Temporal and Spatial Variation in Habitat Quality in Guangxi Based on PLUS-InVEST Model

Abstract

Despite Guangxi’s unique ecological diversity and its important role in land-based ecological security and conservation, research on the assessment and prediction of its habitat quality under the influences of rapid urbanization and environmental pressures remains limited. This study systematically analyzes the spatial and temporal dynamics of land use and habitat quality in Guangxi from 2000 to 2020 using the PLUS-InVEST model and simulates future scenarios for 2030. These scenarios include the Natural Development (ND) scenario, Urban Development (UD) scenario, and Cropland and Ecological Protection (CE) scenario. The results indicate the following: (1) Over the past two decades, rapid urban and construction land expansions in Guangxi intensified their negative impact on habitat degradation. Additionally, the disproportionate change between rural settlement land and rural population warrants attention. (2) Although ecological restoration measures have played a positive role in mitigating habitat degradation, their effects have been insufficient to counterbalance the negative impacts of construction land expansion, highlighting the need for balanced land use planning and urbanization policies. (3) The expansion of rural residential areas had a greater impact on regional habitat quality degradation than urban and infrastructure expansion. Moderate urbanization may contribute to habitat quality improvement. (4) The CE scenario shows the most significant improvement in habitat quality (an increase of 0.13%), followed by the UD scenario, which alleviates habitat degradation by reducing pressure on rural land. In contrast, the ND scenario predicts further declines in habitat quality. Furthermore, land use planning, restoration measures, and sustainable development policies are key factors influencing habitat quality changes. These findings emphasize the importance of integrating land use strategies with ecological restoration measures to balance economic growth and biodiversity conservation, especially in rapidly urbanizing regions.

1. Introduction

Habitat quality refers to the ability of a habitat to consistently provide suitable conditions for the survival of individuals or populations [1]. The quality of a habitat reflects the ecosystem service value and biodiversity level of a region [2]. The habitat quality index, as a key metric for assessing environmental conditions, is crucial for evaluating the suitability of human living spaces [3]. Land serves as the carrier for various ecological systems on Earth’s surface [4]. Changes in land use and land cover (LULC) are a direct reflection of how human activities interact with the environment [5,6]. Alterations in land use structure and intensity can profoundly impact the exchange processes of materials and energy between habitat patches, influencing regional climate, water quality, vegetation, and the structure and function of ecosystems [7,8,9]. These changes can, in turn, lead to shifts in regional habitat quality [10,11]. Therefore, from the perspective of land use change, exploring the spatiotemporal evolution of habitat quality and its ecological responses to land use changes, as well as analyzing their spatial distribution characteristics, influencing factors, and driving mechanisms, plays a critical role in biodiversity conservation and the maintenance of ecosystem balance [10].

The impact of land use transformation on the ecological environment is a critical aspect of global change research [12]. Currently, balancing the need for land development to meet socio-economic demands with the protection of the ecological environment has become an increasingly urgent issue. As a result, there is growing scholarly interest in studying how land use change affects the environment. The Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) model [5,13], which evaluates the impact of land use changes on ecosystem services, has gained widespread applications in land planning and ecological conservation decision-making. Compared to other models such as Ecotourism Sustainability Maximization (ESM) [14], SLEUTH (Slope, Land use, Exclusion, Urbanization, Transportation, and Hillshade) [15], and Dynamic Conversion of Land Use and its Effects (Dyna-CLUE) [16], InVEST places greater emphasis on assessing ecosystem services and has been successfully applied in evaluating the impact of urban expansion on habitat quality [17] and the factors driving changes in habitat quality in small watersheds [18]. Additionally, studies have examined landscape sustainability in rural areas [19] and provincial-level quality changes and their influencing factors in China [20]. Despite these advancements, the majority of existing research has focused on the impacts of land use changes in urban, rural, and small watershed environments. In contrast, studies assessing the extent to which changes in built-up land use—such as urban land, rural settlement areas, and transportation and industrial land—affect habitat quality are relatively limited. This gap necessitates further attention. Additionally, while common land use simulation models, such as Conversion of Land Use and its Effects at different Spatial Scales (CLUE-S) [21], Cellular Automata Models and Markov models ( CA-Markov) [22], and Future Land Use Simulation (FLUS) [23], have been widely employed for predictive modeling, there is limited research that has integrated the PLUS-InVEST model to assess the impact of various land use transitions on regional habitat quality, as well as to project future trends in habitat quality evolution.

Guangxi, as a crucial ecological barrier and biodiversity conservation area in China, plays a key role in the construction of ecological civilization. It is endowed with rich natural resources and a unique ecological environment. However, it also faces significant challenges in environmental protection. This study utilizes the PLUS-InVEST model, combined with dynamic spatiotemporal assessments, to reveal the spatiotemporal evolution patterns of habitat quality in Guangxi from 2000 to 2030 and the factors influencing these changes. The findings provide a scientific reference for local governments in balancing economic development with ecological protection, enabling the formulation of more scientific and rational land management and resource utilization strategies.

2. Materials and Methods

2.1. Study Area

Guangxi, located in the southern border region of China and sharing borders with Southeast Asian countries, is characterized by significant natural geographical diversity, with an intricate mix of mountains and plains, as well as relatively fragile ecosystems (Figure 1). This study not only examines changes in the ecological environment quality of Guangxi over the past two decades but also focuses on the variation in habitat quality within the karst rocky desertification prevention zone and areas undergoing rapid urbanization. The karst rocky desertification prevention zone is marked by economic underdevelopment and ecological vulnerability, making it a primary target for policies such as returning farmland to forests and grasslands, as well as ecological migration. In contrast, Nanning, the capital of Guangxi, is situated in the central region of the province and features diverse topography, primarily consisting of hills and basins. The urbanization rate of Nanning increased from 52.6% in 2010 to 71.4% in 2023, marking a growth of 18.8 percentage points, the highest among all cities in Guangxi. In addition, Beihai is a coastal city characterized by flat topography, predominantly formed by coastal alluvial plains with saline–alkali soils. Its humid climate fosters a unique ecological environment. As one of China’s 14 coastal open cities, Beihai’s urbanization rate reached 60.47% by the end of 2023, reflecting an increase of 11.86 percentage points since 2010. Both cities exemplify rapid urbanization trends and serve as key representatives of Guangxi’s economic development and urban construction advancements.

2.2. Data

This investigation incorporated an array of data sources, encompassing land use/land cover (LULC) datasets, natural environmental parameters, and socio-economic variables (

Table 1. The data source and descriptions in this study.

Data Types	Data Name	Accuracy	Source and Description
LULC data	LULC in 2000, 2010, 2015, and 2020	30 m	Resource and Environmental Science Data Center of Chinese Academy of Sciences (https://www.resdc.cn, accessed on 10 July 2024) [24]. By using the “Mask” tool in ArcGIS 10.8, the remote sensing raster data for the Guangxi area were clipped out.
Socio-economic factors	Population density	1 km	Resource and Environmental Science Data Center of Chinese Academy of Sciences (https://www.resdc.cn, accessed on 10 July 2024). By using the ’Resample’ tool in ArcGIS 10.8, the data were transformed into a resolution of 30 × 30 m.
	Gross domestic product (GDP)	1 km
	Distance from railway, expressway, national highway, provincial highway, county roads, or county center	30 m	China National Bureau of Geomatics Information System (CNBGIS: https://www.webmap.cn, accessed on 15 July 2024) [25]. Includes the vector data of China’s road network, administrative centers, and waterways for the year 2020. Calculated using the Buffer Analysis tool in ArcGIS10.8
Natural factors	Distance from water area	-
	Karst rocky desertification areas in Guangxi	-	Guangxi Zhuang Autonomous Region Meteorological Disaster Mitigation. Institute Remote Sensing Basic Database (http://gx.cma.gov.cn, accessed on 15 August 2024) [26].
	Elevation	30 m	Resource and Environmental Science Data Center of Chinese Academy of Sciences (https://www.resdc.cn, accessed on 15 July 2024). The elevation and slope were extracted from the DEM; meteorological data were sourced from the annual spatial interpolation dataset of Chinese meteorological elements. Soil type data were resampled to a 30 m × 30 m resolution using ArcGIS 10.8.
	Slope	30 m
	Soil type	1 km
	Normalized difference vegetation index (NDVI)	250 m
	Mean annual temperature	1 km
	Mean annual precipitation	1 km

). The LULC data for the years 2000, 2010, and 2020 were procured from the “China Multi-Period Land Use Remote Sensing Monitoring Dataset (CNLUCC)” accessible at https://www.resdc.cn/DOI/DOI.aspx?DOIID=54 (accessed on 10 July 2024). Employing ArcGIS 10.8 software, the LULC categories were systematically reclassified into eight distinct types: cultivated land, forest, grass, water, urban land, rural residential land, other construction land, and bare land. In an effort to delineate the differential impacts of urban and rural areas on habitat quality, this study retained the secondary land use categories of residential areas, which include urban land, rural residential land, and other construction land (including transportation land, mining land, and industrial land).

2.3. Methods

This study investigated the impacts of land use changes driven by human activities on habitat quality in Guangxi over the past two decades and forecasts the habitat pattern for 2030. The dynamics of land use transitions were analyzed using the land use transfer matrix method, while the InVEST model was applied to evaluate the ecological impacts of these transitions. Furthermore, the PLUS model was employed to predict land use patterns for 2030. In consideration of the multifaceted factors influencing land use change, and adhering to the principles of data availability and applicability, we selected a range of natural geographic and socio-economic factors. The natural geographic factors included the (a) soil type, (b) NDVI, (c) mean annual precipitation, (d) mean annual temperature, (e) elevation, and (f) slope. The socio-economic factors comprised (g) the GDP and (i–o) distances from railways, expressways, national highways, provincial highways, county roads, county centers, and water bodies. Based on the simulated land use data, this study assessed the potential effects of these changes on habitat quality. The research’s conceptual framework and methodology are illustrated in Figure 2.

2.3.1. Land Use Transfer Matrix

The land use transition matrix is an application of the Markov model in land use change studies [27]. The land use transition matrix, represented in matrix form, quantifies the conversion relationships between land cover types in two distinct periods, providing a comprehensive depiction of the values and direction of transitions among land cover types in a given region. This method involves the overlay analysis of land use type raster data from two different time periods to derive the land use transition matrix for a specified time interval [28]. In this study, we used this matrix to investigate the changes in different land use types in Guangxi between 2000 and 2020. The calculation of the land use transition matrix in this research followed the methodologies established in previous studies [29,30,31].

2.3.2. Multi-Scenario Simulation

The PLUS model, developed by the China University of Geosciences (Wuhan), is a land use simulation system based on a cellular automata framework that operates at the patch scale [32]. By integrating rule mining and a multi-type random seed mechanism, the model identifies the drivers of urban expansion and predicts land use evolution at the patch level [33]. Due to its integration of land use change drivers, the PLUS model has been widely applied to forecast future land use changes, demonstrating practicality and versatility in various fields [17,34]. In this study, 30 m resolution land use raster data for Guangxi from 2015 and 2020 were used, with the random forest algorithm employed to calculate transition probabilities between land use types. Land use patterns for 2030 were then simulated under three different scenarios.

(1)

Land Expansion Analysis Strategy (LEAS)

The LEAS is an analytical methodology that examines land use data from two different dates. By analyzing the growth patches of various changing land use types, it identifies the patterns of land use change over a specific time interval, allowing for the characterization of land use changes. This approach employs the random forest classification (RFC) algorithm to investigate the relationships between the growth of different land use types and various driving factors, enabling the estimation of development probabilities for each land use type. The parameter settings for the LEAS included the number of decision trees being set to 20, with a default sampling rate of 0.01. The parameter mTry was set to 15, and the number of parallel threads was configured to 1. The calculation formula is as follows:

$$P_{i,k\left(X\right)}^d={\displaystyle \frac{{\displaystyle \sum _{n=1}^{M}I}\left[h_n\left(X\right)=d\right]}{M}}$$

(1)

In Equation (1), X represents a vector composed of driving factors; M denotes the number of decision trees; and d can take values of either 0 or 1, where 1 indicates that other land use types can transition to the land use type k, and 0 signifies that they cannot transition to the type k. The term $h_n\left(X\right)$ refers to the predicted land use type calculated when the number of decision trees is n; $I\left[h_n\left(X\right)=d\right]$ is the indicator function for the decision trees; and $P_{i,k\left(X\right)}^d$ represents the probability of growth for the land use type k for the spatial unit i.

(2)

Cellular Automata Model Based on Multi-Class Random Patches (CARS)

CARS is a land use simulation model grounded in the principles of cellular automata. It primarily simulates land use distribution patterns by estimating the development probabilities for various land use types, functioning in a scenario-driven manner. The parameter settings for CARS include a default neighborhood size of 3, a thread count of 1, a decay threshold coefficient of 0.5, a diffusion coefficient of 0.1, and a random patch seed probability of 0.0001. The total conversion probability for the land use type k, represented as $\left(P_{0,i,k}^{d=1,t}\right)$, is calculated using the following formula:

$$P_{0,i,k}^{d=1,t}=P_{i,k}^{d=1}\times {\mathrm{\Omega }}_{i,k}^t\times D_k^t$$

(2)

In Equation (2), $P_{i,k}^{d=1}$ represents the growth probability of the land use type k in the cell i; ${\mathrm{\Omega }}_{i,k}^t$ denotes the neighborhood effect of the cell i; and $D_k^t$ reflects the influence of future demand for the land use type k. The calculation formula is as follows:

$$\begin{array}{l}{\mathrm{\Omega }}_{i,k}^t={\displaystyle \frac{\mathrm{con}\left(c_i^{t-1}=k\right)}{n\times n-1}}\times w_k,\hfill \end{array}$$

(3)

$$D_k^t=\left\{\begin{array}{l}{D}_k^{t-1}\hspace{1em}\left(\left|G_k^{t-1}\right|\le \left|G_k^{t-2}\right|\right)\hfill \\ D_k^{t-1}\times {\displaystyle \frac{G_k^{t-2}}{G_k^{t-1}}}\left(G_k^{t-1}<G_k^{t-2}<0\right)\hfill \\ D_k^{t-1}\times {\displaystyle \frac{G_k^{t-1}}{G_k^{t-2}}}\left(0<G_k^{t-2}<G_k^{t-1}\right)\hfill \end{array}\right.$$

(4)

In Equations (3) and (4), con represents the total number of grid cells occupied by the land use type k in the n × n window during the last iteration; $w_k$ denotes the weight between different land use types, with a default value of 1; and $G_k^{t-1}$ and $G_k^{t-2}$ indicate the current demand and future demand differences for the land use type k during the t − 1 and t − 2 iterations, respectively.

(3)

Accuracy Validation

To validate the simulation accuracy of the PLUS model, land use data from 2010 and 2015 were utilized. The projected land use scales for various types in Guangxi for 2020, generated using the Markov model, served as the basis for simulating land use changes under a comprehensive spatial optimization scenario. The results were then compared to the actual values for 2020, allowing for the calculation of overall accuracy and the Kappa coefficient. The calculation formulas is as follows:

$$\mathrm{Kappa}={\displaystyle \frac{P_0{-P}_c}{P_p{-P}_c}}$$

(5)

In the equation, $P_0$ represents the proportion of correctly simulated results; $P_c$ denotes the expected proportion of correct simulation results under random conditions; and $P_p$ indicates the proportion of correct results under ideal conditions. Generally, a Kappa value greater than 0.75 signifies a high level of consistency, a range between 0.4 and 0.75 indicates moderate consistency, while a value lower than 0.4 reflects poor consistency and, consequently, lower simulation accuracy.

(4)

Scenario Simulation Parameters

Natural Development scenario (ND): The ND scenario extends the land use change patterns observed from 2015 to 2020. It does not take development policy requirements into account and uses linear regression to calculate the demand for each land transition type in 2030.

Urban Development scenario (UD): As Guangxi is an underdeveloped region in China, urbanization and economic growth are expected to accelerate over the coming years. In the Urban Development scenario, the transfer rate of cultivated land, forest land, and grassland to urban construction land is increased by 5%, and the probability of rural residential and other construction land transitioning to urban land is increased by 10%. Additionally, the transfer probability of construction land to forest land, grassland, and other land use types is reduced by 10%, and the transfer probability to rural residential land is reduced by 20%. The probability of cultivated land transitioning to rural residential land is decreased by 10%.

Cropland and Ecological Protection scenario (CE): Cropland serves as the foundation of agricultural production and is critical for ensuring national food and sugar security. Guangxi, as China’s largest sugarcane production base and a major agricultural region, plays a vital role in the nation’s agricultural development. Simultaneously, Guangxi is a crucial ecological barrier and biodiversity hotspot in southern China, making ecological protection a key focus for future decades. Based on Guangxi’s 14th Five-Year Plan and 2035 Vision Goals for cropland protection and ecological civilization, this scenario was simulated with the following assumptions: The probability of cultivated land transitioning to water bodies and urban land is reduced by 30%, to bare land by 20%, and to grassland and other construction land by 10%. The probability of cultivated land transitioning to rural residential land is reduced by 5%. The probability of forest land transitioning to cultivated land and grassland is reduced by 10%, to urban land and bare land by 20%, to other construction land by 15%, and to rural residential land by 5%. The probability of rural residential land transitioning to forest land and grassland is increased by 5%. The probability of bare land transitioning to forest land and grassland is increased by 20%. The setup of the multi-scenario transition cost matrix is shown in

Table 2. Multi-scenario LULC transfer cost matrix setting.

LUCC Types	ND Scenario								UD Scenario								CE Scenario
	Ⅰ	Ⅱ	Ⅲ	Ⅳ	Ⅴ	Ⅵ	Ⅶ	Ⅷ	Ⅰ	Ⅱ	Ⅲ	Ⅳ	Ⅴ	Ⅵ	Ⅶ	Ⅷ	Ⅰ	Ⅱ	Ⅲ	Ⅳ	Ⅴ	Ⅵ	Ⅶ	Ⅷ
Ⅰ	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	0	0	0	0	0	0
Ⅱ	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	0	1	0	0	0	0	0	0
Ⅲ	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	0	0	0	0	0
Ⅳ	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Ⅴ	0	0	0	0	1	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	1	0	0	0
Ⅵ	1	1	1	1	1	1	1	0	1	1	1	1	1	1	1	0	1	1	1	1	1	1	1	0
Ⅶ	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1
Ⅷ	1	1	1	1	0	0	0	1	1	1	1	1	0	0	0	1	0	1	1	0	0	0	0	1

Notes: Classes I, II, III, IV, V, VI, VII, and VIII represent eight types of land use: cultivated land, forest, grass, water, urban land, rural residential land, other construction land, and bare land, respectively. The value of 1 indicates that a land class is subject to conversion, while a value of 0 indicates that conversion is not permitted.

.

2.3.3. Habitat Quality Evaluation

The InVEST model, developed by Stanford University, is designed to assess changes in the quantity and value of ecosystem services, enabling spatially explicit evaluations of ecosystem services’ function and value. The Habitat Quality module evaluates the impact of human activities on the environment by identifying potential threats to habitat quality and assessing the sensitivity of different land use types to these threats. Based on previous research, this study identified cultivated land, urban land, rural residential land, other construction land, and bare land as the primary threats to habitat quality [35,36] (

Table 3. Threat factor parameters.

Threat Factors	MAX_DIST (km)	Weight	Spatial Attenuation Types
Cultivated land	2	0.3	Linear
Urban land	5	1	Exponential
Rural residential land	2	0.5	Exponential
Other construction land	3	0.6	Exponential
Bare land	1	0.3	Linear

). Taking into account the variations in natural and cultural factors across different regions, we referenced parameter-setting methods from relevant studies [37,38] and conducted multiple rounds of validation and adjustments based on the remote sensing data of land use in Guangxi. This rigorous process ultimately allowed us to determine the sensitivity of each land use type to these threats and its habitat adaptability (

Table 4. Sensitivity parameters of different land types to habitat threat factors.

Land Use Types	Habitat Suitability	Threats Factors
		Cultivated Land	Urban Land	Rural Residential Land	Other Construction Land	Unused Land
Cultivated land	0.5	0	0.8	0.6	0.5	0.5
Forest	1	0.6	0.7	0.6	0.7	0.2
Grass	0.8	0.7	0.5	0.5	0.5	0.7
Water	0.75	0.2	0.3	0.3	0.3	0.5
Urban land	0	0	0	0	0	0
Rural residential land	0	0	0.3	0	0	0
Other construction land	0	0	0	0	0	0
Bare land	0.2	0.2	0.3	0.2	0.2	0

). The degree of habitat degradation can be calculated using the following formula:

$$D_{xj}={\displaystyle \sum _{r=1}^r{\displaystyle \sum _{y=1}^y\left(\frac{w_r}{{\displaystyle \sum _{n=1}^n{w}_r}}\right)}}{r}_y{\beta }_x{S}_{jr}\ast \left\{\begin{array}{c}\left(1-{\displaystyle \frac{d_{xy}}{d_{rmax}}}\right)\\ \exp \left(-{\displaystyle \frac{2.99d_{xy}}{d_{rmax}}}\right)\end{array}\right.$$

(6)

where $D_{xj}$ is the total threat level in the grid cell x with the land type j; r is the number of threat factors; y is the set of grid cells on r’s raster map; $w_r$ is the impact weight of the threat r; ${\beta }_x$ is the accessible grid cell x; $S_{jr}$ is the relative sensitivity of the land type j to the threat factor r; $d_{xy}$ is the distance between the pixel x and pixel y; and $d_{rmax}$ is the maximum impact distance of the threat r originated in the pixel y. The value of $D_{xj}$ is between 0 and 1, where higher values indicate greater habitat degradation [39].

Habitat quality was calculated by the following formula:

$$Q_{xj}=H_j\left(1-D_{xj}^2/\left(D_{xj}^2+k^2\right)\right.$$

(7)

where $Q_{xj}$ is the habitat quality of the cell x in the land type j; $H_j$ is the habitat suitability of the land use type j; and k is the half-saturation constant, usually set at k = 0.05 in the biodiversity model [40]. The habitat quality was quantified on a scale from 0 to 1, where higher values denoted superior habitat quality, based on insights from a prior study [13].

3. Results

3.1. Spatiotemporal Pattern of Land Use Changes in Guangxi from 2000 to 2020

Based on the land use classification system adopted in this study, the primary land use types in Guangxi, ranked by area from largest to smallest, are forest, cultivated land, grass, water, rural residential land, other construction land, and bare land. The land use transfer results from 2000 to 2020 are illustrated in

Table 5. Land use transfer matrix in Guangxi from 2000 to 2020. Unit: km².

Land Use Types	2000–2010
	Cultivated Land	Forest	Grass	Water	Urban Land	Rural Residential Land	Other Construction Land	Bare Land	Total
Cultivated land	31,448.97	14,879.65	2320.97	842.70	210.39	1916.69	161.32	4.01	51,784.69
Forest	14,684.18	132,496.52	6608.37	1125.20	68.91	480.45	97.72	11.02	155,572.37
Grass	2460.29	6710.41	11,277.30	249.56	27.03	113.17	20.17	2.97	20,860.89
Water	868.44	1031.49	206.11	1302.40	51.00	97.20	15.21	1.99	3573.83
Urban land	113.88	35.05	7.04	31.19	619.51	3.06	5.97	0.00	815.70
Rural residential land	1977.58	499.75	104.70	95.05	32.05	732.35	4.96	1.99	3448.43
Other construction land	72.95	32.94	7.93	7.93	13.93	4.99	124.90	0	265.57
Bare land	7.04	4.96	4.93	3.09	0	0.98	0	15.98	36.98
Total	51,633.33	155,690.77	20,537.34	3657.12	1022.82	3348.88	430.25	37.96	236,358.47
Land Use Types	2010–2020
	Cultivated Land	Forest	Grass	Water	Urban Land	Rural Residential Land	Other Construction Land	Bare Land	Total
Cultivated land	28,815.25	16,171.05	2634.62	949.76	360.75	2086.23	612.60	4.08	51,633.33
Forest	15,785.74	130,341.23	7176.77	1236.00	92.70	570.03	479.20	7.09	155,690.77
Grass	2435.15	7279.83	10,340.94	243.18	29.97	104.67	99.62	3.98	20,537.34
Water	834.83	1176.50	246.76	1189.82	36.24	110.91	58.96	3.09	3657.12
Urban land	91.99	56.05	17.05	51.24	777.44	5.97	23.08	0	1022.82
Rural residential land	1917.64	567.65	131.20	93.21	33.92	572.19	33.06	0	3348.88
Other construction land	87.92	57.95	10.07	19.19	34.87	16.04	204.21	0	430.25
Bare land	8.02	3.00	4.00	3.00	0	1.01	0	18.93	37.96
Total	49,976.55	155,659.27	20,557.41	3785.41	1365.88	3467.05	1510.72	37.17	236,358.47

and Figure 3. In terms of the total area and structural composition, forest land accounted for the largest proportion of land use in Guangxi during the period, representing approximately 65% of the total area. This was followed by cultivated land and grassland, which constituted approximately 21% and 8.8% of the total area, respectively. Artificial construction land, including urban land, rural residential land, and other construction land, occupied a minimal proportion. Although its extent increased over the past 20 years, it remained below 3% of the total territorial area. Among the construction land types, rural residential land occupied a larger area compared to urban, transportation, and industrial or mining land, indicating a relatively low level of urbanization in Guangxi during this period.

Over the past 20 years, various land use types in Guangxi underwent mutual conversions (Table 5), with simultaneous increases and decreases observed across different categories. Between 2000 and 2010, urban land and other construction land (including industrial and transportation infrastructure land) experienced the most significant expansion among all land use types, growing by 25.39% and 62.01%, respectively. However, this expansion was primarily at the expense of cultivated land, which contributed to 52.17% and 52.83% of the net land increases for urban and other construction land, respectively. Conversely, during this period, cultivated land, grassland, and rural residential land experienced varying degrees of reduction, decreasing by 0.29%, 1.55%, and 2.89%, respectively.

From 2010 to 2020, the growth rate of other construction land further accelerated (from 62.01% to 238.13%), reflecting rapid development in industrial and transportation infrastructure in Guangxi during this period. Urban land also maintained steady expansion, with growth rates increasing from 25.39% to 39.08%. Notably, rural residential land transitioned from negative growth in the previous decade to a 3.43% increase, indicating the success of rural revitalization and poverty alleviation strategies during 2010–2020. Improvements in rural infrastructure and increased farmer incomes likely contributed to this expansion. However, it is critical to recognize that the expansion of rural residential land often came at the cost of cultivated land. Over the past decade, Guangxi’s rural population decreased by 4,652,106 (a 14.18% reduction), while rural residential land increased, potentially contradicting sustainable development principles. This period also witnessed a continuous decline in cultivated land (−0.29% to −3.21%), further supporting this observation. Additionally, bare land decreased by 1.98%, while grassland grew by 0.4%, suggesting effective desertification control measures in Guangxi.

3.2. Spatio-Temporal Pattern of Habitat Quality in Guangxi from 2000 to 2020

Over the past two decades, the overall habitat quality in Guangxi showed a continuous declining trend, primarily driven by the expansion of construction land, which outpaced the growth of forest and grassland. We calculated the spatial raster data of the habitat quality index using the InVEST model (with a resolution of 30 m × 30 m). We employed an equal interval method to classify the habitat quality in Guangxi into five distinct categories: very low (0–0.2] (Level 1), low (0.2–0.4] (Level 2), medium (0.4–0.6] (Level 3), high (0.6–0.8] (Level 4), and very high (0.8–1] (Level 5). Lv1 and Lv2 corresponded predominantly to construction land and bare land (including urban areas, rural residential areas, transportation facilities, industrial land, and rocky areas), Lv3 and Lv4 were primarily cultivated land and grassland, and Lv5 represented forest and water bodies (Figure 3a–c).

Between 2000 and 2010, the average habitat quality index in Guangxi decreased from 0.7597 to 0.7557, a reduction of 0.0040. The spatial extent of Lv1 and Lv2 expanded by 0.39% and 0.29%, respectively, while the extent of Lv3 to Lv5 showed a decreasing trend, with reductions ranging from 0.03% to 0.39%. From 2010 to 2020, the average habitat quality index further declined from 0.7557 to 0.7382, a decrease of 2.32%. During this period, the proportions of Lv1, Lv2, and Lv4 continued to increase, whereas Lv3 and Lv5 experienced further reductions. Spatially, Lv1 and Lv2 primarily expanded outward from urban centers (Figure 3a–c). Temporally and spatially, the most significant changes in habitat quality occurred in the urban peripheries, with habitat quality index changes ranging from −1 to 0.2 (Figure 4a,b). It is important to note that, despite the overall declining trend in habitat quality, both improvement and degradation occurred simultaneously. Improvements in habitat quality (index changes between 0.4 and 1) were evident in areas further from urban centers, as shown in Figure 4a,b. This suggests that the expansion of construction land had a profound impact on regional habitat quality, with stronger degradation occurring closer to urban areas.

Land use transitions significantly influence regional habitat quality. Using geospatial mapping techniques and zonal statistics, changes in the habitat quality index (HQI) were calculated for each type of land use transition. A comparative analysis of the contribution rates of different transitions revealed the intensity of their impact on the regional habitat quality and identified the dominant factors driving these changes (Figure 5a,b). From 2000 to 2010, the conversion of forest to cultivated land emerged as the primary driver of habitat degradation, with a negative contribution rate of 57.31%. Concurrently, habitat improvement and degradation occurred simultaneously, with cultivated land reverting to forest being the dominant factor for habitat improvement, showing a positive contribution rate of 58.28%. Remarkably, the positive contribution rate of cultivated land-to-forest transitions exceeded the negative contribution rate of forest-to-cultivated land transitions (58.28% > 57.31%). Additionally, the positive contribution rate of cultivated land-to-grassland transitions was 5.38 times greater than the negative rate of cultivated land encroachment on grassland, highlighting the significant role of the “Grain for Green” policy in improving habitat quality in Guangxi during this period. However, the expansion of rural residential land was identified as a major threat to habitat quality, with a negative contribution rate of 16.26%, primarily attributed to the encroachment of cultivated land by rural residential development. This underscores the dual challenges of balancing rural development and habitat conservation.

Between 2010 and 2020, the expansion of urban land, rural residential land, and other construction land had an increasingly significant impact on habitat degradation in Guangxi. During this period, the contribution of cultivated land expansion to habitat degradation decreased from 57.31% to 53.36%, while the positive contribution of cultivated land-to-forest transitions to habitat improvement slightly increased from 58.28% to 58.56%, emphasizing the continued role of the “Grain for Green” policy in ecological restoration. Concurrently, rapid urbanization and infrastructure development led to a 39.8% increase in urban land and a 238.13% surge in transportation and industrial land, which contributed 0.52% and 5.80% to habitat degradation, respectively. Much of this was driven by the encroachment of construction land into cultivated land and forest land, with the negative contribution of construction land on cultivated land rising from 0.65% to 11.99% and on forest land increasing from 0.54% to 2.98%. Despite urban land expanding faster than rural residential land, the latter had a disproportionately greater negative impact on habitat quality, with its contribution rate being 30–50 times higher, likely due to its dispersed nature and proximity to natural ecosystems. These findings highlight the complex interplay between land use transitions and habitat quality, emphasizing the need for sustainable land management policies to mitigate ecological degradation.

3.3. Habitat Quality Changes in Typical Areas in Guangxi

To further explore the patterns of habitat quality changes in typical areas of Guangxi, we selected the regions experiencing the most rapid urbanization over the past two decades. These included the provincial capital, Nanning, and the coastal open city Beihai, representing areas of accelerated urban development. Additionally, considering that Guangxi is one of the provinces in China most severely affected by rocky desertification, we identified the Guangxi rocky desertification control area as a typical case study. This selection aims to analyze the patterns and influencing factors of habitat quality changes in both rapidly urbanizing regions and areas undergoing ecological restoration interventions.

3.3.1. Rapidly Urbanizing Areas in Guangxi

As the capital city of Guangxi, Nanning experienced a 2.13-fold expansion in urban land and a 9.77-fold increase in transportation and industrial land over the past two decades, making it the fastest urbanizing city in Guangxi. Similarly, Beihai, one of China’s 14 coastal open cities, saw a 1.92-fold increase in urban land and a 2.74-fold expansion in transportation and industrial land, accompanied by a 7.28% reduction in rural residential land, indicating rapid urban development during this period.

Over the past 20 years, the average HQI in Nanning showed a gradual decline, with a more pronounced decrease in the latter decade, falling by 3.73% (from 0.6533 to 0.6290). The spatial extent of very low habitat quality (0–0.2, Lv1) expanded outward from urban centers, increasing by 30.32% (Figure 6a–c). Conversely, areas with very high habitat quality (0.8–1, Lv5) consistently decreased by 0.65%. In Beihai, the HQI initially increased by 0.61% from 2000 to 2010 but then declined sharply by 7.32% from 2010 to 2020. This pattern closely correlates with changes in land use; during the first decade, urban and rural residential land slightly decreased (by 0.05–1.75%), whereas in the latter decade, urban land expanded by 95.34%, and other construction land (transportation and industrial) surged by 174.61%. This rapid urban development came at the expense of cultivated land, with approximately 10% of cultivated land converted into rural residential land. The total area of cultivated land lost to rural residential expansion was three to four times larger than Beihai’s urban land area. Such a trend underscores the need for greater attention to cultivated land loss, which may have more significant ecological consequences than urban expansion.

Although urban, transportation, and industrial land expanded rapidly in Nanning and Beihai over the past two decades, they may not be the primary drivers of regional habitat quality decline. Analysis of land use change contribution rates to habitat quality degradation (Figure 7a,b) provided deeper insights into the impacts of land use transitions in these rapidly urbanizing areas of Guangxi. Between 2000 and 2010, habitat quality decline in Nanning was primarily attributed to cultivated land expansion, particularly the conversion of forest land to cultivated land, which had a negative contribution rate of 48.53%. Rural residential land expansion was the second major factor, contributing 27.57% to habitat degradation. However, with the acceleration of urbanization and infrastructure development, this trend improved during 2010–2020, with their respective negative contribution rates decreasing by 3.84% and 0.22%. In contrast, Beihai’s habitat quality decline was mainly driven by rural residential land expansion, followed by the encroachment of cultivated land on forests and water bodies. The most notable transition was the conversion of cultivated land to rural residential land, with a significant negative contribution rate of 45%. Both regions showed a similar pattern between 2010 and 2020, as the negative contribution rates of rural residential land and cultivated land expansion decreased, while those of urban, transportation, and industrial land expansion increased. This indicates a growing impact of urbanization and infrastructure development on habitat quality. Nevertheless, rural residential land expansion remained the critical factor driving habitat quality decline in these areas.

3.3.2. Karst Rocky Desertification Region in Guangxi

Guangxi ranks third in China in terms of rocky desertification land area, following Guizhou and Yunnan. This phenomenon has resulted in the significant degradation of karst ecosystem functions, including diminished water retention capacity, exacerbated soil erosion, and worsened ecological and living conditions, posing a critical environmental challenge. Rocky desertification leads to a fragile socio-ecological system, threatening biodiversity and undermining sustainable development in the region. Land degradation neutrality (LDN) refers to a state in which the quantity and quality of ecosystem functions, services, and healthy land resources remain stable or are enhanced through management and restoration measures. This concept emphasizes the maintenance or improvement of land resource health and productivity without exacerbating land degradation pressures. One of its primary objectives is to achieve zero net land degradation globally by 2030 [41,42]. Over the past decades, the Guangxi provincial government has implemented proactive measures to combat rocky desertification, aiming to restore ecosystem stability and improve local livelihoods. However, understanding the spatio-temporal evolution of habitat quality over the past 20 years and identifying the driving factors behind these changes is crucial for assessing the effectiveness of these restoration efforts and guiding future strategies. Such research provides valuable insights into the interactions between land use changes and habitat quality in the context of a rapidly transforming karst environment.

Land use transition serves as the foundation for changes in habitat quality. Over the past two decades, land use changes in karst rocky desertification areas were significant. Cultivated land exhibited a pattern of a slight initial increase followed by a gradual decline, with area changes ranging from an increase of 0.01% to a decrease of 2.23%. Forests, water bodies, and rural residential areas initially decreased but later increased, while barren land consistently declined at a rate of approximately 2.35% per decade. In contrast, urban and other construction land expanded rapidly, with other construction land (primarily for transportation and industrial purposes) increasing by approximately 245% from 2010 to 2020. The mean habitat quality index in these areas decreased from 0.7185 to 0.6976 over the 20-year period, with the rate of decline accelerating between 2010 and 2020.

Spatially, areas with very low habitat quality (0–0.2] (Lv1) expanded outward from existing construction land centers, increasing from 2.89% to 4.51% of the total area (Figure 8a–c). This expansion was likely driven by the growth of urban, other construction, and rural residential land. In contrast, areas with very high habitat quality (0.80–1] (Lv5) saw a slight reduction, decreasing by 0.96%, further validating the overall decline in the habitat quality index. Temporal–spatial analysis (Figure 8d,e) revealed that rural and peri-urban areas experienced the most significant habitat quality degradation (index decreases of −1 to −0.4), while forested areas far from urban centers showed a notable improvement (index increases of 0.2 to 1). Evidently, during the 2010–2020 period, the dynamics of habitat quality were more pronounced, with both improvements and degradations intensifying. This suggests that rapid urbanization and ecological restoration efforts occurred simultaneously in the region. However, the benefits of ecological recovery were insufficient to offset the negative impacts of construction land expansion.

To further investigate the underlying causes of habitat quality changes in the karst rocky desertification control areas of Guangxi, the contribution of land use transitions to habitat quality was analyzed (Figure 9c). Over the past two decades, the decline in habitat quality in these areas was primarily driven by cultivated land expansion. However, as cultivated land gradually decreased, its negative impact on habitat quality weakened, with its contribution to habitat degradation declining from 69.93% to 63.67%. Conversely, the expansion of urban land, rural residential areas, and other construction land increasingly contributed to habitat degradation. Among these, the impact of other construction land expansion was particularly significant, with its contribution to habitat degradation during 2010–2020 being 8.48 times greater than in the previous decade. This aligned with the substantial increase in other construction land during that period. Notably, despite the rapid expansion of urban, transportation, and industrial land, their contributions to habitat degradation were still smaller than those of rural residential land. Furthermore, the impact of rural residential land expansion on habitat degradation showed an upward trend, increasing from 13.13% to 13.83%. These findings highlight the need for stricter rural land use policies to curb the uncontrolled expansion of rural residential land and mitigate its adverse effects on the ecological environment.

3.4. Spatio-Temporal Pattern of Land Use Change and Habitat Quality Under Different Scenarios in Guangxi in 2030

3.4.1. Land Use Change Under Different Scenarios in Guangxi in 2030

This study employed the PLUS model to predict land use changes in Guangxi by 2030. Using raster land use data from 2015 and 2020 as the baseline, the model’s Validation tool was applied to compare the 2020 simulation results against the observed 2020 raster data for accuracy assessment. The model achieved a Kappa coefficient of 0.753 and an overall accuracy of 0.806, demonstrating that the simulation met the required precision standards for reliable land use predictions.

Based on multi-scenario parameter settings (Section 2.3.2), this study simulated land use changes in Guangxi Province by 2030 under different development scenarios (

Table 6. Land use changes under different scenarios in 2030.

Land Use Types	2020	2030			2020–2030
		ND	UD	CE	ND	UD	CE
	Unit: km²	Unit: km²	Unit: km²	Unit: km²	Unit: km²	Unit: km²	Unit: km²
Cultivated land	49,973.66	49,930.76	49,957.64	50,065.76	−42.902	−16.016	92.099
Forest	155,659.27	155,600.68	155,608.89	155,714.69	−58.591	−50.376	55.426
Grass	20,557.41	20,616.31	20,594.22	20,561.26	58.901	36.809	3.846
Water	3785.41	3802.45	3800.40	3791.34	17.044	14.994	5.934
Urban land	1367.77	1421.47	1539.72	1383.30	53.693	171.943	15.521
Rural residential land	3467.05	3463.67	3360.27	3321.75	−3.384	−106.781	−145.304
Other construction land	1484.97	1461.45	1435.64	1464.63	−23.517	−49.329	−20.336
Bare land	37.17	35.93	35.93	29.99	−1.245	−1.245	−7.185
Land Use Types	2020	2030			2020–2030
		ND	UD	CE	ND	UD	CE
	Unit: %	Unit: %	Unit: %	Unit: %	Unit: ‰	Unit: %	Unit: %
Cultivated land	21.145	21.127	21.139	21.184	−0.086	−0.032	0.184
Forest	65.864	65.840	65.843	65.888	−0.038	−0.032	0.036
Grass	8.699	8.723	8.714	8.700	0.287	0.179	0.019
Water	1.602	1.609	1.608	1.604	0.450	0.396	0.157
Urban land	0.579	0.601	0.652	0.585	4.069	12.727	1.275
Rural residential land	1.467	1.466	1.422	1.406	−0.098	−3.080	−4.191
Other construction land	0.628	0.618	0.607	0.620	−1.584	−3.322	−1.369
Bare land	0.016	0.015	0.015	0.013	−8.025	−8.025	−23.233

Notes: since the area of some islands and coastal waters is not counted, the total area is not exactly consistent with the official land area of Guangxi.

and Figure 10a–c). The results indicate that under the ND scenario, grassland, water bodies, and urban land are expected to increase, with growth rates ranging from 0.287% to 4.069%, whereas arable land, forest, rural residential areas, other construction land, and bare land are projected to decrease by 0.086% to 8.025%. Bare land, having a relatively small area, exhibits the most significant reduction (8.025%). In the UD scenario, urban expansion is anticipated to occur at the expense of cultivated land, forest, other construction land, and rural residential areas, with rural residential land experiencing the most significant reduction, shrinking by 106.781 km² (−3.08%). Under the CE scenario, the reduced encroachment of construction land on ecologically functional land types such as cultivated land, forest, and grassland leads to significant expansions in cultivated land (92.099 km², 0.184%) and forest (55.426 km², 0.036%), while rural residential areas show the largest reduction (−4.191%). Notably, in the UD scenario, despite the higher probability of other land types transitioning to urban land and the largest increase in the urban land area (12.727%), the reduction in forest area remains smaller compared to the ND scenario, suggesting that further urbanization does not necessarily result in proportional reductions in ecological land. In other words, even with continued urbanization, its impact on ecologically functional land remains limited.

3.4.2. Habitat Quality Changes Under Different Scenarios in Guangxi in 2030

The extent of ecological environment changes varies depending on the trajectory of human activities (Figure 11a–c). Under the UD scenario, urban land expansion by 2030 is projected to be the most pronounced, extending outward from the 2020 urban land base. This expansion is primarily driven by the conversion of rural residential land to urban land, aligning with Guangxi’s current urbanization trends. Analysis of the spatial distribution of habitat quality in Guangxi for 2030 reveals that the expansion of construction land (including urban, rural residential, and other construction land) will lead to a further increase in the area of regions with a very low-quality habitat (Lv1: 0–0.2). In 2020, the mean HQI for Guangxi was approximately 0.76817. By 2030, the mean HQI is expected to decline to 0.73811 under the ND scenario and 0.73815 under the UD scenario. Conversely, under the CE scenario, habitat quality improves with a mean HQI increase of 0.001347. These results indicate a varying degree of habitat quality degradation under the ND and UD scenarios, whereas the CE scenario achieves significant improvements.

Notably, the decline in the HQI under the UD scenario is less severe compared to the ND scenario. This suggests that while urban land expansion accelerates under the UD scenario, habitat quality degradation occurs at a slower pace than during historical development trends. This phenomenon may be attributed to the concentration of rural populations in urban areas, leading to a reduction in rural residential land and mitigating human impact on ecosystems such as forests, grasslands, and cultivated land. Furthermore, this trend may reflect the influence of China’s ecological civilization policies and green development strategies, emphasizing the protection of ecologically functional land types such as forests and grasslands. Consequently, future urbanization in Guangxi appears to prioritize sustainable practices to balance urban growth with ecological conservation.

4. Discussion

This study, based on the PLUS-InVEST model, explores the spatial and temporal variation in habitat quality in Guangxi, China, and systematically assesses the impact of land use change on ecosystem services, particularly habitat quality. The results indicate significant spatial and temporal variations in habitat quality, reflecting the complex interactions between land use change, urbanization, agricultural activities, and conservation measures.

4.1. Spatial and Temporal Variation in Habitat Quality

The spatial pattern of habitat quality in Guangxi is significantly influenced by the expansion of rural settlements and agricultural activities. In other regions of China, land use/cover changes, particularly the conversion of agricultural land to urban areas, have led to a significant decline in ecosystem services [43,44]. This study shows that, compared to urban expansion, the expansion of rural settlements has a greater impact on habitat degradation, which diverges from previous studies on the impact of land use change during urbanization on habitat quality [45,46]. The temporal variation in habitat quality highlights the effects of both short-term and long-term land use changes. Over the past 20 years, Guangxi has experienced rapid urbanization, which has intensified the fragmentation of natural habitats. This trend aligns with the global acceleration of habitat degradation due to urban expansion [44,47]. Rural settlements are often more dispersed and closer to natural ecosystems such as forests and grasslands, making them more prone to threatening habitat quality. Additionally, this study indicates that, despite improvements in habitat quality in some areas due to reforestation and ecological protection policies, the effectiveness of ecological restoration will be insufficient to offset the impacts of urban development.

4.2. Land Use Change and Ecological Services’ Spatial and Temporal Variation in Habitat Quality

Our findings indicate that land use change, particularly the loss of cultivated land and the expansion of built-up areas, has had a significant impact on habitat quality in Guangxi. The loss of cultivated land is a critical issue in many regions of China, and recent studies on cultivated land loss in China provide supporting evidence for this trend [48]. The conversion of agricultural land to urban areas typically results in a decline in ecosystem service values, such as biodiversity, hydrological regulation, and soil fertility [43,49]. Further analysis highlights the importance of preserving rural landscapes and promoting sustainable land use practices to mitigate the negative effects of urban expansion, consistent with the findings of Liu et al. [50], who emphasize the need for balanced land use policies to ensure the long-term sustainability of rural environments.

The results of this study also emphasize the necessity of implementing policies that prioritize the protection of high-quality habitats, especially in areas prone to karst rocky desertification. Similar conclusions have been made by other studies, which argue that integrating ecological conservation into land use planning is an effective means of preventing the further degradation of natural habitats [51].

4.3. Limitations and Outlook

Despite the fact that this study provides important insights into the spatial and temporal variations in habitat quality in Guangxi, there are several limitations. For example, while the predictions made by the PLUS model closely align with actual conditions, with a Kappa coefficient of 0.785, the model lacks long-term validation data and does not fully account for external driving factors such as climate change and policy shifts, which may affect the accuracy of the model’s predictions. Additionally, although the habitat quality module of the InVEST model is widely utilized in ecological research, resource management, and sustainable development, it is important to acknowledge certain limitations in its parameter selection and applicability. The parameters of the model are often based on empirical data or the literature, lacking standardized definitions and measurement criteria, which can affect their suitability across different regions or ecosystems. Additionally, while some parameters significantly influence the model’s outcomes, the model may not conduct comprehensive sensitivity analyses for these critical parameters, leading to uncertainties in the results. The InVEST model simplifies the complex relationships between land use changes and habitat quality, which may not fully capture the dynamic nature of the ecosystem in Guangxi.

Future research could address these limitations by incorporating higher-resolution data and long-term monitoring data. More precise dynamic scenario analyses, including the consideration of climate change and policy shifts, would enhance the model’s predictive accuracy. Furthermore, the integration of advanced techniques, such as machine learning and remote sensing technologies, could improve the sensitivity and precision of habitat quality predictions. This would deepen our understanding of the impacts of habitat change on biodiversity conservation, offering more comprehensive insights into ecosystem dynamics and providing a solid foundation for future conservation strategies.

5. Conclusions

Utilizing the InVEST model, spatiotemporal patterns of land use and habitat quality across Guangxi over the past two decades were systematically analyzed. Selecting rapidly urbanizing areas and karst rocky desertification control regions as typical case studies, this research provided a comprehensive analysis of the factors influencing habitat quality changes in Guangxi, from a regional to a more localized perspective. Additionally, considering natural factors, socio-economic elements, and future development policies in Guangxi, the PLUS model was employed to simulate and predict land use patterns for 2030 under three development scenarios, while exploring the trends of habitat quality changes under these scenarios. The key findings of the study are as follows:

(1)

There were significant changes in land use types, characterized by the rapid expansion of urban land and other construction land. Specifically, from 2010 to 2020, urban land increased by 39.8%, while transportation and industrial land experienced a remarkable growth of 238.13%. This expansion primarily came at the expense of agricultural land and forests; during this period, the conversion of agricultural land to urban and other construction uses accounted for 52.17% and 52.83% of the net increase in land, respectively. Consequently, the degradation of habitat quality was most pronounced in areas surrounding urban centers (Figure 4a,b), confirming the intensified impact of urbanization and construction land expansion on habitat degradation in Guangxi over the past decade. Furthermore, despite a rapid decline in the rural population of Guangxi over the past ten years (with a decrease of 14.18%), the area of rural residential land has paradoxically increased by 3.43%. This counterintuitive phenomenon warrants attention from local management authorities.

(2)

Although ecological restoration measures such as the “Grain-for-Green” program have played a significant role in alleviating habitat degradation, especially in rural and karst desertification areas, the expansion of built-up land still outweighed the positive effects of ecological restoration. The expansion of forests and grasslands contributed over 58% to habitat improvement in these areas. However, ecological restoration will be insufficient to offset the impacts of urban development. These findings highlight the importance of targeted land use planning and balanced urbanization policies to achieve a harmonious trade-off between economic development and ecological protection.

(3)

The expansion of rural settlement land was found to have a significantly greater negative impact on regional habitat quality than the expansion of urban or infrastructure-related land. From 2000 to 2020, the contribution rates of rural residential land expansion to habitat quality degradation increased to 16.26% and 15.38%, respectively. The negative contribution rate associated with the expansion of rural residential areas was found to be 30 to 50 times greater than that of urban and other types of construction land, with urban expansion contributing negatively at rates of 0.52% to 0.62% and other construction land at rates of 1.4% to 5.8%. Furthermore, as the rate of urbanization increased, the contribution of rural land expansion to the degradation of habitat sub-quality diminished. This suggests that appropriate urbanization might be beneficial for improving overall habitat quality in the region.

(4)

Under the Natural Development scenario, habitat quality is expected to continue to decline. In contrast, the Urban Development scenario, which emphasizes compact urbanization, reduces the pressure on rural land and slows the pace of habitat degradation. This is likely due to accelerated urbanization, which facilitates the migration of rural populations to urban areas, thereby reducing human encroachment on ecosystems. Additionally, the Cropland and Ecological Protection scenario shows potential for improvement, with the habitat quality index projected to increase by 0.13%, further validating the effectiveness of sound land management policies and ecological restoration measures.