Socio-Economic Viability of the High Nature Value Farmland under the CAP 2023–2027: The Case of a Sub-Mediterranean Region in Slovenia

Abstract

Our study aimed to analyse socio-economic sustainability and the drivers of land abandonment in the Kras region of Slovenia, a representative eastern Mediterranean farmland area. We also sought to provide policy recommendations for supporting biodiversity conservation and facilitating the sustainable transition of similar High Nature Value (HNV) farming systems across Europe. The Slovenian Typical Farm Model (SiTFarm) was used to assess the economic performance of representative livestock and wine-growing farm types. Additionally, in-depth interviews with farmers were conducted to understand their perspectives on these farming systems and their preferences for alternative management strategies and policy instruments. Our findings indicate that, due to the introduction of basic income support for sustainability and complementary voluntary coupled payments, budgetary support for the livestock sector in the region is projected to increase by 27–55% in estimated gross margins during the 2023–2027 Common Agricultural Policy (CAP) period, depending on the farm type. Furthermore, farms can enhance their economic performance by converting to organic farming and enrolling in agri-environmental schemes that promote extensive grasslands management, which is crucial for biodiversity conservation. This suggests that Slovenia’s current CAP strategic plan adequately addresses the maintenance of the existing farming systems. However, the region faces significant challenges, particularly in restructuring small farms and adding value to primary farm products. These issues appear to be insufficiently addressed by the current CAP strategic plan, implying that limited progress is expected in mitigating land abandonment in the long term. Comprehensive strategies for the development of feasible HNV farming systems, aligned with biodiversity conservation recommendations, and a well-managed system of supporting institutions and policy instruments is needed to facilitate more market-oriented and sustainable development of agriculture at the local level.

1. Introduction

One of the key societal challenges of the 21st century is to provide quality and affordable food for a growing human population, while also ensuring economic sustainability of farming and biodiversity conservation [1]. In Europe, the development of food systems has led to significant farmland biodiversity declines in recent decades [2]. Agricultural intensification coupled with the abandonment of farming in less productive areas is leading to the loss of traditional High Nature Value (HNV) farmlands, which support habitats for a large diversity of species [3]. These habitats are maintained through low-intensity farming practices, highly diversified land cover, and the preservation of semi-natural vegetation [4]. Moreover, HNV farmlands provide a range of nature’s contributions to various aspects of human well-being and quality of life [5,6].

Despite public investment aimed at halting biodiversity decline and preserving HNV farming systems that sustain it [7], low-intensity farmland in Europe continues to disappear [8]. The primary drivers of this trend are typically linked to biophysical constraints on modern agricultural production, remoteness from major urban centres, and various socio-economic factors, including structural transitions in agriculture, rural depopulation, and broader political and societal changes [9]. Compared to conventional agriculture, HNV systems are often characterized by lower use of pesticides and fertilizers, higher labour intensity and lower yields, making them less economically competitive [10,11,12]. In more productive European regions, HNV systems have gradually been replaced by more intensive and specialized farming systems; a transition sometimes accelerated by income support and reforms under the EU’s Common Agricultural Policy (CAP) [13]. Conversely, land abandonment frequently occurs in low-intensive and less profitable farmland, where modern agricultural practices are challenging to implement [14]. According to the European Commission’s Joint Research Centre estimate [15], over 20 million hectares (11%) of agricultural land in the EU is at high risk of abandonment between 2015 and 2030. Particularly vulnerable areas include parts of the Mediterranean, eastern Europe, northern Europe and the Balkan Peninsula [16], with negative consequences for both biodiversity and society [6]. Therefore, there is an urgent need for contemporary farming systems that enable economically efficient, environmentally sustainable and socially acceptable agricultural production, ensuring the long-term maintenance of these multifunctional landscapes [17].

To date, the CAP has had limited success in preventing the abandonment of HNV systems [11,18,19], despite the availability of policy instruments such as additional income support for marginal areas with natural constraints, voluntary agri-environmental measures, investment support and organic farming payments [20]. Several reasons for this have been identified. Many EU Member States have failed to develop effective and targeted measures to support low-intensity farming [11,21,22]. As a result, farms in HNV areas often receive less income and agri-environmental support than farms in areas with more intensive agricultural production [23]. Additionally, some agricultural land in these areas is deemed ineligible for support due to the high proportion of semi-natural vegetation [19]. Furthermore, evidence suggests that major shifts in the CAP’s income support structure, such as the decoupling of direct payments under the 2003 CAP reform, may have had negative effects on certain HNV farming systems [24,25]. To develop more targeted policy support, a better understanding of local drivers of change is necessary. This includes analysing the socio-economic sustainability and conservation effectiveness of typical HNV farm holdings [26,27], as well as farmers’ decision-making processes [28].

Economic, political and technological factors are recognized as the key drivers of change in traditional HNV farmland systems [29,30]; however, studies focusing on the economic aspects of these farming systems are relatively scarce [31]. Bio-economic farm models are frequently used to assess the economic conditions of farms and to simulate the impacts of various policies and market changes [32]. These models are typically based on mathematical programming and use detailed farm-level data to optimize production activities and to achieve maximum profit [33]. In recent years, farm models have evolved to incorporate environmental dimensions of farming [34], allowing for a more comprehensive analysis. For instance, they can be used to study the financial and environmental effects of conservation measures on farms [35,36]. Additionally, farm models facilitate the evaluation of how different policy instruments and reforms might influence farm incomes, production decisions, and the resilience of HNV farming systems; e.g., in Mediterranean [37], mountainous [38,39] and upland regions [24,40]. These models can thus serve as essential tools in understanding and supporting the sustainability of these complex and often vulnerable agricultural systems.

Despite the ambitious EU targets and the introduction of a new CAP Green Architecture, concerns have been raised about the post-2023 CAP’s ability to effectively halt biodiversity decline and prevent the loss of HNV farmland in Europe [41]. Moreover, it remains uncertain how the choices made by Member States will impact the economic situation of the various HNV farming systems [42]. This study seeks to address these concerns by focusing on the socio-economic sustainability of HNV farmland during the CAP post-2023 period. Using mathematical programming, the study examines the economic performance of a typical HNV farming system in the Kras region, a sub-Mediterranean area in Slovenia that experiences high rates of land abandonment [43]. An ex-ante modelling approach is employed to simulate the impacts of the post-2023 CAP reform on economic indicators of selected HNV farm types. This includes assessing the effects of different policy instruments on these farms’ economic performance. Using a mixed methods approach, the study complements the quantitative results with a qualitative analysis of the historical development of farms and farmers’ preferences regarding alternative farm management strategies and policy instruments. This component aims to provide an in-depth interpretation of the historical, social and legal reasons for the current economic performance of different farm types. Furthermore, it identifies conditions that could improve the economic outcomes of HNV farms, as well as key potential barriers. The study thus evaluates the performance of different farming systems that could support the long-term preservation of one of the typical HNV farming systems in European agriculture. Furthermore, it provides a comprehensive analysis of how CAP reforms and farm management strategies could influence the sustainability of HNV farmlands, ensuring that policy adjustments support both environmental and economic viability.

2. Study Area

2.1. Geographical Characteristics and Historical Development of Kras

The study area was selected in Slovenia, a Central European country that currently hosts one of the largest relative areas of HNV farmland in Europe [8,44]. However, over the last century, many of these areas have experienced land abandonment [43] and transformation into more intensively managed production systems [45,46].

Kras is a low limestone karst plateau in southwestern Slovenia, designated as a Natura 2000 site (SI3000276, SI5000023) (Figure 1). It is characterized by a sub-Mediterranean climate with hot, dry summers, limited precipitation and strong bora winds [47]. The site is a classic example of a karst landscape, featuring stony limestone terrain and shallow soils, which significantly affect farming conditions and limit agricultural development [48]. Over the past two millennia, the area has been almost completely deforested and transformed into dry, extensive grasslands and Mediterranean mosaic farmland, known for its high biodiversity [49]. However, since the 19th century, systematic afforestation with black pine and the abandonment of agricultural land have accelerated forest overgrowth, with forests now covering approximately 73% of the area [43].

Based on our own analysis of the Land Parcel Identification System (LPIS) data, in 2022, 1104 farms operated in the study area, cultivating a total of 9215 ha of agricultural land. The farmland is predominantly characterized by permanent grasslands and vineyards, with arable land being scarce (

Table 1. Analysis of land use types in Kras in 2022, by zones in %.

	Study Area		Zone 1		Zone 2		Zone 3
Land use type	ha	%	ha	%	ha	%	ha	%
Forest	37,858	59.3	17,626	63.1	17,568	60.9	2664	37.5
Successional stages	10,142	15.9	3494	12.5	3673	13.1	2975	41.9
Grassland	11,551	18.1	4638.7	16.6	6083	21.1	829	11.7
Arable land	1141	1.8	473	1.7	394	1.4	273	3.9
Vineyards	693	1.1	606	2.2	37	0.1	51	0.7
Other permanent crops	331	0.5	150	0.5	122	0.4	59	0.8
Other	2173	3.4	951	3.4	973	3.4	248	3.5
Total	63,889	100.0	27,938	100.0	28,850	100.0	7100	100.0

Note: Data source: MKGP land use types, 2022.

). The majority of farms (84%) managed less than 10 ha of land, indicating a prevalence of small-scale farming. According to the last national census of agricultural holdings in 2020, most farms in the municipalities within the study area identified as specialized wine producers (41%), while others were mixed farms (27%), specialized cattle breeders (19%) or arable farms (13%). The arable farms, however, are mostly located in neighbouring lowland areas, as arable production is not feasible in Kras. Therefore, arable farms were excluded from this study. Approximately 65% of the farms were fully or predominantly self-subsistent [50].

Historically, the area has been influenced by the Habsburg Empire (and partly the Venetian Republic), where the inheritance system led to equal land division. After World War II, the region was affected by Yugoslavia’s socialist policy, which limited the size of privately owned farms to 10 hectares [51,52]. This resulted in a highly fragmented land structure with complex ownership patterns, significantly slowing the restructuring of agriculture due to the high transaction costs associated with expanding farm sizes [22].

2.2. Biodiversity Conservation Objectives and Management Guidelines

The Natura 2000 management plan for the study area [53] can serve as a guideline for identifying conservation targets for Natura 2000 habitat types and species, as well as recommended management and farming practices (Appendix A). The most prominent farmland habitat type is the eastern sub-Mediterranean dry grasslands (Natura 2000 habitat type 62A0; order Scorzoneratalia villosae Kovačević 1959) [53]. This habitat includes plant communities from the alliance Chrysopogono-Saturejon Horvat and Horvatić 1934, which is more sclerophyllous and found primarily in extensive semi-natural pastures, and a more mesophyllous Scorzonerion villosae Horvatić 1963 alliance, which consists of extensive unfertilized hay meadows [54]. Many endangered wildlife species, such as Skylarks (Alauda arvensis) and now almost extinct Ortolan bunting (Emberiza hortulana), depend on these open dry grassland and stony habitats [55,56]. Additionally, species such as the Woodlark (Lullula arborea), Hoopoe (Upupa epops) and several butterfly and beetle species, also require a mix of dispersed trees, hedgerows, shrubs and dry stonewalls [Appendix A; [57]], typical of traditional Mediterranean mosaic farming systems [58].

Historically, dry grasslands in the Kras region were managed through extensive mowing or grazing, including with transhumance along the Balkan peninsula. Grasslands were maintained with low livestock density (mostly sheep, four to seven per hectare) [49], or by mowing one to two times per year, without the use of mineral fertilizers. The main conservation threat today is the abandonment of grassland management, leading to shrub encroachment and forest succession [43]. Grazing still occurs in parts of Kras, but now with fenced and rotational grazing paddocks and higher livestock density [59]. Conservation efforts focus on restoring a mosaic landscape with extensive grasslands and early successional stages, which support greater species diversity [57]. Extensive land use is also critical for preserving the unique cave fauna, including Proteus (Proteus anguinus), which requires low nitrogen and pesticide levels in groundwater (Appendix A).

3. Methods

3.1. Identification of Representative Farm Types in the Study Area

The analysis was based on economic modelling of representative farm types specifically developed for the study area. In the first step, farm types were identified and validated by analysing data from a survey of 263 farms in Kras (Figure 1) conducted by the authors in 2019. The survey gathered information on farm size, land use types, production activities, and participation in the CAP income support and agri-environmental instruments [22]. Based on these data, representative farm types were subsequently designed using the SiTFarm tool (Slovenian Typical Farm Model). The detailed characteristics of these representative farms were identified through two focus group discussions with local agricultural advisors held in December 2021 and October 2022, supplemented by additional consultations to refine the production characteristics of the farms. Local agricultural advisors were selected as key informants due to their extensive knowledge of the agricultural landscape and local farming systems in the study area.

Two basic farming systems were identified: wine-growing and livestock production (

Table 2. Livestock, land area and labour on the typical farms.

	BEEF40	BEEF10	SHP240	SHP45	HAY5	WINE1	WINE5
Livestock (no. animals)
Suckler cows	40	10
Heifers	8	2
Sheep			240	45
Plant production (ha)
Pasture	70.2	12.2	66.0	12.0
Hay meadow (unfertilized)	44.8	5.1			5.0	1.7
Hay meadow (fertilized)			4.8	8.0
Meadow + autumn pasture		14.7	49.2
Viticulture (ha)
Vineyard						1.0	2.0
Vineyard—superior quality						0.5	2.5
Utilized agri. area (ha)	115.0	32.0	120.0	20.0	5.0	3.2	4.5
Own permanent grassland	115.0	32.0	3.0	8.0	5.0	1.7
Leased grassland			117.0	12.0
Own vineyard						1.5	4.5
Own area overgrown with shrubs (ha)	0.5	0.5	3.0	0.5	0.5	1.5	1.5
Own forest area (ha)	3.0	0.5	0.5	0.5	0.5	0.5	0.5
Labour (FTE)	1.55	0.58	1.24	0.41	0.06	0.41	1.18

Note: ha—hectares, FTE—full-time equivalent equals 1800 working hours, BEEF40—a large cattle farm focusing on suckler cow rearing, BEEF10—a small cattle farm focusing on suckler cow rearing, SHP240—a large sheep farm, SHP45—a small sheep farm, HAY5—a small farm without livestock, which produces only hay on permanent grasslands, WINE1—a small mixed farm with one hectare of vineyards and some grasslands, WINE5—a medium-sized specialized viticulture farm.

). Wine-growing farms are predominantly located in the northern part of the area (Matični Kras, zone 1) (Figure 1). These farms can be categorized into two types: a medium-sized specialized viticulture farm with 4.5 hectares of vineyards (WINE5) and a small mixed farm with one hectare of vineyards and some grasslands (WINE1). Livestock farms, which specialize in either cattle or sheep rearing, are more prevalent in the central (zone 2) and the southernmost part (Kraški rob, zone 3) of the study area (Figure 1). Cattle farms focus on suckler cow rearing, with two types identified: a large farm with 40 suckler cows (BEEF40) and a small farm with 10 suckler cows (BEEF10). Among sheep farms, two types were distinguished: a large farm with 240 sheep (SHP240) and a small farm with 45 sheep (SHP45). Lastly, a small farm without livestock, which produces only hay on 5 hectares of permanent grasslands (HAY5), was defined.

Subsequently, we conducted a literature review, analysed existing data, and organized four focus groups with experts in botany, butterflies and birds, who were familiar with the study area. This expert analysis was used to evaluate whether the identified farming practices on typical farms aligned with the recommended management practices outlined in the Natura 2000 management plan (Appendix A) [53].

3.2. Modelling of the Farming Systems

3.2.1. Modelling Approach

Socio-economic sustainability was assessed through the economic analysis of selected farm types. The analysis focused on the performance of individual farms and whether their economic results support the sustainability and further development of each specific farm type. Our analysis primarily focused on monitoring the expected hourly gross margin (EGM) as a relative economic indicator. The hourly rate is determined by aggregating all market revenues in the initial phase, adding any budget payments received by the farm, and deducting total variable costs, including those related to herd renewal. Based on the total number of required working hours, we determined the EGM per hour invested. This approach enables direct comparisons between farms, allowing us to benchmark performance against results of other, or similar, farm types. By using hourly EGM, we can assess farm efficiency and economic viability in a manner that accounts for variations in production scale and operating conditions, offering a clearer understanding of relative performance across diverse farming systems. The concept of socio-economic sustainability in relation to the hourly EGM rate can be understood in that a low hourly rate signifies that income generated from farming activities is insufficient to sustain farming in the long term. This implies that investments are not only unfeasible, but the income itself is unattractive, potentially leading to the abandonment of farming. This conclusion is further supported by qualitative research. Consequently, economic and social sustainability have been combined into a single indicator (EGM/hour).

For this study, we utilized the SiTFarm tool [60]. SiTFarm is a bioeconomic farm model that uses a mathematical programming approach to facilitate various analyses, including sector-level evaluations [60], risk management strategies at the farm level [61], and resilience measurement through production plan adjustments [62]. It has a modular structure and has been developed in Microsoft Excel with automation implemented via Visual Basic for Applications (VBA) [60]. SiTFarm incorporates budget calculations from the Agricultural Institute of Slovenia [63], allowing for real-time adjustments to individual budget calculations concerning technology, intensity and price–costs relationships. This capability makes the model calculations system a key reference source of analytical and economic data at the level of individual production activities (enterprises) of analysed farms. The enterprise budgets are based on independent production simulation models that estimate technological and economic data. Consequently, the production costs of agricultural products are influenced by production technology, intensity, plot size and technological parameters [64]. SiTFarm allows for adaption of enterprise budgets to the specific conditions of individual farms, enabling flexibility in various areas, such as production or breeding technology, the number of pesticide application, the forage harvesting ratio and the quality of forage, among others. This adaptability ensures that the model can adapt to the unique circumstances and practices of each farm.

The model is based on classical linear programming (LP) [65]. In this context, the LP model represents the production plan of an agricultural holding as a linear combination of various production activities that adhere to the specific constraints of the analysed case. Each activity is characterized by technological coefficients that quantify its contribution to the objective function under the given constraints. The model’s output in this study was an optimized production plan that allocates land, labour, and fertilizer to maximize the EGM. In this context, the model’s critical function is to provide a comprehensive assessment of the farm’s economic situation, while facilitating the reconstruction of a technologically consistent production plan. This includes regulating and optimizing nutrient flows and ensuring that all production limitations are thoroughly addressed. Thus, the primary purpose of using LP in this study is not to optimize the overall production plan but rather to reconstruct the baseline production plan and balance it according to the key information for each typical farm. The basic idea, therefore, is to estimate or calculate the missing data—variables ($x_j$) —through the linear programme, which maximizes the EGM. The EGM is defined as the difference between total revenues (including market revenues and budgetary payments) and the associated variable costs.

Mathematically, the model can be described as follows:

$$max\mathrm{E}\mathrm{G}\mathrm{M}={\sum }_{j=1}^n{c}_j{x}_j+{\sum }_{f=1}^n{c}_f{x}_f$$

(1)

so that:

$${\sum }_{j=1;f=1}^{n;r}{a}_{ij}{x}_j+a_{if}{x}_f\le b_i \mathrm{for} \mathrm{all} i=1 \mathrm{to} m$$

(2)

$$x_f=b_f \mathrm{for} \mathrm{all} f=1 \mathrm{to} r$$

(3)

$$x_j\ge 0 \mathrm{for} \mathrm{all} j$$

(4)

where: i—sequential constraint in the model; j—sequential production activity, marketing activity, technological activity; f—sequential activity that defines type of a farm (e.g., number of suckler cows in the herd).

The objective function (1) represents the maximization of EGM at the farm level. The variables describing production and market activities ($x_j$) are determined using the simplex algorithm. Given the assumption of no switching between farm types, some production activities ($x_f$) that define the farm type (e.g., number of sheep, number of suckler cows) are fixed with additional constraints ($b_f$). The vector $b_i$ represents resource endowments (e.g., different land types and categories, available labour, stable capacity) and technical constraints (e.g., ration optimization, nutrient flows at the farm level, forage harvest ratio, restrictions related to grape processing, conditions for CAP measures and landscape features) within the model. To ensure the socio-economic sustainability of HNV agricultural land, the model integrates additional constraints ($b_i$) that define the permissible use on individual plots, the proportions of different crops and the intensity of land use. These constraints are aligned with traditional farming practices, such as limiting land use to one cut in late summer, thereby preserving existing biodiversity-friendly management regimes. This approach ensures that farming activities remain compatible with biodiversity conservation and landscape preservation goals, thereby maintaining the ecological integrity of HNV farmland while enhancing its socio-economic viability. The coefficients $c_j$ and $c_f$ in the objective function represent the EGM or cost per unit of each production or market activity ($x_j$ and $x_f$). These coefficients are critical in evaluating the economic performance of different farming activities within the model, helping to identify strategies that balance economic returns with environmental sustainability. This section also incorporates agricultural policy measures, including adjustments between different program periods (2014–2022 and 2023–2027). These measures are considered in the model when minimum conditions are met, even though farming practices on individual farms may not perfectly align with specific policy requirements. The payment amounts associated with these measures are integrated into the EGM at the farm level, thereby influencing the overall economic evaluation of the farm. This approach ensures that policy impacts are accurately reflected in the model, capturing the financial contributions of agricultural subsidies and support payments in the economic performance of farms.

Budget calculations served as the primary source of economic and technological data [64]. To mitigate the impact of annual price fluctuations, economic coefficients were calculated as the 3-year averages for variable costs and revenues, with each year’s probability being equal (1/3). Consequently, the average prices from the 3-year period of 2020–2022 were used in the analysis. The optimal solution for the LP model was obtained using Analytic Solver V2021 (21.0.0.0) from FrontlineSolvers^®.

Calibrating and validating a model are critical steps to ensure its accuracy and reliability. To achieve this, we reconstructed production plans based on the presumed data for each farm type (e.g., production intensity and size, share of available agricultural land, tillage and conservation technologies, allocation of key production inputs, etc.). While this allocation of resources may not represent the optimal scenario, it reflects real-world conditions, where deviations from optimal allocation can occur due to various factors. Additionally, agricultural consultants with in-depth knowledge of on-the-ground conditions were involved in the validation process by reviewing the baseline solutions.

3.2.2. Technological Assumptions of Production Activities

For each farm, we defined the utilized agricultural area and its structure, production technology and inputs, feeding requirements, available workforce, expected yields and intensity of production. Additionally, we accounted for market-based revenue, CAP and other types of budgetary support. These elements are described in detail below.

The beef production models were based on the rearing of suckler cows of Charolais/Limousine crossbreed. The final body weight of a cow was set at 600 kg, while pregnant heifers at the time of delivery weighed approximately 500 kg. Calving occurs between February and April, with a calf mortality rate of 8%. Suckler cows are replaced every 5 years from within the herd. Calves are sold as young animals (6–7 months old), with an average weight of 240 kg per animal, typically at the end of the main grazing season in autumn. Both cows and heifers are kept outdoors year-round, and fed by grazing and hay, with no additional concentrated feed purchased. Labour requirements are estimated at 61 h of effective work per cow per year in smaller herds, and 27 h per cow per year in larger herds. Market revenue primarily consists of sales from live calves and culled cows. Costs per suckler cow include those related to home-produced forage, bedding, pasture maintenance, insemination, insurance, capital and veterinary care. Additionally, expenses related to breeding heifers for replacement are considered.

The sheep farms in the study rear improved Jezersko-Solčava sheep, with lambing occurring in spring. On the small farm (SHP45), an average of 1.3 lambs per ewe are born, while on the large farm (SHP240), each ewe produces one lamb. Herd renewal occurs every 5 years on SHP45 and every 6 years on SHP240. Lambs are reared to a final weight of 36 kg and are sold either directly on the farm or for slaughter. Lambs exhibit an average growth rate of 200 g per day. The herd grazes on pastures for 9 months per year and is housed indoors during the winter months, where they are fed hay. The large sheep farm (SHP240) supplements feeding with additional concentrated barley feed (5 tons per year). Labour input requirements are higher in smaller herds, with 12 h of work per sheep per year, compared to larger herds, where the input is reduced by half per sheep. Both sheep farms have their own rams, which they replace every 2 years. Market revenue primarily comes from the sale of live lambs and culled ewes due to the herd renewal. Costs per ewe include expenses for home-produced forage, mineral and vitamin supplements, bedding, veterinary services, insurance and breeding of ewes and rams.

All livestock farms rely on permanent grasslands, with animals grazing on pastures and farmers producing hay. Due to the dry climate and stony karst landscape, production is generally extensive. Grazed feed and hay are typically of low nutritional value. Pastures are grazed at low livestock densities, averaging 0.2 livestock units per ha (LU/ha) on sheep farms and up to 0.5 LU/ha on cattle farms, with an average dry matter yield of 1.5 t/ha. On the large sheep farm (SHP240), the average yield is 1.2 t/ha. Hay meadows are mown once a year in late June before the summer drought. Most meadows are unfertilized, resulting in an average yield of 1.7 t/ha. Some meadows on sheep farms are fertilized with manure produced by the animals during the winter, resulting in slightly higher yields averaging 2.1 t/ha. A third type of grassland has mixed use, being mown once a year in June and then grazed briefly in autumn. The average yield for this grassland type is 1.9 t/ha on the small beef farm (BEEF10) and 1.7 t/ha on the large sheep farm (SHP240). Additional parameters for grassland production activities vary between farms. Plot sizes range from 0.5 and 1.5 ha, with distances from the farm location varying from 1 to 12 km and average distances between plots ranging from 0.5 to 1.5 km. The slope of plots is generally between 8% and 15%, except for vineyards, where a slope of 20% is common.

In addition to the livestock farms, several small farms in the Kras region maintain grasslands without rearing animals (HAY5). These farms mow their meadows once a year and sell the hay to dairy, beef, and horse farms located either within Kras or, more frequently, in other regions in Slovenia and Italy. These meadows are managed extensively without fertilization or maintenance beyond mowing and harvesting.

Wine-growing farms in the region predominantly cultivate Terran grapes. The model considered three wine quality classes: table wine (sold in bulk), quality wine (bottled in 1 L) and superior quality wine (bottled in 0.75 L). The proportion of each wine class varies by farm size. On a small farm (WINE1), 20% of grapes are used for quality wine, compared to up to 50% on a larger farm (WINE5). Superior quality wine is produced only in selected vineyards with lower yields. The majority (63%) of wine is sold in bottles, with the remaining 37% sold as bulk wine. Vineyards produce an average yield of 7000 kg of grapes per hectare, with a planting density of 4500 vines per hectare. In vineyards with 4000 vines per hectare, used for superior quality wine, the average yield is 4000 kg of grapes per hectare. Both small and larger wine-growing farms need to hire additional labour, primarily for harvesting (70% on WINE1 and 50% on WINE5), with larger farms also employing workers for green cutting (25%). Besides labour, costs include fertilizers, phytopharmaceuticals, bottling, and vineyard establishment.

All farms rely on family labour, with detailed data on labour expressed in full-time equivalent (FTE), corresponding to 1800 effective working hours per year (Table 2). Family labour is not treated as a cost in the EGM model, as it is unpaid. If additional labour is required, farms may hire workers for short-term tasks, such as harvesting, at an average cost of 5.5 €/h, and 6 €/h on wine-growing farms.

3.2.3. Budgetary Support and Scenarios

The model considered and compared agricultural policy interventions available during the programming periods 2014–2022 [66] and 2023–2027 [67] (

Table 3. Budgetary payments on typical farms in the programming period 2014–2022 and 2023–2027.

		CAP Pillar I				CAP Pillar II
Typical Farms	CAP Period	Direct Payments a	Voluntary Coupled Support		Eco-Scheme	Animal Welfare Payments		ANC	Organic Farming b	AECM b
			Suckler Cows	Sheep	Extensive Grassland	Cattle	Sheep
		(€/ha)	(€/LU)	(€/head)	(€/ha)	(€/LU)	(€/LU)	(€/ha)	(€/ha)	(€/ha)
BEEF40	2014–2022	167.87				53.4		108.68	155.57	258.10
	2023–2027	186.15	99.58		46.7	52.9		101.96	159.00	325.51
BEEF10	2014–2022	161.79						117.97	155.57	258.10
	2023–2027	191.22	99.58		46.7			115.15	159.00	325.51
SHP240	2014–2022	131.68					27.6	108.24	155.57	378.00
	2023–2027	186.07		18.52	46.7		92.47	101.33	159.00	331.12
SHP45	2014–2022	134.64					27.6	122.47	155.57	378.00
	2023–2027	195.43		18.52	46.7		92.47	126.90	159.00	331.12
HAY5	2014–2022	108.00						115.45		258.10
	2023–2027	211.58						82.51		362.03
WINE1	2014–2022	162.84						113.56	692.74 c	411.38
	2023–2027	211.58						78.76	888.00 c	581.12
WINE5	2014–2022	225.00						111.67	692.74	411.38
	2023–2027	211.58						75.01	888.00	581.12

Note: BEEF40—a large cattle farm focusing on suckler cow rearing, BEEF10—a small cattle farm focusing on suckler cow rearing, SHP240—a large sheep farm, SHP45—a small sheep farm, HAY5—a small farm without livestock, which produces only hay on permanent grasslands, WINE1—a small mixed farm with one hectare of vineyards and some grasslands, WINE5—a medium-sized specialized viticulture farm, ANC—Payments for Areas with Natural Constraints, AECM—Agri-environmental-climate measure, LU—livestock units). ^a Basic income support for sustainability in 2023–27 is 184.2€/ha for all hectares. In addition, farms receive 27.38 €/ha for the first 8.2 ha as complementary redistributive income support. ^b Organic farming and AECM payments were not included in the models as they are voluntary for farms. In AECM, maximum possible payment combinations are shown. However, in AECM, not all farmland is eligible for enrolment. Furthermore, if a farmer combines both instruments, AECM payments are reduced to avoid double funding (e.g., in BEEF, the amount is approx. 50% lower in case of combinations). ^c Only for 1.5 ha of vineyards as permanent grasslands are not eligible.

). Direct income support (CAP Pillar I) included both coupled and decoupled payments. In the 2023–2027 programming period, farms are eligible for coupled payments for sheep (18.52 EUR/head) and suckler cows (99.58 EUR/LU), which were not available during 2014–2022. In Slovenia, decoupled payments in the period 2014–2022 were entitlements, which were still partly based on historical payments and the distribution between permanent grassland and arable land (including vineyards). Therefore, they differed between farms and ranged between 108 EUR/ha and 225 EUR/ha (Table 3). In the 2023–2027 period, these entitlements were replaced by a new intervention, the Basic Income Support for Sustainability, set uniformly at 184 EUR/ha of utilized land. Additionally, there is Complementary Redistributive Income Support for Sustainability (27.4 EUR/ha), which is distributed to all farms in the study area, but only for the first 8.2 ha of land. This is also why the payment amounts vary between farms and are generally lower on larger farms (Table 3). Farms in the study area can also participate in the Eco-schemes introduced in 2023–2027, a voluntary CAP instrument that allows yearly enrolment [41]. According to the Slovenian CAP Strategic Plan, farms can engage in the Extensive Grasslands Eco-scheme, which restricts livestock intensity to 0.2 to 0.9 LU/ha and prohibits mineral fertilizers and pesticides on grasslands [68]. Due to its relevance, it is assumed that this Eco-scheme will be widely adopted by livestock farms in the study area (Table 3). The model also accounts for the national refund of excise tax, which varies by farm and depends on fuel consumption associated with machinery operations.

Farms in the study region also benefit from several rural development policy instruments under CAP Pillar II (Table 3). In the model, all typical livestock farms, with the exception of a small cattle farm, were assumed to receive payments for the Animal welfare measure. These payments are contingent upon meeting the minimum required standards, which focus primarily on the duration of the grazing period (at least 120 days for cattle and at least 210 days for sheep), and these standards were met by the cattle and sheep farm types in our study in both CAP periods [69,70]. Payments to Areas Facing Natural or Other Constraints (ANC) were also included in the simulation. Between the two CAP periods, there were changes in the ANC payment calculation methodology in Slovenia, which now considers factors such as the total number of difficulty points per farm, land size, and the presence of herbivorous animals. Consequently, there has been a slight decrease in ANC payments for most of the analysed farms, with the exception of the small sheep farm, which experienced a slight improvement (Table 3).

Organic Farming Payments and Agri-Environment Climate Measure (AECM) were not directly considered in the models, as these instruments are voluntary for farmers and participation can vary greatly among farms of the same type. Organic Farming Payments saw a modest increase in the 2023–2027 period compared to the previous one (Table 3) [71,72]. However, livestock rearing was set as a condition for receiving support for permanent grasslands in both periods, thus precluding farms of types HAY5 and WINE1 from benefiting from Organic Farming Payments. For AECM, optimal combinations of suitable schemes for permanent grasslands and vineyards were identified based on consultations with local agricultural advisors and a review of national legislation [71,72]. These schemes include measures such as delayed mowing or grazing of grasslands (after 30 May) and preventive measures against wolf and bear attacks (OGRM) on permanent pastures. In vineyards, several schemes aimed at reducing pesticide and mineral fertilizer use were available. Combinations between organic farming and AECM were not always permitted by national legislation, or the AECM payments were adjusted to prevent double funding [71,72].

3.3. Interviews with Farmers

We conducted 15 in-depth interviews with local farmers, whose contacts were provided by the local agricultural advisors. The interviewees were selected to represent different types of farms found in the study area, all of which were modelled in this paper. The interviews began with a conversation based on the interview guide, typically followed by a farm walk. The structure of the interview included: (1) a description of the farmer, farm characteristics and farm history, (2) characteristics and values related to farming in the local area, (3) perceptions of the karst landscape and agricultural land, (4) perceptions of biodiversity and the inclusion of conservation practices in farm management, and (5) farming in the context of the biodiversity conservation, including protected area management and access to markets.

The interviews were conducted in March 2023, with an average duration of 1 h and 45 min. All interviews were recorded, except for one, in which detailed notes were taken by the interviewer. Prior to each interview, farmers were informed about the study’s aims and data protection protocols, and they signed an informed consent from to participate. The recordings were transcribed and analysed using Atlas.ti (Cleverbridge, version 9). The analysis included multiple readings and a subsequent coding process to identify and organize relevant segments [73].

4. Results

4.1. Optimal Production Plan per Each Representative Farm Type

We present the results for the seven representative farms analysed using the SiTFarm tool, which integrates a linear programming (LP) algorithm, as detailed in Equations (1)–(4). The model’s results represent optimal production plans that account for all production constraints, including practical ones (3), which define the farm type and help bridge the gap between theoretical modelling and real-world conditions. The LP model, explained in more detail in Section 3.2.1, enables the consideration of a complex system of constraints while maximizing the objective function, which aims to optimize EGM when establishing the production plan of each farm.

4.2. Large Cattle Farm (BEEF40)

Historically, large cattle farms (BEEF40) in the region evolved from small dairy or mixed farms that have adapted to the political, structural and market changes in recent decades [74]. These farms expanded their operations and restructured into specialized cattle rearing. In our model, a typical cattle farm manages 40 lactating suckler cows including Charolais, Limousin, and occasionally Cika or Angus breeds. The farm produces sufficient heifers to maintain the herd, while surplus calves are sold at 5 to 7 months of age, usually coinciding with the end of the grazing season or the onset of grazing shortages in autumn. A breeding bull is maintained within the herd to facilitate natural breeding and enhance the sale of weaned calves for further fattening, along with some heifers reserved for future breeding. The farm operates on a combination of owned and leased land, forming a contiguous grassland area that supports a grazing season of 7 to 9 months. It manages approximately 69 hectares of grazing land and around 49 hectares of meadows (Table 2). The cattle are kept outdoors throughout the year. During the winter and for occasional supplementary feeding during the grazing season, the farm produces hay in the form of large bales, and no concentrated feed is purchased. In years with adverse conditions, such as drought, the farm attempts to compensate for feed shortages with surplus hay from previous years. If these reserves are inadequate, adjustments are made by reducing herd size, selling younger calves, or prematurely drying off cows.

According to the optimized production plan (SiTFarm), which considers all constraints, the workload for this farm is equivalent to approximately 1.5 FTE (Table 2). Typically, the farmer is fully engaged in agricultural activities and is insured as a farmer, while other family members contribute to the farm work to a lesser extent. Economically, the farm demonstrates a strong performance, allowing for necessary investments (

Table 4. Economic indicators for the typical beef, sheep and hay-producing farms in the programming periods 2014–2022 and 2023–2027 without potential enrolments in the AECM and organic farming.

		BEEF40		BEEF10		SHP240		SHP45		HAY5
CAP Period		2014–2022	2023–2027	2014–2022	2023–2027	2014–2022	2023–2027	2014–2022	2023–2027	2014–2022	2023–2027
Total Revenues	(EUR)	67,344	78,000	17,181	20,521	59,112	77,188	12,397	15,905	2406	2759
Market revenues	(EUR)	31,308	31,308	7827	7827	28,582	28,582	6868	6868	1234	1234
Budgetary payments	(EUR)	36,036	46,692	9354	12,694	30,530	48,606	5529	9037	1171	1525
Direct payments	(EUR)	19,306	21,408	5177	6119	15,802	22,329	2693	3909	540	1058
Voluntary coupled support	(EUR)	0	3983	0	996	0	4445	0	833	0	0
Eco-scheme	(EUR)	0	5365	0	1493	0	5598	0	933	0	0
Repayment of excise duty	(EUR)	1841	1841	402	402	746	746	200	200	54	54
ANC Payments	(EUR)	12,498	11,726	3775	3685	12,988	12,160	2449	2538	577	413
Animal welfare payments	(EUR)	2392	2370	0	0	994	3329	186	624	0	0
Variable costs	(EUR)	28,372	28,372	7173	7173	26,251	26,251	5617	5617	703	703
EGM	(EUR)	38,972	49,628	10,008	13,348	32,861	50,937	6780	10,288	1703	2056
	(EUR/h)	14.0	17.8	9.5	12.7	14.7	22.8	9.1	13.9	16.4	19.8
	(EUR/ha)	339	432	313	417	274	424	339	514	341	411
	(EUR/FTE)	25,208	32,100	17,178	22,910	26,445	40,992	16,463	24,983	29,468	35,581

Note: BEEF40—a large cattle farm focusing on suckler cow rearing, BEEF10—a small cattle farm focusing on suckler cow rearing, SHP240—a large sheep farm, SHP45—a small sheep farm, HAY5—a small farm without livestock, which produces only hay on permanent grasslands, FTE—full-time equivalent equals 1800 working hours, ANC—payments to areas facing natural or other specific constraints, EGM—expected gross margin.

). Total annual revenues are approximately 67,000 EUR, with potential additional revenue from organic farming and AECM payments. Notably, at least 54% of the revenues are derived from budgetary payments, which primarily cover all variable costs. The gross margin is around 39,000 EUR, or approximately 25,000 EUR per FTE.

The farm’s economic position could be further improved by converting to organic farming, as it already meets most of the required conditions, or by enrolling portions of its permanent grassland in AECM aimed at conserving ecologically valuable grasslands. According to interviewed farmers and agricultural advisors, this strategy is commonly pursued, as it can significantly enhance economic outcome, especially during drought years, which, as farmers have observed, are becoming more frequent in the region due to climate change. However, the decision to enrol can be deterred if farmers perceive the transaction costs as too burdensome or inconsistent with their strong sense of independence. Consequently, despite the economic incentives, some farms have opted to discontinue their organic farming certifications or AECM participation. Additionally, the lack of horizontal and vertical cooperation among producers has limited the establishment of value chains in the region that could add value to organically or locally produced meat. As a result, organic products often fail to achieve a market premium and are sold at a price comparable to conventional products.

“I got out of [all schemes], absolutely all of them. I was an organic farmer, one of the first in Kras. That was more than 20 years ago. At the time, everyone laughed at me for going into it, but I saw the future in organic farming. Over the years, however, it has become increasingly bureaucratic, politically or how to say. There is organic farming in the field and organic farming in the office. This is a big difference. Organic farming should be more about life than about papers. /…/ For organic farmers, the animal feed has a considerably [higher] price. However, the [difference] in price between my organic meat and animal feed defies logic. If I tell a customer that I have organic meat, I can’t raise [the price of the meat] as much as I paid more for feed. It doesn’t match, and it doesn’t go together. I can’t stand this pressure anymore, although I have many ideas.” (K3)

With the latest changes in the CAP in 2023, this type of farm experienced a significant increase in budgetary support, totalling at least 10,500 EUR (Table 4). With market revenues remaining stable, the share of budgetary payments in total revenues is expected to rise to 60%. This increase is primarily due to production-related direct payments (coupled support) and eco-schemes. Consequently, the hourly EGM rate per hour improves from 14 EUR to 17.8 EUR. Despite the extensiveness of the area under low-intensity farming, the average EGM per hectare remains relatively modest at 339 EUR/ha; although this figure is still above the average for livestock farms in the study area and is projected to improve by about 27% during 2023–2027. A similar positive trend is observed in the increased payments for the AECM, which can further enhance overall farm income. This suggests that government policies have successfully supported the development of this farm type in the region. However, it also indicates that these farms are becoming increasingly reliant on budgetary support, as little progress has been made in incentivizing and enhancing their cooperation for better market access. Consequently, market revenues are likely to continue stagnating. Despite the relatively favourable economic incentives, further growth of these farms and restructuring of other farms, such as BEEF10, into this farm type are significantly hindered by issues related to land access, as the land in the region is highly fragmented and difficult to purchase or lease, largely due to the existing land use policies in Slovenia [22].

From a biodiversity conservation perspective, the farming model of these farms is estimated as suitable for achieving conservation objectives, provided that grazing pressure on the dry grasslands remains relatively consistent with historical levels. This stability is likely due to the climatic and soil conditions that limit the potential for more intensive grazing practices. However, targeted small-scale measures are needed to preserve existing landscape features that are crucial for biodiversity and cultural heritage. While solitary trees and hedgerows are often retained in pastures as a shade for cattle, features such as dry stonewalls and small water ponds are now maintained by only a few farmers. The traditional boundaries between fields are gradually being removed, and water for cattle is increasingly supplied through closed cisterns rather than open water ponds.

4.3. Small Cattle Farm (BEEF10)

The small cattle farm (BEEF10) is a typical cattle-rearing operation found across the three zones of the Karst region (Figure 1). It raises 10 suckler cows of a mixed breed (possibly crossbred with a meat breed) and their calves. They primary focus is on selling calves, with no engagement in fattening. Historically, these farms were involved in milk production, but they have since transitioned to selling calves to local traders and direct meat sales from the farm. The farm exclusively maintains permanent grasslands, as small arable plots—once used for small-scale vegetable and wheat production—have been largely converted to grasslands in recent decades due to the loss of market access in urban centres such as Trieste (Italy) and Istria (Croatia). The farm manages approximately 32 hectares of grassland (Table 2), most of which is leased. Fodder production is exclusively meadow-based and generally of low quality, with no purchase of concentrated feed. Consequently, the growth rate of young animals is below the Slovenian average [74], at approximately 850 g/day. In terms of biodiversity conservation, this farming model is similar to the large cattle farm (BEEF40) in its suitability for achieving conservation objectives.

The farm requires a labour input of approximately 0.6 FTE to manage all necessary tasks, regarding optimized production plan (Table 2). Typically, this work is performed by a retired individual or the farm owner, who operates the farm alongside other employment, with occasional assistance from family members. Over half of the farm’s income (54%) is derived from budgetary payments, while variable costs reach around 42% of the total income. Although this farm has poorer economic outcomes compared to the large cattle farm (BEEF40), it achieves EGM of around 10,000 EUR and an EGM per hour of 9.5 EUR, representing average economic performance for this sector (Table 4).

For the current generation of owners, these farms often represent a way of life, preserved as part of family tradition. They can also serve as a supplemental income during regular employment or retirement and can support small investments to maintain basic assets. However, its social sustainability is questionable, as these farms often lack successors willing to continue farming. This reluctance is frequently due to a lack of capital necessary for significant investments in farm revitalization or restructuring. Unless successors are particularly innovative in securing funding or exploring on-farm diversification opportunities such as tourism, they often choose to leave the farm or abandon farming altogether, in favour of other employment opportunities.

“You often run out of money for all the investments, and you have to pay for everything yourself first, before [the government] pays you back [to cover part of the investment costs]. That’s a big ask. We, who farm on a small scale, don’t have any savings either, so it’s hard to afford it. /…/ Until now, we received enough subsidies to cover [variable costs]. We put so much [effort] in this land but hardly gain anything out of three-quarters of it. You need to renew the machinery, and the tractor works all the time. So if [the government] cancels the subsidies, it won’t pay for the oil, which means we will invest less. In the end, you have to save enough for the bills and for the schooling of your children. /…/ I would be very pleased if [the farm] continued [in the next generation]. But you can’t force anyone. /…/ If the boys take over, they will have to increase [the farm size] if they want to stay at home. Otherwise, by keeping only 6 animals, we have enough meat for all our wider family, so we have taken care of ourselves. /…/ I am not here to farm for others. If I did, [the farm] should have more surface area, but this is very difficult in Kras.” (K2)

Similar to BEEF40, CAP measures in the 2023–2027 period are expected to improve the economic situation of small cattle farms (BEEF10) by around 33% in EGM, mainly due to coupled payments for cows and Eco-scheme (Table 4). Despite this positive impact, however, this farm type still achieves a relatively modest EGM per hectare (313 EUR) compared to a larger cattle farm. This policy change is thus anticipated to provide the necessary support to maintain the existing farming model, and may help to slow down the trend of farm abandonment. However, the long-term development challenges of these farms are unlikely to be resolved through budgetary support alone. Targeted advisory services and investment support will be needed to facilitate either the expansion of farm size towards the BEEF40 model or farm diversification strategies, which can potentially add value to the primary products and supplemental income from cattle rearing. Economic outcomes could be further enhanced by converting to organic farming or enrolling in the AECM aimed at preserving grasslands, which is important for nature conservation. However, these owners seem to be more reluctant than their BEEF40 counterparts to participate in these schemes, perceiving transaction costs, including administrative burden and compliance controls, as too high [75].

4.4. Large Sheep Farm (SHP240)

Large sheep farms (SHP240) are predominantly located in zone 2 (Figure 1), where they manage extensive grassland areas, typically around 120 ha. Most of their land is leased, often from local agrarian communities or municipalities, with minimal land ownership and no arable land. These farms maintain a flock of approximately 240 sheep of the improved Jezersko-Solčava meat breed, along with two breeding rams and young stock. Lambs that are not retained for the flock, typically weighing up to 18 kg, are sold directly to slaughterhouses. The grasslands are primarily used for grazing (66 ha), with the remaining 54 hectares being mowed once a year (Table 2). Due to the poor quality of the pasture and hay, these farms need to purchase additional concentrated feed, leading to relatively low growth rates among the lambs.

The average livestock density (0.2 LU per ha) is lower than on cattle farms, which may be advantageous for managing targeted dry limestone grasslands, provided proper livestock rotation is implemented to prevent over- or under-grazing. However, similar to beef farms, landscape features such as dry stonewalls and water ponds are rarely maintained on these sheep farms, as they are no longer required and are not supported by targeted conservation incentives (Appendix A). One of the main challenges these farms face are conflicts with large carnivores, particularly wolves. To mitigate this, most farms have invested in electric fences and/or trained livestock-guarding dogs. These investments, together with trainings supported by several EU projects [76], have significantly reduced sheep mortality in the last decade, although wolf attacks still occur annually. Many farmers have enrolled in the AECM with targeted schemes that compensate for the higher costs of active preventive measures for herd protection [77].

Labour input on these farms amounts to approximately 1.2 FTE, with the owner fully engaged in farming activities (Table 2). The farm generates just under 60,000 EUR in annual revenue, with more than half (52%) derived from budgetary payments (Table 4), excluding potential AECM and organic farming support. As these farms meet the conditions for organic farming, enrolment in this measure can significantly boost revenue. Therefore, many local farmers have decided to convert. According to the optimal production plan derived from the SiTFarm model, variable costs account for 44% of total revenue, and the EGM per hour of work is 14.7 EUR, equating to just over 26,000 EUR per FTE per year.

The voluntary coupled support for sheep introduced in the 2023–2027 period has significantly improved the financial standing of this farm type, potentially boosting revenues by up to 55%. This increase elevates the EGM hourly rate to nearly 23 EUR, positioning it in the upper quartile of results for livestock farms in Slovenia [74]. Despite having below-average payment entitlements (131 EUR/ha) in the CAP period (2014–2022), the new basic income support system introduced substantial gains in direct payments for these farms, totalling around 6500 EUR annually due to their large size in terms of managed areas. Combined with coupled payments, these changes result in an annual increase of approximately 11,000 EUR, providing sufficient support to ensure the socio-economic and environmental sustainability of this farm type. Consequently, these farms can play an important role in preserving the agricultural landscape and targeted wildlife species in the region, with relatively modest labour and capital inputs.

4.5. Small Sheep Farm (SHP45)

Small sheep farms (SHP45) typically rear a herd of 45 Jezersko-Solčava breed sheep, along with one breeding ram and all the young stock. These farms manage around 20 ha of grasslands, with 12 ha dedicated to pastures and 8 ha to meadows that are mown once per year (Table 2). Small sheep farms are distributed across all zones of the Karst region (Figure 1). Despite the poor quality of pasture and manure, no concentrated feed is purchased. The biodiversity conservation impacts of this farming system are likely similar to those of large sheep farms (SHP240).

The effective labour input for all necessary tasks is around 0.3 FTE (Table 2). According to the optimal production plan derived from the SiTFarm model, the farm generates over 12,000 EUR in total annual revenue, with 45% derived from budgetary payments (Table 4). Variable costs approximately equal the budgetary payments received, making up 45% of total revenues. The EGM is just under 600 EUR per month, or slightly over 9 EUR per hour of work. Like the large seep farms (SHP240), the economic situation of small sheep farms improved significantly in the 2023–2027 program period, with a 52% increase in EGM. However, due to the small scale of these farms, this increase likely results in only a slight slowdown in structural changes, as this farm type remains economically unsustainable. Fixed costs are not covered in most years, preventing the farm from making necessary investments in machinery and farm buildings. To address these challenges, farms often rely on additional income from non-farm activities. For long-term economic viability, improvements in breeding technology and husbandry are necessary, along with increases in herd size and land area. Revenues could also be enhanced through conversion to organic farming and direct sales of lambs to consumers. However, the social sustainability of small sheep farms is questionable, as they are often maintained by retired individuals who engage in farming primarily for additional income or to preserve tradition and lifestyle. As seen with small cattle farms (BEEF10), a similar trend of farm abandonment is expected in the next decade, despite the increased budgetary support.

4.6. Hay-Producing Farm without Livestock (HAY5)

This farm type represents small farms in the region that have not successfully restructured or expanded their farm size in the last decades, leading many to abandon animal husbandry. Consequently, these farms now primarily focus on mowing their meadows once per year, often justifying this practice as a means of preserving cultural heritage and preventing the disappearance of the traditional agricultural landscape due to overgrowth. These farms are present across all three zones (Figure 1), with each managing about 5 ha of mown meadows (Table 2). The hay is sold in bales to other regions in Slovenia or Italy, with prices ranging from 80 EUR to 120 EUR per ton of hay bale during the analysed period. Mowing traditionally occurs around early June or follows historical regional practices. The meadows are unfertilized, resulting in modest yields of approximately 1.7 t of dry matter per hectare. Additionally, these farms often own about one hectare of land that was abandoned years ago due to remoteness, now experiencing natural overgrowth with forest vegetation.

According to the optimal production plan derived from the SiTFarm model, a labour input of approximately 0.1 FTE per year is sufficient for all necessary tasks on such farms, allowing them to be maintained alongside regular employment or with the older generation handling most of the work (Table 2). The farm generates an annual income of around 2400 EUR, half of which comes from budgetary payments. Variable costs, which include expenses for baling and transporting hay, total approximately 700 EUR (Table 4). This results in an EGM of 1703 EUR, equivalent to just over 16 EUR per hour of effective work. The CAP measures introduced in 2023–2027 are expected to somewhat improve the farm’s financial outlook. Although modest in absolute terms, these improvements could increase the hourly rate to just under 20 EUR. With these measures, budgetary payments would exceed market revenues, leading to a 21% enhancement in EGM.

As long as such a farm can sustain itself solely by maintaining existing fixed assets without requiring new investments, its economic outlook remains viable, allowing this management approach to continue for some time. However, without external inputs, the farm lacks the capability to invest in necessary renewals of its fixed assets, making this farming system likely transitional, with agricultural activities expected to be abandoned over the next decade or two. Local livestock farms willing to expand may acquire or lease some of these grasslands if support is provided to develop larger and consolidated farm areas. Where this process does not occur, it is reasonable to expect that most of this land will be subject to overgrowth with shrubs.

Currently, these farms play a crucial role in preserving biodiversity and open grassland landscapes as they represent a predominant farm type across the study area. Consequently, they are ideally suited for the AECM aimed at conserving grasslands biodiversity, given their ability to readily fulfil the management requirements. However, in practice, these farmers are the least likely to enrol in these measures due to their small farm size (and thus low payments received) and the obligations associated with inspections, record-keeping, and other administrative requirements [22].

4.7. Wine-Growing Farms

Small wine farms (WINE1) produce wine on 1.5 ha of fully owned vineyards using vertical production methods (Table 2). All vineyard tasks, including harvesting, are performed by family members, who also manage the processing and marketing of the wine. The total workload amounts to approximately 0.4 FTE. Wine is sold at prices ranging from 2.5 to 3.0 EUR/L, and 4.5 EUR/L for premium wine. Wine production is typically not the primary income source for the household, and often not even for the farm itself, as family members usually have regular employment outside of agriculture or are involved in tourism. According to the optimal production plan derived from the SiTFarm model, these farms generate annual revenues of approximately 20,000 EUR, with variable costs slightly exceeding 6000 EUR (

Table 5. Economic indicators for the typical wine-growing farms in the programming periods 2014–2022 and 2023–2027 without potential enrolments in the AECM and organic farming.

		WINE1		WINE5
CAP Period		2014–2022	2023–2027	2014–2022	2023–2027
Total revenues	(EUR)	20,617	20,661	64,693	64,468
Market revenues	(EUR)	19,568	19,568	62,777	62,777
Budgetary payments	(EUR)	1049	1093	1917	1691
Direct payments	(EUR)	521	677	1013	952
Eco-scheme	(EUR)	0	0	0	0
Repayment of excise duty	(EUR)	164	164	402	402
ANC payments	(EUR)	363	252	503	338
Variable costs	(EUR)	6351	6351	20,999	20,999
EGM	(EUR)	14,266	14,311	43,694	43,469
	(EUR/h)	19.5	19.6	20.6	20.5
	(EUR/ha)	4458	4472	9710	9660
	(EUR/FTE)	35,099	35,209	37,079	36,888

Note: WINE1—a small mixed farm with one hectare of vineyards and some grasslands, WINE5—a medium-sized specialized viticulture farm, FTE—full-time equivalent, ANC—payments to areas facing natural or other specific constraints, EGM—expected gross margin.

). Nearly all revenues come from market sales, with budgetary payments contributing only around 5%. The EGM surpasses 14,000 EUR, equating to 19.5 EUR per hour of effective work. As a family-run enterprise, particularly when combined with tourism, these farms are economically and socially sustainable. Given the minimal impact of budgetary payments on total revenues and EGM, the financial outlook for these farms is unlikely to change significantly in the 2023–2027 period. Small wine farms are present across all three zones of the Karst region (Figure 1).

Apart from the vineyard, this farm type also typically manages approximately 1.7 ha of grassland near the village, maintained as part of local tradition. The grassland is mown once a year, and the hay is sold to other regions in Slovenia or to Italy. Through the maintenance of meadows, these farms play a crucial role in preserving the cultivated and open cultural landscape, particularly in the broader vicinity of the village, thereby positively impacting biodiversity. Additionally, they own about 2 ha of more remote land that was previously used for cattle grazing but has since been abandoned and is now heavily overgrown or characterized by low-quality forest.

By contrast, medium-sized vineyard farms (WINE5) are primarily located in zone 1 (Figure 1), managing around 4.5 ha of vineyards (Table 2). All grapes are processed into wine, resulting in an annual production of approximately 77,000 L. Most of this wine is bottled, with 37% sold in bulk and 22% as premium wine. These farms typically own up to 15 ha of additional land, some of which they acquired or used in the past decades when they operated as mixed farms, including cattle rearing. However, due to poor economic results from cattle rearing, management of these grasslands has since been abandoned, and the farms have restructured to focus exclusively on wine production. This land is now in various stages of overgrowth or has already completely forested.

Viticulture operations on these farms (WINE5) require approximately 1.2 FTE, with 1.0 FTE attributed to family labour and additional seasonal hired labour employed for tasks such as pruning and harvesting. Typically, the farm owner manages viticulture professionally or coordinates with several individuals, often combining this activity with tourism. According to the optimal production plan derived from the SiTFarm model, annual income from viticulture is just under 65,000 EUR, with approximately 3% derived from budgetary payments (Table 5). Variable costs amount to around 21,000 EUR, resulting in an EGM exceeding 43,500 EUR, equivalent to nearly 21 EUR per hour of effective work. This financial performance demonstrates the farm’s economic sustainability, allowing for investments in the development and renovation of fixed assets. Despite a 6% reduction in direct payments per hectare expected in the 2023–2027 period, this change reflects only a marginal decrease of −0.5%.

Wine-producing farms in Kras currently likely have a limited impact on biodiversity in the study area due to their small size. However, those that still manage grasslands can also contribute to achieving conservation targets, and such activities should be promoted. The maintenance of landscape features, such as stonewalls and hedgerows that often surround meadows and vineyards, also provides important structural elements for biodiversity [78]. Few existing wine-growing farms have converted to organic farming, although this system seems to have been relatively well promoted by the local advisory service in recent years, as most producers are well informed about its principles. The primary reason for not converting appears to be that established local markets, through direct sales to consumers and local restaurants provide sufficient revenue to achieve favourable economic outcomes, leading farmers to perceive the added value of organic production as being too low to outweigh additional costs.

5. Discussion

This study examines socio-economic sustainability and drivers of agricultural abandonment in a typical eastern Mediterranean European landscape, where extensively managed farming systems support HNV farmland, playing a crucial role in biodiversity conservation within the EU [3]. The study also contributes to the existing literature in the southeastern European region, where the ecological and socio-economic impacts of land abandonment remain relatively poorly understood [6]. Given that dry grasslands represent the predominant agricultural land use [43], we identified four cattle and sheep-rearing farm types, along with a small hay-producing farm. In the northernmost part of the study area, these systems are complemented by small- and medium-sized wine-growing farms, which are relatively more intensively managed.

Our assessment of the conservation impacts showed that current livestock farming systems appear to support local biodiversity, particularly in areas characterized by eastern sub-Mediterranean dry grasslands [49,54], contributing to the achievement of targets set in the Natura 2000 management plan (Appendix A). Due to the natural constraints such as shallow soils, a dry climate, and the prevalence of surface limestone, intensifying these farming practices is likely unfeasible for most farmers. However, some unsustainable practices, such as overly intense grazing regimes or fertilization of meadows, have been observed in the past. While existing farming systems still allow for the preservation of many characteristic landscape features—such as water ponds, hedgerows and dry stonewalls—which are important habitats for biodiversity and species of high conservation importance [79,80]—these features are not sufficiently addressed by Slovenia’s CAP strategic plan. There are no targeted AECM or similar conservation instruments available for this region [67]. Furthermore, the lack of ecological research and data has hindered the development of detailed management recommendations at the level of individual farming systems, limiting the creation of targeted measures and advisory support [22]. This issue is common in many eastern European Member States, particularly in HNV farmland and even Natura 2000 sites [21,81]. Detailed management recommendations must also account for the expected impacts of climate change, especially extreme events such as droughts and wildfires. Farmers in our study noted that these events have become increasingly frequent in the past decade. However, due to a lack of data, we did not consider climate change impacts in our models. Future research is needed to evaluate the effects of these events on both farm economics and biodiversity conservation.

In contrast, wine-growing farms in our study area are characterized by more intensive management practices. However, given their small size, scarcity, and scattered presence in the local landscape, their overall impact on local biodiversity is likely minimal, and their importance as habitat features for some species may be limited [82]. As climate change drivers the anticipated global expansion of large-scale wine production in the Mediterranean region, there are concerns about potentially significant negative impacts on biodiversity and local natural resources [83]. Future development of this sector should therefore be carefully planned to address potential trade-offs between biodiversity conservation and economic goals. Encouraging the conversion of at least some wine-growing farms to organic production or other systems that reduce phytopharmaceutical use may be a viable strategy if the prevalence of these farms increases [84].

The primary conservation priority in the region is the revitalization of agriculture, particularly grassland management. On livestock farms, the low production intensity results in lower incomes from basic agricultural activities. However, CAP budgetary payments significantly improve income indicators (EGM/FTE), with income support contributing 45 to 63% of total revenue on livestock farms. Additionally, in the 2023–27 CAP period, a substantial increase in support for both beef farms (27% to 33% of EGM) and sheep farms (51% to 55%) is expected. This is primarily due to the abolition of historical payment rights [64] and the (re)introduction of voluntary coupled support for suckler cows and sheep [67]. The latest CAP reform in Slovenia has thus improved the economic sustainability of the existing livestock farming systems in these HNV farmlands. As a result, the two large livestock farm types in our study can provide full-time employment for a farmer, indicating social sustainability as well. Many of these farms also stand to benefit from participation in environmental instruments, such as organic farming and AECM, for extensively managed grasslands, particularly on large beef and sheep farms. When designed in alignment with conservation goals, such CAP environmental instruments can significantly enhance the socio-economic viability of extensive HNV farming systems [24], while also offering a safety net in case of unforeseen events such as extreme weather. These findings align with studies from other marginal and extensively managed systems in Europe, where the livestock sector has often been found to rely heavily on farm income support, making it sensitive to policy change [24,40].

However, the high dependence on budgetary support poses a potential risk in terms of public expenditure, especially given possible reductions in the future CAP budget. This issue is further exacerbated by stagnant market revenues for many farms. Few efforts have been made to increase the value of farm products in regional markets or to foster better cooperation among producers in Kras, which could improve market access. The current CAP strategic plan lacks effective strategies to address these concerns, despite the availability of some targeted instruments [68]. In contrast, wine-growing farms in the region have been more successful in developing supplementary farm activities, boosting their market revenues. Nonetheless, many wine growers reported limited cooperation with each other, often relying on independent sales channels. Development patterns among wine-growing farms suggest spatial agglomeration, where neighbouring farmers’ decisions strongly influence individual choices [85]. Collective approaches to farm development, particularly those driven by local actors such as cooperatives, could encourage cooperation in investments, market access, and the adoption of environmentally friendly practices [86,87], which could be particularly valuable for marginal farming systems [88]. A comprehensive local market strategy, potentially including branding and supported by CAP instruments, along with advisory support and incentives for local leaders is needed to facilitate these initiatives [46].

The main driver of land abandonment in this HNV system, however, remains the cessation of agricultural activities on small farms, which continue to dominate the region. Our analysis shows that the economic and social sustainability of small farms is precarious, as they struggle to cover fixed costs like investments in new machinery or infrastructure. These farms also demand a significant labour input, and they are predominantly managed by an aging, often retired generation, motivated by family tradition and a desire to maintain the traditional landscape. However, in the qualitative analysis, we found that maintenance of these farms is usually less attractive for the young generation of farmers, who either want to develop the farm so that it ensures at least one full-time occupation or they prefer to seek job opportunities elsewhere. Current CAP income support is essential for keeping these farms operational, as it often helps cover variable costs. However, simply maintaining the status quo is unlikely to prevent further land abandonment and subsequent biodiversity loss [89], which may be further exacerbated by climate change and water scarcity. The key challenge over the next two decades is how to facilitate the restructuring of the small farms in a socially, economically and environmentally sustainable way. Based on our farming model analysis, several pathways appear necessary for these farms to thrive: (i) increasing their farm size and product value [90], (ii) transitioning into specialized sectors such as wine-growing, or (iii) diversifying activities, such as tourism or breeding sport horses for riding. These are indeed the strategies that were taken by many of the large farms which were included in the interviews. However, incentives for farm specialization and restructuring must be carefully managed to prevent the adoption of more intensive farming systems that could harm local biodiversity [13]. Therefore, rural and agricultural development should be closely aligned with the conservation policy, supported by effective legal regulation and policy instruments to prevent undesired agricultural development pathways [25].

In the absence of a regional development strategy and land policy instruments, the currently fragmented land structure in Kras significantly slows down the restructuring process [22]. Targeted AECM for biodiversity conservation, alongside other instruments, could play a crucial role in providing supplementary income support for small farms, while ensuring that this transition benefits biodiversity and achieves conservation objectives. At present, these schemes are rarely adopted by small farms, as farmers often perceive transaction costs to be too high [22]. However, innovations in AECM design, such as result-based schemes [91] and locally based incentives for farm cooperation [76], could better align farmers’ preferences and improve the perceived costs and benefits associated with enrolment.

Additionally, strategies are needed to revitalize agricultural land that has already been abandoned [58]. The current extent of extensively managed grasslands needs to increase to successfully meet conservation objectives (Appendix A). There are also broader benefits, such as wildfire prevention, which has become increasingly important over the last decade. Adaptation to these climate-related risks might offer significant opportunities for funding the development of agriculture and biodiversity conservation in such HNV areas in the future [92].

Our study demonstrates that combining farm management modelling and qualitative research is a promising methodological approach for exploring the complex economic, social, and biodiversity conservation conditions and interactions in the study area. We used the SiTFarm tool to demonstrate the suitability of bioeconomic farm modelling in assessing the economic performance and perspectives of marginal farming systems commonly found in the HNV farmland [9]. The evaluation of the results shows that the model performs relatively well as long as suitable data and farming calculations are available. In our case, some estimations had to be adjusted, as standard calculations prepared for Slovenian agriculture did not always accurately reflect the farming systems and conditions found in the study. This highlights that these systems are relatively poorly researched and thus are rarely included in the evaluations of the impacts of the CAP and other major trends in agriculture, which is probably similar in at least some other EU countries [11]. In addition, specific calculations are needed for some biodiversity-friendly farming practices and the restoration of grasslands. The study would thus benefit from an integrated model that simultaneously considers both economic and biodiversity factors. The current approach, which treats these aspects separately, does not fully capture the complex interplay between economic activities and biodiversity conservation. Finally, it is important to note that the SiTFarm tool currently does not account for fixed costs, which can significantly impact the estimated economic sustainability due to the variations in capital intensity and needs between farms. However, we believe that the expert assessment, supplemented by qualitative analysis from farmer interviews, has enabled reasonably accurate interpretation of the likely impacts of fixed costs on the overall economic performance of different farm types.

6. Conclusions

Open and mosaic landscapes in the Mediterranean basin, which have developed over thousands of years of extensive human use, are recognized as one of the global biodiversity hotspots [93]. This study employs a multidisciplinary approach, combining farm modelling with qualitative research to assess the socio-economic performance of a selected HNV farming system in Slovenia. The study area has experienced a high level of farm abandonment in recent decades [43]; a trend observed in several other European countries [14,16]. Our findings suggest that extensive farming systems with low yields do not necessarily lead to land abandonment. Instead, this process is more likely influenced by other sector characteristics, such as the small size of farms, which currently fails to provide adequate incomes and full-time employment, rendering them socially unsustainable. Limited access to land, advisory support, investment, and market opportunities [89] means that even motivated small farms often lack the conditions necessary for expansion and sustainable development.

Our findings suggest that harmonizing agricultural production, nature conservation, and rural development in the Mediterranean HNV farmland areas is feasible and can be complementary, provided that appropriate development models for sustainable farming systems, targeted income support and other policy instruments are established. Additionally, a well-managed regional system of supporting institutions is essential to offer advisory support, facilitate land access [22], and promote market-based and sustainable agricultural development at the local level that can add value to local products [94]. However, further research is required to understand how CAP reforms and farm management strategies can influence the sustainability of HNV farmlands, ensuring that policy adjustments support both biodiversity conservation and economic viability and avoid potentially harmful effects of agricultural development [39]. This necessitates stronger integration of economic and ecological research, which is often limited by the lack of fine-scale data on biodiversity and farm management practices.

Furthermore, bio-economic farm modelling tools, such as the SiTFarm tool used in this study, need to be enhanced with appropriate economic assessments of biodiversity-friendly farming practices and anticipated impacts of climate change. Since socio-economic effects were not simulated in this research, it is clear that this represents an important area for future investigation. The model should be upgraded to support dynamic analyses, which could identify farm types that are likely to expand or undergo structural change over time. Such advancements would represent a valuable contribution to agricultural modelling, and we strongly recommend pursuing this approach in future research to gain deeper insights into the socio-economic dynamics of farming systems.