Low Residents’ Satisfaction with Wetland Leisure Demand in Typical Urban Areas of the Semi-Arid Region in Western China: Spatial Variations and Their Causes

Abstract

Wetlands, as a crucial component of urban green spaces, provide important leisure services for residents. Construction of wetlands has increased with the rapid urban expansion and population growth in China over recent decades, especially in semi-arid regions with scarce water resources. However, the residents’ satisfaction with wetland leisure demand remains unclear. This study evaluated the residents’ satisfaction with wetland leisure demand by a framework of physical calculation, taking Yinchuan City, the capital city of Ningxia Hui Autonomous Region, in the semi-arid region of Western China as a case study area. Spatial variations in residents’ satisfaction and their causes were revealed by a supply–demand relationship between the population capacity of wetlands and the total population of communities under a framework of physical calculation. The results indicated that 4.22% of the study area, which covered 7.38% of the total population, was fully satisfied with wetland leisure demand. Residents’ satisfaction in the urban area as a whole is low at 0.49, with a concentric distribution pattern increasing as the distance from the central urban area increases. The high population density and scanty wetlands mainly induced a relative-low residents’ satisfaction in the central urban area, accounting for 12.02% of the area and 32.70% of the population. Meanwhile, a relative-high residents’ satisfaction in the outer ring of the central urban area accounting for 59.10% of the area and 20.63% of the population was primarily due to the adequate capacity of wetlands. Medium residents’ satisfaction was mainly attributed to the road network density, which partially transferred local wetland leisure demand to adjacent areas. Wetland constructions and internal structural optimization to improve the capacity for leisure in densely populated central urban areas could provide a feasible path to alleviate unbalanced conditions. These results could deepen the understanding of supply–demand dynamics in the wetland leisure service and provide valuable information for optimizing wetland allocation in urban construction in semi-arid regions.

1. Introduction

Wetlands are a crucial component of urban green spaces, typically comprising marsh and water bodies in cities of Western China [1,2,3]. The adequate space in and around these wetlands allows for unrestricted movement, providing vital leisure areas primarily in internal pathway areas for residents. The wetland leisure service can effectively enhance the quality of life and well-being of residents, aligning closely with the Sustainable Development Goals (SDGs) defined by the United Nations [4,5,6]. The number of wetlands with leisure functions has significantly increased, associated with the development of wetland constructions over recent decades [7,8,9,10]. Concurrently, residents’ demand for wetland leisure has been rising due to rapid urban expansion and population growth, especially in semi-arid regions with scarce water resources like Western China [11,12,13]. However, the understanding of residents’ satisfaction with wetland leisure demand remains poor due to a lack of reasonable evaluations. Exploring variations in residents’ satisfaction and their causes is still challenging, but it is crucial for understanding the supply–demand mechanisms of the wetland leisure service in semi-arid areas [4,8,11]. Furthermore, this understanding is imperative to improve wetland constructions and the quality of life for residents in Western China currently.

Numerous studies have evaluated residents’ satisfaction with wetland leisure demand in semi-arid regions using subjective methods such as questionnaire surveys and interviews [14,15,16,17]. Evaluations of residents’ satisfaction with the half-an-hour life circle have emerged in large numbers, especially in the post-COVID-19 era [16,18,19,20]. These evaluations provided insight into residents’ subjective feelings but are inherently susceptible to the influence of differences in the values and knowledge levels among those interviewed or surveyed [15,16]. Consequently, the previous evaluations struggled to objectively quantify residents’ satisfaction. Furthermore, there remains a limited understanding of spatial variations in residents’ satisfaction and their causes [14,21,22]. The supply–demand relationship between the population capacity of wetlands and the total population of communities provides a means to objectively reflect residents’ satisfaction with wetland leisure demand [16]. This relationship is effectively measured through a physical network composed of population density, wetland area, and roads. Subsequently, spatial variations in residents’ satisfaction and its dominant factors could be revealed quantitatively under the frame of physical calculation.

Yingchuan City in Western China is located in a typical semi-arid region. This city has undergone extensive urban wetland constructions over a prolonged period [23,24]. Currently, it has a relatively large number of wetlands compared to other cities in the same region. The half-an-hour life circle urban plan has been implemented recently [25,26]. However, residents’ satisfaction with wetland leisure demand, especially in urban areas, remains unclear. These points provide favorable conditions for exploring spatial variations in residents’ satisfaction with wetland leisure demand and their causes in urban areas of semi-arid regions.

This study aims to evaluate residents’ satisfaction with wetland leisure demand in a semi-arid region under a framework of physical calculation, taking the urban area of Yinchuan City in Western China as a case study area. The spatial variations in residents’ satisfaction and their causes were revealed by spatial overlay analysis and variable importance diagnosis with a Random Forest model. The results are expected to provide insight into the supply–demand dynamics of the wetland leisure service and support the optimized allocation of urban wetland resources in this semi-arid region.

2. Materials and Methods

2.1. Case Study Area

Yinchuan City is the capital city of Ningxia Hui Autonomous Region in the semi-arid region of Western China, framed by latitudes 37°29′ N to 38°53′ N and longitudes 105°49′ E to 106°53′ E (Figure 1). The terrain is high in the west and south, and low in the north and east. It is characterized by a typical temperate continental climate, with an annual average temperature of around 8.5 °C and an average annual precipitation of about 200 mm [27]. The Yellow River traverses this city, with a historical record of multiple course changes [28], nurturing numerous lakes and marshes. Moreover, the number and area of artificial wetlands have experienced rapid expansion due to urban expansion and population growth over the past few decades [23,29]. Currently, this city has a relatively large wetland area of approximately 578 km² compared to other cities in the same region. These wetlands are rich in biodiversity, inhabited by 192 plant species, 153 wild animal species, 24 nationally protected waterfowl species, and 14 endangered wild animals and plants [30]. The urban population comprises 81.74% of the total population, primarily concentrated in the urban area. The transportation network shows significant spatial differences in both level and density [23,24].

2.2. Wetland Data

A map of the wetlands was derived from the East Asia wetlands map with a 10 m resolution, provided by Wang et al. (2023) [31]. These data were produced by two-stage object-based Random Forest and hierarchical decision tree algorithms on Sentinel-1/2 images, and have a high quality in the semi-arid region. A total of 273 publicly accessible wetlands in the case study area were extracted from this data, including rivers, lakes, ponds, and marsh (Figure 2).

2.3. Population Data

The population of residential communities within the urban districts of Yinchuan City for the year 2023 was derived by multiplying the total number of households by the average number of permanent members per household. The quantity, number of permanent members, and locations of these households were obtained from a community household statistical database (https://yinchuan.anjuke.com/, accessed on 20 January 2024) and the Bulletin of the 7th Population Census of Yinchuan City. The population proportion by age group was also extracted from the Bulletin to estimate the willingness of different age groups to visit wetlands within a week (

Table 1. The proportion of different age groups in Yinchuan City and the probability of visiting wetlands within a week.

Age Groups	0–20	21–59	>60
Proportion	21.57%	65.64%	12.79%
Willingness to travel (times/week)	0.7996	1.2916	3.825

) using the method proposed by Mao et al. (2020) [32].

2.4. Road Data

The road network of Yinchuan City in 2023 was obtained from Open Street Map (OSM, https://www.openstreetmap.org/, accessed on 22 January 2024). The road network data show good performance in urban traffic patterns and have been widely applied in urban geography [33]. The roads were categorized as expressways, main roads, secondary roads, and minor roads, with travel speeds of 50 km/h, 40 km/h, 30 km/h, and 20 km/h, respectively.

2.5. Measuring Residents’ Satisfaction with Wetland Leisure Demand

Residents’ satisfaction (RS_i) with wetland leisure demand was measured by the supply–demand relationship between the total population capacity (S_i) of wetlands, considering the competition among visitors, and the total population (D_i) of the community (i), as shown in Equation (1). RS_i was divided into 3 levels of relative-low, relative-medium, and relative-high, using its mean ± standard deviation as the gradient threshold, in order to explore its spatial variations.

$${RS}_i=\frac{S_i}{D_i}$$

(1)

D_i was quantified by multiplying the resident population of i with an average daily turnover rate of 2.0 visitors, which was determined by the specific conditions of Yinchuan’s wetlands. Meanwhile, S_i was quantified by the population capacity (PC_j) within the wetland (j) and competitiveness (C_ij) of the visitors from i in the leisure service of j, as follows.

$$S_i={\sum }_{j=1}^k{PC}_j\times C_{ij}$$

(2)

PC_j was calculated by the ratio of internal pathway area within j to personal available space. The internal pathway area was extracted from the wetland data provided in Section 2.2. The personal available space was set to 70 m², following the Guidelines for Planning and Design of Urban Wetland Parks (Trial) formulated by the Yinchuan municipal government. Simultaneously, C_ij was measured by the proportion of the time cost (TC_ij) to the visitors from i to j, compared to the total time cost of all visitors to j, as shown in Equation (3). The proportion depicts the level of competitiveness among the visitors in the leisure service due to the inherent capacity constraints of j.

$$C_{\mathit{ij}}=\frac{{TC}_{ij}}{\sum _{i=1}^n{TC}_{ij}}$$

(3)

TC_ij was quantified by D_i, the probability (P_ij) of residents visiting wetlands, and the shortest travel time (T_ij) from i to j within 30 min, as shown in Equation (4). T_ij was calculated by the cost matrix between i and j. P_ij was estimated by T_ij and PC_j, as shown in Equation (5). Generally, the closer a location is to a wetland and the larger the wetland’s population capacity, the higher the probability of residents visiting that wetland.

$${\mathit{TC}}_{\mathit{ij}}={T_{\mathit{ij}}^{-\alpha }\times D}_i\times P_{\mathit{ij}}$$

(4)

$$P_{\mathit{ij}}=\frac{{\mathit{PC}}_j^{\beta }}{T_{\mathit{ij}}^{\gamma }}$$

(5)

where α represents the distance attenuation coefficient, which was determined by Wang et al. (2023) [16]. β and γ serve as the weight coefficients that account for the influence of PC_j and T_ij on P_ij, respectively. These two coefficients were determined by the range normalization of their respective variables, PC_j and T_ij.

2.6. Spatial Load of Wetland Leisure Resources

The spatial load (SL_j) of a wetland serves as a metric to reflect the utilization degree of wetland leisure resources. SL_j was quantified by the ratio of the total demand of visitors for the leisure service of j to PC_j, as shown in Equation (6). The total demand was measured by the number of visitors in j, which was calculated by P_ij and D_i. An SL_j greater than 1.0 indicates an overload of the wetland, that is, the leisure demand of visitors exceeds the wetland’s supply capacity, and vice versa.

$${SL}_j=\frac{\sum _{i=1}^n{D}_i\times p_{\mathit{ij}}}{{\mathit{PC}}_j}$$

(6)

2.7. Exploring Spatial Variations in Residents’ Satisfaction and Their Causes

The spatial variations in residents’ satisfaction with wetland leisure demand were explored by comparing high-level, medium-level, and low-level areas of RS_i. The dominant factors of residents’ satisfaction were identified by the relative importance of wetland areas, total population of the community, and road network traffic capacity to RS_i using the Random Forest model [34]. This model treated these three factors as independent variables and RS_i as the dependent variable, producing a ranking of their relative importance. The performance of the ranking results was measured using the determination coefficient (R²). Meanwhile, a spatial overlay analysis was employed to determine the matching degree of these three factors. Finally, the causes of the spatial variations in residents’ satisfaction were revealed by the aforementioned analysis.

3. Results

3.1. Spatial Differences in Wetland Load of Residents’ Leisure Demand

Figure 3 shows the wetland load of residents’ leisure demand in Yinchuan City, which was quantified using the spatial load metric. When the load value exceeds 1.0, it indicates that the demand for wetlands surpasses their supply capacity, thereby exposing the spatial load pressure faced by the wetlands. In the urban area of Yinchuan City, 59 wetlands have load values exceeding 1.0, accounting for 65.56% of the total. This reveals the overall characteristics of the wetland spatial load in Yinchuan City as “high in the center, low in the periphery”. Particularly in the central urban area and some surrounding areas, a high level of wetland spatial load is observed, indicating significant usage pressure on wetlands in these regions. The wetland spatial load level in the city center is particularly prominent, reflecting the contradiction between high population density and limited wetland resources. Among them, wetlands with load values exceeding 3 are mainly concentrated in areas such as Lijing Block, Shanghai West Road Block, Great Wall Middle Road Block, and Beijing Middle Road Block. These high-spatial-load areas exhibit a clustering trend spatially, especially in commercial and residential areas in the city center, primarily due to the dense population in these areas and the mismatch between wetland resources and population growth due to limitations in urban planning. In contrast, areas on the outskirts of the city and far from the center exhibit lower spatial loads.

3.2. Residents’ Satisfaction with Wetland Leisure Demand and Its Spatial Heterogeneity

Figure 4 illustrates a low residents’ satisfaction with wetland leisure demand in the urban area as a whole, with a mean value of 0.49, mainly distributed between 0.26 and 0.60. Yet, residents’ satisfaction shows an obvious spatial difference, with a concentric distribution, increasing as the distance from the central urban area increases. The result shows that 4.22% of the urban area, which covers 7.38% of the total population, is capable of fully meeting residents’ needs for the wetlands. This highlights a significant spatial distribution imbalance in resident satisfaction with wetland visitation within the urban area. Areas with high satisfaction levels and communities fully meeting residents’ needs are primarily concentrated in the peripheral zones of the urban area, such as Fengdeng Town, Daxin Town, Great Wall Middle Road Block, and Ninghua Road Block. Additionally, they are sporadically distributed in central Blocks like Beijing Middle Road Block and Huanghe East Road Block. Particularly, in the eastern part of the urban area, residents’ satisfaction exhibits an increasing spatial distribution pattern, with satisfaction levels decreasing closer to the central area. Conversely, areas with low satisfaction levels are mostly centralized in the central positions of the urban area, although there are sporadic communities fully meeting residents’ needs. In the northern and western parts of the urban area, overall satisfaction levels are relatively good, with most communities reaching high or above levels of satisfaction. Furthermore, these areas have a higher number of communities that can fully meet residents’ needs.

3.3. The Dominant Factors of Spatial Differences in Residents’ Satisfaction

The dominant factors of residents’ satisfaction showed an obvious difference between the low-level, medium-level, and high-level areas, as shown in

Table 2. Relative importance of community population, road networks, and wetland area to residents’ satisfaction.

Levels of Residents’ Satisfaction	Total Population of Communities	Road Traffic Capacity	Wetland Area	R2
Relatively low	20%	27%	52%	0.50
Relatively medium	32%	37%	31%	0.32
Relatively high	25%	35%	40%	0.50

. Areas of low-level satisfaction (12.02%) with the wetland visiting experience are negligible for a population of over 520,000 (32.70% of the total population). The road network density in this area meets urban standards, but the wetland infrastructure is inadequate in both quantity and capacity. The average spatial load of the four available wetlands is 3.32, significantly higher than the sustainable load for recreation and ecological balance. This spatial load reflects a significant gap between the available wetland area and residents’ demand for wetland visiting experiences. The mechanisms causing this low satisfaction are twofold: insufficient wetland quantity unable to serve a large number of residents, and existing wetlands facing immense pressure, unable to provide the services they cannot sustain (Figure 5b,c). Additionally, congestion within these few wetlands inevitably leads to a decline in their recreational value, increased maintenance challenges, and impacts on wetland lifespan and ecological functionality. Furthermore, the dense road network (Figure 5d), while facilitating travel, results in a rapid influx of visitors to the wetlands. Moreover, the greater distance of peripheral wetlands increases the time cost, reducing residents’ competitiveness in peripheral wetlands and paradoxically exacerbating the issue. Existing data paint a picture of wetland systems being overly burdened, unable to effectively meet the ecological and recreational needs of the local population.

The medium-level satisfaction area spans 113.9 km² (28.88%), comprising 18 wetlands, and can generally meet the demands of over 740,000 people. The average spatial load for wetlands in this area is 3.2. Compared to the low-level satisfaction area, the spatial load is slightly lower, but the demand for wetland resources remains significant. The road network density in this area is slightly lower at 4.6 km/km². However, the ability to quickly move outward to nearby wetland areas facilitated by higher-level road networks is a key factor. This acts as a relief valve, redistributing visitor pressure to a broader wetland network, including wetlands in neighboring areas. However, this redistribution is insufficient to significantly improve satisfaction, as indicated by the high spatial load still present on wetlands in this area. This suggests that although wetland utilization is more evenly distributed than in the low-satisfaction area, they still face significant usage close to capacity, resulting in a medium level of satisfaction.

The high-level satisfaction area covers 233.1 km² (59.10%), with a low population density of 330,000 (20.63% of the total population), encompassing 68 wetlands, with an average spatial load of 1.4. This favorable spatial load leads to a comfortable carrying capacity and reduced occurrence of overcrowding, directly related to higher satisfaction levels. Although the road network density is low at 2.87 km/km², it does not hinder the accessibility and enjoyability of wetland resources, compensated by the spatial distribution of large-capacity wetlands within reachable distances. The high satisfaction in this area is attributed to ample green spaces, allowing population dispersion across various wetland environments. This distribution minimizes the likelihood of visitor congregations, alleviating pressure from overuse. Consequently, each wetland in the area can offer a serene and sustainable interaction with nature, preserving the integrity of wetland ecological functions. The abundance of wetlands also indicates a degree of redundancy, beneficial for distributing wear, typically concentrated on fewer wetlands in denser or less resource-rich areas. The availability of larger peripheral wetlands further enhances satisfaction, providing additional recreational options and alleviating demand on urban wetlands. This creates a buffering system that enhances high satisfaction through diversified recreational points of interest and ecological service provisions.

Within the medium-level satisfaction area, a high-satisfaction anomaly area appears near the northern boundary adjacent to the low-satisfaction area. This anomaly, characterized by a cluster of five wetlands, deviates from the broader medium-satisfaction environment, creating a highly satisfactory island. High-value areas within the medium-satisfaction area indicate localized optimization, with abundant wetland resources and strong visitor accommodation capabilities. The phenomenon of high satisfaction within the medium-satisfaction area can be attributed to several factors. Firstly, the larger wetland capacity surrounding this specific area suggests recent expansions or targeted urban development initiatives. Secondly, the clustering of multiple wetlands within walking distance forms a micro-network of recreational spaces, enhancing residents’ ability to engage in various wetland experiences without venturing far, contrasting sharply with the broader medium-satisfaction area. These data present a scenario where infrastructure and resource allocation are adequate, but utilization rates indicate potential overburdening of resources. This situation requires careful monitoring to prevent satisfaction from peaking due to overuse, potentially leading to rapid wetland degradation. Maintaining a balance between high satisfaction and ecological sustainability becomes the most critical issue in these high-value areas.

4. Discussion

Spatial variations in residents’ satisfaction with wetland leisure demand and their causes in semi-arid regions remain poorly characterized [7,14]. This study evaluated residents’ satisfaction in a typical urban area of a semi-arid region in Western China based on the supply–demand relationships between the population capacity of wetlands and the total population of communities under a framework of physical calculation. Compared with previous studies using subjective methods such as questionnaire surveys and interviews [15,16,17], our study could objectively reflect residents’ satisfaction and its spatial variations. Moreover, the causes of these variations were revealed quantitatively by a mathematical and statistical analysis. The results can deepen the understanding of dynamics in cultural services of wetland ecosystems in semi-arid regions and support the optimized allocation of urban wetland resources.

Our results revealed low residents’ satisfaction with wetland leisure demand in the urban area as a whole. But residents’ satisfaction showed a concentric distribution, increasing as the distance from the central urban area increased. Compared to the areas with relative-low residents’ satisfaction, areas with relative-high residents’ satisfaction have a low population density and adequate wetland area. This kind of spatial pattern of wetlands was seen in most cities in the semi-arid region of Western China where urban areas have rapidly expanded over the past few decades [14,26,29]. The results indicated an unbalanced and inadequate supply–demand contradiction between the population capacity of wetlands and the total population of communities locally. This finding has promoted a deep understanding of dynamics in wetland leisure services in semi-arid regions.

Actually, it is difficult to solve the supply–demand contradiction of more wetland constructions in the outer ring of central urban areas with relatively low population densities [15,16]. Moreover, regulating the distribution of population density is often challenging. More wetland constructions, such as area expansion and structural optimization of internal road networks in densely populated central urban areas, could provide a feasible path to alleviate the supply–demand contradiction and could be highly cost-effective [16,35]. Therefore, improving wetland constructions in the central urban area is an important strategy for solving the supply–demand contradiction of wetlands in semi-arid urban areas such as the case study area in the future. Carrying out wetland constructions blindly and endlessly to improve their leisure service is strictly forbidden, and should meet the premise that the wetland ecosystem as a whole does not degrade. This requires the joint participation of the public and the government. Furthermore, it is also necessary to deeply consider the optimization of urban wetlands under the optimal goals of multiple wetland service functions in the above strategies.

Residents’ satisfaction with wetland leisure demand was evaluated by a framework of physical calculation, which considered the delay of wetland attractiveness to community residents. The delay of this study was depicted by the statistical distribution characteristics of survey data from similar studies [36], which could largely reflect changes in the wetland attraction of the case study area. This point provides a crucial opportunity to optimize the delay function based on mobile trajectory data. Additionally, our assessment could not fully consider individual choices, which need to be addressed in future work.

5. Conclusions

Residents’ satisfaction with wetland leisure demand in an urban area in the semi-arid region of Western China was evaluated by the supply–demand relationships between the population capacity of wetlands and the total population of communities under a framework of physical calculation. A low residents’ satisfaction in the urban area as a whole was shown with a concentric distribution that increased as the distance from the central urban area increased. The significant spatial heterogeneity was induced by the unbalanced allocation of wetland resources in the urban area. More wetland constructions appeared in the outer ring of the central urban area with a currently relatively low population. Wetland constructions and internal structural optimization in densely populated central urban areas could provide a feasible path to improve the leisure capacity and alleviate the unbalanced condition in semi-arid urban areas of Western China such as the case study area.