Spatial Distribution of Cultural Ecosystem Services in Rural Landscapes Using PGIS and SolVES

Abstract

Cultural ecosystem services (CES) play a vital role in rural well-being, yet their spatial patterns and local perceptions remain underexplored in many regions, including Türkiye. This study aims to assess the social values of CES in rural landscapes by focusing on the Şarkikaraağaç and Yenişarbademli districts of Isparta Province. Using Participatory Geographic Information Systems (PGIS) and the Social Values for Ecosystem Services (SolVES) models, we collected and analyzed spatial data from 836 community surveys, mapping 3771 CES value points. Sentinel-2A imagery and derived indices (NDVI, NDWI, SAVI, NDBI) were used to classify landscape infrastructures into green, blue, yellow, and grey categories. The results show that aesthetic and recreational services were most highly valued, followed by biodiversity, spiritual, and therapeutic values. Chi-square and Kruskal–Wallis tests revealed significant demographic and spatial variation in CES preferences, while Principal Component Analysis highlighted two key dimensions of value perception. MaxEnt-based modeling within SolVES confirmed the spatial distribution of CES with high predictive accuracy (AUC > 0.93). Our findings underscore the importance of integrating CES into sustainable land-use planning and suggest that infrastructure type and proximity to natural features significantly influence CES valuation in rural settings.

1. Introduction

The Millennium Ecosystem Assessment (MEA) brought attention to the extensive influence of CES on the overall population, particularly emphasizing the substantial adverse effects faced by rural communities due to their direct dependence on nature, resulting in a decline in the functionality of CES [1]. Since the CES concept emerged, several studies have been conducted to comprehend, assess, categorize, and visualize CES through various methods [2,3,4,5,6,7]. Regardless of the methods employed, studies commonly focus on subjects that address the benefits and values of CES. The benefits depend on the individual’s needs, preferences, and values. As a result, these subjective advantages are tied to locations and are likely to vary across geographical areas [8]. At this point, participatory mapping is an effective tool for understanding and evaluating the benefits and values of CES related to human interactions and ecological systems [9].

Despite the widespread increase in global research on CES, the published research on this topic is notably growing more slowly, especially in rural areas of Türkiye. While much of the research on CES has focused on urban and peri-urban areas, rural landscapes in Türkiye have been underexplored in this context. Rural areas, particularly in regions like Isparta Province, are rich in natural resources and cultural heritage; however, they have received limited attention in academic literature [10,11]. This gap exists because prior studies have often emphasized urbanization trends, agricultural intensification, or environmental degradation [8,12]. At the same time, the integration of cultural services with ecological services in rural landscapes remains a topic that has been insufficiently studied.

In Türkiye, rural landscapes play a crucial role in sustaining local livelihoods through agriculture, livestock, and traditional practices that are closely intertwined with nature [13]. These ecosystems are not only important for biodiversity but also hold deep cultural and spiritual significance for local communities. Despite their importance, these ecosystems have been overlooked in discussions surrounding CES, particularly in understanding how local populations perceive and value these services [10].

In this manner, this paper uses participatory techniques to evaluate the differing importance of specific CES benefit categories in a rural landscape context. In addition, the study aims to shed light on how these values are distributed spatially with different infrastructure types related to land use and land cover (LULC) [10]. This study evaluates which CES is typical for different infrastructures for CES-based planning and management [11].

The study’s goal is to establish a practical understanding of the potential of an integrated landscape approach in supporting sustainable landscape development. Arslan et al. [10] bring about a concept of landscape infrastructure to connect it to LULC. In this context, green infrastructure refers to areas covered by vegetation, including forests, green spaces, pastures, and parks. Similar terminology can be applied to other land cover types, like green infrastructure, to identify their large-scale distribution. Agricultural lands, permanent crops, pastures, and arable lands can be identified as yellow infrastructure, water bodies and wetlands as blue, and built-up areas as grey infrastructure. This understanding of landscape infrastructure is crucial for our study, as it helps us categorize and analyze the land cover types in the study villages. A map can be generated using the landscape infrastructure concept to portray the physical landscapes and how people use the land. The map contains green, blue, gray, and yellow infrastructure types related to CES.

Non-monetary values of CES need spatial representation to connect these values to the LULC. Various methods have been used to identify CES in the last decades. These methods incorporate applications of spatial mapping [14,15,16], PGIS [5,17,18,19], SolVES [12,20], MaxEnt modeling [21], and deep learning [22]. Various methods utilize various types of data, such as social media [23], remote sensing [24,25], and location data [26] to figure out how people affect the landscape and how the landscape affects the people.

Participatory Geographic Information System (PGIS) offers a distinct approach to understanding human-environment interactions. The involvement of local stakeholders is crucial for comprehending experiential environmental practices, meanings, values, and preferences. Only through participation can information from actual actors and users be effectively captured and utilized. In the PGIS approach, the experiences of local communities are regarded as equivalent to those of experts. Participatory mapping facilitates the analysis and interpretation of local knowledge within a geographical context. Spatial information technologies enhance stakeholder involvement in decision-making [26,27]. One of these spatial information technologies, SolVES, brings a social perspective to community knowledge. This approach provides spatially explicit indicators useful for understanding community preferences [28]. The SolVES model was created as a custom toolbar for ArcGIS^® to evaluate, quantify, and map the social values of ecosystem services [12,29]. The SolVES studies were primarily conducted in urban areas [20,29,30], with a focus on rural areas being limited.

However, a growing body of recent studies, conducted between 2023 and 2025, has begun to address this gap by applying participatory digital tools and spatial models to assess CES in rural contexts. Chen and Wu [31] employed a MaxEnt-based supply demand model to identify mismatches in recreation services within national parks, highlighting how geospatial analysis can inform management planning. Similarly, Haern and Fagerholm [32] integrated land cover data and social values through participatory mapping in Eastern Finland, revealing how local spatial knowledge informs the multifunctional characterization of landscapes. Yang et al. [33] investigated how older populations contribute to the preservation of cultural heritage in Zhejiang, highlighting the role of intergenerational memory in maintaining rural landscape identity. These approaches reflect a methodological shift toward combining community engagement with digital spatial analysis in CES research.

Further contributions underscore the importance of integrating local perceptions, stakeholder knowledge, and landscape infrastructure data into spatial decision-making frameworks. Duan and Xu [34] applied SolVES to Chinese rural landscapes to analyze the spatial distribution of spiritual and aesthetic values, demonstrating how cultural dimensions influence land-use preferences. Gao et al. [24] employed remote sensing indicators to map forest ecosystem service supply, linking satellite data with on-the-ground values. Adebayo [34] discussed resilience strategies that incorporate local ecological knowledge into sustainable development planning.

Rural settlements in the Şarkikaraağaç and Yenişarbademli districts, located in Isparta Province in southwestern Türkiye, have been chosen as the study area. This study identifies the social benefits of natural and cultural environments related to some of the CES using the SolVES model. The practical implications of our research include performing integrated spatial analyses of the landscape services perceived by the local communities and the LULC coverage, as well as identifying the most significant connections between location-based landscape services and LULC patterns. Based on our findings, we discuss how integrated landscape analysis enhances our comprehension of the patterns of association in multifunctional landscapes across various scales. Finally, we examine the methodological challenges and risks associated with spatially analyzing LULC and service patterns, particularly when extrapolated to larger scales. The specific objective is to identify the benefits of CES in rural landscapes and explore their spatial distribution across different infrastructure types.

CES are indispensable for the livelihoods of local inhabitants, and comprehending the connection between these services and the physical landscape is vital for developing sustainable landscape strategies. Due to specific objectives, the research is focused on CES, addressing the following research questions:

(1)

Which benefits are essential to villagers, and how are these benefits spatially distributed in different types of infrastructure?

(2)

Which relations can be identified to provide CES and various types of infrastructure?

2. Materials and Methods

2.1. Study Area

Şarkikaraağaç and Yenişarbademli districts, located in Isparta Province in the Southwest of Türkiye, mainly cover agricultural areas with many rural settlements; some are located near the Beyşehir Lake, while others have a strong relationship to the forest landscape. The study villages are situated near Kızıldağ National Park (KNP), the second-largest national park in Türkiye, known for its invaluable natural resources. The study area includes 16 rural settlements (Armutlu, Belceğiz, Beyköy, Çaltı, Çeltek, Fakılar, Fele, Gedikli, Karayaka, Kıyakdede, Salur, Sarıkaya, Yeniköy, Gölkonak, Pınarbaşı, and Çiçekpınar) with a strong relationship with the KNP, reflecting natural and cultural values (Figure 1).

2.2. Data Preparation

The rural settlements‘ digital layers and environmental variables were collected and prepared to identify infrastructure types and understand the landscape’s natural characteristics (

Table 1. Digital data sources and environmental variables.

Layer	Definition/Resolution/Unit	Source/Databases/Target
Digital elevation model	Meter (m)
Elevation	Percentage (%)	The elevation of the study villages was calculated using a digital elevation model and the slope analysis tool in the QGIS 3.40 program.
Distance to roads	Horizontal distance to the closest road (m)	https://download.geofabrik.de/europe/turkey-latest-free.shp.zip (accessed on 25 June 2023).
Distance to water	Horizontal distance to the closest water body (m)	https://download.geofabrik.de/europe/turkey-latest-free.shp.zip (accessed on 25 June 2023).
Land use and land cover	CORINE (2018)	https://corinecbs.tarimorman.gov.tr/ (accessed on 25 June 2023).
NDWI (Normalized difference water index)	NDWI was used to identify blue infrastructure	Sentinel 2A satellite images were used for remote sensing.
NDVI (Normalized difference vegetation index) and EVI (Enhanced vegetation index)	NDVI was used to identify green infrastructure	Sentinel 2A satellite images were used for remote sensing.
SAVI (Soil-adjusted vegetation index)	SAVI was used to identify green and yellow infrastructure	Sentinel 2A satellite images were used for remote sensing.
NDBI (Normalized difference built-up index)	NDBI was used to identify green infrastructure	Sentinel 2A satellite images were used for remote sensing.

).

Sentinel 2A satellite images generated the NDVI, NDWI, EVI, SAVI, and NDBI indices. Images from 25 June 2023, corresponding to the most prominent vegetation period, were obtained from Copernicus database web page (t.ly/HRf3a) Pre-processing, supervised classification, and validation analyses were then conducted. The infrastructure types were illustrated on a map and attached to the survey spatial data using remote sensing technologies. CORINE and field observations were used to verify the accuracy of the identified infrastructures. Ground truth data were collected through direct observations and GPS-based location recordings, ensuring spatial correspondence with the classified pixels. Each land cover class identified in the satellite image was cross-validated with in-situ measurements, including CORINE 2018 and photographic documentation and descriptive notes on vegetation type, land use practices, and surface characteristics.

NDVI (Normalized Difference Vegetation Index) was employed to measure vegetation health, helping us identify green infrastructure, such as forests, parks, and other vegetated areas. NDWI (Normalized Difference Water Index) was used to detect blue infrastructure, which includes water bodies like rivers, lakes, and wetlands. SAVI (Soil-Adjusted Vegetation Index), which accounts for soil brightness, helped distinguish areas with sparse vegetation and enabled us to identify yellow infrastructure, such as agricultural lands and grasslands. Lastly, NDBI (Normalized Difference Built-Up Index) was used to identify grey infrastructure, including urban settlements, roads, and other built-up areas.

It is a way to understand which kind of infrastructure provides more CES. On the other hand, non-spatial data (population, mainly sources of income, photos from essential areas, etc.) were provided from databases and field studies.

2.3. PGIS Data Collection

PGIS data were collected from July 2023 to October 2024. The study employed random sampling to gather information. In this study, survey days were organized separately in each village within the Şarkikaraağaç and Yenişarbademli districts. During these days, we made an effort to reach every individual actively living in the village, ensuring a wide representation of the community. This approach aimed to capture a broad set of perspectives on CES from various demographic groups within each village, rather than relying solely on random sampling.

The survey followed the guidelines of the Science and Engineering Ethics Committee at Süleyman Demirel University and received ethical approval (Ethical Identity Number: E-87432956-050.99-216507). Semi-structured surveys, including participatory mapping, were conducted in the study area to engage community members.

A team of two to six data collectors conducted the surveys. The questionnaire began with questions about socio-demographic background, including gender, age, education, household size, and sources of livelihood. Participants were also introduced to a satellite image map of their village. The survey continued with questions about the social values of CES (

Table 2. List of social values and other PGIS survey questions.

Values	Definitions
Socio-demographic characteristics	Gender, age, education level, household size, source of livelihood
Social values	Aesthetic–I value these places because they have attractive or pleasing landscapes.
	Biodiversity–I value these places for the plants, animals, wildlife, or ecosystem
	Cultural-I value these locations because they are a place for me to continue to pass down the wisdom and knowledge, traditions, and way of life of my ancestors.
	Economic–I value these places because they provide tourism opportunities, produce agricultural products, and support local businesses.
	Learning-I value these locations because we can learn about the environment through scientific observation or experimentation.
	Life-sustaining-I value these locations because they help produce, preserve, clean, and renew air, soil, and water.
	Recreation–I value these places because they provide outdoor recreation opportunities.
	Spiritual–I value these places because they have religious and spiritual meaning.
	Therapeutic–I value these places because they support mental and physical well-being.

).

Participants were asked to map the ecosystem service benefits they identified, and the mapping process ensued. The mapping was conducted using the most recent satellite image map, consisting of high-resolution Google Earth images from 2023, at scales of 1:5000, 1:10,000, and 1:25,000, with print sizes A1, A2, and A3. Different-colored pencils were used for mapping, and respondents were encouraged to mark as many locations for each benefit as they deemed relevant. Descriptions of these places and responses to background questions were recorded manually (face-to-face survey) during the study. They were later input into an online database using the map-based survey tool Partimap (https://www.partimap.eu/en/p/KMP_koy_anketi/1) (created on 12 June 2023).

2.4. Data Analysis and Mapping

Descriptive analyses were applied to understand the demographic background of the participants. Non-spatial data were analyzed using the Chi-Square and Kruskal–Wallis tests. The Chi-Square and Kruskal-Wallis tests were employed to assess the statistical significance of differences in various social values of CES across rural settlements and demographic groups.

Chi-Square Test was used to assess the independence of categorical variables, such as differences in CES preferences across different villages or demographic groups (e.g., gender, age group, or village). The Chi-Square test was chosen because the data were nominal, and the test helped determine whether significant differences existed between these categories.

The Kruskal–Wallis test was used to analyze differences in ordinal or non-parametric data, which is suitable for comparing more than two groups without assuming a normal distribution. This test was selected because our data were non-normally distributed, especially for comparing CES preferences across multiple villages. The Kruskal–Wallis test allowed us to assess the variation in social values across different rural settlements.

The assumptions of normality and homogeneity of variance were thoroughly examined. The Chi-Square test assumes that the sample size is sufficiently large for each category to ensure valid results. For the Kruskal–Wallis test, the assumption of homogeneity of variance was checked. Since our data were non-parametric, the test was deemed appropriate for our analysis. Additionally, Principal Component Analysis (PCA) was performed to group respondents based on their CES preferences.

The analysis of spatial value signifies the application of diverse methodologies. This process requires a comprehensive methodology capable of objectifying social, economic, and environmental values, which can be qualitatively different and subjectively obtained from multiple stakeholders. To meet this requirement, the SolVES tool was selected for the spatial expression of spatial values. SolVES version 3.0 was applied to assess the spatial distribution of participants’ perceptions of the social values of CES in the study area. SolVES is an open-access Geographic Information System (GIS) application developed by the United States Geological Survey. Collected PGIS data were used to map the social values of CES and to determine the weighted importance of these social values and spatial environmental information as inputs for the model. SolVES can use these models to transfer social values to similar values. Within SolVES, the kernel density analysis illustrates core and dispersal areas in the CES where CES are concentrated in terms of social value. Kernel density analysis enables spatial analyses, such as statistics on dominant environmental variables and regional statistics.

The SolVES procedure unfolds in two stages. In the first stage, a model of values is created using the survey data analysis tool, while in the second stage, user preferences are associated with environmental layers using the transition values tool [35,36]. MaxEnt reveals the relationship between PGIS survey data and environmental data, enabling the modeling of the social value of CES (Figure 2).

In the context of summarization, the PGIS methodology and the SolVES tool were utilized sequentially for data collection and spatial analysis in this study. First, PGIS was employed to collect participatory mapping data from local communities to identify the social values of CES. These data were then transferred into SolVES for further spatial analysis. SolVES utilized the spatial data collected by PGIS to map the social values of CES. In this process, the data provided by PGIS served as input for SolVES, enabling the visualization of the spatial distribution of CES values across different landscape features.

Furthermore, the MaxEnt model was integrated into the SolVES process. MaxEnt was used to model the spatial distribution of CES based on environmental variables and the social values generated by SolVES. This integration enabled the high-accuracy prediction of CES spatial patterns, providing a better understanding of the relationship between CES benefits and landscape features and infrastructures. Figure 3 visually represents the flow of the SolVES model and its interaction with the PGIS data collection process.

3. Results

3.1. Descriptive Analysis

Eight hundred thirty-six (836) surveys were conducted in all villages, and 3771 social values were identified from the spatial data. According to the survey findings, 39% of the participants are female, while 61% are male (Figure 4a). Most participants, 59%, are individuals aged 55 and above (Figure 4b). The educational attainment levels predominantly consist of primary education (58%), secondary education (16%), and middle school education (12%) (Figure 4c). Forty-four percent of participants live with fewer than three people in their households, while thirty-nine percent live with between three and five people (Figure 4d).

Additionally, 88% of participants reported being satisfied with village life. Approximately 34% of the local demographic consists of retirees. Their principal sources of income are agriculture, livestock, and fishing.

3.2. Statistical Analysis

The chi-Square (χ²) test statistics show that most villages show significant differences (p < 0.01) in Recreation, Life-sustaining, Spiritual, Biological Diversity, Aesthetic, and Therapeutic values. Economic and Cultural values show fewer significant differences, indicating a more uniform perception across villages. Beyköy, Çaltı, and Gedikli exhibit the highest chi-square values, suggesting substantial regional variations in how people value different social aspects. In Fele and Sarıkaya, some social values were not significantly different, indicating a more homogenous perception. Figure 5 represents a heatmap of some social values across villages. Darker blue shades indicate higher Chi-Square values, suggesting more significant statistical differences across villages. Beyköy, Çaltı, and Salur show the most variation, particularly in Recreation, Biological Diversity, and Aesthetic values.

The Kruskal–Wallis test results indicate statistically significant differences across multiple variables. When we examined the differences in the favorite locations of rural inhabitants, some locations showed statistically significant differences in usage patterns. These findings suggest substantial variations in park visits based on location, with the Kızıltepe Lakeside Viewing Area, Beyşehir Lake, Mada Island, Karayaka, and Geledost villages being highly significant at p < 0.01. At the same time, Sindel Plateau and Kokurdanlık Road show moderate significance at p < 0.05.

There were significant differences in the preferences of male and female visitors for specific locations. These results indicate that male visitors show stronger preferences for visiting Karayaka village, Beyşehir Lake, and the Kızıltepe Lakeside Viewing Area. At the same time, female visitors also visit these places, albeit with slightly lower statistical significance. Age also played a crucial role in determining preferences. The most significant destinations for male visitors over 65 were Beyşehir Lake, Karayaka, and Sindel Plateau. Female visitors aged 55–64 were more likely to visit Cedar Forests, while older male visitors preferred lakes and mountainous regions. The education level of visitors also influenced their visitation preferences. Findings highlight that men with primary education prefer Kızıltepe Lakeside Viewing Area, Beyşehir Lake, and Karayaka, while women with primary education prefer Kızıltepe Lakeside Viewing Area and Yenişarbademli Pond. Figure 6 provides a comprehensive visual summary of the Kruskal–Wallis test results across different locations and groups.

The highest H-scores were observed in Karayaka, Beyşehir Lake, and Sindel Plateau, suggesting substantial statistical differences in visitor distribution. Male visitors tend to show higher statistical significance in more locations than female visitors.

The results of the PCA demonstrate that two principal components account for a significant proportion of the variance observed within the dataset. The first principal component (PC1) represents 49.232% of the variance, while the second principal component (PC2) contributes 12.834%. This leads to a cumulative variance explanation of 62.066%. These findings suggest that these two components encompass a significant portion of the dataset’s information, facilitating dimensionality reduction while preserving key patterns. According to the results, PC1 is primarily associated with natural, aesthetic, and spiritual values. Biological diversity, life-sustaining functions, aesthetic value, and spiritual significance are all firmly tied to this component. PC2 is more closely related to economic and cultural values, as indicated by the higher loadings of economic value and cultural significance. This distinction suggests that, while PC1 captures environmental and intrinsic values, PC2 reflects human-centered and economic values.

In Figure 7a, the scree plot demonstrates that PC1 accounts for nearly 50% of the variance, while PC2 contributes approximately 13%, reinforcing that these two components capture most of the dataset’s meaningful structure. Figure 7b, the PCA biplot, illustrates how the variables align with the two principal components. PC1 (horizontal axis) is strongly associated with Aesthetic, Spiritual, and Biological Diversity, indicating that it signifies natural and intrinsic values. PC2 (vertical axis) is more connected to economic and cultural values, highlighting its role in capturing economic and social aspects. This visualization supports the interpretation that PC1 represents environmental and aesthetic importance, while PC2 is more aligned with socio-economic factors.

3.3. Analysis of the Infrastructure Types in the Study Area

Using Sentinel 2A satellite imagery data, comprehensive maps have been generated for the normalized vegetation index (NDVI), normalized water index (NDWI), soil-adjusted vegetation index, and enhanced vegetation index. The analysis of the NDVI reveals that values ranging from −1 to −0.1, which correspond to water-covered areas, encompass an area of 6591.74 hectares. In contrast, regions exhibiting NDVI values exceeding the 0.2 threshold, indicative of vegetative cover, have been quantified at 50,880.18 hectares. Similarly, the evaluation of the NDWI demonstrates that values within the 0–1 range represent water surfaces, which have been recalculated to encompass 8140.27 hectares following a reclassification. The map delineating landscape infrastructure types within the study area is presented in Figure 7. According to the information in Figure 8, it has been determined that 48% of the study area is constituted by green infrastructure, 8% comprises blue infrastructure, 30% is attributed to grey infrastructure, and 14% is classified as yellow infrastructure. The overall classification accuracy was calculated to be 85.19%, and the overall Kappa statistic was 0.7782, as per the accuracy assessment results.

3.4. Spatial Distribution Analysis of Social Values

Digital data sources and environmental variables shown in

Table 3. AUC values of the SolVES model.

Social Value	AUC	Social Value	AUC	Social Value	AUC
Aesthetic	0.931	Economic	0.941	Recreational	0.927
Biological diversity	0.941	Learning	0.950	Spiritual	0.934
Cultural	0.967	Life-sustaining	0.952	Therapeutic	0.940

were used to apply the SolVES 3.0 version procedure. The environmental variables consisted of 30 m resolution rasters, including elevation, slope, distance to roads (DTR), distance to water (DTW), and land cover and land use data, which were maintained for the modeling process. The social values of CES were attributed to the local population living around KMP. Spatial correlations were tested using the MaxEnt 3.4 version. The area under the curve (AUC) was calculated for each type of social value of CES. AUC measures the model’s success across all potential thresholds. If the area under the curve (AUC) exceeds 0.5, then the model performs better than random selection prediction [36]. Table 3 presents the Area Under the Curve (AUC) values, which indicate the model’s effectiveness.

In the context of environmental variables, some of the most significant factors influencing all social values of CES are distance to water (DTW), distance to road (DTR), and elevation. Accordingly, in line with the SolVES procedure, spatial data were analyzed based on information obtained from the local population according to the social value index. In this context, it was determined that nine social values of CES were provided and attributed value by the local population in KMP. The spatial distribution of CES is shown in Figure 9. Figure 9 shows that recreational, aesthetic, and biological diversity values are the most represented CES for all inhabitants in the study area.

4. Discussion

This study aimed to explore the social values of CES from the perspective of local communities living in rural areas of the Şarkikaraağaç and Yenişarbademli districts in Isparta Province, located in southwestern Türkiye. The research mapped the spatial distribution of social values associated with CES by applying the PGIS and SolVES models. It examined the relationships between various types of infrastructure and these services. The findings are consistent with, and at times contrast against, recent studies in the field, highlighting the intricate relationship between local communities and their environment.

The statistical analyses performed in this study, encompassing chi-Square (χ²) and Kruskal–Wallis tests, indicated significant regional discrepancies in the perception of various social values associated with CES across different villages. Notably, recreation, aesthetic, biological diversity, spiritual, and therapeutic values were particularly prominent in certain villages, such as Beyköy and Çaltı. Conversely, the economic and cultural values exhibited more uniformity across the villages. These findings align with recent research that underscores the significance of place-based values in influencing the perception of CES. For instance, Fagerholm et al. [27] illustrated how demographic and geographic factors shape individuals’ perceptions of ecosystem services. This observation corroborates our findings of regional discrepancies in the perception of CES in rural Türkiye landscapes.

Moreover, Arslan et al. [10] also found significant differences in how local communities value different CES across rural settlements, with the aesthetic and recreational values aligning with our findings in areas where natural landscapes, such as lakes and forests, are prominent. However, our study found a more pronounced differentiation between biological diversity and aesthetic values, reflecting a deeper connection to the natural environment. In contrast, other studies (e.g., Sherrouse et al. [12]) have primarily focused on urban or peri-urban areas, where social values are often more economically driven.

This study’s spatial analyses, which employed the SolVES model and MaxEnt algorithm, demonstrated that green infrastructure (forests, green spaces) was strongly linked to aesthetic and biological diversity values. In contrast, blue infrastructure (water bodies and wetlands) strongly correlated with recreational and therapeutic values. These findings align with recent research, such as that of Koschke et al. [11], who identified green infrastructure as a key contributor to aesthetic and biodiversity values, which aligns with our findings. However, our study extends their work by demonstrating how blue infrastructure (water bodies) also plays a significant role in providing recreational and therapeutic services, which were less emphasized in their research. Likewise, Shi et al. [20] showed that blue infrastructure, especially wetlands and water bodies, offers significant recreational and therapeutic services, further reinforcing the relationship observed in our study. However, our study’s findings diverged in the case of yellow infrastructure (agricultural lands), which showed a stronger relationship with economic and cultural values. Fagerholm et al. [8] and Plieninger et al. [26] have pointed out that agricultural landscapes are often linked to economic benefits and cultural practices. Yet, our findings suggest that in the studied region, agricultural areas are perceived as providing more economic services with limited cultural value attributed. This divergence may be due to regional differences in agricultural practices or the perception of the role of agriculture in the local culture.

PGIS in this study allowed us to capture local knowledge, which is crucial for understanding how communities interact with their landscapes. Brown and Fagerholm [5] and Fagerholm et al. [27] have emphasized the value of PGIS as a tool for engaging local communities and incorporating their perceptions into environmental management. Our findings confirm this, as the participatory mapping process revealed a strong connection between local values and the landscape. Moreover, this study corroborates the importance of participatory mapping in bridging local ecological knowledge with spatial data, a point also supported by Sherrouse et al. [29], who used similar participatory methods in national forest settings in the United States.

The mapping identified the locations of various CES and highlighted how people perceive and value their environment differently, depending on their demographic background. This aligns with Xiang and He [19], who found that PGIS can effectively capture the diversity of local knowledge and environmental perceptions, contributing to more inclusive decision-making processes.

While this study contributes to the understanding of CES in rural areas, it also highlights some of the gaps in the existing literature. While Sherrouse et al. [12] found that aesthetic and recreational values were the most prominent in urban areas, our study found similar trends in rural landscapes, particularly in settlements near water bodies and forests. However, we also observed that spiritual and therapeutic values were more pronounced in our study, which could be attributed to the rural context and the close connection between local communities and nature. Similarly, Duan and Xu [28] applied the SolVES model in urban contexts; however, its application in rural areas, particularly in Türkiye, has been limited. This research fills the gap by providing a spatial representation of CES values in a rural setting, contributing to the growing body of literature that emphasizes the need for CES mapping in less-studied areas, as noted by Koschke et al. [11] and Arslan et al. [10]. Future studies should extend these findings by exploring how CES are perceived in other rural areas of Türkiye and comparing them with urban settings.

Furthermore, additional studies are necessary to investigate the role of remote sensing technologies in enhancing CES mapping. Recent works, such as Liao et al. [22], have used deep learning techniques to map landscape values, which could be integrated with PGIS and SolVES in future research. These innovations could provide more accurate models of CES distribution, particularly in complex landscapes with multiple types of infrastructure.

Nevertheless, relying on local perceptions introduces limitations rooted in socio-economic heterogeneity, varying education levels, and potential bias. Previous studies have shown that perceptions of CES may differ significantly by gender, age, and land tenure, potentially skewing the representation of actual ecological functions [33,37]. These limitations necessitate triangulation with biophysical indicators and long-term monitoring to validate and refine spatial CES assessments. Future research could employ mixed-method approaches, such as PGIS combined with remote sensing and automated land-cover classification to cross-verify perceived service value distributions. Such integrative strategies have been effective in other contexts for balancing traditional ecological knowledge with empirical land-system modeling [38,39].

One of the key limitations of the study arises from the demographic characteristics of the local population. The sample is predominantly composed of older adults and individuals with primary education, which may influence the distribution of CES preferences. However, this demographic structure reflects the actual socio-spatial profile of the rural villages in the study area, where younger and more highly educated individuals typically reside elsewhere due to urban migration. As a result, the perceptions documented in this study predominantly represent permanent residents and long-term land users, which should be considered when interpreting the findings.

In terms of policy implications, integrating CES into rural development strategies in Türkiye can significantly contribute to maintaining ecological integrity while supporting local livelihoods. Recent research emphasizes the need to design participatory frameworks that incorporate the social values of ecosystems, thereby ensuring culturally relevant and equitable planning outcomes [34,38]. For example, rural planning in Türkiye could benefit from community-based land-use zoning that recognizes areas of cultural and recreational significance, much like successful riparian management initiatives in semi-arid Zimbabwe that aligned local perceptions with ecological functionality [38]. Additionally, aligning rural green infrastructure development with identified CES hotspots can bolster place-based stewardship, as seen in sustainable farming and intercropping systems that elevate both provisioning and CES [39].

Furthermore, future studies should focus on the temporal aspects of CES. As Gao et al. [24] suggested, the dynamics of CES values may change over time due to environmental or socio-economic changes. Understanding these temporal shifts could inform more sustainable land management practices.

5. Conclusions

This study adds to the expanding literature on CES by evaluating the perceived social values of rural landscapes with the SolVES model. By combining georeferenced public perception data with land use and environmental factors, we illustrated how various rural landscape structures—especially differences in topography, vegetation cover, and settlement patterns—affect the distribution and intensity of CES values, including aesthetics, recreation, and cultural significance heritage.

Our findings indicate that the values of CES are unevenly distributed across the landscape, as they tend to cluster in regions characterized by distinct environmental and landscape features. For example, traditional agricultural areas and landscapes that preserve a visual or functional connection to nature were consistently linked to higher social value indices. This implies that maintaining such landscape elements is essential for sustaining CES in rural areas.

From a policy viewpoint, these results emphasize the importance of incorporating cultural and perceptual aspects of ecosystem services into rural planning frameworks. Decision-makers should consider the spatial variability of CES values when crafting land use strategies, ensuring that interventions do not unintentionally jeopardize regions with high perceived worth. Furthermore, incorporating local insights and stakeholder values into planning can enhance social acceptance and support the long-term sustainability of rural development policies.

This study highlights how spatially explicit modeling tools, such as SolVES, can effectively connect ecological data with community landscape values. Future research could enhance this method by incorporating a broader range of demographic datasets and participatory mapping techniques to better align environmental management with local cultural priorities.

To improve the practical relevance of CES in rural land-use planning, this study recommends incorporating CES mapping results into local zoning and development strategies. Identified CES hotspots can be prioritized for cultural conservation or designated as multifunctional buffer areas. Land-use policies should also embed CES criteria within environmental assessments and rural investment programs. Furthermore, participatory planning frameworks involving community members, planners, and ecologists can ensure that cultural values are adequately reflected in land management decisions. These integrative actions would support more culturally informed and socially accepted planning outcomes.