Next Article in Journal
Transcriptional Regulations and Hormonal Signaling during Somatic Embryogenesis in the Coconut Tree: An Insight
Next Article in Special Issue
Pocket Parks: A New Approach to Improving the Psychological and Physical Health of Recreationists
Previous Article in Journal
Does Soundscape Perception Affect Health Benefits, as Mediated by Restorative Perception?
Previous Article in Special Issue
Using a Novel Green Index to Support Ecosystem Services in a Megacity
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Forests
Volume 14
Issue 9
10.3390/f14091799
forests-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 thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Impact of Forest Parkification on Color Authenticity

by
Wenjing Han
1,2,
Chang Zhang
1,2,*,
Cheng Wang
1,2,
Songsong Liu
3,4,
Decai Shen
3,4,
Haiqi Zhou
3,4 and
Dan Han
5
1
Research Institute of Forestry, Chinese Academy of Forestry, Beijing 100091, China
2
Key Laboratory of Tree Breeding and Cultivation and Urban Forest Research Centre, National Forestry and Grassland Administration, Beijing 100091, China
3
Dongguan Institute of Forestry Science, Dongguan 523006, China
4
Forest Ecosystem Research Station in City Cluster of the Pearl River Estuary, Dongguan 523006, China
5
School of Design and Architecture, Zhejiang University of Technology, Hangzhou 310023, China
*
Author to whom correspondence should be addressed.
Forests 2023, 14(9), 1799; https://doi.org/10.3390/f14091799
Submission received: 26 June 2023 / Revised: 20 August 2023 / Accepted: 1 September 2023 / Published: 3 September 2023
(This article belongs to the Special Issue Urban Forestry and Sustainable Cities)
Browse Figures
Versions Notes

Abstract

:
Preserving the authenticity of forest colors is essential to highlight regional characteristics and promote the sustainable development of forest landscapes. However, the factors and mechanisms influencing forest color remain unclear. We quantified 1422 forest color images from 43 parks across seven biogeographic regions in China to capture the forest color composition among regions. A generalized linear mixed-effects model was used to analyze the effects of meteorological and anthropogenic disturbance factors on forest color. Meteorological factors included accumulated sunshine hours, average temperature, accumulated precipitation, frost-free period, average wind speed, and average air quality index. Anthropogenic disturbance factors included park feature indicators (area, elevation, and perimeter-area ratio) and human activity indicators (distance to urban areas, building density, and road density). We calculated p-values and relative effect estimates to determine the sensitivity and degree of sensitivity of color to each factor. The results indicated the following: (1) forest color composition varied significantly among different regions in China with variations observed particularly in the proportions of primary (green), secondary (yellow and yellow-green), and accent colors (orange and blue-green); (2) forest colors were sensitive to all meteorological factors; (3) orange, yellow, purple, and red were all sensitive to anthropogenic disturbance factors; and (4) forest accent colors were more strongly influenced by anthropogenic disturbance factors, particularly park features. To protect the authenticity of forest colors, it is necessary to avoid excessive borrowing of forest color schemes from different regions, control park features, reduce building area within the park buffer zones, and optimize park tourism routes.

1. Introduction

Forest color authenticity refers to the extent to which various colors in the forest landscape remain in their native state despite natural and human disturbances [1]. Protecting forest color landscape authenticity is essential to show regional characteristics and achieve sustainable development of forest color landscapes. It can provide better long-term and stable landscape services for people [2]. National forest parks were established to protect the authenticity and integrity of representative large-scale forest ecosystems [3]. However, with the increasing ecological and cultural needs of people, national forest parks integrate scientific research, education, and leisure functions, which inevitably involve diversified human activities, such as monitoring, protection, and recreational experiences, leading to increased anthropogenic disturbance in the forest landscape [4]. Increasing anthropogenic disturbance can weaken or even eliminate the authenticity of forest colors [5]. Therefore, factors that influence forest colors should be identified to better protect and restore authentic forest color landscapes.
Forest colors originate from changes in the growth and development of plant organs and adaptive changes resulting from environmental variability [6]. Climate is considered the primary environmental factor affecting forest color changes [7], including light, temperature, humidity, precipitation, and air quality [8]. Plant coloration depends on the relative pigment content of the plant body [9]. Previous studies have mainly explored meteorological effects on plant pigment content from the perspective of plant organs or individual levels. Generally, longer sunshine hours favor chlorophyll synthesis [10], resulting in a green appearance. When plants are subjected to stressors, such as drought and low temperatures, the chlorophyll in their bodies decomposes easily [11], while carotenoids and anthocyanins remain relatively stable [12], leading to an orange or yellow appearance. In addition, some studies have examined the impact of meteorological factors on the quantity and duration of forest colors at the landscape scale during seasons when color changes are particularly pronounced (spring and autumn) [13,14]. Furthermore, forest color landscapes are affected by a combination of temperature, photoperiod, and precipitation [15,16]. However, the impact of each meteorological factor on forest colors varies according to the season. For example, while higher temperatures favor the emergence of new leaves in spring, resulting in an increased proportion of red in the forest, they negatively affect leaf aging and coloring in autumn, leading to a reduced proportion of red [14]. Therefore, to predict the composition of forest landscape colors under changing climatic conditions, the impact of meteorological factors on the annual forest color index requires further exploration.
With the acceleration of urbanization and development of forest tourism, human activities are increasingly affecting forest color landscapes [17,18]. The expansion of human activity has reduced forest areas and increased the fragmentation of forest color landscapes [19]. Additionally, human activities have indirectly affected forest color landscapes by altering the regional climate [20] and hydrological cycles [21]. Most studies suggest that human activities have caused the start of the forest growing season to advance and become delayed by the end of the growing season [22,23,24], resulting in a decreased proportion of non-green colors in forests. Considering that they are major tourist areas, forest parks may experience excessive disturbance of forest ecosystems and degradation of tourism resources due to human activities [25]. Previous studies showed that the larger the park area and the lower the elevation, the more concentrated the visitor activity paths within the park [26,27], thus increasing the degree of interference with local forest color landscapes. The complexity of the park shape [28] and degree of land-use changes [29,30,31] lead to more severe fragmentation of forest color landscapes. The more fragmented the forest color landscape, the more threatened is the forest color authenticity [30]. In summary, human activities influence the authenticity of forest colors; however, the mechanism through which anthropogenic disturbance factors affect each forest color remains unclear.
Assuming that climate change and anthropogenic disturbances threaten the authenticity of forest color landscapes after forest park development, this study aimed to explore the key factors affecting the authenticity of forest colors, thereby offering guidance for protecting and restoring forest color landscape authenticity. Recent developments in big data technology have made it possible to conduct nationwide research on forest colors in China. We selected 43 national forest parks in seven biogeographical regions as research objects and obtained forest images of each park from Weibo, a popular social media platform in China [32]. We then quantified the color information of each image [33] to determine the characteristics of forest color composition in various regions after forest park establishment. Subsequently, we collected meteorological information (temperature, precipitation, etc.) and anthropogenic disturbance factors (park features, changes in land use inside and outside the park, etc.) for each park to investigate the impact of these factors on forest color. The main issues addressed in this study were as follows: (1) what is the forest color composition in different regions, and which colors show significant differences among them; (2) which environmental factors (meteorological and anthropogenic) do forest color sensitivity depend on; and (3) what is the degree of sensitivity of each forest color to meteorological and anthropogenic disturbance factors?
We hypothesized that: (1) forest color composition changes with parkification; (2) temperature, sunshine hours, and precipitation are critical meteorological factors affecting forest color, whereas lower temperatures, shorter sunshine hours, and less precipitation will decrease green percentages and increase orange and yellow percentages in the forest; (3) both park features and human activities impact forest color authenticity; and (4) different forest colors exhibit varying degrees of sensitivity to meteorological and anthropogenic disturbance factors. Validating these hypotheses could help assess the impacts of climate change and human activities on the forest color landscape composition. It can guide forest landscape planning, design, and sustainable management of parks and other urban green spaces.

2. Methods

2.1. Study Area and Study Population

2.1.1. Study Area

China is a vast country with diverse climates and vegetation types [34], leading to different forest color compositions among regions. To determine the influence of forest park establishment on color authenticity in each region, we divided China into seven biogeographic regions [35] (henceforth, regions) based on climate and vegetation distribution: northeast China, north China, northwest China, Mongolian Plateau, Tibetan Plateau, eastern Himalayas, and southeast China [36,37]. Using the inventory of national forest parks (henceforth, parks) in China (https://www.maigoo.com/goomai/167110.html (accessed on 17 February 2023)), we randomly selected one to two parks from each provincial administrative region for a total of 43 parks from diverse regions (Figure 1).

2.1.2. Study Object

Forest image acquisition. We used a focused web crawler [38] to collect images between 1 December 2018 and 30 November 2019 with Weibo as the data source and the name of each park as the keyword (Figure A1). A total of 187,149 park images were obtained.
Forest image screening. First, we removed images dominated by non-forest elements such as human portraits, streams, buildings, food, and weather. Then, we manually screened the park images based on the principle that (1) the forest part comprised at least 60% of the total image area and (2) the color situation matched the physical object [33]. Finally, 1422 high-quality images of forest landscapes from various regions were obtained (Table A1). There were 52 images from northeast China, 224 from north China, 122 from northwest China, 551 from southeast China, 118 from the eastern Himalayas, 268 from the Mongolian Plateau, and 87 from the Tibetan Plateau. There are significant seasonal differences in forest color landscapes [39]. To avoid differences between images and seasonal meteorological data, we divided the seasons according to the upload time and content of each image (Table A2).
Forest image processing. We uniformly applied adaptive gamma correction [40] and automatic color equalization [41] to the forest color images, which reduced the problem of inconsistent image quality caused by the image equipment and environment. To eliminate the impact of non-forest elements on the color extraction results, we used Photoshop to remove color components unrelated to the forest, such as buildings, sky, and water bodies [42].

2.2. Forest Color Classification and Analysis

As the primary basis for distinguishing different colors [43], hue (H) is relatively less affected by external factors, such as light and viewing distance, than saturation and value [44]. Therefore, in this study, color classification was performed based on a range of hue values. We used a secondary K-means clustering approach for color classification [33] with the following steps: (1) determine the optimal k-value using the elbow method [45] to cluster the color information of each forest image, and (2) set the k-value to eight to cluster the H-value of the clustered colors of each image [46]. The secondary clustering results were then adjusted according to human eye color sensitivity [42,47]. Finally, we quantified the forest colors as orange [0, 25] or (345, 360], yellow (25, 55], yellow-green (55, 75], green (75, 140], blue-green (140, 165], blue (165, 220], purple (220, 290], and red (290, 345] (Figure 2), and performed the color index calculation:
H i = A H i A n × 100 %
where A H i is the total number of pixels occupied by hue i, and A n is the total number of pixels in the plant part of that image.

2.3. Environmental Factor Selection and Data Acquisition

2.3.1. Meteorological Factors

Meteorological conditions are the main factors causing changes in plant color [7]. We obtained meteorological data for each park from 1 December 2018 to 30 November 2019 from the Chinese historical weather website (Table A3). From the weather website (https://rp5.ru (accessed on 25 February 2023)), we collected meteorological data for the closest meteorological station (Table A4) for each park, including daily average temperature (T), daily average wind speed (F), and daily precipitation (R). The China Meteorological Information Service Center (https://data.cma.cn/ (accessed on 25 February 2023)) and the air quality historical data website (http://www.aqistudy.cn/historydata/ (accessed on 25 February 2023)) provided the monthly cumulative sunshine hours (SH) and average air quality index (AQI) for each park’s prefecture-level city (Table A4). We calculated the average or cumulative values of the meteorological variables for each season (Table A2). The frost-free period (FFP) was obtained from the official website of each park (accessed on 10 March 2023).

2.3.2. Anthropogenic Disturbance Factors

Forest parkification refers to the process of transforming a natural forest area into a designated forest park through anthropogenic disturbances. This involves modifying the landscape to create recreational spaces while aiming to preserve the natural resources within the park’s boundaries. The primary objectives of forest parkification include providing areas for leisure activities, cultural events, scientific research, and educational opportunities within a controlled and managed environment [3]. Anthropogenic disturbances affect the natural environment and ecosystems owing to human products, presence, and other social activities [48,49]. To measure the extent to which parkification impacts forest color landscapes, we selected six anthropogenic disturbance variables from park features and human activities.
Park features, including park size, shape, and location, significantly affect vegetation diversity [50]. Therefore, we considered that park features also affected the authenticity of forest color and selected the following three variables (Table 1 and Table A3). (1) The area of the park (AREA), which reflects the horizontal extent of human activities, was obtained from the official website of each park (accessed on 10 March 2023). (2) The perimeter area ratio of the park (PAR) describes the complexity of the park shape; the more complex the park shape, the stronger the edge effect [28]. This was calculated using park perimeter/park area × 100. The boundaries of each park were obtained from the Baidu map (https://map.baidu.com (accessed on 25 March 2023)). (3) Elevation of the park (ELEV), which conveys the vertical extent of human activity, was obtained from each park’s official website (accessed on 10 March 2023).
Human activities refer to the development, utilization, and protection of the natural environment during human social development [51]. Buildings, roads, and towns, as products of human social development, have increased human influence on forest color landscapes [52,53]. Three human activity variables were selected for this study (Table 1). (1) The distance to the nearest city center (D) reflects the impact of urbanization on forest color [54]. We identified the city center nearest to the park using ArcGIS 10.6 and obtained the minimum linear distance to the park. (2) Building density (BD), which reflects the intensity of human modification of nature [53], was calculated as the building area/3 km buffer area of the park × 100. We obtained building data from the RiverMap 4.1 software. (3) The road network density (RD) reflects park [55]. The higher the road network density, the more significant the impact of human activities on the forest color landscape [52]. It was calculated as the road length/3 km buffer area of the park × 100. We obtained road network data from the RiverMap 4.1 software.

2.4. Data Analysis

All statistical analyses and mapping were performed using R, version 4.1.3 (R Core Team, Vienna, Austria).
The forest color index is the ratio of the data converted from count data. The Kruskal–Wallis test was used to analyze the differences in forest color indices among regions with p < 0.05 considered statistically significant. In addition, we divided forest color into three categories: primary, secondary, and accent colors (color indices of 30%–100%, 10%–30%, and 0%–10%, respectively) [57]. In this study, we used the color composition of unparked forests in the same biogeographic region as baseline data for color authenticity. The color composition of unparked forests was obtained from previous studies.
A generalized linear mixed model with a binomial error structure was used to investigate the influence of environmental factors on forest color indices. We fitted the model to each of the seven color indices (except the green index) using meteorological and human activity indicators as fixed effects and parks as random effects. The model was constructed as follows:
Model   < - glmer ( H i ~ BD + RD + D + AREA + ELEV + SH + T + R + F + AQI + FFP + ( 1 | park ) , data = data ,   family = “ binomial ” ,   weights = A H i )
where H i is the color index of hue i, park is the code of each park, and A H i is the total number of pixels occupied by hue i.
We normalized the z-scores of all explanatory variables for fixed effects to compare the model parameter estimates. Collinearity among variables was tested using the variance inflation factor (VIF), ensuring that all explanatory variables satisfied the criterion of VIF < 10 [58] (Table A5). The heterogeneity of the residuals was tested using the Durbin-Watson test [59]. R2m (marginal R2, accounting for fixed effects) and R2c (conditional R2, accounting for full model effects) were calculated to estimate the goodness of fit [60]. The significance of each explanatory variable was assessed using a χ2 test [61], taking p < 0.05 as statistically significant. The p-value was used to determine whether each color was sensitive or insensitive to the variable. The sensitivity or insensitivity of the forest color was employed to determine whether its authenticity was affected.
To evaluate the relative importance of the predictors as drivers of forest color indices, we calculated the effect of the parameter estimates for each explanatory variable relative to all parameter estimates in the model [62]. The following three identifiable variance fractions were examined: (1) meteorological variables, (2) park feature variables, and (3) human activity variables. Relative effect estimates were used to determine the sensitivity of each forest color to meteorological and anthropogenic disturbance factors (park features and human activities). The sensitivity of the forest color was used to measure the degree to which its authenticity was affected.

3. Results

3.1. Forest Color Composition by Region after the Establishment of Forest Parks

From a national perspective, green (36.30%) was the primary color of forest landscapes, yellow (26.94%) and yellow-green (19.04%) were secondary colors, and the remaining colors were accent colors (Figure 3). Orange (H1), yellow (H2), yellow-green (H3), green (H4), and blue-green (H5) indices differed significantly among regions (p < 0.01). Orange was a secondary color in the forest color landscape in northeast China (20.86%) and north China (11.04%) and an accent color in other regions. Yellow was the primary color in the forest landscape of the Mongolian Plateau (45.96%) and northeast China (38.10%), and a secondary color in other regions. Yellow-green was the primary color in the forest landscape of the Tibetan Plateau (34.89%), an accent color in northeast China (8.66%), and a secondary color in other regions. Green was a secondary color in the forest landscape of the Mongolian Plateau (25.02%) and northeast China (20.90%), and a primary color in other regions. Blue-green was a secondary color in the forest landscape of north China (10.20%), and an accent color in other regions. The blue (H6), purple (H7), and red (H8) indices were not significantly different among regions (p > 0.05), and all showed accent colors in the forest landscape (Table 2).

3.2. Effect of Meteorological Factors on the Forest Color Authenticity

The sensitivity of different colors to meteorological factors varied (Figure 4, Table A6). T, R, F, and FFP had significant negative effects (p < 0.05) on H1, indicating that orange was sensitive to these four meteorological factors. The proportion of orange in the forest decreased with higher temperatures, more precipitation, higher wind speed, and a longer frost-free period. However, SH and AQI had no significant effects on H1 (p > 0.05) (Figure 4A). SH, T, R, F, and AQI had significant effects on H2 and H3 (p < 0.01), indicating that these indices were sensitive to these five meteorological factors. The proportion of yellow in the forest increased with longer sunshine hours, lower temperatures, less precipitation, lower wind speeds, and better air quality (Figure 4B). The proportion of yellow-green in the forest increased with shorter sunshine hours, higher temperatures, lower precipitation, higher wind speeds, and better air quality (Figure 4C). FFP had no significant effects on H2 and H3 (p > 0.05). SH, R, and AQI had significant effects on H5 (p < 0.01), indicating that blue-green was sensitive to these three meteorological factors. The proportion of blue-green in the forest increased with shorter sunshine hours, increased precipitation, and poorer air quality. Nevertheless, T, F, and FFP had no significant effects on H5 (p > 0.05) (Figure 4D). SH, T, F, and FFP had significant effects (p < 0.05) on H6, indicating that blue was sensitive to these four meteorological factors. The proportion of blue in the forest increased with longer sunshine hours, lower temperatures, higher wind speeds, and longer frost-free periods. R and AQI had no significant effect on H6 (p > 0.05) (Figure 4E). SH, R, F, AQI, and FFP had significant effects on H7 (p < 0.05), indicating that purple was sensitive to these five meteorological factors. The proportion of purple in the forest increased with shorter sunshine hours, less precipitation, higher wind speed, poorer air quality, and shorter frost-free periods. However, T had no significant effect on H7 (p > 0.05) (Figure 4F). All meteorological factors had significant effects (p < 0.01) on H8, indicating that red was sensitive to SH, T, R, F, AQI, and FFP. The proportion of red in the forest increased with shorter sunshine hours, higher temperatures, less precipitation, higher wind speeds, poorer air quality, and shorter frost-free periods (Figure 4G).

3.3. Effect of Anthropogenic Disturbances on Forest Color Authenticity

Anthropogenic disturbance factors from forest parkification only showed significant effects (p < 0.05) on the H1, H2, H7, and H8 (Figure 4, Table A6). AREA, ELEV, and BD had significant effects (p < 0.05) on H1, indicating that orange was sensitive to these three anthropogenic disturbance factors. The proportion of orange in the forest increased with larger park area, lower park elevation, and higher building density. However, PAR, RD, and D had significant effects (p > 0.05) on H1 (Figure 4A). ELEV had a significant negative effect on H2 (p < 0.05), indicating that yellow was sensitive to the elevation of the park. The proportion of yellow in the forest decreased with an increase in park elevation. AREA, PAR, BD, RD, and D had no significant effects (p > 0.05) on H2 (Figure 4B). Park feature factors (AREA, ELEV, and PAR) had significant effects (p < 0.05) on H7, indicating that purple was sensitive to park feature factors. The proportion of purple in the forest increased with a larger area, lower elevation, and simpler shape of the park. Human activity factors (BD, RD, and D) had no significant effects (p > 0.05) on H7 (Figure 4F). AREA, ELEV, PAR, BD, and RD had significant effects on H8 (p < 0.01), indicating that red was sensitive to these five artificial disturbance factors. The proportion of red in the forest increased with larger area, lower elevation, more complex shape of the park, less building density, and more road network density within the park buffer. D had no significant effect on H8 (p > 0.05) (Figure 4G).

3.4. Degree of Impact of Meteorological and Anthropogenic Disturbances on Forest Color Authenticity

The fixed effect R2m showed (Figure 4) that environmental factors explained the H1, H2, H7, and H8 (0.688, 0.450, 0.767, and 0.622, respectively) better than the H3, H5, and H6 (0.154. 0.144, and 0.269, respectively), indicating that orange, yellow, purple, and red were more sensitive to environmental factors. From each variable group, the relative effect estimates of anthropogenic disturbances from forest parkification for H1 (62.44%), H5 (58.31%), H6 (58.09%), H7 (51.56%), and H8 (58.02%) were greater than those of meteorological factors (37.56%, 41.69%, 41.91%, 48.44%, and 41.98%, respectively), indicating that orange, blue-green, blue, purple, and red were more sensitive to anthropogenic disturbances than meteorological factors. The relative effect estimates of meteorological factors for H2 (60.30%) and H3 (63.97%) were greater than those for anthropogenic disturbances (39.70% and 36.03%, respectively), indicating that yellow and yellow-green were more sensitive to meteorological factors than to anthropogenic disturbances. Among the anthropogenic disturbance factors, the relative effect estimates of park features for H1 (33.63%), H2 (27.85%), H5 (32.71%), H6 (46.76%), H7 (49.21%), and H8 (44.39%) were greater than those of human activities (28.81%, 11.86%, 25.61%, 11.34%, 2.34%, and 13.63%, respectively), indicating that orange, yellow, blue-green, blue, purple, and red were more sensitive to park features than human activity factors. The relative effect estimates of human activity for H3 (19.90%) were greater than those for park features (16.13%), indicating that yellow-green was more sensitive to human activity than to park features.

4. Discussion

4.1. Differences in Forest Color Authenticity among Regions

Different regions have developed different forest color landscapes under the combined influence of the natural environment and vegetation types [63]. Previous studies have indicated that green is the primary forest color in all regions, owing to the limitations of plant pigments [64]. By contrast, this study found that the primary forest color in the Mongolian Plateau and northeastern China was yellow, whereas green was only a secondary color. This can be attributed to the establishment of forest parks that have increased human influence on the forest color landscape [25], resulting in a change in the color composition of these two regions. In addition, the vegetation of the Mongolian Plateau and northeastern China is dominated by temperate grasslands and temperate mixed coniferous forests with a cold, dry climate and long autumn and winter, which make plants prone to yellow, resulting in a greater proportion of yellow than green in the forests [65,66]. Consistent with the findings of previous studies, the predominance of green was particularly prominent in southeastern China and the eastern Himalayas, probably because subtropical evergreen broad-leaved forests dominate the vegetation in these two regions [67] where plants exhibit green all year round. Yellow-green was expressed as the primary color in the Tibetan Plateau, which may be because alpine grasslands, and scrub are the dominant vegetation types in this region with a large proportion of herbaceous plants [68] comprising a large proportion of yellow-green in the forest. Orange and blue-green had large proportions in northern China, and both appeared as secondary colors, which may be because temperate deciduous broad-leaved mixed forests dominate the vegetation in this region and seasonal changes in forest color are apparent [69], resulting in diverse secondary colors. Blue, purple, and red were not significantly different among regions and were all accent colors, probably because these three colors mainly appear in flowers in the forest [36], and flower colors account for a minor proportion of the forest color landscape.
In summary, forest parkification may change the forest color composition. The differences in primary and secondary forest colors among different regions mainly depend on the vegetation type. Vegetation types result from joint selection by humans and nature in a region [70]. In forest color landscape creation, it is necessary to avoid excessive borrowing among different regions and increase the proportion of native tree species, which will protect the forest color authenticity in each region to highlight regional characteristics.

4.2. Impact of Climate Change on Forest Color Authenticity

Previous studies have reported that forest color landscapes are influenced by temperature, precipitation, and sunshine hours [15,16]. Higher temperatures lead to increased yellow-green and red indices in forest areas, possibly because of the promotion of new leaf growth (yellow-green and red) and flowering (red) [14]. Increased precipitation provides ample water for plant growth and promotes flourishing vegetation [71], leading to an increase in blue-green indices in forested areas. Previous studies have indicated that longer sunshine hours promote chlorophyll formation in plants, resulting in a greener appearance [10]. However, this study found that shorter sunshine hours result in higher yellow-green and blue-green indices in forests. This can be attributed to various factors such as latitude, weather, and elevation [72], with areas experiencing shorter sunshine hours tending toward evergreen tree species, such as those found in southeastern China. In addition, the results of this study indicate that changes in forest color are associated with frost-free periods, wind speed, and air quality. The longer the frost-free period, the longer the plant growth period, and the more likely they are to appear green [73], resulting in decreased orange, purple, and red indices. Increased wind speeds within the tolerance range of plants promote flower unfolding [74], resulting in an increase in purple and red indices in forests. However, increased wind speeds can also cause the color-changing leaves of deciduous trees to detach from the plant body [75], reducing the orange and yellow indices. As air quality worsens, blue-green, purple, and red indices in forests increase. This is likely owing to increased haze, which causes more mixed colors in the forest [76].
Climate change poses a significant threat to the authenticity and protection of forest color landscapes. Increasing human activity and emissions of greenhouse gases and air pollutants have resulted in problems such as local temperature increases and poorer air quality in forest parks [20,77], causing decreased orange and yellow indices. First, it is necessary to adopt eco-friendly travel methods and strictly adhere to tourism regulations when visiting forest parks. To address the issue of decreased orange and yellow indices caused by human-induced climate change, appropriate deciduous native tree species and flowering plants with orange and yellow coloring should be planted in forest parks, thereby increasing the likelihood and proportion of orange and yellow.

4.3. Impact of Anthropogenic Disturbances on Forest Color Authenticity

This study found that park area, elevation, perimeter area ratio, building density, and road network density all significantly affect forest color indices. Park elevation and area often have opposite effects on forest color indices, which may be related to varying levels of human activity. Larger parks and lower elevations tend to have higher levels of human activity [26,27]. Deciduous trees with colorful leaves are highly appreciated for their long display periods and contribution to park landscapes [78]. Accordingly, areas with high human activity tend to have a larger proportion of colorful deciduous trees, leading to an increase in color indices (orange, yellow, purple, and red). The perimeter area ratio of parks reflects the extent of interaction between human societies and forest environments [79] and has a significant impact on the purple and red color indices found in forests. Furthermore, increasing building density leads to an increase in the orange index but a decrease in the red index. This could be because plants in regions of high-building density are more susceptible to environmental stressors, such as soil pollution and water scarcity [80], resulting in a shortened growing season [81] and an increased proportion of orange in the forest. In forests, red is mostly present in the form of flowers with light being the primary regulatory factor of plant flowering [82]. An increase in road network density results in an increase in streetlights, which provide favorable conditions for flower blooming. Accordingly, road network density has a significant positive influence on the red index.
In summary, balancing the scope, manner, and intensity of human activities is key to preserving the authenticity of forest color [4]. The size of park visiting areas should be controlled to maintain the natural properties of forest landscapes. Higher elevations have relatively fragile plant habitats and are more vulnerable to human activities [83]. Accordingly, close-range human visits to these areas should be minimized. In highly scenic forest landscapes with rich colors, artificial constructions should be limited; moreover, it is necessary to follow the principle of “controlling and reducing community and resident numbers as much as possible” [84]. Optimizing human travel paths within parks [85] and minimizing road network density can significantly reduce the impact of human activities on the authenticity of forest colors.

4.4. Different Degrees of Sensitivity to Forest Color Authenticity

As we hypothesized, a combination of climate change and anthropogenic disturbances influences the authenticity of forest color. Orange, yellow, purple, and red were more sensitive to the external environment than yellow-green and blue-green. This may be because orange, yellow, purple, and red are associated with plant phenological changes (ecological responses by plants to changes in the external environment, such as flowering and leaf discoloration) [14]. Simultaneously, yellow-green and blue-green depend more on other factors, such as plant growth status. Furthermore, the results of this study rejected the conclusion that meteorological factors are the primary regulators of forest color [7]. We found that meteorological factors were the key factors affecting secondary forest colors (yellow and yellow-green), and human disturbances were the key factors affecting forest accent colors (orange, blue-green, blue, purple, and red). Among the anthropogenic disturbance factors, park features had the most significant influence on forest accent color. This may be because forest accent colors mainly appear in flowers and are less sensitive to meteorological factors [36] but are susceptible to human activities and anthropogenic selection. Although accent colors account for a minor proportion of the forest color landscape, they often serve as color harmonizers, making the landscape more colorful and layered [57]. Therefore, it is necessary to stringently control the influence of human interference factors on forest color. Before a park is established, we must fully grasp the scope and intensity of local human activities to determine the park’s scale, location, and shape.

4.5. Limitations

To the best of our knowledge, this was the first study to explore the impact of parkification on forest color landscape authenticity from a nationwide perspective. We clarified the forest color composition of each region after parkification and provided new ideas for forest color landscape planning, design, and sustainable management. However, this study had limitations.
Due to data availability and geographical limitations, we used the Sina Weibo platform as the data source for forest imagery rather than collecting primary data on-site. However, the forest images obtained from the Sina Weibo platform exhibit inconsistencies in quality, which manifest as follows: (1) image quality is affected by human factors such as equipment and shooting modes; and (2) forest image screening is inevitably affected by the preferences of image uploaders and researchers [33]. In future research, it is advisable to utilize consistent equipment and shooting modes (such as the same camera and accessories, same aperture, and shutter speed) to conduct regular on-site sampling of forest color landscapes from different biogeographic regions. Sampling should occur at the exact locations, in the same directions, angles, and observation distances, while documenting on-site conditions such as illumination and color temperature. Furthermore, adopting an automatic image selection algorithm minimizes the influence of individual preference on study results.
Regarding the baseline determination of forest color authenticity, forest color data before parkification was unavailable, making it impossible to quantify forest color authenticity accurately. However, compared with previous studies, we found that forest color composition changed after parkification. Therefore, we suggest continuously monitoring forest color landscapes and adopting the color data from the initial monitoring period as baseline data to explore the degree of influence of forest parkification on color authenticity.

5. Conclusions

This national-scale study quantified forest colors using Weibo images as the data source. We found that the proportion of primary and secondary colors in forest composition was affected by factors such as vegetation type, resulting in significant differences among regions. However, there was no significant variation in the proportion of accent colors (blue, purple, and red) across regions. We further explored the impact of environmental factors in parks, including meteorological and anthropogenic disturbances, on the forest color landscape. The results showed that forest color was sensitive to all meteorological factors with the proportion of orange and yellow decreasing under anthropogenic climate change. Some forest colors (orange, yellow, purple, and red) were sensitive to anthropogenic disturbances in the following ways: (1) larger park areas, lower elevations, and larger building areas led to a higher proportion of orange; (2) lower park elevations resulted in a higher proportion of yellow; (3) larger park areas, simpler shapes, and lower elevations resulted in a higher proportion of purple; and (4) larger park areas, more complex shapes, lower elevations, lower building densities, and higher road network densities led to a higher proportion of red. The sensitivity of forest secondary colors to meteorological factors was higher than that to anthropogenic disturbances. Forest accent colors were more sensitive to anthropogenic disturbances than to meteorological factors with the greatest sensitivity being to park features. In conclusion, to protect the authenticity of forest main and accent colors, excessive borrowing among different regions should be avoided during forest color landscape construction, and the proportion of native tree species should be increased. To protect the authenticity of forest accent colors, it is necessary to prioritize park area, shape, and location control; reduce the construction of land in park buffer zones; and optimize park tour routes.
This study analyzed the impact and degree of influence of meteorological factors and anthropogenic disturbances on different forest colors, which provided new ideas for authentic protection and sustainable management of forest color landscapes. However, some limitations need to be addressed. In future studies, we will enhance image quality by improving image acquisition and screening. We can continuously monitor forest color landscapes to grasp the dynamic changes and anthropogenic disturbance degree of forest color, which will enhance protection of forest color landscapes’ authenticity. Moreover, we will investigate the effects of climate change and human interference on the spatial patterns of forest color landscapes to provide in-depth guidance for planning and constructing sustainable forest color landscapes in parks and other urban green spaces.

Author Contributions

Conceptualization, C.Z., C.W. and W.H.; methodology, C.Z. and W.H.; software, C.Z. and W.H.; validation, C.Z., C.W. and W.H.; formal analysis, C.Z. and W.H.; resources, C.W. and C.Z.; data curation, W.H. and H.Z.; writing—original draft, W.H.; writing—review and editing, C.Z., C.W., S.L. and W.H.; visualization, W.H. and D.H.; supervision, C.Z. and C.W.; project administration, C.Z. and C.W.; funding acquisition, C.W., C.Z. and D.S. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by National Key R&D Program of China (grant number: 2021YFE0193200); National Non-Profit Research Institutions of the Chinese Academy of Forestry (grant number: CAFYBB2020ZB008); Social development science and technology projects of Dongguan (grant number: 20231800936142).

Data Availability Statement

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Acknowledgments

We thank Ting Liu for writing advice and Yi Li for image processing advice.

Conflicts of Interest

The authors declare no conflict of interest.

Appendix A

Forests 14 01799 g0a1
Figure A1. Acquisition of forest color image and quantification of forest color.
Figure A1. Acquisition of forest color image and quantification of forest color.
Forests 14 01799 g0a1
Table A1. Summary data of the images collected from the different forest parks.
Table A1. Summary data of the images collected from the different forest parks.
RegionProvinceNational Forest ParkNumber of ImagesTotal
SpringSummerAutumnWinterSubtotal
Northeast ChinaJilinLafashan National Forest Park012102252
HeilongjiangWuying National Forest Park278118
HeilongjiangXianglushan National Forest Park155112
North ChinaBeijingLabagou Origin Forest Park2424030224
TianjinJiulongshan National Forest Park386017
HebeiYesanpo National Forest Park21710130
ShanxiTaihang Canyon National Forest Park32015240
LiaoningDaheishan National Forest Park8108026
LiaoningDalian Tianmen Mountain National Forest Park327214
ShandongTaishan National Forest Park4125526
HenanBaiyunshan National Forest Park371011
ShaanxiTaibaishan National Forest Park01510530
Southeast ChinaGansuMaiji National Forest Park2252130551
ShanghaiSheshan National Forest Park31110024
JiangsuZijinshan National Forest Park2132540
ZhejiangYandang Mountain National Forest Park2121116
AnhuiHuangshan National Forest Park5128328
AnhuiTachuan National Forest Park1744456
FujianFuzhou National Forest Park8211030
JiangxiYangling National Forest Park81011130
HubeiShennongjia National Forest Park3203430
HunanZhangjiajie Tianmen Mountain National Forest Park1204227
GuangdongGuanyinshan National Forest Park5165026
GuangdongWutongshan National Forest Park213010768
GuangxiDarongshan National Forest Park550414
GuangxiDebao Red Leaves National Forest Park1116119
HainanJianfengling National Forest Park3116020
ChongqingGeleshan National Forest Park4104018
SichuanTiantaishan National Forest Park0145120
GuizhouFenghuangshan National Forest Park366015
GuizhouLeigongshan National Forest Park235010
ShaanxiJinsixia National Forest Park2244030
Mongolian PlateauShanxiHengshan National Forest Park6167130268
HebeiSaihanta National Forest Park61041360
Inner MongoliaArshaan National Forest Park121459792
GansuTulugou National Forest Park81311335
QinghaiBeishan National Forest Park71128551
Eastern HimalayasYunnanXishuangbanna National Forest Park1134141372118
TibetSegyi La National Forest Park101813546
Northwest ChinaXinjiangTianshan Grand Canyon National Forest Park2115141868122
XinjiangNalati National Forest Park141420654
Tibetan PlateauTibetNimu National Forest Park981433487
QinghaiKanbula National Forest Park91719853
Total2255375371231422
Table A2. Seasonal division of forest images.
Table A2. Seasonal division of forest images.
Start Date of Image UploadEnd Date of Image UploadSeason
1 December 201828 February 2019Winter
1 March 201931 May 2019Spring
1 June 201931 August 2019Summer
1 September 201930 November 2019Autumn
Table A3. Data sources for environmental factors.
Table A3. Data sources for environmental factors.
TypeVariablesData Source Website or Software
MeteorologicalCumulative sunshine hours (SH, hour)https://data.cma.cn/ (accessed on 25 February 2023)
Average temperature (T, °C)https://rp5.ru (accessed on 25 February 2023)
Average wind speed (F, m/s)https://rp5.ru (accessed on 25 February 2023)
Cumulative precipitation (R, mm)https://rp5.ru (accessed on 25 February 2023)
Average air quality index (AQI, μg/m3)http://www.aqistudy.cn/historydata/ (accessed on 25 February 2023)
Frost-free period (FFP, day)the official website of each park (accessed on 10 March 2023)
Park featureArea of the park (AREA, ha)the official website of each park (accessed on 10 March 2023)
The perimeter area ratio of the park (PAR, km/km2)https://map.baidu.com (accessed on 25 March 2023)
Elevation of the park (ELEV, m)the official website of each park (accessed on 10 March 2023)
Human activityDistance to the nearest city center (D, km)ArcGIS 10.6 software
Building density (BD, %)RiverMap 4.1 software
Road network density (RD, km/km2)RiverMap 4.1 software
Table A4. Prefecture-level city and its adjacent meteorological station of each park.
Table A4. Prefecture-level city and its adjacent meteorological station of each park.
No.National Forest ParkCityMeteorological Station
1Lafashan National Forest ParkJilin54,172
2Wuying National Forest ParkYichun50,774
3Xianglushan National Forest ParkHarbin50,953
4Labagou Origin Forest ParkBeijing54,511
5Jiulongshan National Forest ParkTianjin54,527
6Yesanpo National Forest ParkBaoding54,602
7Taihang Canyon National Forest ParkChangzhi53,882
8Daheishan National Forest ParkChaoyang54,324
9Dalian Tianmen Mountain National Forest ParkDalian54,662
10Taishan National Forest ParkTaian54,827
11Baiyunshan National Forest ParkLuoyang57,073
12Taibaishan National Forest ParkBaoji57,016
13Maiji National Forest ParkTianshui57,006
14Sheshan National Forest ParkShanghai58,362
15Zijinshan National Forest ParkNanjing58,238
16Yandang Mountain National Forest ParkWenzhou58,659
17Huangshan National Forest ParkHuangshan58,437
18Tachuan National Forest ParkHuangshan58,437
19Fuzhou National Forest ParkFuzhou58,847
20Yangling National Forest ParkGanzhou57,993
21Shennongjia National Forest ParkShennongjia57,259
22Zhangjiajie Tianmen Mountain National Forest ParkZhangjiajie57,554
23Guanyinshan National Forest ParkDongguan59,289
24Wutongshan National Forest ParkShenzhen59,493
25Darongshan National Forest ParkYulin59,453
26Debao Red Leaves National Forest ParkBaise59,211
27Jianfengling National Forest ParkLedong59,940
28Geleshan National Forest ParkChongqing57,516
29Tiantaishan National Forest ParkChengdu56,294
30Fenghuangshan National Forest ParkZunyi57,713
31Leigongshan National Forest ParkQiandongnan57,832
32Jinsixia National Forest ParkShangluo57,067
33Hengshan National Forest ParkDatong53,487
34Saihanta National Forest ParkChengde54,423
35Arshaan National Forest ParkHinggan50,727
36Tulugou National Forest ParkLanzhou52,889
37Beishan National Forest ParkHaidong52,875
38Xishuangbanna National Forest ParkXishuangbanna56,959
39Segyi La National Forest ParkLinzhi56,312
40Tianshan Grand Canyon National Forest ParkUrumqi51,463
41Nalati National Forest ParkYili51,431
42Nimu National Forest ParkLasa55,591
43Kanbula National Forest ParkHuangnan52,984
Table A5. Variance inflation factors (VIF) for the generalized linear mixed models including all variables.
Table A5. Variance inflation factors (VIF) for the generalized linear mixed models including all variables.
H1H2H3H5H6H7H8
BD4.692.652.632.713.021.531.00
RD5.973.743.653.694.25-1.07
D1.301.331.341.331.331.331.00
AREA1.661.461.471.511.571.901.00
PAR1.301.271.271.251.781.281.07
ELEV2.131.861.911.901.902.131.00
SH5.082.662.043.043.367.131.00
T6.593.742.655.264.645.771.00
R2.461.852.341.881.673.291.00
F1.771.421.171.201.131.581.00
AQI1.721.491.291.771.631.571.00
FFP1.991.551.571.591.692.851.00
Table A6. Summary of the generalized linear mixed models for forest color indices.
Table A6. Summary of the generalized linear mixed models for forest color indices.
H1H2H3H5
PredictorsORCIpORCIpORCIpORCIp
(Intercept)0.0050.003–0.008<0.0010.2110.125–0.358<0.0010.1330.084–0.210<0.0010.0440.027–0.071<0.001
BD2.7981.123–6.9700.0271.4460.619–3.3740.3940.610.295–1.2590.1811.6060.751–3.4370.222
RD0.3150.090–1.1040.0711.2250.449–3.3420.6921.5530.655–3.6860.3180.8760.352–2.1800.776
D0.5110.261–1.0020.0510.8070.397–1.6410.5541.1310.608–2.1040.6971.3790.723–2.6330.33
AREA2.041.110–3.7490.0221.4970.731–3.0620.270.7760.416–1.4490.4260.8560.445–1.6480.642
PAR1.2980.776–2.1740.320.9670.550–1.7030.9091.1110.679–1.8180.6760.7010.399–1.2320.217
ELEV0.2560.133–0.493<0.0010.4820.235–0.9890.0471.6430.873–3.0910.1241.7450.894–3.4030.103
SH0.9140.651–1.2840.6041.2771.130–1.441<0.0010.5620.513–0.615<0.0010.7340.642–0.840<0.001
T0.7260.563–0.9370.0140.860.774–0.9560.0052.6122.337–2.921<0.0011.060.895–1.2550.502
R0.0750.045–0.124<0.0010.2590.231–0.290<0.0010.3850.350–0.423<0.0012.1791.884–2.519<0.001
F0.5010.386–0.652<0.0010.1990.173–0.228<0.0011.8921.684–2.126<0.0010.9990.837–1.1920.991
AQI0.9720.833–1.1330.7160.5320.495–0.572<0.0010.7620.689–0.841<0.0011.4381.274–1.623<0.001
FFP0.370.177–0.7770.0090.5070.256–1.0060.0521.0030.553–1.8170.9930.8890.474–1.6700.715
Random Effects
σ23.293.293.293.29
τ00 park1.642.82.132.27
ICC0.330.460.390.41
Npark43434343
Observations1422142214221422
Marginal R2/Conditional R20.688/0.7920.450/0.7030.154/0.4860.144/0.493
H6H7H8
PredictorsORCIpORCIpORCIp
(Intercept)0.0110.006–0.021<0.00100.000–0.000<0.00100.000–0.000<0.001
BD0.7960.279–2.2710.6691.0080.385–2.6390.9870.2080.208–0.209<0.001
RD1.3940.393–4.9500.608 1.1061.102–1.110<0.001
D0.960.419–2.2000.9241.4490.448–4.6930.5360.5840.150–2.2710.438
AREA0.9370.406–2.1630.8783.4541.146–10.4130.02810.10910.078–10.140<0.001
PAR0.3050.072–1.3030.1090.0680.006–0.7320.0271.8951.889–1.901<0.001
ELEV1.1680.483–2.8230.730.1660.050–0.5520.0030.0890.089–0.089<0.001
SH1.2531.016–1.5460.0350.0580.024–0.136<0.0010.2190.218–0.219<0.001
T0.2490.197–0.314<0.0011.1520.605–2.1950.6661.4721.467–1.476<0.001
R1.2020.989–1.4610.0650.0760.028–0.205<0.0010.1110.110–0.111<0.001
F0.6790.541–0.8520.0013.9322.146–7.205<0.0013.4793.468–3.490<0.001
AQI1.0330.864–1.2350.7212.4661.415–4.2970.0014.2394.226–4.252<0.001
FFP2.9191.257–6.7790.0130.1060.018–0.6290.0130.1620.162–0.163<0.001
Random Effects
σ23.293.293.29
τ00 park3.394.455.61
ICC0.510.570.63
Npark434343
Observations142214221422
Marginal R2/Conditional R20.269/0.6400.767/0.9010.622/0.860
OR is odds ratio and CI is 95% confidence interval of OR.

References

  1. Zhao, Z.; Yang, R. The concept of national park authenticity and integrity in China and its evaluation framework. Biodivers. Sci. 2021, 29, 1271–1278. [Google Scholar] [CrossRef]
  2. De Jesus França, L.C.; Júnior, F.W.A.; e Silva, C.S.J.; Monti, C.A.U.; Ferreira, T.C.; de Oliveira Santana, C.J.; Gomide, L.R. Forest landscape planning and management: A state-of-the-art review. Trees For. People 2022, 8, 100275. [Google Scholar] [CrossRef]
  3. Chen, Z.; Fu, W.; Konijnendijk van den Bosch, C.C.; Pan, H.; Huang, S.; Zhu, Z.; Qiao, Y.; Wang, N.; Dong, J. National Forest Parks in China: Origin, Evolution, and Sustainable Development. Forests 2019, 10, 323. [Google Scholar] [CrossRef]
  4. Peng, K. Basic characteristics, management problems and adjusted strategies of human activities for China’s national parks. Biodivers. Sci. 2019, 29, 278–282. [Google Scholar] [CrossRef]
  5. Fuller, J.L.; Foster, D.R.; McLachlan, J.S.; Drake, N. Impact of Human Activity on Regional Forest Composition and Dynamics in Central New England. Ecosystems 1998, 1, 76–95. [Google Scholar] [CrossRef]
  6. Keenan, T.F.; Gray, J.; Friedl, M.A.; Toomey, M.; Bohrer, G.; Hollinger, D.Y.; Munger, J.; O’Keefe, J.; Schmid, H.P.; Wing, I.S.; et al. Net carbon uptake has increased through warming-induced changes in temperate forest phenology. Nat. Clim. Change 2014, 4, 598–604. [Google Scholar] [CrossRef]
  7. Estrella, N.; Menzel, A. Responses of leaf colouring in four deciduous tree species to climate and weather in Germany. Clim. Res. 2006, 32, 253–267. [Google Scholar] [CrossRef]
  8. Zhang, S. The Research on the Cities’ Colors Based on Regional Characteristics. Ph.D. Thesis, Wuhan University, Wuhan, China, 2014. [Google Scholar]
  9. Zhang, X.; Wang, Y. Progress of research on the mechanism of color presentation and factors affecting anthocyanin in colored foliage plants. J. Sichuan For. Sci. Technol. 2015, 36, 28–30. [Google Scholar]
  10. Cornah, J.E.; Terry, M.J.; Smith, A.G. Green or red: What stops the traffic in the tetrapyrrole pathway? Trends Plant Sci. 2003, 8, 224–230. [Google Scholar] [CrossRef]
  11. León-Chan, R.G.; López-Meyer, M.; Osuna-Enciso, T.; Sañudo-Barajas, J.A.; Heredia, J.B.; León-Félix, J. Low temperature and ultraviolet-B radiation affect chlorophyll content and induce the accumulation of UV-B-absorbing and antioxidant compounds in bell pepper (Capsicum annuum) plants. Environ. Exp. Bot. 2017, 139, 143–151. [Google Scholar] [CrossRef]
  12. He, M.; Wang, H.; Hu, Y.; Zhao, P.; He, S.; Mei, Y. Research Status, Development and Utilization of Color-Leaf Trees. Acta Agric. Univ. Jiangxiensis 2018, 40, 1134–1144. [Google Scholar]
  13. Archetti, M.; Richardson, A.D.; O’Keefe, J.; Delpierre, N. Predicting climate change impacts on the amount and duration of autumn colors in a New England Forest. PLoS ONE 2013, 8, e57373. [Google Scholar] [CrossRef] [PubMed]
  14. Klosterman, S.; Richardson, A.D. Observing spring and fall phenology in a deciduous forest with aerial drone imagery. Sensors 2017, 17, 2852. [Google Scholar] [CrossRef] [PubMed]
  15. Keenan, T.F.; Richardson, A.D. The timing of autumn senescence is affected by the timing of spring phenology: Implications for predictive models. Glob. Change Biol. 2015, 21, 2634–2641. [Google Scholar] [CrossRef] [PubMed]
  16. Koide, D.; Ide, R.; Oguma, H. Detection of autumn leaf phenology and color brightness from repeat photography: Accurate, robust, and sensitive indexes and modeling under unstable field observations. Ecol. Indic. 2019, 106, 105482. [Google Scholar] [CrossRef]
  17. Ma, R.; Cai, L. Visual analysis of forest sports and health tourism based on artificial intelligence. J. Electron. Imaging 2022, 31, 062008. [Google Scholar] [CrossRef]
  18. Kuvan, Y. Assessing the impacts of tourism on forests: Mass tourism and policy in Turkey. Environ. Eng. Manag. J. 2012, 11, 1415–1424. [Google Scholar] [CrossRef]
  19. Noble, I.R.; Dirzo, R. Forests as human-dominated ecosystems. Science 1997, 277, 522–525. [Google Scholar] [CrossRef]
  20. Rosenzweig, C.; Karoly, D.; Vicarelli, M.; Neofotis, P.; Wu, Q.; Casassa, G.; Menzel, A.; Root, T.L.; Estrella, N.; Seguin, B.; et al. Attributing physical and biological impacts to anthropogenic climate change. Nature 2008, 453, 353–357. [Google Scholar] [CrossRef] [PubMed]
  21. Liu, Y.; Hua, Y.; Zhou, H.; Ye, L.; Wang, G.; Jin, J.; Bao, Z. Precipitation variation and trend projection in the eastern monsoon region of China since 1470. Adv. Water Sci. 2022, 33, 1–14. [Google Scholar]
  22. Li, D.; Stucky, B.J.; Deck, J.; Baiser, B.; Guralnick, R.P. The effect of urbanization on plant phenology depends on regional temperature. Nat. Ecol. Evol. 2019, 3, 1661–1667. [Google Scholar] [CrossRef]
  23. Qiu, T.; Song, C.; Zhang, Y.; Liu, H.; Vose, J.M. Urbanization and climate change jointly shift land surface phenology in the northern mid-latitude large cities. Remote Sens. Environ. 2020, 236, 111477. [Google Scholar] [CrossRef]
  24. Lindh, B.C.; McGahan, K.A.; Bluhm, W.L. Changes in urban plant phenology in the Pacific Northwest from 1959 to 2016: Anthropogenic warming and natural oscillation. Int. J. Biometeorol. 2018, 62, 1675–1684. [Google Scholar] [CrossRef] [PubMed]
  25. Liu, Y.; Tao, Y. The Analysis of the Development Obstacles for Chinese Forest Tourism. Probl. For. Econ. 2003, 23, 49–52. [Google Scholar]
  26. Niu, X.; Kang, N. Spatio-temporal Characteristics and Influencing Factors of Tourist Activities in Shanghai Country Parks—A Study Based on Mobile Phone Signaling Data. Chin. Landsc. Archit. 2021, 37, 39–43. [Google Scholar]
  27. Zhang, Y.; Liu, L.; Bai, W.; Shen, Z.; Yan, J.; Ding, M.; Li, S.; Zheng, D. Grassland Degradation in the Source Region of the Yellow River. Acta Geogr. Sin. 2006, 61, 3–14. [Google Scholar]
  28. Xiao, X.; Ding, M.; Dong, L.; Wang, Y.; Wang, L. Study on the Influence of Park Characteristics and Distance from the Water Area on Summer Temperature and Humidity Effect: A Case Study of Suzhou Industrial Park. Chin. Landsc. Archit. 2021, 37, 48–53. [Google Scholar]
  29. Lu, S.; Qin, F.; Chen, N.; Yu, Z.; Xiao, Y.; Cheng, X.; Guan, X. Spatiotemporal differences in forest ecological security warning values in Beijing: Using an integrated evaluation index system and system dynamics model. Ecol. Indic. 2019, 104, 549–558. [Google Scholar] [CrossRef]
  30. Zhang, X.; Wang, X.; Zhou, Z.; Li, M.; Jing, C. Spatial Quantitative Model of Human Activity Disturbance Intensity and Land Use Intensity Based on GF-6 Image, Empirical Study in Southwest Mountainous County, China. Remote Sens. 2022, 14, 4574. [Google Scholar] [CrossRef]
  31. Cai, B.; Shao, Z.; Fang, S.; Huang, X. Quantifying Dynamic Coupling Coordination Degree of Human–Environmental Interactions during Urban–Rural Land Transitions of China. Land 2022, 11, 935. [Google Scholar] [CrossRef]
  32. Shao, Q.; Mu, D.; Wang, P.; Jin, C. Identifying Subjects of Online Opinion from Public Health Emergencies. Data Anal. Knowl. Discov. 2020, 4, 68–80. [Google Scholar]
  33. Han, W.; Zhang, C.; Wang, C.; Yin, L. Constructing a Forest Color Palette and the Effects of the Color Patch Index on Human Eye Recognition Accuracy. Forests 2023, 14, 627. [Google Scholar] [CrossRef]
  34. Xu, Y.; Zang, R. Theoretical and practical research on conservation of Wild Plants with Extremely Small Populations in China. Biodivers. Sci. 2022, 30, 22505. [Google Scholar] [CrossRef]
  35. Zhao, S.; Fang, J.; Lei, G. Global 200: An approach to setting large-scale biodiversity conservation priorities. Biodivers. Sci. 2000, 8, 435–440. [Google Scholar] [CrossRef]
  36. Wang, Q.; Guo, Q.; Chi, X.; Zhu, S.; Tang, Z. Evolutionary history and climate conditions constrain the flower colours of woody plants in China. J. Plant Ecol. 2022, 15, 196–207. [Google Scholar] [CrossRef]
  37. Wang, Z.; Fang, J.; Tang, Z.; Lin, X. Relative role of contemporary environment versus history in shaping diversity patterns of China’s woody plants. Ecography 2012, 35, 1124–1133. [Google Scholar] [CrossRef]
  38. Kumar, M.; Bindal, A.; Gautam, R.; Bhatia, R. Keyword query based focused Web crawler. Procedia Comput. Sci. 2018, 125, 584–590. [Google Scholar] [CrossRef]
  39. Lin, W.; Mu, Y.; Zhang, Z.; Wang, J.; Diao, X.; Lu, Z.; Guo, W.; Wang, Y.; Xu, B. Research on cognitive evaluation of forest color based on visual behavior experiments and landscape preference. PLoS ONE 2022, 17, e0276677. [Google Scholar] [CrossRef] [PubMed]
  40. Huang, S.C.; Cheng, F.C.; Chiu, Y.S. Efficient contrast enhancement using adaptive gamma correction with weighting distribution. IEEE Trans. Image Process. 2012, 22, 1032–1041. [Google Scholar] [CrossRef]
  41. Rizzi, A.; Gatta, C.; Marini, D. A new algorithm for unsupervised global and local color correction. Pattern Recognit. Lett. 2003, 24, 1663–1677. [Google Scholar] [CrossRef]
  42. Zhang, Z.; Qie, G.; Wang, C.; Jiang, S.; Li, X.; Li, M. Relationship between Forest Color Characteristics and Scenic Beauty: Case Study Analyzing Pictures of Mountainous Forests at Sloped Positions in Jiuzhai Valley, China. Forests 2017, 8, 63. [Google Scholar] [CrossRef]
  43. Qin, Y.; Fang, L.; Zhang, L.; Shi, J.; Wang, B. Aesthetic Effects of Individual Variation of Three Forest Color Elements. J. Chin. Urban For. 2016, 14, 26–32. [Google Scholar]
  44. Cao, Y.; Xu, C.; Cui, Y.; Yue, Y.; Ren, Y. Effects of viewing distance and light conditions on the color of Cotinus coggygria var. cinerea forest landscape. J. Cent. South Univ. For. Technol. 2019, 39, 22–29+48. [Google Scholar]
  45. Sammouda, R.; El-Zaart, A. An optimized approach for prostate image segmentation using K-means clustering algorithm with elbow method. Comput. Intell. Neurosci. 2021, 2021, 4553832. [Google Scholar] [CrossRef]
  46. Xu, P.; Mao, H.; Zhang, Y.; Gu, B.; Zhang, Y. Study on analysis method for color composition of ethnic costumes. J. Silk 2019, 56, 24–29. [Google Scholar]
  47. Cheng, D.; Zhang, H.; Jiang, M.; Kou, Q. Color Image Retrieval Method Fusing Principal Curvature and Color Information. J. Comput.-Aided Des. Comput. Graph. 2021, 33, 223–231. [Google Scholar] [CrossRef]
  48. Wang, D.; Chen, J.; Zhang, L.; Sun, Z.; Wang, X.; Zhang, X.; Zhang, W. Establishing an ecological security pattern for urban agglomeration, taking ecosystem services and human interference factors into consideration. PeerJ 2019, 7, e7306. [Google Scholar] [CrossRef]
  49. Wang, X.; Zhao, X. Impacts of human activities on ecosystem services in national parks: A case study of Qilian Mountain National Park. J. Nat. Resour. 2023, 38, 966–982. [Google Scholar] [CrossRef]
  50. Yin, L.; Zhang, W.; Ren, B. Research on the Distribution Characteristics of Plant Diversity in Beijing Urban Parks. J. Inn. Mong. Agric. Univ. (Nat. Sci. Ed.) 2020, 41, 21–26. [Google Scholar]
  51. Xu, Y.; Xu, X.; Tang, Q. Human activity intensity of land surface: Concept, methods and application in China. J. Geogr. Sci. 2016, 26, 1349–1361. [Google Scholar] [CrossRef]
  52. Mo, W.; Wang, Y.; Zhang, Y.; Zhuang, D. Impacts of road network expansion on landscape ecological risk in a megacity, China: A case study of Beijing. Sci. Total Environ. 2017, 574, 1000–1011. [Google Scholar] [CrossRef] [PubMed]
  53. Gounaridis, D.; Newell, J.P.; Goodspeed, R. The impact of urban sprawl on forest landscapes in Southeast Michigan, 1985–2015. Landsc. Ecol. 2020, 35, 1975–1993. [Google Scholar]
  54. Canedoli, C.; Manenti, R.; Padoa-Schioppa, E. Birds biodiversity in urban and periurban forests: Environmental determinants at local and landscape scales. Urban Ecosyst. 2018, 21, 779–793. [Google Scholar]
  55. Han, W.; Zhang, C.; Zhang, X.; Wang, C. Tourists Spatial Distribution Pattern and Accessibility in Hunan Forest Botanical Garden. J. Chin. Urban For. 2021, 19, 34–39. [Google Scholar]
  56. Wang, J.; Zhou, G. The climatic suitability and climatic impact factors affecting the wine grapes (Vitis vinifera L.) planting distribution in China. Acta Ecol. Sin. 2021, 41, 2418–2427. [Google Scholar]
  57. Zhou, L.; Feng, Y.; Wang, J. The Study on Color Quantitative Characteristics of Plant Communities in Xihua Park, Kunming. J. Southwest For. Univ. (Soc. Sci.) 2023, 7, 42–54. [Google Scholar]
  58. Volařík, D.; Svátek, M.; Šenfeldr, M.; Kučera, A.; Šrámek, M.; Dreslerová, J.; Matula, R. Variation in canopy openness among main structural types of woody vegetation in a traditionally managed landscape. Folia Geobot. 2017, 52, 15–32. [Google Scholar] [CrossRef]
  59. Turner, P. Critical values for the Durbin-Watson test in large samples. Appl. Econ. Lett. 2020, 27, 1495–1499. [Google Scholar] [CrossRef]
  60. Lembrechts, J.J.; Pauchard, A.; Lenoir, J.; Nuñez, M.A.; Geron, C.; Ven, A.; Bravo-Monasterio, P.; Teneb, E.; Nijs, I.; Milbau, A. Disturbance is the key to plant invasions in cold environments. Proc. Natl. Acad. Sci. USA 2016, 113, 14061–14066. [Google Scholar] [CrossRef]
  61. Eskelinen, A.; Harrison, S.P. Resource colimitation governs plant community responses to altered precipitation. Proc. Natl. Acad. Sci. USA 2015, 112, 13009–13014. [Google Scholar] [CrossRef]
  62. Han, D.; Wang, C.; Sun, Z.; She, J.; Yin, L.; Bian, Q.; Han, W. Microhabitat preferences of butterflies in urban parks: Both vegetation structure and resources are decisive. Urban For. Urban Green. 2022, 71, 127552. [Google Scholar] [CrossRef]
  63. Song, X.; Su, Y. Study on the Color Aesthetics of Chinese Royal Gardens. J. Appl. Sci. Eng. Innov. 2019, 6, 61–67. [Google Scholar]
  64. Du, H.; Zhuang, Y. Optical remote sensing images feature extraction of forest regions. In Proceedings of the 2019 IEEE International Conference on Signal, Information and Data Processing (ICSIDP), Chongqing, China, 11–13 December 2019; pp. 1–5. [Google Scholar]
  65. Wang, X. Leaf Color Interation Properties of Harbin Woody Plants and Their Communication System Establishment. Master’s Thesis, Northeast Forestry University, Harbin, China, 2008. [Google Scholar]
  66. Yi, H. The Regional Characteristics Study on the Environmental Color of Traditional Village in Baotou, Inner Mongolia. Master’s Thesis, lnner Mongolia University of Science and Technology, Baotou, China, 2021. [Google Scholar]
  67. Wei, X.; Liu, Y. Geographical overview of forest landscape in Zhejiang Province. J. Zhejiang For. Coll. 1996, 13, 316–321. [Google Scholar]
  68. Zhu, Y.; Zhang, H.; Ding, M.; Yu, Z. Changes of vegetation greenness and its response to drought-wet variation on the Qingzang Plateau. Chin. J. Plant Ecol. 2023, 47, 51–64. [Google Scholar] [CrossRef]
  69. Xing, X.Y.; Hao, P.Y.; Dong, L. Color characteristics of Beijing’s regional woody vegetation based on Natural Color System. Color Res. Appl. 2019, 44, 595–612. [Google Scholar] [CrossRef]
  70. Zhu, H.; Wang, J.; Xu, Y.; Qi, J.; Luo, H.; Peng, Z. Analysis on Forest Vegetational Cover Changes in Yuxi and its Driving Force. For. Resour. Manag. 2021, 1, 121–131. [Google Scholar]
  71. Lu, J.; Wang, X.; Cao, Y. Effects of climate and human activities on vegetation variation in Liaoning Province. Adv. Sci. Technol. Water Resour. 2022, 42, 7–14+38. [Google Scholar]
  72. Xiao, F.; Zhang, X.; Liao, Y.; Liu, Q. Sunshine Duration in China: Variation Characteristics and Its Influence. Chin. Agric. Sci. Bull. 2020, 36, 92–100. [Google Scholar]
  73. Chou, J.; Xu, Y.; Xu, H. Definition of growing period for major grain crops in China: A review. J. Beijing Norm. Univ. (Nat. Sci.) 2022, 58, 945–949. [Google Scholar]
  74. Pan, C.; Liu, X.; Ye, L.; Xue, Q. A preliminary study on the Oenothera littaralis biological characteristics of blossom. J. Fujian Coll. For. 2014, 34, 21–25. [Google Scholar]
  75. Wang, X.; Li, Y.; Luo, Y.; Lian, J.; Niu, Y.; Gong, X.; Yang, H. Inter-and intra-annual dynamics of vegetation litter at different habitats in Horqin Sandy Land, China. Chin. J. Appl. Ecol. 2018, 29, 1494–1502. [Google Scholar]
  76. Huang, X. Analysis of Photoshop in digital image post. Comput. Knowl. Technol. 2011, 7, 6240–6241+6252. [Google Scholar]
  77. Williams, J.W.; Jackson, S.T.; Kutzbach, J.E. Projected distributions of novel and disappearing climates by 2100 AD. Proc. Natl. Acad. Sci. USA 2007, 104, 5738–5742. [Google Scholar] [CrossRef]
  78. Wang, W. Investigation and Analysis of Color-leafed Plants in Shrub and Herb in Park of Shenzhen. J. Anhui Agric. Sci. 2019, 47, 129–132. [Google Scholar]
  79. Wu, D.; Xiao, L.; Zhang, L.; Liu, J.; Chen, C.; Deng, C. Relationship between plant richness, similarity, and spatial characteristics of islands around Pingtan. Subtrop. Agric. Res. 2020, 16, 35–41. [Google Scholar]
  80. Guo, Z.; Ma, Y.; Li, H.; Liu, W. Impact of Human Activities on Forest Landscape in the Liushahe Watershed in Xishuangbanna, Southwest China. Res. Soil Water Conserv. 2007, 14, 38–42. [Google Scholar]
  81. Ziska, L.H.; Gebhard, D.E.; Frenz, D.A.; Faulkner, S.; Singer, B.D.; Straka, J.G. Cities as harbingers of climate change: Common ragweed, urbanization, and public health. J. Allergy Clin. Immunol. 2003, 111, 290–295. [Google Scholar] [CrossRef] [PubMed]
  82. Liu, J.; Li, Y.; Dong, Y.; Han, D. Role of PRRs in Photoperiod Regulation of Flowering in Different Plants. Mol. Plant Breed. 2023, 21, 1175–1181. [Google Scholar]
  83. Shen, J.; Chen, D.; Hong, Y.; Sun, X.; Zhang, S. Prediction on potential distribution areas of Larix kaempferi in China based on MaxEnt model. J. Plant Resour. Environ. 2019, 28, 19–25. [Google Scholar]
  84. Funch, R.R.; Harley, R.M. Reconfiguring the boundaries of the Chapada Diamantina National Park (Brazil) using ecological criteria in the context of a human-dominated landscape. Landsc. Urban Plan. 2007, 83, 355–362. [Google Scholar] [CrossRef]
  85. Xiong, P.; Xu, Y.; Zhu, Q. Application of an algorithm for scenic route recommendations during selective sightseeing. J. Beijing Univ. Chem. Technol. (Nat. Sci.) 2017, 44, 99–106. [Google Scholar]
Forests 14 01799 g001
Figure 1. Distribution of 43 national forest parks.
Figure 1. Distribution of 43 national forest parks.
Forests 14 01799 g001
Forests 14 01799 g002
Figure 2. Classification of forest colors.
Figure 2. Classification of forest colors.
Forests 14 01799 g002
Forests 14 01799 g003
Figure 3. Forest color composition in China.
Figure 3. Forest color composition in China.
Forests 14 01799 g003
Forests 14 01799 g004
Figure 4. Relative effects for (A) orange index (H1), (B) yellow index (H2), (C) yellow-green index (H3), (D) blue-green index (H5), (E) blue index (H6), (F) purple index (H7) and (G) red index (H8). The parameter estimates of the model predictors with their associated 95% confidence intervals are shown on the horizontal axis. The relative effect estimates for each variable group are shown on the vertical axis. * p < 0.05; ** p < 0.01.
Figure 4. Relative effects for (A) orange index (H1), (B) yellow index (H2), (C) yellow-green index (H3), (D) blue-green index (H5), (E) blue index (H6), (F) purple index (H7) and (G) red index (H8). The parameter estimates of the model predictors with their associated 95% confidence intervals are shown on the horizontal axis. The relative effect estimates for each variable group are shown on the vertical axis. * p < 0.05; ** p < 0.01.
Forests 14 01799 g004
Table 1. Environmental variables of the forest image.
Table 1. Environmental variables of the forest image.
TypeVariablesDescription
MeteorologicalCumulative sunshine hours (SH, hour)Cumulative sunshine hours for the corresponding season of the forest image
Average temperature (T, °C)Average temperature for the corresponding season of the forest image
Average wind speed (F, m/s)Average wind speed for the corresponding season of the forest image
Cumulative precipitation (R, mm)Cumulative precipitation for the corresponding season of the forest image
Average air quality index (AQI, μg/m3)Average air quality index for the corresponding season of the forest image
Frost-free period (FFP, day)Number of days between final frost day and first frost day in a year [56]
Park featureArea of the park (AREA, ha)The maximum horizontal extent of human activity in the park
The perimeter area ratio of the park (PAR, km/km2)The perimeter of the park divided by its area
Elevation of the park (ELEV, m)The maximum vertical extent of human activity in the park
Human activityDistance to the nearest city center (D, km)The minimum linear distance between the park and the nearest prefecture-level city center
Building density (BD, %)Building area within the 3 km buffer of the park divided by its buffer area
Road network density (RD, km/km2)The total length of roads within the 3 km buffer of the park divided by its buffer area
Table 2. Kruskal–Wallis test for forest color index among different regions.
Table 2. Kruskal–Wallis test for forest color index among different regions.
RegionH1H2H3H4H5H6H7H8
Northeast China20.8638.108.6620.905.612.452.191.22
North China11.0427.5211.8532.0610.204.481.561.30
Mongolian Plateau7.1245.9612.9625.024.652.301.220.78
Northwest China2.9723.7828.0131.316.494.632.110.69
Tibetan Plateau4.9226.9234.8931.970.180.740.390.00
Southeast China2.4919.7120.4444.558.642.970.790.42
Eastern Himalayas2.8614.8023.5846.546.562.461.801.39
H61.276 **96.235 **68.498 **64.536 **32.307 **12.41410.85312.304
The non-bold numbers in the table are color indices. H is the statistic for the Kruskal-Wallis test. The eight color indices include orange (H1), yellow (H2), yellow-green (H3), green (H4), blue-green (H5), blue (H6), purple (H7) and red (H8). ** p < 0.01.
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

Han, W.; Zhang, C.; Wang, C.; Liu, S.; Shen, D.; Zhou, H.; Han, D. Impact of Forest Parkification on Color Authenticity. Forests 2023, 14, 1799. https://doi.org/10.3390/f14091799

AMA Style

Han W, Zhang C, Wang C, Liu S, Shen D, Zhou H, Han D. Impact of Forest Parkification on Color Authenticity. Forests. 2023; 14(9):1799. https://doi.org/10.3390/f14091799

Chicago/Turabian Style

Han, Wenjing, Chang Zhang, Cheng Wang, Songsong Liu, Decai Shen, Haiqi Zhou, and Dan Han. 2023. "Impact of Forest Parkification on Color Authenticity" Forests 14, no. 9: 1799. https://doi.org/10.3390/f14091799

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