The Effects of Urban Landscape Pattern Evolution on Habitat Quality Based on InVEST Model

A Case of Harbin City

Abstract

With the acceleration of urbanization, urban landscape patterns have been rapidly adjusted and reorganized, which has resulted in drastic changes in the quality of urban habitats. To achieve the objective of a new type of eco-conservative urbanization, it is urgent that we strengthen the research on the relationship between the habitat quality and landscape patterns. Based on the InVEST model and FRAGSTATS landscape pattern index, in this study, we evaluated the habitat quality evolution and landscape pattern change trend of Harbin based on remote sensing images from 2000, 2010, and 2020, and we explored the correlation between them using Pearson’s correlation coefficient. The results showed that: (1) From 2000 to 2020, the habitat quality in Harbin decreased from 0.68 to 0.65, with an overall downward trend. The low-quality habitat increased by 1.25%, while the high-quality habitat decreased by 6.43%. The spatial heterogeneity was substantial, and the distribution pattern was high in the middle, middle in the east, and low in the west; (2) The landscape pattern index of Harbin obviously changed, and the overall landscape pattern had a trend of further fragmentation and gradual irregularity, decreases in the diversity and evenness, and spatial heterogeneity; (3) For most of the land types, there were substantial correlations between the habitat quality and landscape pattern indices. The PD, COHESION, and LSI had the most substantial correlations with the habitat quality. The research results provide a theoretical basis for the ecological protection and sustainable development in Harbin.

Introduction

In recent years, China's urbanization process has been accelerating, with the urbanization rate rising from 17.92% in 1978 to 64.72% in 2021, and it is expected to reach 70% by 2030 (United Nations in China, n.d.). The rapid growth of the urban population has resulted in urban expansion and an increased demand for land resources (Cheng, Chen, et al., 2019). Moreover, large areas of ecological land, such as forest, grassland, and water, have been transformed into built-up land (Chopra, Singh, et al., 2022), which has resulted in a series of environmental problems, such as the urban heat island effect, environmental pollution, soil erosion, biodiversity reduction, and ecosystem imbalance (Paul and Bardhan, 2022). Biodiversity plays a vital role in the global economy (Rands, Adams, et al., 2010) and human wellbeing (Duffy, 2009). However, the land-use change that is caused by urbanization leads to the fragmentation of habitats that are suitable for living organisms, affecting the energy flow and material circulation between habitat patches, which is one of the most important factors in the decline of biodiversity (Sala, Stuart Chapin, et al., 2000). Therefore, to meet the needs of urban development and construction, biodiversity conservation has become an increasingly urgent issue.

Researchers often use the habitat quality to characterize the regional biodiversity level (Kareiva, Tallis, et al., 2011). The habitat quality refers to the ability of the environment to provide appropriate resources for the sustainability of individuals and populations, including survival and reproduction (Hall, Krausman, et al., 1997). The habitat quality is directly related to the abilities of the regional ecosystem services (Y. Lin, W. Lin, et al., 2017). Changes in landscape patterns can lead to corresponding changes in the ecosystem composition and biodiversity (Mairota, Cafarelli, et al., 2015; Yan, Wang, et al., 2017). The landscape pattern refers to the shape, proportion, and spatial arrangement of the ecosystem or land-use type (Hu, Wang, et al., 2008), and it is a key driving factor in habitat quality evolution (Vögeli, Serrano, et al., 2010). Research on the relationship between the habitat quality and landscape pattern change is an important basis for regional ecosystem service management and sustainable landscape planning. In recent years, scholars have conducted empirical studies on the habitat quality and landscape pattern changes in river basins (Berta Aneseyee, Noszczyk, et al., 2020), mountain areas (Yang, 2021), coastal zones (Cozzoli, Smolders, et al., 2017), cities (Li, Duo, et al., 2021), and urban agglomerations (Wang, Tang, et al., 2020). For example, the expansion of agricultural land has led to its invasion of large areas of native vegetation, which has resulted in the loss, fragmentation, and reduced quality of habitats (Dai, Li, et al., 2019). According to Berta Aneseyee, Noszczyk, et al. (2020)'s study, the increase in cropland and built-up land has resulted in soil erosion, decreased vegetation coverage, and environmental pollution, which have reduced the forest, shrub, and pasture habitat types and affected the local habitat quality. (Liu, Zhang, et al., 2004) also report that land-use change has resulted in the rapid decline of waterfowl and plant species, the loss and fragmentation of natural wetlands and the degradation of wetland ecosystems. Although researchers have conducted many studies on landscape patterns and the quality of habitats, we lack studies on the impact of the different urban landscape patterns on the habitat quality and temporal evolution characteristics. Therefore, we studied the temporal and spatial evolution of the habitat quality, and we explored the impact of the landscape pattern change on its evolution to provide a basis for regional ecological environment restoration and biodiversity conservation.

At present, we can divide the methods that scholars commonly use to evaluate the habitat quality at the landscape scale into three categories: (1) The selection of field surveys and indices to evaluate the habitat quality, such as the Shannon diversity index and species richness. For example, Monteiro, Fava, et al. (2013) studied the grassland habitat quality in the southern Alps using statistical analyses of the Shannon diversity index and species richness of the permanent meadows in the region. Evans, Rodrigues, et al. (2006) studied the relationship between the proportion of nature reserves and species richness by calculating the total species richness and threatened species richness of birds in protected areas in South Africa. However, this method is usually only suitable for small-scale areas or nature reserves; (2) The construction of an index system. For example, Ding, Shan, et al. (2015) selected 10 indicators (e.g., riparian zone width, land use pattern outside the riparian zone, riparian vegetation coverage, etc.) to evaluate the quality of the river channel habitat in China's Haihe River basin. Yu, Liu, et al. (2022) used 4 categories of 20 indicators, including the water environment status, river morphology, riparian zone, and human disturbance, to evaluate and classify the channel habitat quality of the Chishui River basin of China. However, such methods require a large amount of data, and the impacts of the surrounding landscape structure and potential threats are not considered; thus, they are only appropriate for descriptive assessments of the habitat suitability; (3) Model-based assessment methods, such as Habitat Suitability Index (HSI) (Jung, Shimizu, et al., 2016), Social Values for Ecosystem Services (SoIVES) (Sherrouse, Semmens, et al., 2014), Integrated Valuation of Ecosystem Services and Trade-offs (InVEST) (Sharp, Wood, et al., 2018). We can use the habitat quality module of the InVEST model to establish the relationship between the different land-use changes and biodiversity threat sources, and to evaluate the distribution and degradation of the habitat quality according to the sensitivity of different habitats to threat sources (Terrado, Sabater, et al., 2016). The InVEST model uses less data, is easy to obtain, and it does not require species distribution data, and we can use it to spatially visualize the calculation results (Polasky, Nelson, et al., 2011). In habitat quality module, not only the threat source and its impact distance are considered, but also the geographical distance and sensitivity between habitats (Sharp, Wood, et al., 2018). In general, the InVEST model is more suitable for the habitat quality assessment process at the urban scale (or larger) than other models. Researchers have successfully applied it to habitat quality studies at the regional, watershed, and city scales, and it has been widely recognized by scholars in related fields. For example, Dai, Li, et al. (2019) evaluated the spatial-temporal characteristics of the land-use and habitat quality in Northeast China from 1990 to 2010 using the InVEST model to assess the impact of the land-use change on the habitat quality. Berta Aneseyee, Noszczyk, et al. (2020) used the InVEST model to map and identify the habitat quality changes in the Winike watershed in southern Ethiopia from 1988 to 2018, and studing the correlation between the influencing factors within the basin and the habitat quality. Sharma, Nehren, et al., (2018) used the Land Change Modeler (LCM) and InVEST model to evaluate the habitat quality and spatial biodiversity distribution under different land-use scenarios in Pulang Pisau, India.

As one of the important old industrial bases in northeast China, Harbin is the capital of Heilongjiang province and an important grain production base. With the proposal of the Belt and Road Initiative and the continuous promotion of northeast Asia's opening-up and cooperation, Harbin's urbanization process is accelerating, and land development efforts are increasing. Although the economy has been rapidly improving, the disorderly spread of construction land is an encroachment on the natural resources around the city, which has resulted in the fragmentation of the ecological land, landscape pattern changes, and an increasingly serious threat to the urban habitat. The 14th Five-Year Plan of National Economic and Social Development of Harbin and Vision Goal of 2035 require that we build a solid barrier for ecological security, hold the bottom line for the use of ecological space, improve the quality of the ecological, urban, and rural living environments, and build beautiful, ecological, and liveable cities. According to the development orientation and realistic demand of Harbin, the biodiversity and ecosystem functions must adapt to the new landscape configuration and regional environment. Harbin has rarely been the subject of scholarly work, which makes it a good case study to explore and quantify the landscape pattern evolution impact on the habitat quality.

Many scholars prefer to select the habitat quality and landscape pattern index for one year. However, we believe that in the process of urbanization, the landscape pattern impact of the same land on the habitat quality is not invariable, and the evolution direction and intensity of the relationship between the two need further demonstration. Therefore, in this study, our main objectives were to evaluate the spatial and temporal variations in the habitat quality and the differentiation of the landscape patterns in Harbin, and to investigate the correlation between the two in this area. The aims of this study were as follows: (1) to quantitatively evaluate the spatiotemporal changes and evolution of the habitat quality in Harbin in 2000, 2010, and 2020 using the habitat quality module of the InVEST model; (2) to use the standard method and moving window method of the FRAGSTATS platform to calculate the various landscape pattern indices for the three periods to reflect the landscape pattern differentiation in the study area; (3) to analyze the correlation between the habitat quality and the landscape patterns using Pearson's correlation coefficient. From the main results of this study, we were able to clarify the positive and negative effects of the landscape pattern evolution on the habitat quality, and especially the temporal changes on the correlation between the landscape pattern index and the habitat quality of the same land type over time, which enriches our understanding of the habitat quality evolution characteristics and the optimization of landscape pattern. This study also serves as a reference for local land space planning and the implementation of ecological protection strategies.

Methods and Materials

Study area

Harbin is located on the northeast plain of China in the south of Heilongjiang province (44°06'-46°67' N, 125°68'-130°23' E). Harbin consists of 9 municipal districts (Daoli; Nangang; Xiangfang; Daowai; Hulan; Songbei; Acheng; Pingfang; Shuangcheng), 2 county-level cities (Shangzhi, Wuchang), and 7 counties (Tonghe; Mulan; Bayan; Yanshou; Bin; Yilan; Fangzheng), with a total area of 53,100 km2 (Figure 1). The western part of Harbin is flat, while the eastern part is undulating, with peaks and hills. The main rivers are the Songhua, Hulan, Ash, and Lalin Rivers. Harbin has the continental monsoon climate of the mid-temperate zone and four distinct seasons, with long and cold winters and short and cool summers. The precipitation is mainly concentrated from June to September, with average annual precipitation of 569.1 mm. According to statistical data, from 2000 to 2020, the total population of Harbin increased from 9.4 to 10 million people, and the GDP increased from CNY 100.2 to 518.38 billion. Due to the continuous acceleration of urbanization in Harbin, the rapid increase in built-up land has changed the original land structure and landscape pattern, which has affected the quality of the habitat and caused substantial pressure on the ecosystem. Therefore, assessing the spatial–temporal evolution and landscape pattern response of the habitat quality in Harbin for biodiversity and ecosystem conservation in the city is urgent.

Figure 1. The study area.

Data sources

For the research data, we selected the land-use and road data for Harbin from 2000, 2010, and 2020. We obtained the land-use data from the Data Center for Resources and Environmental Sciences, the Chinese Academy of Sciences (http://www.resdc.cn, accessed on 13 July 2021). The data was obtained from 30 m multispectral Landsat-TM/ETM and Landsat-8 images based on a high- resolution remote sensing ground survey and observation technology, combined with the human-computer interaction interpretation method. The accuracy is over 85%, and the interpretation accuracy is reliable. According to the dominant land-use types in the study area, we can classify the land uses into 6 first-level land-use types and 16 second-level land-use types (Table 1). We took the road data from the China National Basic Geographic Information Center (http://nfgis.nsdi.gov.cn, accessed on 13 July 2021), including railway, expressway, national road, provincial road, and county road data. We obtained the boundary data from the Bigemap GIS Office map downloader.

Table 1. Land use classification

Class1	Class2	Class1	Class2
Cropland	Paddy field	Water	River
	Dryland		Lake
Forest	Woodland		Reservoir
	Shrubwood		Marshland
	Sparse forest	Built-up land	Urban land use
	Other woodland		Rural residential areas
Grassland	High-coverage grassland		Other built-up land
	Moderate-coverage grassland	Bare land	——
	Low-coverage grassland

Methods

Habitat quality assessment based on InVEST model

The InVEST model is a set of modeling systems that was developed at Stanford University in the United States. Researchers use it to quantitatively evaluate the function of ecosystem services by simulating the combined impacts of the land-use changes and human activities on ecosystems (Goldstein, Caldarone, et al., 2012; Nelson, Sander, et al., 2010; Sharp, Wood, et al., 2018). Based on the habitat quality module in InVEST 3.8.5, in this study, we generated habitat degradation and habitat quality maps based on land-use data and biodiversity threat information to obtain the regional biodiversity status. The model is based on the following two assumptions: (1) the legal protection of the land is effective, and all threats to the landscape are cumulative; (2) areas with higher habitat quality scores have higher biodiversity, while areas with lower habitat quality scores have lower biodiversity (Sharp, Wood, et al., 2018). In this model, we determined the habitat quality assessment according to four factors: (1) the relative impact of the threat source; (2) the distance between the habitat and threat source; (3) the legal/social/physical protection of the land; (4) the relative sensitivity of each habitat type to each threat (Liu, Liao, et al., 2022; Wang, Tang, et al., 2020). The final map generated by the model contains the degradation degrees and habitat quality scores of all the land-use types in the study area. The score is a continuous variable that range from 0 to 1: the closer the value is to 1, the better the habitat quality, which is conducive to the maintenance of biodiversity (Sharp, Wood, et al., 2018). The habitat quality module primarily includes evaluations of the habitat degradation degree and habitat quality. The calculation formula for the habitat degradation degree is as follows:

$i_{\mathrm{\text{rxy}}}=1-\left(\frac{d_{\mathrm{\text{xy}}}}{d_{\mathrm{\text{rmax}}}}\right)$ if linear (1)

$i_{\mathrm{\text{rxy}}}=exp(-{\left(\frac{2.99}{d_{\mathrm{\text{rmax}}}}\right)d}_{\mathrm{\text{xy}}})$ if exponential (2)

$D_{\mathrm{\text{xj}}}={\sum }_{r=1}^R{\sum }_{r=1}^{Y_r}\left(\frac{{\omega }_r}{{\sum }_{r=1}^R{\omega }_r}\right)r_y{i}_{\mathrm{\text{rxy}}}{\beta }_x{S}_{\mathrm{\text{jr}}}$ (3)

where $i_{\mathrm{\text{rxy}}}$ is the influence of the threat source ( $r$ ) in grid $y$ on grid $x$ ; $d_{\mathrm{\text{xy}}}$ is the linear distance between grid $x$ and $y$ ; $d_{\mathrm{\text{rmax}}}$ is the maximum influence range of the threat source ( $r$ ); $D_{\mathrm{\text{xj}}}$ is the habitat degradation degree of grid x in land-use type $j$ ; R is the number of threat sources; $Y_r$ is the total number of threat sources; ${\omega }_r$ is the weight; $r_y$ is the number of threat sources in the grid cells; ${\beta }_x$ is the anti-interference level of grid $x$ ; $S_{\mathrm{\text{jr}}}$ is the sensitivity of land-use type $j$ to the threat sources. The value of the habitat degradation was between 0 and 1: the higher the value, the more obvious the degradation degree.

The habitat quality calculation formula is as follows:

$Q_{\mathrm{\text{xj}}}=H_j[1-\left(\frac{D_{\mathrm{\text{xj}}}^z}{D_{\mathrm{\text{xj}}}^z-k^z}\right)]$ (4)

where $Q_{\mathrm{\text{xj}}}$ is the habitat quality of grid $x$ in land-use type $j$ ; $H_j$ is the habitat suitability of grid $x$ for land-use type $j$ ; $z$ is the scale constant; $k$ is the half-full and constant.

The main reasons that we selected this model for the study are as follows: (1) We only collected a few data sources in this study. Compared with other models, this model has a strong analytical ability and is useful in cases of poor species observation data (He, Huang, et al. 2017); (2) We can easily apply the model to specific environments or global data (Gao, Ma, et al., 2017); (3) We can use the model to analyze the spatial habitat quality trends and connectivity of each land-use type and quantify the sensitivity to the threats (Berta Aneseyee, Noszczyk, et al., 2020); (4) The model is reliable. Terrado, Sabater, et al. (2016) found that the calculated results of the model were substantially correlated with the biodiversity observation results.

According to the field survey of Harbin, the cropland, built-up land, and traffic roads, along with the large population in this area, substantially impact the landscape pattern change and pose a certain degree of threat to the natural ecosystem. Forests are the most suitable land-use type for species survival (Terrado, Sabater, et al., 2016), while bare land has low vegetation coverage and poor environmental conditions. Therefore, we established cropland, urban land, rural residential areas, other built-up land, and traffic roads severely disturbed by human activities as the threat sources in this study. The sensitivities of the different land-use types to the various threat sources are different; we primarily determined the sensitivity according to the basic principles of biodiversity conservation in landscape ecology (Lindenmayer, Hobbs, et al., 2007). On this basis, we referred to the InVEST model manual (Sharp, Wood, et al., 2018) and related studies (Bai, Xiu, et al., 2019; Gao, Ma, et al., 2017; He, Huang, et al., 2017) to assign the maximum impact distance, weight, habitat suitability of each land-use type, and sensitivity to various threat sources in the habitat (Table 2, Table 3).

Table 2. The sensitivity of different landscapes to threatening factors

Stress factors	Maximum impact distance /km	Weight	Decay type
Cropland	8	0.7	exponential
Urban land use	10	1.0	exponential
Rural residential areas	5	0.6	exponential
Other built-up land	8	0.8	exponential
Road	3	0.5	linear

Table 3. The sensitivity of different landscapes to threatening factors

Name	Habitat suitability	Cropland	Urban land use	Rural residential areas	Other built-up land	Road
Paddy field	0.6	0.3	0.5	0.35	0.4	0.6
Dryland	0.4	0.3	0.5	0.35	0.4	0.2
Woodland	1	0.7	0.85	0.8	0.7	0.65
Shrubwood	0.9	0.6	0.7	0.6	0.6	0.4
Sparse forest	0.8	0.85	0.85	0.7	0.55	0.6
Other woodland	1	0.9	0.85	0.85	0.6	0.7
High-coverage grassland	0.8	0.5	0.6	0.45	0.5	0.3
Moderate-coverage grassland	0.7	0.5	0.65	0.5	0.55	0.35
Low-coverage grassland	0.6	0.5	0.7	0.55	0.55	0.3
River	0.9	0.65	0.9	0.7	0.8	0.45
Lake	1	0.7	0.9	0.75	0.8	0.5
Reservoir	1	0.7	0.9	0.75	0.8	0.5
Marshland	0.65	0.75	0.95	0.8	0.7	0.55
Urban land use	0	0	0	0	0	0
Rural residential areas	0	0	0	0	0	0
Other built-up land	0	0	0	0	0	0
Bare land	0.1	0.1	0.2	0.2	0.2	0.15

Landscape pattern index method

Scholars often use the landscape pattern index to quantitatively describe landscape pattern changes (Liu, Chen, et al., 2020; Long, Nelson, et al., 2010; Nowosad and Stepinski, 2018). The landscape pattern index is the digital information on the landscape pattern, and it effectively reflects the structural characteristics and spatial allocation (Uuemaa, Antrop, et al., 2009). In this study, we used the FRAGSTATS 4.2 platform to calculate the landscape pattern index at the overall level of the study area using the standard method. From the two dimensions of the class level and landscape level, we selected eight representative indices that reflect the degree of the landscape fragmentation, aggregation, connectivity, and diversity. Among them, the selected indices at the type level were as follows: the patch density (PD), average patch area (AREA_MN), maximum patch index (LPI), cohesion index (COHESION), and landscape shape index (LSI). At the landscape level, the selected indices were as follows: the sprawl index (CONTAG), Shannon diversity index (SHDI), and Shannon evenness index (SHEI) (Shuangao, Padmanaban, et al., 2021; Yang, 2021). Then, we used the moving window method to calculate the local-level landscape pattern index (Hagen-Zanker, 2016), and we finally obtained its spatial distribution. If the radius of the moving window is too large or too small, then it cannot accurately reflect the landscape characteristics of the area. We attempted to use 500 m, 1000 m, and 2000 m as the window sizes (Kassawmar, Murty, et al., 2019). After the experiment, we determined that the most stable window radius was 1000 m, and that the landscape index grid map obtained under this window radius could accurately reflect the spatial landscape pattern changes in the study area.

Statistical analyses

In this study, we used the SPSS 23 platform to conduct the Pearson’s correlation analysis at a significance level of P < 0.01. In the study area, we established 464 grids with 10*10 km as the unit. We extracted the habitat quality and landscape pattern indices through the grids to evaluate the correlation between the landscape pattern change characteristics and habitat quality changes (Berta Aneseyee, Noszczyk, et al., 2020). We can use Pearson’s correlation analysis to measure the linear relationship between two continuous variables: the larger the absolute value of the correlation coefficient, the stronger the correlation. The calculation formula is as follows (Biswas and Si, 2011):

$r=\frac{N{\sum }_{}^{}{x}_i{y}_i-{\sum }_{}^{}{x}_i{\sum }_{}^{}{y}_i}{\sqrt{N{\sum }_{}^{}{x}_i^2-{\left({\sum }_{}^{}{x}_i\right)}^2}\sqrt{N{\sum }_{}^{}{y}_i^2-{\left({\sum }_{}^{}{y}_i\right)}^2}}$ (5)

In the formula, x represents the landscape pattern index, y represents the habitat quality, and r represents the correlation degree between -1 and 1. The greater the absolute value of r, the greater the correlation. If the r value is greater than 0, then the landscape pattern index is positively correlated with the habitat quality; if the r value is less than 0, then the relationship between the two is negatively correlated.

Results

Spatial and temporal patterns of habitat quality

By using the habitat quality module of the InVEST model, we obtained the spatial distribution of the habitat quality in Harbin for 2000, 2010, and 2020. To better compare and explain the changes in the habitat quality, we used ArcGIS software to divide the habitat quality results in each period into four intervals: (1) 0–0.3; (2) 0.3–0.6; (3) 0.6–0.9; (4) 0.9–1. According to these four intervals, we divided the habitat quality into four grades: (I) low-quality areas; (II) medium-quality areas; (III) higher-quality areas; (IV) high-quality areas (Figure 2). We calculated the score interval of each grade, the proportion of the habitat quality mass area in the three periods, and the mean value of the habitat quality in each grade (Table 4), and we present the results as follows.

In terms of the time changes, the mean values of the habitat quality in Harbin were 0.68, 0.66, and 0.65, which indicates a downward trend. In general, the habitat quality in Harbin is high, and the medium- and high-quality areas are large. Over the past 20 years, the proportion of low- and medium-quality habitats has continued to increase (by 1.25% and 3.21%, respectively). The proportion of higher-quality habitats substantially increased at first, and then slightly decreased, with an overall increase of 1.95%. However, the proportion of high-quality habitats sharply declined (by 6.43%). In addition, the changes in the medium-quality habitats were greater in the last decade than in the first decade, while the changes in the low-quality, higher-quality, and high-quality habitats were more substantial in the first decade. The above changes indicate that the habitat quality in Harbin has changed from high quality to medium quality and low quality, and that the original habitat area has decreased and is in the process of degradation.

Table 4. The level of habitat quality of Harbin City from 2000 to 2020

Level	Value interval	2000		2010		2020
		Area weight/%	Average value	Area weight/%	Average value	Area weight/%	Average value
I	0—0.3	4.70%	0.68	5.41%	0.66	5.95%	0.65
II	0.3—0.6	44.58%		45.54%		47.79%
III	0.6—0.9	9.69%		13.95%		11.64%
IV	0.9—1	41.04%		35.09%		34.61%

Figure 2. Spatial distribution of habitat quality in Harbin City from 2000 to 2020

In terms of the spatial pattern, the habitat quality of Harbin presents a trend of high in the middle, middle in the east, and low in the west. Among them, the high- and higher-quality areas are mainly concentrated in north Tonghe, central Fangzheng, eastern Shangzhi, and southern Wuchang. The primary land-use types of these areas are forest, grassland, and water, with the distribution of multiple forest parks and nature reserves, a high rate of forest coverage, rich biodiversity, a high ecological protection degree, and less human activity. Moreover, these areas are not easily affected by farmland and built-up land. The medium-quality areas are mainly distributed in the west, northeast, and central river valleys. Most of these areas are plains, with cropland as the primary land-use type and rural residential areas as the distribution. Human activities are relatively frequent, which cause ecological damage. The low-quality areas are distributed among a large area in the west and are sporadically distributed in the middle of the region. The primary land-use types in this region are urban land, rural residential areas, and industrial land. With rapid economic development and a relatively dense population, the expansion of built-up land, road construction, and other human activities have substantially intensified the habitat loss and degradation in these regions.

To further study the temporal and spatial variation characteristics of the habitat quality in Harbin, we calculated the difference values of the habitat quality for the three periods, and we used the segment point method (Jenks) to divide the calculated results into five categories: (1) dramatic decrease; (2) slight decrease; (3) no change; (4) slight increase; (5) dramatic increase. We thereby obtained the change chart of the habitat quality in Harbin from 2000 to 2020 (Figure 3). As can be seen from the figure, from 2000 to 2020, with the intensification of urbanization, the most substantial decreases in the habitat quality were around the central urban areas and along the rivers. Among them, the most prominent areas were the Songbei, Hulan, Daoli, Bayan–Mulan–Bin, and Tonghe–Yilan sections of the Songhua River. The quality of the habitats near the nature reserves, such as Tuanshanzi in Hulan, Pingdingshan in Tonghe, and Anxing Wetland in Yilan, was substantially improved. From 2000 to 2010, the degradation of the habitat quality was substantial, and the degradation areas were scattered, mainly due to the expansion of the central urban area to the periphery and the expansion of some rural residential areas and industrial and mining lands, which further stressed the surrounding habitats. The habitat degradation was not substantial from 2010 to 2020, and it primarily took place in the urban development zone and Bayan–Mulan–Bin area of the Songhua River, while the habitat quality recovered in the vicinity of the Yangdali Waterfowl Nature Reserve and other parts of Tonghe.

Figure 3. Spatial distribution of habitat quality change in Harbin City from 2000 to 2020

Landscape pattern index change

We present the changes in the landscape pattern index at the type level and landscape level of the different land-use types in Harbin from 2000 to 2020 in Figure 4 and Figure 5, respectively. We obtained the following results: (1) In terms of the landscape fragmentation degree, the PD of the built-up land was always the largest at each time node, and it’s related to the human activities and economic development in the study area. The LPI of the cropland and forest were the largest, as they represent the dominant land-use types of the region; (2) In terms of the landscape aggregation degree, the cropland and forest had the highest degrees of aggregation, which indicates that the distributions of these two land types have been relatively concentrated; (3) In terms of the landscape connectivity, the wetland LSI was the lowest, which indicates that the water spaces have become occupied and segmented, with complex shapes, an increasing degree of dispersion, and decreasing landscape dominance; (4) In terms of the landscape diversity, the SHDI and SHEI had the same trend. In terms of the timeseries, the landscape fragmentation degree in this region was strong in the first decade and gradually decreased in the last decade.

Figure 4. Disaggregated changes in class-level landscape metrics

The LPI of the forest and grassland substantially decreased, which indicates that the landscape areas of contiguous grassland and forest have shrunk with the spatial fragmentation. The agglomeration degree of the built-up land obviously increased, which reflects the increasing trend of building up land and the change in its spatial form in the process of urbanization. However, the aggregation degrees of the other land types demonstrated decreasing trends, which indicates that, with the passage of time, the landscape in this region has undergone spatial segmentation, which has resulted in a large number of small patches in the landscape and a decreased aggregation degree. The patch connectivity in the study area decreased first, and then increased with time. The overall connectivity increased, which indicates that, over the past 20 years, the landscape in this area has tended towards irregularity. The decrease in the landscape diversity in the study area indicates that the complexity and evenness of the landscape has decreased, and that the area proportion difference of the landscape types has slightly increased.

Figure 5. Disaggregated changes in landscape-level landscape metrics

Figure 6. Spatial distribution of Landscape Pattern change in Harbin City from 2000 to 2020

We processed the normalized grid graphs of the various landscape patterns using the GIS platform, and we superimposed them with equal weights according to the types. We obtained the spatial pattern distribution results of the landscape fragmentation, aggregation, connectivity, and diversity in the study area (Figure 6). From the perspective of the spatial scale, the spatial distribution characteristics of the four landscape types are similar. The landscape fragmentation, convergence, and diversity of Mulan and Tonghe in the north, Acheng in the middle, Shangzhi in the east, and Wuchang in the south are low. The patches in these areas are larger and complete, with regular shapes and good connectivity, and they are occupied by some landscape types with higher dominances. These areas are distributed among wetlands, geological forest parks, national forest parks, and national and provincial nature reserves, with high forest coverage rates, strong ecological environments, and relatively rich species. The landscape pattern of the central urban area in the west is relatively concentrated, with low fragmentation and diversity, which is because the built-up land in the central urban area is concentrated and contiguously distributed, and the landscape type is single. However, Bayan, Yanshou, Bin, Yilan, and other counties have higher landscape fragmentation, connectivity, and diversity because the landscape types in these areas are mainly cultivated land, and due to the strong human interference, the landscape fragmentation degree and heterogeneity are higher. From the timescale perspective, the landscape fragmentation degrees in the nature reserves and their surrounding areas decreased from 2000 to 2020, and the landscape pattern became more concentrated, better connected, and less diverse. Due to the promotion of biodiversity protection and restoration in Harbin, the project of returning farmland to forest, grassland, and wetland, and the construction of the Three North Shelterbelt Project, the local ecological environment has been restored. The landscape fragmentation, connectivity, and diversity for the city center and various districts and counties of the urban land and rural residential areas demonstrate increasing trends, which is because, against the background of rapid urbanization, erosion, and the mass sprawl of built-up land and other landscape patch types, a variety of land distribution types have emerged and become more diffuse. In Bayan, Yanshou, Bin, and Yilan, the fragmentation degrees of the cultivated landscape slightly increased, while the aggregation degrees decreased and the diversity degrees increased, which indicates that a large number of small patches of cultivated land appeared in these areas during the study period, which were closely related to the long-term implementation of the project of returning cultivated land to forest, grassland, and wetland.

Relationship between habitat quality and Landscape pattern

We used the Pearson’s correlation coefficient to test the correlation between the habitat quality and landscape pattern in Harbin for 2000, 2010, and 2020. We present the results in Figure 7. In general, 73.7% of the land-use types had significant correlations with the landscape pattern index (P < 0.01). (1) In terms of the landscape fragmentation degree, the PD of all the land-use types, except for the built-up land, was substantially positively correlated with the habitat quality. Forest and cropland were the two main landscape types in the study area, and the LPI of them were substantially positively and negatively correlated with the habitat quality, respectively, which indicates that they were the landscape types with the most intensive human activities during the whole study period, while the forests had the least intensive human activities. (2) In terms of the landscape aggregation degree, the AREA_MN and COHESION of the forest and grassland were substantially positively correlated with the habitat quality. The higher the concentration of forest and grassland, the better the habitat quality of the region. (3) In terms of the landscape connectivity, the CONTAG and LSI were strongly correlated with the habitat quality. The water and built-up land had substantial negative correlations with the habitat quality, which indicates that the habitat quality has decreased with the increase in the complexity and connectivity of these two landscape types. (4) In terms of the landscape diversity, the SHDI and SHEI were substantially negatively correlated with the habitat quality, which indicates that, with the diversification of the landscape types and the homogenization of their distributions, the habitat quality has declined. In terms of the timeseries, 36.4% of the Pearson’s correlation coefficients of the landscape pattern indices decreased first and then increased, while 24.2% continuously increased from 2000 to 2020. This change from 2000 to 2010, with the advancement of urbanization and the transformation of large areas of cropland and forest into built-up land, created more human disturbance, inevitable changes in the landscape pattern, the production of more uncertain landscape information, and a tendency towards homogenization, which caused decreases in most of the land landscape pattern indices and their correlations with the habitat quality. The maximum LPI and AREA_MN area were the most obvious. From 2010 to 2020, with the comprehensive promotion of the project of returning farmland to forest and the afforestation of barren hills and wasteland in Harbin, large areas of soil erosion and desertification lands were treated, and the forest coverage rate gradually improved. During this period, the correlations between the landscape pattern indices and habitat quality were enhanced in more than half of the indices. The PD and COHESION indices most strongly reflected this enhancement.

Figure 7. Pearson’s correlation between landscape pattern index and habitat quality in Harbin City

Discussion

Temporal and spatial evolution of habitat quality in Harbin city

From 2000 to 2020, the habitat quality in Harbin decreased from 0.68 to 0.65, presenting an overall downward trend with substantial spatial heterogeneity, which is consistent with the research of Li, Duo, et al. (2021). The low-quality habitat increased by 1.25%, while the high-quality habitat decreased by 6.43%, with a general distribution pattern of "high in the middle, middle in the east and low in the west". The central region contains many forest parks and nature reserves, and the primary land-use type is forestland. The forest ecosystem structure is complex, it has a self-recovery ability in the face of human disturbance, and it can maintain high habitat quality. In the northeast, west, and central river valleys, cropland is the primary land-use type, and there are rural residential areas with frequent human activities, which have resulted in no obvious improvement effect on the habitat quality. The deterioration of the habitat quality was most substantial in the built-up land in the west and along the river, which is mainly due to the distribution of the urban and rural residential areas and farmland along the river, the expansion of built-up land to the periphery, and other human activities that cause damage to the regional ecosystem.

Temporal and spatial evolution of landscape pattern in Harbin

From 2000 to 2020, the landscape pattern index of Harbin substantially changed. In the process of urbanization, the overall landscape pattern has shown a trend of further fragmentation, and the overall landscape has become irregular and spatially heterogeneous, with reduced diversity and uniformity, which is consistent with the research of Wang, Tang, et al. (2020). Against the background of rapid urbanization, the large-scale disordered expansion of the built-up land has eroded other landscape-patch types, and the landscape fragmentation, connectivity, and diversity demonstrate trends of continuous increases. To improve the ecological environment, the government has implemented a series of local ecological protection policies. As a result, the forest coverage rates in the northern, central, eastern, and southern regions have increased, the fragmentation degrees of the landscape patches have been continuously reduced, and the landscape pattern has become more concentrated, with better connectivity and less diversity. However, the landscape types in the northeast and western regions are mainly cropland. Under the implementation of the project of returning cropland to forest, grassland, and wetland, many small patches of cultivated land have appeared, and the landscape fragmentation degree and diversity have slightly increased, while the aggregation degree has decreased.

Correlation between habitat quality and landscape pattern in Harbin

We used Pearson’s correlation coefficient to explore the correlation between the habitat quality and landscape pattern in the city of Harbin. According to the results, there was a correlation between the habitat quality and landscape pattern index, which is consistent with previous research conclusions (Dai, Li, et al., 2019; Yang, 2021). The habitat quality and landscape pattern index had a substantial correlation for 73.7% of the land types. The PD, COHESION, and LSI had the most substantial correlations with the habitat quality. Urbanization and urban expansion have led to frequent human activities and developed road networks, which have increased the habitat fragmentation, with more dispersed landscape patterns that contain more edges. However, these patches are too small to meet the survival needs of individual species or entire populations, which has resulted in the deterioration of the surrounding habitat quality. In the past 20 years, 36.4% of the Pearson’s correlation coefficients of the landscape pattern indices initially decreased and then increased, which indicates that the urbanization speed accelerated in Harbin in the first decade, a large amount of cultivated land and forest were converted to built-up land, and the landscape tended towards homogeneity, which led to the weakening of the correlation between most of the land landscape pattern indices and the habitat quality. In the last ten years, with the comprehensive promotion of the project of returning farmland to forest and the afforestation of barren mountains and wasteland in Harbin, the forest coverage rate has gradually increased, and the correlations between more than half of the land landscape pattern indices and the habitat quality have increased.

Optimization strategy

The contradiction between biodiversity conservation and economic development is an urgent challenge for ecological restoration. Based on the spatial heterogeneity of ecological environments, landscape regulation that is based on the combination of biodiversity conservation and economic development is becoming the mainstream view in restoration ecology. Based on the analysis of the temporal and spatial evolution of the habitat quality and the landscape pattern response, we propose the following development strategies to promote the in-depth implementation of the new urbanization and the improvement in the regional habitat quality in Harbin:

(1) Wetlands, forest parks, and nature reserves play a key role in regional ecological security. We need to focus on protecting the forest, grassland, waters, and other ecological sources, strictly observing the red line for ecological protection, and protecting ecologically fragile and sensitive areas. By reducing the landscape fragmentation and improving the landscape connectivity, we can increase the large patches, build ecological corridors and biodiversity conservation networks, and improve the quality and stability of ecosystems;

(2) The increase in agricultural activities will lead to the disordered expansion of cropland, the encroachment of forests and grassland, and the degradation of habitats. We need to continue to implement the policy of returning farmland to forest and grassland in areas with fragmented landscapes, restore areas with degraded habitats by ecological means, organically integrate shelter forests, timber forests, economic forests, and other forest species, and build a complex agricultural and forestry network that combines trees, grass, and crops to maintain the balance of the ecosystem in the region;

(3) Urban sprawl poses a serious threat to habitats. In the process of urbanization, the scale of the built-up land should be well managed, and the development and construction activities should be strictly controlled within the boundaries of urban development to avoid the disorderly expansion of urban built-up land. We need to strengthen the construction of blue and green ecological infrastructures and give full play to the ecological function of green spaces.

Conclusions

Habitat quality and landscape pattern assessments are an important tool for optimizing the regional landscape pattern and maintaining the regional ecological security. In this study, we evaluated the spatial–temporal evolution of the habitat quality in the city of Harbin for 2000, 2010, and 2020 using the InVEST model, and we analyzed the response of the habitat quality to the landscape pattern change. According to the results, the overall habitat quality in Harbin indicates a downward trend, with substantial spatial heterogeneity. The distribution pattern was high in the middle, middle in the east, and low in the west. The overall landscape pattern indicates a trend of further fragmentation, and the fragmentation degrees of the dominant landscape types have increased, while the diversity and evenness of the dominant landscape types have decreased and present spatial heterogeneity. The habitat quality and landscape pattern indices had substantial correlations for most of the land types. The PD, COHESION, and LSI had the most substantial correlations with the habitat quality.

The InVEST model, despite its advantages for habitat quality assessments, has some limitations that require further improvement: (1) The model lacks further information on the diversity of organisms, such as animals, plants, and insects, observed over time, and the use of remote sensing observations alone is not sufficient; (2) No standard calculation method for the parameter setting of the InVEST model exists. We set the relevant parameters in this study according to the model manual, references, and expert judgment; thus, subjective judgment was inevitable; (3) Data mismatch increases the habitat quality uncertainty for historical periods, and we were unable to collect all the open geographic datasets required for the InVEST model. For example, we used road data from 2020 as a threat source for all three periods; (4) In this study, we only considered the impact of the threat sources on the habitat area within the study area; however, the habitats at the edge of the study area are also affected by other threat sources outside the boundary, which may have led to inaccurate assessment results. In future studies, researchers should focus on conducting comprehensive investigations and research, collecting detailed data sources, and establishing objective parameter settings to draw more accurate habitat quality maps and provide a more scientific basis for guiding ecological environmental protection.

With the acceleration of urbanization in Harbin, the population increase, urban expansion, and other problems will damage the regional ecological environment. Harbin is one of the areas in Heilongjiang province with serious habitat quality degradation. How to improve the ecosystem of Harbin is a key issue for future decision makers. In the future territorial space planning of Harbin, attention should be paid to the habitat patches, such as woodland, grassland, and wetland, that are highly suitable to enhancing the complexity of the community structure to strengthen the self-recycling ability of the system and form an organic whole. The habitat quality in the main urban area of Harbin is relatively weak. We need to improve the ecological condition through moderate manual intervention, strengthen the supplement of urban green microspaces, such as urban greenways and pocket parks, and promote the improvement in the habitat quality of the surrounding construction land. Along with economic development, the landscape pattern should be optimized to improve the overall habitat quality in Harbin.

Author Contributions

Conceptualization, G. R., L. S. and F. J.; methodology, L. S. and F. J.; software, L. S. and F. J.; investigation, L. S.; data curation, L. S.; writing—original draft preparation, L. S.; writing—review and editing, L. S. and F. J.; supervision, G. R.. All authors have read and agreed to the published version of the manuscript.

Ethics Declaration

The authors declare that they have no conflicts of interest regarding the publication of the paper.

References

•  Bai,  L.,  Xiu,  C., et al. (2019). “Influence of Urbanization on Regional Habitat Quality:A Case Study of Changchun City”. Habitat International, 93, 102042. doi: https://doi.org/10.1016/j.habitatint.2019.102042.
• Berta Aneseyee, A., Noszczyk, T., et al. (2020). “The InVEST Habitat Quality Model Associated with Land Use/Cover Changes: A Qualitative Case Study of the Winike Watershed in the Omo-Gibe Basin, Southwest Ethiopia”. Remote Sensing, 12(7), 1103. doi: https://doi.org/10.3390/rs12071103.
• Biswas, A., and Si, B. C. (2011). “Scales and Locations of Time Stability of Soil Water Storage in a Hummocky Landscape”. Journal of Hydrology, 408(1–2), 100–112. doi: https://doi.org/10.1016/j.jhydrol.2011.07.027.
•  Cheng,  R.,  Chen,  W., et al. (2019). “Spatiotemporal Variation Characteristics of Ecosystems and Their Driving Forces in the Min Delta Urban Agglomeration”. International Review for Spatial Planning and Sustainable Development, 7(1), 177–194. doi: https://doi.org/10.14246/irspsda.7.1_177.
•  Chopra,  B.,  Singh,  B., et al. (2022). “Spatio-Temporal Analysis of Land Use/Land Cover Changes in Urban Forest Ecosystem”. International Review for Spatial Planning and Sustainable Development, 10(3), 128–147. doi: https://doi.org/10.14246/irspsd.10.3_128.
•  Cozzoli,  F.,  Smolders,  S., et al. (2017). “A Modeling Approach to Assess Coastal Management Effects on Benthic Habitat Quality: A Case Study on Coastal Defense and Navigability”. Estuarine, Coastal and Shelf Science, 184,67–82. doi: https://doi.org/10.1016/j.ecss.2016.10.043.
•  Dai,  L.,  Li,  S., et al. (2019). “The Influence of Land Use Change on the Spatial–Temporal Variability of Habitat Quality between 1990 and 2010 in Northeast China”. Journal of Forestry Research, 30(6), 2227–2236. doi: https://doi.org/10.1007/s11676-018-0771-x.
•  Ding,  Y.,  Shan,  B., et al. (2015). “Assessment of River Habitat Quality in the Hai River Basin, Northern China”. International Journal of Environmental Research and Public Health, 12(9), 11699–11717. doi: https://doi.org/10.3390/ijerph120911699.
•  Duffy , J. E. (2009). “Why Biodiversity Is Important to the Functioning of Real-World Ecosystems.” Frontiers in Ecology and the Environment, 7(8), 437–444. doi: https://doi.org/10.1890/070195.
•  Evans , K. L.,  Rodrigues , A. S., et al. (2006). “Protected Areas and Regional Avian Species Richness in South Africa”. Biology Letters, 2(2), 184–188. doi: https://doi.org/10.1098/rsbl.2005.0435.
• Gao, Y., Ma, L., et al. (2017). “Constructing Ecological Networks Based on Habitat Quality Assessment: A Case Study of Changzhou, China”. Scientific Reports, 7(1), 46073. doi: https://doi.org/10.1038/srep46073.
•  Goldstein , J. H.,  Caldarone,  G., et al. (2012). “Integrating Ecosystem-Service Tradeoffs into Land-Use Decisions”. Proceedings of the National Academy of Sciences, 109(19), 7565–7570. doi: https://doi.org/10.1073/pnas.1201040109.
•  Hagen-Zanker,  A. (2016). “A Computational Framework for Generalized Moving Windows and Its Application to Landscape Pattern Analysis”. International Journal of Applied Earth Observation and Geoinformation, 44, 205–216. doi: https://doi.org/10.1016/j.jag.2015.09.010.
•  Hall , L. S.,  Krausman , P. R., et al. (1997). “The Habitat Concept and a Plea for Standard Terminology”. Wildlife Society Bulletin (1973-2006), 25(1), 173–182.
•  He, J.,  Huang,  J., et al. (2017). “The Evaluation for the Impact of Land Use Change on Habitat Quality: A Joint Contribution of Cellular Automata Scenario Simulation and Habitat Quality Assessment Model”. Ecological Modelling, 366, 58–67. doi: https://doi.org/10.1016/j.ecolmodel.2017.10.001.
•  Hu, W.,  Wang,  G., et al. (2008). “Advance in Research of the Relationship between Landscape Patterns and Ecological Processes”. Progress in Geography, 27(1), 18–24.
•  Jung,  J.,  Shimizu,  Y., et al. (2016). “Developing and Testing a Habitat Suitability Index Model for Korean Water Deer ( Hydropotes Inermis Argyropus ) and Its Potential for Landscape Management Decisions in Korea”. Animal Cells and Systems, 20(4), 218–227. doi: https://doi.org/10.1080/19768354.2016.1210228.
• Kareiva, P., Tallis, H., et al. eds. (2011). Natural Capital: Theory and Practice of Mapping Ecosystem Services. Oxford University Press.
•  Kassawmar,  T.,  Murty , K. S. R., et al. (2019). “Making More out of Pixel-Level Change Information: Using a Neighbourhood Approach to Improve Land Change Characterization across Large and Heterogeneous Areas”. Geocarto International, 34(9), 977–999. doi: https://doi.org/10.1080/10106049.2018.1458252.
•  Li, Y.,  Duo,  L., et al. (2021). “Assessment and Estimation of the Spatial and Temporal Evolution of Landscape Patterns and Their Impact on Habitat Quality in Nanchang, China”. Land, 10(10), 1073. doi: https://doi.org/10.3390/land10101073.
•  Lin, Y.-P.,  Lin,  W.-C., et al. (2017). “Systematically Designating Conservation Areas for Protecting Habitat Quality and Multiple Ecosystem Services”. Environmental Modelling & Software, 90, 126–146. doi: https://doi.org/10.1016/j.envsoft.2017.01.003.
•  Lindenmayer,  D.,  Hobbs , R. J., et al. (2007). “A Checklist for Ecological Management of Landscapes for Conservation”. Ecology Letters, 071010211025003-??? doi: https://doi.org/10.1111/j.1461-0248.2007.01114.x.
•  Liu,  H.,  Zhang,  S., et al. (2004). “Impacts on Wetlands of Large-Scale Land-Use Changes by Agricultural Development: The Small Sanjiang Plain, China”. AMBIO: A Journal of the Human Environment, 33(6), 306–310. doi: https://doi.org/10.1579/0044-7447-33.6.306.
•  Liu,  L.,  Chen,  X., et al. (2020). “Identifying the Impact of Landscape Pattern on Ecosystem Services in the Middle Reaches of the Yangtze River Urban Agglomerations, China”. International Journal of Environmental Research and Public Health, 17(14), 5063. doi: https://doi.org/10.3390/ijerph17145063.
•  Liu,  S.,  Liao,  Q., et al. (2022). “Spatial and Temporal Variations of Habitat Quality and Its Response of Landscape Dynamic in the Three Gorges Reservoir Area, China”. International Journal of Environmental Research and Public Health, 19(6),3594. doi: https://doi.org/10.3390/ijerph19063594.
•  Long,  J.,  Nelson,  T., et al. (2010). “Regionalization of Landscape Pattern Indices Using Multivariate Cluster Analysis”. Environmental Management, 46(1), 134–142. doi: https://doi.org/10.1007/s00267-010-9510-6.
• Mairota, P., Cafarelli, B., et al. (2015). “Very High Resolution Earth Observation Features for Testing the Direct and Indirect Effects of Landscape Structure on Local Habitat Quality”. International Journal of Applied Earth Observation and Geoinformation, 34, 96–102. doi: https://doi.org/10.1016/j.jag.2014.07.003.
•  Monteiro , A. T.,  Fava,  F., et al. (2013). “Landscape Context Determinants to Plant Diversity in the Permanent Meadows of Southern European Alps”. Biodiversity and Conservation, 22(4),937–958. doi: https://doi.org/10.1007/s10531-013-0460-1.
•  Nelson,  E.,  Sander,  H., et al. (2010). “Projecting Global Land-Use Change and Its Effect on Ecosystem Service Provision and Biodiversity with Simple Models”. PLoS ONE, 5(12), e14327. doi: https://doi.org/10.1371/journal.pone.0014327.
•  Nowosad,  J., and  Stepinski , T. F. (2018). “Global Inventory of Landscape Patterns and Latent Variables of Landscape Spatial Configuration”. Ecological Indicators, 89, 159–167. doi: https://doi.org/10.1016/j.ecolind.2018.02.007.
•  Paul,  S., and  Bardhan,  S. (2022). “Overview: Framework for Quantitative Assessment of Urban-Blue-and-Green-Spaces in a High-Density Megacity”. International Review for Spatial Planning and Sustainable Development, 10(3), 280–301. doi: https://doi.org/10.14246/irspsd.10.3_280.
•  Polasky,  S.,  Nelson,  E., et al. (2011). “The Impact of Land-Use Change on Ecosystem Services, Biodiversity and Returns to Landowners: A Case Study in the State of Minnesota”. Environmental and Resource Economics, 48(2), 219–242. doi: https://doi.org/10.1007/s10640-010-9407-0.
•  Rands , M. R. W.,  Adams , W. M., et al. (2010). “Biodiversity Conservation: Challenges Beyond 2010”. Science, 329(5997), 1298–1303. doi: https://doi.org/10.1126/science.1189138.
•  Sala , O. E.,  Stuart Chapin,  F., et al. (2000). “Global Biodiversity Scenarios for the Year 2100.” Science, 287(5459),1770–1774. doi: https://doi.org/10.1126/science.287.5459.1770.
• Sharma, R., Nehren, U., et al. (2018). “Modeling Land Use and Land Cover Changes and Their Effects on Biodiversity in Central Kalimantan, Indonesia”. Land, 7(2), 57. doi: https://doi.org/10.3390/land7020057.
•  Sharp,  R.,  Wood,  S., et al. (2018). “InVEST User’s Guide”. Retrieved from https://invest-userguide.readthedocs.io/en/latest/index.html on July 10, 2021.
•  Sherrouse , B. C.,  Semmens , D. J., et al. (2014). “An Application of Social Values for Ecosystem Services (SolVES) to Three National Forests in Colorado and Wyoming”. Ecological Indicators, 36, 68–79. doi: https://doi.org/10.1016/j.ecolind.2013.07.008.
•  Shuangao,  W.,  Padmanaban,  R., et al. (2021). “Using Satellite Image Fusion to Evaluate the Impact of Land Use Changes on Ecosystem Services and Their Economic Values”. Remote Sensing, 13(5), 851. doi: https://doi.org/10.3390/rs13050851.
•  Terrado,  M.,  Sabater,  S., et al. (2016). “Model Development for the Assessment of Terrestrial and Aquatic Habitat Quality in Conservation Planning”. Science of The Total Environment, 540, 63–70. doi: https://doi.org/10.1016/j.scitotenv.2015.03.064.
• United Nations in China. n.d. “United Nations in China.”. Retrieved from https://www.un.org/zh on July 5, 2021.
•  Uuemaa,  E.,  Antrop,  M., et al. (2009). “Landscape Metrics and Indices: An Overview of Their Use in Landscape Research”. Living Reviews in Landscape Research, 3. doi: https://doi.org/10.12942/lrlr-2009-1.
•  Vögeli,  M.,  Serrano,  D., et al. (2010). “The Relative Importance of Patch Habitat Quality and Landscape Attributes on a Declining Steppe-Bird Metapopulation”. Biological Conservation, 143(5),1057–1067. doi: https://doi.org/10.1016/j.biocon.2009.12.040.
•  Wang,  H.,  Tang,  L., et al. (2020). “Assessing the Impacts of Urban Expansion on Habitat Quality by Combining the Concepts of Land Use, Landscape, and Habitat in Two Urban Agglomerations in China”. Sustainability, 12(11), 4346. doi: https://doi.org/10.3390/su12114346.
•  Yan,  S.,  Wang,  X., et al. (2017). “Investigation of the Spatio-Temporal Dynamics in Landscape Variations in a Shallow Lake Based on a New Tendency-Pattern-Service Conceptual Framework”. Journal of Cleaner Production, 161, 1074–1084. doi: https://doi.org/10.1016/j.jclepro.2017.05.114.
•  Yang,  Y. (2021). “Evolution of Habitat Quality and Association with Land-Use Changes in Mountainous Areas: A Case Study of the Taihang Mountains in Hebei Province, China”. Ecological Indicators, 129, 107967. doi: https://doi.org/10.1016/j.ecolind.2021.107967.
•  Yu, F.,  Liu,  F., et al. (2022). “Classification and Assessment Methods for Mountain Channel Habitats in the Chishui River Basin, China”. Water, 14(4), 515. doi: https://doi.org/10.3390/w14040515.