Profile of the different user groups
The average age of respondents was 47 years, with local respondents having a higher average of 53 years (see Online Resource 2 for descriptive statistics). Our sample consists of slightly more females than males, especially for local inhabitants and experts. These results were expected for local inhabitants, as Castro Laboreiro inhabitants are mainly women (62%) and elderly, with most in the 65–69 years age category (Instituto Nacional de Estatistica 2012). The respondents generally had a high education level, especially among visitors. The locations of origin, as expected, differ per user group. The vast majority of local inhabitants (78%) originated from Castro Laboreiro. Less of the expert group originated from the region, although all were familiar with Castro Laboreiro and the Peneda-Gerês National Park. This group included varied occupations related to land management, such as local and regional policy makers (Ministry of Agriculture), employees related to the National Park and researchers from different institutions.
Visitors of Castro Laboreiro had a varied profile, with the majority being a domestic tourist (75%). Time spent in the area is equally divided between first-time visitors and those visiting 2–5 times and > 10 times (each 32% of our sample). Almost half the visitors came to Castro Laboreiro on a daytrip, followed by those visiting for multiple days but less than a week (32%), with the remaining visiting for a week or longer. Individuals reported multiple purposes for their visit, of which the main ones were sightseeing (60%), followed by recreation (38%) and visiting family or friends (24%). Two visitors specifically mentioned a visit to see the local dog breed (Cão de Castro Laboreiro). Twenty-nine respondents clarified their recreation purposes, with most focusing on walking/hiking (27 respondents), followed by other outdoor sports (13 respondents) and a nature focus (11 respondents).
The profile of the local inhabitants highlights the traditional agricultural origin of Castro Laboreiro, with high percentages of inhabitants having a connection with agriculture and finding human influence necessary in natural areas. Although only a limited number of respondents list farming as their primary profession, 73% of all local respondents owned fields and/or animals (e.g. cattle, sheep, goats). Agriculture is mainly focused on home consumption and the majority of land and animal owners gain less than half their income from agriculture. Fifty-five percent of local inhabitants report that they make use village communal lands for livestock grazing. More than half of the land owners (65%) mentioned that they have abandoned (part of) their land, due to retirement, reduced physical ability linked with age, the lack of subsidies and the limited financial payback.
Current situation and landscape preference
We analysed the current landscape preferences of the respondents by classifying the aspects of the Castro Laboreiro landscape that the respondents liked the most (Fig. 2). The majority of the respondents would take a photo of the small villages (heritage category), followed by the mountain landscape (physiography category) and the traditional agriculture (land use category). Traditional agriculture was mentioned significantly more often by local inhabitants than by the other user groups, when testing for all classes with an n above 10 (Fisher’s test p = 0.0001974, local—visitor). The view from the Castle (aesthetic category) and the mountain/rock landscape (physiography category) was mentioned most often by visitors, who mentioned significantly more aspects from aesthetic categories (Fisher’s test p = 0.003776). Oak forest (land cover category) was most often brought up by experts, as they mentioned significantly more aspects from the land cover categories than the other user groups (Fisher’s test p = 0.008502).
We also classified the aspects of the Castro Laboreiro landscapes that the respondents disliked the most. Here, the overall highest categories were modern houses disrespecting the traditional style (17%), abandoned villages (10%) and litter (10%), as well as signs of farming discontinuation (8%) such as stone walls falling apart or unkempt hedges. It should also be noted that more than 20 respondents said that they like everything in the area.
When comparing these preferences to the way that local inhabitants describe the “Castrejo way of being” to imaginary outsiders of the area, this shows a large overlap. Traditional agriculture is mentioned the most, often accompanied by an image of a women working in the fields wearing a traditional outfit, within a mountain setting. As land cover, oak forest was specifically mentioned, while reported heritage features included, e.g. pilgrimage places, the local dog breed and baking bread in a traditional way. When focusing on the local culture, the hard-working attitude, social cohesion and closed society, with some distrust to outsiders, were among the most mentioned aspects.
Opinions on dimensions of land abandonment
Perception of land abandonment
Abandonment in the area of Castro Laboreiro has clear visual effects on the local landscape, with respondents reporting the increase of shrubland and specific shrub species, decrease of domestic animals, increase of alien species and the visible collapsing of houses and other structures (e.g. stone walls). Several respondents report on the increase of wild animals, with a number of local inhabitants discussing their negative attitude towards the wolf (Iberian wolf, Canis lupus signatus), which they perceive as dangerous and a threat to the domestic animals.
The overall impact of the discontinuation of traditional farming is perceived negatively among all user groups, although there are some differences between the groups. Experts give a significant more positive rating, 2.8 on a 5-point Likert scale ranging from very negative to very positive, compared to both local inhabitants and visitors (2.0 and 2.1, respectively), as revealed through a post-hoc Dunn’s test (χ2 = 9.2865, d.f. = 2, p = 0.01). There is no discernible difference between local inhabitants and visitors. Regression results indicate that respondents from the expert group are more likely to be positive about abandonment. Also, more positive attitudes to abandonment are found with those that indicate that human influence is not a necessary aspect of a natural area, those that are not from Castro Laboreiro or surroundings and those that have a connection with agriculture. The last two are, however, not significant (for details see Online Resource 2).
The overall negative attitude towards land abandonment is reflected in the words or emotions that respondents linked to abandonment, which is most often associated with words as sadness, (the feeling of being) abandoned and left behind and emotions related to the loss of a former agricultural lifestyle (e.g. nostalgia, saudade—the latter is a Portuguese term for a deep nostalgic of melancholic feeling) (see Fig. 3). A 54-year-old local inhabitant for instance notes: “Sadness, because we can’t work them [the fields] and our kids had to leave because they didn’t have a future here”. A smaller number (12% of all respondents) reported mixed or more positive feelings, for instance stating: “On [the] one hand sadness and emptiness. On the other hand relief, because I associate the local population with the [forest] fires and with environmental problems” (local inhabitant, 51 years). During the interviews, multiple respondents stated that they believe that the traditional farming practices are problematic for fire risk, especially regarding the ignition of new fires, related to the burning of shrublands to enhance grazing potential.
Trade-off of different dimensions of land abandonment
Scores assigned to a selection of consequences and gains of land abandonment revealed that the loss of heritage is regarded as the most important negative consequence, followed by increased fire risk and the loss of income sources (Fig. 4). The second point relates to the fact that forest fires will be more threatening when extensive areas are under succession, although some respondents also linked traditional agricultural practices to increased forest fire risk (see section 3.3.1). The increased area for wild animals and the increase of oak forest are seen as the most important gains. Tourist potential is listed under both the positive and the negative consequences. As a 31-year-old expert commented: “Tourism is slowly starting, but not equal to the loss of agriculture”.
All user groups had remarkable uniform responses to the negative consequences, while for the positive consequences, the increase in oak forest is significantly higher when rated by the experts (χ2 = 12.146, d.f. = 2, p = 0). Local inhabitants clearly are more negative about the gains of abandonment and give a high value for “no gains”, but this effect was not significant.
Preference for abandoned landscape types
To assess the preference for the different trajectories of landscape development after abandonment, respondents ranked pictures from different types of abandoned landscapes in Castro Laboreiro based on attractiveness. Overall, oak forest was clearly the most preferred landscape future, followed by low shrubland and tall shrubland. Acacia encroachment was the least preferred, both for the valleys and for the plateau part of the landscape (valley χ2 = 183.8546, d.f. = 4, p = < 2.2 × 10−16, plateau χ2 = 178.1069, d.f. = 3, p = < 2.2 × 10−16). When evaluating the differences between the valley and the plateau, we find that for the five landscape options in the valley, plantations and oak forest were both preferred the most, with an average of 2.1 versus 2.2, respectively. A more detailed analysis shows that 42 respondents (35%) have oak forest as their first pick for both the valley and the plateau part of the landscape, while 30 (25%) have plantation as their first pick for the valley in combination with oak forest for the plateau.
Although oak forest is clearly the most dominant pick for the plateau, it currently only occurs in a limited number of locations, due to grazing and a limited seedbank. This is confirmed by some of the local respondents. A 58-year-old local inhabitant notes: “There is no oak in the plateau. So tall shrubs and giestas [broom] is what we can have and it is beautiful. [Also useful for] firewood”. On the other hand, oak forest is fast developing in the valley area, as a result of limited grazing and agriculture. As noted by a 57-year-old expert: “In the valley, [the] agriculture is marginal. Only hay to feed the cows is viable […] so the oak can easily regenerate in there”.
In terms of the user groups preferences, visitors scored the Acacia encroachment more positive than both the local inhabitants and the experts, with significant results for both the plateau and the valley (plateau χ2 = 5.5697, d.f. = 2, p = 0.01; valley χ2 = 12.369, d.f. = 2, p = 0). An opposite effect is visible for oak forest in the valley. Here, visitors scored the oak forest significantly lower than local inhabitants and experts (χ2 = 12.369, d.f. = 2, p = 0). These results are in line with our expectations, as both local and expert respondents are more familiar with the effect of Acacia encroachment (an invasive species) on the landscape, which quickly became a plague after initial planting.
We also collected the reasons for abandonment landscape preferences in an open question (Fig. 5). Aesthetic reasons were most commonly given, followed by reasons based on natural value, utility and being traditional/typical for the area. Overall, oak forest is mainly associated with a positive aesthetic, being a native and symbolic species for the area and having useful benefits. A specific category for the valley emerged on planning/human influence, referring to the structure and human origin of forest plantations. Plantations were perceived positively by the respondents, especially those indicating reasons for their preferences that were related to aesthetics, the organization of the plantations, the human origin and the financial benefit. A local respondent (58 years) for instance stated: “This is income. In the valley it can be a good option for the abandoned fields”.
Utility as positive feature of landscapes is significantly more mentioned by local inhabitants for the valley landscapes (Fisher’s test p = 0.01868, for local—visitor). For the plateau section of the landscape, both local respondents and experts mention usefulness as an important feature (Fisher’s test p = 0.0246). Here, experts mainly refer to the use of the low shrubs area for grazing. A focus on natural features of the landscape was most dominant in the expert group, followed by visitors and locals, with significant group differences for the valley (Fisher’s test p = 0.00436). Other classified features of the abandonment landscapes did not have significant group differences for the user groups.
Management preferences
To assess possible management directions of Castro Laboreiro, we assessed the preferred overall management direction. When asked about specific landscape management aimed at more farmed areas in the valley, local inhabitants clearly had the highest agreement, followed by visitor and experts (4.5 for locals, 4.0 for visitors and 2.9 for experts on a 5-point Likert scale). The group differences were significant (χ2 = 18.8017, d.f. = 2, p = 0). When the overall preferred management direction was assessed as a semantic differential question from “more traditional farming” to “more natural areas”, both visitors and locals indicated a slight preference for traditional farming (2.5 and 2.9 respectively on a 5-point scale), while experts had a slight preference for more natural areas (3.3 on a 5-point scale), although individual experts were divided on this question. The group differences were significant (χ2 = 6.2087, d.f. = 2, p = 0.04, visitor—local p = 0.0414). The correlation (Spearman’s rank) between respondent characteristics and the preferred management direction showed weak but significant relationships for several characteristics. Respondents who preferred a management focused on traditional agriculture often have either a higher age, a low income (< 500 €), are from Castro Laboreiro and/or have a connection with agriculture (see Online Resource 2 for more information).
Specific management practices noted as important by respondents included fire risk management (68%), active management for key plants and animals (56%), low intensity grazing (35%) and the development of the wilderness character of the area (32%). Several respondents also mentioned the need for better cleared hiking trails and roads. For specific preferences per user group, see Online Resource 2. Furthermore, the comments on the management showed distrust for the park authority, for instance in handling the fire risk. When discussing the suitable levels of responsibility, about half of the respondents indicated that land management is a multilevel issue with connected responsibilities for park authorities, local governance and the national level. Here, most respondents saw the national level as mainly having a financial role. Regional and European level responsibilities were indicated less (33 and 18%, respectively).