Are Wolves the Real Problem? Challenges Faced by Livestock Farmers Living Alongside Wolves in Northwestern Greece

Abstract

Mitigating human–wolf conflict is crucial, yet conventional approaches often overlook the broader socioeconomic challenges faced by farming communities. Wolves frequently become scapegoats for deeper rooted issues such as economic disadvantages, policy deficiencies, and rural depopulation. We conducted semi-structured interviews with 118 livestock farmers to examine (a) farmer profiles and wolf-related interactions, (b) professional challenges and proposed solutions, (c) reasons for perceiving wolves as a major problem, and (d) the impact of wolf presence on job dissatisfaction. Farmers reported low specialized education and job satisfaction, particularly regarding income. Many struggled to afford or find shepherds, especially sheep/goat farmers. Guardian dog poisoning incidents and dissatisfaction with the damage compensation system were prevalent. Key challenges included economic marginalization, wolf presence, climatic factors, inadequate grazing policies, infrastructure deficits, distrust in policy, rural depopulation, and a lack of services. Farmers who perceived wolves as a major problem implemented weaker preventive measures and moved herds seasonally over longer distances. Job dissatisfaction was linked to wolf presence, livestock type, and economic marginalization. Our findings emphasize that while wolves impact farmers, economic and policy-related factors play a greater role. Educational initiatives, supportive policies, effective depredation mitigation, and fair compensation systems are essential for sustainable livestock farming and coexistence with wolves. By tackling socioeconomic challenges, enhancing policies, and supporting farmers to adapt to evolving circumstances, the livestock farming sector can thrive while minimizing conflicts associated with wolves.

1. Introduction

Throughout human history, the relationship between gray wolves (Canis lupus) and humans has been characterized by tension and unease, primarily due to the perception of wolves as a hazard to human lives and livelihoods [1]. As a result, wolves were eradicated from large parts of their global range [2]. Identifying the drivers of human–wildlife conflict and mitigation strategies is a topical issue [3,4]. In the traditional definition of human–wildlife conflicts, the human role is often excluded from the analysis, with solutions focusing solely on wildlife [5]. While implementing effective mitigation measures is always beneficial [6], it is important to recognize that the causes of conflict are often more complex and deeply rooted and require more comprehensive approaches to address them effectively [7]. Recently, there has been growing recognition of the complexity of socio-ecological systems in addressing human–wildlife conflicts, leading to the proposal of multidisciplinary approaches that integrate ecological, economic, and social perspectives [4,8,9,10]. These approaches draw upon methods and concepts from diverse disciplines, including conservation biology, anthropology, economics, social psychology, and development studies, with equal attention given to wildlife management and the human dimensions of the issue [4,11,12].

To understand the attitudes of people toward wolves and other large carnivores, previous researchers have investigated various socioeconomic factors. These factors include the age, gender, education level, and involvement in hunting/farming activities [13,14,15,16,17], as well as income/wealth [18,19,20] and a rural or urban residence [21]. The most positive attitudes toward wolves are held by younger and more educated people [4,16,17], urban residents [4], and those with better ecological knowledge of wolves [14,16,17]. Farmers and hunters generally have the most negative attitudes towards wolves [22]. Livestock farmers tend to have more negative attitudes towards wolves compared to other sympatric carnivores, such as bears, lynxes, snow leopards, and wolverines, likely due to higher incidences of livestock depredation by wolves [15,20]. However, in regions like Norway and Sweden, even though wolf attacks on livestock are less frequent than those by other large carnivores, they provoke strong controversy and negative feelings [23].

Negative interactions with wildlife can have significant impacts on the livelihoods of local communities, contributing to increased poverty and declines in overall well-being [24]. These interactions result in direct financial costs for stakeholders, such as the loss of livestock and crops [3]. Additionally, stakeholders face indirect costs, investing time and money in preventing wildlife-related damages [3]. Both direct and indirect costs contribute to poverty or worsen existing poverty levels, ultimately undermining human well-being [24]. Rural livelihoods are also influenced by various external factors, including social, economic, contextual, and policy factors [24], including large-scale agricultural policies [25]. However, the magnitude of the impacts of human–wildlife conflict on the well-being of livestock farmers, relative to other external factors, is often overlooked, and research in this area remains limited [18,26,27]. Social research on human–wildlife conflict tends to focus solely on the conflicts themselves, without integrating the broader socioeconomic context and the other challenges faced by the farming communities. This narrow focus can exaggerate the perceived impacts of human–wildlife conflict and can even distract from more pressing issues faced by the farming sector. Wolves, in particular, are perceived as a major threat to livestock farming [28,29] and have become an ideal scapegoat for other problems in rural areas, such as economic disadvantages, depopulation, and the loss of basic services [30]. Livestock farmers often see wolves as something tangible and easily blamed, making it simpler to focus on them rather than dealing with more complex socioeconomic and political issues [30]. Consequently, wolves have been used as political symbols and scapegoats, instrumentalized by people for political gain [28,30,31]. To illustrate this, French politicians exaggerated the impact of livestock attacks and opposed wolf recovery while ignoring the underlying market-driven decline in sheep farming, which was caused by imports of lambs from outside the EU [32]. Similarly, the recent proposal to reduce the protection status of wolves in Europe cites increasing socioeconomic challenges and harm to livestock as justifications [33], but without proper scientific evidence. To effectively address these issues, it is essential for conservationists to bring attention to the pressing issues faced by farmers and redirect the public and political focus toward these more significant concerns.

Here, we focused on gray wolves in NW Greece as a case study, an area with persisting wolf populations and a traditional husbandry system with free-ranging herds in unfenced pastures with varying levels of surveillance. We built upon previous work that examined the influence of husbandry practices on wolf depredation losses on sheep/goat and cattle herds, which proposed the subsidization of preventive measures through the Common Agricultural Policy [34]. Social research on wolves in Greece in scarce, with only three published papers available. The first, published nearly 30 years ago, explored wolf hunts as a symbolic practice among livestock farmers [35]. The second investigated teenagers’ beliefs about wolves in both rural and urban areas [21]. The third examined narratives, rumors, and social representations of wolves among livestock farmers [36]. Additionally, in recent years, there has been an increase in social research on the sustainability well-being, and challenges of livestock farmers in European countries, including Greece [37,38,39,40]. In this study, we integrated concepts from agroeconomic, social, and ecological disciplines to examine the characteristics of livestock farmers in relation to both wolves and well-being. Specifically, we aimed to (a) examine the profile of farmers in the study area, including wolf-related interactions for both cattle and sheep/goat farmers, (b) identify and prioritize the major challenges faced by livestock farmers and their suggestions to improve their professional conditions, (c) investigate the drivers of the perception that wolves are a major problem in livestock farming, and (d) assess the relative contribution of wolf presence as a driver of job dissatisfaction. Our work contributes to understanding the complex dynamics between wolves and livestock farmers and provides valuable insights for developing effective conservation strategies to promote successful coexistence.

2. Materials and Methods

2.1. Study Area

The study area covered over 6800 km² in NW Greece (see also [34]) (Figure S1). The area is mountainous, with a mean elevation of 935 m, and is covered by natural grasslands and shrublands (35%), broadleaved, coniferous, and mixed forests (31%), and agricultural land (29%) [41]. This region falls within the geographical range of wolves in Greece, including 15 wolf packs [42], while a recent camera trap study showed that wolves are widespread in a large part of the study area [43]. Livestock farming, including free-ranging sheep/goat and cattle herds, is a significant economic activity in the region. Wolf predation poses a considerable challenge, leading to estimated annual losses of €140,000 [44]. Farmers employ non-lethal methods, such as shepherding, night enclosures, guardian dogs, and the confinement of young animals [34]. Protected areas constitute 60% of the study area, including 21 sites within the European Natura 2000 network. Hunting restrictions apply to only 6.3% of the study area, distributed across 24 wildlife reserves.

2.2. Interviews with Livestock Farmers and Derived Variables

The farmers to be interviewed were selected under a stratified random approach to represent a gradient from very low to very high livestock losses, as determined by official compensation data (Figure S1), following the methodology of Petridou et al. [34]. During 2018–2020, we conducted semi-structured interviews of 118 farmers and collected their responses on 22 questions relevant to socioeconomic and wolf-related issues, the main challenges that they face in their profession, and the suggested measures (Table S1). We used data on livestock losses caused by wolves and prevention measures (2 questions) from a previous study undertaken in the same period and involving the same farmers [34].

To describe the socioeconomic profile of farmers, we considered 11 socioeconomic variables, related to age, education specific to livestock farming, age at farming initiation, farming generation, livestock type and size, and seasonal movements (Table 1). We categorized the age into five groups: 18–29, 30–39, 40–49, 50–59, and ≥60 years. We classified the education levels as elementary, secondary, or higher education. We categorized the education specific to livestock farming as none, short-term seminars, cheese/meat school, or zootechnical school. We classified the age at the initiation of livestock farming as an adult (≥18 years old) or non-adult (<18 years old). We categorized the livestock type as sheep/goats or cattle. We classified the farming generation as first, second, and three or more generations, and for cattle farming, we recorded the generation of cattle farming initiation separately to capture transitional changes. To homogenize the livestock type and cattle ages, we recalculated the herd size using Livestock Unit (LSU) coefficients, by assigning the following values: cattle < 1 year = 0.4, cattle aged 1–2 years = 0.7, cattle > 2 years = 0.8, and sheep or goats = 0.1 [45]. We categorized seasonal movements between summer and winter pastures as stationary (0 km), short-distance (less than 10 km), and long-distance (over 20 km). Farmers who perform seasonal herd movements typically move their herds between higher elevation pastures in summer and lower elevation pastures during winter. To evaluate job satisfaction among farmers, we assessed their satisfaction with their income and profession and their desire for their children to continue in the same profession. We measured satisfaction levels on a Likert scale ranging from 1 (not at all satisfied) to 5 (very much satisfied), while we categorized the desire for their children to continue in the profession as yes, no, no children, and do not know.

To collect data on wolf-related aspects of livestock farming, we considered 5 variables related to livestock depredation, prevention measures, shepherd hiring, guardian dog poisoning incidents, and satisfaction with the damage compensation system (Table 1). Petridou et al. [34] provided a detailed discussion of the losses caused by wolves and their relationship with husbandry practices in the study area. For this study, we used two indexes for livestock losses: killed LSUs per weighted year [34] and the percentage of LSUs killed per weighted year.

We developed a preventive measures score using a 10-point scoring system that combined three main measures that significantly influenced depredation in both sheep/goat and cattle herds: night confinement, shepherd surveillance, and the number of guardian dogs per 100 livestock animals [34]. Each farmer was assigned individual scores for each one of the three preventive measures based on their level of application (Table S1). For sheep/goat herds, we considered the number of Greek guardian dogs per 100 sheep/goats, and for cattle herds, the number of guardian dogs per 100 cattle. We developed a scoring system (see Table S1) using the results obtained by Petridou et al. [34] that identified the optimal number of guardian dogs per 100 livestock animals. Each farmer’s scores for the three preventive measures were summed to obtain their overall preventive measures score.

We also examined whether farmers employed a shepherd and, if not, their reasons (e.g., not needed, affordability, unavailability). Additionally, we explored the use of modern deterrents, such as light and sound deterrents and electric fences, to mitigate wolf depredation. Moreover, we asked farmers about incidents of guardian dog poisoning over the past five years, including the specific animals targeted and their opinions on who was responsible. Lastly, we assessed farmers’ satisfaction levels with the damage compensation system using a Likert scale ranging from 1 (not at all satisfied) to 5 (very much satisfied).

To understand the perceived magnitude of conflict with wolves relative to other challenges, we asked farmers to identify the three primary problems they encounter in livestock breeding, along with their three main suggestions for improving livestock farming. We then categorized these problems and suggestions into broader categories based on expert opinion. Notably, during the interviews, several farmers mentioned more than three problems and suggestions as the conversation progressed. We chose to retain these additional responses as many could be grouped into the broader categories established during the subsequent analysis (refer to Table 2 and Table 3).

To ensure ethical data collection, we obtained written consent from all interviewees after explaining the purpose of the study and assuring them of the anonymity and security of their personal information following the General Data Protection Regulation. The final database used for analysis comprised a total of 118 farmers (59 sheep/goat and 59 cattle farmers).

2.3. Data Analysis

We analyzed survey data in the program R, version 4.2.3 [46]. An illustration of the methodological approach and data analysis is provided in Figure S2. We used descriptive statistics to assess the data and present the results. To test whether socioeconomic variables differed significantly between sheep/goat and cattle herds, we applied the Wilcoxon signed-rank test for numeric variables and Pearson’s chi-square test for ordinal, nominal, and binary variables.

To examine the influence of farmer characteristics on the perceived conflict with wolves, we used Generalized Linear Models (GLMs) with a binomial distribution, using responses to the question “What are the three main challenges you face in livestock farming?” as the response variable. If a farmer mentioned wolves as one of their challenges, the response was coded as “yes” (1); otherwise, it was coded as “no” (0). Prior to GLM modeling, we conducted four preparatory steps. First, we checked for multicollinearity among numeric variables using the Pearson correlation coefficient, ensuring |r| < 0.5. If variables were correlated, they were included in a separate set of models. Second, we standardized numerical variables by subtracting their mean and dividing them by 2 standard deviations [47]. Third, we set the lowest level of ordinal variables as the reference and merged the levels of categorical variables with minimal observations. Fourth, considering the limited sample sizes, we followed the rule of thumb of having a minimum of 10 outcome events per predictor to avoid overly complex models and overfitting [48], resulting in a maximum of 11 predictors in our models. To achieve this, we implemented the Least Absolute Shrinkage and Selection Operator (LASSO), a regression analysis technique that conducts variable selection and regularization [49], employing 100 permutations and a grouped penalty option using the “grpreg” package in R [50]. We then proceeded to GLMs, using the variables selected through the LASSO process (Figure S3). We performed the model selection using a multi-model inference approach that compared all possible combinations of variables using the ‘MuMIn’ R package [51]. We ranked the models using the Akaike Information Criteria for small samples (AICc) and focused on models with ΔAICc < 2 [52]. We calculated the model-averaged coefficients (β), standard errors (SEs), p-values, and 95% confidence intervals (CIs). We considered variables to significantly influence the response variable if their 95% CI did not overlap with zero.

To study job satisfaction among livestock farmers, we used a composite response variable called the “job satisfaction score” derived from three variables. First, we summed the scores of satisfaction with their profession and income. Next, we assigned scores of −1, 0, or 1 to the responses “no”, “don’t know/no children”, and “yes”, regarding the farmers’ desire for their children to continue in the same profession, and we incorporated this score into the job satisfaction score. Job satisfaction scores ranged from 2 to 10. As predictor variables for our models, we used the challenges faced by farmers. To do so, we categorized the challenges into broader categories, as outlined in Table 2. We then created binary predictor variables, coded as “1” if a farmer mentioned a problem belonging to the respective broader category and “0” otherwise. Additionally, we included the livestock type as another predictor variable. For constructing our models, we followed the same approach with GLMs, first using LASSO regression (Figure S4) and then Ordinal Logistic Regression, an extension of logistic regression that is used to model the relationship between an ordinal response variable and one or more predictor variables.

3. Results

3.1. Profile and Wolf-Related Interactions of Farmers

3.1.1. Profile of Farmers

In total, we used 118 interviews for the analysis. A detailed descriptive profile of the respondents is provided in Table 1. The majority of respondents were male (93.2%). The age distribution was diverse, covering different age groups: 9.3% were aged 19–29, 22.9% were aged 30–39, 28.0% were aged 40–49, 24.6% were aged 50–59, and 15.3% were more than 60 years old. The education attainment was relatively low, with most farmers only having attended elementary (33.1%) or secondary school (61%) and very few holding a higher education degree (6%). Notably, most farmers lacked formal education in livestock farming, with 75% reporting no specialized training and 13% having attended only short-term seminars. Only 12% of respondents had attended a cheese/meat school or a zootechnical school.

Knowledge of farming practices seemed to be transferred from generation to generation, as most farmers (91.0%) followed a multi-generational tradition of livestock farming and started to be involved in the family business at a young age, even before reaching adulthood (75.0%). Different types of seasonal movements were well represented in our sample, covering all three types (i.e., stationary: 22%; short-distance: 39%, and long-distance: 39%). These figures underline the importance of transhumance in the region, as the majority of herds participated in either short- or long-distance seasonal movements.

When comparing sheep/goat and cattle farmers, we found significant differences (p < 0.05 for Wilcoxon and X² tests) only in certain aspects (Table 1). The most notable difference was the one related to the livestock type transition, with nearly all cattle owners being first-generation (64.4%) or second-generation (33.9%) farmers.

3.1.2. Wolf-Related Interactions

Farmers reported an average of 3.6 ± 5.1 Livestock Units (LSUs) killed by wolves per weighted year, equivalent to 5% ± 5.1% of their herds (Table 1). Cattle farmers experienced higher absolute numbers of LSUs killed per weighted year compared to sheep/goat farmers, but the percentage of the herd loss was similar in both livestock types (5%). Sheep/goat farmers demonstrated higher scores for implementing preventive measures, indicating a greater focus on proactive protection.

Regarding shepherd hiring, 43.0% of farmers employed a shepherd and 30.0% did not need one, while 27.0% could not afford or find a skilled one. Significantly more sheep/goat farmers reported difficulties in affording or finding shepherds compared to cattle farmers (42.4% vs. 11.9%, p < 0.0001). The adoption of modern deterrents was limited (Table S2), with farmers using devices such as small solar lights on livestock pen fences (n = 22), propane cannons (n = 11), flashing road lights (n = 5), firecrackers (n = 3), and radios (n = 1). Only a few farmers used carnivore-proof electric fences around pens or overnight spots (n = 6). Notably, many cattle farmers (n = 22) used one-line electric fences in pastures to restrict cattle movement.

Among the 108 farmers with guardian dogs, half of them reported incidents of guardian dog poisoning over the past five years (Table 1), with an average of 5.4 poisoned dogs per farmer (ranging from 1 to 40; 290 poisoned dogs in total). According to farmers’ beliefs, illegal poisoned baits were primarily targeted at guardian dogs (reported by 21 respondents), followed by foxes (11 respondents), hunting dogs (5 respondents), and wolves (4 respondents; Figure S5). Farmers attributed the use of illegal poisoned baits to hunters (reported by 24 respondents) and other livestock farmers (8 respondents) (Figure S5). Notably, sheep/goat farmers experienced significantly higher rates of guardian dog poisoning compared to cattle farmers (64.4% vs. 27.1%, p < 0.0001, Table 1).

Farmers expressed a general dissatisfaction with the damage compensation system, reporting a mean satisfaction score of 2.2 out of 5 (Table 1). Sheep/goat farmers were significantly more dissatisfied than cattle farmers (mean scores of 1.9 vs. 2.5, p = 0.03). Farmers cited multiple reasons contributing to their dissatisfaction, including insufficient compensation amounts paid for verified depredations, the exclusion of certain livestock mortality causes from the compensation scheme, the requirement of visible signs of depredation on killed animals for compensation eligibility, and difficulties in detecting carcasses (Table S3).

Table 1. Socioeconomic and wolf-related characteristics of 118 livestock farmers in NW Greece, with comparison between sheep/goat and cattle farmers. Results are presented as mean values (±SD) with Wilcoxon test (W) results for numeric variables and as percentages with chi-square test (X²) results for ordinal, nominal, and binary variables. Significant differences (p < 0.05) are highlighted in bold. Five variables (*) are wolf-related. Maximum values are highlighted in gray.

AA	Variables	Type	Unit	All Farmers	Sheep/Goat	Cattle	W/X2 Test	p-Value
1	Age categories	Numeric; count of 1–5	1 = 18–29	11 (9.3)	8 (13.6)	3 (5.1)	6.50	0.17
			2 = 30–39	27 (22.9)	10 (16.9)	17 (28.8)
			3 = 40–49	33 (28.0)	20 (33.9)	13 (22.0)
			4 = 50–59	29 (24.6)	14 (23.7)	15 (25.4)
			5 = ≥60	18 (15.3)	7 (11.9)	11 (18.6)
2	Level of education	Ordinal	Elementary	39 (33.1)	23 (39.0)	16 (27.1)	1.90	0.38
			Secondary	72 (61.0)	33 (55.9)	39 (66.1)
			Higher	7 (5.9)	3 (5.1)	4 (6.8)
3	Education on livestock farming	Ordinal	None	89 (75.0)	49 (83.1)	40 (67.8)	4.08	0.13
			Seminars	15 (13.0)	6 (10.2)	9 (15.3)
			Cheese/meat school	12 (10.0)	3 (5.1)	9 (15.3)
			Zootechnical school	2 (2.0)	1 (1.7)	1 (1.7)
4	Age at initiation of livestock farming	Binary	Non-adult	89 (75.0)	47 (79.7)	42 (71.2)	0.73	0.39
			Adult	29 (25.0)	12 (20.3)	17 (28.8)
5	Farming generation (b)	Ordinal	First generation	8 (6.8)	3 (5.1)	38 (64.4)	102.95	<0.0001
			Second generation	3 (2.5)	0 (0.0)	20 (33.9)
			Many generations (≥3)	107 (90.7)	56 (94.9)	1 (1.7)
6	Livestock type (a)	Binary	Sheep/goat	59 (50.0)	-	-	-	-
			Cattle	59 (50.0)	-	-
7	Herd size (a)	Numeric	Livestock Units (LSUs)	79.6 ± 100.1	39.4 ± 21.5	119.7 ± 128.4	390.0	<0.0001
8	Seasonal movements between summer and winter pastures	Ordinal	Stationary (0 km)	26 (22.0)	13 (22.0)	13 (22.0)	0.17	0.91
			Short-distance (<10 km)	46 (39.0)	22 (37.3)	24 (40.7)
			Long-distance (>20 km)	46 (39.0)	24 (40.7)	22 (37.3)
9	Satisfaction	Ordinal;	1—Not at all	33 (28.0)	28 (47.5)	5 (8.5)	27.83	<0.0001
	with income	Likert	2—Only a little	30 (25.4)	16 (27.1)	14 (23.7)
		scale	3—Moderately	33 (28.0)	8 (13.6)	25 (42.4)
			4—Much	21 (17.8)	7 (11.9)	14 (23.7)
			5—Very much	1 (0.8)	0 (0.0)	1 (1.7)
		Numeric	1–5	2.4 ± 1.1	1.9 ± 1.1	2.9 ± 1.1	871	<0.0001
10	Satisfaction	Ordinal;	1—Not at all	18 (15.3)	12 (20.3)	6 (10.2)	3.82	0.43
	with profession	Likert	2—Only a little	15 (12.7)	9 (15.3)	6 (10.2)
		scale	3—Moderately	36 (30.5)	15 (25.4)	21 (35.6)
			4—Much	22 (18.6)	10 (16.9)	12 (20.3)
			5—Very much	27 (22.9)	13 (22.0)	14 (23.7)
		Numeric	1–5	3.2 ± 1.4	3.1 ± 1.4	3.4 ± 1.2	1520.5	0.23
11	Desire for	Nominal	Yes	29 (24.6)	12 (20.3)	17 (28.8)	1.23	0.54
	children’s		No	46 (39.0)	25 (42.4)	21 (35.6)
	involvement		Do not know	5 (4.2)	2 (3.4)	3 (5.1)
			No children	38 (32.2)	22 (37.3)	21 (35.6)
12 *	Livestock losses caused by wolves (a)	Numeric	Killed LSUs per weighted year	3.63 ± 5.1	1.6 ± 1.5	5.6 ± 6.5	916	<0.0001
		Numeric	% killed LSUs per weighted year	4.98 ± 4.7	5.0 ± 5.0	5.0 ± 4.4	1667	0.70
13 *	Preventive measures score (a)	Numeric	Aggregation of levels of three preventive measures using a 10-point system (Section 2.2, Table S1)	6.1 ± 2.2	7.4 ± 1.5	4.8 ± 2	2918	<0.0001
14 *	Shepherd hire	Nominal	Yes	51 (43.0)	19 (32.2)	32 (54.2)	14.15	<0.001
			No—Does not need one	35 (30.0)	15 (25.4)	20 (33.9)
			No—Cannot afford/cannot find one	32 (27.0)	25 (42.4)	7 (11.9)
15 *	Satisfaction with damage compensation system (ELGA)	Ordinal; Likert scale	1—Not at all	47 (40.0)	30 (50.8)	17 (28.8)	8.81	0.03
			2—Only a little	22 (19.0)	12 (20.3)	10 (16.9)
			3—Moderately	32 (27.0)	10 (16.9)	22 (37.3)
			4—Much	10 (8.0)	5 (8.5)	5 (8.5)
			5—Very much	7 (6.0)	2 (3.4)	5 (8.5)
		Numeric	1–5	2.2 ± 1.2	1.9 ± 1.2	2.5 ± 1.2	1653	0.007
16 *	Guardian dogs lost due to illegal poisoned baits, last 5 years	Nominal	Yes	54 (45.8)	38 (64.4)	16 (27.1)	9.02	0.003
			No	53 (44.9)	21 (35.6)	32 (54.2)
			No dogs in the herd	11 (9.3)	0 (0.0)	11 (18.6)

(a) Variables derived from Petridou et al. ([34]). (b) Values in the “All Farmers” column for the farming generation do not represent the sum of the sheep/goat and cattle farmers columns, due to the shift from sheep/goat to cattle farming. These specific columns were subsequently used in the models.

Table 1. Socioeconomic and wolf-related characteristics of 118 livestock farmers in NW Greece, with comparison between sheep/goat and cattle farmers. Results are presented as mean values (±SD) with Wilcoxon test (W) results for numeric variables and as percentages with chi-square test (X²) results for ordinal, nominal, and binary variables. Significant differences (p < 0.05) are highlighted in bold. Five variables (*) are wolf-related. Maximum values are highlighted in gray.

AA	Variables	Type	Unit	All Farmers	Sheep/Goat	Cattle	W/X2 Test	p-Value
1	Age categories	Numeric; count of 1–5	1 = 18–29	11 (9.3)	8 (13.6)	3 (5.1)	6.50	0.17
			2 = 30–39	27 (22.9)	10 (16.9)	17 (28.8)
			3 = 40–49	33 (28.0)	20 (33.9)	13 (22.0)
			4 = 50–59	29 (24.6)	14 (23.7)	15 (25.4)
			5 = ≥60	18 (15.3)	7 (11.9)	11 (18.6)
2	Level of education	Ordinal	Elementary	39 (33.1)	23 (39.0)	16 (27.1)	1.90	0.38
			Secondary	72 (61.0)	33 (55.9)	39 (66.1)
			Higher	7 (5.9)	3 (5.1)	4 (6.8)
3	Education on livestock farming	Ordinal	None	89 (75.0)	49 (83.1)	40 (67.8)	4.08	0.13
			Seminars	15 (13.0)	6 (10.2)	9 (15.3)
			Cheese/meat school	12 (10.0)	3 (5.1)	9 (15.3)
			Zootechnical school	2 (2.0)	1 (1.7)	1 (1.7)
4	Age at initiation of livestock farming	Binary	Non-adult	89 (75.0)	47 (79.7)	42 (71.2)	0.73	0.39
			Adult	29 (25.0)	12 (20.3)	17 (28.8)
5	Farming generation (b)	Ordinal	First generation	8 (6.8)	3 (5.1)	38 (64.4)	102.95	<0.0001
			Second generation	3 (2.5)	0 (0.0)	20 (33.9)
			Many generations (≥3)	107 (90.7)	56 (94.9)	1 (1.7)
6	Livestock type (a)	Binary	Sheep/goat	59 (50.0)	-	-	-	-
			Cattle	59 (50.0)	-	-
7	Herd size (a)	Numeric	Livestock Units (LSUs)	79.6 ± 100.1	39.4 ± 21.5	119.7 ± 128.4	390.0	<0.0001
8	Seasonal movements between summer and winter pastures	Ordinal	Stationary (0 km)	26 (22.0)	13 (22.0)	13 (22.0)	0.17	0.91
			Short-distance (<10 km)	46 (39.0)	22 (37.3)	24 (40.7)
			Long-distance (>20 km)	46 (39.0)	24 (40.7)	22 (37.3)
9	Satisfaction	Ordinal;	1—Not at all	33 (28.0)	28 (47.5)	5 (8.5)	27.83	<0.0001
	with income	Likert	2—Only a little	30 (25.4)	16 (27.1)	14 (23.7)
		scale	3—Moderately	33 (28.0)	8 (13.6)	25 (42.4)
			4—Much	21 (17.8)	7 (11.9)	14 (23.7)
			5—Very much	1 (0.8)	0 (0.0)	1 (1.7)
		Numeric	1–5	2.4 ± 1.1	1.9 ± 1.1	2.9 ± 1.1	871	<0.0001
10	Satisfaction	Ordinal;	1—Not at all	18 (15.3)	12 (20.3)	6 (10.2)	3.82	0.43
	with profession	Likert	2—Only a little	15 (12.7)	9 (15.3)	6 (10.2)
		scale	3—Moderately	36 (30.5)	15 (25.4)	21 (35.6)
			4—Much	22 (18.6)	10 (16.9)	12 (20.3)
			5—Very much	27 (22.9)	13 (22.0)	14 (23.7)
		Numeric	1–5	3.2 ± 1.4	3.1 ± 1.4	3.4 ± 1.2	1520.5	0.23
11	Desire for	Nominal	Yes	29 (24.6)	12 (20.3)	17 (28.8)	1.23	0.54
	children’s		No	46 (39.0)	25 (42.4)	21 (35.6)
	involvement		Do not know	5 (4.2)	2 (3.4)	3 (5.1)
			No children	38 (32.2)	22 (37.3)	21 (35.6)
12 *	Livestock losses caused by wolves (a)	Numeric	Killed LSUs per weighted year	3.63 ± 5.1	1.6 ± 1.5	5.6 ± 6.5	916	<0.0001
		Numeric	% killed LSUs per weighted year	4.98 ± 4.7	5.0 ± 5.0	5.0 ± 4.4	1667	0.70
13 *	Preventive measures score (a)	Numeric	Aggregation of levels of three preventive measures using a 10-point system (Section 2.2, Table S1)	6.1 ± 2.2	7.4 ± 1.5	4.8 ± 2	2918	<0.0001
14 *	Shepherd hire	Nominal	Yes	51 (43.0)	19 (32.2)	32 (54.2)	14.15	<0.001
			No—Does not need one	35 (30.0)	15 (25.4)	20 (33.9)
			No—Cannot afford/cannot find one	32 (27.0)	25 (42.4)	7 (11.9)
15 *	Satisfaction with damage compensation system (ELGA)	Ordinal; Likert scale	1—Not at all	47 (40.0)	30 (50.8)	17 (28.8)	8.81	0.03
			2—Only a little	22 (19.0)	12 (20.3)	10 (16.9)
			3—Moderately	32 (27.0)	10 (16.9)	22 (37.3)
			4—Much	10 (8.0)	5 (8.5)	5 (8.5)
			5—Very much	7 (6.0)	2 (3.4)	5 (8.5)
		Numeric	1–5	2.2 ± 1.2	1.9 ± 1.2	2.5 ± 1.2	1653	0.007
16 *	Guardian dogs lost due to illegal poisoned baits, last 5 years	Nominal	Yes	54 (45.8)	38 (64.4)	16 (27.1)	9.02	0.003
			No	53 (44.9)	21 (35.6)	32 (54.2)
			No dogs in the herd	11 (9.3)	0 (0.0)	11 (18.6)

(a) Variables derived from Petridou et al. ([34]). (b) Values in the “All Farmers” column for the farming generation do not represent the sum of the sheep/goat and cattle farmers columns, due to the shift from sheep/goat to cattle farming. These specific columns were subsequently used in the models.

3.2. Challenges and Suggestions for the Livestock Farming Profession

3.2.1. Challenges

Livestock farmers identified several key challenges affecting their profession (Table 2, Figure 1). The primary concern, reported by 55.9% of farmers, was economic marginalization, attributed to factors such as the low prices of milk and meat, high feed costs, insufficient subsidies, high taxes, high shepherd costs, and high fuel costs. The presence of wolves was the second most frequently reported challenge, mentioned by 41.5% of farmers. Climatic factors, particularly water shortages and harsh weather conditions, were identified as challenges by 39.8% of farmers.

Approximately one-third of farmers (32.2%) highlighted issues related to grazeland planning policy, such as the lack of management plans and poor grazeland quality. Infrastructure deficiencies were noted by 30.5% of farmers, who pointed to poor road conditions, inadequate communal pen infrastructure, and a lack of electricity in mountain pastures. A lack of trust in policy implementation was expressed by 22.9% of farmers, who referenced insufficient state support, inefficient damage compensation schemes, bureaucracy, inadequate inspections for subsidy allocation, and limited information regarding potential funding programs.

Additionally, 22.9% of farmers expressed concerns related to a lack of services and rural depopulation, mentioning isolation, poor product distribution, a shortage of available spouses, and the scarcity of skilled shepherds and veterinarians. While 17.8% of farmers expressed concerns about bears and their predation, it is noteworthy that all farmers who mentioned bears also mentioned wolves. Notably, only 8% of farmers identified livestock diseases as a challenge.

3.2.2. Suggestions

Livestock farmers provided several suggestions to improve the farming sector (Table 3, Figure 1). The majority (62.7%) emphasized the importance of addressing economic marginalization by increasing subsidies, raising milk and meat prices, and reducing feed costs. Nearly half (47.5%) emphasized the need for policy improvement, calling for increased state support, enhanced inspections for subsidy allocation, the promotion of farmers’ cooperatives, and reduced bureaucracy.

Farmers also highlighted the need to improve infrastructure (45.8% of farmers), suggesting road condition improvement, improving pen infrastructure in communal mountain grazelands, and renovating or constructing water points and reservoirs, the latter of which is related to addressing climate-related challenges. Farmers deemed access to electricity as essential, either through a grid connection or subsidizing solar panels.

Farmers also noted the grazeland planning policy (23.7% of farmers) by emphasizing the need for the completion and implementation of planned grazeland management plans while ensuring the pasture quality. As a response to the challenge of rural depopulation and the lack of services, farmers underlined the need for cheeseries, veterinary care, improved product distribution, and the construction of slaughterhouses (12.7% of farmers).

Although the presence of wolves was highly ranked as a challenge, only 11.0% of farmers raised predator-related solutions, suggesting the need for prevention measures’ funding, carnivore removal, and the implementation of zoning systems to separate predators from livestock grazing areas.

Table 2. Livestock farmers’ perceptions of the main challenges affecting their profession (n = 118 farmers in total; 59 sheep/goat farmers and 59 cattle farmers). The results are presented as the number of reports (N reports) for each item for all farmers and separately for sheep/goat and cattle farmers. The results are also shown as the number and percentage of farmers reporting at least one item belonging to the specific category (Farmers’ N (%)). Numbers in parentheses indicate the ranking order of the challenge categories.

Challenge Category	Farmers’ N (%)	Item	N Reports	Sheep/Goat	Cattle
Economic	66 (55.9%)	Low prices of milk/meat	53	37	16
marginalization		High feed costs	24	15	9
[1]		Insufficient subsidies	19	9	10
		High taxes	9	8	1
		High social insurance costs	5	4	1
		High shepherd costs	4	3	1
		High fuel costs	3	3	0
		High transhumant movement costs	3	1	2
		High medicine costs	2	1	1
		High grazeland rent costs	1	0	1
Predators	49 (41.5%)	Wolf predation	49	24	25
[2]		Bear predation	21	6	15
Climatic factors	47 (39.8%)	Water shortage	30	11	19
[3]		Harsh weather conditions	26	14	12
Grazeland	38 (32.2%)	Lack of grazeland management plans	23	11	12
planning policy		Poor grazeland quality	15	7	8
[4]
Infrastructure	36 (30.5%)	Poor road condition	28	16	12
[5]		Poor pen infrastructure in communal mountain pastures	13	6	7
		Lack of electricity (mountain pastures)	9	6	3
Rural	27 (22.9%)	Isolation	7	5	2
depopulation		Poor product distribution	7	2	5
and lack of		Lack of wife	6	4	2
services		Lack of skilled shepherds	4	1	3
[6]		Lack of cooperative initiatives	2	0	2
		Lack of veterinarians	3	2	1
		Lack of milk concentration stations	3	2	1
		Lack of cheeseries	2	2	0
		Lack of slaughterhouse facilities	1	0	1
Distrust in	27 (22.9%)	Lack of state support	13	9	5
policy		Inefficient compensation scheme	5	3	2
[7]		Bureaucracy	3	3	0
		Challenges with pen permit acquisition	3	1	2
		Low educational level among farmers	2	2	0
		Insufficient info on funding programs	2	2	2
		Inadequate inspections for subsidy allocation	2	2	0
		No promotion of indigenous livestock breeds	2	2	0
		Inadequate support for young farmers	1	1	0
		Issues with social insurance	1	1	0
		High pension age	1	1	0
		Difficulty in legally hiring shepherds	1	0	1
Diseases [8]	9 (7.6%)	Livestock diseases	9	4	5

Table 2. Livestock farmers’ perceptions of the main challenges affecting their profession (n = 118 farmers in total; 59 sheep/goat farmers and 59 cattle farmers). The results are presented as the number of reports (N reports) for each item for all farmers and separately for sheep/goat and cattle farmers. The results are also shown as the number and percentage of farmers reporting at least one item belonging to the specific category (Farmers’ N (%)). Numbers in parentheses indicate the ranking order of the challenge categories.

Challenge Category	Farmers’ N (%)	Item	N Reports	Sheep/Goat	Cattle
Economic	66 (55.9%)	Low prices of milk/meat	53	37	16
marginalization		High feed costs	24	15	9
[1]		Insufficient subsidies	19	9	10
		High taxes	9	8	1
		High social insurance costs	5	4	1
		High shepherd costs	4	3	1
		High fuel costs	3	3	0
		High transhumant movement costs	3	1	2
		High medicine costs	2	1	1
		High grazeland rent costs	1	0	1
Predators	49 (41.5%)	Wolf predation	49	24	25
[2]		Bear predation	21	6	15
Climatic factors	47 (39.8%)	Water shortage	30	11	19
[3]		Harsh weather conditions	26	14	12
Grazeland	38 (32.2%)	Lack of grazeland management plans	23	11	12
planning policy		Poor grazeland quality	15	7	8
[4]
Infrastructure	36 (30.5%)	Poor road condition	28	16	12
[5]		Poor pen infrastructure in communal mountain pastures	13	6	7
		Lack of electricity (mountain pastures)	9	6	3
Rural	27 (22.9%)	Isolation	7	5	2
depopulation		Poor product distribution	7	2	5
and lack of		Lack of wife	6	4	2
services		Lack of skilled shepherds	4	1	3
[6]		Lack of cooperative initiatives	2	0	2
		Lack of veterinarians	3	2	1
		Lack of milk concentration stations	3	2	1
		Lack of cheeseries	2	2	0
		Lack of slaughterhouse facilities	1	0	1
Distrust in	27 (22.9%)	Lack of state support	13	9	5
policy		Inefficient compensation scheme	5	3	2
[7]		Bureaucracy	3	3	0
		Challenges with pen permit acquisition	3	1	2
		Low educational level among farmers	2	2	0
		Insufficient info on funding programs	2	2	2
		Inadequate inspections for subsidy allocation	2	2	0
		No promotion of indigenous livestock breeds	2	2	0
		Inadequate support for young farmers	1	1	0
		Issues with social insurance	1	1	0
		High pension age	1	1	0
		Difficulty in legally hiring shepherds	1	0	1
Diseases [8]	9 (7.6%)	Livestock diseases	9	4	5

Table 3. Livestock farmers’ suggestions for improving the livestock farming sector (n = 118 farmers in total; 59 sheep/goat farmers and 59 cattle farmers). The results are presented as the number of reports (N reports) for each item for all farmers and separately for sheep/goat and cattle farmers. The results are also shown as the number and percentage of farmers reporting at least one item belonging to the specific category (Farmers’ N (%)). Numbers in brackets correspond to the challenge category (Table 2).

Suggestion Category	Farmers N (%)	Item	N Reports	Sheep/Goat	Cattle
Improve economic conditions [1]	74 (62.7%)	Increase subsidies	42	20	22
		Increase milk and meat prices/impose price floors	37	27	10
		Decrease feed prices	11	6	5
		Implement tax reductions	8	6	2
		Lower fuel prices	3	3	0
		Reduce rent costs for grazelands	1	0	1
		Reduce costs of social insurance	1	0	1
Improve policy implementation [7]	56 (47.5%)	Increase overall state support	16	12	4
		Strengthen inspections for subsidy allocation	12	6	6
		Promote establishment of farmers’ cooperatives	10	5	5
		Reduce bureaucracy	8	5	3
		Increase support for young farmers	7	3	4
		Improve the damage compensation system	4	0	4
		Promote organic livestock farming	4	2	2
		Promote technical seminars and establish farming schools in rural areas	3	2	1
		Enhance information dissemination to farmers	2	1	1
		Simplify the process of obtaining pen permits	2	0	2
		Improve livestock to increase productivity	1	1	0
		Decrease the pension age	1	0	1
		Promote indigenous livestock breeds	1	1	0
Improve infrastructure [3,5]	54 (45.8%)	Improve road conditions	33	19	14
		Upgrade pen infrastructure in communal mountain grazelands	28	21	7
		Improve water resource conditions: renovate/construct water points and reservoirs	21	5	16
		Electricity: connect with network or subsidize solar panels	13	8	5
Implement grazeland planning policy [4]	28 (23.7%)	Design and implement the planned grazeland management plans	27	7	20
		Improve grazeland quality	3	2	1
		Prohibit hunting activities in grazelands	1	0	1
		Use abandoned agriculture land for feed production	1	0	1
Improve services [6,8]	15 (12.7%)	Establish cheeseries	5	5	0
		Hire community vets	5	3	2
		Promote product distribution	4	3	1
		Establish slaughterhouses	3	3	0
		Set up butcheries	1	1	0
		Establish local milk concentration stations	1	1	0
		Construct loading and vaccination ramps	1	0	1
		Build a cereal silo	1	1	0
Manage predator problem [2]	13 (11.0%)	Fund prevention measures against predator attacks	9	2	7
		Implement carnivore removal	5	3	2
		Establish zoning of carnivores and livestock grazing areas	1	0	1

Table 3. Livestock farmers’ suggestions for improving the livestock farming sector (n = 118 farmers in total; 59 sheep/goat farmers and 59 cattle farmers). The results are presented as the number of reports (N reports) for each item for all farmers and separately for sheep/goat and cattle farmers. The results are also shown as the number and percentage of farmers reporting at least one item belonging to the specific category (Farmers’ N (%)). Numbers in brackets correspond to the challenge category (Table 2).

Suggestion Category	Farmers N (%)	Item	N Reports	Sheep/Goat	Cattle
Improve economic conditions [1]	74 (62.7%)	Increase subsidies	42	20	22
		Increase milk and meat prices/impose price floors	37	27	10
		Decrease feed prices	11	6	5
		Implement tax reductions	8	6	2
		Lower fuel prices	3	3	0
		Reduce rent costs for grazelands	1	0	1
		Reduce costs of social insurance	1	0	1
Improve policy implementation [7]	56 (47.5%)	Increase overall state support	16	12	4
		Strengthen inspections for subsidy allocation	12	6	6
		Promote establishment of farmers’ cooperatives	10	5	5
		Reduce bureaucracy	8	5	3
		Increase support for young farmers	7	3	4
		Improve the damage compensation system	4	0	4
		Promote organic livestock farming	4	2	2
		Promote technical seminars and establish farming schools in rural areas	3	2	1
		Enhance information dissemination to farmers	2	1	1
		Simplify the process of obtaining pen permits	2	0	2
		Improve livestock to increase productivity	1	1	0
		Decrease the pension age	1	0	1
		Promote indigenous livestock breeds	1	1	0
Improve infrastructure [3,5]	54 (45.8%)	Improve road conditions	33	19	14
		Upgrade pen infrastructure in communal mountain grazelands	28	21	7
		Improve water resource conditions: renovate/construct water points and reservoirs	21	5	16
		Electricity: connect with network or subsidize solar panels	13	8	5
Implement grazeland planning policy [4]	28 (23.7%)	Design and implement the planned grazeland management plans	27	7	20
		Improve grazeland quality	3	2	1
		Prohibit hunting activities in grazelands	1	0	1
		Use abandoned agriculture land for feed production	1	0	1
Improve services [6,8]	15 (12.7%)	Establish cheeseries	5	5	0
		Hire community vets	5	3	2
		Promote product distribution	4	3	1
		Establish slaughterhouses	3	3	0
		Set up butcheries	1	1	0
		Establish local milk concentration stations	1	1	0
		Construct loading and vaccination ramps	1	0	1
		Build a cereal silo	1	1	0
Manage predator problem [2]	13 (11.0%)	Fund prevention measures against predator attacks	9	2	7
		Implement carnivore removal	5	3	2
		Establish zoning of carnivores and livestock grazing areas	1	0	1

Figure 1. Reported challenges in the livestock profession (left) and suggestions to improve the profession (right), as reported by free-ranging livestock farmers living alongside wolves in NW Greece (n = 118 farmers in total; 59 sheep/goat farmers and 59 cattle farmers).

Figure 1. Reported challenges in the livestock profession (left) and suggestions to improve the profession (right), as reported by free-ranging livestock farmers living alongside wolves in NW Greece (n = 118 farmers in total; 59 sheep/goat farmers and 59 cattle farmers).

3.3. Drivers of the Perceived Conflict with Wolves

We identified four models within our confidence set (ΔAIC < 2;

Table 4. Most highly supported models (out of 31 models) relating the perception of wolves as a major problem to predictor variables. Shown are the Akaike Information Criterion corrected for a small sample size (AICc), the number of parameters (k), the difference in the AICc between a model and the model with the lowest AICc (ΔAICc), and the model weight (w_i). We only report models with a ΔAICc < 2 here.

Rank	Model	k	AICc	ΔAICc	wi
1	Preventive measures score + farming generation + seasonal movements + satisfaction with profession	4	145.4	0.00	0.39
2	Preventive measures score + farming generation + seasonal movements	3	145.8	0.42	0.31
3	Preventive measures score + farming generation + seasonal movements + % losses in LSUs	4	147.2	1.85	0.15
4	Preventive measures score + farming generation + seasonal movements + % losses in LSUs + satisfaction with profession	5	147.2	1.89	0.15

) to assess the perceived importance of wolves as a major issue. The preventive measures score, generation of livestock farming initiation, and seasonal movements were included in all four models, while the satisfaction with profession score and % losses in LSUs were included in two models. We averaged the coefficients across these models and identified significant predictors (

Table 5. Summary of model-averaged coefficients from models relating the perception of wolves as a major problem to predictor variables (average across models with ΔAIC < 2). Shown are the coefficients (β), standard errors (SE), 95% confidence intervals, p-values, and summed model weights (Σw_i). Predictors were considered significant if their 95% confidence intervals did not overlap with zero. Asterisks: significance level of p-values (*: p < 0.05, **: p < 0.01).

Variable	β	SE	95% CI		p-Value	Σwi
			Lower	Upper
(Intercept)	−1.23	0.58	−2.38	−0.08	0.035 *	-
Preventive measures score	−1.04	0.34	−1.72	−0.37	0.003 **	1.00
Farming generation—second generation (reference: first generation)	−1.68	0.77	−3.21	−0.14	0.032 *	1.00
Farming generation—many generations (reference: first generation)	0.58	0.59	−0.59	1.74	0.34
Seasonal movements—short-distance, <10 km (reference: stationary, 0 km)	0.29	0.61	−0.92	1.51	0.64	1.00
Seasonal movements—long-distance, >20 km (reference: stationary, 0 km)	1.63	0.64	0.38	2.89	0.011 *
Satisfaction with profession	−0.35	0.23	−0.79	0.09	0.11	0.53
% losses in LSUs	0.20	0.26	−0.32	0.72	0.45	0.30

).

Farmers who implemented more effective preventive measures (high preventive measures score) had a lower perception of wolves as a major problem (β = −1.04, CI = −1.72 to −0.37). Second-generation farmers expressed a reduced perception of wolves as a problem (second generation: β = −1.68, CI = −3.21 to −0.14). Upon closer examination, all second-generation farmers were cattle farmers, with 90% of them being ≤49 years old. Younger cattle farmers exhibited greater tolerance towards wolves (β = 0.52, p = 0.03). Conversely, age did not appear to similarly influence the attitudes of sheep/goat farmers.

Farmers who practiced long-distance seasonal herd movements (>20 km) perceived wolves as a major issue to a greater extent than those who moved shorter distances between seasons (long-distance movements: β = 1.63, CI = 0.38 to 2.89). There was a trend indicating that farmers with lower satisfaction with their profession perceived wolves more as a major problem, though the relationship was not statistically significant (β = −0.35, CI = −0.79 to 0.09). Notably, losses attributed to wolves did not significantly impact the perception of wolves as a problem (β = 0.20, CI = −0.32 to 0.72).

3.4. Job Satisfaction and the Role of Wolf Presence

Farmers expressed a low satisfaction level with their income (mean score of 2.4), with 59% of farmers reporting minimal or no satisfaction (Table 1). Satisfaction with the profession was moderate (mean score of 3.2), while 58% of farmers with children did not wish for them to pursue the same career. Satisfaction levels were slightly higher among cattle farmers compared to sheep/goat farmers.

Four factors were found to influence the overall job satisfaction levels of farmers in two models within our confidence set (ΔAIC < 2;

Table 6. Most highly supported models out of 15 models relating job satisfaction scores of livestock farmers to predictor variables. Shown are the Akaike Information Criterion corrected for a small sample size (AICc), the number of parameters (k), the difference in the AICc between a model and the model with the lowest AICc (ΔAICc), and the model weight (w_i). We only report models with a ΔAICc < 2 here.

Rank	Model	k	AICc	ΔAICc	wi
1	Economic marginalization + Wolves + Herd type	3	522.1	0.00	0.517
2	Economic marginalization + Wolves + Herd type + Rural depopulation and lack of services	4	522.2	0.14	0.483

). Both models included the variables of economic marginalization, wolves, and the herd type, while one model also included depopulation and a lack of services. By averaging the coefficients of these models, we identified significant predictors influencing job satisfaction (

Table 7. Summary of model-averaged coefficients from Ordinal Logistic Regression models relating the job satisfaction score of livestock farmers to predictor variables (average across models with a ΔAIC < 2). Shown are the coefficients (β), standard errors (SE), 95% confidence intervals, p-values, and summed model weights (Σw_i). Predictors were considered significant if their 95% confidence intervals did not overlap with zero. Asterisks: significance level of p-values (*: p < 0.05, **: p < 0.01).

Variable	β	SE	95% CI		p-Value	Σwi
			Lower	Upper
Wolves—yes	−1.09	0.36	−1.81	−0.39	0.003 **	1.00
Herd type—sheep/goat	−0.96	0.35	−1.64	−0.29	0.005 **	1.00
Economic marginalization—yes	−0.90	0.35	−1.59	−0.21	0.01 *	1.00
Rural depopulation and lack of services—yes	−0.60	0.39	−1.36	0.17	0.48	0.48

). The presence of wolves had the greatest negative impact on job satisfaction for farmers (β = −1.09, CI = −1.81.34 to −0.39), followed by the herd type, with sheep/goat farmers being less satisfied (β = −0.96, CI = −1.64 to −0.29), and economic marginalization (β = −0.90, CI = −1.59 to −0.21).

4. Discussion

4.1. Challenges in the Livestock Farming Sector

4.1.1. Economic Marginalization

Farmers reported that economic marginalization was the top professional challenge that needed addressing (Figure 1, Table 2 and Table 3), significantly decreasing their job satisfaction (Table 6 and Table 7). This trend aligns with general tendencies across Europe, where the uncertainty in meat and milk prices, commodity price volatility, low farm incomes, high dependency on subsidies, and uncertainty about future subsidy changes are major obstacles faced by farmers and render the sector unattractive to young farmers [38]. In our study, farmers identified low meat and milk prices as a primary economic challenge. This issue had an ever greater impact on sheep/goat farmers, who demonstrated lower job satisfaction rates, as their income is largely derived from milk production [39,53]. Sheep/goat farmers in the study area predominantly sold their milk to large private dairy companies at low prices instead of processing it themselves [54]. This situation is further exacerbated by the scarcity of farmers’ associations, which could otherwise help secure better sales prices and support local product processing and marketing. According to our results, farmers might be highly dependent on subsidies, as increasing subsidies was their top suggestion for the profession’s improvement (Table 3). Such a reliance on subsidies is common for both sheep/goat and cattle farmers in several European countries [38,55]. Feed costs were highlighted as another major factor impacting the finances of farmers in this study. In Greece, feed expenses account for a substantial portion of the total production costs of livestock farms, representing 44–48% for sheep/goat farms [37,39]. High feed costs in southern Europe are attributed to the limited availability of perennial pastures, inadequate farmers’ associations, and an imbalanced price transmission throughout the food chain [38,39].

4.1.2. Policy

Farmers emphasized the need to improve policies relevant to the livestock farming sector (Figure 1, Table 3). These policies are interconnected with the poor overall state support for the sector and the economic difficulties faced by farmers. They particularly underlined the need for the development of a grazeland planning policy through the adoption of grazing management plans, as required by Law 4351/2015 to meet Greece’s obligation to the European Union. These plans encompass various measures, such as the geospatial mapping of pastures, evaluation of pasture quality, determination of the livestock carrying capacity, and assessment and improvement of current infrastructure, including adaptation measures to climate change and water shortages (KYA/CMD 1058/71977). The management plans are not yet in place in Greece despite relevant national guidelines having been available since 2017 (KYA/CMD 1058/71977). It is evident that farmers in our study area recognize the need for compiling and implementing grazing management plans as an umbrella action that would address many challenges in the sector (Figure 1, Table 2). Among other beneficial provisions, these plans will deliver proposals on the rangeland units (specific area or permanent pasture) that each farmer will use for the next seven years based on the real number and type of livestock and the grazing capacity of the rangeland unit.

4.1.3. Infrastructure

Farmers also identified poor infrastructure as a major challenge in the sector. Poor road conditions leading to mountainous pastures was commonly referred to as a problem, contributing to increased travel costs and duration, potential harm to livestock [56], and ultimately reducing the farm profitability. Respondents did not suggest the extension of the current road network but rather for improvements in the road safety and quality. Therefore, we suggest improving the quality of existing mountainous road networks rather than extending the road length to enhance farming conditions in remote areas and wilderness areas, given the fragmentation impact of new roads on natural ecosystems of high ecological integrity [57,58]. Road infrastructure improvements should be accompanied by restriction measures (road closure devices) allowing access only to state authorities, as well as farmers during the grazing period [59].

Inadequate communal pen infrastructure was another main challenge. Farmers rely on multiple pens during seasonal movements, typically well-built pens during the winter and communal or makeshift/rustic pens during the summer. Communal pens, many of which were constructed several decades ago, suffer from inadequate maintenance by the state. Farmers are often hesitant to invest in repairs due to uncertainties surrounding grazing permissions and limited financial resources. Additionally, summer pens frequently lack electricity, which restricts the use of essential farming equipment and living amenities. Poorly maintained pens are also more accessible to wolves, leading to destructive depredation events [60] and increasing the vulnerability of livestock due to suboptimal health conditions. To address these issues, relevant authorities should prioritize investments in constructing new communal pens and renovating existing ones. Providing financial assistance to farmers for repair and construction costs would be beneficial. Ensuring access to electricity through infrastructure development or subsidized solar panels is also essential.

4.1.4. Social Marginalization

Farmers highlighted various aspects of social marginalization, including feelings of isolation and difficulties in finding a spouse. Farming communities in European mountainous areas face well-recognized social difficulties. The sector is often unattractive to younger generations and experiences low female participation, and the role of the farmer is often underappreciated by society [38,61]. Additionally, low levels of cooperation among farmers [38,54] exacerbate these challenges. This lack of collaboration not only worsens the economic conditions of farmers but also negatively impacts their social well-being, leaving them without a sense of belonging or community connections.

4.1.5. Job Satisfaction

The moderate level of professional satisfaction, combined with the tendency of many farmers to discourage their children from pursuing a career in livestock farming, underscores the difficulties inherent in the profession. Livestock farming involves demanding working conditions characterized by long hours, exposure to harsh weather, odors, dust, and physical hazards [62], making it physically strenuous [63]. These difficult working conditions are further compounded by a low net income, insufficient state support, and social marginalization, raising concerns about the long-term viability and attractiveness of the profession, as well as future food security [64].

Implementing systems that alleviate these burdens, such as automated equipment and improved facilities for handling livestock, could enhance job satisfaction among farmers [63] and help mitigate the risk of a decline in the sector. However, the effectiveness of such measures may be hindered by farmers’ low educational levels and lack of specialized training in livestock farming (Table 1), as reported by other studies in Greece [37,39] and across Europe [38]. Limited education contributes to the insufficient professionalization and modernization of farms, identified as major weaknesses in the sector [38]. Farmers in the study area largely relied on practical knowledge passed down through generations, with many becoming involved in family businesses at a young age. While this highlights the deeply rooted nature of traditional practices, it can also lead to resistance to adopting new methods and approaches necessary to address emerging economic challenges. We recommend enhancing education and training opportunities as essential strategies for equipping farmers to adapt to changing socioeconomic conditions [65,66].

4.2. Living Alongside Wolves

Our study revealed that the presence of wolves was a significant factor in reducing job satisfaction among livestock farmers, although it ranked low in their suggestions for sector improvement. Living alongside wolves presents complex challenges and various impacts for farmers. Wolves not only impose direct economic costs but also lead to various indirect expenses that can exceed the direct costs [67]. Farmers often need to invest additional time, finances, and effort to protect their herds from wolf attacks, which increases their workload and contributes to psychological stress [68,69,70]. Wolf attacks can also induce stress-related issues in livestock, including abortions, reduced food intake, decreased fertility, lower milk production [71], and reduced milk quality [72]. Additionally, wolves are known to attack and kill various types of dogs, including livestock guardian dogs [73], adding further economic and time burdens for farmers. Retroactive compensation systems typically reimburse farmers only for the direct costs of wolf attacks, overlooking the indirect costs mentioned above [71]. Funding for adequate preventive measures could greatly contribute to reducing livestock depredation, but such support has been insufficient in Greece [34].

4.2.1. Wolf-Related Farmer Characteristics

Permanent, skilled shepherds are essential for mitigating wolf attacks [34]. However, hiring such personnel poses a significant concern, especially for sheep/goat farmers, due to financial constraints and the shortage of skilled labor (Table 1). We recommend providing economic support for hiring shepherds through organizations such as the Hellenic Manpower Employment Organization (OAED), or through the Common Agricultural Policy (CAP). We also suggest enhancing shepherds’ skills through pastoral schools, which have gained popularity in other Mediterranean countries [74,75,76]. These schools often incorporate training on how to manage wolf predation as part of their curriculum [77]. Initiatives aimed at improving shepherding skills are also integrated into LIFE projects, such as LIFEstockProtect and LIFE ShepForBio.

The farmers in our study reported a high incidence of guardian dog poisoning (Table 1), which presents a crucial conservation issue [78,79]. The intentional poisoning of guardian dogs by hunters and other farmers highlights the complex dynamics and competing interests within agricultural and hunting communities (Figure S5, [79]). This detrimental practice not only undermines the protection of herds from large carnivore attacks but also threatens the preservation and re-establishment of traditional Greek guardian dog breeds [80]. Various projects in Greece include the operation of specially trained Anti-poison Dog Units and the distribution of anti-poison first aid kits to farmers (e.g., LIFE AMYBEAR, Egyptian Vulture New LIFE, LIFE ARCPROM). To combat this problem effectively, we strongly advocate for robust law enforcement and strict penalties for the illegal use of poisoned baits.

Farmers in our study area, particularly sheep/goat farmers, expressed dissatisfaction with the current national damage compensation system (Table 1), which is consistent with the situation in other European countries [81]. The main issues with ex-post compensation systems include administrative inefficiencies, difficulties in the timely detection of carcasses, difficulties in accurately determining the cause of livestock deaths, and concerns that these systems may encourage a passive mindset among farmers [82,83]. To improve the effectiveness of compensation schemes, it is widely recommended to revise these systems by linking compensation to preventive measures, incorporating outcome-based performance payments, and speeding up compensation payments [83]. Additionally, we highly encourage the integration of new technologies into compensation schemes, particularly in the Greek case.

4.2.2. Drivers of Perceived Conflict

Our findings indicate that farmers’ perceptions toward wolves were influenced by the use of preventive measures, the farming generation, and the type of seasonal movements. Farmers who did not consider wolves as a major challenge implemented stronger prevention measures, suggesting a better understanding of wolf behavior and a willingness to coexist by minimizing livestock predation. In the study region, livestock farming has historically been conducted in areas with a wolf presence, leading many farmers to adapt their management strategies accordingly. Research from other regions found that effective protection measures were associated with a reduced desire for the retaliatory killing of large carnivores [84], supporting the idea that improved protection strategies can help reduce conflict levels.

Our results indicated that younger cattle farmers exhibited greater tolerance towards wolves, whereas younger sheep/goat farmers did not. Cattle farming is a relatively recent practice in Greece, whereas sheep/goat farming has been a traditional activity for centuries. It is plausible that the opinions of younger individuals in the more traditional sheep/goat sector are shaped by the perspectives of elder men in their community. The mountainous villages of the Balkan regions often feature a strong age-based patriarchal system, where elder men occupy leadership roles and influence younger members of the community [15].

Farmers who engaged in long-distance seasonal movements (>20 km) tended to perceive wolves more as a major problem. These farmers reported higher losses compared to stationary farmers and those making short-distance movements (4.2 LSUs vs. 3.2 and 3.3 LSUs/weighted year, respectively), despite having similar preventive measure scores (p = 0.47, Χ² = 1.5). During interviews, they noted high rates of livestock diseases due to the rapid transport of livestock from low to high altitudes and inadequate summer pen infrastructure, which provided limited protection against harsh weather conditions and increased the livestock vulnerability. We attribute their lower tolerance toward wolves to this combination of factors.

Interestingly, farmers’ attitudes toward wolves were not significantly influenced by the extent of livestock losses, aligning with some studies [85] and contradicting others on large carnivores [18]. Even minimal losses can trigger strong reactions [7]. Farmers’ attitudes may be shaped by the magnitude of the economic loss relative to income from other activities [85,86] or the duration over which the losses occur [4,7]. Additionally, perceived indirect costs, such as psychological costs and a sense of fear, danger, or risk associated with the presence of large carnivores, play a significant role in shaping attitudes [4,7,87]. Personal beliefs, values, upbringing, education, traditions, and cultural factors also contribute to farmers’ perceptions of large carnivores [9,18].

5. Conclusions

Our study evaluated the challenges of the livestock sector in quantitative terms, relying exclusively on farmers’ opinions. It concluded with a prioritization of the measures that should be implemented to improve the livestock sector, as perceived by farmers. It became clear that while wolves do pose substantial challenges to livestock farmers and contribute to job dissatisfaction, they are not the sole or most critical issue in the sector. Combined challenges related to economic pressures, policy deficiencies, and inadequate infrastructure may have a more extensive impact on farmers’ livelihoods, particularly when coupled with the low level of specialized education among farmers. As no change in governmental policies has been noted since the collection of the interview data, the outputs of this study well reflect the current situation of the livestock sector. Only 4% of farmers suggested the removal of carnivores as a measure. We acknowledge that this proportion might be higher in other parts of Greece and Europe with recent wolf re-establishment in the absence of a long-term culture of human–large carnivore coexistence. Our findings align with a European-level study involving experts and other stakeholders [38], which identified economic aspects and insufficient knowledge and training as the most significant challenges in the farming sector, with wildlife conflicts being of lesser concern. Similarly, a participatory study including livestock farmers in a region inhabited by bears and wolves in Italy found that many of the farmers’ proposed actions extended beyond issues related to large carnivores and focused on agricultural policy, grazing regulations, rural development, and the promotion and marketing of local products [88]. Further socioeconomic research focusing on farmers’ opinions is recommended to cover different farmers’ profiles across Europe to inform policy measures for the sustainability and well-being of the sector.

To address the challenges in the farming sector, it is essential to enhance educational opportunities that improve farmers’ knowledge and skills for modernizing their practices. Additionally, supportive policies, such as increasing and stabilizing prices, improving infrastructure, providing financial assistance, and implementing comprehensive grazing management plans, are necessary for the sector’s long-term sustainability and attractiveness. It is important to acknowledge farmers’ concerns regarding wolves and implement effective strategies to mitigate livestock depredation, along with fair and efficient compensation systems. Enhancing farmers’ economic conditions will empower them to invest in preventive measures independently, reducing their reliance on external financial aid. Addressing broader socioeconomic challenges and policy deficiencies and supporting farmers in adapting to changing circumstances will ensure the livestock farming sector can thrive while minimizing the negative impacts associated with wolves.

The recent proposal to reduce the protection status of wolves in Europe, namely the European Union’s initiative to downlist wolves under the Bern Convention and the EU Habitats Directive [33], raises significant concerns. This proposal cites increasing socioeconomic challenges and harm to livestock as justifications. However, as noted by the Large Carnivore Initiative for Europe [89], there is no robust scientific evidence to substantiate a notable rise in livestock damages or public safety risks caused by wolves in recent years [90]. Such decisions [91] should be based on sound science rather than driven by political pressures [89,92]. Here, we showed that most socioeconomic challenges faced by farmers stem from other systemic issues, rather than wolves alone. Decisions on wolf management must be evidence-based to ensure effective policies that genuinely address the root causes of farmers’ difficulties. Downlisting wolves is unlikely to address the deeper socioeconomic challenges faced by farmers and risks undermining conservation success while further polarizing stakeholders [89]. Our work contributes to pinpointing the real challenges in the livestock sector and guiding evidence-based policy decisions.