Introduction
The perception of landscapes is determined by functional ties, roles and social practise of people. Within society, values and perspectives change over time and are subject to Zeitgeist (e.g. shift from a functional to a hedonistic image of landscape). Along with given natural capital, such as mountains, forests or coastal areas, also landscape practices defining agricultural land use structure and cropping pattern influences the amenity value and scenic beauty of the agricultural countryside. 'Landscape aesthetics' as well as 'recreation and health' represent two important ecosystem services and functions delivered by the landscape and how the region profits from the landscape. The scenic beauty and amenity value contribute to the provision of cultural ecosystem services. These services are of direct and indirect importance to human well-being as they represent a considerable economic resource by generating income, jobs and business opportunities in tourism and related business networks.
But better information on the economic importance and valuation of these services is needed. Also policy actions require information about the social demands towards landscape aesthetics. Valuation ( monetary, as well as non-monetary) of these services could enable policy makers to address trade-offs in a rational manner. In rural areas this is of great importance, because landscape is one of the most important public goods and its amenity as territorial asset contributes to the provision of second-order effects (like rural tourism and related businesses), regional development, economy and welfare.
The methods of economic valuation of non-marketed goods can assist to identify, which landscape characteristics foster the cultural function of agricultural landscapes. The most commonly method used is the stated preference method, also referred to as choice experiment method. Users of the landscape are directly asked about their preferences for visual landscape characteristics and hence the marginal value of discrete landscape characteristics can be estimated. Landscape attributes that appeared to be a dominant variable in identifying landscape preference in earlier studies using different methods are for example grazing animals, number of land types, number of patches and land type diversity, hedgerows and tree lines, water courses, woodland, and man-made attributes like farm buildings or cultural buildings.
It is unlikely that respondents are a homogeneous group, but rather a group of individuals with different, even contradictory landscape preferences. This heterogeneity may result from different socio-demographic backgrounds of individuals. Several studies indicate an influence of socio-economic, demographic or sociocultural characteristics on preferences such as age, gender, education or familiarity with the landscape.
Empirical Case Study Evidence
Preference studies were carried out in several case study areas (BG, DE, ES, NL, IT, PL) to explore the value of landscape characteristics. The aim was to evaluate user preferences of the managed landscape, aiming at the identification of inter-group differences and commonalities. The research revealed landscape preference differences and commonalities that are characteristic-related (respectively attribute-related) but also group-related landscape preference differences and commonalities.
Regarding the variations, preference differences for landscape attributes were found between farmers and inhabitants and visitors in the Polish and Italian case studies, but not in the German one. However, major methodological differences need to be taken into consideration when reflecting on the results. But even when applying a comparative research design and methodology, inter-regional preference differences occurred, e.g. between Winterswijk (NL) and the Märkische Schweiz (DE). Landscape characteristics were ranked differently and also the influence of the socio-cultural background differed. Commonalities were found for: (i) the low appreciation of natural landscape elements by some local residents and tourists (BG, IT), (ii) the simultaneous high evaluation of "grey", non-landscape elements, such as buildings and infrastructure (BG, PL), (iii) the decisive role of socio-economic characteristics of individuals for landscape preferences (DE, IT, NL, PL), as well as high appreciation of green, structuring landscape elements (DE, NL).
Two studies included a monetary approach to estimate the economic value and impact of landscape attractiveness due to changes in landscape management (ES, NL). Including an economic parameter in the studies had an impact on landscape preferences. The analysis revealed that the respondents' probability of visiting an area decreases with higher fees that could be taken as payment for cultural service provision (ES). And including financial trade-offs, the marginal preferences for some landscape attributes decreased (NL). In two other case studies (BG, FR), the importance of landscape for the provision of added value of a specific commodity (wine, meat) was confirmed by the producer, as it helps to increase recognition in the regional community and beyond.
Attribute-related landscape preference differences and commonalities
In the preference studies a diverse set of landscape characteristics (attributes) was included ranging from natural green landscape elements such as e.g. hedgerows to "grey" elements such as e.g. historical buildings. In three case study areas (BG, ES, PL) grey elements were included. In the Bulgarian case out of 10 landscape attributes mainly grey landscape elements and infrastructure (e.g. attractive buildings, gastronomy) were evaluated important for visitors. Natural landscape attributes (e.g. vineyards) showed only a low score. Also in the Polish case an especially high appreciation of grey landscape elements (architecture) was found among habitants and visitors. Also the presence of stone walls within olive groves was one of the most valued landscape elements, but got just second place behind the natural element 'green cover' (ES).
The Dutch and German case study areas (CSA) examined the preference for green/agricultural landscape elements. In both CSA a strong appreciation for high levels of landscape attribute occurrence was found. But the comparison of different types of attributes showed, that in the German case point elements (trees, groups of trees, pond vegetation) are most appreciated, ahead of linear element (hedges, tree lines). Presence of livestock as well as landscape diversity ranked lowest. In the Dutch case linear elements (tree lines) were the by far highest ranked landscape element, whereas point elements (group of trees) and land use diversity are ranked lowest. Other studies estimated the economic value of landscape attractiveness of olive orchards due to changes in the management of olive groves, i.e. presence of green cover, stonewalls and woodland islets (ES), but also of characteristics like water bodies and wetlands and related animals (IT).
Group-related landscape preference differences and commonalities
Some studies (DE, IT, NL, PL) included socio-cultural background information of respondents to investigate preference differences and commonalities among groups. In all these cases an influence of socio-economic characteristics of individuals on their landscape preferences was found. This was the case for different agent groups like e.g. farmers differing from other respondents ( PL) in their preferences for field, forest and small ponds.
Especially interesting appear groups of inhabitants and visitors. In the Italian case strong differences were found. Natural elements (e.g. bogs and wetlands) are perceived negatively by local residents, but attract different type of tourists (naturalists, bird-watchers), whereas sea-tourists do not attach specific values to the inner-side agricultural landscape. However, no significant differences between local residents and visitors could be found in the German and Polish cases. Habitants outside and inside the park and visitors of the park showed a similar preference pattern (PL). Being a resident or visitor did not influence the preference for any landscape attribute significantly (DE). Other characteristics that appeared to influence landscape preferences were especially gender (DE), education (DE, IT), age (DE, IT) familiarity with the region (IT) or urbanity and relation to agriculture (NL).
Conclusions & Recommendations
According to our research, natural as well as 'grey' landscape characteristics appeared very important for the amenity value and scenic beauty of the agricultural countryside. But no common preference pattern was found or could be formulated, because landscapes are differing a lot across Europe and each single region has specific characteristics that are perceived valuable by its users and visitors. Additionally, also the cultural imprinting of people that shapes their perception of landscape differs across Europe.
Hence, in each single case an evaluation of landscape preferences is needed to estimate the contribution of landscape to the provision of cultural ecosystem services and therefore also the contribution to regional development, economy and welfare through second-order effects.
The research on landscape preferences can contribute to the understanding of the multifunctionality of landscape. Considering visual qualities and including an aesthetic value perspective (next to environmental values) can help to improve multi-objective targeting of policies, supporting a diverse set of ecosystem services. Additionally, priority landscape measures and areas for the local landscape management from an aesthetic value perspective can be identified.
Policies have a strong impact on landscape structures. The Common Agricultural Policy (CAP) is one of the main drivers for agricultural production, landscape management and hence for landscape change affecting rural landscapes. Targeting of policies at maintaining and establishing the considered preferred landscape attributes could enhance the capacity to provide cultural ecosystem services and therefore support the rural economy and competitiveness.