Rural Tourism Opportunity Spectrum: Linking People and Landscape for Spatial Planning
e | 29 apr 2022
Pubblicato online: 29 apr 2022
Pagine: 101 - 110
Ricevuto: 09 mar 2022
Accettato: 29 apr 2022
DOI: https://doi.org/10.2478/mgrsd-2022-0006
Parole chiave
© 2022 Sylwia Kulczyk et al., published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.
In rural areas, especially where agriculture no longer supports the desired standard of living, tourism is considered an important element of multifunctional development (Przezbórska, Sznajder & Scrimgeour 2009, Ibănescu et al. 2018). At the same time, the popularity of outdoor recreation and leisure in nature is growing significantly, especially in highly developed societies (Winter et al. 2020). Seeking what they consider to be pristine environments, city dwellers spend their free time in rural areas. The traditional rural landscape – integrated systems of fields, meadows, and forests which reflect the diversity of abiotic components – is considered to be “wilderness” in Europe, and is highly appreciated for its natural character (Buijs, Pedroli & Luginbühl 2006). However, the increasing demand for contact with nature does not always meet the available supply. This leads to economic loses and does not fulfil the public’s expectations. Despite support from government strategies and programs (Koreleski 2007), the average occupancy of farm accommodation in Poland remains at only 14% (Central Statistical Office 2016). Similar problems have been identified in other countries, for example, Cyprus and Lithuania (Sharpley 2002; Baležentis et al. 2012). At the same time, the most popular destinations tend to be overcrowded. The negative environmental impact of tourism and outdoor recreation is reflected in many different ways (Gössling & Hall 2006, Huddart & Stott 2019). There is the need to work out a method to describe landscape potential (supply side of the tourism system) on the one hand, and to model tourist penetration of areas (demand side) on the other. This could help to manage tourism in rural areas in a more sustainable way.
An area’s suitability and actual use as a tourism resource depends on its landscape. According to Neef (1967), the concept of landscape embraces both its visual character and the material diversity of its abiotic and biotic components, including their spatial position. Landscape, understood as a holistic nature-cultural system (Naveh 2001), is crucial for tourism and recreation (van der Zee 1990; de Aranzabal, Schmitz & Pineda 2009). As noted by Kienast et al. (2012), the human experience of landscape has a dual character. It is driven both by the perception of the landscape’s physical features (referred to as ‘space’), and the specific meanings of cultural origin attached to them (often referred to as ‘place’). In this paper we focus on aspects concerning space. The patch-corridor-matrix model (Forman 1995) is commonly used to describe landscapes. It is similar to the tourism system model which consists of spatial (destination), linear (transit routes), and point (attractions) elements (Lew & McKercher 2006). Despite this similarity, landscape structure models are rarely used in tourism research (Gkoltsiou & Terkenli 2012).
The importance of landscape components for tourism potential has been identified and studied all over the world. Such components as land cover, relief, and the presence of surface water are widely recognized as key factors in defining an area as being suitable and attractive for tourism (Sołowiej 1986; Bastian et. al 2015). Altitude difference and slope inclination are the most common indicators of relief attractiveness (Gül, Örücü & Karaca 2006, Chhetri and Arrowsmith 2008). The land cover’s suitability for selected activities or its diversity are assessed (Burkhard, Kroll &Müller 2009; Paracchini et al. 2014). A popular water-based indicator for tourism attractiveness is distance from the nearest water body (Hammit, Patterson & Noe 1994; Woźniak et al, 2018). Despite an awareness of the complexity of landscape phenomena, tourism studies usually take into account only selected features, and the landscape mosaic is limited to spatial units.
Landscape spatial structure is also an important factor which influences tourist mobility patterns (Lew & McKercher 2006; Melicher & Špulerová 2022). Taking into consideration the visitors’ mobility is crucial for an understanding of their experience. Information on tourist flows also provides early warning signs of areas which are being used intensively and thus require management action. The deficiency in both the quantity and quality of the data on tourist mobility and preferences is still a serious impediment to assessing spatial patterns of tourism and recreation (Komossa, Wartmann & Verburg 2021). Popular tourist tracking methods have some shortcomings. Data collected from surveys and interviews can be difficult to interpret spatially (Kliskey 2000). Similarly, self-mapping methods are highly informative, but their effectiveness depends upon the skills and commitment of the respondents (Gimblett et al. 2003; Raymond & Brown 2007; Alessa, Kliskey & Brown 2008). Data from social media are both common and readily available, but might not reflect the entire visitor population (da Mota & Pickering 2020). Other promising methods for the determination of tourist flows are based on advanced technical equipment such as global positioning systems (GPS) (Beeco, Hallo & Brownlee 2014; Bauder 2015), cameras (Kammler & Schernewski 2004), and mobile phones (Ahas et al. 2008). However, while GPS and camera tracking systems are suitable for smaller (preferably closed) areas such as national parks or city centres (Shoval & Isaacson 2007; Kajala 2007), they are difficult to use in more open areas. Predicting tourist spatial mobility on the basis of a landscape structure analysis could be an attractive alternative to the above mentioned methods and would improve the efficiency of such studies.
Recognizing the spatial diversity of landscape potential and tourist mobility patterns could help minimize the gap between supply and demand in rural tourism. The recreation opportunity spectrum model (ROS) (Clark & Stankey 1979) combines tourists’ preferences with available resources to manage areas in a sustainable way. It provides a framework for the following:
establishing outdoor recreation management goals and objectives for specific management areas; trade-off analyses of available recreation opportunities, such as characteristic settings, would be changed by other proposed resource management actions; monitoring outputs in terms of established standards for experience and opportunity settings; providing specific management objectives and standards for project plans.
The ROS model, primarily implemented for forest areas, has proved its usefulness at various scales, and in various landscape types and management conditions; as well as for different forms of recreation and tourism activities (Boyd & Butler 1996; Joyce & Sutton 2009; Sæþórsdóttir 2010; Gomes, Figureido & Salvio 2021). Although it has primarily been implemented as a management framework, it is also used for suitability analysis (Weyland & Laterra 2014).
In its original form, ROS is based on the detailed analysis of users’ preferences. This imposes a need for field research (usually surveys, self-mapping, or tourist use monitoring), which makes this method most efficient at the local scale – in areas where users are relatively easy to control (e.g. protected areas). Although ROS has also been implemented at the continental level (Paracchini et al. 2014), the data used and results obtained have a very generalised character.
The goal of this paper is to present the rural tourism opportunity spectrum (RTOS) model. This model of landscape-tourism relation has its roots in the ROS concept but focuses on the multidimensional spatial analysis of landscape structure. RTOS can support environmental management, because it permits the following areas to be distinguished:
areas of high landscape potential and low use, where tourism could be developed; areas which are threatened by overdevelopment; areas where landscape potential is very low and rural tourism is unlikely to be a suitable means of development.
The study area covered 3,317 km2 and was located in northeastern Poland (Fig. 1) in the Lithuanian Lakelands, which forms part of Baltic Lakelands. This part of Poland is widely recognized as an area of great environmental, scenic, and cultural value, which is confirmed by the existence of two protected areas, Wigry National Park and Suwałki Landscape Park. The researched area is of glacial origin and its land cover reflects traditional rural use. It can be divided into two parts. The northern section has a varied relief composed of glacial moraines and numerous small lakes. It is morphologically diverse with a mix of meadows, fields, and deciduous forests. The southern section is dominated by a monotone fluvio-glacial plain covered by coniferous forest and a few larger lakes.
Figure 1
Study area
Source: own study
Paracchini et al. (2014), in their classification of European potential for outdoor recreation, recognized the study area as valuable yet difficult to reach for daily recreational tours. This is because big cities (e.g. Białystok, Warsaw) are far from this area. Therefore, in this study, we refer to tourism and not to daily recreation. We recognize the term
Przezbórska, Sznajder, and Scrimgeour (2009) noted that northeastern Poland offers a unique potential for rural tourism. This is reflected in the facilities: of the 157 localities which offer accommodation, farm stays constitute 84%. The most frequently-visited areas are Wigry National Park (category II under the IUCN classification of protected areas) and Suwałki Landscape Park (category V, IUCN classification). The natural value of these areas had already been recognized at the beginning the twentieth century when the first accommodation facilities were established beside Wigry Lake. The level of tourism development was initially relatively low, but since then the area has been commonly perceived as being attractive to tourism. On the other hand, the natural values deserved to be kept; however, formal protection was not established until the 1970s, with the national park existing since 1989. This resulted in a further increase in the number of tourist and accommodation developments in the vicinity of the protected areas. High natural value, which results in raising tourists’ interest, is a situation of potential conflict and deserves to be properly managed. The number of visitors is estimated at 100,000 (Wigry National Park 2016) and 50,000 (Suwałki Landscape Park 2016) per year. Tourist flow is seasonal, and the high season runs from May to September (Central Statistical Office of Poland 2016).
The area of the study fits within the mesoscale level. As it corresponds well to the spatial range of tourist activities, it is fundamental for studies on landscape potential (Bastian, Krönert & Lipský 2006). The evaluation was based on spatial and tourism data. The spatial data was obtained from the numerical terrain model (NTM) of the Shuttle Radar Topography Mission (SRTM) at a spatial resolution of 80 m, Landsat ETM (enhanced thematic mapper) satellite images from 28 August 2012 at a resolution of 30 m, and topographic maps at 1:25,000 scale, to which the Universal Transverse Mercator zone 34 (UTM 34) was applied. The tourism data was associated with this spatial data: accommodation lists from the Podlaskie Region Tourist Service, the farm stays catalogue from the Suwałki Chamber of Agriculture and Tourism’s website, and tourist attractions from the Podlaskie Voivodeship Information Portal.
The RTOS was organized in five steps (Fig. 2).
The attraction index described the tourism potential of the terrain as a function of the density of tourist attractions, combined with relief and land cover.
The view index reflected the visibility of the terrain and its features as a function of time; it is calculated separately for hikers and car travellers.
The access index was applied to areas which could be reached during a one-hour walk or a one-hour car trip from all accommodation facilities.
The tourism suitability index was calculated by multiplying the previous three results.
Rural tourism opportunity spectrum was described, and the management principles were defined.
Figure 2
Tourism evaluation method: the general approach
Source: own study
The reference unit for the attraction index was a 2,400 m × 2,400 m square. These dimensions were established by assessing two criteria. The first came from the fundamental landscape ecology principle of adjusting the spatial unit to the subject of the study (Wiens & Milne, 1989). Taking this into consideration, the size of the unit reflected the distance covered by a walker in a one-hour round-trip hike. The second criterion was introduced for technical reasons, namely to avoid further processing of raster data. This criterion was the pixel size of the satellite images and the NTM. The dimensions of the square were established in order that a whole pixel of the Landsat TM and NTM data fitted into the reference unit.
The attraction index was calculated as the sum of the spot attractions index and the relief and land cover attraction indices for each unit in the area: this was because all three factors contribute to the overall attraction of the terrain.
The spot attractions index reflects the density of places commonly known as sightseeing attractions. Tourists tend to perceive landscape as a collection of sites to be visited (Chhetri & Arrowsmith, 2008), thus their spatial arrangement form a frame, within which tourist activities are realized. The spot attractions index was based upon freely-accessible data taken from maps of tourist attractions found on the Podlaskie Voivodeship’s information portal: 101 objects were identified. Cultural (53) as well as natural (48) objects were included. The spot attractions were calculated as the number of specific tourist attractions in each reference unit. The spot attraction index (IS) was found by dividing the spot attraction values into six classes (from 0 to 5). The spot attraction index was the only attraction indices to have a minimum value of zero (in units where there were no attractions). This was due the spatial discontinuity of the factor. The areas where at least one spot attraction was present were grouped into classes: the maximum was five (in units where there were five or more attractions).
The diversity of relief and land cover is pointed out as being an important factor also from the use suitability and aesthetic points of view (Dramstad et al. 2006; Sarnowski, Podgórski & Brykała 2016). The relief attraction index was developed using the NTM, which was used to calculate the curvature of the terrain. The curvature map shows the convexity, concavity, and flatness of an area and produces values from around −3 to 3. Higher values indicate more convex terrain; lower values a more concave relief, while zero characterizes flat zones. In order to find the most diverse relief, the standard deviation of the curvature was calculated for each unit. The higher the standard deviation, the greater the divergence from the average, which in turn indicated a more varied and attractive relief. The relief attraction index (IR) was found by dividing the standard deviation of the curvature into five classes using the natural breaks method.
Land cover attraction (LC) was calculated on the basis of land cover tourism suitability and diversity. Suitability values were derived from former studies carried out in the Baltic Lakelands (Bartkowski 1975; Krzymowska-Kostrowicka 1980; Sołowiej 1992). To establish the land cover attraction index, the Landsat ETM image was sorted into nine land cover classes: urban areas, fields, meadows, swamps, lakes, bush land, and coniferous, deciduous and mixed forests. Next, the tourism utility of each class was determined. Useful land cover classes (lakes, meadows, pastures and all types of forests) were given a value of one. Urban areas, fields, swamps, and bush land were regarded as uninteresting and given a value of zero (any interesting urban objects were included in the spot attractions index). In the next step, the suitability of each reference unit was established by summing the number of suitable land cover patches.
A higher land cover attraction meant that the land cover was more varied and the land cover classes were interesting for tourists. Finally, the land cover attraction index (ILC) was calculated by dividing land cover attraction values into five classes using the natural break method. Low values (1) indicated an unattractive unit, and high values (5) indicated exceptional attractiveness.
The view from the road is the first contact a tourist has with the destination landscape. However limited in space and time, it is essential for the quality of human-landscape interaction (Appleyard, Lynch & Meyer 1963; Daniel 2001). The view index was calculated separately for hikers and tourists travelling by car, as they take different routes and move at different speeds. We assumed that a tourist travelling by car moves principally along paved roads at an average speed of 70 km/h, whereas a hiker also uses dirt roads or marked paths and moves at a speed of 5 km/h.
In the first step, the viewsheds were calculated from the NTM for each pixel of terrain, for each type of road, and for both types of tourists. Each pixel represented a fragment of the terrain, and the obtained map showed from how many points on the road the pixel could be seen. Taking into account the speed of travel and how often the pixel was viewed we calculated the duration of the observation for each pixel (OTP). Then, the terrain was divided into three groups: contemplative (observation duration over 8 seconds), snapshot (2–8 seconds observation duration), and invisible (observation duration under 2 seconds) (Kozaczko 2008).
However, tourists do not see ‘pixels’ in the terrain, they look rather at the land cover units. For this reason the observation time for each unit of land cover (OTLC) is more important than that of each pixel. This was, therefore calculated, by summing the observation times for all visible pixels in the land cover unit. On the basis of the values obtained, land cover units were reclassified as invisible (value 0), snapshot (1), or contemplative (2).
The access index shows the range of penetration into the terrain by hikers, and tourists travelling by car. Accessibility was defined as the penetration into an area for each type of tourist in one hour, starting from their accommodation. Areas were defined individually for each accommodation point. The reference area was assumed to be a flat, easily-penetrable terrain, and was defined as a circle with a 70 km radius for a tourist travelling by car and 5 km for a hiker. The analysis was not reduced only to areas adjacent to roads and trails for the following reasons:
the majority of space in Poland is open to the public; according to personal observation, tourist penetration of areas does not stick to the road network, which serves rather as an access network. Mushroom and berry picking, or simply wandering around could be given as examples of popular activities which are strongly spatial and not linearly oriented; the density of roads and trails in the research area was high (mean 1.9 km/km2), so even a short walk done by a car tourist during a travel break could ‘cover’ a whole range area.
We assumed that although car use expands the range of movement, it is a mode of transport rather than an activity in itself. For example, one can reach a forest by car (transport) and then walk through the area (leisure activity).
The first step was to establish the 5 km and 70 km limits for each accommodation point. Then, the cumulative cost of reaching each cell within the delimited zone, starting from the accommodation unit, was calculated on the basis of the slope model from the NTM. This made it possible to calculate the maximum reachable distance and to detect inaccessible areas. Tourist penetration was calculated on the basis of the area which was common to both the delimited zone and the maximum distance calculation. On the basis of these ranges, the availability of each part of the terrain was calculated as the number of accommodation points from which the area could be reached. Finally, the terrain was divided into four zones using the natural break method: inaccessible (value 0), difficult (1), average (2), and easily accessible (3).
The tourism suitability index was evaluated by multiplying the three previously described indices.
This was estimated based on the assumption that all three variables (attraction, view, and access) were equally important and must be positive for the area to be suitable for tourism purposes. Areas with at least one negative factor were considered to be either not suitable or out of range. The higher the attraction, view, and access indices, the more suitable the area for tourism. Next, the terrain was divided into the following suitability classes using the natural break method and excluding 0 value as being out of range class: very low, low, medium, high, and very high. As before, areas were assessed separately for hikers and for tourists travelling by car.
The overall calculation of the tourism suitability index was carried out for both types of tourists. The analysis was based on the 2,400 m × 2,400 m square units. Hiking and car tourism suitability indexes were identified and each unit was assigned a value for each type of tourism. Finally, the mean values were calculated to obtain a general overview.
Comparing the tourism suitability index (present state) and the attraction index (potential), the following classes of RTOS were identified:
Underdeveloped: areas with a very high attraction index, and medium or low suitability index (potential for development); Threatened by overdevelopment: areas with a very high attraction index and very high suitability index (potential need for protection); Incompatible use: areas with a very low attraction index and very low, or out of range suitability index (these could be a potential location for objects incompatible with tourism).
In order to evaluate the method (not for management purposes) we conducted a survey with tourists who visited Suwałki Landscape Park. The survey was carried out at the main tourist attractions, on seven concurrent days in July (peak tourist season). The tourists were asked about the following:
their place of stay; their main forms of transport; their average range of daily penetration; visited places; their opinion on the most valuable aspects of the area.
The 80 questionnaires were obtained and used to validate the model for two aspects. First, to check if the partial assumptions of the model which were derived from literature were valid for the area. We compared those assumptions to the tourists’ answers. Second, we evaluated whether the modelled tourist penetration corresponded to the areas actually visited by the tourists. The declared tourist itineraries were then mapped. Then we checked the overlap between the modelled tourism suitability index and the real area penetration. The evaluation of the modelled penetration for hikers was done for all tourists who declared that they moved only on foot or in a combined way: on foot and by bike, or on foot and by car (in this order). The evaluation of the modelled penetration for car tourists was done for all tourists who declared that they moved only by car or in a combined way: by car and on foot (in this order), or by car and by bike. The overlap of these areas was expressed as the percentage of the length of routes performed for each tourism suitability index class.
We assessed 575 units covering an area of 3,317 km2. These were divided into six classes, which reflected the internal diversity of the area.
For the attraction index, spot attractions were relatively scarce and unevenly distributed – 94% of units did not have any. The most significant group of spot attractions were cultural objects, identified within built-up areas. The relief attraction index classes had the following values: very low – standard deviation of curvature < 0.05, low 0.051–0.065, medium 0.0651–0.0815, high 0.0186–0.106, and very high > 0.106. The relief attraction index had a maximum value of 6.8% of the area, while 16.8% of the space was identified as unattractive. The land cover index classes had the following values: very low < 30, low 31–60, medium 61–90, high 91–120, and very high > 121. The land cover index showed that 1.6% of the area had the maximum attractiveness score, while 12.2% was unattractive. Relief and land cover attraction indices were not consistently found together for a large part of the area. The diverse relief found in the northern section is put to agricultural use, whereas the more attractive forest covers the flat southern section. This pattern highlights the importance of including both factors in the evaluation.
The relatively low relief diversity and dense road network resulted in high view indices. Hikers could observe 97.7% of the area and their low speed meant that the view index of the entire area was classified as contemplative. The view index for car tourists was a function of the density of paved roads, and 11.6% of the area was outside the visual range of motorized tourists. The motorized tourists’ high relative speed meant that 6.9% of the area was classified as snapshot, and 81.6% as contemplative. The low percentage of snapshot units indicates their lack of importance.
The access index classes were established using the following numbers of reachable accommodation places: inaccessible 0, difficult 1–4, average 5–9, and easily accessible > 9. The entire area was accessible by car, which resulted in a high access index. The access index for hikers was very diverse, with 15.7% of the space being out of range. Nevertheless, some of these areas had a high attraction index, and highlighted places which may be suitable for new accommodation infrastructure. The areas with the highest access index for hikers were those which had the most developed tourism infrastructure, which typically surrounded the popular holiday resorts of Augustów and Wigry lake (3.9% of the area).
The tourism suitability index had the following values: out of range 0, very low < 5, low 5–8, medium 9–12, high 13–16, and very high > 16. Figures 3 and 4 show the ranges and values of the tourism suitability index. Figure 5 highlights the index’s dependency on the different forms of tourism. When analysing the distribution of tourism suitability index classes for car tourist and hikers, a negative correlation is clearly seen. The tourism suitability index’s very high and high classes were very rare for hikers, while they were very common for car tourist. The opposite situation can be seen for very low and low classes. For car tourists it is much easier to reach more remote parts of the area which are out of range for hikers. Thus, facilities for them – car parks, look out points etc. – should be sparsely dispersed, whereas hiking trails would be more beneficial when concentrated within the most attractive and accessible parts of the terrain.
Figure 3
Tourism suitability index: hikers
Source: own study
Figure 4
Tourism suitability index: car tourists
Source: own study
Figure 5
Differences in the tourism suitability index between hikers and tourists travelling by car
Source: own study
Figure 6 shows the overall tourism suitability index for both hiking and car tourism, and highlights the rural tourism opportunity spectrum.
Potential new locations for accommodation infrastructure were identified by finding attractive areas with good visibility which lacked accommodation facilities. These occupied 1% of the area. This means that actual supply corresponded well to landscape potential, and that a lack of accommodation has a good correlation with the unattractiveness of the landscape.
Areas with high suitability and intensive use were also identified; these were the areas around Wigry National Park, Suwałki Landscape Park, and the Augustów Lakeland area. They occupy 1.2% of the area. The most intensively visited areas require detailed monitoring of tourist activities in order to properly manage them and, in turn, avoid any losses in their potential. The situation is especially fragile in the Wigry Lake area, where high suitability for tourism corresponds to high natural values and established protected areas. As mentioned above, the tourism development of this area has a longer history than the idea of nature protection, which makes the situation especially difficult to negotiate.
Areas which were identified as invisible or inaccessible and that have a low attraction index should be dedicated to activities which are incompatible with tourism. They occupied 10.6% of the area, mainly in the south-western part. These are agricultural areas without any outstanding natural features or cultural attractions. This results in a lack of accommodation facilities and suggests that pure agritourism without any additional attractions is unlikely to succeed. Such areas are considered not suitable for rural tourism development.
Figure 6
Overall tourism suitability index and rural tourism opportunity spectrum
Source: own study
The evaluation of the method’s assumptions demonstrated that, according to the survey, the most valuable aspects of the area highlighted by tourists were the view (40%), nature in general, and various nature elements: relief, water bodies, forests, animals, etc. (39%); and peace and quiet (21%). All tourists moved between specific spot attractions. The survey results showed that view was an important aspect for tourists. This justified the use of the view index as an input for the tourism suitability index. The high scores obtained for nature and its elements also prove that the variables used in developing the attraction index corresponded to tourists’ expectations. The access index was developed for car tourist and hikers, and, according to the survey, the main declared forms of transport were by car (46%), on foot (40%), and by bicycle (11%). The survey results showed that the average range of penetration was about 5 km for tourists travelling on foot and 70 km for tourists travelling by car, as was adopted in the model.
Table 1 shows the results of the comparison between the penetration of the area declared by tourists in the survey and the modelled tourists’ mobility. It is expressed in percentage of route length for the different classes of tourism suitability index. For hikers, the accordance was found to be very good. The 91.3% of route lengths done by tourists who travelled only on foot was located in high and very high tourism suitability index classes. Very good agreement between the model and the survey data was also achieved for tourists who travelled in combined ways, on foot and by bike, with 95% of route lengths found in high and very high tourism suitability indexes, mapped for hikers. But the tourists who declared a combined method of travelling on foot and by car should be rather considered car tourists. The tourism suitability index for car tourists was also mapped with very high accuracy: 78.2% for those who travelled only by car, 79.2% for those who travelled by car and on foot, and 80.1% for those who combined travelling by car and by bike. In all three transport groups the tourists moved mainly in areas of very high tourism.
The percentage of route lengths done by tourists who used different forms of transport in each class of tourism suitability index, mapped for hikers and car tourists
| only on foot | 0.6% | 0.0% | 0.6% | 7.6% | 80.3% | 11% |
| combined, on foot and by bike | 0.0% | 0.0% | 3.6% | 1.5% | 33.4% | 61.6% |
| combined, on foot and by car | 7.1% | 5.7% | 22.2% | 24% | 31.2% | 9.7% |
| Form of transport | Tourism suitability index mapped for car tourists | |||||
| out of range | very low value | low value | medium value | high value | very high value | |
| only by car | 2.0% | 0.1% | 9.4% | 10.3% | 29.5% | 48.7% |
| combined, by car and on foot | 3.5% | 0.1% | 7.8% | 9.5% | 31.7% | 47.4% |
| combined, by car and by bike | 4.5% | 0.0% | 1.9% | 13.6% | 27.2% | 52.8% |
The RTOS model presented here proved to be useful in assessing landscape for rural tourism at the regional and local level. It clearly describes the relationships between resources and their potential use. The method is suitable for tourists using different modes of transport (car, bicycle, hiking, etc.) and can reflect the specific characteristics of each group. The results obtained depend upon the form of tourism, which is included in the analysis. The model makes it possible to analyse the overall attractiveness of the terrain as well as its individual elements. It can be used to assess how viewable an area is, and identify visible and invisible units. The implementation of the rural tourism opportunity spectrum allows areas of special interest to be highlighted: underdeveloped areas, those threatened by overdevelopment, or areas suitable for activities incompatible with tourism. This helps prepare strategies for their management according to local needs, and to monitor changes.
The presented study has some limitations. The proposed method gives a holistic description of geographical features in the context of rural tourism, but it does not take into account social and economic factors. For complex rural tourism management, it should be combined with other tools. The results permit the comparison of different parts of an area and an estimation of their potential, but cannot be used as current, real data. Potential users, for example, local and regional authorities, should understand that GIS analysis is not an automated decision-making system but a support tool which enhances further enquiry.
The use of worldwide, freely-accessible data means that the method can be applied to other regions without significant modification. If a more detailed analysis is necessary, more precise data from other sources can be used. The method only describes the potential for basic forms of tourism without considering specific ones; for example, sailing, bird watching, or climbing. It can easily be modified by adding other aspects, for example, the length of accessible coastline for water-based tourism.
RTOS helps to detect potential problems quickly and accurately, and can therefore help to assess rural landscape suitability for tourism in a sustainable and efficient way.