The Lycian civilization was a long-standing historical civilization that developed in the region known today as the “Teke Peninsula,” located in southwestern Anatolia along the Mediterranean coast (Figure 1). Situated within the mountainous and coastal areas between the present-day provinces of Antalya and Muğla, Lycia is recognized as one of the unique civilizations of the ancient world due to its distinctive political organization, language, architecture, and funerary culture (Bryce, 1986)(Keen, 1998).

From as early as the 2nd millennium BCE, the Lycians inhabited the region, and by the 1st century BCE, they had established a federative political structure known as the Lycian League, which is considered one of the earliest examples of democratic representation in the ancient world (Republic of Türkiye Ministry of Culture and Tourism, XXXX). Prominent cities of the League included Patara, Xanthos, Letoon, Myra, Tlos, Olympos, Phaselis, and Pinara, while smaller cities such as Aperlae, Kyaenai, Simena, Soura, Theimussa, and Limyra were also represented within the federal structure. The majority of these ancient cities are interconnected by the modern Lycian Way, transforming this historical network into a continuous corridor. See Figure 1 for the spatial distribution of these cities.

Xanthos, known as the capital of Lycia, and Letoon, its religious center, were inscribed on the UNESCO World Heritage List in 1988, marking the international recognition of the civilization’s cultural heritage (Türkiye, XXXX). While Xanthos held historical significance as the political capital of Lycia, Letoon served as the religious center of the Lycian League. Today, Xanthos is located near the village of Kınık in the Kaş district of Antalya Province, approximately 70 kilometers from the Kaş highway. The city's structures, shaped by various civilizations over time, reflect a fusion of Lycian traditions with Hellenistic and Roman architectural features.

Located approximately 4 km from Xanthos, the ancient city of Letoon held a special position as Lycia’s religious center, which represents a unique heritage value, both religiously and architecturally (UNESCO World Heritage Centre, n.d.). Additionally, the artistic production of Xanthos-Letoon had a direct influence on other Lycian cities and surrounding regions, including Patara, Pınara, and Myra (UNESCO World Heritage Centre, n.d.). The Parliament building in Patara, Lycian rock tombs, fortress remains, temple ruins reflecting the belief systems of the society, and urban remnants that offer insights into Lycian social life all contribute to this rich cultural landscape (Republic of Türkiye Ministry of Culture and Tourism, XXXX). Together, these remains reflect the cultural importance of the Lycian civilization and its legacy along the Lycian Way.

The Lycian Way is a cultural route approximately 540 kilometers long, extending across southwestern Türkiye from the Fethiye district of Muğla to the Geyikbayırı area of Antalya. Introduced in 1999 by Kate Clow and her team, the trail gained recognition within the international hiking community and has since become a popular tourism destination among both domestic and international nature enthusiasts (Clow, 2019).

The Lycian Way is situated within a complex cultural landscape that connects numerous ancient cities, mountain villages, forested areas, highlands, and coastal towns. While parts of the trail follow ancient roads and aqueducts, other sections intersect with modern asphalt roads. To aid navigation, the Cultural Routes Society carries out marking and maintenance work along the route.

In addition to trekking, various activities such as swimming, diving, nature and bird watching, climbing, paragliding, and cultural tourism through visits to ancient sites are also available along the route (Bastemur, 2011). Hikers are motivated not only by the desire to socialize but also by personal goals such as self-realization, the pursuit of meaningful experiences, and a desire to explore the world (Akgündüz & Kızılcalıoğlu, 2016). Although some hikers find the route physically demanding—due to steep slopes and sections that require endurance—it is widely regarded as rewarding on physical, emotional, and intellectual levels (Aydın, 2024).

Despite a generally high level of satisfaction, several infrastructure issues have been reported to affect the overall experience negatively. These include a lack of informational panels on flora and fauna, limited access to healthcare and internet, insufficient accommodation and food options, high prices, hygiene problems, environmental pollution, water shortages, and inadequate signage for orientation and guidance (Tütüncü et al., 2020). Within this context, most academic studies on the Lycian Way have focused on user experience.

The Lycian Way is considered an important destination within the scope of alternative tourism, sustainable tourism, and ecotourism (Şenel, 2019); (Kurar & Kavacık, 2022); (Bozhüyük Ardahanlıoğlu & Çınar, 2013). (Kurar & Kavacık, 2022) highlight the route's strengths, including its historical and cultural sites, natural heritage, rich vegetation, authentic accommodation options such as village guesthouses and stone houses, and favorable climate conditions. However, they also highlight structural problems that threaten sustainability, such as insufficient promotion and marketing, inadequate trail signage, limited access to food and water, underdeveloped accommodation infrastructure, lack of data, and shortage of skilled labor.

The authenticity of the region is threatened by unplanned development and marble quarrying, whereas environmental degradation is further exacerbated by illegal excavations, treasure hunting, pollution, and insufficient environmental awareness. These pressures are further intensified by weak legal enforcement, which undermines the protection of both cultural and natural heritage along the route. In this context, it has been suggested that the Lycian Way should be aligned with the standards defined by the European Institute of Cultural Routes, and steps should be taken toward its inclusion in international conservation programs such as the Cultural Routes of the Council of Europe (Bastemur, 2011); (Kurar & Kavacık, 2022).

(Bozhüyük Ardahanlıoğlu & Çınar, 2013), in their ecotourism-focused study on the Fethiye section of the Lycian Way, found that increasing construction in Ölüdeniz since the 2000s has placed pressure on natural areas. They also observed that tourism intensity in areas such as Kabak and the Butterfly Valley has led to dispersed settlement patterns, and in places like Kabak and Patara, tourism density poses threats to both ecosystems and coastal landscapes.

More recent spatial analyses, particularly in the Kaş district, show that between 2000 and 2020, built-up areas increased by 2.5 times, greenhouse areas tripled, and agricultural land was reduced by half (Sünbül & Tonyaloğlu Ersoy, 2021). These findings show a transformation of the natural landscape resulting from non-agricultural land use.

Tourism pressure on coastal areas is especially evident in cultural and natural heritage sites such as Korsan Bay, Ceneviz Bay, and Phaselis. Studies report serious threats in these areas, including pollution, exceeding carrying capacities, and coastal erosion (Erdoğan & Kuter, 2010); (Selçuk et al., 2020). Plans for the construction of concrete day-use facilities in Alacasu (Cennet) and Bostanlık Bays have been criticized by civil society organizations and professional chambers, pointing to shortcomings in the enforcement of conservation decisions (Derneği, 2023); (Bianet, 2023-06-06).

Even specially protected areas like Patara face increasing threats such as illegal construction, environmental pollution, and tourism pressure. Similarly, despite their conservation status, development pressures continue in natural protected areas such as Kaputaş and Fırnaz Bays (Tunçer & Atik, 2023).

3. Materials and Methods

This research adopts a multi-scalar geospatial approach to systematically evaluate long-term landscape transformations and environmental pressures along the Lycian Way corridor over a 26-year period. The overall methodological framework, integrating remote sensing analysis, spatial statistics, and policy review, is illustrated in Figure 2.

To monitor land-use and land-cover (LULC) dynamics over 26 years, Landsat Level-2 Surface Reflectance products with a 30 × 30 m spatial resolution were utilized for four reference years: 1999 (Landsat 7 ETM+), 2009 (Landsat 5 TM), 2019 (Landsat 8 OLI), and 2025 (Landsat 9 OLI-2). To ensure seasonal comparability and minimize phenological variations, all scenes were selected from the summer months (July–August) with cloud-free or low cloud cover. Multispectral composites were generated using six primary bands (Blue, Green, Red, Near-Infrared (NIR), SWIR1, and SWIR2) to maximize spectral separability for identifying major land-use categories.

LULC maps were produced using Supervised Classification in ArcGIS Pro, applying the Random Trees algorithm. As an implementation related to Random Forest, Random Trees allows multiple decision trees to be trained on randomly selected subsets, improving classification accuracy (Gislason et al., 2006). In this study, the classifier was configured with specific hyperparameters to ensure an optimal balance between model complexity and generalization: the maximum number of trees was set to 50, the maximum tree depth was limited to 30, and a maximum of 500 samples per class was utilized for training. Training samples were collected across the study area to represent the classes defined in the legend: Dense Forest, Low-Density Forest, Cultivated Area, Greenhouse, Urban Area, Low-Density Built Area, Bare Soil, and Water. Classification was performed for the full study area; additionally, a 1-km impact corridor around the Lycian Way was used for corridor-based analysis and interpretation of surrounding land-use patterns.

A 1-km impact corridor was defined around the Lycian Way to facilitate the visual interpretation of the spatial relationship between the trail and its surroundings. Empirically, existing practice demonstrates that buffer delineations are highly context-dependent rather than standardized. For example, the ICOMOS advisory body evaluation for the Routes of Santiago de Compostela notes that protection can rely on a “generic 30-m band” and that buffer zones are regulated independently by each Autonomous Community and may be contextual (ranging from 15 m to 100 m, depending on location and context) (Monuments & Sites, 2015). The Via Appia nomination file similarly reflects multi-scalar integrity-oriented approaches, with buffer-zone mapping presented at different scales, ranging from 100 m up to 4 km across sections (Cultura, 2023). In Chios (Greece), (Oikonomopoulou et al., 2017) report a 500-m corridor (each side) for fire protection and segment-specific protection bands for other purposes, including heritage conservation, varying from 70 to 450 m.

(Bambi et al., 2019) explicitly apply a one-kilometer buffer to support landscape and socio-economic characterization along the St. Francis’s Ways, noting its usefulness for assessing marginality and territorial conditions surrounding the route. Rather than implying a fixed, universally ‘correct’ buffer width for cultural routes, this study also treats the 1-km corridor as an analytical landscape setting within which route–territory interactions can be observed consistently. This framing aligns with the ICOMOS Charter on Cultural Routes (I.C.O.M.O.S., 2008), which stresses that the route’s territorial setting provides the framework and atmosphere necessary for understanding the Cultural Route and calls for the setting to be delineated through a regulated buffer zone where appropriate.

Within this broader evidence base, adopting a 1-km impact corridor offers a transparent and replicable compromise for landscape-scale GIS/remote-sensing analysis—wide enough to capture the surrounding territorial patterns that condition route experience and vulnerability, while remaining closely tied to the route corridor. Drawing on these precedents, the 1-km distance in this study is specifically determined to ensure that the 30-meter spatial resolution of satellite imagery yields representative descriptive results. This buffer functions as an analytical frame for understanding spatial dynamics and visual integrity rather than as a regulatory protection strip. Nevertheless, it is acknowledged that for a formal conservation management plan, buffer zones should be delineated differently, reflecting the specific historical and natural unique features of the route segment.

Accuracy assessment was performed only for the 2019 and 2025 classifications, as Sentinel-2 Views imagery (10 m resolution, ESRI Living Atlas service) was available for these years, providing an independent and higher-resolution reference compared to Landsat data. For 1999 and 2009, Sentinel-2 was not yet operational, and validating Landsat with Landsat would not constitute an independent assessment. Therefore, 220 stratified random points were generated for both 2019 and 2025, visually interpreted using Sentinel-2 Views imagery, and compared with the classified maps to produce confusion matrices. The 2019 classification achieved an overall accuracy of 87.2% with a KAPPA (Cohen, 1960) coefficient of 0.84, while the 2025 classification achieved 90% and 0.88. According to the widely accepted benchmarks by (Landis & Koch, 1977), Kappa values between 0.61–0.80 represent 'substantial' agreement, while values between 0.81–1.00 indicate 'almost perfect' agreement. In this study, the Kappa coefficients of 0.84 for 2019 and 0.88 for 2025 fall into the 'almost perfect' category, confirming the high statistical reliability and robustness of the classification results.

Change analysis was performed through categorical change detection to identify transitions between land-use classes over the study period, focusing particularly on transitions from Natural Areas to Urban Areas and Greenhouses, highlighting the expansion of built-up land uses and their encroachment into natural and agricultural landscapes. To minimize spectral noise and enhance the spatial coherence of the classified outputs, a 3 × 3 median filter was applied. The analysis was conducted in two sequential stages:

In the next step, hot spot analysis was conducted on the overall land-use change map (1999–2025) to identify the spatial clustering patterns of land-use conversions across the study area. In the hot spot analysis, greenhouse-related conversions and urban-area conversions were treated as aggregated change classes and analyzed to identify statistically significant clustering patterns. Global spatial autocorrelation was assessed using Global Moran’s I to test whether conversion intensity is randomly distributed; after confirming statistically significant positive clustering, Getis–Ord Gi* hot spot statistics were computed to locate and map local clusters (Mitchell, 2005); (Getis & Ord, 1992). To improve interpretability and to better capture regional effects and fringe patterns, the study area was partitioned into 500 m × 500 m (25 ha) grid cells. This resolution was selected to reduce the influence of sparse, zero-dominated cells (thereby strengthening statistical robustness) while providing a meaningful spatial unit for interpreting landscape-ecological functions. Spatial relationships were defined using a k-nearest neighbors (KNN) scheme (k = 8) with row standardization. For greenhouse-conversion intensity, Global Moran’s I indicated strong, statistically significant positive spatial autocorrelation (Moran’s I = 0.423031; z = 184.812; p ≪ 0.001; expected I = −0.000021); similarly, urban conversion intensity exhibited significant clustering (Moran’s I = 0.455162; z = 198.648702; p ≪ 0.001; expected I = −0.000021). The high z-scores and extremely low p-values indicate that the spatial distribution of land-use transitions is not a result of random chance, but instead confirms a strong, non-random clustering pattern with geographically concentrated high-intensity conversions (Mitchell, 2005); (Getis & Ord, 1992).

Local clustering was then evaluated using Getis–Ord Gi* under the same neighborhood definition (KNN, k = 8) on 500-m grid–aggregated conversion intensities. Gi* computes a z-score (and its associated p-value) for each cell, and ArcGIS summarizes statistical significance using confidence-level (significance) bins; these 90%–95%–99% confidence classes correspond to p-value thresholds derived from the Gi* z-scores under the standard normal distribution (Mitchell, 2005). Separate Gi* outputs were produced for urban area and greenhouse conversions and were overlaid in a single map to compare hot-spot patterns, while simultaneously displaying the Lycian Way polyline (and analytical buffer) to visually emphasize locations where statistically significant conversion clusters intersect the cultural-route environment.

Finally, in addition to the GIS-based change detection, an interpretive policy review was conducted for the major change nodes/hotspots. This review drew on the 1/100,000 scale regional master plans for the Aydın–Muğla–Denizli and Antalya–Burdur–Isparta planning regions, which were used as analytical base maps to delineate settlements, protected areas, and cultural/natural heritage designations along the Lycian Way corridor. For the nodes, the observed LULC transitions were cross-referenced with the relevant national protection designations and associated planning instruments in order to contextualize spatial pressures within the prevailing legislative and administrative framework. This step provides a structured basis for assessing whether the current protection regime is commensurate with the intensity and character of landscape transformation observed along the route corridor.

The analytical approach employed in this study involves several inherent limitations and boundary conditions that should be acknowledged. In the northern mountainous regions of the study area, classification accuracy was limited due to reflectance heterogeneity associated with bare soils, rock outcrops, and seasonal snow cover. However, as these areas lie outside the Lycian Way corridor, they were not considered in the interpretation of results. The 30-m spatial resolution of Landsat data limits the ability to capture small-scale land-use features but remains suitable for assessing broader regional-scale transformations over time. Using multiple Landsat sensors across the study period may introduce inter-sensor differences that affect class separability and temporal comparability. All scenes were selected from summer months to improve seasonal comparability; consequently, the analysis does not capture intra-annual variability, although detailed vegetation dynamics were not considered necessary given the study’s primary focus on built-up change.

Supervised classification outcomes also depend on the representativeness of training samples and class definitions, which may introduce uncertainty—particularly where land-use classes exhibit spectral similarity; to mitigate this, potentially controversial areas were cross-checked using contemporaneous high-resolution Google Earth imagery. Finally, the 3×3 median filter improves spatial coherence but may smooth small patches and narrow features, potentially under-representing fine-grained change. Because this choice was intended to tolerate resolution-related noise, the resulting spatial data provide a conservative estimate of landscape transformation, as small-scale and isolated anthropogenic footprints may have been removed alongside spectral noise. Accordingly, the results are discussed with a focus on key nodes. Findings are also conditioned by the chosen analytical corridor (1-km impact corridor), and alternative buffer widths could yield different estimates of surrounding land-use patterns.

The supervised classification results obtained through the Random Trees algorithm illustrate the spatial distribution of land-use categories along the Lycian Way for the years 1999, 2009, 2019, and 2025. The Mediterranean vegetation cover, dominated by maquis shrublands, required a distinction between dense forest (closed-canopy, non-maquis woodlands) and low-density forest (more open, maquis-dominated areas) to improve the training sample accuracy.

Although the 30 × 30 m resolution of Landsat imagery limits the detection of small-scale rural settlements, the classification adequately captures the major categories relevant to regional change dynamics, including forest, cultivated areas, greenhouses, and urban expansion. The 1-km impact corridor around the Lycian Way emphasizes the interaction between the trail corridor and surrounding land uses. As observed, different sections of the Lycian Way intersect with diverse land-use mosaics, ranging from dense forests to intensified greenhouse and urban development along the coastal plains.

The change detection analysis was conducted for three consecutive periods: 1999–2009, 2009–2019, and 2019–2025 (Figure 3). The results highlight the timing and intensity of land-use transitions from natural environments to built environments, specifically urban areas and greenhouses.

The period-specific change detection maps (Figure 3) reveal spatially concentrated transitions from natural to built environments and delineate several key transformation nodes along the Lycian Way corridor. Primary nodes are highlighted on the overview maps, and these nodes are shown as enlarged inset panels for closer inspection in the following figures (Figure 4, Figure 5, Figure 6).

Period of 1999–2009

Regarding the primary nodes of change during the 1999–2009 period, the area encompassing Gelemiş and Kalkan Figure 4a exhibits distinct transition patterns. For descriptive purposes, the cluster comprising the neighborhoods of Gelemiş, Kınık, Ova, and Yeşilköy is hereafter referred to as the Gelemiş area. Within this area, and particularly across the existing agricultural and greenhouse zones, transitions from Low-Density Forest and Cultivated Area to Greenhouse are dominant. Furthermore, the conversion of Low-Density Forest and Bare Soil into Urban Area is observed, particularly toward the coastal sections. Simultaneously, in Kalkan—a coastal tourist settlement—the transformation from Low-Density Forest and Bare Soil into an Urban Area is prominently observed.

In the Demre area Figure 4b, transitions are characterized by the intensive conversion of Cultivated Area into Greenhouse. Additionally, Urban Area transformations from Low Dense Forest, Cultivated Area, and Bare Soil are evident. Within Kumluca Figure 4c, the Cultivated Area to Greenhouse transformation is the primary trend, while the Cultivated Area to Urban Area transformation is also significantly observed. It is noteworthy that the spatial extent of change in this area is distributed across a wide territory.

In the area covering Çamyuva and Kemer Figure 4d, the conversion of Dense Forest, Low Dense Forest, and Cultivated Area into Urban Area constitutes the most frequent transition type. In the Göynük area Figure 4e, Urban Area transformation remains the most prominent feature, with the conversion of Low Dense Forest into Urban Area acting as the primary driver.

Period of 2009–2019

Between 2009 and 2019, while the established trends in Kumluca, Çamyuva, and Göynük persisted, the spatial patterns of change diversified and intensified in specific focal points.

In the Gelemiş area, which encompasses the Kınık, Ova, Yeşilköy, and Gelemiş neighborhoods Figure 5a, Greenhouse conversion accelerated during this decade. Specifically, along the Göl Street corridor in Ova Neighborhood, the transition of Cultivated Area and certain Low-Density Forest patches into Greenhouse is clearly visible. In the region encompassing Kalkan and the rural settlements of Bezirgan and Sarıbelen Figure 5b, a marked transformation into an urban area is observed. This transition is particularly significant in Bezirgan, while Sarıbelen also exhibits transitions into the greenhouse.

In the area containing the Kaş city center Figure 5c—a coastal touristic town—spatial transformation follows a fragmented pattern, with Bare Soil to Urban Area transitions occurring across various parts of the district. An extensive conversion to Greenhouse is observed in the area covering the villages of Kasaba and Dirgenler Figure 5d. Significant transitions from Low-Density Forest, Bare Soil, and Cultivated Area to Greenhouse are evident. Although this area is not directly intersected by the Lycian Way route, it is crucial for demonstrating the general characteristic of greenhouse expansion in the region. In Demre Figure 5e, the conversion from Low-Density Forest, Cultivated Area, and Bare Soil into Greenhouse is concentrated northwest of the Beymelek Lagoon, effectively encircling the water resource with greenhouse areas.

Period of 2019–2025

The analysis for the 2019–2025 period indicates that Greenhouse expansion in the Gelemiş area has continued throughout the district. In Kalkan and its surroundings Figure 6a, the expansion of the Urban Area persists at the city fringes, characterized primarily by Low-Density Forest to Urban Area transformations.

In the villages of Kasaba and Dirgenler Figure 6b, the Greenhouse transition trend has spread across the entire area, representing the persistence of intensive greenhouse cultivation. In the Demre area Figure 6c, the trend of conversion to Greenhouse remains focused around the Beymelek Lagoon, primarily involving Cultivated Area and Bare Soil, while fragmented transitions continue across the wider Demre district.

In the Finike and Kumluca areas Figure 6d, fragmented Cultivated Area to Greenhouse transitions are observed throughout the region. Notably, in Finike, Urban Area expansion from Low Dense Forest is concentrated and intensified within coastal zones. Finally, in the Çamyuva and Kemer areas Figure 6e, Cultivated Area to Urban Area and Low-Density Forest to Urban Area transitions are observed in a scattered manner, with an increasing trend towards the urban peripheries.

Particularly in Demre, Gelemiş, and Kumluca, the expansion of greenhouses has significantly contributed to the total increase in the built areas. Across the plains, the spread of greenhouse-based agriculture within croplands is evident in the satellite classifications. This finding is consistent with national statistics: Türkiye’s total greenhouse area expanded from 536,000 decares in 2002 to 811,000 decares in 2022, representing a 51% increase. Antalya alone accounts for the largest share nationwide, with 308,427 decares of greenhouse cultivation (Türkiye et al., 2023).

These findings are also consistent with previous research in Kaş district, which documented a 2.5-fold increase in built-up areas, a tripling of greenhouse cultivation, and a significant reduction of agricultural lands between 2000 and 2020 (Sünbül & Tonyaloğlu Ersoy, 2021). The conducted land use/land cover change analysis likewise demonstrates marked increases in greenhouse areas and substantial decreases in cultivated lands between 1999 and 2025. The concentration of greenhouse and urban transformations along particular segments of the Lycian Way corridor indicates that regional-scale dynamics are strongly manifested at the local scale.

The cumulative land-use change map for the period 1999–2025 provides an integrated perspective on the long-term transitions from natural environments to built environments along the Lycian Way corridor. In particular, the cumulative land-use change analysis within the 1-km impact corridor of the route (Table 1) reveals a spatially concentrated pattern of landscape transformation. While transitions to Urban Areas account for 58.9% (417 ha) of the total transitions, transitions to Greenhouses comprise 41.1% (290 ha) in total. These transitions show that urban development emerges as the primary driver of landscape transformation, largely fueled by the encroachment of built-up areas into natural maquis cover and existing croplands. This trend suggests a progressive fragmentation of the Mediterranean landscape, particularly where residential and tourism-related infrastructure displaces low-density forested areas. Simultaneously, the transition from traditional cultivation to greenhouse-based production reflects a shift toward industrial-scale agriculture in the surrounding area of the route. Although the aggregate footprint of change constitutes a small fraction of the total corridor area, its high spatial density at specific nodes represents a significant threat to the route’s visual integrity and traditional rural character. While the total converted area of 707.4 ha accounts for only 0.7% of the 96,300 ha 1-km impact corridor, its high spatial concentration at critical nodes signifies irreversible anthropogenic impact within the trail’s immediate environment. Since these figures represent a conservative estimate due to noise-reduction filtering, they effectively underscore the most established and intense zones of localized pressure.

To examine the pressure of regional change trends on the Lycian Way in greater detail, a hot spot analysis was conducted to identify spatial clustering patterns of Greenhouse and Urban Area transformations. These clusters were evaluated in relation to the Lycian Way route and its 1-km impact corridor. As illustrated in Figure 7, Greenhouse transitions are concentrated in Gelemiş, Demre, Finike, and Kumluca (indicated by green hatched areas), whereas Urban Area transitions cluster in Gelemiş, Kalkan, Finike, the northern part of Kumluca, Çamyuva, Kemer, and Göynük (shown in pink). Current land use in these areas was cross-checked using high-resolution Google Earth imagery to evaluate classification accuracy. The results confirm that the analysis is generally consistent; however, the cluster identified in northern Kumluca was determined to be invalid. In this specific area, remote sensing tools were unable to accurately distinguish the highly fragmented distribution of agriculture, greenhouses, and residential areas, leading to misleading clustering results.

The statistical magnitude of these hot spots underscores a profound and systemic transformation of the regional landscape. With transition to urban area hot spots covering approximately 872.0 km² (87,200 ha)—representing approximately 7.36% of the total study area—and transition to greenhouse clusters totaling 645.5 km² (64,550 ha), it is evident that built-up expansion and industrial agriculture exert parallel and massive pressures on the territory. The high concentration of these transformations within 99% confidence cores—nearly 50,000 hectares each (492.5 km² for greenhouses and ~547.0 km² for urban areas)—indicates that these changes are not isolated or random events but are instead highly localized, intense, and persistent spatial shifts.

For a cultural heritage route like the Lycian Way, these figures represent more than just land-use data; they signify a narrowing window of preservation. The fact that nearly 1,500 km² (150,000 ha) of the region is now identified as a statistically significant zone of rapid anthropogenic change highlights the urgent need for integrated landscape management. These expansive ‘hot cores’ effectively function as encroachment zones that threaten to diminish the trail's environmental buffer and permanently alter its traditional rural and natural character.

These patterns reflect the dual intensification of agricultural modernization and urban development along the cultural route. In contrast to the period-specific analyses, this cumulative representation captures the aggregated spatial consequences of anthropogenic pressures accumulated over more than two decades. The findings underscore two dominant trajectories: the continuous expansion of greenhouse agriculture across fertile coastal plains and the progressive urban sprawl concentrated around touristic and residential centers.

These site-specific transformations directly affect the quality and character of the natural and cultural landscapes through which the route passes. Many of the corridor’s touristic settlements are characterized by a locally distinctive spatial pattern that is historically integrated with surrounding nature, producing a landscape identity grounded in the close interweaving of built form and Mediterranean ecological features. At the same time, the areas surrounding these tourism nodes are frequently designated for protection due to their environmental significance and sensitive natural attributes. The Lycian Way’s embeddedness within Mediterranean landscapes constitutes one of its key defining characteristics; accordingly, degradation in adjacent natural areas—and the environmental changes associated with the expansion of greenhouses—poses a direct challenge to the route’s landscape integrity and experiential legibility.

Comparable dynamics have been documented in site-specific remote sensing studies of heritage landscapes, where development pressures translate into measurable land-cover conversions around protected cultural assets. While prior work often foregrounds tourism- or settlement-led built-up expansion, our corridor-scale results point to a coupled trajectory in which urban growth co-evolves with greenhouse-based agricultural intensification. For instance, (Butt et al., 2025) report a marked increase in built-up area around the UNESCO World Heritage Site of Taxila between 1990 and 2024, alongside substantial conversions of cropland and vegetation to built-up and bare land—changes discussed in relation to risks for site integrity and documented concerns about urban development and encroachment in UNESCO reporting.

In a similar vein, analyses in cultural-heritage contexts in China emphasize that built-up expansion is commonly associated with declining ecological land and increased landscape fragmentation, and they argue for integrating ecological considerations into urban planning while also examining how policy-driven mechanisms shape land-use trajectories (Cao & Li, 2023). Conceptually, this aligns with planning-oriented discussions that frame cultural landscape preservation as increasingly intertwined with urban planning and sprawl, underscoring that ‘coastalization’-type growth dynamics can generate structural pressures that are not unique to the Lycian context but are clearly visible in its corridor-scale pattern of change (Imene, 2021).

Antalya’s favorable climatic conditions provide a comparative advantage for greenhouse-based agricultural production, accelerating the transformation of fertile coastal lowlands into highly commercialized landscapes. Collectively, these dynamics highlight the dual role of commercial agriculture and urbanization as the primary drivers of landscape transformation along the Lycian Way corridor. In this context, the management of cultural routes requires cross-sectoral coordination to integrate policies on conservation, preservation, regional organization, sustainable development, land use, and tourism. Therefore, joint projects should be developed at national (provincial, regional, and local) and international levels to promote sustainable development and to establish management tools designed to safeguard the route against natural hazards and any risks that may threaten its integrity, authenticity, and overall significance (I.C.O.M.O.S., 2008).

In Türkiye, legally protected areas are defined as zones restricted in terms of usage to conserve natural, ecological, and cultural values. Within this scope, it is observed that the natural, historical, and built environment is protected under various statuses through different classifications, primarily including Special Environmental Protection Areas (SEPA). These areas are defined as zones with high environmental sensitivity, approached through an ecosystem-based perspective. The SEPA status restricts any activity that could disrupt the natural environment, biodiversity, and landscape character by enforcing planning decisions that prioritize conservation. Beyond this, protected areas are defined through various classifications such as national parks, conservation areas, archaeological conservation areas, wetlands, and wildlife conservation areas. These classifications are defined under various legislative frameworks (Türkiye, 1983); (Türkiye, 1983)(Türkiye, 1983)(Türkiye, 1985).

In statutory regulations and urban plans prepared at different scales, approaches similar to the protected area classification defined by the (International Union for Conservation of Nature, 2008) are gradually applied. In this context, strictly protected sensitive areas, qualified natural conservation areas, and sustainable conservation and controlled-use areas, as well as the reclassification of natural sites, reflect a hierarchy ranging from absolute protection to limited use. Special Environmental Protection Areas are particularly defined as regions of national and international ecological significance, where construction and land use must be strictly regulated. While this classification, which encompasses a broader scope, aims to graduate the protection–use balance, the management of these areas is distributed across the authorities of different institutions.

Accordingly, the implementation and spatial reflection of these multi-layered conservation statuses can be traced through statutory master plans prepared at different scales. These plans provide a structured framework for organizing the spatial components of the Lycian Way, linking legal conservation categories with observable land uses and settlements along the route. The Lycian Way and its surroundings have been analyzed under four main categories: settlements, natural and historical sites, and protected areas. Accordingly, the analysis relies on the master plans: (1) the Master Plan of the Aydın–Muğla–Denizli Planning Region (1/100,000 scale), and (2) the Master Plans of the Antalya–Burdur–Isparta Planning Region (1/100,000 scale) (Republic of Türkiye Ministry of Environment, 2023); (Republic of Türkiye Ministry of Environment, 2025)(Republic of Türkiye Ministry of Environment, 2025). In both plans, the region located between the Muğla and Antalya provinces of Türkiye is comprehensively represented in terms of diverse land uses, including settlements, forests, pasture, agricultural, and protected areas. These plans serve as the analytical base map, upon which various spatial features have been delineated along the Lycian Way and its surroundings. These include urban and rural settlements, archaeological sites, ancient cities, natural conservation areas, historical conservation areas, national parks and their conservation areas, special environmental conservation areas, qualified natural conservation areas, strictly protected sensitive areas, wildlife conservation areas, sea turtle breeding and conservation areas, as well as sustainable and controlled use areas (Figure 8).

Aydın–Muğla–Denizli master plan encompasses the western part of the study area, whereas Antalya–Burdur–Isparta master plan encompasses the eastern part, which constitutes the majority of the study area. The urban and rural settlements within the study area include Ölüdeniz, Kaş, Kalkan, Demre, Finike, Tekirova, Kemer, and Göynük. Forest, agricultural, and pasture areas represent the predominant land uses in the study area. In addition, several significant natural conservation areas are located within the study area, including Saklıkent National Park, Çığlıkara Natural Protection Area, Avlan Lake Wetland, Sarıkaya National Park, and Beydağları National Park and its associated conservation zones (Republic of Türkiye Ministry of Forestry and Water Affairs, 2025).

The western part of the Lycian Way begins in Ölüdeniz, Muğla, and the eastern part extends to the northern part of the Göynük district in Antalya. This route encompasses a series of ancient cities, including Telmessos, Kayaköy (Levissi), Patara (the capital of Lycia), Xanthos and Letoon (listed on the UNESCO World Heritage List), Antiphellos, Apollonya, Kyaneai, Simena, Myra, Arykanda, Limyra, Olympos, Phaselis, İdebessos, and Kitanaura (Ancient Cities of Turkey, XXXX). Beyond these major sites, the route and its surroundings also encompass smaller-scale yet historically significant archaeological sites. Numerous additional archaeological remains are dispersed across the Teke Peninsula. Furthermore, natural conservation areas are located between Ölüdeniz and Kalkan, as well as between Tekirova and Göynük, while wildlife protection zones are present to the north of Finike and Kaş.

Based on spatial analyses and existing plans, a review of existing literature, investigation of different land uses and protection statuses along the Lycian Way reveals several critical challenges. One of the main issues is the presence of privately owned lands along the trail, along with ongoing road construction projects (Bolat, 2023) and villa-type second homes near residential areas. While these developments threaten the continuity of the route, the changing landscape also undermines its harmony with the natural environment. The intensive utilization and high density of key tourism zones along the trail and the urbanization pressure they generate have a notable impact on the natural character of the region.

In particular, settlements such as Kaş, Kalkan, Tekirova, and Kuzdere have experienced rapid urban expansion due to growing tourism pressure, threatening the natural landscape in these dispersed tourist areas. Villages like Sarıbelen and Bezirgan, located near Kalkan, are under intense urbanization pressure as a result of their proximity to tourism hubs. The rising demand in these areas has led to the development of second homes and villa-style accommodations aimed at meeting tourism demand, rather than maintaining the rural identity of the villages. This trend puts additional pressure on agricultural lands in the region.

In contrast, agricultural zones in settlements like Gelemiş (also Kınık, Ova, and Yeşilköy neighborhoods), Demre, and Kumluca are largely dominated by greenhouse farming, which has intensified over time. As a result, the sections of the Lycian Way passing through these areas often run along paved roads, creating environments that are isolated, insecure, and lacking interaction with nature due to the industrial nature of the landscape.

Both existing settlements and new developments place increasing pressure on the region’s archaeological and natural character. For example, in the Xanthos-Letoon archaeological zone, a long-standing asphalt road running through the ancient city poses a direct threat to the integrity of the site. Although the Regional Council for the Conservation of Cultural Heritage decided to close this road and install fencing in 2010 under the revised Conservation Legislation enacted in 2004, these measures have not been effectively enforced. Consequently, additional steps are needed to ensure the protection and continuity of the site. Meanwhile, in Letoon, monuments and archaeological remains within the sacred area face seasonal threats from rising groundwater levels. Furthermore, the dominance of greenhouses surrounding the site has resulted in a visually incompatible landscape.

In terms of both local community capacities and the region’s characteristic cultural and ecological features, the Lycian Way offers significant opportunities for sustainable management and protection. In response to increasing pressure from tourism-driven development, several local communities and initiatives have been established throughout the region. These include environmental movements such as Phaselis İnisiyatifi (Phaselis Initiative), which opposes hotel construction and coastal development in Phaselis; Beycik Yöre Sakinleri (Beycik Community Group), formed in response to deforestation; and the Çıralı Ulupınar Çevre Koruma Kooperatifi (Çıralı Ulupınar Environmental Protection Cooperative), which was founded to resist environmental degradation in the Çıralı area. Culturally focused organizations also play a key role, including the Kemer Yörükler Derneği (Kemer Yörüks Association), which seeks to preserve and promote Yörük heritage, and the Kültür Rotaları Derneği (Cultural Routes Society), which works to protect and promote cultural routes (Bolat, 2023). The presence of these local initiatives offers significant potential for the success of site management. According to (U.N.E.S.C.O., 2013), cultural heritage management is not limited to physical conservation but is also a human- and community-oriented process. Therefore, the participation of local stakeholders constitutes a fundamental component of effective heritage management.

The Lycian Way passes through numerous ancient settlements, including Xanthos–Letoon— inscribed on the World Heritage List in 1988 (UNESCO World Heritage Centre, n.d.)—as well as first-degree archaeological and natural conservation zones. This makes the trail a rare example where cultural and natural assets coexist within a single route. Additionally, the region reflects key characteristics of the Mediterranean ecosystem. The presence of olive groves and citrus orchards indicates fertile agricultural land, while large areas are covered by forests and pastures. Inland, protected wildlife zones and national parks exist alongside nesting areas for Caretta caretta (sea turtles) along the coast, underlining the importance of preserving biodiversity. Altogether, the integration of natural and cultural diversity shapes the distinctive identity of the Lycian Way (Republic of Türkiye Ministry of Environment, 2025)(Republic of Türkiye Ministry of Environment, 2025).

The protection statuses defined within the legal framework are accompanied by restrictions and sanctions concerning construction and land use. Activities in protected areas, such as unauthorized construction, road development, alteration of land use, and interference with the natural environment, are subject to administrative fines, demolition orders, and cessation of operations under the relevant legislation. In particular, within Special Environmental Protection Areas and natural conservation sites, construction is permitted only in accordance with conservation-oriented master plans and under strictly limited conditions (Türkiye, 1983)(Türkiye, 1983)(Türkiye, 1983)(Türkiye, 1985). Nevertheless, in practice, the spatial implementation of this multi-layered protection system appears weak. Despite clearly defined protection statuses and legal sanctions, illegal constructions, second-home and villa-type developments, intense tourism pressure, and uncontrolled spatial interventions are observed, particularly in Special Environmental Protection Areas encompassing settlements such as Kaş, which contains sections of first-degree archaeological sites and areas designated for sustainable conservation and controlled use; Kalkan, which contains areas designated as Special Environmental Protection Areas; Tekirova and Kuzdere, which contain sections of first-degree natural sites and parts of a national park; Gelemiş, which contains sections of first-degree archaeological sites; and Demre, which contains sections of both first- and third-degree archaeological sites. These issues are further intensified by weak monitoring, fragmented planning authority, and delays in implementation, which help explain why rapid spatial transformations along the Lycian Way persist despite the protection framework. In light of these considerations, it becomes evident that national conservation policies should be re-evaluated within the frameworks established by international organizations and institutions. As (I.C.O.M.O.S., 2008) emphasizes, the partial, fragmented, or ad hoc protection of heritage attributes is inadequate. Instead, effective stewardship requires the development of strategies to manage and mitigate the impacts of deterioration processes affecting the cultural route and to prevent the negative consequences of neglect. All these efforts require the establishment of a coordinated system of legal measures and appropriate management tools that ensure the preservation of the route and the holistic emphasis of its value and significance. In this context, (U.N.E.S.C.O., 2013) provides a significant guiding framework for the management of Cultural World Heritage sites through the preparation, implementation, and monitoring of management plans.

As (Lin et al., 2024) state, one of the most commonly used strategies for the management of cultural routes in terms of regional conservation is establishing a route’s committee or association. This is followed by integral tourism management and route planning, constructing infrastructure facilities, and community participation. From this perspective, the conservation of this route is maintained by a non-governmental organization, the Cultural Routes Association, with partial support from the Ministry of Culture. Due to the limited scope of this association, no regulation or supervision can be imposed on the interaction between the route and the surrounding natural and cultural landscape areas. In addition, infrastructure facilities are provided through individual initiatives in rural establishments along the route, and there is no dedicated unit responsible for improving, regulating, or controlling the conditions of these facilities. Furthermore, in the absence of a route-specific conservation strategy, community participation in the protected areas surrounding the route is realized at a limited level. Higher-level conservation decisions are determined not by local units but by the central administration. In this context, planning, implementation, and supervision authorities are assigned to the central administration. While Special Environmental Protection Areas are designated by the Ministry of Environment, Urbanization, and Climate Change, national parks, and wildlife protection zones fall under the authority of the Ministry of Agriculture and Forestry, and natural conservation sites are evaluated by conservation boards affiliated with the Ministry of Culture and Tourism. Taken together, these institutional arrangements and management limitations reveal the segments of the route where its integrated cultural and natural values are most vulnerable, setting the stage for the subsequent discussion.

The abstract of this paper was presented at the Landscapes Across the Mediterranean (CrossMED) - 2nd Edition conference, which was held on the 09^th-11^thof December 2026.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors/individuals.

Not applicable.

The author(s) declare that there is no competing interest.