Response of Plant Diversity of Urban Remnant Mountains To Surrounding Urban Spatial Morphological Features: A Case Study In Guiyang of Guizhou Province, China


 To explore response of plant diversity of urban remnant mountains (URMs) in the built environment to the surrounding urban spatial morphological features during urban expansion, 9 typical URMs were selected as the research objects, the spots in each sample URM were set by the combination way of the slope direction and slope position, a total of 99 plots for plant diversity survey. Taking the edge line of the sample URMs as the datum, annular buffer zones were set successively outward at step lengths of 100 m, a total of 16 buffer zones with a total width of 1600 m.The spatial morphological characteristics within each buffer zone were analyzed by using spatial syntax, then the relationship between spatial morphological characteristics and plant diversity of URMs were analyzed. The results indicated that: ① There were significant differences in plant diversity among different URMs, and there were also significant differences in plant diversity in different slope positions or different directions of the same URM. ② The spatial morphology around the URMs was different, and the road density (Dn) around the URMs tended to be stable with the increase of spatial scale. The space syntactic indices were positively linearly correlated with the buffer width. ③ On the whole, there was a positive correlation between spatial morphology indices and URMs plant diversity indices. Connectivity (Ci), integration (Ii) and road density (Dn) were more comprehensive and specific, and the correlation increased with the increase of spatial scale. However, choice (C), connectivity (Ci) and mean depth (MDi) were not comprehensive and unstable in response to plant diversity indices. ④There were differences in the response of different slope positions or different directions of the same URM to the spatial morphology. The response intensity of plant diversity in different slope position of URMs to urban spatial morphology was the foot of mountain> mountainside > mountaintop; There was a weak and unstable relationship between road density (Dn) and plant diversity indices in different directions. The results of this study could provide important scientific basis for the conservation and management of urban plant diversity and urban planning and construction.


Introduction
According to the statistics and projections of the United Nations, by 2050, the urban population will increase by 2.5 billion and the urbanization rate will reach 63 percent. The growth of the world's rural population has been slow since 1950 and is expected to peak within a few years (United Nations, 2018).
It can be said that the spatial distribution structure of human economic and social activities has entered a new stage dominated by urban areas. As an important part of urban ecosystem, urban plant diversity provides a series of ecological, economic and social bene ts for cities, such as protection of urban natural ecology and native plants, mitigation of urban heat island effect, puri cation of urban pollution, beauti cation of urban environment, etc. (Bolund & Hunhammar, 1999;Miller, 2005;Sushinsky, 2011). However, with the continuous advancement of urbanization, the invasion of alien species, the extinction of native species and the homogeneity of urban species composition have led to the decline of plant diversity and other problems (Miller, 2006;McKinney, 2002). In the context of rapid urbanization, how to protect and maintain urban plant diversity and make it play a powerful role in the ecosystem has become a hot issue in current research.
Urbanization process is the process of land use type transformation, and land use transformation is the main driving factor affecting urban biodiversity (Zhu et al, 2019). Previous studies have systematically elucidated the effects of urbanization on the distribution patterns of biodiversity, exotic/indigenous plants, and plant diversity homogenization at different scales, discussed that the effects of urbanization on plant diversity are mainly land use change, climate change, urban landscape pattern change, urban social economy activity, such as urban environment complex changes caused by the factors of urbanization (Mao et al., 2013). For example, Wirth et al. (2020) monitored the ora of Hungary for 70 years and found that woody plants or alien annual and biennial plants increased with urbanization. Vakhlamova et al. (2014) used the urban-rural gradient method to study changes in plant species composition and richness in Kazakhstan, Siberia, and found that plant diversity increased with distance from the center, at the same time, also affected by land use type and building coverage within the 500 m radius. In addition, study has shown that the combined effect of urbanization and changes in agricultural land use has changed the composition of plant species. Some studies have taken the percentage of total impervious surface area (PTIA) as the key predictor of urban plant diversity in Wuhan, China. Urban plant diversity decreases with the increase of the percentage of total impervious surface area (PTIA). When PTIA reaches 40% or above, plant diversity declines sharply (Yan et al, 2019). The effect of landscape pattern on plant diversity has been veri ed in many studies. Peng et al. (2019) discussed the impact of landscape metrics on native plant diversity in Shunyi District, Beijing, and determined that landscape units within a radius of 600-700m in Shunyi District, Beijing, were the most optimal spatial scale range for the conservation of native plant diversity. Besides, increasing temperatures in cities encourage the migration of thermophilic plants to urban areas, and adaptations to the environment may make these plants more aggressive, increasing the probability of alien plant invasion (Parmesan & Hanley, 2015). And a study by Čeplová et al. (2017) on 45 Central European settlements showed that species composition was signi cantly more in uenced by local habitat conditions than by urban size, highlighting the important role of habitat conditions on biodiversity of native and alien plant communities. In addition, social, economic and cultural factors are also closely related to urban plant diversity (Monteiro et al., 2013). However, few studies have considered the potential impact of urban spatial morphology on urban plant diversity.
Urban spatial morphology is the spatial arrangement of various urban elements in urban region (Feng & Zhou, 2003), it determines the distribution of people's social and economic activities in the city, and then affects other urban elements, such as transportation, land use function and form of architecture, etc., and ultimately affects the formation and evolution of urban morphology (Han et al, 2018). There is a correlation between urban spatial morphology and environmental change (Barau et al, 2015;Yang et al., 2019). At present,some scholars have explored the relationship between urban spatial morphology and urban vegetation, for example, Bigsby et al. (2014) studied the relationship between tree cover patterns with urban morphology (housing density, parcel size), socioeconomic factors (education, income, lifestyle characteristics), and historical heritage in Baltimore, Maryland, and Raleigh, North Carolina, found that urban morphology is more effective than socioeconomic factors in predicting tree cover patterns at parcel and neighborhood scales, and concluded that urban morphology as the main driving factors of urban tree cover patterns, may lead to the homogenization of tree canopy. Road system is the carrier of urban spatial morphology, and changes in road density and landscape pattern can explain variables related to land use, land cover and environmental factors (Hawbaker et al, 2005). Cai et al. (2013) found that an increase in road density is often accompanied by an increase in construction land area and a decrease in forest coverage, leading to a signi cant decline in ecosystem health. Some studies also proved that the non-native perennial grass cover is signi cantly positively correlated with the road density across the urban landscape (Zeeman et al., 2018). Christen et al. (2009) investigated nonnative plant species along roads in deciduous forest sites in southeastern Ohio, USA, and discovered that roads are both habitats and a conduit for population expansion, its rate of spread depends on the life history of the individual species, these results demonstrated that the hierarchical process of regional invasion, with different dispersal mechanism in different spatial scales. therefore, on the basis of the in uence of urban spatial morphological structure on urban plant diversity, it is of great signi cance to explore whether urban spatial morphology has scale effect on plant diversity, which is of great signi cance to urban spatial planning and plant diversity conservation.
The city is a complex of spatial arrangement and combination of various elements, and species diversity is affected by multiple factors. Island biogeography is no longer applicable to urban landscapes isolated by natural/semi-natural ecosystems (Niemela, 1999). In mountainous areas of central Guizhou province, China, a large number of natural or near-natural Karst mountains have constantly been embedding into the arti cial urban environment in the process of urban expansion, forming a special urban spatial form of "city among the mountains, mountains in the city", and the URMs embedded in urban arti cial built environment are the main carrier of urban native biodiversity. We asked: 1) Whether urban spatial morphology has an impact on the plant diversity of URMs? 2) If the impact exists, what are the factors and the range of the in uence? In this study Guiyang, a typical mountainous city in central Guizhou, was taken as the research area. 9 URMs in the built-up area were selected as the research object, and the plant diversity in the URMs in Guiyang, the urban spatial morphological structure around the URMs, and the response relationship between them were studied to explore the relationship between urban spatial morphological structure and plant diversity and their spatial scale dependence, so as to provide a scienti c basis for urban spatial planning, biodiversity conservation and maintenance, and eco-city construction.

Study area
Guiyang (26°11 '--26°55N, 106°07' --107° 17'E) is located in central Guizhou Province, in the middle of the Yunnan-Guizhou Plateau and in the watershed zone between the Yangtze River and the Pearl River. The landform belongs to the hilly basin area and is mainly composed of karst mountains and hills, the whole terrain is high in the southwest and low in the northeast, with an altitude of about 1100 m, it belongs to subtropical humid mild climate, the annual mean temperature is 15.3 ℃, the annual mean total precipitation is 1129.5 mm. By the end of 2018, it has jurisdiction over 6 districts, 3 counties and 1 county-level city, with a permanent resident population of 4.8819 million and an urban population of 3.6824 million, with an urbanization rate of 75.43%. The built-up area of the central urban area is 368.68 km 2, and there are 527 Karst remnant mountains in the urban area, with a total area of 44.94 km 2 , and 416 small and medium-sized URMs smaller than 10 hm 2 . This study takes the central urban area of Guiyang as the study area ( Fig. 1).

Subjects
In order to explore the in uence of urban spatial morphological structure on plant diversity of URMs, on  Table 1. 12 plots were set for each URM according to the method of direction + slope positions (i.e., the sample plot was set respectively at the mountaintop mountainside and mountain foot of south,east, north and west four directions), each plot was 30 m × 30 m. 5 tree quadrates of 10 m × 10 m were set in each sample plot and ve shrub quadrates of 3 m × 3 m were set in each tree quadrate according to the "5 points method", randomly set 1 m×1 m herbs quadrate in shrub sample. Some sample plots that could not be sampled were removed because the mountain was mined and the rock was exposed seriously and the degree of human disturbance was serious. A total of 99 sample plots were effectively investigated in the whole region. The species name, number of each species, DBH, height and crown width of the trees in each tree quadrate, the species name, number of each species and average coverage of shrubs in each shrub quadrate and the species name, plant number and average coverage of herbs in each herbs quadrate were recorded. At the same time, the geographic coordinates, elevation, slope direction and other information of each quadrate were measured by hand-held GPS.
The Shannon-Wiener index (H'), the Simpson index (D), the Pielou index (Jh) and the Margalef index (R) were used to describe plant species diversity: Where N is the total number of individuals; S is the total number of species; And In is the natural log base e; P i = n i /N, where n i represents the number of individuals of the ith species.

Measurement of urban spatial morphological structures
Taking the high-resolution Pleiades satellite image (0.5 m spatial resolution) of Guiyang in 2018 as the image data source, through image enhancement, geometric correction, map projection and other preprocessing, visual interpretation of the preprocessed remote sensing image was performed, and the spatial attribute database of the study area was established based on ArcGIS 10.2 software. Taking the edge line of the sample URMs as the baseline, annular buffer zones were successively set outward at step lengths of 100 m. A total of 16 buffer zones were set, with a total width of 1600 m.The road information in each gradient buffer zone around 9 sample URMs were extracted and processed in relevant software (Auto CAD, UCL Depthmap) to obtain the road axis models around 9 sample URMs in Guiyang, which can be used for subsequent measurement of relevant indices of spatial morphology..
Space syntax and road density were used to characterize the urban spatial morphological structure. Space syntax characterizes the relationship between space and its organization and the interaction between human society through a quantitative description of the spatial morphological structure of human settlements, including architecture, settlements, cities and even landscapes (Hillier, 1993). Road density is the main index to re ect the activity degree of regional tra c, economy and commerce, and it is also an important index to explain the urban spatial morphological features. These space formed by the development of human society will inevitably affect the plant diversity in the space. In this study, relevant Page 7/24 spatial morphological structure indices in space syntax and road density index were selected to explore the in uence of the spatial morphological structure of buffer zones on the plant diversity of URMs. The selected indices are as follows (Table 2):   (p < 0.05) lower than the east and west orientation, Margalef and Pielou indices were only signi cantly (p < 0.05) lower than the east direction. The Margalef index of NM8 in the north was the highest and signi cantly higher than that in the south and east, but there was no signi cant difference in the plant diversity indices in different directions. The Shannon-Wiener, Simpson and Pielou indices in the northern region of NM9 were the highest and signi cantly (p < 0.05) higher than the east and south orientation, and the Margalef index was only signi cantly (p < 0.05) higher than the east direction.

Urban spatial morphological features around the sample URMs
Regression analysis was used to detect the relationship between the surrounding spatial morphology indices and buffer gradient of 9 URMs. Figure 5 shows that there were differences in the spatial morphology around the URMs, and the road density (D n ) tends to be the same with the increase of spatial scale. The space syntax index was generally in a positive linear correlation with the buffer gradient, but the choice (C) uctuates with the increase of spatial scale. and plant diversity at different slope positions all started from 400 m, and the correlation increased with the enlargement of spatial scale, the response relationship between different slope positions was not completely consistent (Fig. 7).

Plant diversity in different directions and urban spatial morphology indices in corresponding directions
There was no obvious relationship between plant diversity and road density (D n ) at different directions in the 9 URMs (Fig. 8)  URMs did not show obvious variation patterns, which may be due to the fact that human activities in mountainous cities tend to choose slope sections with better site conditions to reclaim for cultivated land or orchard, and anthropogenic disturbance affects spatial differentiation of mountain vegetation (Ma et al., 2002). It has been con rmed that the species composition, distribution and diversity of plant communities are not only affected by natural environmental factors such as light, soil nutrients, moisture and terrain, but also affected by surrounding human activities and environmental changes (Xu et al., 2014). Guiyang is a city developed under the special Karst landform. The urban spatial morphology with the city and mountain inlaid each other makes the contradiction between people and mountains prominent and human interference strong, so human factors have become the key factors affecting the diversity of plants in urban remnant mountains (Vollstädt et al., 2017). Besides, in the complex urban ecosystem, human socio-economic activities and decision-making behaviors also greatly affect urban biodiversity (Pandey, 2021).

Effects of urban spatial morphology on URMs plant diversity
It was showed that there was a signi cant correlation between plant diversity of URMs and urban spatial morphology, which con rmed that there is some mapping relationship between urban spatial form, social function and ecological environment quality ( integration (I i ) and road density (D n ) were more comprehensive. Therefore, the plant diversity was higher in the urban spatial agglomeration, large ow of people, strong spatial permeability and developed tra c areas. The reason for this may be that the road network, as a transmission corridor, played a key role in the spread of alien plants in the urban matrix (Carlton & Ruiz, 2005; Von der Lippe & Kowarik, 2008; Zeeman et al., 2018). Also, with the increase of road density (D n ), human entry and interference effects tend to increase (Forman, 2014;Forman & Alexander, 1988). The response of plant diversity to choice (C) and the mean depth (MD i ) was not comprehensive, possibly because the choice (C) expressed the probability of a certain spatial node being selected, which is usually used to measure the tra c potential of commercial roads (Ma et al., 2019).In the small scale buffer zone near the URMs, the tra c attraction brought by the increasing choice (C) would promote the material exchange and increase the plant diversity, but on a larger scale, the potential service range of road choice (C) goes beyond the threshold of the attraction of a mountain to surrounding tra c, which attracts the tra c ow around the mountain to other spaces, so that the in uence on plant diversity decreases with the enlargement of the space.
On the whole, plant diversity was positively correlated with urban spatial morphology. Locally, there were differences in response to urban spatial morphology at different slope positions and directions of the same mountain. The response intensity of plant diversity to urban spatial morphology on different slope positions of URMs was the foot of mountain> mountainside > mountaintop, indicating that the change of human activity intensity from low altitude to the high altitude gradually weak (Sharma et al., 2009), it is con rmed that there was certain correlation between the species richness and management strength and structure characteristics of the road (such as the edge of the road disturbance characteristics) (Pourrezaei et al., 2021).while the response relationship in different directions was not stable, indicating that the same problem has different effects on the ecological factors at different scales (Arteaga et al, 2009 within the range of 600-800m, and it is believed that the radius of 600-700m is the optimal range for the protection of plant diversity in Beijing (Peng et al., 2019). These ndings emphasize that plants and lowmobility species are signi cantly responsive to urbanization at the medium spatial scale, while highmobility species are signi cantly responsive to urbanization at the larger scale when discussing biodiversity response to urbanization (Concepción et al., 2015).

Conclusion
There were signi cant differences in plant diversity among different URMs, and there were also signi cant differences in plant diversity at different slope positions or different directions of the same URM. Spatial morphology around the URMs was different, and the road density (D n ) around the URMs tends to be stable with the increase of spatial scale. Space syntactic indices were positively linearly correlated with the buffer gradient. The plant diversity and the difference of spatial morphology were the premise to discuss the response relationship. On the whole, there was a positive correlation between spatial morphology and URMs plant diversity. Connectivity (C i ), integration (I i ) and road density (D n ) were more comprehensive and speci c, and the correlation increased with the increase of spatial scale.
However, choice (C), connectivity (C i ) and mean depth (MD i ) were not comprehensive and unstable in response to plant diversity. There are differences in the response of different slope positions or different directions of the same URM to the spatial morphology. The response intensity of plant diversity and urban spatial morphology on different slope positions of URMs was the foot of mountain > mountainside > mountaintop; There was a weak and unstable relationship between road density (D n ) and plant diversity in different directions.
In the future, the planning of urban spatial morphological structure and the planning of plant diversity should be coordinated, the urban land use should be rationally planned, the urban spatial morphology should be optimized, and the urban planning and landscape design should adopt more indigenous plants, re ect the regional characteristics, and strengthen the protection of urban plant diversity. Also, it is necessary to consider the scale-dependence effect, select the appropriate spatial morphology index and consider the in uence of various spatial scales, which are different in different parts of the world.    Plant diversity indices at different slope positions.

Figure 5
Urban spatial forms under different spatial scales.

Figure 6
Space syntax and plant diversity correlation coe cient. Correlation coe cient between road density and plant diversity at different heights.

Figure 8
Correlation coe cient between road density (Dn) at different directions and plant diversity at different directions.