1. Introduction
In order to meet the public’s demand for nature, the Chinese government has established a national forest park (NFP) system across the country [
1], which refers to forest areas with important natural and cultural values that are of national conservation significance [
2]. However, the NFP system has revealed many problems in its long-term development, such as overdevelopment that impairs tourism carrying capacity, infrastructure shortages that reduce tourists’ perceived well-being, and landscape degradation that affects the public’s perceived experience and health [
3,
4,
5]. Therefore, in order to maintain public well-being and perceived experience, there is a need for sufficient evidence of the role of public opinion and indigenous values NFP management and planning [
6,
7,
8], and wider public participation as the basis of resource management.
The Ecosystem Service concept has been incorporated into the assessment frameworks and decision-making systems of many forest parks. The concept promotes collaboration between researchers, the public and decision-making bodies and categorizes the benefits generated by ecosystem services into four categories: regulation, support, provisioning and culture [
9,
10]. Among them, cultural ecosystem services (CES) refer to the intangible benefits that individuals derive from ecosystems through spiritual enrichment, cognitive development, reflection, recreation, and aesthetic experiences [
11,
12]. CES are often associated with forest environments and can significantly improve public health [
13]. In addition, CES can draw public attention to forest ecosystems and lead them to participate in environmental governance [
14].
However, existing theoretical frameworks fail to specify what specific benefits CES bring to individuals and ignore the biophysical process of interaction between individuals and the environment [
15,
16,
17]. In addition, CES involve subjective experiences such as spirituality and aesthetics, which are influenced by subjective perceptions and cultural contexts, and are not directly linked to changes in the natural environment. Therefore, they lack clear measurement boundaries and internal consistency [
18]. For accurate understanding, many scholars have deconstructed the attributes of CES and outlined specific types and sources of benefits. The UK National Ecosystem Assessment (NEA) emphasizes the need to distinguish between ‘services’ and ‘benefits’ [
19]. Fish et al. [
20] argue that the main components of CES are environmental spaces and cultural practices, and that the interactions between the two produce individual-level cultural ecosystem benefits (CEBs), including identity, experience, and capability. For NFP, its ecosystem provides the public with a rich environmental space, which can effectively improve physical and mental health, individual competence as well as social identity through practices such as environmental education, forest therapy, consumption experiences and sports [
21,
22,
23,
24,
25]. Incorporating CEB into the assessment framework of NFP helps decision makers to quantify the well-being gained by the public, and one of the objectives of our study is to explore what CEB is perceived by the public in NFP and to establish a universal indicator system.
In the assessment process of evaluating ecosystems, public views may diverge from expert opinions due to differences in values, knowledge structures, and economic interests [
26,
27,
28], and can provide guidance from a societal perspective [
29,
30,
31]. Public participation methods have been widely used to guide environmental policies, such as biodiversity policymaking in the European Union [
32], Ecosystem governance in the Taihu Lake Basin in China [
33]. Some countries have established tools or procedures for public participation, such as the Ecosystem Management Program in Northern Lower Michigan, USA and the Nature 2000 network in Europe [
34,
35]. Unlike traditional interview-based surveys, public participatory geographic information systems (PPGIS) allow cultural services to be better visualized and expressed in resource management processes [
36], helping researchers to relate cultural ecosystem services in a given space to perceived human benefits [
37]. Currently, PPGIS is mainly used in the study of ecosystem services to assess the relationship between the objective physical environment and cultural ecosystem benefits, such as natural landscape features [
38], landscape accessibility [
39], land use [
40]. Considering that cultural ecosystem benefits do not rely on the interpretation of a single sensory dimension [
41], and that the perceived preferences of policy makers, managers, and the public are inconsistent [
42], investigating nature-based sensory perceptions of experiences during public engagement can help to provide insights into the factors influencing people’s concerns about ecosystem benefits [
43].
A particular challenge at present is to clarify how the public’s perception of the environment influences cultural ecosystem benefits (CEB) [
37]. While the effects of cultural ecosystem services (CES) provided by the physical environment of forests on individuals have been demonstrated [
13], it remains unclear how national forest parks (NFPs) drive the public’s perceived sensory dimensions to provide cultural ecosystem benefits (CEBs). We must recognize that the physical environment is only the basis of perception, intangible values are assigned by the superposition of cognition and awareness [
30], and people’s evaluation of nature is not determined by purely objective landscape features [
44], but rather from the interplay of the physical environment and Perceived Sensory Dimensions (PSD) [
45].
PSD is an important way to assess environmental perception. People’s evaluation of the real environment does not rely on a single physical environment such as visual or auditory [
41], and PSD has the advantage of categorizing the public’s perceived value of green space into several categories, which to a large extent correspond to the public’s different cultural practices respectively [
46,
47,
48]. Given that all cultural ecosystem services require the involvement of human sensory organs and the brain to interpret the information generated by ecosystem components and structures, respectively [
49], we believe it is necessary to establish a link between PSD and CEB. Furthermore, as a basis for evidence-based design, PSD has been used to design nature-based recreation policies in Denmark and healing gardens in Sweden [
47], but such studies have not answered a key question: for NFP or similar forested environments, is it better to enhance the integrated quality of environmental perceptions or to focus on improving the experience of a particular perceptual dimension? In this regard, Stoltz and Grahn [
50] modeled the synergies/tensions between different perceived sensory dimensions, arguing that a close collaboration of multiple perception dimensions enhances the aesthetic value of green spaces. However, our study is not limited to aesthetic experience, but is oriented to the broader human well-being, how to enhance the public’s environmental perceptual experience can lead to more CEB? We hope to answer the above questions to better understand the comprehensive value of national forest parks, and to propose refined solutions for planning and maintenance.
We constructed an evaluation method for the cultural ecosystem benefits of national forest parks (NFPs), specifying which factors influence the cultural ecosystem benefits of NFPs through the perceived sensory dimensions (
Figure 1). We integrated PPGIS into the management process of NFP to facilitate decision-making through visual representation. The objectives of our study are as follows:
What cultural ecosystem benefits are perceived by the public within the NFP?
Is there a link between perceived sensory dimensions and cultural ecosystem benefits in NFP?
Are there differences in the extent to which different perceived sensory dimensions and different levels of perception influence cultural ecosystem benefits in national forest parks?
3. Materials and Methods
This study combines various methods such as questionnaire survey, PPGIS and hotspot analysis to design the study process (
Figure 2). First, the Cultural Ecosystem Benefits Scale for assessing NFP was constructed. A PPGIS survey was conducted in Fuzhou NFP to record the coordinates of visited sites, self-reported PSD levels, CEB perception ratings, and personal information of visitors. Then, the geographic coordinates were visualized and analyzed using ArcMap 10.8. To answer the purpose of the study, the questionnaire data were subjected to correlation analysis, principal component analysis, and multifactor ANOVA. Finally, the visualization analysis results and questionnaire analysis results were combined to give refined management advice.
3.1. Study Area
This study was conducted in the Fuzhou National Forest Park, which is located in Fuzhou City, Fujian Province in southern China and is the earliest established and most representative national forest park in the city. The park is located in the northern mountainous area of Fuzhou City, covering an area of 8.59 km
2, and is popular among the public because it is only about 5–6 km from the city center (
Figure 3).
We assessed the feasibility of Fuzhou National Forest Park as a study area. First, the park’s visitors are demographically and sociologically universal, as the park combines the functions of scientific research, education, tourism, and convalescence in one place. The park’s infrastructure is relatively well-developed, with hiking trails, waterfront trestles, and ancient stagecoach routes connecting most of the area, as well as open activity spaces such as lawns and camping areas, which can meet the needs of different groups of people for cultural practices. Secondly, the park has typical landscape elements in the NFP definition, including natural landscapes such as lakes, forests and wildlife, as well as human resources such as monuments and temples. It is also a national forestry and grassland science base with national conservation value. Finally, the park has gradually revealed deficiencies in management and maintenance during its more than 30 years of operation, such as aging facilities, deterioration of soil resources, and reduction of native vegetation [
60], which may be detrimental to the well-being of visitors and local residents. Combining these three factors, we believe that a PPGIS survey can be conducted in Fuzhou National Forest Park.
3.2. Questionnaire
The questionnaire of this study contains three scales, namely, the Socio-Demographic Survey, the Cultural Ecosystem Benefits Scale, and the Perceived Sensory Dimensions Scale. All indicators and descriptions are presented in simplified Chinese.
3.2.1. Socio-Demographic Survey
At the end of the survey, respondents were asked to provide personal information including gender, age, educational background, occupation, place of residence, income level, duration of visit, and number of visits to Fuzhou National Forest Park. In addition, respondents were asked to record their opinions on the management and maintenance of Fuzhou National Forest Park.
3.2.2. Cultural Ecosystem Benefits (CEB) Indicators
A CEB scale (1 = completely disagree, 5 = completely agree) was developed based the theoretical framework of Fish et al. [
20] and some classic scales designed by previous authors. This scale is applicable to forest parks (
Table 1). The indicators of the Identities dimension refer to the Place Attachment Scale, which was designed by Williams et al. [
53] and Kyle et al. [
54]. In order to contextualize the topics under discussion, we drew upon the work of Xu et al. [
61], who had previously applied the Place Attachment Scale in a forest environment. The indicators of the Experiences dimension are based on the Perceived Restorative Scale (PRS) developed by Hartig et al. [
62]. Given the difficulty of directly measuring the physically perceived benefits of the PRS, we also introduced the Six-Dimensional Health State Classification System (SF-6D), a standardized scale for measuring health states. This is suitable for measuring the health level of Chinese people, as evidenced by previous research [
63]. The indicators of vigor and physical functioning were extracted from this source. The indicators of the capabilities dimension refer to the three high-level categories proposed by O’Brien [
51], namely learning, health, and economy.
In order to obtain a reliable CEB scale, 167 questionnaires were randomly selected to analyze their reliability and validity. The results showed that physical function led to a decrease in the overall reliability of the questionnaire. In addition, the results of the factor analysis showed that the rotated component matrix loaded A16 (physical function) onto only one component. We deleted this indicator and obtained a three-dimensional matrix with each indicator loading on its respective component with a value greater than 0.7. Finally, by deleting A16, we obtained a CEB scale with 18 indicators.
3.2.3. Perceived Sensory Dimensions Indicators
The PSD scale is used to assess individuals’ perceptions of different natural environments and has been shown to be reliable [
45]. We refer to Li et al. [
64] for the application of the PSD scale to forest environments, which consists of eight different dimensions: Serene, Natural, Species, Open, Cohesive, Refuge, Social, and Cultural, with additional descriptions following each of these dimensions to facilitate participants’ understanding. This section included eight questions (
Table 2) on a five-point Likert scale (1 = completely disagree, 5 = completely agree).
3.3. Public Participatory GIS
PPGIS typically ask respondents to mark on a map the geographic location and the benefits they perceive in that location [
29,
31,
65]. PPGIS focus on what individuals think, feel, and do in space [
66], and help to assess the multilevel human-place relationships of places by deepening the respondent’s sense of space [
67]. However, web-based PPGIS surveys have methodological limitations, and respondents may have a digital divide and digital exclusion, making the demographic-sociological structure of respondents young and highly educated [
68]. Respondents are unable to perceive the situation in the field with online maps alone [
69], and the data obtained lacks sufficient geographic precision. Conducting participatory mappsing in the field deepens respondents’ understanding of the questions and yields higher response rates [
70], while using points instead of polygons to map spatial attributes yields more accurate results when the sample size is large enough [
71]. Therefore, we returned to the traditional field survey method of asking respondents to identify places they had visited on a map, recording the geographic coordinates and their feelings about being there.
3.4. Data Collection
We collected data from 22 March 2024–27 March 2024 within the Fuzhou National Forest Park. The main tools used were tablet computers and the OvitalMap application. First, researchers randomly invited tourists or local residents to participate in the survey, and respondents were required to visit the park for more than 30 min to ensure that they had a basic understanding of the park. Next, respondents identified a visited site on a satellite map and recorded the geographic coordinates using the Ovi Interactive Maps application 10.0. To ensure that the coordinates were accurate, coordinates were only recorded if the map scale was less than 1:1500. Finally, respondents completed a questionnaire to record their feelings about the site. The questionnaire consisted of 34 questions and included CEB scales, PSD scales, and a socio-demographic survey.
3.5. Data Analysis
We initially screened the data from 881 PPGIS surveys and obtained 853 valid data after deleting coordinate points outside the study area and invalid questionnaires. Statistical analysis of the questionnaires was performed using SPSS 26.0, and visualization of the geographic coordinate points were performed using ArcMap 10.8.
3.5.1. Statistical Analysis of Questionnaires
First, the reliability of the revised CEB scale was tested. Exploratory factor analysis was used to test whether the structure of the questionnaire met the expectations of the theoretical framework. Second, principal component analysis was used to test the public’s perception of different CEBs in the NFP, and the combined score coefficients and weights of each indicator in the CEB scale were obtained. Then, Pearson correlation analysis was used to test the relationship between the eight perceived sensory dimensions of NFP and CEB, and the strength of the correlation was categorized into three levels: high (0.6 < |r| ≤ 0.8), medium (0.4 < |r| ≤ 0.6), low (0.2 < |r| ≤ 0.4), and irrelevant (|r| ≤ 0.2). Finally, a multifactor ANOVA was used to test the effects of different PSDs on CEB in national forest parks. The degree of influence was categorized into three levels: strong (0.14 < ), moderate (0.06 < ≤ 0.14), weak (0.01 < ≤ 0.06), and no effect ( ≤ 0.01). In addition, post hoc tests can further analyze the effect of different perceived levels of PSD on the CEB overall. When the null hypothesis of the multifactor ANOVA was rejected (i.e., all eight categories of PSD had a significant effect on CEB overall), the different perceptual levels of PSD were compared pairwise using the Bonferroni method.
3.5.2. Data Visualization and Analysis
The collected geographic coordinates were visualized and analyzed using ArcMap 10.8. The collected georeferenced points were transformed into point elements and each point element was assigned a CEB and PSD score. Then, hotspot analysis was performed using the Getis-Ord Gi* tool. This tool is based on the principle that high scoring elements in space are more likely to be noticed, but may not be statistically significant. To be a statistically significant hotspot, elements should have high scores and be surrounded by other elements that also have high scores. The local sum of each element and its neighboring elements will be compared to the sum of all the elements and finally a statistically significant z-value will be obtained. The z-value can be used to determine the location in space where clustering of high- or low-value elements occurs; when the positive z-value is larger, the higher the degree of clustering of high-value elements; and when the negative z-value is smaller, the higher the degree of clustering of low-value elements. The principle of localization statistics can be expressed as
where
is the attribute value of element
,
is the spatial weight between elements
and
,
is the total number of elements, and the
statistic is the z-score, which does not need to be calculated further.