Spatiotemporal Evolution of Ecosystem Services in the Wanhe Watershed Based on Cellular Automata (CA)-Markov and InVEST Models

Abstract

The evaluation of habitat quality and its genesis is of great significance to ecological protection of the watershed. Based on land use data, Digital Elevation Model (DEM), and road network data and population data, the Cellular Automata (CA)-Markov model and InVEST model were used to analyzed the land use change in the Wanhe Watershed, predicting the land use in 2025. Based on this, the degree of the habitat degradation and habitat quality in 2000, 2005, 2010, 2015, 2020, and 2025 were predicted and analyzed, and combined with the particularity of the terrain in the study area, the topography was introduced. Landform relief gradient was used to discuss the relationship between habitat quality and topographic factors in the Wanhe Watershed, and to reveal the distribution law. The result shows that from 2000 to 2025, farmland and forestland are the main land use types in the study area, and the main change is due to the expansion of the construction land, whereby the area increased by 62.86 km², with an increase of 34.41%, mainly from farmland and forestland. From 2000 to 2020, mainly due to the expansion in urban land and the reduction in forestland, the habitat degradation tends to be serious, and the habitat quality generally shows a downward trend, with areas with low habitat quality having had the largest increase from 452.67 km² in 2000 to 526.15 km² in 2025. The topography of the study area affects the distribution of natural landscapes and the intensity of human activities, resulting in significant differences in the landscape pattern of habitat degradation and habitat quality. The western mountains are relatively better. After 2020, due to the implementation of environmental protection policies, the habitat quality has tended to improve. This study can promote the adjustment of land use planning policies in the study area, maintain the biodiversity in the watershed, and realize the coordinated development of environmental benefits and social and economic development. The research results have theoretical significance and practical value for ecological environmental protection and land use layout in Wanhe Watershed.

1. Introduction

The damage caused by rapid urbanization outpaces the rate at which the ecological environment can repair itself, leading to ecological degradation at all scales worldwide [1]. Habitat Quality (HQ) refers to the ability of an ecosystem to provide suitable living conditions for individuals and populations [2]. Its change is a direct result of the interaction between human activities and the natural environment [3]. It affects biodiversity and reflects the status of the ecological environment [4,5]. Land use is closely related to habitat quality because of its impact on material and energy cycles. The change of land use will inevitably bring about the change of habitat quality [6]. The impact of land use change, especially on ecosystem functions, has become a research hotspot in ecology, geography, and other fields [7]. The rapid urbanization has led to changes in land use patterns. Some forests, grasslands, and water areas have been converted into urban land. This has resulted in the deterioration of an ecological environment and that of ecological habitat quality [8]. Therefore, analyzing the degradation degree of habitat quality and the spatiotemporal evolution characteristics of habitat quality can provide a scientific basis for the construction of ecological civilization [9].

The index system method is based on landscape pattern [10] and the model-based method [11] are mostly used to evaluate habitat quality at present. The latter plays a prominent role in predicting future habitat distribution and urban development planning in combination with land use change. Scholars often use habitat suitability models, such as Social Values for Ecosystem Services (SolVES) model [12], to evaluate regional habitat quality. Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) model (https://naturalcapitalproject.stanford.edu/software/invest, accessed on 3 May 2022) is jointly developed by Stanford University, Nature Conservancy and World Wildlife Fund (WWF). InVEST model is a tool for simulating dynamic changes in habitat quality by using land use change scenarios. It has many advantages: relatively small data demand, convenient operation, strong spatial expression ability, and objectivity of output results. The model includes many patterns evaluating ecosystem services such as water conservation, habitat quality assessment, water supply calculation, and soil erosion and is widely used in the world [13,14]. Gomes et al. [15] simulated and predicted future habitat quality in different scenarios. Xu et al. [6] assessed the spatiotemporal evolution of habitat quality in Taihu Lake Watershed from 1985 to 2015. Lin et al. [13] evaluated the degradation of habitat quality in Three Gorges Reservoir, Hubei province, China. Tang et al. [15] evaluated the evolution of habitat quality in Huaihe River Economic Belt, China. Scholars have studied the functions and characteristics of different ecological services at different scales, and it confirmed the obvious advantages of InVEST model in evaluation and estimation research.

Many cases may lead to the changing in land use. Land use change caused by industrialization, urban expansion, and other human activities is the main driving force of habitat deterioration [16]. Therefore, the assessment of habitat quality evolution and distribution based on predicted future land use can contribute to the formulation of land use and environmental protection policies. There are various models simulating and predicting land use change. Common prediction models include Future Land Use Simulation (FLUS) model [17], Conversion of Land Use (CLUE-S) model [17], Cellular Automaton (CA) random model, etc. CA-Markov model integrates the Markov chain model of CA and stochastic model, and is used for the simulation and prediction of land use at different spatiotemporal scales. It has the advantages of high simulation accuracy and visual output of prediction results [8].

Topography is also an important factor resulting in the spatial differentiation of land use and habitat quality patterns due to its influence on human activities, vegetation distribution, and other natural and human geographical factors. Landform relief gradient is an important index to describe topographic features. Different land use types and their conversion and transfer areas have obvious differences in the areas with different gradients, which thus leads to the obvious correlation between habitat quality and landform relief gradient. Therefore, analyzing the relationship can contribute to study the causes of such distribution patterns [18].

Most of the spatial scales selected by scholars in recent years are towns [19], metropolitans [20], and reserves [21], etc. However, there are few studies on the habitat quality of watersheds of the Yangtze River in a relatively small scale. Large-scale studies are more macroscopic on the whole, but they tend to ignore details of small scale and lack pertinence of small scale. Furthermore, the ecology of small watersheds is relatively vulnerable, affected by many factors such as vegetation, precipitation, etc. [22]. Therefore, the study of habitat quality in the watershed of that scale is of practical significance for ecological restoration. Wanhe Watershed is located in the Yangtze River Delta urban agglomeration and plays a pivotal role in the development of the Yangtze River Delta Urban agglomeration. The western part of Wanhe Watershed is located in the Dabie Mountains with complex terrain, diverse biological [23] environments, underdeveloped economy, limited land resources, and frequent natural disasters [24]. In recent years, the rapid development of urbanization, transportation, and tourism has become an important pillar in the region, playing an important role in promoting and supporting regional social and economic development [25].

The study of land use and habitat quality is of great significance to regional assessment and future planning. In order to better study the temporal and spatial changes in the landscape pattern of the Wanhe Watershed, this paper aims to study land use and habitat quality by combining the CA-Markov model and the InVEST model and forecast. We evaluated habitat degradation and habitat quality in the Wanhe Watershed and explored the influencing mechanism of the change. Landform relief gradient was used to discuss the relationship between habitat quality and topographic factors in the Wanhe Watershed, and revealed the cause of such distribution. We aimed to provide a scientific basis for ecological protection planning and land management policy in the Wanhe Watershed.

2. Materials and Methods

2.1. Study Area

The Wanhe Watershed (30°03′–31°01′ N, 115°45′–117°02′ E) is an important part of the Yangtze River Watershed and plays a pivotal role in the development of the Yangtze River Delta Urban agglomeration. The Wanhe Watershed is located in the southwest of Anhui Province in China, between the southern slope of the Dabie Mountains and the main stream of the Yangtze River. It is an important part of Anqing City. The upper reaches of the Wanhe River system are composed of three tributaries: the Changhe River, the Qianshui River, and the Wanshui River (Figure 1).

The terrain of the Wanhe Watershed varies greatly, and the overall pattern is high in the northwest and low in the southeast. The area of the mountainous area is 3337 km², accounting for 51.8% of the area. The area of hilly area is 1689 km², accounting for 26.2%. Wanhe Watershed passes through the Tan-Lu fault [26] zone, which is one of the famous deep fault zones in Eastern Asia. The mountainous part of the watershed is the Dabie Mountains, with various landform types, including hills, mountains, and inter-mountain pelvises, undulating hills, large cutting depths, and terrain slopes of 20–45°, which are prone to landslides and other geological disasters. The problems of flood disaster and soil erosion are more prominently affected by climate, topography, and human activities in the region. The Wanhe Watershed is an important part of the Yangtze River Watershed. The Wanhe Watershed has a GDP of CNY 40.93 billion and a population of 4.16 million. Yuexi, Qianshan, and Taihu counties within the watershed are important political, economic, industrial, and commercial centers as well as power and transportation hubs. In addition to traditional farming, forestry, and breeding industries, the watershed also includes processing and manufacturing industries such as machinery, textiles, and building materials, as well as mining industries such as copper, iron, and coal. Ecological and environmental problems such as flood disaster and soil erosion often occur due to the special topographic and geomorphic features in the watershed, as well as the effects of economic and social factors. Therefore, it is of great necessity and practical significance to carry out habitat quality assessment in the Wanhe Watershed and put forward relevant suggestions.

2.2. Data Resources and Processing

Land use data of the Wanhe Watershed in 2000, 2005, 2010, 2015 and 2020 were obtained from Resource and Environment Science and Data Center (http://www.resdc.cn/, accessed on 12 May 2022), with a spatial resolution of 30 m (Table 1). The landscape was classified on the reference of Land Use Category System of the Chinese Academy of Sciences and according to research needs. Table 2 shows the landscape category system for use in our study, which was divided into 6 classes of the first class and 15 classes of the second class. In this paper, the area change of land use types on the first class in different periods was counted, and then the land use data in 2000 and 2025 were integrated and superimposed, obtaining the land use transfer matrix, which described the trend of land use mutual transfer.

Digital elevation model (DEM) data of the study area include ASTER GDEM V2 digital elevation data with a resolution of 30 m of Geospatial Data Cloud (http://www.gscloud.cn/ accessed on 13 May 2022). Slope and aspect data were calculated according to the DEM data. The optimal statistical unit size of landform relief gradient obtained by DEM data was 0.2916 km², using Geographic Information System (GIS) window increment analysis [22]. Road network and town center datasets were obtained from Open Street Map (https://download.geofabrik.de/ accessed on 10 May 2022). Then, they were sorted out and the Euclidean distance was calculated via the GIS method. The sub-datasets of the study area were extracted by mask. The population spatial distribution data are from WorldPop (https://www.worldpop.org/ accessed on 13 May 2022) with a resolution of 100 m. Due to different resources, all the data need to be normalized standard processing through projection conversion and resampling. The spatial reference was unified as WGS 84/UTM zone 50 N, with a spatial resolution of 30 m (Table 1).

2.3. Methods

2.3.1. Land Use Simulation Using CA-Markov

Cellular Automata (CA) model is a kind of system dynamics model, which can predict the time and space at the same time. In the discrete cellular space, the state of the cell at t + 1 is predicted based on its own and the surrounding critical state at t time. Based on this, the complex evolution of land use is simulated. It has strong spatial prediction ability [8,27]. The model is as follows:

(1)$S\left(t,t+1\right)=f\left[S\left(t\right),N\right]$

In the formula, S is a finite two discrete state sets of cells; t, t + 1 for different time; N is cell neighborhood; f is the rule of cell transformation in local space.

Markov model is a long-term prediction method based on a dynamic random model, which studies the initial state of random events and the transition probability between different states, determines its change rule, and predicts its next state. Land use change under Markov process is as follows:

(2)$P_{ij}=\left[\begin{array}{cccc}{P}_{11}& P_{12}& \cdots & P_{1n}\\ P_{21}& P_{22}& \cdots & P_{2n}\\ \cdots & \cdots & \cdots & \cdots \\ P_{n1}& P_{n2}& \cdots & P_{nn}\end{array}\right],\mathrm{and}{\displaystyle \sum }_{j=1}^n{P}_{ij}=1 \left(i,j=1,2,\dots ,n\right)$

(3)$S_{t+1}=P_{ij}\times S_t$

S_t and S_t+1 are the state of ecosystem land cover at time of t and t+1, respectively; P_ij is the transition probability matrix of a certain state; n is the type of land use.

Both CA model and Markov model have limitations: the CA model has a weak ability to explain spatiotemporal changes of expansion, whereas the Markov model cannot predict spatial distribution changes when predicting land use changes. The CA-Markov model is the coupling of cellular automata model and the Markov model. It is used to simulate and predict the future land use pattern, which can improve the accuracy of land use prediction in quantity. It can also effectively simulate the spatial evolution trend of land use in the structure. It is more practical and scientific and can better predict and simulate the spatiotemporal pattern of land use in quantity and time.

The Kappa consistency test is a method used to verify the accuracy of the model results, which is widely used to simulate the accuracy of spatial distribution maps of land use [28,29]. Kappa coefficient calculation method is as follows.

(4)$Kappa=\frac{P_0-P_c}{P_p-P_c}$

P₀ is the correct proportion of land use distribution simulation; P_c is the correct proportion of simulation in random case; P_p is the perfect ratio for the model.

F1 score (also known as F-measure, or balanced F-score) is an error metric which measures model performance by calculating the harmonic mean of precision and recall for the minority positive class [30]. It is a popular metric to use for classification models as it provides accurate results for datasets and takes into account both the precision and recall ability of the model.

(5)$precision=\frac{True Positive}{True Positive+False Positivie}$

(6)$Recall=\frac{True Positive}{True Positive+False Negative}$

(7)$F-Score=2\times \frac{Precision\times Recall}{Precision+Recall}$

2.3.2. Habitat Quality

Habitat quality index is a comprehensive index to evaluate habitat suitability and degradation degree in the study area [31]. The level of habitat degradation index reflects the extent to which the land class is affected by threat factors under the current protection level [32,33]. It reveals the possibility of potential habitat destruction and habitat quality declined in the study area. It has important reference value for delimiting habitat protection area. The data of land use and threat factor were imported into the habitat quality assessment module of the InVEST model. The habitat degradation and habitat quality of the current study area can be obtained by running the model. In this paper, the Habitat Quality Model in InVEST 3.10.2 software was used to evaluate the habitat quality of the Wanhe Watershed. The core of this method is to establish a link between habitat quality and threat sources, calculate the negative impact of threat sources on habitats, and obtain the habitat degradation degree, so as to calculate habitat quality through habitat suitability and degradation degree. Among them, the calculation formulas of index decline and habitat degradation degree are as follows:

(8)$i_{rxy}=exp\left(-\left(\frac{2.99}{d_{rmax}}\right)d_{xy}\right)$

(9)$D_{xj}={{\displaystyle \sum }}_{r=1}^R{{\displaystyle \sum }}_{y^r=1}^Y\left(\frac{w_r}{{\sum }_{r=1}^R{w}_r}\right)r_y{i}_{rxy} {\beta }_x{S}_{jr}$

w_r is the number of threat factors; y_r is the weight of threat factor; r_y is the number of grids of threat layer on the map layer; i_rxy is the strength of the threat factor; β_x is the number of threat factors on each grid. i_rxy is the threat level of threat factor to habitat; S_jr is the sensitivity of j land types to threat factors; d_xy is the distance between grid x (habitat) and grid y (threat factor); d_rmax is the maximum influence range of threat factors, and the habitat degradation degree is between 0 and 1. The larger the value is, the higher the habitat degradation degree. The calculation formula of habitat quality is as follows:

(10)$Q_{xj}=H_j\left(1-\frac{D_{xj}^z}{D_{xj}^z+k^z}\right)$

H_j is the habitat suitability of land type j; k is the semi-saturation constant, namely half of the maximum degradation degree; D_xj is the habitat degradation degree of grid x in category j; z is the default parameter of the model. The habitat quality value is between 0–1, and the higher the Q_xj value, the better the habitat quality.

The main parameters required for the habitat quality model include the distance and weight of threat factors, habitat suitability, and sensitivity to each threat factor. Paddy field, dry-farm field, urban land, the rural residential area, industrial land, and traffic land are selected as threat factors according to the reference value recommended by the model and the situation of the study area.

The influence scope and weight are assigned according to the influence degree of human activities on the natural environment (Table 3). The suitability of each habitat type and its sensitivity to threat factors were assigned referring to the sample data of the InVEST model and the studies of Foresman et al. [2,32,33,34] (Table 4).

2.3.3. Landform Relief Gradient

Landform relief gradient is an important index for quantitative description of terrain morphology [37]. It is mainly extracted by ArcGIS neighborhood tool, and its calculation formula [18] is as follows:

(11)$C=C_{max}-C_{min}$

C is landform relief gradient, C_max is the maximum elevation (m) in the analysis window, and C_min is the minimum elevation (m) in the analysis window.

3. Result and Analysis

3.1. Model Accuracy Verification

This study uses both Kappa consistency test and F1 Score [38] to verify the accuracy of land use forecast data in 2025. Use 2000 and 2005 to simulate 2010 land use data, use 2005 and 2010 to simulate 2015 land use data, and use 2010 and 2015 to simulate 2020 land use data. Compare all simulation results with actual land use data, then calculate the Kappa index and F1 Score (Table 5). According to the Kappa coefficient grading standard [39] and the general rule of thumb for F1 Score [30,40], the Kappa coefficient is greater than 90%, and the highest weighted of F1 Score is 85%, indicating that the simulated data is highly consistent with the real data [41], indicating that the simulation results are credible, and the verified CA-Markov model rules can be used for land use forecasting in 2025.

3.2. The Spatiotemporal Distribution Characteristics of Land Use

Between 2000 and 2025, the main land use types in the Wanhe Watershed were farmland and forestland (Figure 2), accounting for about 85% of the watershed area (Table 6). During the study period, the farmland and forestland decreased, which was the most significant type of net loss. The farmland area of the Wanhe River decreased from 2408.6 km² to 2369.18 km². The forestland area decreased from 3105.03 km² to 3085.84 km², with the proportion of 0.62%. Grassland showed a trend of decreasing while fluctuating. It decreased from 2000 to 2015, increased slightly from 2015 to 2020, and decreased from 2020 to 2025. Water fluctuated as well. They increased slightly from 2000 to 2005, slightly decreased from 2005 to 2020, and slightly increased again from 2020 to 2025, with an overall decrease of 0.41% from 2000 to 2025. The most significant increase in the land type was the construction land, which increased from 182.69 km² to 245.55 km². The unused land area was relatively stable.

Figure 3 is the Sankey diagram of the land use transition matrix in the Wanhe Watershed from 2000 to 2025. It can be seen from the figure that cultivated land, forestland, and grassland all showed a decreasing trend, whereas construction land increased significantly. According to the matrix (Table 7), the proportion of construction land transferred is the largest, and the area increased by 62.86 km², with an increase of 34.41%, mainly from farmland and forestland. The areas of farmland and forestland decreased by 39.42 km² and 19.19 km², respectively, indicating that during the economic development of the Wanhe Watershed, the expansion of construction land occupied a large amount of cultivated land and forests. Due to the diversion of rivers in some areas, the maintenance of farmland and the return of farmland to lakes, the proportion of water in and out is relatively large, but the total area is relatively stable.

3.3. The Spatiotemporal Distribution Characteristics of Habitat Degradation

According to the situation of the study area, habitat degradation was classified as follows: very slight degradation (0–0.005), slight degradation (0.005–0.015), moderate degradation (0.015–0.025), severe degradation (0.025–0.04), and very severe degradation (0.04–0.075) so as to carry on our study (Figure 4).

The degradation of habitats in the Wanhe Watershed became increasingly serious from 2000 to 2020 (Table 8). The maximum value increased from 0.0628 to 0.0720, and the average value of the habitat degradation gradually increased from 0.0154 in 2000 to 0.0175 in 2020, and the most severe areas were still deteriorating. From 2000 to 2020, very slight and slight degradation area of the Wanhe Watershed decreased by 393.44 km². Additionally, from 2020 to 2025, it increased by 170.02 km². The severe and very severe degradation area increased from 1304.26 km² in 2005 to 2826.29 km² in 2025 with an increase of 1564.03 km², accounting for up to 15% of the initial area. In the Wanhe Watershed from 2000 to 2025, habitat degradation degree showed an overall upward trend; however, the maximum and average values of habitat degradation in 2025 are lower than those in 2020.

Habitat degradation in the Wanhe Watershed presented a clear dividing line at the junction of the plains and mountains. The degradation of habitats in plain areas was more obvious, while that in mountainous areas was lighter. This suggests that areas with high vegetation cover have less potential threat of habitat degradation. In the central area of Yuexi County in the northwest of the watershed, there was a pattern of severe habitat degradation centered on the Yuexi North Hub, formed by Jinguang Expressway and Yuewu Expressway and radiating to the surrounding area. The habitat quality in the Wuchang Lake area in the southeast of the watershed was relatively good, but as the lake area decreased, part of the water surface degenerated into beaches, and the habitat degradation tended to be serious.

3.4. Spatiotemporal Evolution of Habitat Quality

According to the equal spacing method, the habitat quality was divided into five grades: very low (0–0.2), low (0.2–0.4), moderate (0.4–0.6), high (0.6–0.8), and very high (0.8–1) (Figure 5). From the perspective of the time pattern, in the five years, very high habitat quality areas accounted for the largest proportions. However, there was still a slight decline, from 42.19% in 2000 to 37.31% in 2020. The proportion would increase to 40.46% in 2025. The areas with moderate habitat quality accounted for the second largest, with a relatively large decline in 2000 to 2020, from 38.01% to 34.27%. The two types of habitat quality areas accounted for more than 70% of the Wanhe Watershed. Areas with high habitat quality took the third place, and that area had an upward trend, indicating that the overall habitat quality of the Wanhe Watershed, especially the Dabie Mountains, was relatively good. However, areas with low habitat quality had the largest increase: from 452.67 km² in 2000 to 526.15 km² in 2025. Both very low and low habitat quality areas had a trend of increasing while fluctuating. From 2010 to 2020, rapid urban expansion near Yuxiu Avenue in Pingshan Town, Huaining County led to a significant decline in habitat quality. The upstream of Hualiangting Reservoir improved the quality of its habitat due to the protection of water (Table 9).

From the perspective of the spatial pattern (Figure 5), the overall habitat quality of the watershed had such a distribution: high in the western mountainous area, middle in the central and eastern plain area, and low in rivers in the southeast. The habitat quality of forestland and grassland was higher than that of farmland and construction land. As the western area of the study area was covered by large forestland, the trend of urban land expansion was limited, and the habitat quality was generally high. The eastern plain accounted for a large proportion of arable land. Economic and trade activities were more frequent and convenient resulting in large population density, accelerating land urbanization. Therefore, areas with lower habitat quality tend to spread.

3.5. Analysis of Habitat Quality and Habitat Degradation in Response to Landform Relief Gradient

Terrain is an important factor affecting the spatial difference of ecological services. In order to analyze the spatial distribution of ecological services with different topography, the landform relief gradients were divided into five classes, using the natural discontinuity point method [41]. They are the first class (0–54 m), second class (24–119 m), third class (119–180 m), fourth class (180–253 m), and fifth class (more than 253 m). The area of different habitat quality under different classes of landform relief gradients in each year was counted.

Table 10 shows that the proportion of areas with moderate habitat quality and below in the first and second classes was significantly higher than that of other classes. This was mainly because the cultivated and construction land in the Wanhe Watershed was mainly concentrated in areas under relatively flat terrain. The transformation of human activities on land types affected the habitat quality. With the passage of time, its impact would have an obvious trend of expansion. The fourth and fifth classes are mainly covered by forestland and grassland with good habitat conditions, so areas with high habitat quality are mainly concentrated here. Very low habitat quality increased slightly in the first- and second-class areas, indicating that the worst habitat conditions are still deteriorating. The low habitat quality of the first class increased rapidly by 11.45% in 2010 compared with 2005, and continued to deteriorate in the following 10 years, but improved rapidly in 2025 and returned to 6.55%, which may be related to the controlled growth of construction land and rural settlements that are closely related to area reduction. From 2000 to 2025, the area of the first-grade topography will decrease by 10.55%. The area below the third grade will increase slowly, and the area below the fourth grade will increase slightly. The area of all kinds of habitat quality areas showed a general downward trend. In the fourth class, the proportion of very high habitat quality decreased by 2.18%, and the very high habitat quality also decreased slightly, indicating that the high-quality environment at higher altitudes also deteriorated.

In spatial distribution, low habitat quality areas were mainly distributed in flat terrain areas. Fourth and fifth classes are mainly covered by high and very high habitat quality. In general, with the increase in gradients, the area of the very high habitat quality increased significantly, and that of the moderate, low and very low habitat quality decreased (Figure 6).

Therefore, topography played an important role in the spatial pattern of habitat degradation and habitat quality in the study area, because topographic differences would lead to the distribution of natural landscapes and the scope and intensity of human activities. In addition, the Tan-Lu Fault zone and mountainous areas are prone to cause geological disasters and cause damage to the ecological environment [26]. Attention should be paid to the assessment and prevention of disasters. In order to protect the ecological environment according to local conditions, it is necessary to focus on the influence of terrain factors.

4. Discussion and Conclusions

4.1. Discussion

It was revealed that forestland and grassland were the main land use types, accounting for nearly 85% of the total area with a comprehensive analysis of different land use changes in the Wanhe Watershed from 2000 to 2025. Ecological restoration of grassland and forestland could effectively improve and maintain the habitat quality in watersheds due to their impacts on the local habitat quality. The habitat quality of the Wanhe Watershed was generally at a high level, but there are geographical differences in the decline. The urbanization process has accelerated, and the urban land area has continued to grow, especially the rapid expansion of construction land in parts of the eastern watershed that has led to a sharp decline in its habitat quality. The northeastern part of the watershed was close to the center of Anqing City and Sinopec Anqing Petrochemical Company, and habitat quality was low, but habitat degradation declined.. Industrial land growth slowed down, and the area of water that is considered a habitat has expanded, which implies habitat quality has a trend of improvement.

The improvement of habitat quality of some reservoirs was closely related to ecological protection measures such as returning farmland to lake, Hualiangting Reservoir, in Taihu County with functions of flood control, irrigation, and aquaculture significantly improved the habitat quality in 2020, indicating that manual intervention and protection are conducive to the improvement of habitat quality and could preserve the diversity of aquatic life. The habitat quality of Wuchang Lake in Wangjiang County was significantly higher than that of the cultivated land surrounding the lake. Habitat quality in the lake area remained almost unchanged from 2010 to 2020. However, due to the implementation of the Yangtze River Protection Plan, it was predicted that habitat quality would improve in 2025. The high habitat quality of Qingcao Lake in the eastern part from 2010 to 2015 was related to Qingcao Lake Drainage Station Project. The habitat quality degradation increased in 2015–2020, so the risk of habitat deterioration in Qingcao Lake should be prevented. The habitat degradation and habitat quality of Matanghu Reservoir and the surrounding areas in Huaining County worsened significantly. The main reason was that the rapid expansion of Shipai Town occupied part of the forestland. There were many ditches and scattered settlements between the Jintan River and the Dahe River, and the areas with low habitat quality were relatively fragmented. The southern part of the Jintan River was mainly paddy field, so the overall habitat was slightly higher than the urban land area. In 2017, the implementation of the ecological environment protection plan for the Yangtze River Economic Belt will improve the overall ecological environment. The habitat quality of the eastern plain area will be greatly improved after 2020. Due to the government’s restrictions on the spread of industrial and mining land and the protection of water areas, while the distribution of the population also tends to rationalize, it shows that the ecological environment protection planning of the Yangtze River Economic Belt will have a promising effect on environmental protection.

Due to land classification and parameter settings, there may be some errors in the interpretation and classification of the five-phase land use data. This may affect the prediction results of the model. In addition, due to different land use prediction principles of different models, the results may be slightly different, thus affecting habitat quality prediction. In addition, the InVEST model only considers the influence of the influence factors within the study area, but the Watershed edge area is also affected by outside factors, which may cause some errors in the research at the Watershed edge.

4.2. Conclusions

The expansion of urban land in the Wanhe Watershed would bring some damage to habitat quality. We should also pay attention to the protection of ecological environment while pursuing social and economic development. In this study, the land use transfer matrix, InVEST model, and CA-Markov model were used to simulate and predict the spatiotemporal changes of land use structure, habitat degradation, and habitat quality in the Wanhe Watershed in 2000, 2005, 2010, 2015, 2020, and 2025. The following conclusions were drawn:

• (1). Since 2000, the area of grassland and forestland has gradually decreased, and the area of construction land has increased significantly. Farmland and Waters decreased before 2015 and increased afterwards. Land use change mainly occurs between forestland, cultivated land, and construction land. Due to the expansion of urban land, some farmland, woodland, and grassland are occupied.

• (2). The habitat degradation increased from west to east in the Wanhe Watershed. That was closely related to the spatial difference of land cover caused by topography and human activities. The habitat quality around the river decreased significantly, and the degradation around construction land was serious.

• (3). The habitat degradation tended to be serious from 2000 to 2020, and the overall habitat quality showed a downward trend. The spread of urban and rural land and the reduction in forestland were the main reasons for the decline of habitat quality in the Wanhe Watershed.

• (4). Topography affects the distribution of the natural landscape and the area and intensity of human activities. That led to the difference and change in landscape pattern. The area of the very high habitat quality increased and that of the relatively low habitat quality decreased with the increase in landform relief gradient.

• (5). As a result of the implementation of the environmental protection policy and the ecological environment protection plan for the Yangtze River Economic Belt, the quality of the habitat will improve in 2025. However, the degree of habitat degradation is still relatively high, and it is necessary to prevent the habitat quality from declining again.

Due to the special topography of Wanhe Watershed, the habitat degradation and habitat quality show obvious east–west differences, and it is crucial to distinguish the area for planning. The degradation degree of urban land and waters is not optimistic. The waters, especially the main stream and tributaries of the Anhui River, are easily damaged by threat sources. There are many nature reserves of animals and plants in the Wanhe Watershed, and precious animals and plants are widely distributed, so strict monitoring and planning is required for conservation and restoration. In the trend of rational planning of urban land expansion, the establishment and improvement of the green and circular development mechanism of the river watershed can better promote ecological environmental protection and achieve a balance between social and ecological benefits. Based on our assessments and predictions, we can adjust ecological policies and land use planning in a timely manner, make scientific predictions and assist decision-making on regional development, and promote both economic and social development and biodiversity conservation in the Wanhe Watershed.

Author Contributions

Conceptualization H.G.; methodology C.Z.; software C.Z.; validation C.Z. and Y.B.; formal analysis P.Z.; investigation C.Z. and Y.B.; writing—original draft preparation C.Z. and Y.B.; visualization C.Z.; supervision P.Z.; project administration H.G. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflict of interest.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures and Tables

Figure 1. The map of the study area.

Figure 2. Land use change of the Wanhe Watershed from 2000 to 2025. (a): land use of 2000; (b): land use of 2005; (c): land use of 2010; (d): land use of 2015; (e): land use of 2020; (f): land use of 2025.

Figure 3. Sankey diagram of the land use transfer matrix.

Figure 4. Degradation of habitat quality in the Wanhe Watershed. ((a): degradation of habitat quality in 2000; (b): degradation of habitat quality in 2005; (c): degradation of habitat quality in 2010; (d): degradation of habitat quality in 2015; (e): degradation of habitat quality in 2020; (f): degradation of habitat quality in 2025).

Figure 5. Spatial distribution of habitat quality in the Wanhe Watershed from 2000 to 2025. ((a): habitat quality in 2000; (b): habitat quality in 2005; (c): habitat quality in 2010; (d): habitat quality in 2015; (e): habitat quality in 2020; (f): habitat quality in 2025).

Figure 5. Spatial distribution of habitat quality in the Wanhe Watershed from 2000 to 2025. ((a): habitat quality in 2000; (b): habitat quality in 2005; (c): habitat quality in 2010; (d): habitat quality in 2015; (e): habitat quality in 2020; (f): habitat quality in 2025).

Figure 6. Proportions of areas of different habitat quality under different gradients in each year.

Figure 6. Proportions of areas of different habitat quality under different gradients in each year.

Table 1

Data resource.

Data	Data Sources	Description	Format
Land use data	30 m resolution land use maps were from Data Center for Resource and Environment Science, Chinese Academy of Sciences (http://www.resdc.cn/, accessed on 12 May 2022)	Land use data was divided into 15 land use types (6 classes of the first class and 15 classes of the second class) of the Wanhe Watershed in 2000, 2005, 2010, 2015, and 2020.	TIFF
DEM data	30 m resolution DEM data was from ASTERGDEM 30m resolution digital elevation model of Geospatial Data Cloud (http://www.gscloud.cn/, accessed on 13 May 2022)	DEM was used to calculate the slope, aspect data, and landform relief gradient in 2020.	TIFF
Population data	100 m resolution population data was from WorldPop (https://www.worldpop.org/, accessed on 13 May 2022)	Population data was used for the prediction of land use.	TIFF
Road network data	Traffic land data was obtained from Open Street Map (https://download.geofabrik.de/, accessed on 10 May 2022)	Road network within Wanhe Watershed in 2000, 2005, 2010, 2015, and 2020.	TIFF
Town center data	Town center data was from Open Street Map (https://download.geofabrik.de/, accessed on 10 May 2022)	Town centers within Wanhe Watershed in 2000, 2005, 2010, 2015, and 2020.	Shapefile

Table 2

Landscape category system.

First Class	Second Class
Farmland	Paddy field, dry-farm field
Forestland	Forestland, shrub land, sparse forestland, other forestland
Grassland	Grassland
Waters	Rivers and lakes, reservoirs and ponds, beaches
Construction land	Urban land, rural settlements, industrial land, traffic land
Unused land	Barren land

Table 3

Attributes of threat data.

Threat	Threat Maximum Effective Distance (km)	Weight	Decay	Reference
Paddy field	1	0.4	exponential	Hu et al., 2022 [33]
Dry-farm land	1	0.4	exponential
Urban land	4	1	exponential	Yan et al., 2021 [35]
Rural settlements	3	0.7	exponential
Industrial land	2.5	1	exponential	Tang et al. [36]
Traffic land	2	0.7	exponential

Table 4

Landscape types and its sensitivity to each threat.

Type	Habitat Suitability	Paddy Field	Dry-Farm Land	Urban Land	Rural Settlements	Industrial Land	Traffic Land
Paddy field	0.6	0.3	0.3	0.3	0.4	0.2	0.3
Dry-farm land	0.4	0.3	0.3	0.3	0.4	0.2	0.3
Forestland	1.0	0.6	0.6	0.8	0.8	0.7	0.6
Shrub land	0.8	0.6	0.6	0.6	0.5	0.5	0.6
Sparse forestland	0.7	0.5	0.6	0.8	0.6	0.4	0.6
Other forestland	0.7	0.6	0.6	0.9	0.8	0.7	0.5
Grassland	0.9	0.4	0.4	0.6	0.5	0.3	0.3
Rivers and lakes	0.9	0.6	0.6	0.4	0.4	0.4	0.4
Reservoirs and ponds	0.7	0.6	0.6	0.4	0.4	0.4	0.4
Beaches	0.5	0.6	0.6	0.7	0.7	0.4	0.4
Urban land,	0.0	0.0	0.0	0.0.	0.0	0.0	0.0
Rural settlements	0.0	0.0	0.0	0.0	0.0	0.0	0.0
Industrial land	0.0	0.0	0.0	0.3	0.3	0.0	0.0
Barren land	0.0	0.0	0.0	0.0	0.0	0.0	0.0
Traffic land	0.0	0.0	0.0	0.0	0.0	0.0	0.0

Table 5

The result of model accuracy verification.

Year	2010	2015	2020
F1-Score	0.85	0.84	0.77
Balanced Accuracy	0.92	0.93	0.89
Kappa Coefficient	0.96	0.92	0.9
Overall Accuracy	0.97	0.94	0.93

Table 6

Area change of different types of land in the Wanhe Watershed from 2000 to 2025 (km²).

Land Use Type	2000	2005	2010	2015	2020	2025
Paddy field	2184.66	2181.35	2151.47	2143.04	2144.33	2147.02
Dry-farm land	223.94	224.06	222.92	221.78	222.01	222.17
Forestland	2524.79	2524.19	2522.25	2518.70	2518.22	2520.29
Shrub land	571.22	570.90	567.25	564.77	560.61	557.24
Sparse forestland	7.68	7.69	7.70	7.71	7.41	6.96
Other forestland	1.34	1.35	1.35	1.35	1.35	1.35
Grassland	416.41	415.76	413.87	411.68	411.40	408.66
Rivers and lakes	131.01	131.00	130.46	174.09	129.50	115.38
Reservoirs and ponds	86.74	86.86	86.37	87.94	90.82	85.33
Beaches	105.64	105.61	105.63	60.40	99.90	122.28
Industrial land	10.47	13.51	24.54	31.67	38.82	40.82
Urban land	167.81	168.56	180.32	183.75	184.04	184.22
Rural settlements	4.40	5.29	22.00	28.79	23.56	20.51
Barren land	1.22	1.23	1.23	1.85	1.40	1.14

Table 7

Land use change transfer matrix of the Wanhe Watershed from 2000 to 2025 (km²).

Land Use Type	Farmland	Forestland	Grassland	Waters	Construction Land	Unused Land
Farmland	2277.90	59.46	4.12	9.14	57.65	0.00
Forestland	62.06	3010.54	11.80	3.71	14.22	0.06
Grassland	5.85	11.30	389.26	3.34	5.48	0.19
Waters	9.12	2.80	2.36	306.50	2.47	0.00
Construction land	14.24	1.49	0.93	0.29	165.73	0.00
Unused land	0.00	0.14	0.18	0.00	0.00	0.90

Table 8

Habitat degradation in the Wanhe Watershed from 2000 to 2025.

Year	Very Slight Degradation	Slight Degradation	Moderate Degradation	Severe Degradation	Very Severe Degradation	Maximum	Average
2000	824.18	2769.74	1539.16	1142.02	162.24	0.0154	0.0629
2005	954.21	2990.89	1931.95	489.88	70.50	0.0135	0.0527
2010	760.09	2594.23	1445.44	1417.68	219.35	0.0166	0.0640
2015	751.92	2579.76	1427.53	1442.32	234.93	0.0168	0.0660
2020	692.49	2507.99	1452.71	1529.23	256.35	0.0175	0.0720
2025	740.58	2629.92	1561.69	1306.60	194.58	0.0163	0.0635

Table 9

Habitat quality grade in the Wanhe Watershed from 2000 to 2025 (km²).

Year	Very Low	Low	Moderate	High	Very High
2000	184.07	268.60	2446.47	822.12	2715.81
2005	188.52	250.87	2373.02	795.74	2826.15
2010	235.77	571.63	2323.69	854.42	2456.26
2015	250.75	628.35	2286.45	825.38	2447.17
2020	255.12	725.50	2205.69	850.80	2401.67
2025	248.91	277.24	2460.99	841.46	2604.41

Table 10

Changes in the total area of each habitat quality grade by landform relief gradients.

HQ/Landform Relief Gradient	Year	Proportions (%)
		First Class	Second Class	Third Class	Fourth Class	Fifth Class
Very low	2000	6.31	0.68	0.11	0.03	0.33
	2005	6.44	0.71	0.15	0.03	0.33
	2010	8.05	0.93	0.18	0.03	0.33
	2015	8.56	1.01	0.19	0.03	0.33
	2020	8.68	1.04	0.25	0.04	0.33
	2025	8.49	1.01	0.22	0.04	0.26
Low	2000	6.25	3.63	3.14	1.41	0.43
	2005	5.61	3.64	3.14	1.41	0.43
	2010	17.06	4.01	3.24	1.41	0.41
	2015	19.09	4.08	3.25	1.42	0.45
	2020	22.58	4.17	3.26	1.44	0.41
	2025	6.55	3.60	3.18	1.46	0.46
Moderate	2000	69.13	30.64	11.89	4.28	1.10
	2005	66.87	29.86	11.78	4.30	1.10
	2010	63.34	33.12	12.38	4.31	1.09
	2015	61.79	33.66	12.41	4.36	1.04
	2020	58.58	33.95	12.72	4.39	1.03
	2025	69.55	30.95	12.03	4.21	1.03
High	2000	11.50	13.16	12.83	15.16	15.21
	2005	11.27	12.01	12.39	15.06	15.20
	2010	8.69	17.91	15.82	16.82	15.53
	2015	7.65	17.43	16.09	16.97	15.68
	2020	7.66	18.45	16.69	17.49	15.00
	2025	11.13	14.80	13.51	15.46	15.20
Very High	2000	4.28	49.63	71.07	78.83	83.05
	2005	9.80	53.78	72.53	79.20	82.93
	2010	2.86	44.03	68.39	77.42	82.64
	2015	2.91	43.81	68.06	77.22	82.50
	2020	2.50	42.39	67.09	76.65	82.32
	2025	6.80	51.89	72.03	79.12	82.93

References

1. Li, F.; Wang, L.; Chen, Z.; Clarke, K.C.; Li, M.; Jiang, P. Extending the SLEUTH model to integrate habitat quality into urban growth simulation. J. Environ. Manag.; 2018; 217, pp. 486-498. [DOI: https://dx.doi.org/10.1016/j.jenvman.2018.03.109]

2. Song, Y.; Wang, M.; Sun, X.; Fan, Z. Quantitative assessment of the habitat quality dynamics in Yellow River Basin, China. Environ. Monit. Assess.; 2021; 193, 614. [DOI: https://dx.doi.org/10.1007/s10661-021-09404-4]

3. Ding, Q.; Chen, Y.; Bu, L.; Ye, Y. Multi-Scenario Analysis of Habitat Quality in the Yellow River Delta by Coupling FLUS with InVEST Model. Int. J. Environ. Res. Public Health; 2021; 18, 2389. [DOI: https://dx.doi.org/10.3390/ijerph18052389] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33804509]

4. Yohannes, H.; Soromessa, T.; Argaw, M.; Dewan, A. Spatio-temporal changes in habitat quality and linkage with landscape characteristics in the Beressa watershed, Blue Nile basin of Ethiopian highlands. J. Environ. Manag.; 2020; 281, 111885. [DOI: https://dx.doi.org/10.1016/j.jenvman.2020.111885] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33385905]

5. Polasky, S.; Nelson, E.; Pennington, D.; Johnson, K.A. The Impact of Land-Use Change on Ecosystem Services, Biodiversity and Returns to Landowners: A Case Study in the State of Minnesota. Environ. Resour. Econ.; 2010; 48, pp. 219-242. [DOI: https://dx.doi.org/10.1007/s10640-010-9407-0]

6. Xu, L.; Chen, S.; Xu, Y.; Li, G.; Su, W. Impacts of Land-Use Change on Habitat Quality during 1985–2015 in the Taihu Lake Basin. Sustainability; 2019; 11, 3513. [DOI: https://dx.doi.org/10.3390/su11133513]

7. Di Pirro, E.; Sallustio, L.; Capotorti, G.; Marchetti, M.; Lasserre, B. A scenario-based approach to tackle trade-offs between biodiversity conservation and land use pressure in Central Italy. Ecol. Model.; 2021; 448, 109533. [DOI: https://dx.doi.org/10.1016/j.ecolmodel.2021.109533]

8. Li, Y.; Duo, L.; Zhang, M.; Wu, Z.; Guan, Y. Assessment and Estimation of the Spatial and Temporal Evolution of Landscape Patterns and Their Impact on Habitat Quality in Nanchang, China. Land; 2021; 10, 1073. [DOI: https://dx.doi.org/10.3390/land10101073]

9. Gomes, E.; Inácio, M.; Bogdzevič, K.; Kalinauskas, M.; Karnauskaitė, D.; Pereira, P. Future scenarios impact on land use change and habitat quality in Lithuania. Environ. Res.; 2021; 197, 111101. [DOI: https://dx.doi.org/10.1016/j.envres.2021.111101]

10. Wan, X.; Yang, T.; Zhang, Q.; Wang, W.; Wang, Y. Joint effects of habitat indexes and physic-chemical factors for freshwater basin of semi-arid area on plankton integrity—A case study of the Wei River Basin, China. Ecol. Indic.; 2020; 120, 106909. [DOI: https://dx.doi.org/10.1016/j.ecolind.2020.106909]

11. Zhu, Z.; Huai, W.; Yang, Z.; Li, D.; Wang, Y. Assessing habitat suitability and habitat fragmentation for endangered Siberian cranes in Poyang Lake region, China. Ecol. Indic.; 2021; 125, 107594. [DOI: https://dx.doi.org/10.1016/j.ecolind.2021.107594]

12. Zhang, Z.; Zhang, H.; Feng, J.; Wang, Y.; Liu, K. Evaluation of Social Values for Ecosystem Services in Urban Riverfront Space Based on the SolVES Model: A Case Study of the Fenghe River, Xi’an, China. Int. J. Environ. Res. Public Health; 2021; 18, 2765. [DOI: https://dx.doi.org/10.3390/ijerph18052765] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33803264]

13. Chu, L.; Sun, T.; Wang, T.; Li, Z.; Cai, C. Evolution and Prediction of Landscape Pattern and Habitat Quality Based on CA-Markov and InVEST Model in Hubei Section of Three Gorges Reservoir Area (TGRA). Sustainability; 2018; 10, 3854. [DOI: https://dx.doi.org/10.3390/su10113854]

14. Kadaverugu, A.; Rao, C.N.; Viswanadh, G.K. Quantification of flood mitigation services by urban green spaces using InVEST model: A case study of Hyderabad city, India. Model. Earth Syst. Environ.; 2020; 7, pp. 589-602. [DOI: https://dx.doi.org/10.1007/s40808-020-00937-0]

15. Tang, F.; Fu, M.; Wang, L.; Song, W.; Yu, J.; Wu, Y. Dynamic evolution and scenario simulation of habitat quality under the impact of land-use change in the Huaihe River Economic Belt, China. PLoS ONE; 2021; 16, e0249566. [DOI: https://dx.doi.org/10.1371/journal.pone.0249566] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33819303]

16. Wu, L.; Sun, C.; Fan, F. Estimating the Characteristic Spatiotemporal Variation in Habitat Quality Using the InVEST Model—A Case Study from Guangdong–Hong Kong–Macao Greater Bay Area. Remote Sens.; 2021; 13, 1008. [DOI: https://dx.doi.org/10.3390/rs13051008]

17. Tang, F.; Fu, M.; Wang, L.; Zhang, P. Land-use change in Changli County, China: Predicting its spatiotemporal evolution in habitat quality. Ecol. Indic.; 2020; 117, 106719. [DOI: https://dx.doi.org/10.1016/j.ecolind.2020.106719]

18. Xu, C.; Gong, J.; Yan, L.; Yan, L.; Gao, B. Spatial Distribution Characteristics of Typical Ecosystem Services Based on Terrain Gradients of Bailongjiang Watershed in Gansu. Acta Ecol. Sin.; 2020; 40, pp. 4291-4301.

19. Bai, Y.; Li, S.; Zuo, Q.; Zhang, W. Study on Dynamic Change of Land Use in Qingzhen City Based on GIS Technology and CA-Markov Model. E3S Web Conf.; 2021; 271, 02017. [DOI: https://dx.doi.org/10.1051/e3sconf/202127102017]

20. Gao, X.; Wang, J.; Li, C.; Shen, W.; Song, Z.; Nie, C.; Zhang, X. Land use change simulation and spatial analysis of ecosystem service value in Shijiazhuang under multi-scenarios. Environ. Sci. Pollut. Res.; 2021; 28, pp. 31043-31058. [DOI: https://dx.doi.org/10.1007/s11356-021-12826-9] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33598839]

21. Babbar, D.; Areendran, G.; Sahana, M.; Sarma, K.; Raj, K.; Sivadas, A. Assessment and prediction of carbon sequestration using Markov chain and InVEST model in Sariska Tiger Reserve, India. J. Clean. Prod.; 2020; 278, 123333. [DOI: https://dx.doi.org/10.1016/j.jclepro.2020.123333]

22. Balocchi, F.; Flores, N.; Arumí, J.L.; Iroumé, A.; White, D.A.; Silberstein, R.P.; de Arellano, P.R. Comparison of streamflow recession between plantations and native forests in small catchments in Central-Southern Chile. Hydrol. Process.; 2021; 35, 14182. [DOI: https://dx.doi.org/10.1002/hyp.14182]

23. Qiu, S.; Yan, Y.-J. New species and new records of Tinodes Curtis (Trichoptera: Psychomyiidae) from Dabie Mountains, East-central China. J. Asia-Pac. Èntomol.; 2018; 21, pp. 394-401. [DOI: https://dx.doi.org/10.1016/j.aspen.2018.01.012]

24. Zhou, H.Q.; Ren, Q.Y. Strategies for E-Commerce Development of Local Agricultural Products—A Case Study in Dabie Mountain. Appl. Mech. Mater.; 2014; 651–653, pp. 1717-1720. [DOI: https://dx.doi.org/10.4028/www.scientific.net/AMM.651-653.1717]

25. Tian, S.; Xu, J.; Wang, Y. Human infrastructure development drives decline in suitable habitat for Reeves’s pheasant in the Dabie Mountains in the last 20 years. Glob. Ecol. Conserv.; 2020; 22, e00940. [DOI: https://dx.doi.org/10.1016/j.gecco.2020.e00940]

26. Qin, H.; Hao, W.; Deng, C.; Zhao, P.; Shen, Z.; Han, F.; He, H.; Pan, Y.; Zhu, R. Sinistral displacement along the Tan–Lu Fault during the Cretaceous induced by Paleo-Pacific subduction: Constraints from new paleomagnetic and U–Pb geochronological data. J. Southeast Asian Earth Sci.; 2022; 237, 105362. [DOI: https://dx.doi.org/10.1016/j.jseaes.2022.105362]

27. Han, H.; Gao, T.; Yi, H.; Yang, M.; Yan, X.; Ren, G.; Yang, J. Extraction of Relief Amplitude Based on Change Point Method: A Case Study on the Tibetan Plateau. Sci. Geogr. Sin.; 2012; 32, pp. 101-104. [DOI: https://dx.doi.org/10.13249/j.cnki.sgs.2012.01.013]

28. Mansour, S.; Al-Belushi, M.; Al-Awadhi, T. Monitoring land use and land cover changes in the mountainous cities of Oman using GIS and CA-Markov modelling techniques. Land Use Policy; 2019; 91, 104414. [DOI: https://dx.doi.org/10.1016/j.landusepol.2019.104414]

29. Fu, X.; Wang, X.; Yang, Y.J. Deriving suitability factors for CA-Markov land use simulation model based on local historical data. J. Environ. Manag.; 2018; 206, pp. 10-19. [DOI: https://dx.doi.org/10.1016/j.jenvman.2017.10.012] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29055845]

30. Zhang, E.; Zhang, Y. F-Measure. Encyclopedia of Database Systems; Liu, L.; Özsu, M.T. Springer: Boston, MA, USA, 2009; 1147. ISBN 978-0-387-35544-3

31. Zhang, Z.; Hu, B.; Jiang, W.; Qiu, H. Identification and scenario prediction of degree of wetland damage in Guangxi based on the CA-Markov model. Ecol. Indic.; 2021; 127, 107764. [DOI: https://dx.doi.org/10.1016/j.ecolind.2021.107764]

32. Liu, S.; Liao, Q.; Xiao, M.; Zhao, D.; Huang, C. Spatial and Temporal Variations of Habitat Quality and Its Response of Landscape Dynamic in the Three Gorges Reservoir Area, China. Int. J. Environ. Res. Public Health; 2022; 19, 3594. [DOI: https://dx.doi.org/10.3390/ijerph19063594]

33. Hu, J.; Zhang, J.; Li, Y. Exploring the spatial and temporal driving mechanisms of landscape patterns on habitat quality in a city undergoing rapid urbanization based on GTWR and MGWR: The case of Nanjing, China. Ecol. Indic.; 2022; 143, 109333. [DOI: https://dx.doi.org/10.1016/j.ecolind.2022.109333]

34. Yang, Y. Evolution of habitat quality and association with land-use changes in mountainous areas: A case study of the Taihang Mountains in Hebei Province, China. Ecol. Indic.; 2021; 129, 107967. [DOI: https://dx.doi.org/10.1016/j.ecolind.2021.107967]

35. Li, Y.; Duo, L.; Zhang, M.; Yang, J.; Guo, X. Habitat quality assessment of mining cities based on InVEST model—A case study of Yanshan County, Jiangxi Province. Int. J. Coal Sci. Technol.; 2022; 9, 28. [DOI: https://dx.doi.org/10.1007/s40789-022-00498-w]

36. Tang, F.; Wang, L.; Guo, Y.; Fu, M.; Huang, N.; Duan, W.; Luo, M.; Zhang, J.; Li, W.; Song, W. Spatio-temporal variation and coupling coordination relationship between urbanisation and habitat quality in the Grand Canal, China. Land Use Policy; 2022; 117, 106119. [DOI: https://dx.doi.org/10.1016/j.landusepol.2022.106119]

37. Bin Zhao, B.; Xu, X.Y.; Ou, W.H. Effect Analysis of Statistical Unit on Relief Amplitude in Chongqing, China. Adv. Mater. Res.; 2014; 1065–1069, pp. 2273-2277. [DOI: https://dx.doi.org/10.4028/www.scientific.net/AMR.1065-1069.2273]

38. Yuan, G.; Li, J.; Liu, X.; Yang, Z. Weakly supervised road network extraction for remote sensing image based scribble annotation and adversarial learning. J. King Saud Univ.-Comput. Inf. Sci.; 2022; 34, 9. [DOI: https://dx.doi.org/10.1016/j.jksuci.2022.05.020]

39. Feinstein, A.R.; Cicchetti, D.V. High agreement but low Kappa: I. the problems of two paradoxes. J. Clin. Epidemiol.; 1990; 43, pp. 543-549. [DOI: https://dx.doi.org/10.1016/0895-4356(90)90158-L]

40. Barbato, M.P.; Napoletano, P.; Piccoli, F.; Schettini, R. Unsupervised segmentation of hyperspectral remote sensing images with superpixels. Remote Sens. Appl. Soc. Environ.; 2022; 28, 100823. [DOI: https://dx.doi.org/10.1016/j.rsase.2022.100823]

41. Tang, G.; Song, J. Comparison of Slope Classification Methods in Slope Mapping from DEM. J. Soil Water Conserv.; 2006; 20, pp. 157-160.