Road Landscape Design: Harmonious Relationship Between Ecology and Aesthetics

Abstract

In view of global climate and environmental challenges, exploring sustainable urban vegetation management and development is crucial. This study aims to investigate the design strategies of urban road green space plants under the guidance of the dual theories of carbon sequestration and cooling eco-efficiency and aesthetics. In this study, Yangling, a representative small- and medium-sized city, was selected as the study area, and road green space plants were identified as the research objects. The assimilation method was employed to ascertain the carbon sequestration and oxygen release, as well as the cooling and humidification capacities of the plants. The aesthetic quality of the plants was evaluated using the Scenic Beauty Estimation and Landscape Character Assessment. Finally, we propose design strategies for landscapes with higher aesthetic and carbon sequestration and cooling benefits. The results demonstrate a clear nonlinear positive correlation. The carbon sequestration and cooling benefits of plants and the aesthetic quality, with correlation coefficients of 0.864 and 0.922, respectively. Across the same sample points, the rankings of standardized values for carbon sequestration, cooling benefits, and aesthetic quality vary minimally. This indicates that eco-efficient plants with harmonious colors and elegant forms can boost the aesthetic appeal and ecological function in road green spaces. Furthermore, the Sophora japonica Linn., Ligustrum lucidum Ait., Koelreuteria paniculata Laxm., Prunus serrulata Lindl., Prunus cerasifera Ehrhar f., Ligustrum sinense Lour., Photinia × fraseri Dress, Ligustrum × vicaryi Rehder, Sabina chinensis (L.) Ant. cv. Kaizuca, and Ophiopogon japonicus (L. f.) Ker Gawl. are proved to be ecologically dominant plants. They can be employed as the principal selected species for plant design. This study summarizes applicable design strategies for three types of green spaces: avenue greenbelts, traffic separation zones, and roadside greenbelts. The nonlinear regression model developed here provides a reference for scientifically assessing and optimizing urban planting designs.

1. Introduction

The issues of ecological destruction and environmental pollution have gotten worse with the fast pace of industrialization and urbanization, which has a direct impact on public health and quality of life. Environmental quality is directly related to public health and quality of life. Plants in urban green spaces can mitigate urban diseases such as stormwater flooding and the urban heat island effect, which are essential for improving urban ecological services [1,2]. In recent years, the focus of researchers has gradually shifted from investigating the basic ecological functions of urban plants (shade and noise reduction) to evaluating their contribution to combating global climate change [3,4]. Notably, scientific studies on the functions of plants in carbon and oxygen sequestration, as well as temperature and humidity reduction, have been a prominent area of research. At present, research on the ecological benefits of plants in terms of carbon sequestration, cooling benefits, and temperature and humidity reduction is being conducted in two dimensions: micro and macro levels. At the macro level, researchers have analyzed the abundance of plant species, spatial structure, and ecological service function of the community in ecosystems at different scales using the Citygreen, I-tree, and UFORE models under the “3S” technology. The results indicate that the value of ecological services provided by plant communities is positively correlated with the number and complexity of tree species; however, there are significant geographical constraints on the basic model parameters, and the application of the model has certain limitations [5]. At the micro level, the ecological capacity of plants in different regions and different types of green spaces for carbon fixation and oxygen release, cooling and humidification, dust and noise absorption, and other ecological capabilities have been extensively studied using biomass and assimilation methods [6,7]. Studies reveal that plant eco-efficiency exhibits spatio-temporal dynamics and shows variability in different regions [8,9]. Therefore, systematic measurements of the ecological benefits of plants in different geographical areas are needed to explore universal laws. The assimilation method measures the feedback effect of plants in improving the urban microclimate by estimating the values of carbon sequestration, oxygen release, temperature reduction, and humidity increase per m² of greening area of a single plant in the surrounding 100 m² of 100 m thick atmosphere, which is suitable for small-scale and short-time ecological research [10], and it can accurately determine the carbon sequestration and temperature reduction level of different plants in urban areas.

An extensive review of landscape professionals found that despite the wealth of research on plant ecology available today to provide a theoretical underpinning for design, the practical application of plant ecology is limited to a few concepts. The conclusion suggests that “landscape ecology theory is not yet deeply integrated into the planning process” and that professionals are aware of applying landscape ecology theory but are not clear about how to transfer the knowledge, which is the main obstacle [11,12]. Other studies have shown that some high-ecological-value landscapes exhibit low aesthetic value, while appropriate landscape design can achieve both aesthetic and ecological goals [13]. Therefore, studying how to enhance the aesthetic value of highly eco-efficient landscapes may become a key step in realizing the effective integration of landscape ecology theory into the actual planning process.

The current discussion on the relationship between plant ecology and aesthetic quality is reflected, on the one hand, in the fact that ecology is one of the indicators that are routinely applied in aesthetic evaluation studies [14]. It is generally recognized that landscapes of higher ecological quality are more attractive to the public [15]. On the other hand, ecological aesthetics theory plays a role in regulating the perception of design aesthetics to enhance ecological awareness [16]. This is highlighted in the ecological restoration project but rarely mentioned in everyday planning and design [17]. A synthesis of environmental aesthetics theories mentions that design interventions and science education might promote the unification of landscape aesthetics and ecological values [13]. Tribot et al. suggested integrating aesthetic research with an understanding of ecological functions at the species and landscape level [18]. In developing green eco-indicators for evaluating the aesthetic value of urban green ecological landscapes, researchers suggested that urban vegetation is an important resource for building green ecological landscapes [19]. Scholars have used the I-tree model to determine the ecological benefits of park plants from a macroscopic perspective and the SBE method to assess the effects of plant landscapes from a subjective perspective. After analyzing the data from both aspects, it is proposed that dominant tree species contribute more to the total ecological benefits of green space in urban parks and that the ecological and landscaping effects of native tree species are better [20]. It also demonstrates that an integrated study of plant ecology and aesthetics is an effective way to explore ways to enhance the aesthetic value of highly eco-efficient landscapes. However, that study only provides suggestions for urban vegetation landscape planning and design from the macro perspective of native tree species, and further exploration is needed to refine the design strategies. Therefore, the question arises as to how to implement some specific means to enhance the aesthetic effect of high-efficiency planted landscapes in design.

Currently, the Scenic Beauty Estimation (SBE) method is a common tool for studying the aesthetic quality of plants and is usually used to assess various aspects such as urban micro-renewal, traditional village conservation, and scenic forest quality assessment. This facilitates the analysis of the site’s surrounding environment preferences, societal requirements, human social desires, etc. As well as the site’s post-design landscape quality, tourism value, and sustainability, measuring the quality of landscape aesthetics by assessing the consistency of preferences for each landscape factor is a fairer and more accurate method of evaluating plant aesthetics [21,22,23,24]. Scholars believe that landscape aesthetic attributes are a joint product of biophysical elements and public perceptual, cognitive, and affective processes; a balanced view to formulate a comprehensive evaluation of objective indicators and subjective human preferences can provide a more comprehensive response to landscape aesthetic effects [25]. Terkenli et al. suggested that Landscape Character Assessment (LCA), as a widely recognized and established landscape assessment method, can be used to help evaluators make choices with more objective data, and the difficulty of its application lies in the screening of evaluation indicators [26]. The results of previous studies have shown that plant life type, color difference, growth status, and green coverage are important factors affecting the aesthetic effect of plants [27,28] and are closely related to the strength of the ecological capacity of plants in sequestering carbon and releasing oxygen and cooling and humidifying [29,30]. From this point of view, it may be possible to construct a subjective–objective combination of aesthetic evaluation index system with the help of the above basic physiological factors of plants in order to assess the aesthetic quality of the landscape more rigorously.

On the construction of a coupled model of plant ecology and aesthetics, a study of interviews and questionnaires with landscape experts concluded that the development of tools for evaluating the aesthetic and ecological quality of green spaces is necessary and available [31]. Correlation analyses can be used to quantify the degree of correlation between the aesthetic quality of plants and the variables of carbon sequestration and cooling in urban road green spaces [32]. However, subjective factors may influence the landscape’s aesthetic quality, and there is no fixed slope between it and ecological values. Scholars have suggested that such data can be fitted using nonlinear models [33]. Others have proposed that polynomial regression can fit the data more accurately by increasing the number of polynomials to increase the prediction accuracy of curvilinear models [34,35].

Starting from a micro perspective, we comprehensively analyze the ecological function and aesthetic value of carbon sequestration and cooling by plants in road green spaces, aiming to discuss three aspects: the strategy of constructing highly eco-efficient plant landscapes, the aesthetic quality of highly eco-efficient plant landscapes, and the design strategy of plant landscapes under the guidance of the dual theories of ecology and aesthetics, and finally, to propose a design strategy with practical significance. At the same time, we constructed nonlinear multinomial binary regression model models to help landscape professional practitioners further scientifically assess and optimize plant allocation options, thus promoting the sustainable management of urban vegetation and achieving the core goal of ecological construction for human well-being.

2. Materials and Methods

2.1. Research Framework

In this study, we based on the classification standard of “Urban Road Greening Design Standard CJJ/T75-2023” [36] to classify road green spaces. As there are few green spaces in urban areas for traffic islands and parking lots, between March and May 2023, we focused our field research on the current status of the flora, growth pattern, and configuration of three types of green spaces: roadside greenbelts, traffic separation green zones, and avenue greenbelts. By counting the frequency of use of plants in all road green spaces, we identified 17 plants with a frequency of use greater than 15% as the main research objects experimentally measured their carbon sequestration and cooling benefits and further analyzed the data and shortlisted ten eco-efficient plants. After that, we photographed the sample sites in the urban area where high-eco-efficiency plants dominated. Thirteen sample points that were not identical were selected based on road density and landscape composition. We calculated the carbon sequestration and temperature reduction values per unit land area of the 13 sample sites by formulas, used SBE and LCA methods to assess the aesthetic quality of the sample sites, and discussed strategies for designing planted landscapes guided by the dual theories of ecology and aesthetics. Finally, a nonlinear regression model was established (Figure 1).

2.2. Overview of the Research Area

The Yangling Demonstration Area (34°14′–34°20′ N, 107°56′–108°08′ E) is located in the Xianyang City administrative region of Shaanxi Province, China (Figure 2). The region has a warm temperate, semi-humid continental monsoon climate. With a significant year-round temperature variation, the multi-annual average temperature is 12.8 °C. The multi-annual average precipitation is 635.1 mm, whereas the total multi-annual solar radiation is 114.86 kcal·cm⁻². The distinctive hilly topography between the Qinling and Weibei mountain ranges results in suboptimal air pollution dispersion conditions in the Yangling metropolitan region, exacerbating the heat island effect. In accordance with the “Yangling Demonstration Zone to implement the provincial air pollution control special inspector report rectification program Yangfa No. 11” [37] notification document, Yangling persists in executing environmental remediation efforts, thereby providing environmental conditions for research on the aesthetic value and ecological function of road plants [38].

2.3. Measurement of Ecological Values

The frequency of application of road green space plants was defined as the percentage of occurrence of a particular plant in the sample plots studied [39]. We selected 17 road green space plants with an application frequency of more than 15% (Figure 3), of which three are evergreen (Pinus bungeana Zucc. ex Endl., Eriobotrya japonica (Thunb.) Lindl., Ligustrum lucidum Ait.); seven are deciduous (Sophora japonica Linn., Koelreuteria paniculata Laxm., Prunus persica Batsch. var. duplex Rehd, Prunus serrulata Lindl., Aesculus chinensis Bunge, Ginkgo biloba L., Prunus cerasifera Ehrhar f.); five are shrubs (Ligustrum × vicaryi Rehder, Berberis thunbergii cv. Atropurpurea, Photinia × fraseri Dress, Ligustrum sinense Lour., Sabina chinensis (L.) Ant. cv. Kaizuca); and two are grasses (Poa annua L., Ophiopogon japonicus (L. f.) Ker Gawl.). The six plants were selected for each tree, six 3 m × 3 m sample plots were set up for each shrub, and six 1 × 1 m sample plots were set up for each grass. In June–September 2023, the photosynthesis indexes of the plant growing season were measured using the Li-6400 portable photosynthesis meter. Four adjacent sunny days with sufficient light, no wind, and few clouds were selected each month, and measurements were taken at 2 h intervals from 8:00 to 18:00, using a standard transparent leaf chamber. Five fully expanded leaves were randomly selected from the midpoint of the outer 1/3 layer on the west side of the plant canopy or off the inner part of the plant. Five sunlit leaves were selected from each sample plot for shrubs and herbs, and five instantaneous photosynthesis indices were recorded for each leaf and averaged when the system was stabilized, and the order of the measurements of the tree species remained constant in each period. We used assimilation equations to calculate carbon fixation and oxygen release, cooling, and humidification per unit land area of plants [40] (see Appendix D). We further analyzed and shortlisted ten ecologically efficient plants.

2.4. Aesthetic Evaluation Sample Points

Sample spots with ten highly eco-efficient plants as the main body were photographed between 5 and 8 May 2024 from 9:00 to 15:00 within the road green spaces across the district. Using an OPPO smartphone, the apparatus was positioned at eye level (1.7 m above the ground), with horizontal shots and a consistent resolution of 300 dpi to minimize the effect of extraneous elements [41]. Each plant grouping was photographed more than three times, resulting in a total of 60 plant sample points and 350 photographs. We categorized the sample points according to green space type, plant composition, and life type, selected additional sample points in the dense road network area based on the road network density analysis in the urban area (Figure 4A), and finally identified 13 sample points with non-exactly the same species composition. The distribution of the sample points in the urban area is shown in Figure 4B. A clear and comprehensive photo was selected for each sample site, and the photos were numbered according to the type of green space in which the sample site was located and the characteristics of the life type, as shown in Figure 5.

2.5. Aesthetic Evaluation Methods

For the SBE evaluation, we randomly distributed 220 questionnaires to the public in August 2024 and obtained 201 valid questionnaires. Eight indicators were selected to evaluate the plant landscape based on the components of the road plant landscape and other evaluation indicators in the literature (

Table 1. Quantitative evaluation table of landscape characteristics.

Marking Scheme	1	2	3	4	5	Focus
Floristics	1	2	3	4	>4	Richness of element variety
Configuration structure	Grass/Shrub	Shrub/Grass	Tree/Grass	Tree/Shrub	Tree/Shrub/Grass	Lifestyle
Canopy density	≤20%	20%–40%	40%–60%	60%–80%	≥80%	Degree of ground coverage
Growing status	Declining	Fair	Better	Healthy	Lush	Health status and growth dynamics
Leaf characteristic	Unattractive	Less attractive	Average	Rather attractive	Very attractive	Physiological and morphological characteristics of leaves
Vertical greening effect	None	Single	Rather single	Eather diverse	Diverse	Vertical spatial coverage and dispersion impacts
Spatial characterization	Dispersed	Rather dispersed	Medium	Rather coordinated	Coordinated	Sample points openness and relative position of elements
Environmental harmonization	Dysfunctional	Rather dysfunctional	Moderate	Rather harmonious	Harmonious	Interrelationships between sample points and their surroundings

Note: The specific scoring criteria are shown in Appendix C.

). Forty professionals were selected to score the photos of the 13 sample points according to the evaluation criteria presented in Table 1. We have harmonized the SBE evaluation and quantitative landscape character evaluation scales to facilitate the calculation of SBI values. Five levels of evaluation indexes were selected, with the scores 1, 2, 3, 4, and 5 representing very dislike, not too much, average, more like, and like, respectively.

2.6. Ecological Data on Plant Sample Points

Given the differences in total land area among the 13 sample sites, we quantified the level of eco-efficiency by calculating the average land area carbon sequestration and cooling benefits of the sample sites. Specifically, we first measured the area covered by each plant in each sample site. We multiplied it by its corresponding carbon sequestration and cooling per unit of land area to obtain the total ecological value of each sample site, which was then divided by the total area of the sample site to calculate the average land area carbon sequestration and cooling value of the site. In August 2024, we used a tape measure to determine the total area of the sample site (S) and an XR850 laser altimeter (Xinrui, China) to measure the area covered by each plant in the sample site (C), which was calculated using the following formula:

$${CEB}_i={\sum }_{o=1}^n\left(W_o\times C_o\right),$$

(1)

$${AEB}_i={CEB}_i/S_i$$

(2)

In the formula, AEB_i is the average ecological value of the i-th sample point; CEB_i is the total ecological value of the i-th sample point; W_o is the carbon sequestration cooling benefit per unit land area for the o-th plant in the sample point; C_o is the cover of the o-th plant in the sample point; n is the total number of plant species in the i-th sample point; and S_i is the total green space area of the i-th sample point.

2.7. Data Processing

2.7.1. Standardization of Aesthetics

This study standardized the data in order to eliminate aesthetic differences between participants:

$$Z_{\mathit{ij}}=\left(R_{\mathit{ij}}-{\overline{R}}_i\right)/S_j,$$

(3)

$${\overline{Z}}_i=\sum _{j=1}^{N_i}{Z}_{\mathit{ij}}/N_i$$

(4)

In the formula, Z_ij is the j-th observer’s standardized value for the i-th sample point; R_ij is the evaluation score of the j-th participant for the i-th sample point; $\overline{R}$_i is the average of all evaluation scores for the i-th participant; S_j is the standard deviation of all evaluation scores for the j-th participant; N_i is the total number of participants; and $\overline{Z}$_i is the SBE value of the i-th sample point.

2.7.2. Scenic Beauty Index (SBI)

The SBI value is calculated by combining the results of the SBE evaluation and landscape character evaluation with the following formula:

$${\mathit{SBI}}_i={\overline{Z}}_i\times 50\%+{\overline{Y}}_i\times 50\%$$

(5)

In the formula, SBI_i is the Scenic Beauty Index of the i-th sample point, and $\overline{Y}$_i is the standardized mean value of landscape features at the i-th sample point.

2.7.3. Data Statistics and Analyses

Data were processed using the following software: SPSS 26.0 for Pearson and cluster analyses and Origin 2022 software for plotting correlation coefficients, pie charts, and dot plots.

2.7.4. Establishment of Regression Model

The ecological data on plant sample points and the Scenic Beauty Index served as dependent variables for a nonlinear multinomial binary regression model implemented using the Matlab R2024a software program via the Spearman process.

3. Results

3.1. Plant Ecological Benefits

3.1.1. Survey of Road Green Spaces

We used Engler’s systematic classification method to scientifically classify all the plant resources surveyed. Our findings showed that there are 115 species of plants in the green space of Yangling Road green spaces, belonging to 44 families and 82 genera. The 115 plant species in the road green space of the Yangling urban area exhibited a nonuniform, with only 17 plant species having an application frequency of more than 15% (Figure 6). Although some plant species that are appropriate for local cultivation were introduced to improve the landscape features and plant species diversity of the road green space, native species nevertheless dominated the space’s overall performance.

3.1.2. Carbon Sequestration and Cooling of Plants

The daily mean values of carbon sequestration and oxygen release per unit of land area and the amount of cooling and humidification of 17 native plants with more than 15% application frequency in June–September 2023 are shown in

Table 2. The average daily carbon sequestration and cooling of plants in June–September.

Species	Assimilation (mmol·m−2·s−1)	Carbon Sequestration (g·m−2·d−1)	Oxygen Release Quantity (g·m−2·d−1)	Transpiration (mol·m−2·d−1)	Cooling Values (°C)	Humidification Amount (g·m−2·d−1)
Sophora japonica Linn.	100.18	8.10	5.89	100.18	0.80	4142.73
Ginkgo biloba L.	102.16	12.55	9.13	26.63	0.32	1673.15
Ligustrum lucidum Ait.	151.23	15.98	11.62	45.00	0.47	2431.29
Koelreuteria paniculata Laxm.	248.06	23.05	16.76	64.93	0.60	3084.74
Eriobotrya japonica (Thunb.) Lindl.	148.85	13.15	9.56	37.31	0.33	1685.17
Prunus persica Batsch. var. duplex Rehd	160.06	7.59	5.52	59.95	0.28	1454.69
Prunus serrulata Lindl.	174.15	18.70	13.60	59.10	0.63	3245.16
Aesculus chinensis Bunge	90.91	8.66	6.30	33.37	0.31	1625.82
Prunus cerasifera Ehrhar f.	130.36	14.81	10.78	86.58	0.97	5030.01
Pinus bungeana Zucc. ex Endl.	103.77	11.12	8.09	66.78	0.71	3659.09
Average value of trees	140.97	13.37	9.73	53.29	0.54	2803.19
Ligustrum × vicaryi Rehder	173.81	19.66	14.30	90.58	1.01	5239.19
Berberis thunbergii cv. atropurpurea	106.17	15.89	11.55	62.36	0.92	4772.18
Photinia × fraseri Dress	244.10	31.47	22.89	65.81	0.84	4339.01
Ligustrum sinense Lour.	168.95	20.65	15.02	100.88	1.22	6305.44
Sabina chinensis (L.) Ant. cv. Kaizuca	111.33	18.82	13.69	86.13	1.43	7445.09
Average value of shrub	160.87	21.30	15.49	81.15	1.08	5620.18
Poa annua L.	92.11	3.54	2.57	85.64	0.32	1679.56
Ophiopogon japonicus (L. f.) Ker Gawl.	89.51	9.60	6.98	44.05	0.47	2416.45
The average value of grass	90.81	6.57	4.78	64.85	0.40	2048.00

. Shrubs have 1.59 times higher mean daily values of carbon fixahhhhhtion and oxygen release values than trees and 3.24 times higher than grasses. Similarly, shrubs have mean daily values for cooling and humidification that are 2.70 times higher than grasses and 2.00 times higher than trees.

Cluster analysis was performed on the June–September carbon sequestration and cooling values per unit area of the plant, and subdivisions were made at a distance of 5 to divide the capacity into three levels (Figure 7). The first level of carbon sequestration was between 23.05 g·m⁻²·d⁻¹ and 31.47 g·m⁻²·d⁻¹, including Photinia fraseri and Koelreuteria paniculata, with high carbon sequestration capacity. The second level of carbon sequestration was between 12.55 g·m⁻²·d⁻¹ and 19.66 g·m⁻²·d⁻¹, including Ligustrum vicaryi, Sabina chinensis, and others, with moderate carbon sequestration capacity. The third level of carbon sequestration ranged from 3.54 g·m⁻²·d⁻¹ to 8.10 g·m⁻²·d⁻¹, including Poa annua, Pinus bungeana, and others, with weaker carbon sequestration capacity (Figure 7A).

The first level of cooling value was between 1.01 °C and 1.43 °C, including Sabina chinensis, Ligustrum sinense, and Ligustrum vicaryi, with strong cooling capacity. The second level of cooling value was between 0.80 °C and 0.97 °C, including Prunus cerasifera, Berberis thunbergii, and others, with moderate humidification capacity. The third level of cooling value was between 0.71 °C and 0.32 °C, including Pinus bungeana, Prunus serrulata, and others, with weaker humidifying capacity (Figure 7B). Overall, Chinese scholartree, Ligustrum lucidum, Koelreuteriapaniculata, Prunus serrulata, Prunus cerasifera, Ligustrum sinense, Photinia fraseri, Ligustrum vicaryi, Sabina chinensis, and Ophiopogon japonicus are the ecologically dominant plants in the road green spaces of Yangling City.

3.2. Ecological Values

The carbon sequestration and cooling values of the 13 sample points were calculated using Equations (1) and (2) (for sample points with repetitive plant configurations, calculations were made in separate and intact plant groups). In terms of mean carbon sequestration, TG-A and TS-A stood out with values of 305.77 g·m⁻²·d⁻¹ and 110.31 g·m⁻²·d⁻¹, respectively, followed by TSG-R1 located in the roadside green zone with a value of 69.94 g·m⁻²·d⁻¹, followed by TS-Z, TSG-A, S-Z, TSG-R2, TG-R, SG-A2, SG-A1, TSG-R3, S-R, and G-R. In terms of mean cooling value, the quantitative values of TG-A and TS-A were still outstanding, with values of 11.07 °C and 8.87 °C, respectively, and then, in order from highest to lowest, the values of TSG-R1, TSG-R2, TS-Z, TSG-A, TG-R, S-Z, SG-A1, SG-A2, TSG-R3, S-R, and G-R (

Table 3. Ecological data on plant sample points.

Sample	Position	Green Area (m2)	Carbon Sequestration (g·m−2·d−1)			Cooling Values (°C)
			Total	Mean Value	Standardized	Total	Mean Value	Standardized
G-R	Roadside	91.000	322.140	3.540	−0.611	29.120	0.320	−0.634
S-Z	Dividing strip	30.000	999.609	33.320	−0.246	31.537	1.051	−0.421
S-R	Roadside	10.300	71.449	6.937	−0.570	4.221	0.410	−0.608
TS-Z	Dividing strip	25.330	1104.198	43.593	−0.120	44.233	1.746	−0.216
TS-A	Street tree strip	10.440	1151.665	110.313	0.698	92.633	8.873	1.865
TG-R	Roadside	73.800	2241.063	30.367	−0.282	97.699	1.324	−0.342
TG-A	Street tree strip	1.000	305.772	305.772	3.094	11.072	11.072	2.508
SG-A1	Roadside	63.000	520.801	8.267	−0.553	34.876	0.554	−0.567
SG-A2	Roadside	38.500	372.000	9.662	−0.536	18.612	0.483	−0.588
TSG-A	Street tree strip	73.600	2595.580	35.266	−0.222	126.593	1.720	−0.225
TSG-R1	Roadside	63.000	4406.290	69.941	0.203	159.476	2.531	0.012
TSG-R2	Roadside	70.000	2125.738	30.368	−0.282	132.208	1.889	−0.175
TSG-R3	Roadside	107.300	747.480	6.966	−0.569	44.351	0.413	−0.608

).

The avenue greenbelts have a high mean value of total landscape ecological quantity due to their limited green area and high tree canopy density. The traffic separation green zones, due to the lack of large trees, have a slightly lower total ecological quantity than the avenue greenbelts. The overall ecological quantity of roadside greenbelts varies with plant species richness; in general, the higher the richness, the higher the overall ecological quantity.

3.3. SBE Standardization

The SBE evaluation scores were standardized according to Formulas (3) and (4) (

Table 4. Evaluation results of SBE.

Sample	Mean Value	Standardized Mean	Sample	Mean Value	Standardized Mean	Sample	Mean Value	Standardized Mean
G-R	3.32	−0.296	TG-R	3.49	−0.138	TSG-A	3.76	0.341
S-Z	3.53	−0.022	TG-A	3.59	0.120	TSG-R1	3.57	0.045
S-R	3.47	−0.187	SG-A1	3.52	−0.066	TSG-R2	3.60	0.089
TS-Z	3.57	0.083	SG-A2	3.52	−0.137	TSG-R3	3.52	−0.007
TS-A	3.61	0.175

Appendix E shows the reliability and validity of the questionnaire, the consistency analysis of the population structure, and the landscape evaluation of different groups. Participant aesthetic stability analyses are presented in Appendix B.

). The scenic beauty rating of 13 sample points is medium, with six sample points rated positive and seven rated negative. The highest SBE standardized mean values are 0.341, and the lowest is −0.296. The road green spaces in Yangling City have some variation in SBE among the different plant landscape types but with minor differences. In comparison, landscapes with rich plant species, complete vertical structure, high leaf color saturation, and ordered space have a higher ranking in scenic beauty degree; landscapes with chaotic or empty space, few plant species, and little change in plant color have a lower ranking.

3.4. Landscape Characteristic Values

Floristics, configuration structure, and canopy density are quantitative indicators with precise values, while other indicators are qualitative and show greater variability (

Table 5. Landscape characteristic score.

Sam.	Floristics	Config-Struct	Growing Status	Leaf Characteristic	Canopy Density	Vertical Greening Effect	Spatial Characterization	Environmental Harmonization	Mean Value	Standardized Mean
G-R	1.000	1.000	3.450	2.660	1.000	1.580	3.113	3.321	2.141	−0.732
S-Z	2.000	1.000	3.792	3.620	3.420	3.190	3.755	3.943	3.090	0.155
S-R	1.000	1.000	3.245	2.736	2.700	2.920	2.981	2.962	2.443	−0.607
TS-Z	3.000	4.000	3.717	3.580	4.620	3.280	3.604	3.566	3.671	0.038
TS-A	3.000	4.000	4.230	3.491	5.000	4.210	4.250	4.280	4.058	0.711
TG-R	3.000	3.000	3.620	3.450	4.460	3.660	3.264	3.300	3.469	−0.086
TG-A	2.000	3.000	4.132	3.910	5.000	3.620	4.080	3.755	3.687	0.398
SG-A1	4.000	2.000	3.321	3.400	2.530	3.320	3.019	3.434	3.128	−0.239
SG-A2	2.000	2.000	3.400	3.870	2.300	3.300	3.415	3.430	2.964	−0.147
TSG-A	5.000	5.000	2.679	3.075	4.680	4.040	3.698	3.887	4.007	−0.074
TSG-R1	4.000	5.000	3.906	3.750	5.000	3.830	3.434	3.509	4.054	0.166
TSG-R2	3.000	5.000	4.470	4.245	4.760	4.430	3.679	4.210	4.224	0.743
TSG-R3	3.000	5.000	3.810	2.887	1.000	3.720	2.358	3.547	3.165	−0.326

). An analysis of the means, trends, and standard deviations of the landscape characteristic scores was carried out (Figure 8). The landscape characteristics at the 13 sample points have generally high means; the top three are growth status, landscape coordination, and canopy density; leaf characteristics, configuration structure, and floristics have the lowest means. This indicates that Yangling’s urban road green areas have comparatively acceptable landscape plant growth status, with good stand quality and proper maintenance, but the distribution of plant species could be more varied, and the plant color configuration tends to be monotonous. The comparison of the standard deviations of the sample points shows that there are certain differences in the visual perception of the evaluators, with the configuration structure having the largest standard deviation and the leaf characteristics having the smallest, indicating that the differences in the perception of plant leaf characteristics between them are smaller.

3.5. SBE and Feature Evaluation Correlation

In the Pearson correlation analysis, we investigated the relationships between SBE and eight landscape characteristic values (Figure 9). Each point represents an individual sample point, and the lines connecting these points signify the degree of correlation. Green lines show a negative correlation, and red lines show a positive correlation; the stronger the correlation, the thicker and deeper the line. The SBE exhibits a positive correlation with floristics, configuration structure, canopy density, vertical greening effect, spatial characterization, and environmental harmonization, with more pronounced correlations observed with vertical greening effect, canopy density, and floristics.

3.6. SBI Values

Using Equation (6), the Scenic Beauty Index (SBI) for the sample points was determined (

Table 6. SBI values the sample points.

Sample	SBI	Sample	SBI	Sample	SBI	Sample	SBI	Sample	SBI
G-R	−0.514	TS-Z	0.061	TG-A	0.259	TSG-A	0.134	TSG-R2	0.416
S-Z	0.066	TS-A	0.443	SG-A1	−0.153	TSG-R1	0.106	TSG-R3	−0.166
S-R	−0.397	TG-R	−0.112	SG-A2	−0.142

). The highest SBI value is attributed to TS-A (0.443), whose rich floristics and vertical greening effect create a strong sense of space and volume, and the overall coordination is good. The lowest SBI value is for G-R (−0.514), a graminaceous life form with lower scores in floristics, leaf characteristics, and spatial distribution.

3.7. SBI and Ecological Values

Using correlation analysis, this study explored the harmonious and unified relationship between the ecological benefits and aesthetic quality of urban vegetation. The standardized values of carbon sequestration (W_C) and cooling values (T) per unit of land area represent the ecological benefits of the plants, while the SBI represents the aesthetic quality. The scatter plot data were discontinuous and did not meet the assumptions of normal distribution and linearity (Figure 10), hence the choice of Spearman correlation analysis. The Spearman correlation coefficient between SBI and W_C was 0.864, with a p-value less than 0.01, and between SBI and T was 0.922, with a p-value less than 0.01. According to the definition of Spearman correlation coefficients, both W_C and T have significant positive correlations with SBI, suggesting that improving the aesthetics of road green spaces may be an essential component of urban ecological design.

The Matlab software package was used to run the Spearman program for a nonlinear multivariate binary regression model, with W_C and T as independent variables and SBI as the dependent variable. The linear regression formula was solved in Equation (6):

$$\mathit{SBI}=0.3311-0.6531W_C+1.3598T+0.2404W_C^2-0.5989T^2({\mathrm{R}}^2=0.7964)$$

(6)

In the formula, the increase in plant landscape W_C and the removal of T can explain 79.64% of the variation in SBI. The regression curve graph was obtained by curve fitting (Figure 11), and the predicted value interval of SBI (Y1) can be obtained by manually inputting the predicted values of W_C (X1) and T (X2), with exported rmse = 0.1578, indicating that this regression model has an excellent predictive effect.

4. Discussion

4.1. Plant Ecology and Aesthetic Assessment

4.1.1. Construction of Highly Eco-Efficient Landscapes

The analysis of carbon sequestration and cooling benefits of road green space plants indicates significant variability among different plant species, generally following the pattern of shrubs > trees > grasses, aligning with previous research [30]. The probable cause is that the shrubs in the urban road green spaces of Yangling are artificially cultivated and maintained, with minimal inter-plant spacing, dense leaf distribution, and a high leaf area index, thus leading to high carbon sequestration and cooling benefits per unit of land area. Among the 17 plant species, Chinese scholartree, Ligustrum lucidum, Koelreuteria paniculata, Prunus serrulata, Prunus cerasifera, Ligustrum sinense, Photinia fraseri, Ligustrum vicaryi, Sabina chinensis, and Ophiopogon japonicus exhibit high levels of carbon sequestration and cooling. This may be associated with their growth indices and strong adaptability to the green space environment. Chinese scholartree, Prunus cerasifera, and Photinia fraseri are light-loving and have a high light saturation point, indicating a substantial photosynthetic potential under the high-temperature conditions of road green spaces. Ligustrum sinense, Ligustrum vicaryi, and Ophiopogon japonicus can perform photosynthetic activities at a high frequency under low light conditions beneath the canopy of tall trees. Ligustrum lucidum, Prunus serrulata, and Sabina chinensis possess a high average leaf area index per individual plant, resulting in strong carbon sequestration and cooling capabilities per unit land area. Plants with high saturation points and growth indices have certain growth advantages in road green spaces, which is similar to related research findings [42,43]. Further, differences in plant carbon sequestration and cooling are also related to their leaf mass. Plants with leathery leaves, such as Photinia fraseri and Ligustrum vicaryi, reduced moisture loss due to thicker cuticles, so their cooling ability is strong. On the contrary, plants with papery leaves, such as Poa annua, Prunus persica, and Aesculus megaphylla, are less capable of cooling. The serrated edges of the leaves of Prunus serrulata and Koelreuteria paniculata help to enhance the reflection and refraction of light, which in turn enhances the intensity of photosynthetic activity. The high ecological efficiency of Prunus serrulata and Koelreuteria paniculata has also been confirmed in related studies [40].

The ecological value research results indicate that the highest carbon sequestration and cooling benefits per unit of land area are found in sample points TG-A and TS-A, both of which are avenue greenbelts with large trees. The lowest sample points are TSG-R3, S-R, and G-R, all of which are roadside greenbelts. The highest TG-A carbon sequestration and cooling values are 305.77 g·m⁻²·d⁻¹ and 11.07 °C, respectively, which are more than 10 times higher than the lowest sample G-R. It can be seen that constructing more sample sites with high carbon sequestration and cooling benefits can have a significant effect on the improvement of the road green space environment. Take a campus as an example: trees at Amity University, Noida, can sequester 139.9 tons of carbon a year [44]. The rules for the construction of high-eco-efficiency sample sites can be approached from the perspective of different green space types. The overall ecological value of avenue greenbelts is slightly lower than that of traffic separation green zones, while the overall ecological value of roadside greenbelts varies. However, the more complete the life form of the sample, the higher the average level. This may be due to the lack of large trees in avenue greenbelts, and plant species and structural types influence the overall ecological value of roadside greenbelts. Comparisons reveal that sample points with large canopies, sprawling branches, and continuous shrub belts have higher carbon sequestration and cooling levels, similar to related research findings [45]. The possible reason is that trees with sprawling branches can better utilize light, and continuous shrub belts can form dense vegetation coverage, thereby improving carbon sequestration efficiency. In summary, when configuring road green space vegetation, it is advisable to select upper layer trees with strong adaptability and large canopies in traffic separation green zones, use high-ecological-benefit plants such as Sabina chinensis and Photinia fraseri in avenue greenbelts, and enhance the vertical effect of the landscape in roadside greenbelts, using shade-tolerant plants such as Ligustrum sinense and Ligustrum vicaryi to create continuous shrub belts.

4.1.2. Exploration of Aesthetic Value

In this study, the SBE and landscape characteristic value evaluation methods were used to assess the quality of roadside plant landscapes. These methods complement and combine subjective and objective aspects, which helps designers comprehensively consider public aesthetics and is of practical significance [21].

The SBI value of TS-A is the highest (0.443), and the SBI value of G-R is the lowest (−0.514). Generally, sample points with rich plant species and complete vertical structure, high leaf color saturation and closure, orderly space, and beautiful forest edge lines have higher SBI rankings. Sample points with lower rankings usually have chaotic or empty spaces, low closure, and thin plant leaves with fewer color changes. In this study, professional evaluators are more sensitive to the configuration structure of the landscape and have less difference in the perception of growing status, which may be because the maintenance management of urban road plants is in place, and the difference in the growing status of sample points is not obvious. Evaluators are more likely to notice the diverse changes in the design form of road green space landscapes, which is similar to other research findings [46]. A study on the aesthetics of urban parks shows that the public has the highest preference for urban park images with a tree canopy coverage of more than 45%–50% [47]. This study explores the correlation between SBE and landscape characteristic values among 13 sample points and finds that SBE is significantly positively correlated with the vertical greening effect, canopy density, and floristics. It can be seen that sample points with more complete horizontal and vertical structures are more likely to produce better landscape effects, consistent with related research findings [48]. In summary, on the basis of ensuring road traffic safety, attention should be paid to improving the closure and vertical change effects of sample points, and finally, timely and effective maintenance management is needed to ensure the landscape effect [49].

4.2. Interaction Between SBI and Ecological Benefits

This study evaluates the correlation between ecological benefits and aesthetic quality in urban vegetation construction. The ecological value and SBI do not conform to a linear relationship, but Spearman’s analysis shows a highly significant correlation between the two. The nonlinear multiple binary regression model between SBI and ecological value has been accurately predicted, which can provide scientific guidance for the green space construction of typical small- and medium-sized cities.

We explored plant landscape design strategies guided by the dual theory of ecology and aesthetics by analyzing changes in rankings between the SBI of sample points and the standardized values of ecological values (W_C, T). Overall, there was little variation in the ranking between SBI and ecological values for the same sample points (Figure 12). There is only one sample, TSG-R2, with a difference of more than three positions between SBI and W_C and SBI and T, accounting for 7.69% of the total sample points. TSG-R2 uses a wide variety of colorful leaf plants to create a strong vertical effect landscape, but the coverage rate of trees and shrubs is slightly low, which is manifested as a higher SBI ranking than the ecological value.

Looking at the road green space types, the same life form landscape follows different design guidelines in different types of green spaces, which is similar to other studies [50]. TS-Z and TS-A are both tree and shrub double-layer structures. TS-A has a higher overall ecological value and SBI than TS-Z because TS-Z is located on greenbelt avenue. The transparency of the driving line of sight needs to be considered, and it is not suitable for planting tall trees. However, TS-A is located in the traffic separation green zone and needs trees with good shade, suggesting that avenue greenbelts should use plants such as Photinia fraseri, Ligustrum vicaryi, and Sabina chinensis that are anti-glare and have high ecological benefits for strip planting to enrich the middle and lower layer landscape effects. TG-A is a 1 × 1 m “tree pit” planting method with a small green space area but a high ecological average value. It can be seen that the “tree pit” planting method should make good use of grassy plants or shrubs with dense leaves and evergreen all year round, such as Ophiopogon japonicus, to cover the ground and improve the utilization rate of road green space.

In terms of plant species of the sample points, the richness of species and the configuration structure have a significant impact on landscape aesthetics and ecological benefits. G-R, SG-A2, and TSG-R3 are lawn landscapes with different configuration methods. G-R is a lawn with large sculptures, SG-A2 is a lawn with shrub groups, and TSG-R3 is a large area of blank space with a lawn dotted with tree and shrub groups. The SBI evaluation results are TSG-R3 > SG-A2 > G-R, and the ecological value results are SG-A2 > TSG-R3 > G-R. This may be because grassaceous plants dominate TSG-R3 and G-R, and their carbon sequestration and cooling capabilities are weak, hence the lower ecological value. G-R has a crowded space, resulting in a poor visual experience. In the design of lawn landscapes, attention should be paid to the creation of spatial comfort and landscape layering. In small lawn spaces, the use of sculptures should be reduced, and shaped shrubs should be used to create vertical effects. SG-A1 and SG-A2 are shrub and grass landscapes with obvious differences in plant species richness, both located in roadside green spaces. The SBI and ecological value rankings of the two do not change significantly, which may be because roadside green spaces are close to pedestrians, and sample points with fewer species but reasonable configurations have a better visual experience from this viewing angle. The newly added Sabina chinensis and Ligustrum vicaryi in SG-A1 have weaker carbon sequestration and cooling capabilities than Photinia fraseri, which in turn reduces the average ecological value of the sample. When comparing our research results with previous studies [25], it must be pointed out that landscapes with rich plant species and complete life forms do not necessarily have high ecological and aesthetic values.

A study on visual aesthetic preferences and ecological quality shows that pursuing high-ecological-quality landscapes requires sacrificing certain aesthetic effects [51]. In this study, TG-R and TSG-A contain the coniferous plant Pinus bungeana. TG-R is dominated by Pinus bungeana, with a cold color, while TSG-A is dominated by warm-colored plants Ligustrum vicaryi and Aesculus megaphylla, with Pinus bungeana as a supplement, resulting in a more harmonious overall landscape with higher attractiveness and ecological value ranking. Using high ecological benefits, plant elements with harmonious colors and elegant forms can effectively promote the harmonious unity of aesthetic performance and ecological function of green space landscapes on the road.

In terms of landscape effects, the public prefers the visually intuitive feeling brought by landscapes. TSG-A, TSG-R1, and TSG-R2 are life forms of trees, shrubs, and grasses. The SBI evaluation results are TSG-R2 > TSG-A > TSG-R1, and the ecological value results are TSG-R1 > TSG-A > TSG-R2. TSG-R1 uses a large number of ecological advantage plants to create a space change from open to closed, but it still needs to be more popular than the strong leaf color contrast in TSG-R2 and the obvious undulating forest edge line change in TSG-A. However, existing studies have shown that landscapes with high openness and a strong sense of order are more popular [52,53]. The possible reason is that the strip shrubs in TSG-R2 have a stronger spatial aggregation effect than the spherical group shrubs in TSG-R1, which further increases the landscape attractiveness. Pedestrians’ attention to road green space landscape appreciation is lower than that of parks and other green spaces, so landscapes with stronger visual, intuitive feeling, such as TSG-R2, have higher aesthetic value in road green spaces.

The practical discussion on the harmonious integration of plant ecology and aesthetics into landscape design should not only focus on design methods but also pay attention to the social level. When public aesthetic preferences and ecological quality conflict, they usually manifest as different needs and expectations for landscape design. This conflict is particularly prominent in cities with limited land resources. On the one hand, the public hopes to enjoy the comfort brought by landscapes, and on the other hand, they pursue the ultimate aesthetic experience. Some researchers have found the necessity of ecological education on natural vegetation. For example, the public’s acquisition of ecological benefit information of summer grass can improve their tolerance for the bare landscape in winter, thereby coordinating the feelings of different residential groups on plant ecology and aesthetics [54]. The integration of plant ecology and aesthetics is achieved through a large number of researchers continuously exploring the scale of landscape ecological functions, proposing reasonable strategies, and then combining natural education to strengthen the public’s cognition and understanding of ecological values, thereby enhancing protection awareness and identity [18]. Summarizing and promoting practical experience in different places helps cities learn and draw lessons from each other and promotes the common progress of green space evaluation work.

4.3. Further Improvements of Experiments

Due to time and experimental equipment limitations, this study only measured the ecological capacity of plants to regulate the urban environment directly, represented by carbon sequestration and cooling effects. It did not measure other ecological benefits, such as dust retention, noise reduction, and water conservation, nor did it compare the seasonality of these benefits [55]. Future research could include these aspects for a more comprehensive and accurate assessment of the ecological benefits of plants.

The SBE evaluation process can be affected by weather conditions and participants’ views of the landscape. Future experiments could incorporate panoramic photographs or virtual reality (VR) as visual carriers, using virtual reality to replace actual scenes and enhance the evaluators’ experience. Gao et al. investigated [56] the effects and differences of three landscape perception methods—field surveys, photo induction, and virtual reality—on landscape preferences and found significant differences. This suggests that selecting the best landscape perception approaches for specific seasons and landscapes can provide a scientific basis for future landscape perception and preference assessments.

5. Conclusions

There was a significant positive correlation between the carbon sequestration and cooling benefits of road green space plants and aesthetic quality, and the nonlinear regression model constructed in this study can scientifically evaluate and optimize the planting design in urban areas. The overall effect of carbon sequestration and temperature reduction in urban road green space was as follows: shrubs > trees > grasss. Chinese scholartree, Ligustrum lucidum, Koelreuteriapaniculata, Prunus serrulata, Prunus cerasifera, Ligustrum sinense, Photinia fraseri, Ligustrum vicaryi, Sabina chinensis, and Ophiopogon japonicus are ecologically superior plants and can be used as the primary species of choice for plant design. SBE was highly significantly and positively correlated with vertical greening, canopy density, and floristics, and sample sites with more complete horizontal and vertical structures were prone to producing better landscapes. This study recommends that traffic separation green zones should be populated with tall trees by high light saturation points or possessing leathery foliage to enhance spatial closure, and “tree pool” planting is recommended for ≤1 m wide traffic separation green zones. The avenue greenbelts should make use of plants with both anti-glare and high ecological benefits, such as Photinia fraseri, Ligustrum vicaryi, and Sabina chinensis, to provide a ribbon configuration to enrich the middle and lower landscape levels. The spatial arrangement of roadside greenbelts should be meticulously planned: near the pavement area to simplify the types of plant applications, a small area of lawn with more modeling shrubs and reduce the sculpture; away from the pavement roadside greenbelts through the purple leaf plum and gold leaf chaste tree and other foliage color distinctive plants to strengthen the visual focus. Furthermore, timely and effective maintenance management is required to ensure the landscape effect, encourage public participation in plant cultivation and management, and jointly create a beautiful ecological environment and cultural landscape to enhance human well-being.