Introduction

In China’s northwestern region, Zhongwei serves as a critical ecological security barrier, maintaining biodiversity, conserving soil and water, purifying the air, and controlling wind and sand. Located in the transitional zone between arid and semiarid climates, Zhongwei features a fragile ecological environment characterized by desert landscapes and limited water resources. Consequently, it is susceptible to ecological challenges, including vegetation degradation, land desertification, biodiversity loss, and soil erosion1,2. Land, a core natural resource, profoundly impacts human economic activities and the health and stability of ecosystems in Zhongwei. Habitat quality is crucial, indicating the survival conditions and suitability for biological species in biomes3. Changes in land use and patterns driven by population growth and urbanization significantly affect the structure and function of natural environments and biological habitats in Zhongwei4. These findings highlight the deep connection between socioeconomic activities and the natural environment, which plays a pivotal role in influencing habitat quality5. Therefore, analyzing the characteristics of land use change and factors driving land expansion in Zhongwei is vital for protecting the local ecosystems and managing ecological risks. This analysis also aids in predicting trends and causes that affect habitat quality6. Land use changes in Zhongwei, driven by urbanization, agricultural expansion, and ecological restoration projects, pose significant challenges to the region’s habitat quality. These changes can lead to habitat fragmentation, loss of biodiversity, and degradation of ecological services, threatening the long-term sustainability of the local ecosystem7. Moreover, the arid climate and limited water resources exacerbate ecological vulnerability, making the region particularly susceptible to the adverse effects of land use transformations. These changes not only reflect economic progress but also pose challenges to sustainable resource management and ecological balance8. On the one hand, urban expansion and the establishment of industrial zones have contributed to economic growth and job creation. On the other hand, they have led to increased pressure on natural resources, such as water and land, thereby exacerbating environmental degradation and biodiversity loss9. Concurrently, ecological factors, particularly aridity, soil degradation, and climate variability, play crucial roles in shaping land use decisions in Zhongwei. Changes in precipitation patterns, temperature, and soil type directly influence agricultural productivity, forest cover, and grassland health, which in turn affect land use strategies and policies. The region’s fragile ecosystems are highly susceptible to human activities, making it imperative to balance economic development with ecological conservation10,11,12.

Land use prediction refers to the analysis of historical data and the prediction of future land use trends, which is vital for stipulating scientific land management policies13. With technological developments in GIS, remote sensing, machine learning, and artificial intelligence, significant breakthroughs have been made in the theory, method and application of land use prediction14. In current research, the most commonly used models for land use prediction are coupled models such as Markov-FLUS and CA-Markov. These methods excel in processing large-scale spatial data, simulating time series changes in land use, integrating multiple influencing factors, and providing dynamic simulations. However, these models still exhibit significant limitations in their strategy for mining transformation rules and in the simulation of landscape dynamics15. The PLUS model, an emerging tool in land use change simulation, integrates deep learning and multiscale spatial analysis to enhance adaptability and accuracy in managing complex landscape dynamics while preserving detailed spatial data. This model addresses the deficiencies of traditional coupled models in transformational rule mining strategies and simulations of landscape dynamics, thus providing a scientific basis for land resource management and planning16. Habitat quality assessment involves the systematic analysis and evaluation of the conditions of biological habitats within a specific area, assessing factors such as ecosystem structure and function, species diversity, and overall habitat health. Typically, this assessment is conducted by constructing ecological indicators for habitats using various ecological models, supported by field surveys and remote sensing techniques17. Recently, the InVEST model, which accounts for threatening factors, has been widely used in habitat quality assessment due to its advantages in spatial visualization and accuracy18. Combining the PLUS model with the InVEST model provides a powerful framework for simulating land-use change scenarios on the basis of complex socioeconomic and ecological factors. This integrated model is capable of incorporating various driving forces of change and generating spatially explicit land use maps, making it highly suitable for exploring the potential impacts of different development strategies on the natural environment. The quantification of the spatial distributions of regions differing in habitat quality and degree of habitat degradation allows for a detailed assessment of how land use decisions influence biodiversity and ecological integrity19.

The distinctive natural conditions of Zhongwei provide a representative and valuable basis for research on land use, ecological protection, and agricultural development. Recent economic and social development, particularly in tourism and agriculture, has led to significant changes in land use, placing considerable pressure on land resources20. Consequently, the region faces a critical challenge: how to utilize land resources scientifically while protecting and improving habitat quality. Simulating and predicting land use changes and their ecological impacts under different development scenarios are crucial for assessing and optimizing conservation measures, thereby promoting ecological security and sustainable development in Zhongwei. This promotes ecological security and sustainable development in the region. Therefore, we utilized land use data from 1980, 2000, and 2020 to couple the PLUS and InVEST models. This allowed us to predict and analyze the spatial and temporal evolution, characteristics, and drivers of changes in land use and habitat quality in Zhongwei. We considered three scenarios: natural development, cultivated land protection, and ecological protection. In this study, we aim to (1) analyze the characteristics and drivers of land use changes in Zhongwei, (2) assess the ecological impacts of land use changes on habitat quality, (3) simulate and predict land use changes under various development scenarios, and (4) propose conservation measures to mitigate ecological risk and promote sustainable development. Notably, our study revealed that prioritizing ecological conservation reduces the need to improve land use structure and habitat quality, thereby promoting sustainable development in Zhongwei. These findings contribute to a deeper understanding of land use dynamics in arid and semiarid regions and provide a scientific basis for policy-makers and stakeholders to develop effective land management strategies.

Materials and methods

Study area

The study area is the city of Zhongwei, Ningxia Hui Autonomous Region, in northwestern China (104°17′–106°06′ E, 36°06′–37°50′ N; Fig. 1). Situated at the intersection of the Loess Plateau and the Inner Mongolia Steppe, the area covers 13,648.12 km2. It experiences an arid and semiarid continental climate, characterized by cold, long winters and hot, windy summers. The mean annual temperature is approximately 10 °C, with the lowest average temperature in January reaching − 10 °C and the highest temperature in July reaching 33 °C. The annual precipitation is sparse, ranging from 150 to 250 mm, and is predominantly concentrated between June and September. Sunshine conditions are favorable, with the mean annual sunshine duration exceeding 2,800 h, and evaporation rates varying from 180 to 367 mm. The frost-free period lasts approximately 160 d, with the growing season lasting from April to the end of September. Vegetation and soil types exhibit distinct transitional characteristics. Moving from north to south, zonal soils transition from wind-formed sandy soils to dark loess and mildly saline soils. The main vegetation types include shrub communities, dry grasslands, and meadows, which are dominated by poplar, sea buckthorn, and various drought-tolerant herbs.

Fig. 1
figure 1

Overview of the study area. The figure was generated using ArcGIS 10.6 (https://www.esri.com/zh-cn/arcgis/products/arcgis-desktop/overview).

Full size image

Data source and processing

Building on the foundation of previous studies and considering the characteristics of the study area, along with the availability and timeliness of data, 12 factors driving land use change were selected. This study utilized the following data sources: (1) Land use, temperature, precipitation, information on administrative divisions, population density, GDP, and soil data were collected from the Resource and Environmental Sciences and Data Center, Chinese Academy of Sciences (http://www.resdc.cn, accessed on 4 February 2024). (2) Digital elevation model (DEM) data with a 30 m resolution, derived from ASTER GDEM, were obtained from the Geospatial Data Cloud (http://www.gscloud.cn, accessed on 5 February 2024) and used to determine the slope and aspect. (3) Road vector data were sourced from the National Basic Geographic Database (https://www.webmap.cn, accessed on 5 February 2024) and used to compute Euclidean distances. Following the LUCC system of the Chinese Academy of Sciences and adapting to the unique land use traits of the Hexi Corridor, land cover was classified into six categories: farmland, woodland, grassland, watershed, construction land, and unused land. The classification results for each period of 1000 sample points were randomly selected and verified for accuracy. We used the software Bigmap, and the overall accuracy was 85%, which met the requirements of the study (Table 1).

Table 1 Table of driving factors.
Full size table

Methodology

Land use dynamic degree

The land use dynamic degree reflects the change rate of land resources and uses and quantifies the transformation process of land use types in the studied region. It is calculated via the following formula21:

$$\:{K}_{i}=\frac{{U}_{b}-{U}_{a}}{T\times\:{U}_{a}}$$
(1)

where \(\:{K}_{i}\) represents the dynamic degree of land use type \(\:i\) in Zhongwei, Ningxia; \(\:{U}_{a}\) represents the area of a specific land type at the beginning of the study; \(\:{U}_{b}\) represents the area of a specific land type at the end of the study; \(\:T\) represents the duration of the study, in years.

Land use dynamic degree

Land use transfer matrices are key for studying the impact of human activities on land use changes and used to describe the conversion between different land use types. The formula is as follows22:

$$\:{A}_{ij=}\left[\begin{array}{cc}\begin{array}{cc}{A}_{11}&\:{A}_{12}\\\:{A}_{21}&\:{A}_{22}\end{array}&\:\begin{array}{cc}\dots\:&\:{A}_{1n}\\\:\dots\:&\:{A}_{2n}\end{array}\\\:\begin{array}{cc}⋮&\:⋮\\\:{A}_{n1}&\:{A}_{n2}\end{array}&\:\begin{array}{cc}\dots\:&\:⋮\\\:\dots\:&\:{A}_{nm}\end{array}\end{array}\right]$$
(2)

where \(\:{A}_{ij}\) refers to the area of land transferred from land use types \(\:i\) to j; \(\:i\) and \(\:j(i,j=\text{1,2},3\dots\:,\:\:n)\) refer to the land use types before and after transfer; and \(\:n\) is the number of land use types before and after the transfer.

PLUS model

The Markov model is a widely used model in land transfer studies to predict the probability of a system changing from one state to another23. It is assumed that future states depend only on the current state, not on the events that occurred before it. It is formulated as:

$$\:{S}_{(t+1)}={P}_{ij}\times\:{S}_{t}$$
(3)

where \(\:{S}_{t}\) and \(\:{S}_{(t+1)}\) represent the land use states at moments \(\:t\) and \(\:t+1\), respectively; and \(\:{P}_{ij}\) is the land use transfer matrix. In this study, we implemented stochastic fitting of land use types and their drivers based on the Markovian model prediction of the PLUS model.

The PLUS model introduces the LEAS module, which combines the strengths of the target area strategy (TAS) and probability adjustment strategy (PAS). The LEAS model converts land use type dynamics into a simple binary classification. Through ensemble learning methods, the classification results of the training samples are combined from a multitude of decision trees, to select random subsamples from the original dataset. Finally, the model predicts the growth probability of land use types24. The formula is as follows:

$$\:{P}_{i,k}^{d}\left(x\right)=\frac{{\sum\:}_{n=1}^{M}I\left({h}_{n}\left(x\right)=d\right)}{M}$$
(4)

where \(\:{P}_{i,k}^{d}\left(x\right)\) represents the growth probability of land use type in cell \(\:i\); \(\:d\) is either 0 or 1: (a value of 1, indicates that other land use types are transformed to land use type \(\:k\), and a value of 0, indicates that there are other transformations); \(\:x\) represents a vector composed of driving factors; \(\:I\left(x\right)\) represents an indicator function for decision number sets; \(\:{h}_{n}\left(x\right)\) represents the \(\:n\)th predicted type of the decision tree of the vector \(\:x\); and \(\:M\) represents the sum of the decision trees.

The PLUS model calculates the total probability based on a multitype random patch seeding mechanism with descending thresholds, in order to simulate the patch evolution of each land use type. This mechanism adopts the Monte Carlo method to calculate the “seeds” of land use types25 that change in terms of neighborhood effect, which is formulated as follows:

$$\:If{\sum\:}_{k}^{N}\left|{G}_{c}^{t-1}\right|-{\sum\:}_{k=1}^{N}\left|{G}_{c}^{t}\right|<Step\:Then,\:l=l+1$$
(5)
$$\:\left\{\begin{array}{c}Change\:P{d}_{i,c}^{d=1}>\tau\:\:and\:T{M}_{k,c}=1\\\:no\:change\:P{d}_{i,c}^{d=1}\le\:\tau\:\:or\:T{M}_{k,c}=0\end{array}\:\:\:\tau\:={\delta\:}^{l}\times\:rl\right.$$
(6)

Where Step is the step size of the land use demand fitted by the PLUS model; \(\:T{M}_{k,c}\) is the transfer matrix, which determines whether land use type k can be transformed to land use type c; δ is the decay coefficient of the descending threshold \(\:\tau\:\), \(\delta \in \left[ {0,~1} \right]\), whose exact value is set by experts; l refers to the number of decay steps; rl is a normally distributed random value with the mean of 1, and \(\:rl\in\:[0,\:\:1]\). This descending threshold can maximize the chance of changes in grid cells with higher overall probability. The CA model based on multitype random patch seeds shows spatiotemporal consistency and allows new land use patches to develop freely, subject to development probabilities.

After the simulation, we used the kappa coefficient test for accuracy, which is formulated as follows26:

$$\:Kappa=\frac{{P}_{0}-{P}_{c}}{{P}_{a}-{P}_{c}}$$
(7)

where \(\:{P}_{0}\) is the number of simulated correct grids/total grids; \(\:{P}_{c}\) is the number of randomly simulated correct grids/total grids; and \(\:{P}_{a}\) is the number of randomly simulated correct grids/total grids. Kappa coefficients between 0.6 and 0.8 represent good modeling results. The land use data of Zhongwei in 1980 and 2000 were used to predict those in 2020, and the predicted results was then compared with the actual data from the same year according to the Validation module of the PLUS model. Through this process, we obtained the kappa coefficient of 80.32%, indicating that the land use predictions for 2040 are credible.

Multiscenario setup for PLUS model simulation of future land use

  1. 1.

    Natural Development Scenario (NDS). This scenario sets a continuation of the land use change trend from 2000 to 2020, without considering the conversion probability between various land use types or the development policy requirements involved. With an interval of 20 y, we used the Markov-Chain in the PLUS model to predict the land use demand in 2040 under this scenario, in which parameters such as the land expansion capacity, land use transfer matrix, ___domain factor weights, and habitat quality index were kept consistent with those in the 2000–2020 models.

  2. 2.

    Cultivated Land Protection Scenario (CLPS). The security of cultivated land is the foundation for food security. Cultivated land can be protected by controlling the conversion probability of farmland to other land use types and the expansion of construction lands. The conversion probability of cultivated land to other land use types is controlled at a 60% reduction, to ensure that the cultivated land protection requirements are strictly satisfied.

  3. 3.

    Ecological Protection Scenario (EPS). Considering food security and the resource environmental bearing capacity in order to protect ecosystems and biodiversity, the Chinese government set under in the EPS a 50% reduction in the probability of woodland, grassland and watershed converted to unused land and a 50% increase in the probability of unused land being converted to woodland, grassland and watershed. The watersheds in this area are constrained from arbitrary conversion.

Habitat quality evaluation based on the InVEST model

The Habitat Quality Module of the InVEST model was used to evaluate habitat quality through the Habitat Quality Index (HQI) and the Habitat Degradation Index (HDI). The Habitat Degradation Index (HDI) is commonly used to describe the degree of habitat degradation in a region, with higher values indicating higher threat levels. Its value ranges from 0 to 1. The formula is as follows27:

$$\:{D}_{xj}=\sum\:_{r=1y=1}^{R}\sum\:_{y=1}^{{r}_{y}}\left(\frac{{w}_{r}}{{\sum\:}_{r=1}^{R}{w}_{r}}\right){r}_{y}{i}_{rxy}{\beta\:}_{x}{S}_{ir}$$
(8)

where \(\:{D}_{xj}\) is the degree of habitat degradation; \(\:R\) is the number of threat factors; \(\:r\) is the number of grids in the threat layer on the base map; \(\:w\) is the weight of the threat factor; \(\:{r}_{y}\) is the threat intensity; \(\:{i}_{rxy}\) is the (linear or exponential) effect of \(\:r\) on each grid of the habitat; \(\:{\beta\:}_{x}\) is the effect of local conservation policies; and \(\:{S}_{ir}\) indicates the relative sensitivities of the habitats to different threat sources.

The Habitat Quality Index (HQI) measures the suitability of habitat conditions in a region for one or more species, and ranges from 0 to 1, with the higher values indicating better habitat quality. The formula is as follows:

$$\:{Q}_{xj}={H}_{j}(1-\frac{{D}_{xj}^{z}}{{D}_{xj}^{z}z+{k}^{z}})$$
(9)

where \(\:{Q}_{xj}\) is the habitat quality index of raster \(\:x\) in land use type \(\:j\); \(\:{H}_{j}\) is the habitat suitability of land use type \(\:j\); \(\:{D}_{xjz}\) is the habitat degradation of raster \(\:x\) in land use type \(\:j\); \(\:z\) is a default parameter; and \(\:k\) is a half-saturation constant, which was 0.2 in this study. With reference to previous research, this study considered farmland, construction land, and unused land as threat factors and set the impact distance, weighting and sensitivity for different threat factors (Tables 2 and 3).

Table 2 Threat feed parameters.
Full size table
Table 3 Sensitivity of threat factors.
Full size table

Results

Land use analysis

Land use changes in Zhongwei from 1980 to 2020

Farmland and grassland are the dominant land use types in Zhongwei, accounting for more than 80% of the total area (Fig. 2; Table 4). From 1980 to 2020, the total area of farmland in Zhongwei increased by 336 km². However, there was a decrease of 304 km2 in farmland area from 2000 to 2020. The area of woodland continuously increased from 1980 to 2020. However, the dynamics of land use from 1980 to 2000 were less intense than those from 2000 to 2020. Additionally, the recovery rate of woodland has slowed since 2000. The areas of grassland and watershed decreased overall from 1980 to 2020, although an increase was observed from 2000 to 2020. The area of construction land has continued to grow. The dynamic changes in land use between 2000 and 2020 were significantly greater than those observed from 1980 to 2020, indicating a sharp increase in the area of construction land since 2000. Additionally, the area of unused land increased from 1980 to 2000 and decreased from 2000 to 2020. The degree of land utilization has increased since 2000 (Table 4).

Fig. 2
figure 2

Land use change: (a) land use in 1980; (b) land use in 2000; (c) land use in 2020. The figure was generated using ArcGIS 10.6 (https://www.esri.com/zh-cn/arcgis/products/arcgis-desktop/overview).

Full size image
Table 4 Land use area changes and dynamics in Zhongwei from 1980 to 2020.
Full size table

In terms of land transfer (Table 4; Fig. 3), construction land expanded gradually from 1980 to 2020 into the eastern part of Shapotou District and the central part of Zhongning County, with a large area of farmland and grassland being converted to construction land. During the study period, the increased area of woodland mostly came from grassland, mainly located in Shapotou District. The area of farmland decreased by 304.89 km2 from 2000 to 2020, primarily through conversion to grassland. Between 1980 and 2000, the area of watershed area significantly decreased due to conversion to farmland. However, since 2000, there has been a resurgence in the area of watershed. Between 2000 and 2020, the area of unused land decreased by 80.1 km2, primarily through conversion to farmland and grassland, resulting in an optimized land use structure.

Fig. 3
figure 3

Land use type transfer in Zhongwei: (a) land use type transfer from 1980 to 2000; (b) land use type transfer from 2000 to 2020; (c) land use type transfer from 1980 to 2020. The figure was generated using Origin 2022 (https://www.originlab.com/2022).

Full size image

Land use expansion factors

The land use data of Zhongwei with 12 factors driving land use change were imported into the PLUS model for spatial overlay, and the expansion factors were analyzed using the LEAS module of the PLUS model to obtain the contribution values of the land use type expansion drivers (Fig. 4). DEM had a significant effect on grassland expansion. GDP notably influenced woodland expansion. The distance from highways significantly affected the expansion of farmland and construction land. Precipitation had a pronounced effect on the expansion of all land use types. Distance from water significantly influenced the expansion of grassland, watershed, and construction land. Temperature notably affected the expansion of farmland, woodland, and watershed. Distance from railways had a significant effect on the expansion of grassland and unused land. Population density had a notable effect on the expansion of farmland, watershed, and unused land. Distance from government facilities significantly affected the expansion of farmland, woodland, grassland, construction land, and unused land. The soil type, slope, and aspect had relatively minor impacts on the expansion of various land use types.

Fig. 4
figure 4

Contributions of the factors driving land use change drivers in Zhongwei. The figure was generated using Origin 2022 (https://www.originlab.com/2022).

Full size image

Change in habitat quality in Zhongwei from 1980 to 2020

The habitat degradation index (HDI) in the study area was calculated using the habitat quality module of the InVEST model. The results were classified into five classes: low (0, 0.089), relatively low (0.089, 0.18), medium (0.18, 0.26), relatively high (0.26, 0.35), and high (0.35, 1) (Fig. 5). From 1980 to 2020, the mean values of the HDI consistently increased and were 0.210, 0.214, and 0.235 in 1980, 2000, and 2020, respectively, with a continuous increase in regions with high and relatively high habitat degradation and a continuous decrease in areas with low and relatively low habitat degradation. The regions with high degradation lied in the eastern part of Shapotou District, the central part of Zhongning County and the southern part of Zhongwei, where the grassland and farmland are under greater threat. In addition, the HDI of grassland in the southern part of Zhongwei increased continuously.

Fig. 5
figure 5

Change in habitat degradation: (a) habitat degradation in 1980; (b) habitat degradation in 2000; (c) habitat degradation in 2020. The figure was generated using ArcGIS 10.6(https://www.esri.com/zh-cn/arcgis/products/arcgis-desktop/overview).

Full size image

The Habitat Quality Index (HQI) in the study area was calculated using the habitat quality module of the InVEST model. The results were classified into five classes: low (0, 0.29), relatively low (0.29, 0.49), medium (0.49, 0.60), relatively high (0.60, 0.78), and high (0.78, 1) (Fig. 6). The HQI decreased annually, with mean values of 0.38, 0.35, and 0.33 in 1980, 2000, and 2020, respectively. From 1980 to 2020, the area of regions with low, relatively low and high HQIs increased, whereas the area of regions with medium and relatively high HQIs decreased, indicating that the land use structure in Zhongwei became more compact and that land conservation improved.

Fig. 6
figure 6

Changes in habitat quality: (a) habitat quality in 1980; (b) habitat quality in 2000; (c) habitat quality in 2020; (d) habitat quality in 2040 under the natural development scenario; (e) habitat quality in 2040 under the cultivated land protection scenario; (f) habitat quality in 2040 under the ecological protection scenario. The figure was generated using ArcGIS 10.6 (https://www.esri.com/zh-cn/arcgis/products/arcgis-desktop/overview).

Full size image

Land use projections for 2040 under different scenarios

Based on the land use data from 2000 to 2020, the coupled Markov-Plus model was used to predict the distributions of land use types in Zhongwei in 2040 under the following three scenarios (Fig. 7; Table 5). Under the natural development scenario (NDS), the overall land use trend remained unchanged, with the area of farmland decreasing to 5909.13 km2, the area of woodlands increasing to 524.45 km2, the area of construction land increasing to 788.90 km2, the areas of grassland and watershed basically remaining unchanged, and the area of unused land decreasing. In the cultivated land protection scenario (CLPS), the total area increased by 511.37 km2 compared with that in 2020, and the areas of woodland, grassland, construction land and unused land decreased, compared with those under the NDS. Under the ecological protection scenario (EPS), the areas of woodland, grassland and water were greater than those under the NDS; the area of farmlands was less than that under the CLPS but greater than that under the NDS; and the area of construction land was less than that under the NDS but greater than that under the CLPS.

Fig. 7
figure 7

Land use change: (a): land use in 2040 under the Natural development scenario; (b): land use in 2040 under the Cultivated land protection scenario; (c): land use in 2040 under the Ecological protection scenario. The figure was generated using ArcGIS 10.6(https://www.esri.com/zh-cn/arcgis/products/arcgis-desktop/overview).

Full size image
Table 5 Changes in land use area under different scenarios.
Full size table

Habitat quality projections for 2040 Zhongwei in under different scenarios

The mean values of the HDI under the three different scenarios predicted for 2040 ranked as follows: Cultivated land protection scenario (0.289) > Natural development scenario (0.266) > Ecological Protection scenario (0.264). The HDI values under all scenarios decreased compared to those in 2020. The proportions of regions differing in degree of degradation under the in NDS and EPS were similar; the area of regions with relatively high and high degradations under the CLPS was greater than that under the NDS and EPS, whereas the of regions with relatively low and low degradations was less than that under the NDS and EPS (Fig. 8).

Fig. 8
figure 8

(a) Habitat degradation in 2040 under the Natural development scenario; (b) habitat degradation in 2040 under the Cultivated land protection scenario; (c) habitat degradation in 2040 under the Ecological protection scenario. The figure was generated using ArcGIS 10.6 (https://www.esri.com/zh-cn/arcgis/products/arcgis-desktop/overview).

Full size image

The mean values of the HQI under the three scenarios were ranked as follows: ecological protection scenario (0.35) > cultivated land protection scenario (0.34) > natural development scenario (0.33), The HQI values under, all scenarios increased compared with those in 2020, suggesting a potential increase in the ecological quality of Zhongwei. The proportions of regions differing in habitat quality under the NDS and EPS were similar; the area of regions with low and relatively low habitat quality values under the CLPS was significantly greater than that under the NDS and EPS, whereas the area of regions with medium habitat quality values under the CLPS was significantly lower than that under the NDS and EPS (Fig. 9).

Fig. 9
figure 9

Change in habitat quality: (a) habitat quality in 2040 under the natural development scenario; (b) habitat quality in 2040 under the cultivated land protection scenario; (c): habitat quality in 2040 under the ecological protection scenario. The figure was generated using ArcGIS 10.6 (https://www.esri.com/zh-cn/arcgis/products/arcgis-desktop/overview).

Full size image

To determine the spatial distribution of regions differing in habitat quality, we mapped the differences between the habitat quality measured in 2020 and that predicted under the three scenarios. The results were categorized into drastic decline [0.5, 1), decline[0.1, 0.5), basically unchanged [– 0.1, 0.1), rise [– 0.5, – 0.1), and drastic rise [– 1, – 0.5) (Fig. 10; Table 6). Under the natural development scenario, the regions with the decreased habitat quality values were located north and south of Zhongwei, whereas the regions with increased habitat quality values were located in the central part of Shapotou District and the eastern part of Haiyuan County. Under the cultivated land protection scenario, the overall habitat quality in the study area greatly declined, especially in the eastern part of Shapotou District and the northern part of Zhongning County. Under the ecological protection scenario, the regions with decreased habitat quality values were the least abundant, and the regions with increased habitat quality values were concentrated in the central part of Shapotou District.

Fig. 10
figure 10

Spatial distribution of regions differing in habitat quality under different scenarios. The figure was generated using ArcGIS 10.6 (https://www.esri.com/zh-cn/arcgis/products/arcgis-desktop/overview).

Full size image
Table 6 Area percentage (%) of regions with defined habitat quality levels under different scenarios.
Full size table

Discussion

Causes of land use change

Land use change is not only a key driver of change in ecological service value, but also a significant source of ecological risks28. Overall, the changes in land use types in Zhongwei are influenced by many factors, such as precipitation, temperature, population density, and distance from government facilities (Fig. 4). As Zhongwei has an arid and semiarid climate, precipitation has a continuous and wide-ranging impact on the expansion of all land use types, and these impacts directly affect the abundance of water resources, especially the expansion of farmland and woodland. Considering the impact of precipitation, water management strategies to increase water conservation in the agricultural areas of Zhongwei are essential, such as the construction of reservoirs and the improvement of irrigation systems29. Temperature affects the expansion of primarily farmland, woodland and watershed. The thermal preferences of specific plant species and the influence of climate on ecological succession are indicated. Temperature affects not only plant growth cycles and planting decisions but also the evaporation and recharge of watershed, thus indirectly influencing the distribution of land use types8. Population density has emerged as a key driver of changes to farmland, watershed, and unused land, indicating that human settlement patterns and demographic pressures directly shape land use transitions. In recent years, population growth in Zhongwei has increased the demand for residential and commercial land, further driving the expansion of construction land30. Distance from government facilities similarly affects multiple land use types, reflecting policy interventions, infrastructure development, and the allocation of land resources in proximity to administrative hubs. Especially when considering the development of construction land and unused land, government policies and infrastructure developments are often decisive factors. Distance from government facilities should be leveraged to enhance agricultural services such as research and extension services, making them more accessible to farmers and fostering innovation in sustainable farming practices31. Notably, distance to railways is linked to grassland and unused land expansion, possibly due to transportation costs and development patterns along railway corridors. The improvement of transportation facilities not only facilitates the flow of people and improves logistics but also promotes the development and utilization of suburban areas, enhancing their economic value and attractiveness32. The DEM, a proxy for elevation and topographic features, has emerged as a notable factor influencing grassland expansion. This can be attributed to the suitability of certain elevations for grass species growth and the associated ecological conditions. Conversely, GDP, an indicator of economic activity, strongly influences woodland expansion, likely due to afforestation policies and investment in forestry driven by economic growth and environmental sustainability goals. Soil type, slope, and aspect relatively less impact on land use changes across different categories. This suggests that while the biophysical properties of land affect land use change, their effects in shaping contemporary land use patterns are perhaps overridden by those of more dominant socioeconomic and climatic factors. In total, land use change in Zhongwei is the result of a combination of ecological and socioeconomic factors.

Relationship between habitat quality change and land use change

The distribution of land use types is highly correlated with the spatial pattern of habitat quality values33. The changes in land use and habitat quality in Zhongwei reflect the complex relationship between the ecological protection and economic development of this region. From 1980 to 2020, the land use changes in the region had a remarkable impact on habitat quality. The continuous increase in the HDI and decrease in HQI indicated that this region experienced a slight overall decline in habitat quality. However, improvements in ecological quality have been observed in localized areas, such as Shapotou District (Figs. 5 and 6). The simulation results indicate a trend of improvement in habitat quality in Zhongwei (Figs. 8 and 9). The area of regions with low habitat quality values has continued to increase, a growth that aligns with the expansion of urban land and construction land. Therefore, the expansion of construction land is the primary cause of the decline in habitat quality, reflecting the rapid urbanization and industrialization of the study area. This is attributed to the reduction in farmland occurring alongside the rapid expansion of construction land, reflecting rapid urbanization and industrialization. In addition, the area of farmland decreased significantly between 2000 and 2020, as a result of policies of returning farmland to woodland and grassland. While these policies, are committed to restoring ecosystem services such as soil and water conservation and carbon fixation, they may have short-term impacts on the food supply. Protection therefore needs to be pursued in a number of ways. First, strict policies should be implemented to limit urban sprawl, ensuring that land conversion for nonagricultural purposes is strictly regulated and that comprehensive environmental impact assessments are conducted. Implementing land use zoning plans that prioritize farmland and designate specific areas for urban development can effectively curb urban sprawl34,35,36. Zhongwei, renowned for its efforts in sand prevention and desertification control, has conducted many relevant projects in Shapotou District and eastern Zhongning County and along the Yellow River. The applied measures have included planting trees and grasses, constructing windbreak and sand-fixing forest systems, and installing sandproof barriers, which have effectively controlled the speed of desertification. The success of these projects can be embodied by the increase in the area of woodland and the improvement in habitat quality in these regions. However, the overall decline trend in habitat quality reflects the fact that ecological restoration measures have not yet completely offset other negative impacts.

Relationship between habitat quality and land use change in multi-scenario modeling

Habitat quality and land use structure were projected to change to varying degrees in different scenarios. Under the natural development scenario (NDS), further expansion of construction land led to a decline in habitat quality north and south of Zhongwei, as urbanization and industrialization continued to advance. Habitat quality showed improved in the central part of Shapotou District and the eastern part of Haiyuan County, where the government should continue strengthening ecological protection and restoration. Under the cultivated land protection scenario (CLPS), the area of farmland increased, whereas the areas of woodland, grassland, construction land and unused land decreased, compared with those under the NDS. Such land use restructuring resulted in an overall decline in habitat quality, especially in the eastern part of Shapotou District and the northern part of Zhongning County. This might be attributed to the overemphasis on farmland protection at the expense of other land use types, leading to an imbalance in the ecosystem. Our findings contrast with those of other studies37,38 reporting cultivated land protection scenario are more appropriate for local development. However, the difference can be attributed to the varying scales and contexts of the studies, as well as the specific land use policies implemented in the respective regions. Under the ecological protection scenario (EPS), the areas of woodland, grassland and watershed increase the most; the decrease in the area of farmland is less than that under the cultivated land protection scenario (CLPS); and construction land area is greater than that under the CLPS. The habitat quality values are better under the ecological protection scenario (0.35) than those under the natural development scenario (0.34) and the cultivated land protection scenario (0.33). Habitat degradation values are lower under the ecological protection scenario (0.210) than those under the natural development scenario (0.214) and the cultivated land protection scenario (0.235). These findings indicate that the EPS is able to balance industrial expansion, arable land and ecological protection; control the growth of industrial and urban land use; and ensure good habitat quality. Overall, the EPS is the most suitable scenario for sustainable development in Zhongwei. These findings align with those reported by previous studies39,40 that emphasized the importance of integrated land use planning in maintaining good habitat quality. Our study contributes to this field of study by analyzing specific scenarios tailored to the local context of Zhongwei, thereby offering practical insights for policy-makers. However, some studies41,42 have argued that the expansion of ecological land use has come at the expense of construction land, hindering social progress. The optimal development pattern for different regions changes over time in response to changes in local economic and ecological conditions.

Suggestions

The greatest problem facing Zhongwei is environmental deterioration caused by the large-scale conversion of farmland into construction land, which requires more effort in terms of farmland protection. Under the cultivated land protection scenario, the decrease in the area of farmland was effectively mitigated, and it is recommended that farmland protection policies continue to be promoted and improved. Under the ecological protection scenario, the expansion of woodland, grassland and watershed significantly improved habitat quality; protection efforts are recommended to enhance in ecologically sensitive areas and more ecological restoration projects are suggested. Given that climatic and environmental factors have an impact on land use change, adaptive management measures, such as adjusting planting structure, are recommended to increase the resilience of agricultural systems to climate change.

Conclusions

From 1980 to 2020, the areas of construction land and woodland increased significantly, whereas the areas of watershed and unused land decreased slightly, and that of the other land use types remained essentially unchanged. In general, precipitation, temperature, population, and distance from government facilities significantly impacted land use change in Zhongwei. Under the natural development scenario, construction land continued to expand, the area of farmland decreased, and the area of woodland increased. The ecological conservation scenario was the most effective scenario for improving habitat quality and sustainable development in Zhongwei; the cultivated land protection scenario substantially increased the area of ecological land use and the protection of farmland and produced the best habitat quality and the lowest degree of habitat degradation. In contrast, the cultivated land protection scenario, despite improving farmland protection, failed to uniformly enhance habitat quality. These findings underscore the critical importance of strategic land use planning in Zhongwei. The ecological protection scenario has emerged as the preferred approach, aligning with sustainability goals by safeguarding natural ecosystems and enhancing habitat quality. Our findings have important implications for future land use planning and sustainability in Zhongwei. The implementation of ecological protection measures should be prioritized in order to mitigate habitat degradation and improve habitat quality. This approach will contribute to the sustainable development of Zhongwei by ensuring the protection and expansion of woodland, grassland, and watershed while also addressing the threats faced by farmland. We recommend that policy-makers and planners consider these findings when developing effective land use strategies that balance economic growth with ecological conservation.