Characteristics of Urban Parks in Chengdu and Their Relation to Public Behaviour and Preferences

Abstract

Urban parks can offer a variety of ecosystem services such as beautifying the landscape, ecological regulation, leisure and recreation, and maintaining biodiversity. Understanding how urban parks affect people’s lives can help guide the construction and planning of parks in a direction that is more beneficial to the public. Therefore, it is worth studying the extent to which different urban parks with different characteristics affect public behaviour and preferences. This paper takes five typical urban parks in Chengdu and analyses the relationship between characteristics of the park ecosystem and public behaviour and preferences. The characteristics include the park scale, blue-green space ratio, plant diversity, and degree of re-wilding. Visit frequency, stay time, and park preference characterise public behaviour and preferences. The results show: (1) There are obvious differences in the ecosystem characteristics of the five parks: Qinglong Lake Wetland Park is the largest; the proportion of blue-green space in Jiangjiayiyuan Garden is the highest; the degree of re-wilding in Bailuwan Wetland Park is the highest; the proportion of green space and plant diversity in Guixi Ecological Park is the highest; and the proportion of blue space in Jincheng Lake Wetland Park is the highest. (2) There are differences in public behaviour and preferences for different parks. Tourists visit Qinglong Lake Wetland Park the most in spring and autumn and they choose Guixi Ecological Park instead in summer and winter. The public stays longer in Qinglong Lake Wetland Park and shorter in Jincheng Lake Wetland Park. (3) The scale of urban parks, the proportion of blue-green space, and the degree of re-wilding, especially the proportion of blue space, have a positive impact on the public’s evaluation and promote public visits. The results of the study could help improve public awareness of the relationship between park characteristics and ecological services and well-being.

1. Introduction

Urban parks are an important part of urban ecosystems and urban landscapes [1]. They serve ecological functions such as water conservation, climate regulation, environmental purification, and biodiversity maintenance. They can provide urban residents with scenery, leisure, exercise, entertainment, social networking, and other ecosystem services [2,3] that ensure urban ecological security and promote the sustainable development of human and natural systems [4]. Disturbed by different degrees of human activity, urban parks have dual attributes of both natural and artificial ecosystems. There are also great differences in the characteristics of different types of park ecosystems. Parks dominated by natural ecosystems are larger in scale, with high vegetation coverage and rich biodiversity [5], while parks dominated by artificial ecosystems are smaller in scale, with fragmented landscapes and more infrastructure [6].

Urban parks can affect people’s emotions and perceptions [7] and change public behaviour [8]. These effects may be associated with park features such as flowers [9], dense trees [10], and convenient leisure facilities [11]. Parks with the above characteristics are more popular among tourists. Natural landscapes are generally more attractive to the elderly, while young people prefer artificial facilities [12]. Most researchers study the influence of urban park characteristics on public behaviour through the relationship between spatial structure characteristics of parks and visitors’ behaviour. Greg et al. [13] used the public participatory geographic information system (PPGIS) to study the relationship between urban park types and public participation in sports activities. They found that linear parks provide the most significant overall physical benefit. Hou et al. [14] obtained data such as morning exercise types and space allocation using the behavioural annotation method. Long et al. [15] used space syntax theory to quantify the characteristics of urban spatial organisation, obtaining differential characteristics of walking and non-walking spaces and walking space preference in the elderly. They found that people tend to choose parks with high accessibility and large open spaces for morning exercises, especially those who use group dance as a morning exercise program. Research into the relationship between urban park characteristics and public preferences is based on public physiological and psychological indicators. Song et al. [16] measured tourists’ heart rates to obtain heart rate variability and tested the influence of urban parks on people’s minds and bodies through the semantic differential (SD), Profile of Mood States (POMS), and State-Trait Anxiety Inventory (STAI). This research shows that heart rate was significantly lower while walking in the urban park than in the city street. Furthermore, the urban park walk led to higher parasympathetic nervous activity and lower sympathetic nervous activity. Wu et al. [17] explored the influence of colour characteristics on tourists’ perception through 11 evaluation factors of the three spatial dimensions of the South Part of Minjiang Park. The factors included ambient color characteristics (coordination, color, richness, memory, and attractiveness) and tourist perception (comfort, pleasure, security, dullness, annoyance, and depression). Their results showed that factors correlated with tourist perception are all related to coordination. At present, research on ecosystem services of urban parks mainly adopts computer technology and network technology, including social media, geographic information technology, and modeling, as research methods [12,18,19]. They demonstrate that urban parks have an impact on the public and provide a theoretical basis for this study.

People are the main beneficiaries of urban parks. The comprehensive evaluation of physiology, psychology, and geographic information systems are used to obtain public behaviour trajectories. The differences in public behaviour and preferences among different parks can be revealed in this way but the relationship between park characteristic factors and the public is not intuitive. The current research aims to study the degree of influence of urban park characteristics on public behaviour and preferences and reveal which characteristic factor has which influence. It will make the construction of future urban parks better meet public needs; in this way, the government will improve the quality and quantity of urban parks. Urban parks are an important provider of ecosystem services [2,3,4], so they will enhance park ecosystem services and promote sustainable urban development. Therefore, this paper combines field research with participatory questionnaires (Appendix A) to identify the ecosystem characteristics of typical parks in Chengdu, reveal the differences in public behaviour and preferences for different types of parks, and explain the relationship between important characteristic factors and public behaviour. It also provides a scientific basis for park construction and management.

2. Materials and Methods

2.1. Study Area

Chengdu (102°54′ E~104°53′ E, 30°05′ N~31°26′ N) was the first place to propose the concept of the “Park City” in China. It is also the national demonstration area of China’s park cities, showing a rapid increase in the number and area of its parks [20]. This study selected five urban parks (Figure 1), namely Jincheng Lake Wetland Park, Qinglong Lake Wetland Park, Guixi Ecological Park, Bailuwan Wetland Park, and Jiangjiayiyuan Garden (Figure 1). The above parks are all parts of Chengdu Huancheng Ecological Park. The construction of Chengdu Huancheng Ecological Park is a project in progress. These five parks are relatively complete (Figure 2), while most others are still under construction. Their construction time, distance from the main urban area, infrastructure, and management levels are similar; however, the park scale, ecosystem composition and structure, and biodiversity are different. As the park conditions are similar or identical except for the few characteristics to be studied, they offer the representativeness and typicality needed for this research. Moreover, these five parks are the most popular in Chengdu, which is beneficial for sending out questionnaires.

2.2. Investigation and Analysis of Park Ecosystem Characteristics

Park characteristics are mainly divided into three categories: ecological characteristics, service characteristics, and aesthetic characteristics [21]. Usually, the diversity and spatial distribution of green plants have a greater impact on public behaviour and preferences [22]; therefore, this study analysed the characteristics of urban parks from the scale of the parks, their proportion of blue-green space [23], plant diversity [24], and degree of re-wilding [25]. This study used ArcGIS 10.8 software to obtain the total scale of the study area. The proportion of blue-green space was calculated based on land use data [26]. Green space was divided into “parks and green spaces”, “bushland”, “orchards”, “adjustable orchards”, “woodland”, “bamboo forest”, “other grassland”, “other woodland”, and “other gardens”. Blue space was divided into “dry canal”, “ditch”, “river surface”, “pond water surface”, “reservoir water surface”, and “aquaculture fish pond”. The area of land which belongs to the blue-green space was first calculated and then the obtained value was divided by the total area to obtain the proportion of blue-green space. Richness is used as an evaluation index of plant diversity [27]. The average and standard deviation can represent the level of diversity. In this study, a total of 37 quadrat surveys (10 × 10 m quadrat) were conducted in the five parks. The number of plant species in each quadrat was calculated, and the average and standard deviation were calculated among quadrats in each park. The area occupied by native plants [28] and the number of native species [29] are the keys to re-wilding; the degree of re-wilding was calculated using the following formula: The degree of re-wilding = the proportion of native plant distribution area × 0.5 + the proportion of native species × 0.5. The native plant land includes “bushland”, “woodland”, and “bamboo forest”. We calculated the area of native plant land and divided the obtained value by the green space area to obtain the proportion of native plant distribution area. Then, we counted the number of native species in each park and divided it by the number of all species in the same park to obtain the proportion of native species.

2.3. Participatory Questionnaire Survey and Analysis

From March to April 2021, a participatory questionnaire survey was carried out in the five city parks. A total of 400 questionnaires were sent out, and 377 valid completed questionnaires were collected, resulting in a response rate of 94.25%. The 377 questionnaires consisted of 160 from Qinglong Lake Wetland Park, 104 from Jiangjiayiyuan Garden, 40 from Bailuwan Wetland Park, 22 from Guixi Ecological Park, and 51 from Jincheng Lake Wetland Park. We arrived at the parks and stratified respondents by gender and age. Then, we had conversations with the respondents face to face and provided them with questionnaires which they returned as soon as they completed. In order to motivate people to participate in the survey, respondents who completed the questionnaires were each given a bottle of water.

This survey adopted stratified random sampling [30]. Stratified random sampling is to divide the units of the population into various categories according to specific standards, and then select samples from each category according to the random principle. The categories in this research included basic demographic information (gender, age, occupation, education, and accommodation), public behaviour (visit frequency and length of stay), and public preferences (behavioural activity preferences, landscape preferences, bio-aesthetic preferences, and functional value preferences). The research used a five-point Likert scale (from “1 = very dislike” to “5 = very much like”) [31] to assign scores to respondents’ preferences in a positive order. Besides this, a descriptive statistical analysis was carried out after summarising the results. According to the results, we calculated the average $\overline{X_i}$, standard deviation SD, coefficient of variation CV [32], and median M in order to study the differences in public preferences in different parks.

2.4. Data Analysis

This study used IBM SPSS Statistics (IBM SPSS Statistics for Windows. Version 26.0.0.0, IBM corp, Chicago, IL, USA, released 2019) to test the significance and undertake correlation analyses of the survey results. The public preferences (behavioural activity preferences, landscape preferences, bio-aesthetic preferences, and functional value preferences) of this survey were characterised and compared using a non-parametric test (e.g., Chi-square tests). A canonical correlation analysis (CCA) [33,34] was used to study the relationship between the independent variables (different park characteristics) and the dependent variables (public preferences).

3. Results

3.1. Differences in the Characteristics of Urban Park Ecosystems

The characteristic factors of different park ecosystems were different (

Table 1. The Characteristics of Urban Park Ecosystems.

	Total Scale/km2	Proportion of Blue Space	Proportion of Green Space	Proportion of Blue-Green Space	Plant Diversity Index	Re-Wilding Degree
Qinglong Lake Wetland Park	9.42	17.09%	60.11%	77.20%	6.71 ± 2.50	0.51
Jiangjiayiyuan Garden	1.23	5.96%	86.66%	92.59%	8.67 ± 2.89	0.39
Bailuwan Wetland Park	2.11	24.80%	63.69%	88.49%	8.47 ± 2.35	0.61
Guixi Ecological Park	1.02	1.98%	86.84%	88.82%	11.00 ± 3.52	0.39
Jincheng Lake Wetland Park	1.05	35.28%	45.91%	81.19%	10.50 ± 2.38	0.33

). The area of Qinglong Lake Wetland Park was the largest (9.42 km²), while the area of Guixi Ecological Park was only 1.02 km². In terms of the proportion of blue-green space in the park, the proportion of all five parks was more than 75%. The proportion of blue space in Jincheng Lake Wetland Park and Bailuwan Wetland Park was relatively high, while the proportion of green space in Jiangjiayiyuan Garden and Guixi Ecological Park was relatively high. There were also differences in plant diversity among the parks. The abundance of Guixi Ecological Park was the highest (11.00 ± 3.52), while the value of Qinglong Lake Wetland Park was relatively low (6.71 ± 2.50). The re-wilding degree of Bailuwan Wetland Park was the highest, reaching 0.61; the re-wilding degree of Jincheng Lake Wetland Park was the lowest, at only 0.33.

3.2. Analysis of Public Behaviour and Preferences

3.2.1. Public Demographic Information

The proportion of males and females among the respondents was 48.54% and 51.46%, respectively. The 21–40 age group accounted for the highest proportion, at 50.93%. The proportion of those with a bachelor’s degree or above was highest, at 37.67%, and the largest proportion of the population was employed, at 41.64%. A total of 61.01% of the respondents lived in the central city of Chengdu (see Appendix B 

Table A1. Demographic Information about Park Visitors.

Item	Classification	Number	Percentage
Gender	Male	183	48.54%
	Female	194	51.46%
Age Group	<20	49	13.00%
	21–40	192	50.93%
	41–60	85	22.55%
	>61	51	13.52%
Education	Primary School or less	56	14.85%
	Junior and Senior School	133	35.28%
	Technical School	46	12.20%
	Bachelor’s or more	142	37.67%
Occupation	Student	54	14.32%
	On-the-job	157	41.64%
	Free-lancer	98	26.00%
	Retired	68	18.04%
Accommodation	Central City	230	61.01%
	Suburbs	145	38.46%
	Other Cities	2	0.53%

).

3.2.2. Analysis of Public Behaviour

There was little difference in the frequency of public visits and the length of stay among the five study areas. The respondents tended to visit the park in spring and autumn and the length of stay was generally 1–6 h. In spring and summer, the proportion of visitors who travelled to the study area with more than a 15-day interval between visits accounted for 19.63% and 35.28%, respectively, while the proportion in autumn and winter was 26.26% and 40.59%, respectively (Figure 3). Qinglong Lake Wetland Park accounted for 23.13% and 13.75% of the daily visits to the park in spring and autumn, respectively. In summer and winter, Guixi Ecological Park accounted for 13.63% and 13.64%, respectively (Figure 3). Visitors who stayed for 1–3 h accounted for most of the visits, at 56.76%, and visitors who stayed for more than 6 h accounted for the least, at 2.65% (Figure 3). A total of 6.88% of the respondents stayed in Qinglong Lake Wetland Park for more than 6 h, while 15.69% of visitors stayed in Jincheng Lake Wetland Park for less than 1 h (Figure 3).

3.2.3. Analysis of Public Preferences

By calculating the medians and coefficients of variation of all visitors’ preferences, which were divided into “behavioural activity preferences”, “landscape preferences”, “bio-aesthetic preferences”, and “functional value preferences” (Figure 4), the visiting motivation and preferences of the respondents could be studied. The medians of behavioural activity preferences were lowest for “Socialisation” in Jincheng Lake Wetland Park, Qinglong Lake Wetland Park, Bailuwan Wetland Park, and Jiangjiayiyuan Garden and the coefficients of variation were highest; however, in Guixi Ecological Park it was “Photography”. In terms of landscape preferences, the median of “Recreational Park” was lowest and the coefficient of variation was highest in Jincheng Lake Wetland Park, Qinglong Lake Wetland Park, Bailuwan Wetland Park, and Jiangjiayiyuan Garden. In Guixi Ecological Park, the lowest median was for “Wetland”, and “Countryside” had the highest coefficient of variation. The five parks had the highest median score and the lowest coefficient of variation for visitors’ bio-aesthetic preferences in “Viewing Flowers” or “Lawn”. The medians of “Animal Diversity” in the five parks were the lowest and the coefficients of variation were the highest. The medians of “Landscaping” or “Protecting Environment” were the highest of the functional value preferences in Jincheng Lake Wetland Park, Qinglong Lake Wetland Park, Bailuwan Wetland Park, and Jiangjiayiyuan Garden and their coefficients of variation were the lowest. The median of “Enhancing City Image” was the lowest and the coefficient of variation was highest in Jincheng Lake Wetland Park, Qinglong Lake Wetland Park, Guixi Ecological Park, and Jiangjiayiyuan Garden. There are certain differences in public preferences between Guixi Ecological Park and the other four parks.

In this study, Chi-square tests were performed on visitors’ scores for “Behavioural Activity Preferences”, “Landscape Preferences”, “Bio-aesthetic Preferences”, and “Functional Value Preferences” in the five study areas. “Socialisation”, “Recreational Parks”, “Animal Diversity”, and “Enhancing City Image” were the categories most affected by different parks, while “Viewing”, “Lawn Park”, “Lawn”, and “Relaxing” were the least affected (

Table 2. Differences in Public Preferences on Behavioural Activity, Landscape, Bio-aesthetic, and Functional Value Preferences.

	Classification	Differences
Behavioural	Running/Cycling	X2(5) = 486.443, p < 0.01
Activity	Fitness	X2(5) = 460.915, p < 0.01
Preferences	Viewing	X2(5) = 178.103, p < 0.01
	Entertainment	X2(5) = 193.859, p < 0.01
	Socialisation	X2(5) = 778.135, p < 0.01
	Photography	X2(5) = 459.865, p < 0.01
Landscape	Wetland	X2(5) = 175.716, p < 0.01
Preferences	Forest Park	X2(5) = 179.504, p < 0.01
	Lawn Park	X2(5) = 168.045, p < 0.01
	Countryside	X2(5) = 248.958, p < 0.01
	Recreational Park	X2(5) = 326.878, p < 0.01
Bio-Aesthetic	Viewing Flowers	X2(5) = 251.154, p < 0.01
Preferences	Viewing Leaves	X2(5) = 221.997, p < 0.01
	Lawn	X2(5) = 119.950, p < 0.01
	Trees	X2(5) = 267.292, p < 0.01
	Shade	X2(5) = 266.528, p < 0.01
	Animal Diversity	X2(5) = 729.021, p < 0.01
Functional	Landscaping	X2(5) = 80.385, p < 0.01
Value Preferences	Protecting Environment	X2(5) = 120.427, p < 0.01
	Exercising	X2(5) = 140.289, p < 0.01
	Relaxing	X2(5) = 50.496, p < 0.01
	Leisure Activities	X2(5) = 85.987, p < 0.01
	Enhancing City Image	X2(5) = 267.292, p < 0.01

). This result illustrated that differences in public preferences were affected by park characteristics. The greater the coefficient of variation, the higher the level of influence.

3.3. Correlation Analysis

The parks’ characteristic factors are closely related to some public preferences and less related to others (

Table 3. Canonical correlation analysis of park characteristics and public preferences.

	Total Scale/km2		Proportion of Blue Space		Proportion of Green Space		Proportion of Blue-Green Space		Plant Diversity Index (Average)		Plant Diversity Index (Standard Deviation)		Re-Wilding Degree
	Correlation Coefficient	Significance	Correlation Coefficient	Significance	Correlation Coefficient	Significance	Correlation Coefficient	Significance	Correlation Coefficient	Significance	Correlation Coefficient	Significance	Correlation Coefficient	Significance
Running/Cycling	0.158	0.8	0.341	0.574	−0.384	0.523	−0.351	0.562	−0.193	0.755	−0.41	0.493	−0.524	0.365
Fitness	−0.382	0.526	−0.692	0.196	0.65	0.235	0.343	0.571	0.669	0.217	0.953 *	0.012	−0.64	0.245
Viewing	0.617	0.267	−0.082	0.896	−0.071	0.909	−0.381	0.527	−0.645	0.24	−0.237	0.701	−0.172	0.782
Entertainment	−0.31	0.612	0.749	0.145	−0.72	0.171	−0.415	0.487	0.576	0.31	−0.305	0.618	−0.382	0.525
Socialisation	0.121	0.846	−0.574	0.311	0.386	0.521	−0.151	0.808	0.363	0.548	0.787	0.114	−0.499	0.392
Photography	0.697	0.19	0.334	0.583	−0.396	0.509	−0.4	0.505	−0.94 *	0.017	−0.759	0.137	0.587	0.298
Wetland	0.341	0.574	0.7	0.189	−0.679	0.208	−0.409	0.495	−0.592	0.293	−0.931 *	0.022	0.236	0.702
Forest Park	−0.329	0.589	0.43	0.47	−0.346	0.569	−0.048	0.939	0.216	0.728	−0.333	0.583	−0.654	0.231
Lawn Park	0.463	0.432	0.399	0.505	−0.618	0.266	−0.887 *	0.045	0.076	0.903	−0.137	0.826	−0.236	0.703
Countryside	0.142	0.82	0.365	0.546	−0.354	0.559	−0.215	0.729	−0.346	0.568	−0.545	0.342	−0.307	0.615
Recreational Park	−0.488	0.404	−0.655	0.23	0.651	0.235	0.424	0.477	0.696	0.191	0.904 *	0.035	−0.756	0.139
Viewing Flowers	0.618	0.266	0.559	0.327	−0.624	0.261	−0.557	0.33	−0.784	0.117	−0.866	0.058	0.324	0.594
Viewing Leaves	0.135	0.829	0.206	0.739	−0.312	0.609	−0.439	0.459	0.055	0.931	−0.121	0.846	−0.761	0.135
Lawn	−0.434	0.465	−0.014	0.982	0.094	0.881	0.237	0.701	0.588	0.297	0.335	0.581	0.198	0.75
Trees	0.514	0.375	0.419	0.482	−0.59	0.295	−0.766	0.131	−0.295	0.63	−0.433	0.466	−0.405	0.498
Shade	−0.398	0.507	−0.881 *	0.048	0.958 *	0.01	0.807	0.098	0.281	0.647	0.835	0.079	0.015	0.981
Animal Diversity	−0.532	0.356	−0.426	0.475	0.496	0.395	0.483	0.41	0.471	0.423	0.516	0.374	−0.829	0.083
Landscaping	0.068	0.914	0.766	0.131	−0.711	0.178	−0.357	0.556	−0.257	0.676	−0.831	0.081	−0.116	0.853
Protecting Environment	0.243	0.694	0.417	0.485	−0.323	0.596	−0.012	0.985	−0.705	0.184	−0.797	0.106	0.562	0.324
Exercising	−0.189	0.76	−0.402	0.502	0.319	0.6	0.034	0.956	0.622	0.262	0.751	0.143	−0.264	0.668
Relaxing	0.938 *	0.018	−0.13	0.835	−0.084	0.893	−0.521	0.368	−0.904 *	0.035	−0.269	0.662	0.354	0.559
Leisure activities	−0.292	0.633	0.543	0.345	−0.526	0.363	−0.314	0.607	0.448	0.449	−0.241	0.696	−0.77	0.128
Enhancing City Image	−0.028	0.964	0.553	0.334	−0.415	0.487	0.023	0.971	−0.249	0.687	−0.636	0.248	0.795	0.108

* p < 0.05.

). The public’s evaluation of “Shade” was negatively correlated with the proportion of blue space in the park (r = −0.881, p = 0.048 < 0.05) and positively correlated with the proportion of green space (r = 0.958, p = 0.01 < 0.05). The public believed that the larger the total scale of urban parks, the stronger the function of regulating relaxation (r = 0.938, p = 0.018 < 0.05). Besides this, they believed that the richer the biodiversity, the less conducive to relaxation (r = −0.904, p = 0.035 < 0.05). Rich biodiversity led to a decrease in visitors’ interest in taking pictures (r = −0.94, p = 0.017 < 0.05), while the greater the difference in the biodiversity, the more visitors preferred fitness (r = 0.953, p = 0.012 < 0.05) and visiting recreational parks (r = 0.904, p = 0.035 < 0.05). The wetland area’s correlation was smaller (r = −0.931, p = 0.022 < 0.05) under these circumstances. A park with a higher proportion of blue-green space had a worse public evaluation of lawn parks (r = −0.887, p = 0.045 < 0.05) (Figure 5). The correlations between other public preferences and park characteristic factors were insignificant, indicating a loose relationship.

4. Discussion

The results show that the frequency of visitors visiting five different types of urban parks varied by season. Extreme weather is more likely in both summer and winter and will affect public visits to urban parks; therefore, in order to control for the impact of climate variables when studying the effect of park characteristics on the frequency of public visits, the results in spring were compared with those in autumn, and the results in summer were compared with those in winter. The results show that visitors visit more frequently in spring and summer than in autumn and winter, indicating that plant growth promotes public visits to urban parks [35]. Some studies have shown that excessive vegetation can make tourists feel insecure because it may provide hiding spaces. Whether visitors have negative feelings about dense vegetation may depend on the level of safety in the area where the park is located [36,37]. The motivations of people visiting the parks are relatively similar. Walking, running, cycling, and other physical activities, as well as viewing the scenery, are the main reasons the public visit these five urban parks. Lawns and flowers are the most popular urban park landscapes, while wetlands and lawn parks are the most popular forms of urban park space utilisation among visitors. This survey has raised public awareness of the high-level value of urban parks (such as beautifying and protecting the ecological environment or relaxing) which will help urban park managers to further build and plan spaces that are beneficial to urban populations.

The findings also highlight the differences among urban parks and the different effects of urban park characteristics on public perception and behaviour. Although Qinglong Lake Wetland Park occupies the smallest proportion of blue-green space, its area of blue-green space is the largest because of its largest total scale. According to the results of canonical correlation analysis, visitors can relax more in larger parks, indicating that the overall scale of urban parks has a positive effect on public perception and behaviour. Research in Chongqing, China, shows that several small parks promote urban park equity, while large urban parks are conducive to improving quality [38]. In general, the area of blue-green space and the number of biological species increases with the park area, which can provide better ecosystem services; thus, visitors prefer large parks. Guixi Ecological Park occupies the smallest proportion of blue space, with high plant diversity but a low degree of re-wilding. It lacks wetlands which are popular with visitors, so few people associate Guixi Ecological Park with the function of beautifying and protecting the ecological environment. The blue space in Jincheng Lake Wetland Park accounts for the largest proportion and the rest of its features are similar to Guixi Ecological Park. Many visitors believe that its role in beautifying the environment is obvious, indicating that the blue space has a high ornamental value from the perspective of public perception. In addition, people prefer blue spaces, valuing them more than green spaces in Beijing, China. Thus, citizens are more willing to pay for cultural ecosystem services associated with blue space [39]. People are more likely to use the functions of blue space, such as having picnics along rivers or lakes, fishing, and having a boat trip. Therefore, the blue areas are more highly valued. However, the shading function of parks has a close positive relationship with the proportion of green space [40]. The increase in the proportion of blue space leads to a decrease in the proportion of green space and the function of shading decreases.

Bailuwan Wetland Park has the highest degree of re-wilding and Jiangjiayiyuan Garden has the highest proportion of blue-green space. Most visitors in the two places think that their role in protecting the ecological environment is significant (Table 3), indicating that the degree of re-wilding and the total proportion of blue-green space are important indicators of the public’s evaluation of their ecological protection capacity. According to other research results, the higher the degree of re-wilding, the stronger the ecological stability and the less susceptible it is to external disturbances [41]; the higher the proportion of blue-green space, the stronger the ecological restoration [42]. Therefore, the two are the influencing factors of the park’s ability to protect the ecological environment. Qinglong Lake Wetland Park and Bailuwan Wetland Park have the highest degree of re-wilding and the lowest plant diversity, indicating that plant introduction and cultivation [43] increase the plant diversity of parks after the artificial disturbance [44]. Plant introduction and cultivation are based on human behaviour, which can bring in species that cannot enter Chengdu under natural conditions. However, it is necessary to introduce exotic species in a reasonable way in order to reduce the risk of biological invasion [45]. Plant diversity is inversely proportional to visitors’ willingness to take pictures and their degree of relaxation, indicating that visitors are less fond of green spaces that are over-planned and transplanted with lots of non-native plants [46].

5. Conclusions

Based on a field investigation and survey of five different types of parks in Chengdu, this paper calculated and obtained the characteristic factors of these ecosystems. Combined with questionnaires on public behaviour and preferences, our analysis draws the following conclusions: (1) The scale of urban parks has a positive impact on the public’s evaluation and promotes public visits. (2) From the public perspective, the area and proportion of blue space is proportional to its aesthetic value. (3) The degree of re-wilding and the total proportion of blue-green space are important indicators for the public to evaluate the ability of parks to protect the environment. (4) The characteristics of urban parks are closely related to some public behaviour and preferences (“Fitness”, “Photography”, “Wetland”, “Lawn Park”, “Recreational Park”, “Shade”, and “Relaxing”) and linked less to the others. Park ecosystem is a popular form of space utilisation in cities. Revealing the relationship between park characteristics and public behaviour and preferences is conducive to designing and managing urban parks accordingly and thus better meeting public needs.