Urban forests are important as they provide recreation areas and offer ecological services. Both functions determine the status of an urban forest and reflect contradictory aspects of forest tourism development and environment conservation. However, assessment of urban forest health status at a landscape scale remains scarce. Here, we selected the Nanguo Peach Garden, China, as the study area. Urban forest health status at the landscape scale were classified into recreation and eco-conservation services. Sustainability was quantified using the principal component analysis and the Kriging method to map the landscape classification in the study area. With regard to landscape recreation sustainability, some 18.9% of the total study region was classified as “very good”. They were mainly distributed in the north, southwest, and southeast parts of the study area. The central and southeast regions, accounting for 9.5% of the total area, were classified as “very good” for eco-conservation sustainability. Regarding landscape health, the region classified as “very good” accounted for 11.1% of the total study area, and it was mainly distributed in the southern part of the area; the region classified as “very poor” accounted for 16.4% of the total area, and it was located in the northwestern and eastern parts of the study area. With improved landscape health status, the forest/non-forest patch area ratio was increased and the patch number ratio was decreased. A landscape was considered the healthiest when the forest/ non-forest area ratio was 0.65 and the patch number was 0.48. The spatial distribution of landscape recreation sustainability and eco-conservation sustainability differed in the Nanguo Peach Garden, and a close relationship was observed between the landscape health and forest landscape internal structure. Forest/non-forest patch area ratios and patch number ratios were relatively stable and constant, suggesting the urban forest landscapes were healthy. The healthiest forest landscapes were mainly distributed in the forest/non-forest transition zone and the unhealthiest forest landscape was mainly located in a single natural forest.
Urban forests offer valuable forest ecosystem services by enhancing the well-being of human populations and playing a part in establishing livable cities. However, they are among the landscape types that are most severely affected by urbanization (
Previous studies on forest landscape health mainly focused on concepts, standards, and characteristics of landscape health, as well as theoretical issues, such as the establishment of index systems, selection of parameters determining landscape health, and classification of levels for landscape health evaluation (
Considering these knowledge gaps, in this study, we aimed to improve the understanding of evaluation methodology of urban forest landscape health and to provide a theoretical basis for reasonable planning and enhancement of service functions of forest landscapes in future urban forest development projects. We investigated a representative urban forest Nanguo peach garden, located in the Pearl River Delta (China), and analyzed the spatial differences in terms of landscape recreation sustainability and eco-conservation sustainability. Hence, the specific questions addressed in this study were: (i) whether landscape recreation and eco-conservation contradict each other during the construction of urban forest landscape? (ii) how to coordinate the quantity and volume of landscape patches to achieve a healthier urban forest landscape? and (iii) are there any generally healthier urban forest landscape?
The Nanguo Peach Garden is located in Shishan Town, Nanhai District, Foshan City, Guangdong Province, China (113° 04′ - 113° 07′ E, 23° 09′ - 23° 12′ N). It covers a total area of 15.4 km2 and is one of the “Eight New Scenic Areas of Foshan,” comprising Pingdinggang, Jianfengling, and Fenghuanggang. It has become a multi-purpose recreational area that serves several functions, including tourism, recreation, sightseeing, public science education, and biodiversity maintenance. The area has a subtropical monsoon climate characterized by the following: high precipitation in warm seasons; distinct dry and wet seasons; annual average temperature of 22.4 °C; and annual average precipitation of 1305.9 mm concentrated during the rainy season from April to September, accounting for 79% of the total annual precipitation. The soil in the area belongs to the typical lateritic red soil in the southern subtropical zone, and the area is covered by various vegetation types that mainly consist of indigenous broadleaf tree species, including Chinese guger tree (
QuickBird imagery (resolution: 0.6 m) captured on October 2, 2015, and acquired from Google Earth™, was used as the basic data source for this study. Based on GPS control points acquired from field surveys, high-accuracy geometric correction was performed for the QuickBird images using ArcGIS® v. 9.3 to ensure errors were less than 0.5 pixels. The WGS 1984 Web Mercator, a projection used in most web-mapping applications, was adopted as the projected coordinate system. Image classification was mainly based on visual interpretation. To ensure the accuracy of the classification map, a mapping scale of 1:1000 was used. Field surveys and land classification validation were conducted from June to September 2017, and 855 GPS validation points, with approximately 100 points for each landscape type, were acquired via three field surveys over 12 days. The overall accuracy and kappa coefficient of image interpretation were 96.73% and 0.96, respectively, and they fulfilled the requirements of this study.
By referring to the Land Classification Standard of China (Document No. [2001] 255 of the Ministry of Land and Resources) and considering the landscape function characteristics and current landscape types of the study area, the landscape was classified into the following eight types: grassland, waterbody, forest, garden, cropland, constructed facility, road, and bare land.
Landscape structure index is a proxy for landscape function, and it depends on landscape classification (
The current health status of an urban forest landscape generally reflects the effects of previous management strategies. Therefore, the effectiveness of previous management strategies could be assessed by evaluating landscape health based on the current health status. In this study, we combined the factors associated with the landscape composition characteristics of the Nanguo Peach Garden and established various indices that reflected the current status of landscape recreation sustainability and eco-conservation sustainability to determine landscape health. Landscape recreation sustainability was mainly composed of factors related to landscape recreation, while eco-conservation sustainability was mainly composed of factors related to biodiversity function maintenance. Specifically, four patch indices and four corridor indices were used to evaluate the landscape recreation sustainability and eco-conservation sustainability (
The patch area ratio refers to the ratio of the patch area to the sampling area, while the landscape recreation patch accessibility refers to the least cost distance to the nearest landscape recreation patch (
where
Eco-conservation patch fragmentation is the degree of fragmentation of patches with eco-conservation functions, and it was calculated using the following equation (
where
Patch diversity is a measure of area and proportion of patches in each patch type. It reflects the complexity of patch types in the landscape. The formula used to determined patch diversity was as follows (
where
The corridor area ratio is the ratio of the corridor area to the sampling area. The corridor network was distributed over the landscape matrix, and its connectivity exerts a significant effect on the normal functioning of landscape recreation and eco-conservation patches. In the present study, the corridor circuitry index (α), line-node ratio (β), and connectivity index (γ) were selected to describe corridor structures. The corridor circuitry index α indicates selectivity in the pathways for energy flow, material flow, or species migration, and it characterizes the degree of complexity of a network. The line-node ratio β indicates the average number of connecting lines for each node in the network, that is, the degree of difficulty in connecting one node with other nodes. The corridor network connectivity index γ indicates the degree of connectedness of all nodes in a network (
where
where
where
Based on the average areas of various landscape types, basic sampling units (size: 340 × 340 m) were established to collect spatial data, and regions having an area less than 50% of the basic sampling unit area were omitted, yielding 147 sampling units. The calculation and spatialization of all these indices were performed using the “Spatial Analyst” module in ArcGIS®. Extreme value normalization was performed for all indices to scale the index values to the [0, 1] range, and the normalized index values were classified into the following five rankings: very poor [0-0.2]; poor (0.2-0.4]; fair (0.4-0.6]; good (0.6-0.8]; and very good (0.8-1.0]. Finally, the Spatial Analyst module in ArcGIS® was used to perform spatial visualization of the various levels of the landscape recreation sustainability indices and eco-conservation sustainability indices for each basic sampling unit.
To reduce subjectivity in the evaluation process, the principal component analysis (PCA) method was selected to determine the weight of the landscape recreation sustainability and eco-conservation sustainability indices. Before the PCA, all indices were subjected to extreme value normalization to eliminate the dimensional effects. Based on the PCA results, the top three dominant component scores of the landscape recreation sustainability and eco-conservation sustainability indices were selected to compute the component score coefficient matrix (
The landscape recreation sustainability and eco-conservation sustainability evaluation values of each basic sampling unit were calculated by multiplying the normalized value of each evaluation index (
The landscape health status of a forest landscape is positively correlated with landscape recreation sustainability and eco-conservation sustainability, respectively. Besides, a symbiotic relationship exists between landscape recreation sustainability and eco-conservation sustainability, that is, a healthy forest landscape can only exist with a balanced development of sustainable recreational spaces and those with sustainable eco-conservation efforts (
where
The results of the comprehensive evaluation were classified according to the five rankings described earlier (very poor to very good), and the Spatial Analyst module in ArcGIS® was used for the spatial visualization of the various levels of the landscape health evaluation results for each basic sampling unit.
The spatial distribution characteristics of the various indices calculated using the landscape recreation sustainability evaluation method (
The spatial distribution characteristics of various indices calculated using the eco-conservation sustainability evaluation method (
The landscape recreation sustainability evaluation levels of the study area (
The eco-conservation sustainability evaluation levels of the study area (
The landscape health of the Nanguo Peach Garden was determined by evaluating both landscape recreation sustainability and eco-conservation sustainability. The spatial distribution of landscape health in the study area was obtained based on the spatial distribution characteristics of both aspects and the equation for the comprehensive evaluation of landscape health (
A complex relationship exists between the landscape health, landscape recreation, eco-conservation sustainability, and internal landscape factors of a forest. For instance, with the enhancement of landscape health status from “very poor” to “very good,” the forest/non-forest patch area ratio in the Nanguo Peach Garden showed a generally decreasing trend, and the “very good” rating of landscape health was achieved at a patch area ratio of 0.65 (
As shown in
Information on sustainability patterns are essential tools to evaluate urban forest landscape health (
The evaluation of forest landscape health performed in this study provided a snapshot of the situation in the Nanguo Peach Garden; however, spatial differences in landscape health are dynamic in nature (
The scientific arrangement of landscape factors in urban forest is a prerequisite for the healthy development of forests (
It has been reported that there are certain correlations between landscape types and landscape health (
The evaluation of differences in landscape health is, to a certain extent, dependent on the spatial scale (
Significant differences were observed in the spatial distribution of landscape recreation sustainability and eco-conservation sustainability in the Nanguo Peach Garden, and their interaction demonstrated an overall inverse relationship with a certain degree of convergence. In general, regions with a “good” rating for eco-conservation sustainability had a “poor” rating for landscape recreation sustainability, indicating the existence of a mutually constraining relationship between the two aspects. However, a mutual relationship could also exist within regions that have a good balance of recreational services and sustainable eco-conservation, as observed in the peach garden in the southwestern section of the study area. Hence, both landscape recreation sustainability and eco-conservation sustainability should be enhanced to effectively improve landscape health of urban forests and consequently achieve healthy development of urban forests. Scientific planning and arrangement, clear functional orientation, and appropriate management measures could contribute to the enhancement of both landscape recreation and eco-conservation sustainability of urban forests.
Close relationships existed between the landscape health and internal landscape factors of a forest. Within a certain range, a decrease in the forest/non-forest patch area ratio and an increase in the forest/ non-forest patch number ratio will improve forest landscape health. Thus, a balance between forest/non-forest landscape area and patch number is necessary to maintain the optimum landscape health in urban forests. The regions with better landscape health statuses were mostly concentrated in the transition zones between the forest and non-forest landscapes (
ZQ conceived and designed the experiments, analyzed the data, and wrote and revised the manuscript. THH performed the experiments. GCJ supervised the study and revised the manuscript. WYH operated the ArcGIS® 9.3 software and conducted the field experiments. All authors critically contributed to the manuscript and provided the final approval for publication.
We are grateful to all the participants who generously shared their opinions. We also thank Academician Shouzheng Tang from Chinese Academy of Engineering for his thorough, insightful, and affirmative comments. We are thankful to Editage (http://www.editage.cn/http://www.editage.cn/) for English language editing of the manuscript.
This study was supported by the Forestry Science and Technology Innovational Specific Project of Guangdong Province (grant no. 2017KJCX034) and the Special Fund for Forest Scientific Research in the Public Welfare (grant number no. 201404301).
Map of landscape classification in the Nanguo Peach Garden (refer to
Spatial distribution indices of landscape recreation sustainability. (a) Landscape recreation patch area ratio; (b) Landscape recreation patch accessibility; (c) Landscape recreation patch separation; (d) Landscape recreation patch diversity; (e) Landscape recreation corridor area ratio; (f) α index of landscape recreation corridor; (g) γ index of landscape recreation corridor; (h) β index of landscape recreation corridor.
Spatial distribution indices of eco-conservation sustainability. (a) Eco-conservation patch area ratio; (b) Eco-conservation patch separation; (c) Eco-conservation patch fragmentation; (d) Eco-conservation patch diversity; (e) Eco-conservation corridor area ratio; (f) α index of eco-conservation corridor; (g) γ index of eco-conservation corridor; (h) β index of eco-conservation corridor.
Spatial distribution of sustainability level. (a) Landscape recreation sustainability level; (b) Eco-conservation sustainability level.
Spatial distribution of landscape health in the study region.
Landscape classification system of the Nanguo Peach Garden in Guangdong Province, China.
Landscape type | Secondary classification | Description |
---|---|---|
Grassland | Waste grassland | Land dominated by natural herbaceous plants and weeds |
Grass lawn | Artificial grassland or bush fallow used for aesthetics and recreational purposes | |
Waterbody | River and canal | A natural or artificial linear waterway |
Reservoir and pond | Land surface occupied by a broad waterbody; |
|
Forest | Human habitat forest | Trees around settlements planted to maintain good Feng Shui |
Corridor forest | Trees aside roads, railway tracks, rivers, or canals | |
Mountain forest | Natural or artificial forest growing in low mountains with a canopy density of >30% | |
Garden | Fruit garden | Garden dominated by intensively managed herbaceous plants and fruit trees; |
Tea garden | Garden mainly used for growing tea | |
Cropland | Paddy field | Arable land planted with aquatic crops; |
Vegetable field | Land planted mainly with vegetables; |
|
Constructed facilities | - | Artificial facilities including towns, rural settlements, and other artificial buildings |
Road | - | Land for building road or other traffic facilities |
Bare land | - | Land with a canopy density of <5%; |
Classification levels of landscape functions in urban forests.
Primary | Secondary | Tertiary | Description |
---|---|---|---|
Patch | Eco-conservation patch | Natural forest patch | In an eco-conservation area, characterized by a natural or quasi-natural forest landscape, complex landscape structure, rich ecosystem, diverse vertical structure, and horizontal structure of the area; |
Natural non-forest patch | In an eco-conservation area, characterized by a waterbody with a natural bank and natural grassland or forested landscape; |
||
Landscape recreation patch | Artificial forest patch | In a landscape recreational area, characterized by an artificial forest with a simple structure and simple ecosystem structure; |
|
Artificial non-forest patch | In a landscape recreational area, characterized by artificial hard or semi-hard surface structures and crop plots, and ornamental lawns; |
||
Corridor | Eco-conservation corridor | Natural corridor | Characterized by a complex landscape structure and rich ecosystem. The ecological functions involve the movement of species and material, and the ecological function of water storage and flood control; |
Landscape recreation corridor | Artificial corridor | Provides a spatial contact channel for visitors and managers in the landscape recreational patch; |
Indices of landscape health assessment. (+): The higher the value, the better the sustainability; (-): the lower the value, the poorer the sustainability.
Evaluation index | Landscape recreation sustainability index | Eco-conservation sustainability index |
---|---|---|
Patch | X1: Landscape recreation patch area ratio + | Y1: Eco-conservation patch arearatio + |
X2: Landscape recreation patch accessibility + | Y2: Eco-conservation patch fragmentation - | |
X3: Landscape recreation patch separation - | Y3: Eco-conservation patch separation - | |
X4: Landscape recreation patch diversity + | Y4: Eco-conservation patch diversity + | |
Corridor | X5: Landscape recreation corridor area ratio + | Y5: Eco-conservation corridor area ratio + |
X6: α index of landscape recreation corridor + | Y6: α index of eco-conservation corridor + | |
X7: γ index of landscape recreation corridor + | Y7: γ index of eco-conservation corridor + | |
X8: β index of landscape recreation corridor + | Y8: β index of eco-conservation corridor + |
Dominant component score coefficient matrix of landscape recreation and eco-conservation sustainability. For details of “X1-X8” and “Y1-Y8,” see
Landscaperecreation index | Scores of dominant factors | Eco-conservationindex | Scores of dominant factors | ||||
---|---|---|---|---|---|---|---|
1 | 2 | 3 | 1 | 2 | 3 | ||
X1 | 0.018 | 0.535 | -0.035 | Y1 | 0.034 | 0.519 | 0.012 |
X2 | -0.044 | -0.176 | 0.011 | Y2 | -0.017 | 0.079 | 0.114 |
X3 | -0.055 | -0.050 | 1.017 | Y3 | -0.008 | 0.014 | 1.024 |
X4 | -0.003 | 0.537 | -0.039 | Y4 | 0.028 | 0.543 | 0.010 |
X5 | -0.293 | 0.026 | -0.035 | Y5 | -0.163 | 0.100 | -0.021 |
X6 | 0.547 | 0.007 | -0.030 | Y6 | 0.736 | 0.016 | 0.002 |
X7 | 0.186 | 0.018 | -0.025 | Y7 | -0.221 | 0.053 | -0.034 |
X8 | 0.463 | 0.000 | -0.025 | Y8 | 0.516 | 0.031 | -0.010 |
Eigenvalue | 3.229 | 2.251 | 0.961 | Eigenvalue | 3.200 | 1.731 | 1.168 |
Contribution rate | 40.36 | 28.14 | 12.01 | Contribution rate | 40.01 | 21.64 | 14.61 |
Weight coefficient | 0.501 | 0.350 | 0.149 | Weight coefficient | 0.525 | 0.284 | 0.192 |
Ratio of the area and number of forest/non-forest patches in each grade region in Guangdong Province, China. (LH) Landscape health; (EC) Eco-conservation; (LR) Landscape recreation.
Level | Forest/non-forest patch area ratio | Forest/non-forest patch number ratio | ||||
---|---|---|---|---|---|---|
LH | EC | LR | LH | EC | LR | |
Very poor | 0.92 | 0.46 | 5.20 | 0.32 | 0.45 | 0.31 |
Poor | 0.57 | 0.40 | 0.84 | 0.33 | 0.29 | 0.33 |
Fair | 0.67 | 0.63 | 0.55 | 0.35 | 0.40 | 0.33 |
Good | 0.67 | 1.33 | 0.51 | 0.38 | 0.36 | 0.34 |
Very good | 0.65 | 1.85 | 0.49 | 0.48 | 0.45 | 0.47 |