Next Article in Journal
Urban Subway Station Site Selection Prediction Based on Clustered Demand and Interpretable Machine Learning Models
Previous Article in Journal
The Ephemeral Cultural Landscape of an Australian Federal Election
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Land
Volume 14
Issue 8
10.3390/land14081611
land-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Landscape Health Assessment of Suburban Forest Parks with Different Land Use Intensities and Grid Scales

by
Hao Luo
,
Qing Zhao
*,
Wan-Hui Qian
,
Chi Zhang
,
Ling-Yu Zhang
and
Xiao-Jun Wu
Guangdong Provincial Key Laboratory of Silviculture Protection and Utilization, Guangdong Academy of Forestry, 233# 1st Guangshan Road, Tianhe District, Guangzhou 510520, China
*
Author to whom correspondence should be addressed.
Land 2025, 14(8), 1611; https://doi.org/10.3390/land14081611
Submission received: 13 June 2025 / Revised: 1 August 2025 / Accepted: 7 August 2025 / Published: 8 August 2025
Browse Figures
Review Reports Versions Notes

Abstract

Landscape health assessments are crucial for balancing conservation and use practices. However, studies considering different land use intensities and grid scales from a functional perspective are limited. This study aimed to develop a landscape classification and indicator system based on functional characteristics using Xiqiao Mountain National Forest Park and Yunyong National Forest Park as research objects. The entropy weight method was used to determine indicator weights, and grid analysis and spatial interpolation were used to analyze the spatial distribution, impact differences, and factors influencing landscape health. The effect of different grid scales on landscape health under varying land use intensities first weakened and then strengthened with increasing grid scales; the optimal grid scales for the two parks were 54 × 54 m and 66 × 66 m, respectively. The forest/non-forest area ratio decreased as the health level of the landscape increased. At the optimal grid scale, the threshold range of the forest/non-forest area ratio for achieving optimal landscape health was 4.46–5.81 and 0.79–0.87 in the two parks, respectively. The results indicate that improved landscape health in suburban forest parks requires the intensive development of construction land, enhanced forest management, and functional synergistic designs to promote coordinated functional development.

1. Introduction

China has the world’s largest plantation forest area, and it is the largest contributor to global greening, with forests covering over 25% of its land area in 2023. Moreover, China has accounted for one-quarter of the total increase in global greening over the past two decades [1]. Suburban regions, as transitional zones between urban and rural areas, generally have higher forest coverage than urban built-up areas. Suburban forests serve as important green infrastructure that improve suburban living conditions by providing services such as soil and water conservation, climate regulation, and environmental purification [2,3]. However, rapid urbanization and industrialization have significantly impacted the spatial distribution of suburban forests, leading to landscape fragmentation and reduced forested areas [4]. Suburban forest parks are key to maintaining the ecological balance of suburban forest ecosystems [5] and serve as major recreational centers for urban residents [6]. However, as park development expands, issues such as environmental degradation and excessive exploitation have become increasingly prominent. Suburban forest parks are typical areas for human–environment interactions [7]; therefore, coordinated protection and land use planning are required to ensure efficient and sustainable development.
Landscape health refers to the self-sustaining capacity of structures, functions, and processes among ecosystems [8]. Landscape health assessments evaluate the structural components, ecological processes, and functions of landscapes based on systematic classification [9] and are crucial for balancing conservation and land use. Various approaches have been applied for assessing landscape health, including geostatistics, participatory geographical information systems (GISs), and landscape pattern analysis [10,11,12]. In such studies, indicator systems are constructed using frameworks such as Vigor–Organization–Resilience (VOR) and Pressure–State–Response (PSR), while methods such as Analytic Hierarchy Process (AHP), entropy weight method, Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) method, and combination weighting method are applied to determine indicator weights to assess landscape health condition. Particularly, the entropy weight method has been widely applied due to its advantages of a simple calculation process, high accuracy, and strong objectivity. The VOR model emphasizes the integrity and sustainability of ecosystem structures, focusing on the assessment of the ecosystem’s natural attributes and integrity [13], while the PSR model focuses on the causal relationship between ecosystems and human activities and establishes an assessment system from social, economic, ecological, and other aspects [14]. However, most GIS-derived indicators used in these two models (e.g., productivity, diversity, connectivity, and ecosystem recovery or its economic, resource, and environmental dimensions) exhibit a degree of subjectivity and uncertainty [15]. More importantly, indicator classification methods that link landscape function to land use remain limited. Developing evaluation systems from a land use-based functional perspective can provide a more objective assessment of landscape health and better reflect the interactions between human activities and land.
Most studies have focused on ecosystems such as forests [16], wetlands [17], grasslands [18], and water bodies [19] and examined landscape-scale research areas such as national parks [20], scenic areas [21], and wetland parks [22]. However, suburban forest parks have received relatively little attention, and most of their landscape health assessments are single-case studies that do not include cross-case comparative analyses [23,24]. The comparative landscape health studies of suburban forest parks with different land use intensities could further reveal the extent of land use impacts on landscape health because land use directly influences ecosystem health [25,26]. The quantitative analyses of landscape health in relation to forest and non-forest landscape factors have rarely considered influencing factors. Although previous studies have identified the forest/non-forest area ratio as an influencing factor [27], research on its threshold range, which is required to optimize the health of forest landscapes in current suburban forest park planning, remains scarce [28]. Moreover, the spatial scale of landscape health assessments can affect the results. Most current studies use grid-based approaches to evaluate landscape health at a landscape scale. However, in practical research, the variability in spatial analysis results caused by the subjective selection of sampling units at the landscape scale has been overlooked. Although different grid sizes are used as basic sampling units to evaluate landscape health conditions in each sampling area, most research focuses on a single grid scale. Suburban forest park landscape health has not been widely assessed across different grid scales to examine spatial variations and determine the optimal grid scale, because different grid scales serve different purposes: for example, larger grids are used to analyze spatial patterns, whereas smaller grids are used to analyze spatial configurations [29].
To address these research gaps, this study aims to (1) characterize the spatial distribution pattern of suburban forest park landscape health under different land use intensities and grid scales; (2) determine whether grid scale influences the landscape health assessments of suburban forest parks with varying land use intensities; and (3) analyze how suburban forest park landscape health relates to internal forest landscape factors under different land use intensities and grid scales. To achieve these objectives, we examined the spatial distribution, impact variations, and influencing factors of landscape health across five grid scales for two major national forest parks in the Pearl River Delta region: Xiqiao Mountain National Forest Park and Yunyong National Forest Park. These parks are exemplary suburban forest parks that integrate ecological protection, science and education, and leisure. They serve not only as important spaces for leisure and entertainment but also as important biodiversity conservation areas in urban ecosystems. Given the significant difference in land use intensity between the parks, evaluating their landscape health can provide reference data for productive planning and functional enhancement of other suburban forest parks.

2. Materials and Methods

2.1. Study Area

Xiqiao Mountain National Forest Park, designated as a National Scenic Area, a National Geopark, and a National 5A-Level Tourist Attraction, is located in Nanhai District, Foshan City, Guangdong Province (Figure 1). It has an area of approximately 1304.84 hm2 (22°55′–22°57′ N, 112°56′–113°0′ E) and a southern subtropical monsoon climate, characterized by mild winters, cool summers, and abundant rainfall. The annual average temperature is 21.8 °C and annual average precipitation is 1638.5 mm. The vegetation is predominantly of the subtropical evergreen forest type, including major tree species such as Ficus concinna, Ilex rotunda, Schima superb, Castanopsis fissa, Castanopsis carlesii, Bombax ceiba, Bischofia javanica, Cinnamomum camphora, Michelia odora, and Rhodoleia championii.
Yunyong National Forest Park, one of China’s first Forest Experience Bases and a Forest Wellness Base, is situated in Gaoming District, Foshan City, Guangdong Province (Figure 1). It has an area of approximately 2007.8 hm2 (22°41′–22°46′ N, 112°38′–112°42′ E) and has a southern subtropical monsoon climate, characterized by synchronized rainfall and heat, distinct dry and wet seasons, and relatively low sunshine levels. The annual average temperature is 22.6 °C and annual precipitation is 1800 mm. Native broad-leaved tree species dominate the park, including Pinus elliottii, Schima superba, Cinnamomum camphora, Magnolia blumei, Eucalyptus urophylla, Cunninghamia lanceolata, Cinnamomum burmannii, and Castanopsis fissa.
A fundamental indicator for evaluating land use intensity is the proportion of construction land. A larger proportion of construction land (buildings, roads, and service facilities) indicates that more of the natural environment is occupied by human activities, which is typically associated with a higher land use intensity [30]. According to the National Forest Park Master Plan Standards (YL/T 2005–2024) and Guangdong Forest Park Management Regulations, the land used for engineering facility constructions in a national park should not exceed 3% of the total land area (excluding road construction). However, in this study, Xiqiao Mountain National Forest Park exhibited a high land use intensity, with a construction land rate of 15.61% and a construction land-to-forest land ratio of 1:5 (Table A1). In contrast, Yunyong National Forest Park had a low land use intensity, with a construction land rate of 0.67% and a construction land-to-forest land ratio of 1:148 (Table A2).

2.2. Data Sources

This study used QuickBird satellite imagery (spatial resolution: 0.6 m) from DigitalGlobe (USA) as the primary data source. The imagery, obtained from Google Earth, was captured on 13 October 2021. The projection coordinate system used was the WGS-1984 Web Mercator, which is commonly adopted for web-based geographic applications. The image projection coordinate system used the common map projection of web geography (WGS-1984-Web-Mercator). Based on the GPS control points collected from field surveys, high-precision geometric correction was performed on the QuickBird images using ArcGIS 10.6, ensuring a margin of error of less than 0.5 pixels. Image classification was primarily based on visual interpretation, with a screen resolution controlled within 1:1000 to ensure classification accuracy. Field surveys and land cover validation for Xiqiao Mountain National Forest Park and Yunyong National Forest Park were conducted from April to May and August to October 2022, with a total of 1055 and 915 GPS validation points collected, respectively (approximately 100 validation points for each land cover type). The overall classification accuracy of the interpreted images was 95.25%, with a Kappa coefficient of 0.95, for Xiqiao Mountain and 96.73%, with a Kappa coefficient of 0.96, for Yunyong, meeting research requirements.

2.3. Landscape Classification and Landscape Function Classification

Based on the land classification standard from “Technical Specifications for the Third National Land Survey” (TD/T 1055-2019) and the landscape functions and characteristics of the study areas, Xiqiao Mountain National Forest Park was classified into 5 primary landscape types and 12 secondary landscape types (Table A1) and Yunyong National Forest Park was classified into 5 primary landscape types and 9 secondary landscape types (Table A2).
Based on the landscape type classification and land use function results, a combined function category of ecological protection and recreational use was included in this study. The park patches and corridor landscapes were classified from a multifunctional perspective and the difference between forest and non-forest landscapes was added as a category, thus forming a three-level functional classification system (Table A3).

2.4. Index System and Methods for Landscape Health Evaluation

With reference to previous landscape health evaluation research in suburban forest parks [24], a set of indicator parameters for ecological protection, recreational use, and comprehensive functions was developed (Table A4).
Based on the median area (1273 m2) of all patches and corridors in Xiqiao Mountain National Forest Park, the basic sampling unit was set as 36 × 36 m square and sized at 0.5× (18 m), 1.5× (54 m), 2.0× (72 m), and 3.0× (108 m). The area was gridded using five basic sampling units and areas smaller than 1/2 of the sampling unit size were removed. This generated 10,071, 40,280, 4484, 2530, and 1120 sampling units, respectively. Similarly, based on the median area (1096 m2) of all patches and corridors in Yunyong National Forest Park, the basic sampling unit was set as 33 × 33 m2 and sized at 0.5× (17 m), 1.5× (50 m), 2.0× (66 m), and 3.0× (99 m). The area was gridded using five basic sampling units, and areas smaller than 1/2 of the sampling unit size were removed, generating 18,659, 70,111, 8146, 4656, and 2102 sampling units, respectively.
In accordance with the methods for calculating various indicators (Table 1), the ArcGIS spatial analyst module was used to calculate the patch–corridor metric values for ecological protection, recreational use, and comprehensive functions within every sampling unit.

2.5. Methods for the Comprehensive Evaluation of Landscape Health

2.5.1. Determination of Indicator Weights

To reduce subjectivity in the evaluation process and avoid human bias in weight assignment, this study adopted the entropy weight method to determine the weights of the indicators. The calculation process and formulas are as follows:
(1)
Calculate the proportion Pij for each indicator:
P i j = X i j ∑ i = 1 n X i j
where Pij is the proportion of the i th sample under the j th indicator, and Xij is the value of the i th sample under the j th indicator.
(2)
Calculate the information entropy Ej and differentiation degree Di:
E j = − K ∑ i = 1 n P i j ln P i j
D i = 1 − E i
where Ej is the information entropy, K is a constant ( K = 1 ∕ ln n ), n is the number of samples, Pij is the proportion of the i th sample under the j th indicator, and Di is the differentiation degree.
(3)
Calculate the indicator weight Wi:
W i = D i ∑ i = 1 m D i
where Wi is the indicator weight, Di is the differentiation degree, and m is the number of indicators.

2.5.2. Landscape Health Evaluation

The evaluation scores for ecological protection, recreational use, and comprehensive functions in each sampling unit were obtained by multiplying the standardized value of each indicator by its corresponding weight and summing the results. The comprehensive function for landscape health was calculated by summing the three function scores as follows:
Zi = Ei + Ri + Ci
where Zi is the comprehensive function for the landscape health of sampling unit i, Ei is the ecological protection function score of sampling unit i, Ri is the recreational use function score of sampling unit i, and Ci is the comprehensive function score of sampling unit i.
As the comprehensive function contains both ecological protection and recreational use functions, this study followed the landscape health calculation method of Luo et al. [24]. The ecological protection function for the landscape health of each sampling unit was calculated by adding the sum of the ecological protection function score and half of the comprehensive function score (meaning Ei + 0.5Ci), and the same method was used for recreational use function landscape health (meaning Ri + 0.5Ci). Based on the calculation results, ecological protection, recreational use, and comprehensive function landscape health were normalized and classified into one of five levels: very poor [0–0.2], poor [0.2–0.4], ordinary [0.4–0.6], good [0.6–0.8], and very good [0.8–1]. Finally, the Kriging spatial interpolation method in ArcGIS was used to visualize the spatial distribution of landscape health for each unit, and the distribution map of landscape health under different land use intensities and grid scales was obtained (Figure 2 and Figure 3). The forest/non-forest area ratio was also calculated for each sampling unit to analyze the relationship between landscape health levels (ecological protection, recreational use, and comprehensive function) and the forest/non-forest ratio.

3. Results

3.1. Spatial Distribution Pattern of Landscape Health

Under different grid scales, the areas with “very good” and good” ecological protection function landscape health in Xiqiao Mountain National Forest Park were primarily distributed in the northern, central–southern, and southwestern regions of the park and accounted for 1.13%, 1.55%, 1.77%, 2.39%, and 2.25% of the total area, respectively (Figure 2a, Figure 2d, Figure 2g, Figure 2j and Figure 2m). In contrast, the areas rated “very poor” and “poor” accounted for 98.87%, 98.44%, 98.13%, 96.50%, and 95.83% of the total area, respectively. Under different grid scales, the areas with very good and good ecological protection function landscape health in Yunyong National Forest Park were primarily located in the northwestern, northeastern, central, and central–southern regions of the park, accounting for 2.19%, 2.54%, 2.33%, 3.24%, and 4.11% of the total area, respectively (Figure 3a, Figure 3d, Figure 3g, Figure 3j and Figure 3m). In contrast, the areas rated very poor and poor accounted for 97.81%, 97.45%, 97.02%, 96.75%, and 95.89% of the total area, respectively.
Under different grid scales, the areas with very good and good recreational use function landscape health in Xiqiao Mountain National Forest Park were primarily distributed in the central, central–southern, and peripheral regions and accounted for 6.16%, 8.08%, 17.20%, 9.65%, and 5.28% of the total area, respectively (Figure 2b, Figure 2e, Figure 2h, Figure 2k and Figure 2n). Conversely, the areas rated very poor and poor accounted for 86.39%, 76.99%, 69.46%, 75.23%, and 66.46% of the total area, respectively. Under different grid scales, the areas with very good and good recreational use function landscape health in Yunyong National Forest Park were primarily located in the northwestern, central, and central–southern regions of the park and accounted for 0.14%, 0.40%, 0.48%, 0.70%, and 1.81% of the total area (Figure 3b, Figure 3e, Figure 3h, Figure 3k and Figure 3n). In contrast, the areas rated very poor and poor accounted for 97.41%, 93.40%, 94.72%, 95.31%, and 83.03% of the total area, respectively.
Under the 18 × 18 m and 54 × 54 m grid scales, the areas with very good and good comprehensive function landscape health in Xiqiao Mountain National Forest Park were primarily concentrated in the northern and central–southern regions and accounted for 1.08% and 1.07% of the total area, respectively (Figure 2c, Figure 2f, Figure 2i, Figure 2l and Figure 2o). Under the 36 × 36 m, 72 × 72 m, and 108 × 108 m grid scales, the areas classified as very good and good appeared in the same core regions, as well as the edge areas around the park, and accounted for 1.49%, 3.68%, and 1.25% of the total area, respectively. The areas rated very poor and poor under these grid scales accounted for 88.62%, 88.65%, 88.84%, 84.74%, and 94.25% of the total park area, respectively. Similarly, the areas with very good and good comprehensive function landscape health under the 17 × 17 m, 33 × 33 m, and 50 × 50 m grid scales in Yunyong National Forest Park were primarily distributed in the northwestern and central regions of the park and accounted for 0.28%, 0.18%, and 0.27% of the total area, respectively (Figure 3c, Figure 3f, Figure 3i, Figure 3l and Figure 3o). Under the 66 × 66 m and 99 × 99 m grid scales, the areas classified as very good and good expanded further into the northwestern, central, and central–southern regions, accounting for 0.48% and 1.44% of the total area, respectively. In contrast, areas classified as very poor and poor accounted for 97.61%, 98.81%, 98.76%, 98.18%, and 93.90% of the total area, respectively.

3.2. Landscape Health Variation Analysis

In Xiqiao Mountain National Forest Park, the variation in the total area ratio of very poor and poor landscape health levels under different grid scales was relatively small for ecological protection, recreational use, and comprehensive functions (Figure 4a–c). In addition, the variation in the total area ratio of very good and good landscape health levels for ecological protection was relatively small, whereas very good and good areas exhibited higher levels of variation for recreational use and comprehensive functions. As can be seen from Figure 4b, the area ratio of very good and good landscape health levels increased significantly (2.13) when the grid scale increased from 1.0× to 1.5× sampling units but decreased slightly when scaling from 1.5× to 2.0× (0.56). Notably, the area ratio of very good and good landscape health levels increased significantly (3.44) when scaling from 1.5× to 2.0× sampling units but decreased slightly when scaling from 2.0× to 3.0× (0.34) (Figure 4c). Overall, sampling units at grid scales of 0.5×, 1.0×, and 1.5× had less variation for ecological protection, recreational use, and comprehensive function, whereas those at 2.0× and 3.0× had higher variation levels. Therefore, the 1.5× sampling unit (54 × 54 m) is considered more appropriate for analyzing landscape health in Xiqiao Mountain National Forest Park.
In Yunyong National Forest Park, the variation in the total area ratio of very poor and poor landscape health levels for ecological protection, recreational use, and comprehensive functions under different grid scales was relatively low (Figure 5a, 5b, and 5c). In addition, the variation in the total area ratio of very good and good landscape health levels for ecological protection was relatively low, whereas the very good and good areas exhibited higher levels of variation for recreational use and comprehensive functions. As can be seen from Figure 5b,c, the area ratio of very good and good landscape health levels increased significantly (2.59 and 3.00, respectively) when the grid scale increased from 2.0× to 3.0×. Overall, the sampling units at 0.5×, 1.0×, 1.5×, and 2.0× grid scales had a lower impact on ecological protection, recreational use, and comprehensive function than at the 3.0× grid scale, which significantly impacted the recreational use and comprehensive function. Thus, the 2.0× sampling unit (66 × 66 m) is more suitable for assessing landscape health in Yunyong National Forest Park.

3.3. Relationship Between Landscape Health and Forest Internal Landscape Factors

In Xiqiao Mountain National Forest Park, increasing landscape health levels were associated with a decreasing forest/non-forest area ratio. This trend was observed for the ecological protection function under the 18 × 18 m and 36 × 36 m grid scales, recreational use function under all five scales, and comprehensive function under the 36 × 36 m, 72 × 72 m, and 108 × 108 m grid scales (Figure 6). Under the optimal grid scale of 54 × 54 m, the landscape health levels of ecological protection, recreational use, and comprehensive functions were the highest when the forest/non-forest ratio was 4.46–7.22, 0.11–0.40, and 4.46–5.81, respectively.
In Yunyong National Forest Park, the increasing landscape health levels were also associated with a decreasing forest/non-forest area ratio. Ecological protection, recreational use, and comprehensive functions all showed a decreasing trend under the five different grid scales (Figure 7). Furthermore, under the optimal grid scale of 66 × 66 m, landscape health levels of ecological protection, recreational use, and comprehensive functions were the highest when the forest/non-forest ratio was 2.16–2.86, 0.69–0.87, and 0.79–0.87, respectively.

4. Discussion

4.1. Spatial Distribution Patterns of Suburban Forest Park Landscape Health Under Different Land Use Intensities and Grid Scales

The spatial distribution of suburban forest park landscapes with health levels classified as very good and good under different land use intensities and grid scales was scattered, occupying small areas, predominantly in point-like, patchy, and linear forms (Figure 2 and Figure 3). For example, in both Xiqiao Mountain and Yunyong National Forest Parks, areas rated very good or good in ecological protection function appeared in patches in the northern, central–southern, and southwestern regions (each accounting for <2.40% of the total area) and in point-like or patchy forms in the northwestern, northeastern, central, and central–southern regions (each > 2.10%). Recreational use function areas rated very good or good were distributed in patchy and linear forms in the central and central–southern regions (each > 5.20%) and in patches in the northwestern, central, and central–southern regions (each < 1.90%). For the comprehensive function areas, the very good or good areas were predominantly distributed in patches in the northern and central–southern regions (each < 3.70%) and point-like or patchy forms in the northwestern, central, and central–southern regions (each < 1.50%). These patterns suggest that more effort is devoted to enhancing recreational functions in Xiqiao Mountain National Forest Park, where land use intensity is higher but ecological protection plays a more dominant role in the overall landscape health of the park. Conversely, ecological protection is greater in Yunyong National Forest Park, where land use intensity is lower but recreational use has a stronger influence on the comprehensive health of the park. This indicates that ecological protection and recreational use have a complex relationship of mutual restriction and coexistence in suburban forest parks with different land use intensities. Hence, there is a need to coordinate functions under different land use intensities in practical park planning and construction [31,32] to effectively enhance landscape health.
In contrast, suburban forest park areas with landscape health levels classified as very poor and poor under different land use intensities and grid scales were extensively distributed and occupied large areas, generally in surface forms (Figure 2 and Figure 3). For example, under different grid scales, in both Xiqiao Mountain and Yunyong National Forest Parks, areas with very poor or poor ecological protection functions accounted for >95.80% of the total area. However, areas with very poor or poor recreational use functions accounted for >66.40% and 83.00% of the total area, and comprehensive functions areas rated very poor or poor accounted for >84.70% and 93.00% of the total area, respectively. This is predominantly because, in this study, the landscape health of each sampling unit for ecological protection and recreational use functions was calculated as the sum of the evaluation value from the specific function and half of the value from the comprehensive function. Both Xiqiao and Yunyong parks contain a high proportion of comprehensive function forests and account for 67.84% and 92.21% of their total areas, respectively. These findings highlight that planning and arrangement solely through expanding large-scale, single-type forest landscapes are not sufficient to achieve landscape health. Scientific planning of forest spatial layouts, which is based on landscape integrity and intensive forest land use, is crucial to healthy landscape development. Future research should prioritize improving landscape diversity [33,34] and integrating land use simulation and landscape health assessment. Prediction models, such as CA-Markov [35], FLUS [36], and PLUS [37], can be applied to simulate the changes in landscape health under different land use development scenarios, analyze more healthy and reasonable land use planning methods, and provide a decision-making basis for the land use planning of suburban forest parks.

4.2. Influence of Grid Scales on the Landscape Health of Suburban Forest Parks with Different Land Use Intensities

The influence of different grid scales on landscape health assessments in periurban forest parks with varying land use intensities showed a trend that was initially weak and then increased in strength with the increase in sampling unit size (Figure 4 and Figure 5). This study demonstrates how the division of sampling units may affect evaluation results [38,39] and provides reference data for the scale effects for suburban forest parks across multiple grid scales. Moreover, previous studies have shown that the subjective arbitrary selection of unit sizes can lead to deviation in analysis outcomes [40]. This study improved the standardization of grid selection in suburban forest park landscape research and the reliability of the evaluation results by selecting the median area of patch–corridor units as the base grid size and performing expansion research on four additional scales.
Although previous studies have investigated single or multiple grid scales [41,42,43,44], few have identified an optimal grid scale. This study revealed that the landscape health of forest parks with high and low land use intensity may be more precisely assessed by dividing the park using 1.5× (54 × 54 m) and 2.0× (66 × 66 m) sampling units. This difference likely reflects the slower transition rate between various types of landscapes and the relatively low landscape richness and fragmentation in parks with lower land use intensities [45]. If the grid size is too small, the connected landscape within each grid will be overly fragmented, reducing assessment accuracy [46]. Conversely, larger grid scales can capture internal spatial information within grid units [47] and maintain spatial continuity and integrity [48], making them more suitable for landscape health analysis. Therefore, the optimal grid analysis scale for Yunyong National Forest Park (lower land use intensity) is slightly larger than that for Xiqiao Mountain National Forest Park (higher land use intensity). In addition, differences in region, vegetation, and other local factors also lead to differences in optimal grid scales. For example, at the regional scale, a 4 km grid can better reflect spatiotemporal variation patterns in landscape ecological risk in mountainous areas [49]; at the urban scale, a 2.5 km grid is optimal for analyzing the landscape pattern and carbon sink performance of urban green infrastructure [50]; and at the landscape scale, a 20 m grid is the standard for evaluating greening and the health of community parks [51]. Therefore, the selection of a reasonable grid scale is fundamental to scientific analysis, and the assessment of landscape health using optimal grid scales for different land use intensities of suburban forest parks can help improve the accuracy of research results.

4.3. Relationship Between Suburban Forest Park Landscape Health and Internal Forest Landscape Factors Under Different Land Use Intensities and Grid Scales

Under different land use intensities and grid scales, the forest/non-forest area ratio in the suburban forest parks generally decreased as the landscape health level increased. This suggests that single-type areas with densely distributed forest landscapes tend to have poorer landscape health. In addition, under the optimal grid scale of Xiqiao Mountain National Forest Park (high land use intensity), landscape health was the highest at a forest/non-forest ratio of 4.46–5.81 (Figure 6). However, under the optimal grid scale of Yunyong National Forest Park (low land use intensity), landscape health was the highest at a ratio of 0.79–0.87 (Figure 7).
In Yunyong National Forest Park, a forest/non-forest area ratio of 0.65 indicated optimal landscape health [27]; however, ratios between 0.48 and 0.41 are ideal for achieving both biodiversity conservation and recreational services [52]. The optimal ratio range observed for Xiqiao Mountain is different from that in previous research. This could be because the construction land and forest land accounted for 15.61% and 76.74% of the total area, respectively, in Xiqiao Mountain National Forest Park; the increase in built-up land and reduction in forest cover are key factors contributing to landscape health degradation [25,26]. Thus, a higher forest/non-forest ratio is required to maintain a healthy landscape. These findings indicate that rational planning and coordination of land use in suburban forest parks and changing the forest-to-non-forest ratio can improve the overall landscape health.
The optimal forest/non-forest ratios to achieve very good or good landscape health in terms of ecological protection and recreational use did not overlap between the two parks, and the gap between them was significant. For instance, for ecological protection function, the threshold ranges of the forest/non-forest area ratio to achieve very good and good landscape health levels in the Xiqiao Mountain and Yunyong National Forest Parks (4.46–7.22 and 2.16–2.86, respectively) significantly exceeded those for recreational use function (0.10–0.37 and 0.69–0.87, respectively) (Figure 6 and Figure 7). This difference in the overlap degree of the threshold ranges indicates mutual restriction between the two functions. The trade-off between ecological protection and recreational use in both parks is substantial, with weak synergies between the two functions. This could be a factor contributing to the relatively low landscape health observed at both sites.
With the rapid pace of urbanization and intensification of urban–rural transformation, public demand for landscape functions has shifted from single-purpose to multi-functional use, complicating efforts to balance and coordinate different functions in landscapes [53,54]. The trade-offs and synergistic relationships between different functions of suburban forest parks must be further elucidated.

4.4. Limitations and Future Research

This study adopted a multi-functional perspective to classify patch–corridor landscapes, construct an indicator-based evaluation system, and conduct a comparative analysis of suburban forest parks under varying land use intensities and grid scales to enhance the accuracy of landscape health assessments. However, several study limitations that should be acknowledged are as follows:
(1)
The evaluation results were not evaluated. Although grid-based and interpolation methods can effectively and comprehensively reflect the spatial distribution of landscape health within the study area, the accuracy of the calculated results still requires verification [46]. Future studies should incorporate orthophotography images obtained using drones and overlay them with the spatial distribution maps of landscape health to verify the results and further improve the scientific rigor and reliability of the overall assessment process. Additionally, the fixed weight of 0.5 for the comprehensive function score requires validation. In the future, the weights of the comprehensive function scores for different types of patches and corridors with comprehensive functions should be further clarified and sensitivity analysis conducted for validation.
(2)
The evaluation results were not compared longitudinally. The spatial differentiation of landscape health is a dynamic process and relative health can only be determined using a comparison approach [55]. However, the landscape health assessment presented in this study was static. Future studies should focus on long-term dynamic monitoring to further explore the driving factors and mechanisms influencing landscape health. Then, through appropriate interventions and management, the landscape health of suburban forest parks can be maintained sustainably.
(3)
Multi-scale evaluations were lacking. This study analyzed landscape health using five different grid scales, which improved data processing efficiency, error reduction, and fine-scale representation [41]. However, grid-based units cannot determine the overall features. Future work should apply landscape feature unit scales as the evaluation unit to comprehensively reflect the overall landscape background and the spatial relationships of the patches and corridors. Specifically, future research should divide landscape character units based on the landscape character assessment (LCA) method and consider combining grid-based scales with landscape feature unit scales to develop a more comprehensive, multi-scale evaluation framework.

5. Conclusions

This study assessed landscape health in two suburban forest parks with different land use intensities, Xiqiao Mountain National Forest Park (high) and Yunyong National Forest Park (low), across multiple grid scales using a functional indicator system. The main findings are as follows: (1) at different grid scales, areas with very poor and poor landscape health in both parks were predominantly distributed in patches (accounting for 66.46–98.86% and 83.03–98.81%, respectively). In Xiqiao Mountain National Forest Park, areas rated as very good and good landscape health were predominantly distributed in block and linear forms in the north, central, south–central, and southwest regions of the park (accounting for 1.07–17.20%), whereas in Yunyong National Forest Park, areas rated as very good and good landscape health were distributed as points and blocks in the northwest, northeast, center, south–central, and southwest regions (accounting for 0.14–4.11%). (2) The influence of grid scale on landscape health in parks with different land use intensities showed weakening and then strengthening trends with increasing scale. The optimal grid scales for evaluating landscape health were 54 × 54 m for Xiqiao Mountain National Forest Park and 66 × 66 m for Yunyong National Forest Park. (3) As landscape health levels increased, the forest/non-forest area ratio generally decreased across different land use intensities and grid scales. Under optimal grid scales, the best landscape health was observed when the forest/non-forest ratio threshold intervals ranged between 4.46 and 5.81 for Xiqiao Mountain National Forest Park and 0.79 and 0.87 for Yunyong National Forest Park.
It is necessary to integrate suburban forest landscape health evaluation into land use planning because land use changes lead to changes in landscape health. Land use planning strategies such as the intensive development of construction land and strengthening of forest construction [56] can be used to optimize land use patterns and improve the forest ecosystem health level. For example, to optimize forest spatial distribution, forests should be planned in potential ecological corridor sites for the development of parks in the regions. This would enable the forests to naturally connect the core ecological patches around the park with the isolated green spaces in the construction area. In forest–grassland ecological zones, forest areas should be in high-growth corridor zones to maintain the landscape diversity, control the unreasonable development of construction land, and protect the forest boundaries. Furthermore, functionally coordinated development can enhance spatial resource efficiency. Coordinated planning for suburban forest parks can improve both ecological protection and recreational use. For example, increasing the forest/non-forest ratio, maintaining low fragmentation, and promoting ecological connectivity can enhance the ecological conservation functions of recreation-designated construction land in suburban forest parks.

Author Contributions

H.L.: writing—original draft, writing—review and editing, visualization, software, methodology, and formal analysis. Q.Z.: writing—review and editing, validation, supervision, project administration, and funding acquisition. W.-H.Q.: writing—review and editing, investigation, and data curation. C.Z.: software, resources, investigation, and data curation. L.-Y.Z.: writing—review and editing and conceptualization. X.-J.W.: visualization, validation, and data curation. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the Forestry Science and Technology Innovation Project of Guangdong Province (grant no. 2022KJCX009).

Data Availability Statement

No datasets were generated or analyzed during the current study.

Acknowledgments

We thank the editor and reviewers for their valuable feedback, which has helped us improve the quality of our manuscript.

Conflicts of Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A

Table A1. Landscape classification system for Xiqiao Mountain National Forest Park.
Table A1. Landscape classification system for Xiqiao Mountain National Forest Park.
NumberPrimary ClassificationSecondary ClassificationDescriptionLand Area
1FarmlandFarmlandFarmland dedicated to the cultivation of water-dependent crops, such as rice and lotus. In addition, this includes areas that rotate between water-based and dryland crops and cultivated land that relies on artificial irrigation for the cultivation of dryland crops, including vegetables.2.10%
2GrasslandGrasslandArtificially planted grassland with a tree canopy density of <0.1, intended for scenic viewing or recreational relaxation, and barren grassland with a tree canopy density of <0.1, characterized by an exposed soil surface and the growth of various weeds.0.86%
3Water bodyWater bodyNatural or artificially excavated rivers, lakes, ponds, and artificial ditches used for water diversion, drainage, and irrigation.4.69%
4Forestry landProtective forestA forest primarily designed to preserve soil, prevent wind and sand erosion, conserve water sources, regulate climate, reduce pollution, improve the ecological environment, and enhance human production and living conditions.0.52%
Timber forestA forest primarily intended for the production of timber and wood fiber.5.05%
No timber forestA forest primarily intended for the production of non-timber forest products such as fruits, edible oilseeds, beverages, spices, industrial raw materials, and medicinal plants.3.33%
Scenic forestA forest primarily intended for esthetic purposes, providing opportunities for people to relax, play, and enjoy natural scenery.67.84%
5Construction landSquare landA public space primarily intended for recreational activities, fitness, commemoration, gatherings, and refuge.7.28%
Built-up landRefers to residential homes as well as buildings such as restaurants and hotels.1.00%
Landscape and management facilityRefers to leisure landscape facilities, such as pavilions, walkways, and pergolas in forest parks, and management and service facilities, such as restrooms, dining establishments, convenience stores, and visitor centers.0.71%
Special landRefers to land designated specifically for military purposes, religious activities, or burial sites.2.22%
RoadPrimarily refers to roadways for vehicles and pedestrian walkways.4.40%
Table A2. Landscape classification system for Yunyong National Forest Park.
Table A2. Landscape classification system for Yunyong National Forest Park.
NumberPrimary ClassificationSecondary ClassificationDescriptionLand Area
1FarmlandFarmlandFarmland dedicated to the cultivation of water-dependent crops, such as rice and lotus. This includes areas that rotate between water-based and dryland crops and cultivated land that relies on artificial irrigation for the cultivation of dryland crops, including vegetables.0.05%
2GrasslandGrasslandArtificially planted grassland with a tree canopy density of <0.1, intended for scenic viewing or recreational relaxation, and barren grassland with a tree canopy density of <0.1, characterized by an exposed soil surface and the growth of various weeds.1.57%
3Water bodyWater bodyNatural or artificially excavated rivers, lakes, ponds, and artificial ditches used for water diversion, drainage, and irrigation.5.04%
4Forestry landProtective forestA forest primarily designed to preserve soil, prevent wind and sand erosion, conserve water sources, regulate climate, reduce pollution, improve the ecological environment, and enhance human production and living conditions.92.21%
Timber forestA forest primarily intended for the production of timber and wood fiber.0.19%
Scenic forestA forest primarily intended for esthetic purposes, providing opportunities for people to relax, play, and enjoy natural scenery.0.27%
5Construction landSquare landA public space primarily intended for recreational activities, fitness, commemoration, gatherings, and refuge.0.05%
Commercial and service landRefers to the land used for accommodation, catering, entertainment, health, and other facilities.0.53%
RoadPrimarily refers to roadways for vehicles and pedestrian walkways.0.09%
Table A3. Classification levels of landscape functions.
Table A3. Classification levels of landscape functions.
Primary ClassificationSecondary ClassificationTertiary ClassificationDescription
Landscape patchesEcological protection landscape patchesEcological protection
forest patches		The predominant landscape features are naturalized forest landscapes, exhibiting complex landscape structures, diverse ecosystems, and a variety of vertical and horizontal elements within these patches. These include forest patches with specific functions, such as soil and water conservation, water resource conservation, windbreaking, and sand fixation, and stands of parent trees for specialized tree species and environmental protection forests.
Ecological protection non-forest patchesThese are non-forest type patches formed naturally within the landscape, excluding naturalized forest landscapes. These include naturally occurring bodies of water and grasslands dominated by natural herbaceous plants, such as natural lakes and uncultivated grasslands.
Recreational use landscape patchesRecreational use forest patchesThese predominantly artificially planted forest landscapes are characterized by a simple structure and a single ecosystem type, offering various recreational and scenic functions. These patches encompass economic, timber, and experimental forests within special-use and recreational forests.
Recreational use non-forest patchesThey also include artificially excavated water bodies, decorative lawns, and other man-made natural patches, including artificial ponds, man-made grasslands, and similar features. In addition, these include patches characterized by hard or semi-hard artificial surface spatial structures, encompassing areas such as plazas, commercial and service land, special-purpose land, and other types of developed land.
Comprehensive landscape patchesComprehensive forest patchesPrimarily artificially planted quasi-natural forest landscapes with multiple ecological, cultural, scenic, and military functions. These forest patches encompass national defense forests, scenic forests, and revolutionary memorial forests of historical and cultural significance.
Comprehensive non-forest patchesNon-forest patches primarily comprise crop cultivation areas and bare land, including paddy fields, irrigated fields, bare soil, and bare rocky gravel areas.
Landscape corridorsEcological protection corridorsEcological protection forest corridorsA complex landscape structure, diverse ecosystems, and ecological corridors with functions for species and material migration activities. The main categories include forest strips and forest networks in protective forests with a width of ≥12 m.
Ecological protection non-forest corridorsNaturally formed rivers
Recreational use landscape corridorsRecreational use forest corridorsPredominantly comprising artificially planted forest landscapes with simple structures and a single ecological system, these corridors serve various recreational and scenic functions. They primarily consist of forest strips and networks within protective forests <6 m in width.
Recreational use non-forest corridorsArtificially constructed channels and corridors characterized by hard or semi-hard surface structures, primarily designated for road use.
Comprehensive corridorsComprehensive forest corridorsCorridors primarily characterized by artificially planted forest landscapes with simple structures and a single ecological system, providing various recreational and scenic functions. These corridors are typically forest belts and networks within protective forests, with widths ranging from 6 to 12 m, inclusive.
Comprehensive non-forest corridorsArtificially excavated rivers
Table A4. Indices used for landscape health evaluation.
Table A4. Indices used for landscape health evaluation.
Evaluation
Index		Ecological Protection Index Recreational Use IndexComprehensive Function Index
Patch X1 Ecological protection patch area ratio +Y1 Recreational use patch area ratio +Z1 Comprehensive function patch area ratio +
X2 Ecological protection patch density −Y2 Recreational use patch density −Z2 Comprehensive function patch density −
X3 Ecological protection patch edge density +Y3 Recreational use patch edge density +Z3 Comprehensive function patch edge density +
X4 Ecological protection patch fragmentation −Y4 Recreational use patch accessibility +Z4 50% Ecological protection patch fragmentation + 50% recreational use patch accessibility
X5 Ecological protection patch isolation −Y5 Recreational use patch isolation −Z5 Comprehensive function patch isolation −
X6 Ecological protection patch diversity +Y6 Recreational use patch diversity +Z6 Comprehensive function patch diversity +
X7 Ecological protection patch fractal dimension +Y7 Recreational use patch fractal dimension +Z7 Comprehensive function patch fractal dimension +
Corridor X8 Ecological protection corridor naturalness +Y8 Recreational use corridor density −Z8 50% Ecological protection corridor naturalness + 50% recreational use corridor density
X9 Ecological protection corridor curvature −Y9 Recreational use corridor curvature −Z9 Comprehensive function corridor curvature −
X10 Ecological protection corridor width ratio +Y10 Recreational use corridor width ratio +Z10 Comprehensive function corridor width ratio +
X11 Ecological protection corridor loopiness +Y11 Recreational use corridor loopiness +Z11 Comprehensive function corridor loopiness +
X12 Ecological protection corridor point-line ratio +Y12 Recreational use corridor point-line ratio +Z12 Comprehensive function corridor point-line ratio +
X13 Ecological protection corridor connectivity +Y13 Recreational use corridor connectivity +Z13 Comprehensive function corridor connectivity +
X14 Ecological protection corridor fractal dimension +Y14 Recreational use corridor fractal dimension +Z14 Comprehensive function corridor fractal dimension +
Note: (+): the higher the value, the better the function; (−): the lower the value, the poorer the function.

References

  1. Chen, C.; Park, T.; Wang, X.H.; Piao, S.L.; Xu, B.D.; Chaturvedi, R.K.; Fuchs, R.; Brovkin, V.; Ciais, P.; Fensholt, R.; et al. China and India lead in greening of the world through land-use management. Nat. Sustain. 2019, 2, 122–129. [Google Scholar] [CrossRef]
  2. Kindler, E. A comparison of the concepts: Ecosystem services and forest functions to improve interdisciplinary exchange. For. Policy Econ. 2016, 67, 52–59. [Google Scholar] [CrossRef]
  3. Liu, S.; Dai, L.; Wen, Y.; Wang, H. A review on forest ecosystem management towards ecosystem services: Status, challenges, and future perspectives. Acta Ecol. Sin. 2015, 35, 1–9. (In Chinese) [Google Scholar] [CrossRef]
  4. Ren, F.; Chen, L.; Li, T.; Li, M. Spatial changes of suburban forest ecological functions and their impact on ecological equity in the process of urbanization-A case study of Jiangning District, Nanjing, China. Forests 2023, 14, 1308. [Google Scholar] [CrossRef]
  5. Wang, Y.; Yang, Y.; Li, J.; Zhao, M.; Gong, H.; Wang, L.; Zhao, L.; Huang, M. Analysis on the change of ecosystem service value of national forest park and its coupling with social economy in the past 40 years. Pol. J. Environ. Stud. 2022, 31, 1377–1387. [Google Scholar] [CrossRef]
  6. Zhao, M.; Zhang, J.; Xie, J.; Zhang, M. Empirical study on recreation suitability evaluation of suburban forest park: A case study of Baiyun Forest Park in Lishui City, China. Appl. Ecol. Environ. Res. 2019, 17, 3499–3512. [Google Scholar] [CrossRef]
  7. Su, L.; Ji, X. Spatial-temporal differences and evolution of eco-efficiency in China’s forest park. Urban For. Urban Green. 2021, 57, 126894. [Google Scholar] [CrossRef]
  8. Li, Y.; Liu, H.; Zhang, H.; Sun, Y.; Li, Y. The study of landscape health based on landscape structure and water environmen: A case study of Xixi Wetland Park in Hangzhou. J. Nat. Resour. 2015, 30, 761–771. [Google Scholar] [CrossRef]
  9. Yan, S.; Li, H.; Li, H.; Zhang, Y. Ecosystem health assessment method of eco-redline based on land use and landscape pattern in Nanjing. J. Nat. Resour. 2020, 35, 1109–1118. [Google Scholar] [CrossRef]
  10. García-Ayllón, S.; Martínez, G. Analysis of correlation between anthropization phenomena and landscape values of the territory: A GIS framework based on spatial statistics. ISPRS Int. J. Geo-Inf. 2023, 12, 323. [Google Scholar] [CrossRef]
  11. Schüpbach, B.; Kay, S. Validation of a visual landscape quality indicator for agrarian landscapes using public participatory GIS data. Landsc. Urban Plan. 2024, 241, 104906. [Google Scholar] [CrossRef]
  12. Fu, S.; Zhao, L.; Qiao, Z.; Sun, T.; Sun, M.; Hao, Y.; Hu, S.; Zhang, Y. Development of ecosystem health assessment (EHA) and application method: A review. Sustainability 2021, 13, 11838. [Google Scholar] [CrossRef]
  13. Zhang, H.; Han, M.; Kong, X.; Sun, J.; Duan, X. Spatial differentiation and driving factors of ecosystem health in Yellow River Estuary National Park. Environ. Sci. 2025, 46, 3656–3671. (In Chinese) [Google Scholar] [CrossRef]
  14. Shu, H.; Xiao, C.; Ma, T.; Sang, W. Ecological health assessment of Chinese National Parks based on landscape pattern: A case study in Shennongjia National Park. Int. J. Environ. Res. Public Health 2021, 18, 11487. [Google Scholar] [CrossRef]
  15. Wang, J.; Zhang, J.; Wang, P.; Ma, X.; Yang, L.; Zhou, L. Progress in ecosystem health research and future prospects. Sustainability 2022, 14, 15814. [Google Scholar] [CrossRef]
  16. Meng, Y.; Cao, B.; Dong, C.; Dong, X. Mount Taishan forest ecosystem health assessment based on forest inventory data. Forests 2019, 10, 657. [Google Scholar] [CrossRef]
  17. Ekumah, B.; Armah, F.A.; Afrifa, E.K.A.; Aheto, D.W.; Odoi, J.O.; Afitiri, A.R. Geospatial assessment of ecosystem health of coastal urban wetlands in Ghana. Ocean Coast. Manag. 2020, 193, 105226. [Google Scholar] [CrossRef]
  18. Wu, N.; Liu, A.; Wang, Y.; Li, L.; Chao, L.; Liu, G. An assessment framework for grassland ecosystem health with consideration of natural succession: A case study in Bayinxile, China. Sustainability 2019, 11, 1096. [Google Scholar] [CrossRef]
  19. Zhao, X.; Huang, G. Urban watershed ecosystem health assessment and ecological management zoning based on landscape pattern and SWMM simulation: A case study of Yangmei River Basin. Environ. Impact Assess. Rev. 2022, 95, 106794. [Google Scholar] [CrossRef]
  20. Wang, H.; Hou, P.; Jiang, J.; Xiao, R.; Zhai, J.; Fu, Z.; Hou, J. Ecosystem health assessment of Shennongjia National Park, China. Sustainability 2020, 12, 7672. [Google Scholar] [CrossRef]
  21. Wu, B.; Gao, Y.; Luan, C. Developing ecological protection strategies for scenic spots based on ecological sensitivity: A case study of Yaoshan Scenic Spot. J. Gansu Agric. Univ. 2023, 58, 180–188. (In Chinese) [Google Scholar] [CrossRef]
  22. Fang, X.; Liu, S.; Chen, W.; Wu, R. An effective method for wetland park health assessment: A case study of the Guangdong Xinhui National Wetland Park in the Pearl River Delta, China. Wetlands 2021, 41, 44. [Google Scholar] [CrossRef]
  23. Liu, M.; Liang, J.; Shen, Y. Ecological health assessment of National Forest Parks based on the time series analysis of multi-source remote sensing data. J. Nat. Sci. Hunan Norm. Univ. 2024, 47, 55–62. (In Chinese) [Google Scholar]
  24. Luo, H.; Zhao, Q.; Zhang, L.; Gao, C.; Wu, X.; Nie, Y. Landscape health assessment of suburban forest park: A case study based on multiple sampling units and functional characteristics. Forests 2023, 14, 2237. [Google Scholar] [CrossRef]
  25. Wang, Z.; Liu, Y.; Li, Y.; Su, Y. Response of ecosystem health to land use changes and landscape patterns in the Karst Mountainous Regions of Southwest China. Int. J. Environ. Res. Public Health 2022, 19, 3273. [Google Scholar] [CrossRef] [PubMed]
  26. Yuan, Z.; Wang, Q.; Wang, Y.; Li, W.; Gao, J.; Cheng, X.; Zhu, D. Impacts of land use change on ecosystem health in Chongqing under multi-scenario simulation. Acta Ecol. Sin. 2023, 43, 8279–8291. (In Chinese) [Google Scholar] [CrossRef]
  27. Zhao, Q.; Tang, H.; Gao, C.; Wei, Y. Evaluation of urban forest landscape health: A case study of the Nanguo Peach Garden, China. iFor.-Biogeosci. For. 2020, 13, 175–184. [Google Scholar] [CrossRef]
  28. Lausch, A.; Erasmi, S.; King, D.; Magdon, P.; Heurich, M. Understanding forest health with remote sensing-part II-a review of approaches and data models. Remote Sens. 2017, 9, 129. [Google Scholar] [CrossRef]
  29. Qian, J.; Shao, D.; Jiang, W. The scale-effects of different land use types on the distribution pattern of park green space at multiple grid scales: A case study on the main urban area of Nanjing, China. Landsc. Archit. Front. 2024, 12, 7–19. (In Chinese) [Google Scholar]
  30. Wu, A.; Cheng, L.; Zhao, Y. Spatial-temporal pattern evolution of eco-environmental quality and its response to land use change in rapidly urbanized areas: A case study of northern three counties in Langfang City. J. Northwest For. Univ. 2024, 39, 169–176. (In Chinese) [Google Scholar]
  31. Guo, Y.; Ren, Z.; He, X. Multifunctionality of urban forest landscapes and their trade-offs and synergistic. Chin. J. Appl. Ecol. 2025, 36, 837–846. (In Chinese) [Google Scholar] [CrossRef]
  32. Yu, H. An investigation into the theoretical framework governing the multi-dimensional functional coupling of national parks. Acta Ecol. Sin. 2025, 45, 2985–2995. (In Chinese) [Google Scholar] [CrossRef]
  33. Feng, L.; Chen, M.; Bai, Y.; Zhou, L. Urban forest spatial planning in the context of healthy city: A case study of Chengdu Direct Management Area under Tianfu New Area. Landsc. Archit. 2022, 29, 117–123. (In Chinese) [Google Scholar] [CrossRef]
  34. Zhong, J.; Li, X.; Liu, W.; Zhang, F.; Huang, F.; Xu, S.; Xiao, S.; Wang, Q. Effects of urbanization intensity on forest vegetation characteristics and landscape pattern indices in Nanchang. Chin. J. Ecol. 2024, 43, 2285–2294. (In Chinese) [Google Scholar] [CrossRef]
  35. Zhou, L.; Dang, X.; Sun, Q.; Wang, S. Multi-scenario simulation of urban land change in Shanghai by random forest and CA-Markov model. Sustain. Cities Soc. 2020, 55, 102045. [Google Scholar] [CrossRef]
  36. Lin, J.; He, P.; Yang, L.; He, X.; Lu, S.; Liu, D. Predicting future urban waterlogging-prone areas by coupling the maximum entropy and FLUS model. Sustain. Cities Soc. 2022, 80, 103812. [Google Scholar] [CrossRef]
  37. Liang, X.; Guan, Q.; Clarke, K.; Liu, S.; Wang, B.; Yao, Y. Understanding the drivers of sustainable land expansion using a patch-generating land use simulation (PLUS) model: A case study in Wuhan, China. Comput. Environ. Urban 2021, 85, 101569. [Google Scholar] [CrossRef]
  38. Tuson, M.; Yap, M.; Kok, M.; Murray, K.; Turlach, B.; Whyatt, D. Incorporating geography into a new generalized theoretical and statistical framework addressing the modifiable areal unit problem. Int. J. Health Geogr. 2019, 18, 6. [Google Scholar] [CrossRef] [PubMed]
  39. Chen, Y.; Zhang, L.; Ma, R.; Li, G.; Zhuo, Y.; Xu, Z. Spatial match between urban residents’ distribution and green space accessibility and its driving force. Acta Ecol. Sin. 2022, 42, 9971–9980. (In Chinese) [Google Scholar] [CrossRef]
  40. Wang, X.; Shi, Z.; Dai, H.; Xie, H.; Liu, X. Multi-scale evaluation of urban green space fairness based on the modifiable areal unit problem-A case study of Beijing main urban area. J. Northwest For. Univ. 2024, 39, 238–246+264. (In Chinese) [Google Scholar]
  41. Li, X.; Wu, K.; Feng, Z.; Yang, Q.; Qian, J.; Wang, T. Spatial identification and tradeoff/synergy analysis of multiple agricultural land use system functions at grid scale in Henan Province. Trans. Chin. Soc. Agric. Eng. 2023, 39, 242–252. (In Chinese) [Google Scholar]
  42. Ma, C.; Yi, Z.; Yuan, X.; Zhang, S.; Zhang, S. Spatial difference of ecological services and its influencing factors under different scale. Res. Soil Water Conserv. 2025, 32, 389–399. (In Chinese) [Google Scholar] [CrossRef]
  43. Peng, J.; Xu, F. Effect of grid size on habitat quality assessment: A case study of Huangshan City. J. Geo-Inf. Sci. 2019, 21, 887–897. (In Chinese) [Google Scholar]
  44. Wang, Y.; Tang, D.; Yan, X.; Li, S.; Yu, Z.; Wu, L. Research on ecological security evaluation of central Baoshan City based on grid scale. J. Southwest For. Univ. 2021, 41, 80–87. (In Chinese) [Google Scholar]
  45. Yang, K.; Xin, G.; Jiang, H.; Yang, C. Study on spatiotemporal changes of landscape ecological risk based on the optimal spatial scale: A case study of Jiangjin District, Chongqing City. J. Ecol. Rural Environ. 2021, 37, 576–586. [Google Scholar] [CrossRef]
  46. Han, X.; Hu, J.; Liu, C.; Wang, X. Mountain landscape health diagnosis and pattern evolution based on remote sensing: Case study in Xinjiang Tianshan Karajun-Kurdening. Acta Ecol. Sin. 2021, 41, 6451–6465. (In Chinese) [Google Scholar] [CrossRef]
  47. Liu, B.; Tian, Y.; Wang, L.; Xu, D. Spatio-temporal disparity between supply and demand of park green spaces based on grid scale: A case study of the main urban area of Harbin. Landsc. Archit. 2022, 29, 115–121. (In Chinese) [Google Scholar] [CrossRef]
  48. Masek, J.G.; Hayes, D.J.; Hughes, M.J.; Healey, S.P.; Turner, D.P. The role of remote sensing in process-scaling studies of managed forest ecosystems. For. Ecol. Manag. 2015, 355, 109–123. [Google Scholar] [CrossRef]
  49. Zuo, Q.; Zhou, Y.; Li, Q.; Wang, L.; Liu, J.; He, N. Spatial and temporal variations of landscape ecological risk in the mountainous region of southwestern Hubei Province based on the optimal scale. Chin. J. Ecol. 2023, 42, 1186–1196. (In Chinese) [Google Scholar] [CrossRef]
  50. Hong, G.; Wu, X.; Cai, Y. Influence of urban green infrastructure landscape pattern on carbon sequestration performance at the optimal grid unit. Chin. Landsc. Archit. 2023, 39, 138–144. (In Chinese) [Google Scholar] [CrossRef]
  51. Gupta, K.; Kumar, P.; Pathan, S.K.; Sharma, K.P. Urban neighborhood green index-A measure of green spaces in urban areas. Landsc. Urban Plan. 2012, 105, 325–335. [Google Scholar] [CrossRef]
  52. Ge, Y.; Chen, H.; Zhang, M.; Li, X. Area threshold interval of urban forest patches required to maintain the synergy between biodiversity conservation and recreational services: Case study in Beijing, China. Forests 2022, 13, 1848. [Google Scholar] [CrossRef]
  53. Peng, J.; Liu, Y.; Wu, J.; Lv, H.; Hu, X. Linking ecosystem services and landscape patterns to assess urban ecosystem health: A case study in Shenzhen City, China. Landsc. Urban Plan. 2015, 143, 56–68. [Google Scholar] [CrossRef]
  54. Ren, G.; Liu, L.; Li, H.; Yuan, C. Evolution on trade-offs and synergy of multifunction of rural landscape in metropolitan suburbs. Trans. Chin. Soc. Agric. Eng. 2019, 35, 273–285. (In Chinese) [Google Scholar]
  55. Cao, Y.; Xiao, D.; Ou, Y.; Chen, G. Analysis of landscape change drivers in the Ejina natural oasis. Acta Ecol. Sin. 2004, 24, 1895–1902. (In Chinese) [Google Scholar]
  56. Ni, Y.; Liu, Y.; Yan, S.; Ai, X.; Zheng, X. Land use planning based on regional ecosystem health assessment: A case study of Beijing. Chin. Landsc. Archit. 2020, 36, 110–115. (In Chinese) [Google Scholar] [CrossRef]
Land 14 01611 g001
Figure 1. Location of the study areas: (a) Xiqiao Mountain National Forest Park and (b) Yunyong National Forest Park.
Figure 1. Location of the study areas: (a) Xiqiao Mountain National Forest Park and (b) Yunyong National Forest Park.
Land 14 01611 g001
Land 14 01611 g002
Figure 2. Spatial distribution of landscape health in Xiqiao Mountain National Forest Park under different grid scales: ecological protection, recreational use, and comprehensive functions’ spatial distribution at the (a–c) 18 × 18 m grid scale; (d–f) 36 × 36 m grid scale; (g–i) 54 × 54 m grid scale; (j–l) 72 × 72 m grid scale; and (m–o) 108 × 108 m grid scale.
Figure 2. Spatial distribution of landscape health in Xiqiao Mountain National Forest Park under different grid scales: ecological protection, recreational use, and comprehensive functions’ spatial distribution at the (a–c) 18 × 18 m grid scale; (d–f) 36 × 36 m grid scale; (g–i) 54 × 54 m grid scale; (j–l) 72 × 72 m grid scale; and (m–o) 108 × 108 m grid scale.
Land 14 01611 g002
Land 14 01611 g003
Figure 3. Spatial distribution of landscape health in Yunyong National Forest Park under different grid scales: ecological protection, recreational use, and comprehensive functions’ spatial distribution at the (a–c) 17 × 17 m grid scale; (d–f) 33 × 33 m grid scale; (g–i) 50 × 50 m grid scale; (j–l) 66 × 66 m grid scale; and (m–o) 99 × 99 m grid scale.
Figure 3. Spatial distribution of landscape health in Yunyong National Forest Park under different grid scales: ecological protection, recreational use, and comprehensive functions’ spatial distribution at the (a–c) 17 × 17 m grid scale; (d–f) 33 × 33 m grid scale; (g–i) 50 × 50 m grid scale; (j–l) 66 × 66 m grid scale; and (m–o) 99 × 99 m grid scale.
Land 14 01611 g003
Land 14 01611 g004
Figure 4. Landscape health variations analysis under different grid scales in Xiqiao Mountain National Forest Park: (a) ecological protection functions, (b) recreational use functions, and (c) comprehensive functions.
Figure 4. Landscape health variations analysis under different grid scales in Xiqiao Mountain National Forest Park: (a) ecological protection functions, (b) recreational use functions, and (c) comprehensive functions.
Land 14 01611 g004
Land 14 01611 g005
Figure 5. Landscape health variations analysis under different grid scales in Yunyong National Forest Park: (a) ecological protection functions, (b) recreational use functions, and (c) comprehensive functions.
Figure 5. Landscape health variations analysis under different grid scales in Yunyong National Forest Park: (a) ecological protection functions, (b) recreational use functions, and (c) comprehensive functions.
Land 14 01611 g005
Land 14 01611 g006
Figure 6. Forest/non-forest area ratio threshold ranges at different health levels in Xiqiao Mountain National Forest Park at different grid scales.
Figure 6. Forest/non-forest area ratio threshold ranges at different health levels in Xiqiao Mountain National Forest Park at different grid scales.
Land 14 01611 g006
Land 14 01611 g007
Figure 7. Forest/non-forest area ratio threshold ranges at different health levels in Yunyong National Forest Park at different grid scales.
Figure 7. Forest/non-forest area ratio threshold ranges at different health levels in Yunyong National Forest Park at different grid scales.
Land 14 01611 g007
Table 1. Calculation of landscape health evaluation indices.
Table 1. Calculation of landscape health evaluation indices.
IndexCalculation FormulaFormula InterpretationMeaning of the Indicators
Patch area ratio S i = A i ∕ A S i represents the patch area ratio; A i represents the area of the patch type; A represents the area of the basic sampling unit. It reflects the dominant position of patch types.
Patch density P D i = N i ∕ A P D i represents patch density;   N i represents the count of patch types;   A represents the area of the basic sampling unit. It reflects the degree of patch fragmentation. The larger the value, the wider the distribution of patches, indicating a higher degree of fragmentation.
Patch edge density E D i = K i ∕ A i E D i represents the edge density;   K i represents the length of patch edge;   A i   represents the area of the patch type. It reflects the complexity of the patch boundaries; a larger value indicates a more complex patch edge shape.
Ecological protection patch fragmentation F n = M P S × N p − 1 ∕ N c F n represents the patch fragmentation index; N c represents the ratio of the minimum patch area to the basic sampling unit area; M P S represents the ratio of the average patch area to the minimum patch area; N p represents the total number of patches for ecological protection function. It reflects the degree of disruption in the patch landscape structure. A higher value indicates poorer stability in the landscape structure.
Recreational use patch accessibilityRecreational use patch accessibility is quantitatively estimated based on cost distance in ArcGIS and minimum cumulative resistance methods. It reflects the minimum cost distance to reach adjacent patches; the closer the distance, the better the accessibility.
Patch isolation F i = D i ∕ S i
D i = n ∕ A ∕ 2
S i = A i ∕ A 		F i represents the patch separation degree; D i represents the distance index of patch type; S i represents the area ratio of the patch type; A i represents the area of the patch type; A is the area of fundamental sampling units; i is the patch type; n is the total number of patches. It reflects the patch’s dispersion level; a smaller value indicates better connectivity among patch clusters.
Patch diversity H = − ∑ n = 1 k P n ln P n H represents the Shannon–Wiener index; P n represents the patch type; n represents the proportion of the basic sampling unit area it occupies; k represents the total count of patch types. It reflects the complexity of patches; and as the value increases, the diversity and complexity of the landscape structure also increase.
Patch fractal dimension F D p = 2 ln P ∕ 4 ∕ ln A i F D p represents the patch fractal dimension; P represents the patch perimeter; A i represents the area of the patch type. It reflects the deviation of actual patch shapes from standard shapes (circle or square). The closer the value to 1, the simpler the shape, indicating a greater degree of disturbance.
Ecological protection corridor naturalness N = 1 ∕ D N represents the naturalness of the ecological protection corridor; D represents the corridor density. It reflects the naturalness of the corridors; a higher value indicates less disturbance and is more favorable for the survival of wildlife.
Recreational use corridor density D i = L i ∕ A D i   represents the density of recreational use corridors; L i represents the length of corridor type i ; A represents the area of the basic sampling unit. It reflects the degree of corridor fragmentation. A higher value indicates greater landscape fragmentation.
Corridor curvature D q = Q ∕ L D q represents the corridor curvature; Q represents the actual length of the corridor; L represents the straight-line distance from the starting point to the endpoint of the corridor. It reflects the curvature of the corridor, and a higher value indicates longer travel time and greater energy consumption during movement.
Corridor width ratio W R i = W i ∕ l W R i represents the corridor width ratio; W i represents the width of corridor type i ; l represents the side length of the sampling unit. As the width ratio increases, the corridor’s capacity for passage improves, leading to an increased edge, interior species, and enhanced environmental heterogeneity.
Corridor loopiness a = L − V + 1 ∕ 2 V − 5 a represents the corridor loopiness;   L is the count of edges in the network; V   is the count of nodes. It reflects the complexity of the corridor network and characterizes the degree of choice in energy flow, material flow, or species migration routes in the corridor network.
Corridor point-line ratio β = L ∕ V β represents the corridor point-line ratio;   L   represents the count of edges;   V   represents the count of nodes. It reflects the average number of connecting lines for each node in the corridor network, indicating the ease or difficulty of connectivity between nodes.
Corridor connectivity γ = L ∕ 3 V − 2 γ   represents the corridor point-line ratio; L   represents the count of edges; V   represents the count of nodes. It reflects the degree to which all nodes within a corridor network are connected.
Corridor fractal dimension F D c = 2 ln L ∕ 4 ∕ ln A i F D c represents the corridor fractal dimension; L represents the total length of the corridor type; A i represents the area of the corridor type. It reflects the deviation of the actual corridor shape from the standard shape (circle or square). The closer the value to 1, the simpler the shape, indicating a higher degree of disturbance.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Luo, H.; Zhao, Q.; Qian, W.-H.; Zhang, C.; Zhang, L.-Y.; Wu, X.-J. Landscape Health Assessment of Suburban Forest Parks with Different Land Use Intensities and Grid Scales. Land 2025, 14, 1611. https://doi.org/10.3390/land14081611

AMA Style

Luo H, Zhao Q, Qian W-H, Zhang C, Zhang L-Y, Wu X-J. Landscape Health Assessment of Suburban Forest Parks with Different Land Use Intensities and Grid Scales. Land. 2025; 14(8):1611. https://doi.org/10.3390/land14081611

Chicago/Turabian Style

Luo, Hao, Qing Zhao, Wan-Hui Qian, Chi Zhang, Ling-Yu Zhang, and Xiao-Jun Wu. 2025. "Landscape Health Assessment of Suburban Forest Parks with Different Land Use Intensities and Grid Scales" Land 14, no. 8: 1611. https://doi.org/10.3390/land14081611

APA Style

Luo, H., Zhao, Q., Qian, W.-H., Zhang, C., Zhang, L.-Y., & Wu, X.-J. (2025). Landscape Health Assessment of Suburban Forest Parks with Different Land Use Intensities and Grid Scales. Land, 14(8), 1611. https://doi.org/10.3390/land14081611

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop