The methodology applies multiple criteria, including soil quality, agricultural infrastructure, topography, current land use, and cultural features of agriculture or the landscape. The approach utilises widely available datasets to facilitate broader application. The model was tested in Novo mesto Municipality, southeastern Slovenia, illustrating how these land categories can support regulatory enforcement and inform local planning.

2. Materials and Methods

2.1. Study Area

The Municipality of Novo Mesto is located in southeastern Slovenia and covers an area of 236 km² (Figure 1). With a population of approximately 37,587, it is the sixth-largest municipality in Slovenia. The municipality is organised into 23 local communities, which together comprise 98 settlements. This region exemplifies a typical peri-urban context in Slovenia, facing development pressures from urban expansion while still preserving significant agricultural land.

Novo Mesto’s agricultural landscape is marked by diverse topography, ranging from river valleys to hilly terrain, with elevations varying between 164 and 1178 m above sea level. The area’s geology is primarily karstic, with Sub-Pannonian landscapes [15]. The predominant soil types include eutric brown soils on clay and loam substrates, gleys and pseudogleys, colluvial brown carbonate soils on limestone, and anthropogenic soils associated with vineyard cultivation.

Land use patterns within the municipality reflect its peri-urban characteristics, with agricultural areas interspersed amidst urban settlements, forests, and infrastructure corridors. The Krka River valley is home to the municipality’s most productive agricultural soils, while the sloping terrains in the northern and southern sections present greater challenges for intensive agricultural practices. The region experiences a temperate continental climate, with mean annual temperatures around 10 °C and approximately 1200 mm of annual precipitation, creating generally favourable conditions for a variety of agricultural products, including cereals, vegetables, vineyards, and orchards.

The municipality’s strategic location along key transportation corridors connecting Ljubljana (Slovenia) to Croatia creates development pressures that increase the risk of agricultural land conversion. This dynamic is particularly evident in the southeastern extension of Slovenia’s highway network and its connection to the planned third development axis, both of which enhance the region’s attractiveness to the automotive and pharmaceutical industries. Additionally, the study area is traversed by the karstic River Krka, which, along with its numerous tributaries, serves as the region’s primary water source.

According to the 2020 agricultural census, there are 967 agricultural holdings in the municipality, collectively cultivating 4966 hectares of agricultural land, which constitutes 21% of the total area and approximately 1% of the agricultural land in use throughout Slovenia. Forested areas cover 4208 hectares, while arable land accounts for 1485 hectares, permanent grassland for 3198 hectares, and permanent plantations (including orchards and vineyards) for 283 hectares. On average, each agricultural holding encompasses 5.1 hectares of land. The predominant farming activities include dairy and cattle livestock operations, mixed crop-animal production, and specialised crop production [16]. Land use is often organised on sloping terrain, utilising strip terraces that follow natural contours, resulting in a small-parcel structure characterised by heterogeneous land use, a prominent feature of the landscape.

2.2. Databases

Spatial and tabular data used to identify potential strategic areas were sourced from various public databases (

Table 1. Input data used for spatial analysis and sources of the data.

Spatial Layer	Source	Spatial Accuracy
Digital elevation model [17]	Geodetic Survey of the Republic of Slovenia	12.5 m raster resolution
Land cadaster [18]		Parcel-level accuracy (≤1 m positional accuracy)
Land quality credit points [19]
State and Municipality borders [20]
Digital ortho-photo [21]		0.25 m pixel resolution
Actual land use [22]	Ministry of Agriculture, Forestry and Food	Parcel-level accuracy (≤1 m positional accuracy)
Land consolidation [23]
Drainage systems [24]
Irrigation systems [25]
Local characteristics of agricultural production and use of agricultural land—data on areas labelled “cultural landscape.” [26]	Ministry of Culture	1:5000 scale thematic polygons
Municipal Spatial Plan land use [27]	Municipality of Novo mesto	Parcel-level accuracy (≤1 m positional accuracy)
Local characteristics of agricultural production and land use—data on areas labelled “VIN” and “IKR.” [28]
Protected forests [29]	Forestry Institute of Slovenia

). The classification framework integrates multiple nationally available spatial datasets, ensuring consistent application across Slovenia’s municipalities. For the spatial analysis, we employed ESRI ArcGIS Pro 3.6.1 (Redlands, CA, USA) as our primary tool. Tabular data were merged with spatial layers in ArcMap, yielding a refined dataset that delineates potential strategic areas within the study municipality. These spatial layers have been classified according to a land suitability model developed in partnership with the Ministry responsible for agriculture. All spatial data were processed in a geographic information system (GIS) environment, utilising the national coordinate system (D96/TM) to maintain spatial consistency among all input layers.

2.3. Suitability Model Classification Framework

The authors newly developed the classification framework used in this study as part of a research project financed by the Ministry of Agriculture (contract number 2330-15-000110). The system was subsequently transposed into the Slovenian regulatory framework [30,31] for identifying strategic agricultural areas, thereby ensuring methodological consistency between research and policy implementation. Although informed by existing land evaluation practice, the specific combination of criteria and scoring rules originates from the project work. This model uses a comprehensive descriptive and numerical classification framework that categorises land based on a multicriteria evaluation of agricultural suitability and protection priority, thereby identifying strategic areas for agriculture and food production (

Table 2. Descriptive and numerical classification criteria of the land suitability model for determining strategic areas for agriculture and food production with a surface area (ha, %) for the Municipality of Novo mesto.

Land Suitability Model Classification Criteria		Area in the Case Study (Municipality of Novo Mesto)
Descriptive (Class)	Numerical (Point)	ha	%
Land quality credit points (LQCP)
≤35	1	16,244	68.90
36–50	3	4314	18.30
51–60	6	2267	9.61
61–100	8	751	3.19
Total		23,576	100.00
Slope (SL)
≤6	3	3717	15.77
7–11	2	4530	19.21
12–24	0	9335	39.60
≥25	0	5994	25.42
Total		23,576	100.00
Land consolidation (LC)
Yes	1	0	0
No	0	23,576	100.00
Drainage system (DR)
Yes	2	0	0
No	0	23,.576	100.00
Irrigation system (IR)
Yes	1	85	0.36
No	0	23,491	99.64
Total		23,576	100.00
Permanent plantations (PP)
Intensive	2	286	1.22
Extensive	1	392	1.66
No	0	22,898	97.12
Total		23,576	100.00
Local characteristics (LC)
Yes	1	2516	10.67
No	0	21,060	89.33
Total		23,576	100.00

Remark: lower value (1 point)—low suitability; higher value (8 points)—high suitability; 0—not present in the studied area.

). The criteria are represented as spatial areas (georeferenced vector data layer) within the national coordinate system, with attributes formatted in “shp” as topologically correct polygons. The attribute table of the vector data layer includes a unique identifier (ID), the area in square metres, and both descriptive and numerical criteria for the land suitability model used to determine individual strategic area levels, along with the aggregate point tally based on these criteria (Table 2).

This classification framework categorises agricultural land into three strategic categories derived from the multicriteria evaluation of agricultural suitability and protection priority:

(a) High-quality areas of permanently protected agricultural land (PPAL). Strategic areas, which are actual agricultural land and meet at least one criterion set in Table 2. They are primarily suitable for long-term protection from construction and make the greatest contribution to food security.

(b) Other agricultural land (OAL). Strategic areas that are actual agricultural land but do not meet the minimum criteria set in Table 2.

(c) Areas potentially suitable for agriculture (PSA). Areas of agricultural actual land use, but non-agricultural spatial plan land use that meets at least one criterion set in Table 2. Non-agricultural spatial plan land use potentially suitable for agriculture are: (i) designated as areas for urban, forest and other lands and are, according to their actual use, agricultural lands; (ii) defined as forest areas by the spatial plan and actual land use unless they are defined as protection forest or forest of special importance following the regulations governing forests.

This three-tier classification system is designed to comply with Slovenia’s national regulations on strategic agricultural areas and to provide a detailed framework for municipal spatial planning decisions. The methodology uses a hierarchical decision tree to systematically evaluate criteria and assign each land parcel to the appropriate category. This structured process ensures transparency and replicability in the assessment, effectively addressing the intricate, multi-dimensional aspects of agricultural land quality evaluation.

According to the land suitability model (Equation (1)), the minimum attainable score is 1, and the maximum is 20, indicating optimal land for protection. During this phase, we identify strategic areas across all land parcels, regardless of current use, based on seven essential eligibility criteria (Table 2) for primary strategic areas designated for the long-term, permanent protection of agricultural land (PPAL).

SM = LQCP + SL + CL + DS + IS + PP + LC

(1)

where is:

• SM—land suitability model (a sum of points);
• LQCP—land quality credit points;
• SL—slope inclination;
• CL—land consolidated;
• DS—drainage systems in function;
• IS—irrigation systems in function;
• PP—permanent plantations defined from actual land use (vineyard, intensive orchard, olive grove, extensive or meadow orchard, mother plants, and other permanent plantations);
• LC—land with specific local characteristics of agricultural production and use of agricultural land.

The classification framework assigns agricultural land to PPAL, OAL, or PSA based on the total score obtained from the land suitability model, which ranges from 1 to 20 points (Table 2). PPAL areas represent land with the highest suitability and protection priority and therefore fall within the upper part of the scoring range, meeting at least one key eligibility criterion. OAL includes agricultural land that remains suitable for production but does not reach the minimum threshold for PPAL. PSA consists of land currently designated for non-agricultural uses, but that achieves scores indicating potential agricultural suitability. Although the model uses a combination of quantitative indicators, a single comprehensive evaluation index is not used because the framework is designed for categorical classification rather than continuous ranking. However, the presented point system is essential because it enables transparent differentiation of land quality within each class. This finer distinction supports more optimal spatial planning and ensures that parcels with better biophysical or infrastructural characteristics are recognised, even when they fall within the same strategic class. The three-class approach follows national guidelines that prioritise transparency and regulatory applicability, ensuring that each criterion retains its explicit influence rather than being obscured in a weighted composite index.

The suitability evaluation in this study is conducted at the level of georeferenced cadastral polygons, which serve as the fundamental spatial evaluation units in the Slovenian land administration system. Each polygon corresponds to an officially registered land parcel and includes a unique identifier, precise geometric boundaries, and a complete set of attributes required for the land suitability model. These attributes include LQCP, slope, permanent plantations, irrigation systems, and locally specific agricultural characteristics. Using cadastral polygons ensures topological accuracy and allows the model to assign suitability scores to each parcel individually, reflecting real ownership and management structures. This parcel-based approach enables a transparent and spatially explicit classification of PPAL, OAL, and PSAs, supports integration with municipal spatial plans, and ensures consistency with national land-use regulations. By evaluating suitability at this fine spatial scale, the model provides a robust basis for strategic agricultural land protection and planning.

2.4. Suitability Model Evaluation Criteria

The scoring system in Table 2 reflects the relative contribution of each evaluation criterion to agricultural production potential and long-term land-protection priority. The scores represent ordinal suitability values, not statistical weights. Each indicator receives a score proportional to its agronomic relevance, spatial planning importance, and long-term contribution to food production.

Land Quality Credit Points (LQCP) form the central criterion in the model and therefore receive the widest scoring range (1–8 points). LQCPs are the official national measure of agricultural land productivity and are determined in accordance with the Rules on the Management of Real Estate Cadaster Data. They integrate soil properties, climate, topography, and limiting factors such as rockiness, flooding, drought, and shading. Because LQCPs are used in national compensation procedures for agricultural land conversion and represent the most robust indicator of long-term productive capacity, higher LQCP classes receive proportionally higher scores. The thresholds used in the model (≤35, 36–50, 51–60, 61–100) follow established national classifications and ensure consistency with Slovenian spatial planning and agricultural policy.

Slope receives differentiated scores (0–3) because it directly affects the feasibility of mechanisation, erosion risk, and production intensity. Land with slopes ≤ 11% is highly suitable for intensive agriculture, while steeper slopes significantly reduce agricultural potential.

Permanent plantations receive positive scores (1–2) because they represent substantial long-term investments and indicate areas with established, high-value agricultural use. Local characteristics of agricultural production and cultural landscape elements are assigned one point to recognise their importance for maintaining regionally specific agricultural systems, even where biophysical suitability is moderate. Indicators such as land consolidation, irrigation, and drainage systems receive positive scores (1–2) if present in the study area; however, their inclusion ensures methodological consistency with national guidelines and allows scoring to be applied where such infrastructure exists.

Overall, the scoring system reflects a hierarchical expert-based evaluation, assigning higher scores to criteria with the strongest influence on long-term agricultural productivity, land protection value, and policy relevance.

2.4.1. Land Quality Credit Points (LQCP)

The Rules on the Management of Real Estate Cadaster Data (2021) establish a framework for determining the credit rating of land, which reflects its productive capacity as quantified by credit points. This rating is derived from various factors, including soil properties, climate conditions, topography, and specific influences such as rockiness, flooding, drought, and shading. Credit points are assigned on a scale from 0 to 100, with 100 indicating the highest level of productivity (Figure 2).

Management of this data is conducted by the Geodetic Survey of the Republic of Slovenia and is integrated into the Digital Cadaster Plan (DKP). These data are crucial for calculating compensation payments for land-use changes, particularly for converting agricultural land to other uses.

Under the Agricultural Land Act, a credit score of 35 or higher is required to qualify for compensation when agricultural land is repurposed for development. In the Municipality of Novo Mesto, the average number of credit points is 44.5.

Land is categorised into four LQCP classes based on credit points (see Table 2). The classification begins at 35 credit points, denoting the first class. The second class encompasses land with a credit rating of 36 to 50, which typically provides optimal growing conditions for conventional intensive agriculture. Classes three and four include land scoring 51–60 and 61–100, respectively. Notably, the highest recorded credit point value within the Municipality of Novo Mesto is 90.

2.4.2. Slope

The slope is already integrated in the land quality credit points. However, the assessment lacks sufficient differentiation among levels of agricultural land quality, prompting greater emphasis on the slope’s role in the analysis (Figure 2). Utilising a 12.5 m resolution digital elevation model (DEM), we developed a spatial layer representing slope characteristics. The capacity for mechanised land cultivation is a critical factor in identifying strategic agricultural areas.

Land was classified according to slope into four categories: 0–6%, 7–11%, 12–24%, and ≥25%. Optimal agricultural land has a slope of up to 11%, allowing unrestricted use of agricultural machinery. On slopes ranging from 12% to 17%, the use of machinery is limited, with mechanical mowing becoming partially restricted. For slopes between 18% and 24%, the deployment of heavy agricultural machinery is constrained, while land with slopes exceeding 25% is only adaptable for manual mowing; the use of standard tractors is no longer feasible. Notably, over 65% of the land in the study area has slopes greater than 12%, indicating that the use of agricultural machinery in these regions is either suboptimal or impractical. The analysis identifies potential strategic areas for agriculture and food production, encompassing 35% of the land with slopes of up to 11% (Table 2).

2.4.3. Land Consolidation

Land consolidation is a systematic process aimed at reorganising and redistributing land within a designated area to optimise economic productivity and efficiency for landowners. This process not only focuses on the equitable allocation of agricultural land but may also encompass forests, undeveloped construction sites, and associated facilities. Additionally, it ensures the establishment of essential access routes and infrastructure within the agricultural landscape.

In our proposal for strategic development areas, we have identified zones suitable for land consolidation. However, in the Municipality of Novo mesto, according to the available data, land consolidation has not yet been carried out or recorded in national databases.

2.4.4. Drainage Systems

Hydromelioration systems play a critical role in enhancing agricultural productivity by systematically regulating soil moisture levels and can be categorised into two primary types: drainage and irrigation systems. Most existing drainage systems were established in the 1980s. As of now, Slovenia has 476 registered drainage systems. According to 2017 data, 252 of these systems are under regular maintenance, effectively serving 37,279 hectares of agricultural land. However, no drainage systems are present in the study area.

2.4.5. Irrigation Systems

Irrigation systems in Slovenia are classified as large or small and were primarily established during the 1980s and 1990s. As of 2025, a total of 8.364 hectares of land in Slovenia were equipped with irrigation systems. Large-scale irrigation and drainage systems are designed for multiple users and shared use, with management and maintenance typically overseen by the Agricultural Land and Forest Fund of the Republic of Slovenia, municipalities, or farmers’ irrigation associations. The implementation of drainage or large irrigation systems necessitates the consent of landowners representing over 67% of the agricultural land within the proposed irrigation area. Current statistical data indicate that only 1.2% of agricultural land in Slovenia is irrigated.

In our proposal for strategic areas, we have included lands equipped with operational irrigation systems. The Ministry responsible for agriculture prepares a comprehensive map delineating large and small irrigation systems. Within the area under consideration, an extensive irrigation system covers 85 hectares, constituting 0.36% of the total study area (Table 2, Figure 2).

2.4.6. Permanent Plantations

A permanent plantation refers to agricultural land utilised for crops that are not part of a crop rotation, and excludes permanent grasslands that maintain the same crop for a minimum duration of five years and yield multiple outputs. This category also encompasses nurseries and fast-growing stumps, specifically willow and poplar species. A permanent plantation is formally acknowledged within the agricultural holdings register, incorporating various classifications of actual land use, such as hop fields, permanent crops on arable land, greenhouses with fruit plants (excluding strawberries), vineyards, mother plants, intensive orchards, extensive orchards, and olive groves, among other permanent plantations. The actual land use is represented spatially and reflects current utilisation patterns. Alterations to this status are permissible only through official documentation and following a substantive change in land conditions. The Republic of Slovenia maintains records of actual land use in graphical representation, derived from orthophotos and alternative sources. It is crucial to note that actual land use does not dictate the intended land use as outlined in the Municipal Spatial Plan. In the process of classifying permanent plantations, we employed a structured classification system: (a) intensive permanent plantations (vineyards, mother plants, intensive orchards, olive groves) and (b) extensive permanent plantations (extensive or meadow orchards, other permanent plantations) (Table 2, Figure 2). Extensive permanent plantations comprise less than 400 ha (1.66%), while intensive permanent plantations account for 286 ha (1.22%) of the study area (Figure 1, Table 2).

2.4.7. Local Characteristics of Agricultural Production and Use of Agricultural Land

The proposal for strategic areas encompasses suboptimal lands that are significant due to local characteristics of agricultural production and land management, which can be delineated within each municipality. This layer also includes agricultural landscapes of exceptional aesthetic value at both national and local scales, alongside lower-quality agricultural land in municipalities with limited access to high-quality land resources.

Following consultations with municipal representatives, supplementary areas from the Municipal Spatial Plan were incorporated into this layer: (a) “VIN,” representing designated wine-growing regions, and (b) “IKR,” which characterises areas of exceptional scenic qualities. In the wine-growing regions, a diverse spectrum of land uses converges, primarily encompassing agricultural, forest, and urban uses, with agricultural lands primarily cultivated with vines, fruit-bearing trees, or grasses.

Following discussions with the Ministry of Culture, areas identified as “cultural landscapes” under the Act on the Protection of Cultural Heritage were integrated into the local characteristics layer. These landscapes are classified as immovable heritage, comprising open spaces with both natural and anthropogenic components whose structure, development, and use are predominantly influenced by human activities (Figure 1, Table 2). However, upon integrating the areas denoted as “IKR,” “VIN,” and “cultural landscape” into a consolidated layer, it was observed that some areas overlap.

2.5. Evaluation Scenarios

To evaluate the influence of each criterion on the delineation of strategic areas for agricultural and food production, we developed multiple scenarios (

Table 3. Description of scenarios and classification criteria between PPAL, OAL and PSA.

Scenario	Description	Strategic Area Type
A	Meets at least one criterion according to the suitability model, land suitability model (SM) sum of points is >1	PPAL
	Does not meet any criterion, land suitability model (SM) sum of points is ≤1	OAL
B	Scenario A, without LQCP ≤ 35	PPAL
	Scenario A, with LQCP ≤ 35	OAL
C	Scenario A, without slope ≥ 12%	PPAL
	Scenario A, with a slope ≥ 12%	OAL
D	Scenario A, without LQCP ≤ 35 and slope ≥ 12%	PPAL
	Scenario A, with LQCP ≤ 35 and slope ≥ 12%	OAL
E	Scenario A, without LQCP ≤ 35, slope ≥ 12% and VIN	PPAL
	Scenario A, with LQCP ≤ 35, slope ≥ 12% and VIN	OAL
F	The suitability model calculation is >1 point, but for lands that, according to their use in the spatial plan, are designated as buildings, forests and other land areas, but according to their actual use, they are agricultural land and land that is a forest according to its use in the spatial plan and actual land use, unless it is defined as a protected forest or a forest of special importance	PSA
G	Scenario F, without BT ≤ 35 and slope ≥ 12%	PSA
H	Scenario G, only actual agricultural land use	PSA

Remark: (≤) less than or equal; (≥) more or equal; PPAL—high-quality areas of permanently protected agricultural land (PPAL); OAL—other agricultural land; PSA—areas potentially suitable for agriculture; LQCP—land quality credit points; VIN—vineyard areas.

).

The eight scenarios in Table 3 represent all possible combinations of the key eligibility criteria used to determine whether a land parcel qualifies as permanently protected agricultural land (PPAL). Each scenario reflects a different configuration of the seven indicators in the suitability model (LQCP, slope, irrigation, drainage, permanent plantations, land consolidation, and local characteristics). Because parcels may meet one, several, or none of these criteria, the model generates eight distinct logical scenarios that describe how a parcel satisfies the PPAL requirements. Scenarios A–E correspond to PPAL eligible combinations, meaning that parcels falling into these scenarios meet at least one of the legally defined PPAL criteria. These scenarios, therefore, represent different “pathways” through which land can qualify for the highest protection category. In contrast, scenarios that do not meet any PPAL criterion are classified as OAL or PSA, depending on their actual and planned land use. The scenario structure ensures transparency by showing exactly why a parcel is assigned to PPAL and how different criteria contribute to its protection priority.

In scenario A, we identified areas that satisfied at least one criterion and had a land suitability model score > 1 point, which were designated as PPAL. In scenario B, we defined all areas with an LQCP of ≤35 as OAL. In scenario C, we defined all areas with a slope ≤ 11% as PPAL and all areas with a slope > 11% as OAL. In scenario D, we defined PPAL areas where LQCP > 35 and the slope ≤ 11%. Lastly, in scenario E, we excluded vineyards (designated “VIN”) from the PPAL classification because they are few in number, dispersed across the study area, and have limited economic potential or agricultural viability.

The outcomes of the PSAs are delineated in scenario F. To ascertain PSA, we formulated an additional scenario, G, wherein, under the parameters of scenario F, we identified the extent of PSA on land with an LQCP ≥ 35 and a slope ≤ 11%. This approach yields insights into the distribution of flat terrain characterised by relatively high land quality. Ultimately, building on scenario G, scenario H examines only actual agricultural land use. The conceptual definition of PSA requires distinguishing between theoretical land suitability and the practical feasibility of land conversion. Theoretical suitability refers to the biophysical potential of soil, climate, and topographic factors, representing the inherent capacity of land to support sustainable agriculture [32,33]. In contrast, practical feasibility incorporates socio-economic constraints, such as accessibility, infrastructure, market conditions, and policy factors, that influence actual land conversion and use [32]. This conceptual separation aligns with strategic farmland designation studies that validate multicriteria frameworks by comparing theoretical suitability assessments with observed agricultural patterns [33]. In this study, PSA represents areas that are biophysically suitable for agriculture but currently designated for non-agricultural uses, reflecting a potential, rather than immediate capacity for agricultural restoration. This distinction aligns with strategic farmland designation literature, which differentiates between inherent land capability and planning-based conversion constraints [34,35]. By integrating both suitability and regulatory feasibility, the PSA category captures land with strategic long-term value while acknowledging that conversion depends on planning decisions, legal frameworks, and local development priorities.

3. Results and Discussion

3.1. Suitability of Land for Strategic Areas for Agriculture and Food Production

In the designated study area, a total of 13,255 hectares satisfy at least one criterion outlined in the suitability model, achieving a cumulative points score exceeding 1, regardless of land-use classification. It is important to note that these identified areas are not exclusively agricultural, as evidenced by their land-use designations and the municipal spatial planning framework. The observed points range from 1 to 14 (

Table 4. Areas (ha, %) based on the points achieved as determined by the suitability model, which corresponds to the criteria for identifying PPAL regardless of land use type.

Suitability Model Sum of Points	Study Area
	ha	%
1	10,321	43.78
2	666	2.82
3	4909	20.82
4	3005	12.75
5	824	3.50
6	1410	5.98
7	377	1.60
8	648	2.75
9	748	3.17
10	262	1.11
11	328	1.39
12	46	0.20
13	11	0.05
14	20	0.08
Total	23,576	100.00

). Following the established methodology, low-quality land, which scores only 1 point, is categorised as OAL. In contrast, prime-quality land, which can be classified as a PPAL, is characterised by a minimum score of 2 points, with at least one of these points specifically attributed to the LQCP. Notably, land with a score of 1 point constitute 44% of the total land area within the study area (Table 4).

The findings indicate that, according to the established model, the land with the highest suitability scores is predominantly located near the Krka River and its tributaries, characterised by gentle slopes or flat terrain and high soil quality. Conversely, in the Gorjanci hills, particularly in the southern and southeastern sections of the study area, and in forested regions, most land parcels receive only 1 point in the suitability model. This low score is primarily due to the steep topography at these locations (Figure 3). In contrast, areas designated with two points are found in the distinctive landscape of vine-growing regions.

3.2. Permanently Protected (PPAL) and Other (OAL) Agricultural Land

In assessing the total study area with respect to the actual and municipal spatial plans and agricultural land use, the maximum PPAL area was identified in scenario A. As subsequent scenarios are analysed, the criteria implemented increasingly tighten the entry conditions, resulting in a decrease in the area designated as suitable for PPAL, while simultaneously, the area classified as OAL experiences a proportional increase. The findings for all scenarios encompassing the entire study area are presented in

Table 5. Areas (ha) according to the scenarios for the entire study area and separately for actual land use and spatial plan land use according to the municipal spatial plan of Novo mesto.

Scenario	Strategic Area Type	Area (ha)
		Total Study Area	Actual Land Use (Agriculture)	Spatial Plan Land Use (Agriculture)
A	PPAL	13,255	6921	6490
	OAL	10,321	923	1020
B	PPAL	7332	6024	5628
	OAL	16,244	1820	1882
C	PPAL	8247	3073	4750
	OAL	15,329	4771	2761
D	PPAL	3059	2535	2265
	OAL	20,517	5309	5246
E	PPAL	2991	2478	2208
	OAL	20,585	5366	5303

Remark: PPAL—high-quality areas of permanently protected agricultural land (PPAL); OAL—other agricultural land.

.

According to the actual land-use assessments, agricultural land has approximately 6921 hectares of potential PPAL, whereas the spatial plan indicates around 400 hectares less (6490 hectares) (Scenario A). In circumstances of inadequate food self-sufficiency, state intervention may involve converting non-agricultural land (e.g., forested areas) that, according to the suitability model criteria, may qualify for designation as PSA. However, such actions would require substantial financial and labour resources, as well as unpredictable environmental repercussions.

Evaluating all criteria for PPAL determination requires excluding areas with an LQCP below 36; the greatest PPAL variance is observed across the entire study area (Scenario B). The discrepancy in potential PPAL between actual and spatial plan land use is 500 hectares, suggesting that the classification of PPAL for agricultural use is predominantly influenced by land characterised by high production capacity (LQCP above 36).

Compared with Scenario A, Scenario C shows that, based on actual land use, over half of the PPAL is on terrain with slopes exceeding 11%, which limits the use of agricultural machinery and results in higher production costs than on flatlands. Approximately 1700 hectares (actual land use) and 2000 hectares (spatial plan) of agricultural land in areas with adverse conditions, such as slopes above 11%, may be classified as OAL.

Scenario D underscores the significance of land quality credit points and slope criteria in establishing PPAL. For instance, when lands with LQCP below 36 and slopes exceeding 11% are classified as OAL, only 2500 hectares (10.75%) of the study area are designated as PPAL.

The study area is predominantly characterised by small family-owned vineyards, prompting a particular focus on the potential for their permanent protection. Consequently, vineyard areas (“VIN”) were intentionally excluded from the properties analysed in scenario D for scenario E. The disparity between scenarios D and E is marginally less than 100 hectares. This indicates that vineyard areas are primarily located on slopes exceeding 11% or in low-land regions with a quality rating below 36 points. A significant number of vineyard areas were already excluded from the PPAL in scenario D. Previous research on agricultural land protection [36] has illustrated that wine-growing zones occupy a substantial portion of land with high LQCP, a finding corroborated by our analysis. The current land use within the study area comprises 268 hectares of vineyards, of which 236 hectares exhibit an LQCP exceeding 35 and 246 hectares are situated on slopes exceeding 11%. The pronounced slope enhances wine quality; however, it also complicates mechanical vineyard management, potentially leading to increased abandonment of viticulture in the area over time.

Therefore, excluding wine-growing regions from the PPAL within the study area may be justified. Nonetheless, the vineyards in question are extensively managed and more fragmented than those in other wine-growing regions in Slovenia, with a notable presence of land development, such as family wine cellars, holiday, and residential buildings. To adequately substantiate the inclusion of selected wine-growing areas in the PPAL, further comprehensive professional efforts are required, entailing field assessments and meticulous processing of graphic data [36].

3.3. Potentially Suitable Agricultural Land (PSA)

The study area encompasses approximately 5000 hectares of PSA. The majority of this PSA is located on forest land, with a small portion distributed across other land-use categories (

Table 6. Area (ha) of potential land suitable for permanent protection (PSA) according to scenarios F, G and H for the case study area of the Municipality of Novo mesto.

Scenario	Strategic Area Type	Area (ha)—Municipal Spatial Plan Land Use
		Urban Land	Forest Land	Other Lands	Total
Scenario F	PSA	730	4347	0.69	5078
Scenario G	PSA	337	190	0.35	527
Scenario H *	PSA	229	129	0.24	358

Remark: PSA—areas potentially suitable for agriculture; * Scenario G + only actual agricultural land use.

). Areas of future urban development (residential and industrial) are potential candidates for permanent protection. However, these zones exhibit low LQCP, reflecting limited productive capacity despite their current classification as agricultural land (Table 6). The inclusion of forest areas within PSAs requires careful interpretation, particularly regarding the environmental trade-offs associated with potential forest-to-agriculture conversion. Forest ecosystems provide essential services such as carbon sequestration, biodiversity support, soil protection, and climate regulation, all of which may be diminished if forests are converted. The ecosystem services literature highlights that land-use change from forests to agriculture can generate substantial ecological costs, even when agricultural benefits are present [37]. Studies further emphasise that carbon-dense temperate forests play a critical role in climate mitigation strategies, underscoring the importance of avoiding unnecessary forest loss [38].

Additionally, increasing forest disturbances across Europe have been shown to affect carbon storage and long-term ecosystem resilience, reinforcing the need for careful consideration of any land use transitions [39]. In the context of PSA, forested areas are therefore not interpreted as immediate candidates for agricultural expansion but as locations where biophysical conditions indicate potential suitability under hypothetical long-term scenarios. This approach aligns with broader ecosystem services research, which emphasises the importance of evaluating trade-offs and synergies between agricultural production and environmental conservation [40]. Consequently, PSA serves as a strategic planning tool rather than a prescriptive directive for forest conversion.

In alignment with Scenario F criteria, PSAs include forests, excluding protective forests and forests of particular importance. The analysis reveals 4199 hectares of such forested land. These lands are unsuitable for protection, as transitioning them from forest to productive agricultural use would require substantial financial and human resources and entail substantial environmental impacts.

The findings from Scenario G, which exclude low-quality lands (<36 points) and steep slopes (>11%), suggest that only a marginal portion (526 hectares) of potential productive agricultural land (PPAL) in the study area is situated on land with high production capacity and favourable conditions for the mechanical management of farmland.

Lastly, the Scenario H proposition for PSA includes land classified as urban, forest, or other non-agricultural land, according to the spatial plan’s land-use classification. Nonetheless, upon assessing actual land use, the classification as agricultural land with an LQCP exceeding 35 and a slope below 11% applies to only 358 hectares of the study area.

3.4. Suitability of Land for PPAL, OAL and PSA

The delineation of the proposed PPAL, OAL, and PSAs was conducted on agricultural terrain, adhering to the land-use criteria outlined in the spatial plan while satisfying at least one criterion from the established suitability model. The land suitability evaluation reveals that 6490 hectares (28%) of the land within the Municipality of Novo Mesto could potentially be classified as PPAL, with 1020 hectares (4%) identified as OAL. Regarding the suitability model’s criteria for identifying lands eligible for PSA, a total of 5078 hectares are located within the study area. Consequently, in the event of a food crisis, over half of the study zone’s total area (12,588 hectares) could be allocated to enhance agricultural productivity and self-sufficiency (

Table 7. Area (ha) of PPAL, OAL and PSAs determined according to the suitability model for the case study area of the Municipality of Novo mesto.

	Strategic Area Type by a Suitability Model			Total
	PPAL	OAL	PSA
Area (ha)	6490	1020	5078	12,588
Share of the case study total area (%)	27.53	4.33	21.54	53.39

Remark: PPAL—high-quality areas of permanently protected agricultural land (PPAL); OAL—other agricultural land; PSA—areas potentially suitable for agriculture.

, Figure 4). However, this assertion remains largely theoretical because land-use changes require significant financial investment. The more accurate estimate of PSA is 385 hectares, based solely on the suitability model’s specific criteria and the detailed stipulations outlined in scenario H. Food security research consistently shows that land availability alone does not translate into effective production capacity, as conversion feasibility is constrained by ecological limits, labour availability, infrastructure, and governance conditions [41]. Moreover, land system studies emphasise that theoretical maximums often overestimate realistic potential because competing land uses, environmental safeguards, and socio-economic trade-offs restrict rapid or large-scale agricultural expansion [35]. In this context, the identified PSA areas should be interpreted as an upper-bound estimate of biophysical suitability rather than an actionable reserve. Their practical contribution to crisis response strategies would depend on policy decisions, environmental assessments, and the capacity to mobilise resources under emergency conditions.

Comparability in the scientific literature regarding the permanent protection of agricultural land remains limited. A previous study conducted in Slovenia evaluated alternative methodological criteria for determining PPAL quotas in each municipality [36]. In that study, the calculated PPAL quota for the Novo mesto area in 2013 was 4472 hectares, which is 2018 hectares lower than the quota identified in the current analysis (a 60.47% difference). This discrepancy can be attributed to two main factors. First, the previous study considered only agricultural land with LQCP > 40, while the current analysis includes land with an LQCP > 35. Second, the earlier proposal for the PPAL quota did not account for slope and instead considered only water sources suitable for irrigation [36].

The land suitability analysis indicates that the proposed PPAL covers smaller areas than those currently designated as agricultural land in the Municipality of Novo mesto’s spatial plan. This suggests that, using the proposed methodology, not all agricultural lands identified in existing municipal spatial plans will receive permanent protection. According to the model criteria, PPAL areas are anticipated to be more consolidated, while the proportion of OAL is expected to decrease. Each PPAL area is designed to form a contiguous unit, potentially including smaller plots with an LQCP below 36, slopes exceeding 11%, agricultural field paths, or forest plots classified as agricultural in municipal spatial plans. Automated and manual polygon merging can address empty spaces within existing polygons, potentially integrating diverse land uses of varying quality under permanent protection. The spatial aggregation approach used to merge parcels into continuous PPAL units benefits from contextualisation within spatial-planning and protected-area delineation research. Studies in landscape and conservation planning show that how spatial units are aggregated or bounded can significantly influence the feasibility of governance, administrative clarity, and long-term management outcomes [42]. Research on protected-area design further demonstrates that delineation choices affect policy implementation, enforcement efficiency, and stakeholder acceptance, underscoring the need to balance ecological coherence with institutional and cadastral realities [43]. In this context, the aggregated PPAL units should be interpreted as analytically coherent planning zones rather than predefined administrative entities. Their practical application would require subsequent alignment with existing property boundaries, municipal planning frameworks, and regulatory procedures. This distinction helps clarify that the aggregation method supports strategic spatial reasoning while recognising the administrative constraints that shape real-world land-use decisions.

The methodology detailed in this study has been adopted by the Ministry of the Republic of Slovenia for Agriculture, Forestry, and Food as the “Rules on Detailed Conditions for Determining Permanently Protected Agricultural Land.” This approach standardises the spatial planning of agricultural land at the national level, addressing decades of agricultural land loss under a previously decentralised system in which municipalities held primary authority. The new semi-centralised system has introduced two key steps: first, the State conducts baseline spatial analyses of agricultural land for each municipality and provides proposals for PPAL designation; second, municipalities organise public discussions and submit final PPAL proposals, which must obtain Ministry approval. This process ensures shared decision-making and mutual oversight between state and local authorities, helping address challenges in balancing spatial planning across governance levels, as noted internationally [44].

3.5. Assessment of Strategic Implications for Agricultural Land Protection and Spatial Planning

The results of this study demonstrate the potential of transparent, multicriteria evaluation frameworks to support the designation of strategic agricultural areas in municipal spatial planning. Designating 6.490 hectares as PPAL provides a clear spatial foundation for protection policies, but the effectiveness of such designations depends critically on strict regulatory implementation and enforcement. These results align with international evidence demonstrating that explicit spatial designation of strategic agricultural areas, when combined with strong implementation mechanisms, can substantially reduce agricultural land conversion. Scenario modelling in Switzerland found that the revised spatial planning act could reduce the potential consumption of prime cropland for new urban development areas by a factor of six relative to extrapolated past trends. However, strict implementation remains essential to meet protection targets [3]. This Swiss experience demonstrates that clear designation of protected agricultural areas, combined with effective implementation through cantonal and municipal planning processes, can counter development pressures even in regions experiencing significant urbanisation. However, Mediterranean coastal regions show contrasting patterns, with intense urbanisation within 1000 m of the coast and land consumption continuing to grow even where populations stabilise or decline, underscoring the need for proactive agricultural land protection mechanisms in high-pressure contexts [4]. The Novo mesto classification provides a foundation for such protection in Slovenia’s peri-urban regions. However, effectiveness will depend critically on integration into binding spatial planning regulations and sustained commitment to implementation.

A comparative methodological analysis positions the Novo mesto framework within international strategic-area designation practice while validating its distinct contributions. Comparative analysis of agricultural protection frameworks across different national contexts reveals common principles: hierarchical classification systems that distinguish protection stringency levels; integration of biophysical suitability criteria with spatial planning considerations; emphasis on landscape contiguity and functionality; and transparent decision-making processes that facilitate stakeholder engagement and regulatory implementation [11]. The translation of these evaluations into binding regulations necessitates explicit integration into statutory spatial planning processes, as demonstrated by European case studies in which multiple evaluation criteria inform municipal plans and establish legally enforceable protection zones [9]. However, critical differentiations emerge: whereas China employs continuous scoring that requires subsequent threshold determination, Changzhou City integrates complex functional supply-demand assessments, and Novo mesto utilises hierarchical decision-tree logic with predetermined categorical boundaries derived from readily available datasets. This addresses fundamental tensions identified in MCDA sensitivity analyses [13], which show that continuous scoring introduces methodological variance across weighting schemes. The explicit categorical boundaries enhance transparency and reduce parametric sensitivity, making them critical advantages for regulatory applications that require classifications that withstand legal scrutiny [14]. This methodological parsimony enhances replicability in post-transition contexts and smaller municipalities with limited analytical capacity.

Jiangsu Province’s stakeholder-driven spatial targeting [12] demonstrates the potential of participatory MCDA but requires sustained engagement infrastructure. The Novo mesto framework complements this by providing objective classification that establishes a credible foundation for subsequent stakeholder deliberation, separating technical assessment from value-based negotiation. This addresses critiques of MCDA’s deliberative potential [6] by providing transparent, data-driven classification as a starting point for participatory refinement rather than replacing stakeholder processes entirely.

Turkey’s Isparta Province provides cautionary validation [45]: multicriteria evaluation identified suitable agricultural land only after the highest-quality areas had been converted to artificial surfaces, quantifying measurable urbanisation losses. This retrospective limitation reinforces the present study’s proactive temporal positioning, with designation preceding rather than following urbanisation pressures, and is integrated directly with Slovenia’s regulatory framework to ensure classification outcomes inform binding land-use decisions. This temporal positioning represents a critical methodological contribution, as technically sound MCDA applications may arrive too late to inform protection if not integrated into forward-looking spatial planning instruments [46].

Advanced fuzzy MCDA approaches offer greater uncertainty accommodation [6], yet sensitivity analyses demonstrate relatively small variance across methods with different weighting schemes [13], suggesting methodological sophistication may yield diminishing returns in certain contexts. The Novo mesto framework deliberately prioritises transparency and accessibility over mathematical complexity, producing clear categorical boundaries as informed approximations that address the tension between continuous environmental variation and regulatory requirements for definitive spatial boundaries. While landscape-scale simulations and mechanistic models offer potential for integrating ecological considerations and dynamic modelling [47], such approaches require analytical capacity and data infrastructure that may not be universally available.

The methodological contribution lies not in isolated technical innovation but in a systematic synthesis that addresses five critical gaps: transparency (vs. opaque weighting), accessibility (vs. data-intensive assessments), timing (vs. retrospective documentation), integration (vs. academic disconnection), and replicability (vs. context-specific requirements). This demonstrates that effective strategic agricultural area designation requires not only technical sophistication but also careful attention to transparency, accessibility, and regulatory alignment, which determine whether classification outcomes translate into implemented protection measures.

The effectiveness of agricultural land protection extends beyond technical classification to encompass governance, stakeholder engagement, and integration with other land use objectives. The protection of strategically important agricultural areas in peri-urban contexts requires methodological frameworks that integrate multiple evaluation dimensions and connect directly to spatial planning instruments. International experiences demonstrate that effective approaches combine ecosystem services assessment with systematic multicriteria evaluation to identify agricultural landscapes warranting protection based on their multifunctional values [8]. These peri-urban frameworks emphasise the need for differentiated protection strategies that recognise the particular pressures and opportunities in areas near urban centres. Agricultural land also faces competing uses from renewable energy development, environmental protection, and other objectives. Analysis of land-use conflicts and synergies in Brandenburg, Germany, identified renewable energy and environmental protection as key drivers of conflict, while highlighting agroforestry, agri-photovoltaics, and integrated non-productive areas as opportunities for synergy [48]. A spatial multicriteria analysis of land in Italy found 10.7 million hectares eligible for agrivoltaics, with a potential capacity of 6435 GW. Covering only 1.24% of the eligible area would meet an 80 GW target, indicating a large eligible area but relatively low land-occupation needs if strategically sited [49]. Swiss focus zones for bird conservation identified areas with low management impact and high bird potential (covering approximately 18% of farmland) and high-impact, high-potential zones (covering approximately 31%), providing spatial prioritisation for reconciling agricultural production with biodiversity objectives [50]. These competing use analyses demonstrate that strategic agricultural area classification must increasingly address multiple objectives beyond food production, including renewable energy, biodiversity conservation, and climate adaptation. The Novo mesto classification provides a foundation for such multi-objective planning by clearly identifying areas where agricultural use should be prioritised, enabling more informed decisions about where competing uses might be accommodated.

Classifying agricultural lands with the land sustainability model supports better agricultural policy, especially for the EU’s Common Agricultural Policy (CAP). Studies show MCDA helps balance environmental and economic goals under CAP rules. In the context of the Greek CAP 2023–2027, an MCDA model reduced the area allocated to high-water-demand crops and improved gross margins for certain producers, indicating that MCDA approaches can effectively guide land allocation decisions to enhance both sustainability and profitability [51]. This approach could be adapted to the Novo mesto Municipality by using land classification to target CAP agri-environment measures to appropriate land categories. For example, directing intensive production support to PPALs and extensive management and biodiversity measures to OALs. The differentiated classification also supports prioritisation of land consolidation and rural development investments. A MCDA and clustering approach used to prioritise land consolidation projects in Slovakia revealed subjectivity in previous project selections and introduced an adaptable ranking system for prioritisation [52]. The Novo mesto land sustainability model could similarly guide land consolidation efforts by directing investments to high-quality lands where consolidation would improve productivity and support the municipality’s key agricultural areas. This targeted approach would maximise returns from rural development and strengthen protection of strategic, high-quality agricultural land.

3.6. Model Robustness and Limitations

Several limitations of this study warrant acknowledgement. First, the classification relies on static datasets that may not fully capture the temporal dynamics of soil quality, climate conditions, or land-use patterns. Agricultural land quality can change over time due to management practices, erosion, contamination, or climate trends, suggesting the need for periodic updates to classification. Data quality and resolution vary across input datasets, introducing uncertainty in classification boundaries, particularly in transitional areas between categories.

Second, the simple additive structure employed in Equation 1 warrants critical examination through the lens of contemporary multicriteria decision analysis (MCDA) literature, particularly regarding compensatory effects. Additive aggregation models, while computationally straightforward and widely applied, inherently allow full compensation, in which poor performance in one criterion can be entirely offset by strong performance in another [53,54]. This compensatory property has been increasingly scrutinised in recent MCDA research, with scholars questioning whether unlimited trade-offs between criteria accurately reflect real-world decision contexts [54,55,56]. Nonetheless, we acknowledge that non-compensatory or partially compensatory methods (e.g., outranking, threshold-based, or hierarchical approaches) could reduce compensation effects. Recent methodological advances have highlighted the need to control or limit compensatory effects in additive models. Arman et al. [54] demonstrated that modifying the simple additive weighting (SAW) approach to restrict trade-offs between attributes can yield more realistic decision outcomes, while Kheybari [56] proposed frameworks for adjusting trade-offs by defining upper and lower expectation levels for each criterion. Moreover, Schär et al. [53] developed flow insensitivity intervals to explicitly assess and control compensation in MCDA methods, emphasising the importance of transparency in managing compensatory behaviour. The robustness of additive models has been validated through sensitivity analyses in applied contexts [57], though Kujawski et al. [58] cautioned that in life-critical decisions, compensatory effects may yield misleading recommendations. Alternative approaches, such as non-compensatory methods [59] and veto-based mechanisms [60], have emerged to address scenarios where full compensation is inappropriate. For the present application, the additive structure’s validity is supported by the balanced nature of the evaluation criteria and the absence of threshold effects requiring non-compensatory treatment. Nevertheless, sensitivity analyses examining weight variations and alternative aggregation rules would further strengthen confidence in the model’s robustness [55,57].

Third, the threshold values employed in this study (LQCP 35 and slope 11%) warrant validation through comparison with established land evaluation literature and sensitivity analysis to ensure classification robustness. Recent MCDA-based land suitability studies emphasise the critical importance of threshold validation and sensitivity testing [61,62]. Chelariu et al. [61] demonstrated that defining class thresholds for each factor and validating results through sensitivity analysis is essential for robust land suitability estimation. Similarly, Salunkhe et al. [62] identified slope and topography-related thresholds as among the most sensitive parameters in suitability classification, testing 26 weighting schemes to quantify how threshold changes affect class boundaries. The slope threshold of 11% aligns with established agricultural land evaluation frameworks, in which slopes exceeding that value typically indicate increased erosion risk and reduced suitability for intensive cultivation [62,63]. The LQCP value of 35 falls within the moderate suitability range commonly adopted in multicriteria land assessment [63,64]. Sensitivity analyses in comparable studies have shown that threshold variations of ±10–15% produce relatively stable classification outcomes when properly weighted [62,64], suggesting our adopted values lie within acceptable ranges. Nevertheless, systematic sensitivity testing across a range of threshold values (e.g., LQCP 30–40, slope 8–14%) would strengthen confidence in the classification robustness and facilitate comparison with regional land evaluation standards.

Fourth, the framework emphasises biophysical factors (soil quality, topography, infrastructure) while giving limited consideration to economic factors, such as farm profitability, market access, and production costs, that significantly influence actual agricultural land-use decisions. This is a deliberate methodological choice with important implications for interpretation and policy applicability. The agricultural economics literature demonstrates that farm profitability, market access, and production costs significantly influence land-use decisions and farmer behaviour [65,66]. This exclusion means the classification identifies lands with high productive capacity but cannot distinguish whether cultivation is economically viable under current market conditions. Consequently, PPAL designation based solely on biophysical criteria may protect lands that farmers find unprofitable to cultivate, potentially creating implementation resistance or requiring compensatory mechanisms. Economic land evaluation frameworks that integrate profitability analysis with biophysical assessment provide a more comprehensive sustainability evaluation [67]. For policy applicability, this limitation suggests that PPAL implementation should be complemented by economic support instruments, such as direct payments, market access improvements, or agri-environmental schemes, to ensure designated lands remain economically viable for farmers, thereby enhancing voluntary compliance and long-term protection effectiveness.

Fifth, stakeholder input was limited to expert review rather than broad participatory engagement with farmers and landowners whose local knowledge could enhance classification accuracy and acceptance. An assessment of the proposed framework’s stakeholder engagement using the participatory planning literature reveals both strengths and areas for improvement in inclusiveness and legitimacy. The framework’s two-stage process, with State-conducted baseline spatial analyses followed by municipal public discussions and Ministry approval, is a standard process common in developed countries, which only partially addresses legitimacy concerns by establishing institutional accountability mechanisms [68]. However, compared with integrated landscape governance approaches that emphasise multi-stakeholder dialogue throughout the planning process [69], the current framework’s stakeholder involvement occurs primarily at the proposal-refinement stage rather than during initial criterion selection and weighting. This sequential rather than iterative engagement structure may limit the substantive influence of diverse stakeholders on fundamental methodological decisions. While the framework’s transparency in decision rules and data accessibility supports procedural legitimacy, democratic innovations in participatory planning increasingly emphasise earlier and more continuous stakeholder involvement to enhance both input legitimacy and outcome acceptance [70]. Future iterations could strengthen inclusiveness by incorporating stakeholder co-production during the criteria development and validation phases.

Sixth, the classification could, in the future, address ecological dimensions, such as biodiversity, ecosystem services, landscape connectivity, and drinking water protection zones, which are increasingly considered in agricultural land management decisions [71]. The present framework deliberately prioritises agricultural production capacity and food security objectives, focusing on soil quality, topography, infrastructure, and current land use as primary classification criteria. This scope aligns with Slovenia’s regulatory mandate to identify strategically important agricultural areas for protection from urbanisation. However, integrated land evaluation literature increasingly demonstrates the value of incorporating ecological dimensions, including biodiversity, water protection, and carbon storage, into multicriteria agricultural land assessment [72,73]. Multicriteria approaches integrating ecosystem service supply and demand can identify synergies and trade-offs between production and environmental objectives, supporting more holistic land-use decisions [74]. Future methodological expansion could integrate these dimensions through: (a) overlay analysis incorporating biodiversity hotspots, groundwater protection zones, and high-carbon soils as additional classification criteria; (b) ecosystem service assessment modules quantifying multifunctional land contributions; and (c) scenario modelling evaluating protection strategies under combined production-environment objectives. Such integration would enhance the framework’s alignment with EU sustainability targets while maintaining its core agricultural protection function.

Seventh, climate change projections were not formally integrated into classification criteria. However, future climate conditions (water balance, available water resources, and flood or drought exposure) and adaptation will significantly influence long-term agricultural suitability [75,76]. These limitations suggest directions for future research, including periodic classification updates to capture temporal changes, integration of economic viability assessments, development of participatory classification approaches, and incorporation of climate change projections into evaluation criteria. The transferability of this framework to other contexts depends on the availability of comparable datasets and on adapting the criteria to local conditions. However, the fundamental multicriteria evaluation logic and transparent decision rules facilitate such adaptation across diverse European regions working toward agricultural land protection and sustainable spatial planning objectives.

3.7. Research Technical Roadmap

The land sustainability model developed in this study is already integrated into Slovenia’s national framework for identifying and protecting strategic agricultural areas, forming an operational component of municipal and national spatial planning. Building on this established role, further research should focus on strengthening data quality, methodological robustness, and long-term monitoring capacities to ensure consistent implementation across diverse landscapes. In the short term, priority should be given to improving the completeness and resolution of input datasets, particularly those related to land consolidation, drainage systems, and irrigation infrastructure, which remain limited or absent in several municipalities, including Novo mesto. Updating actual land use layers, refining soil and slope datasets, and standardising GIS workflows would enhance the precision and transparency of PPAL, OAL, and PSA classifications. In the medium term, the model could be expanded to incorporate additional criteria that reflect emerging challenges in agricultural land management, such as climate vulnerability indicators, ecosystem service assessments, and socio-economic characteristics of agricultural production. Applying the model across municipalities with contrasting geomorphological and land-use conditions would enable systematic cross-regional validation and sensitivity testing. At the same time, participatory engagement with farmers, planners, and local authorities would ensure that the model remains aligned with practical land management needs. In the long term, research should focus on developing adaptive monitoring mechanisms that enable continuous updates to strategic agricultural area classifications in response to environmental change, land-use dynamics, and technological developments. Integrating remote sensing, automated data pipelines, and climate adaptation strategies would support a dynamic, evidence-based system that guides agricultural land protection over time. Although the methodology is already embedded in national planning instruments, its ongoing refinement will ensure that it remains responsive to evolving policy priorities and contributes effectively to long-term food security and sustainable land management. This roadmap positions the model not only as a regulatory tool but also as a continuously evolving framework that supports sustainable land management and food security.

4. Conclusions

This study has developed the first fully transparent, nationally replicable multicriteria framework for strategic agricultural area classification in post-transition European contexts. Unlike existing approaches that rely on proprietary data or undisclosed expert judgement, the framework explicitly documents all decision rules, threshold values, and evaluation criteria using solely publicly available national datasets. The three-tier classification system (PPAL/OAL/PSA) provides discrete regulatory categories enabling direct integration into binding spatial planning regulations. Applied to Novo mesto Municipality, the framework identified 6490 ha (28%) as permanently protected agricultural land, 1020 ha (4%) as other agricultural land, and 5078 ha (22%) as potentially suitable areas.

The research advances scientific understanding in three areas. Methodologically, it demonstrates that hierarchical decision-tree logic with predetermined categorical boundaries enhances transparency and reduces parametric sensitivity compared to continuous scoring MCDA approaches. Empirically, it provides quantitative benchmarks on agricultural land distribution in peri-urban areas, serving as reference points for municipalities pursuing EU no-net-land-take objectives. Comparatively, it reveals that effective protection requires not only technical sophistication but also transparency, accessibility, and regulatory integration—qualities determining whether classification outcomes translate into implemented protection measures.

Farmers gain clarity on protected areas supporting informed long-term investment decisions and succession planning. Agricultural advisors can direct support programmes toward strategic areas where continued production is essential for food security. Water managers gain spatial insights, facilitating implementation of the EU Water Framework Directive. Spatial planners receive scientifically robust foundations for municipal decisions that comply with EU sustainable land-use requirements, enabling redirection of development from high-quality agricultural land toward brownfield regeneration and urban densification. However, complementary economic support mechanisms—direct payments, market access improvements, agri-environmental schemes—are essential for maintaining farmer voluntary compliance and long-term protection effectiveness.

Policymakers should formally incorporate classification outcomes into municipal spatial plans with binding regulations preventing PPAL conversion, establish economic support instruments ensuring farm viability on protected lands, implement monitoring systems tracking land-use changes and protection effectiveness, develop participatory approaches enhancing legitimacy and acceptance, and integrate protection frameworks with broader agricultural policies addressing climate adaptation and biodiversity conservation.

Future research priorities include systematic national implementation, establishing comprehensive strategic area inventories, methodological refinement incorporating climate change projections and economic viability analyses, longitudinal studies evaluating protection effectiveness, comparative analyses across EU member states identifying best practices, and the development of stakeholder co-production approaches to strengthen input legitimacy. This research establishes a methodological foundation for ongoing efforts to protect agricultural land resources essential to food security and sustainable development across the European Union as member states progress toward the 2050 no-net-land-take target.