Next Article in Journal
Age Sustainability in Smart City: Seniors as Urban Stakeholders in the Light of Literature Studies
Previous Article in Journal
Residents’ Perception of Flood Prediction Products: The Study of NASA’s Satellite Enhanced Snowmelt Flood Prediction
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Sustainability
Volume 17
Issue 14
10.3390/su17146332
sustainability-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Global Warming-Driven Changes in the Suitable Habitat of Ostryopsis davidiana (Betulaceae) Shrubs

by
Huayong Zhang
1,2,*,
Xinxing Cui
1,
Yihe Zhang
3,
Zhongyu Wang
1 and
Zhao Liu
2
1
Research Center for Engineering Ecology and Nonlinear Science, North China Electric Power University, Beijing 102206, China
2
Theoretical Ecology and Engineering Ecology Research Group, School of Life Sciences, Shandong University, Qingdao 250100, China
3
School of Engineering, RMIT University, P.O. Box 71, Bundoora, VIC 3083, Australia
*
Author to whom correspondence should be addressed.
Sustainability 2025, 17(14), 6332; https://doi.org/10.3390/su17146332
Submission received: 15 May 2025 / Revised: 2 July 2025 / Accepted: 7 July 2025 / Published: 10 July 2025
Browse Figures
Versions Notes

Abstract

Ostryopsis davidiana shrubs, widely distributed in northern China, have been significantly affected by global warming. Based on the current geographical distribution data of O. davidiana in China, this study used climate data, soil data, topographic data, human activity data, and the “biomod2” integrated model to conduct an integrated study on the suitable habitat of O. davidiana under the current scenario and three future climate scenarios (SSP126, SSP370, and SSP585). The results showed the following: (1) The suitable habitats of O. davidiana are mainly concentrated in the northwest and north China regions, accounting for about 9.09% of the national area, centered in Shanyin County, Shuozhou City, Shanxi Province. (2) The suitable habitats of O. davidiana are mainly influenced by temperature and precipitation, with precipitation of wettest quarter (Bio16), isothermality (Bio3), and maximum temperature of warmest month (Bio5) being the key driving factors, with contribution rates of 25.69%, 24.31%, and 14.45%, respectively. (3) Under the three future climate scenarios, the suitable habitats of O. davidiana are expected to contract significantly, with only the low suitability areas expanding, while the rest would be contracting, showing a trend of losing most of their original habitat. The centroid of the suitable habitat would be shifting westward, and the suitable habitats would be generally migrating to higher elevation areas. (4) Climate change reduces the aggregation of O. davidiana, leading to gradual habitat fragmentation. This study provides a theoretical basis for the conservation of O. davidiana.

1. Introduction

Over the past few decades, in the pursuit of rapid socio-economic development, human activities have intensified global warming, leading to a significant decline in the resilience and ecological benefits of ecosystems, and the global ecological environment facing tremendous pressure [1,2,3,4]. Future global climate warming may cause species to lose their original habitat and migrate to higher latitudes and elevations, further leading to the loss of biodiversity [5,6,7,8,9,10,11]. Therefore, the migration and distribution changes of vegetation under global warming have become a hot issue in the study of global vegetation distribution [12,13].
Ostryopsis davidiana Decaisne is a deciduous shrub in the Betulaceae family, genus Ostryopsis. Ostryopsis davidiana shrubs have unique biological characteristics and ecological values. Ostryopsis davidiana is cold-resistant, drought-resistant, and adaptable to poor soils, commonly found on slopes at elevations of 800–2400 m. It is a dominant shrub in the Loess Plateau and also seen under mixed forests and Pinus tabulaeformis Carrière forests. It is a Chinese endemic and excellent soil-protecting shrub [14,15,16,17]. When the original forests and shrubs are destroyed, O. davidiana, due to its strong sprouting ability, can form secondary succession. Ostryopsis davidiana is essential for water conservation, soil retention, and soil fertility improvement. It contributes to both ecological protection and sustainable utilization while holding significant economic value [14]. The bark and leaves of O. davidiana can be used to extract tannin, the seeds can be used to make soap, the branches can be used to make durable agricultural tools, and it is also an excellent stock feed [14,18]. Due to the large-scale human destruction and severe soil erosion in the northwest region, its distribution range has gradually decreased. Therefore, the protection of O. davidiana shrubs is of great significance for the greening of wastelands, vegetation restoration, and soil and water conservation in the northwest region. Although recent studies have focused on the community characteristics, ecological evaluation, and resource development of O. davidiana, there are few studies on the suitable habitat of O. davidiana in response to global warming [18,19].
Species distribution models (SDMs) play an important role in the study of vegetation suitable habitats. Based on species distribution data (response variables) and environmental variable data (explanatory variables), SDMs estimate the ecological niche requirements of species, thereby predicting the suitable habitats of species and assessing the contribution of environmental variables. They have been widely used in the study of the impact of environmental changes on species and the evaluation of species habitat suitability [20,21,22]. Some scholars have pointed out that using a single model to predict the suitable habitat of species makes it difficult to obtain accurate results, while integrated models can increase the accuracy of results by combining the results of multiple models, and studies have demonstrated their advantages [23,24,25,26]. The “biomod2” platform based on R language includes many commonly used species distribution models and is currently the most mature multi-model platform [27,28,29]. This model has high accuracy and reliability in predicting species distribution and can provide strong support for ecological research and biodiversity conservation [30,31,32].
Based on the geographical distribution data of O. davidiana in China, this study used climate, topographic, and soil data and selected the current scenario and three future climate scenarios to conduct an integrated study on O. davidiana using the “biomod2” model, aiming to (1) calculate the current suitable habitat of O. davidiana and identify the main driving factors affecting its suitable habitat, (2) predict the contraction and expansion of the suitable habitat of O. davidiana in the future, (3) calculate the centroid shift and elevation conversion of the suitable habitat of O. davidiana, and (4) assess the habitat fragmentation of O. davidiana under different climatic scenarios. This study provides a theoretical basis for the protection of vegetation biodiversity and the restoration of ecosystem structure and function under global warming stress.

2. Data Integration and Calculation Methods

2.1. Data Collection and Processing

The provincial boundary data used in this study were obtained from the Institute of Resource and Environmental Sciences and Data Center of the Chinese Academy of Sciences (www.resdc.cn). The O. davidiana shrub data were sourced from the Vegetation Map of China (1:1,000,000) published by Science Press. We unified the community distribution point data to a spatial resolution of 5 km × 5 km, retaining only one data point in each 5 km × 5 km grid, resulting in a total of 2170 O. davidiana shrub distribution point data (Figure 1).
We selected 41 environmental factors (including 19 bioclimatic factors, 18 soil factors, 3 topographic factors, and 1 human activity factor) as the initial environmental factors. Among them, the 19 bioclimatic factors all came from WorldClim (www.worldclim.org), with a spatial resolution of 5 km × 5 km. Soil and topographic data used the 18 surface soil data and 3 topographic factor data provided by the World Soil Database (www.fao.org), with a spatial resolution of 5 km × 5 km. Human activity data were obtained from the Global Human Interference Dataset version 3 at a resolution of 1 km (www.earthdata.nasa.gov).
We used the second-generation National (Beijing) Climate Center Climate System Model (BCC-CSM2-MR) as the future climate system model and used the three shared socio-economic pathways (SSPs) of SSP126, SSP370, and SSP585 under different greenhouse gas emission concentrations and social development levels for model simulation and prediction. In terms of greenhouse gas emission concentrations, for climate scenario SSP126, the concentration of CO2 in 2100 will be about 421 ppm, the concentration of methane will be less than 2000 ppb, the concentration of nitrous oxide will be around 334 ppb, and the radiative forcing will reach 2.6 W/m2 in 2100, which is a low-emission scenario. For climate scenario SSP370, the CO2 concentration in 2100 will be about 692 ppm, methane concentration will reach 1887 ppb, nitrous oxide concentration will be around 359 ppb, and the radiative forcing will reach 7.0 W/m2 in 2100, which is a medium–high emission scenario. As for the climate scenario SSP585, the concentration of CO2 in 2100 will be around 868 ppm, the concentration of methane will be around 2248 ppb, the concentration of nitrous oxide will be around 357 ppb, and the radiative forcing will reach 8.5 W/m2 in 2100, which is a high emission scenario. In terms of the level of social development, for the SSP126 scenario, the economic development will be green, low-carbon, and sustainable; the global population growth rate will slow down; and renewable energy and clean energy technologies will develop rapidly. In the SSP370 scenario, economic development will be slower, population will continue to grow, technological progress will be relatively slow, and there will be insufficient incentives for energy transition and the research and development of emission reduction technologies. In the SSP585 scenario, economic development will follow the traditional fossil energy-based development path, population will continue to grow, and there will be a relative lack of investment in research and development of renewable energy and emission reduction technologies. These three climate scenarios were chosen to better model future climate scenarios. In each climate scenario, we selected the three time periods of 2041–2060, 2061–2080, and 2081–2100, with a resolution of 5 km [33,34].
Many scholars have studied the changes of species’ suitable areas based on the spatial resolution of 5 km × 5 km, among which there are many endemic species in the Qinghai–Tibet Plateau, which contains many mountainous areas, indicating the reliability of species research under this resolution [35]. Therefore, we used ArcGIS 10.8 software to carry out mask extraction, clipping, resampling, and projection for all environmental factors, so that their resolution matched the resolution of species point data and was unified to 5 km × 5 km for subsequent analysis and modeling.
The contribution of environmental factors reflects the relative importance of each factor to the prediction of species distribution. Multicollinearity can lead to problems such as overestimation or underestimation of the contribution of environmental factors and significant prediction bias in the model. To avoid overfitting of models due to high collinearity among environmental factors, we have eliminated some factors with strong correlations to other factors. We imported all variables into “biomod2”, selected environmental factors with contribution rates greater than zero, and performed a Pearson correlation analysis on the environmental factors using R language. Variables with a correlation coefficient greater than 0.8 were removed, keeping the one with the higher contribution to the models based on jackknife importance values. Finally, we retained 11 environmental factors for subsequent modeling.

2.2. Model Integration

We selected ten individual models (RF, MaxEnt, FDA, GLM, GBM, CTA, ANN, MARS, XGBOOST, and SRE) in “biomod2” to calculate the distribution of O. davidiana. First, we imported the distribution point data and corresponding environmental variables into “biomod2” and then randomly selected 75% of the 2170 distribution points of O. davidiana as the training set and 25% as the test set. To avoid errors caused by single modeling, each model performed the above process 15 times, and a total of 180 modeling results were finally generated. We integrated each round of single models through a weighted average algorithm to obtain an integrated model, which was then used to calculate the suitable habitat of O. davidiana (Table 1).
The importance of each environmental factor used in modeling was analyzed and evaluated using the “biomod2 4.2.4” software package. Model accuracy was assessed using the Area Under the Curve (AUC) of the Receiver Operating Characteristic (ROC), True Skill Statistic (TSS), and Kappa statistic [36]. Generally, models with AUC > 0.8, TSS > 0.7, and Kappa > 0.85 are considered to have good accuracy [37,38].

2.3. Classification of Suitable Habitat

We used the reclassification tool in ArcGIS 10.8 software and the natural break classification method to grade the simulation results. The natural break classification method is a data classification method based on the distribution characteristics, which can divide the data into different levels according to the natural breakpoints of the data, so that the data within each level has high similarity, and the differences between different levels are more obvious. According to this method, we divided the simulation results into four levels, namely not suitable areas (0–0.2), less suitable areas (0.2–0.4), moderately suitable areas (0.4–0.6), and highly suitable areas (0.6–1). Among them, less suitable areas, moderately suitable areas, and highly suitable areas together constitute the total suitable areas, and these three levels of areas represent the geographical range where O. davidiana is suitable for survival to different degrees, providing an important reference for subsequent research and conservation work [39].

2.4. Identification of Driving Factors

In order to explore the key factors affecting the geographical distribution of O. davidiana communities in depth, we selected the 11 environmental variable data involved in the modeling process, which covered climate; soil; topography; and other aspects, such as temperature, precipitation, soil type, slope, etc., and they may have different degrees of impact on the growth and distribution of O. davidiana (Table 2). The jackknife test option was used to obtain the percentage contribution of each environmental variable. We sorted these contribution rates and selected the top three environmental variables with the highest contribution rates, determining them as the main driving factors for the distribution of O. davidiana. These driving factors play a key role in the growth and spread of O. davidiana, and their changes may significantly affect the distribution range and population density of O. davidiana. To more intuitively display the relationship between these driving factors and the distribution of O. davidiana, we drew the corresponding response curves. By observing the response curves, we can analyze the trend of O. davidiana distribution probability under different environmental variable value ranges. Furthermore, we took the variable range corresponding to a probability exceeding 0.5 in the response curve as the range most suitable for the survival of O. davidiana community species. This range determination clarifies the best environmental conditions for O. davidiana growth, which helps us to choose suitable areas for planting and protecting O. davidiana in ecological protection and vegetation restoration work, improving its survival rate and population stability.

2.5. Contraction and Expansion of Suitable Habitat

In the study of the dynamic changes in the suitable habitat of O. davidiana, we used the grid data of the current and future suitable habitats after reclassification and used the grid conversion tool to accurately calculate the area of each suitable habitat. Grid data is a data representation form that divides geographical space into regular grids, and each grid unit stores corresponding attribute information, which represents different levels of suitable habitats in this study. Through the grid conversion tool, we can extract and convert the attribute information in the grid data to calculate the size of each suitable habitat area. By comparing the current suitable habitat area with the future suitable habitat area, calculating the difference between the two, and then calculating the ratio of this difference to the current suitable habitat area, we can finally obtain the proportion of contraction and expansion of the suitable habitat of O. davidiana under different scenarios. This proportion can intuitively reflect the changing trend of the suitable habitat of O. davidiana under different environmental scenarios and is an important indicator to assess the future survival status and development potential of O. davidiana population.

2.6. Calculation of Centroid and Elevation Changes

To further analyze the spatial changes in the suitable habitat of O. davidiana, we used the powerful spatial statistical and spatial analysis tools of ArcGis to accurately calculate the centroid position and average elevation of the suitable habitat under each scenario. By calculating the centroid position of the suitable habitat under different scenarios and comparing it with the centroid position of the current suitable habitat, we can clearly observe the moving trend of the suitable habitat of O. davidiana in the horizontal direction. At the same time, we also calculated the average elevation of the suitable habitat, which helps us understand the vertical distribution changes of O. davidiana. The change in the average elevation of the suitable habitat of O. davidiana can reflect its response to climate change or topographic changes. By comparing the average elevation under different scenarios with the average elevation of the current suitable habitat, we can analyze the upward or downward trend of O. davidiana in the vertical direction, providing an important reference for the study of the ecological adaptability and distribution pattern of O. davidiana.

2.7. Calculation of Habitat Fragmentation

We used “FragStats 4.2” software to assess fragmentation in suitable areas of O. davidiana. The distribution rasters for the different climate scenarios were reclassified to retain only two categories, suitable areas and unsuitable areas. Subsequently, we imported the acquired raster files into FragStats software. At the class level, two of the most commonly used metrics were selected for analysis: the number of patches (NP) and the Aggregation Index (AI). NP indicates the number of patches, and an increase in the number of patches implies an increase in community habitat fragmentation. The AI is a quantification of community aggregation, and as the AI value decreases, the community becomes more fragmented [40]. These metrics effectively responded to the landscape pattern and helped us to analyze the fragmentation process of O. davidiana.

3. Results

3.1. Current Suitable Habitat of O. davidiana

The current suitable habitat of O. davidiana is mainly located in Inner Mongolia, Liaoning, Hebei, Beijing, Shaanxi, Shanxi, Ningxia, and Gansu, with some distribution in Henan, Jilin, Sichuan, Tibet, and Xinjiang. The suitable habitat area accounts for 9.09% of the national area, with Shanyin County, Shuozhou City, Shanxi Province as the center, among which the highly suitable area accounts for 41.87% of the suitable habitat area, the moderately suitable area accounts for 26.89% of the suitable habitat area, and the less suitable area accounts for 31.24% of the suitable habitat area (Figure 2).

3.2. Driving Factors

Among all environmental variables involved in calculating of the suitable habitat of O. davidiana, climatic factors play a dominant role. Among them, precipitation of the wettest quarter (Bio16), isothermality (Bio3), and maximum temperature of the warmest month (Bio5) have contribution rates of 25.69%, 24.31%, and 14.45%, respectively. The cumulative contribution rate of these three climatic factors is 64.45%, and they can be considered as the main driving factors affecting the suitable habitat of O. davidiana (Figure 3). The suitable range of precipitation of the wettest quarter (Bio16) is 211.47 mm–354.19 mm. The suitable range of isothermality (Bio3) is 27.25–32.35. The suitable range of maximum temperature of the warmest month (Bio5) is 21.68 °C–28.55 °C (Figure 4).

3.3. Contraction and Expansion of Suitable Habitat

Regionally, except for the expansion of suitable habitats in Qinghai, Xinjiang, and Tibet, three provinces with medium and high elevations, the suitable habitats in other provinces are contracting. Most of the highly suitable areas in Inner Mongolia, Hebei, Shaanxi, and Shanxi have been lost, becoming les suitable areas and unsuitable areas, and Henan, Sichuan, and Jilin have become unsuitable areas (Figure 5, Figure 6 and Figure 7).
Under the SSP126 climate scenario, the total suitable habitat for O. davidiana will contract by, at least, 18.43% in the 2050s, 19.85% in the 2070s, and 40.13% in the 2090s, compared to the present. Under the SSP370 scenario, the total suitable habitat will contract by 41.84% in the 2050s and by a further 46.64% and 33.19% in the 2070s and 2090s, respectively. Under the SSP585 climate scenario, O. davidiana will experience the greatest contraction in total suitable habitat: 34.48% in the 2050s and a further 49.18% and 55.35% in the 2070s and 2090s, respectively, compared to the present.
In the future, both the moderately suitable areas and the highly suitable areas of O. davidiana will contract, while the less suitable areas will gradually expand, especially in the SSP370 scenario, where the expansion of the less suitable areas is the most obvious. Under SSP370-2090s, the expansion of the les suitable areas is the largest, with an increase of 95.46%, while under SSP585-2090s, the contraction of the moderately suitable areas is the largest, with a contraction of 94.26%. Under the SSP585 scenario, the contraction of the highly suitable areas of O. davidiana is the most obvious. The contraction of the highly suitable areas is greatest when the climate scenario is SSP585-2090s and will contract by 99.71% (Figure 8).

3.4. Centroid Migration and Elevation Changes

In terms of spatial pattern, the centroid of the current suitable habitat of O. davidiana is located in Shanyin County, Shuozhou City, Shanxi Province, with an average elevation of about 1487 m. Under the background of future global warming, the centroid of the suitable habitat of O. davidiana has a trend of migrating to the west and high-elevation areas, and with the increase in emission concentration, the centroid migrates further and the average elevation rises higher (Figure 9 and Figure 10). Under the SSP370 and SSP585 scenarios, the centroid migration distance and average elevation of the suitable habitat of O. davidiana are much greater than those under the SSP126 scenario. When the climate scenario is SSP370-2090s, the centroid of the suitable habitat of O. davidiana migrates the farthest, with the highest average elevation. At this time, the centroid of the suitable habitat of O. davidiana is located in Minqin County, Wuwei City, Gansu Province, with a straight-line distance of about 900 km from the current centroid of the suitable habitat and an average elevation of about 2883 m.

3.5. Fragmentation Assessment of O. davidiana

The suitable areas of O. davidiana show a large degree of fragmentation. Under the current climate scenario, O. davidiana exhibits the lowest degree of fragmentation, the lowest number of patches, and the highest level of aggregation. In future climate scenarios, as carbon emissions increase, the habitat aggregation level for O. davidiana is expected to decrease, reaching the lowest levels during the 2081–2100 period under the SSP585 scenario. The number of patches in the suitable areas increases by 309% compared to current levels, while the aggregation level decreases by 24%, representing the largest proportional change among O. davidiana, during the 2081–2100 period under the SSP585 scenario. Under the SSP126 climate scenario, the number of patches in the suitable areas is expected to increase, while changes in aggregation levels remain relatively insignificant. However, under the SSP370 and SSP585 scenarios, both the number of patches and the aggregation levels show a decline in the landscape quality of its suitable areas (Table 3).

4. Discussion

The Vegetation Map of China (1:1,000,000) was created using a large amount of fieldwork data collected between 1954 and 1980. During this period, a comprehensive nationwide survey was conducted to map the country’s vegetation. The map covers almost all of China’s provinces, municipalities, and autonomous regions, providing a comprehensive overview of the distribution of vegetation types across the country. Many experts in vegetation science compiled and reviewed the map, strictly checking and verifying the data to ensure its scientific validity and accuracy in terms of vegetation classification and distribution range. Despite its wide coverage, there may be data scarcity or sampling bias in some special areas or areas strongly influenced by human activities due to factors such as difficulty obtaining data or human interference. This can lead to an inaccurate and incomplete understanding of vegetation distribution patterns in these areas. Vegetation is closely related to environmental factors such as climate, soil, and topography. Insufficient data or sampling bias may therefore lead to an inaccurate analysis of the relationship between vegetation and the environment.
Species distribution models (SDMs) are widely used in the study of species’ suitable habitats [41]. Different modeling algorithms have distinct advantages and limitations. For example, MAXENT is capable of fitting complex response curves even with limited occurrence data [42]. ANNs excel at processing complex, nonlinear relationships that are challenging to represent via traditional methods [43]. RF combines ensemble learning and parallel computing to achieve high computational efficiency and predictive accuracy [44]. The FDA provides robust discriminant analyses [45], whereas the SRE represents a traditional climate-envelope model with a relatively simple algorithmic structure [46]. Traditional statistical models, such as GLM and GAM, require data to be normal and independent. They are also not flexible enough to handle complex nonlinearities and higher-order interactions. The BIOME model can be used to simulate macroscopic patterns of vegetation distribution. However, it struggles to simulate species-specific distributions and dynamics. It also requires a large number of parameter estimations and makes assumptions about complex physiological and ecological processes. “biomod2” can capture complex nonlinearities and high-dimensional relationships. It is therefore better suited to dealing with large-scale, complex ecological data. It also makes fewer assumptions about data distributions. The “biomod2” focuses on modelling based on the direct relationship between species observations and environmental variables. This approach is more detailed and accurate at the species level and can provide targeted information for biodiversity conservation and species management. Current scholars mostly use single models, but single models have the disadvantage of low accuracy. We combined the predictive outputs of different models into an integrated model, which can effectively improve the predictive ability through weighted balancing [47,48,49]. The integrated model performed better compared to the single model, with AUC, TSS and KAPPA values of 0.989, 0.94, and 0.928, respectively. The integrated model increased the AUC, TSS, and KAPPA values by 20.32%, 45.96%, and 40.6%, respectively, compared with the worst-performing single model. The integrated model predicted the suitable habitat of O. davidiana to a relatively excellent level.
The accuracy of species distribution model predictions typically depends on the quality and quantity of the input data [50]. However, species distribution data are frequently affected by sampling bias and spatial autocorrelation. In regions with complex or remote ecological environments, such data are often sparse, potentially leading to underestimation or overestimation of suitable habitats. Moreover, “biomod2” assumes that habitat suitability is determined solely by environmental factors. This approach has limitations as it does not take into account ecological factors such as habitat fragmentation, interspecific competition, and predation and does not integrate other ecological processes such as dispersal limitation, biotic interactions, and habitat connectivity. And, the accuracy of “biomod2” prediction results will decrease as the prediction time grows longer. Due to the integration of multiple modelling algorithms, the ability of different models to fit and predict the data varies, as do the final combined results, depending on the parameter settings of each model. In integration methods, the determination of model weights is often subjective, which may lead to different weight assignments from different methods, thus affecting the reliability and stability of the integration results. Although the integration method can alleviate the spatial autocorrelation problem of individual models to a certain extent, the integration results may still be affected if all models involved in the integration have spatial autocorrelation. Due to the spatial heterogeneity of different regions in terms of geographic environment, ecological conditions, etc., the performance of each model in different regions may differ significantly. It may be difficult to fully take into account this spatial heterogeneity when integrating these models using the integration method, thus reducing the applicability and reliability of the integration results in different regions. When predicting the impact of climate change on vegetation, it is often necessary to combine different climate and emission scenarios. As these scenarios are inherently uncertain, different assumptions lead to different projections of future environmental conditions, which affect vegetation distribution projections. Existing models focus on macro-distribution prediction, and the balanced relationship between interpretability and precision in specific conservation practices deserves in-depth exploration. By fusing population genetics data with machine learning interpretability techniques, it is expected to realize the leap from prediction to decision support.
Our research results show that the current suitable habitat of O. davidiana is mainly located in Inner Mongolia, Hebei, Shanxi, Shaanxi, Gansu, and other places, mainly dominated by highly suitable areas. The range of highly suitable areas can be attributed to the unique climate and soil conditions of the Inner Mongolia Plateau and the Loess Plateau. Previous studies have shown that the soil in the Loess Plateau and the Inner Mongolia Plateau has a high sand content, fine and soft texture, and serious soil erosion, especially in the eastern part of the Inner Mongolia Plateau, where the risk of secondary soil salinization is relatively large due to the over-extraction of groundwater and agricultural irrigation [51,52]. Ostryopsis davidiana has strong drought and thin-soil tolerance, and its developed root system can achieve soil consolidation and water conservation in these two plateaus. At the same time, O. davidiana has both photophilic and shade-tolerant characteristics and can form dense shrubs or understory communities in a variety of light gradients, and its wide ecological niche characteristics make it dominant in the diverse habitats of the plateau. In all three scenarios, the less suitable areas of O. davidiana will expand and the highly suitable areas will decrease over time. As the less suitable areas expand, the competitiveness of O. davidiana populations decreases, allowing other plants that are better adapted to environmental changes to take advantage of the opportunity to expand and compete with O. davidiana for resources such as living space, light, water, and nutrients. This results in a gradual decrease in O. davidiana’s status within the plant community, and it may even be excluded from the community. This changes the species composition and community structure of the local area. Animals that depend on O. davidiana may face food shortages or habitat changes, resulting in population declines or range shifts. The density of O. davidiana populations may increase in reduced highly suitable areas, leading to increased competition for resources and restricted individual growth and reproductive success. Consequently, genetic diversity may decrease.
Compared to the results of Wen et al. [53], the predicted suitable habitat area for O. davidiana in this study is smaller. This discrepancy may be attributed to differences in model selection and the climate variables included, as the choice of environmental predictors can substantially affect the projected distribution and extent of suitable habitats. Furthermore, while the study by Wen et al. was based on a single model, this study integrates predictions from an ensemble of multiple models, providing a more robust and comprehensive representation of the spatial distribution of suitable habitats for O. davidiana.
Temperature is an important factor that influences the distribution of vegetation. Warming causes areas that were previously unsuitable due to low temperatures to become progressively more suitable for certain types of vegetation, resulting in vegetation migrating to higher latitudes or altitudes. Changes in precipitation directly affect the availability of water to vegetation. Increased precipitation may promote vegetation growth, while decreased precipitation may lead to increased drought stress or even vegetation mortality, thus altering the extent of vegetation distribution. Furthermore, climate change will reshape the biological relationships between different types of vegetation, such as competition, symbiosis, and predation. Some vegetation may expand its range as competitors decline, while others may contract as they lose symbiotic partners or face new biotic stresses.
A large number of studies have shown that climate is the key factor affecting the distribution of most plants [54,55,56,57,58]. Combined with the model’s prediction results, it is also found that climatic variables have the greatest impact on the suitable habitat of O. davidiana. In our study, precipitation of the wettest quarter (Bio16) is the main driving factor affecting the suitable habitat of O. davidiana in China, followed by isothermality (Bio3) and maximum temperature of the warmest month (Bio5). This result shows that suitable temperature and precipitation conditions promote the growth and expansion of O. davidiana [53]. Ostryopsis davidiana has a well-developed root system comprising a large number of horizontal and oblique roots. This helps the plant to absorb more water during periods of drought and enhances its resistance to drought. In areas such as the Loess Plateau, its root system is able to form a dense network in the soil, enabling it to absorb more water and adapt to environments with less precipitation. Yan et al. believe that a reduction in rainfall will cause O. davidiana to allocate more biomass to the underground to enhance the absorption capacity of deep soil moisture, resulting in slower growth of new branches [16]. The importance of precipitation of the wettest quarter (Bio16) in our study highlights O. davidiana’s dependence on sufficient water conditions. Although drought-resistant, sufficient precipitation is crucial for its growth and reproduction. These research results show that rainfall is the main driving factor affecting the distribution of O. davidiana in China.
The buds of O. davidiana are highly resistant to cold and can remain dormant in low winter temperatures to avoid frost damage. When temperatures rise in the spring, dormant buds sprout rapidly, allowing O. davidiana to begin growth during a brief warm period. It adapts to seasonal fluctuations in temperature in this way. Xi et al. found that areas with small temperature fluctuations are more conducive to the growth of O. davidiana because it can reduce the physiological stress caused by large temperature fluctuations and maintain stable photosynthesis and respiration [59]. Our research results show that isothermality (Bio3) reflects O. davidiana’s demand for climate stability, indicating that areas with large temperature fluctuations may limit its growth, which may be related to the potential physiological effects of extreme temperatures on plants. During periods of high temperature, O. davidiana closes its stomata in order to minimize water loss. The experimental results of Xi et al. also showed that with the increase in temperature, the photosynthetic rate of O. davidiana gradually increased, but after reaching a peak, the photosynthetic rate began to decline due to high temperature. It is not difficult to see from the response curve of the highest temperature of the warmest month (Bio5) that too high temperatures may threaten the survival of O. davidiana, and especially under the global warming scenario, this impact may be exacerbated.
In this study, none of the important environmental variables affecting O. davidiana included soil variables, due to the strong adaptability to soil and low requirements for soil conditions. Ostryopsis davidiana can form symbiosis with a variety of ectomycorrhizal fungi, improving the utilization efficiency of various nutrients, thereby helping it survive [60,61,62].
Vegetation is very sensitive to climate change, and many studies have shown that plants will migrate to adapt to changing climate conditions [51,63]. However, most plants lack the ability to migrate effectively and quickly, making it difficult for them to find suitable habitats in the face of rapid climate change [64]. Previous studies have shown that O. davidiana has a trend of migrating westward over time, which is consistent with our conclusion to a certain extent [53]. Our study found that compared with the SSP126 scenario, the centroid migration amplitude of the suitable habitat of O. davidiana under the SSP370 and SSP585 scenarios is larger, and the migration distance increases with the increase in emission concentration and time. From the vertical gradient, under the scenario with the largest migration amplitude, the average elevation rises from 1487 m to 2883 m, showing a trend of migrating to high-elevation areas. These results are consistent with the speculation that global warming leads to the migration of species distribution to high-latitude and high-elevation areas [65,66,67,68,69,70].
Habitat fragmentation significantly contributes to the decline in species diversity and the reduction in suitable habitats [71,72]. In our results, the aggregation of O. davidiana decreases significantly with increasing carbon emissions. Under the SSP370 scenario, the most significant changes are observed in patch number and aggregation. This may be attributed to the heightened volatility of climate change and the increased frequency of extreme weather events under high carbon emissions, which contribute to the instability of landscape indices [73]. The degree of fragmentation shows a consistent pattern with the habitat changes of O. davidiana. It is reasonable to hypothesize that with climate change, the originally continuous habitat of O. davidiana is gradually fragmented from the edges, the number of patches increases, and the degree of aggregation decreases. This will lead to an increase in the degree of O. davidiana fragmentation, which, in turn, will result in a reduction in suitable areas. Landscape fragmentation can divide O. davidiana habitats into isolated patches, leading to population isolation and hindering gene exchange. Populations that used to mate freely in continuous habitats can now only reproduce in isolated patches. In the long term, this can result in genetic drift and random changes in gene frequencies within small populations. This can lead to the loss of beneficial genes and a reduction in genetic diversity.
The seeds of O. davidiana are nutlets, so wind dispersal of seeds is largely ruled out, and seed dispersal may be accomplished by gravity and small mammals [18,60]. Animal transmission is limited by the range and behavior of animals and usually does not cross long distances. This tends to trigger geographic isolation, which further may affect genetic diversity. Natural seed dispersal is a slow process that requires sufficient time for the seeds to grow and establish populations in the new environment. In a short period of time, it is difficult for O. davidiana seeds to complete long-distance dissemination and successful colonization. However, with the assistance of human activities, such as artificial cultivation, O. davidiana may appear in more distant areas in a relatively short time, thus creating new opportunities for its survival and reproduction. Although artificial propagation reduces geographic isolation to some extent, there is a risk of genetic loss if cultivars from a single source are used for introgression. Therefore, a genetic diversity monitoring system must be established and a mixed seed source strategy prioritized in order to promote the spread of species while maintaining their genetic integrity.
Habitat fragmentation disrupts the continuity of species’ living spaces, posing a major threat to biodiversity and genetic diversity [74]. Furthermore, fragmentation of O. davidiana habitats can result in reduced water retention capacity, increased soil erosion and diminished ecosystem stability and services [75]. To better protect O. davidiana habitats, we recommend designating protected areas in regions where the species is densely distributed, such as the Loess Plateau. Habitats that have been significantly fragmented could be artificially reseeded. At the same time, ecological corridors should be planned and constructed. Prioritize connecting the core and marginal habitats of O. davidiana, particularly those of isolated populations. Linking these areas will facilitate gene flow and communication between populations, thereby increasing their genetic diversity and resilience. Long-term dynamic monitoring is important for understanding the evolution of O. davidiana habitats as it provides a scientific basis for developing precise and effective conservation strategies [76]. Advances in science and technology have made it possible to obtain more accurate spatial information using high-resolution satellite remote sensing imagery and unmanned aerial vehicle (UAV) aerial photography techniques [77]. Machine learning and artificial intelligence algorithms can also be used to better identify the complex mechanisms of association between influencing factors and improve the accuracy and reliability of habitat change predictions [78].

5. Conclusions

Ostryopsis davidiana shrubs are an important shrub type in northern China, and they have important values for the protection of biodiversity and the maintenance of ecosystem functions. In this paper, we used the “biomod2” model to determine the key driving factors affecting the suitable habitat of O. davidiana under the current scenario and three future climate scenarios (SSP126, SSP370, and SSP585), and we studied the distribution of O. davidiana in China and predicted its contraction and expansion, as well as the centroid migration and elevation conversion of its suitable habitat. The research results show the following: (1) Under the current scenario, O. davidiana is mainly distributed in most provinces of North and Northwest China, with highly suitable areas concentrated in Inner Mongolia, Liaoning, Hebei, Beijing, Shaanxi, Shanxi, Ningxia, and Gansu. (2) Precipitation of the wettest quarter (Bio16), isothermality (Bio3), and maximum temperature of the warmest month (Bio5) are the key driving factors affecting the geographical distribution of O. davidiana. (3) Under the three future climate scenarios, the suitable habitat of O. davidiana is expected to contract severely, with only the less suitable area expanding. The centroid of the suitable habitat has a trend of migrating to the west and high-elevation areas, and with the increase in emission concentration, the centroid migrates further and the average elevation rises higher. (4) We found that as carbon increases, the previously continuous habitats gradually fragment from the edges. Declining aggregation reduces the stability of O. davidiana, negatively impacting the development of O. davidiana. This paper provides theoretical support for the protection of O. davidiana in China.

Author Contributions

Conceptualization, H.Z.; methodology, H.Z.; software, X.C.; validation, H.Z., Y.Z., and Z.W.; formal analysis, X.C.; investigation, H.Z. and X.C.; resource, H.Z.; data curation, X.C.; writing—original draft preparation, H.Z. and X.C.; writing—review and editing, H.Z., X.C., Y.Z., Z.W., and Z.L.; visualization, X.C.; supervision, H.Z.; project administration, H.Z. and Z.L.; funding acquisition, H.Z. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the National Water Pollution Control and Treatment Science and Technology Major Project (2017ZX07101) and the Discipline Construction Program of Huayong Zhang, Distinguished Professor of Shandong University, School of Life Sciences (61200082363001).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

All links to input data are reported in the manuscript, and all output data are available upon request to the authors.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Messier, C.; Bauhus, J.; Doyon, F.; Maure, F.; Sousa-Silva, R.; Noler, P.; Mina, M.; Aquilue, N.; Fortin, M.-J.; Puettmann, K. The functional complex network approach to foster forest resilience to global changes. For. Ecosyst. 2019, 6, 21. [Google Scholar] [CrossRef]
  2. Barreto, J.R.; Pardini, R.; Metzger, J.P.; Silva, F.A.B.; Nichols, E.S. When forest loss leads to biodiversity gain: Insights from the Brazilian Atlantic Forest. Biol. Conserv. 2023, 279, 109957. [Google Scholar] [CrossRef]
  3. Li, G.; Wen, Z.; Guo, K.; Du, S. Simulating the Effect of Climate Change on Vegetation Zone Distribution on the Loess Plateau, Northwest China. Forests 2015, 6, 2092. [Google Scholar] [CrossRef]
  4. Naranjo, A.A.; Melton, A.E.; Soltis, D.E.; Soltis, P.S. Endemism, projected climate change, and identifying species of critical concern in the Scrub Mint clade (Lamiaceae). Conserv. Sci. Pract. 2022, 4, e621. [Google Scholar] [CrossRef]
  5. Cong, M.; Li, Y.; Yang, W. Potential Distribution of Bryophyte, Entodon challengeri (Entodontaceae), under Climate Warming in China. Diversity 2023, 15, 871. [Google Scholar] [CrossRef]
  6. Gan, Y.; Cheng, L.; Tang, J.; Han, H.; Gan, X. Evaluating the vulnerability of Tetracentron sinense habitats to climate-induced latitudinal shifts. Ecol. Evol. 2024, 14, e11710. [Google Scholar] [CrossRef]
  7. Huang, P.; Xiao, Y.; Sun, Y.; Huang, H.; Gong, Z.; Zhu, Y. Distribution changes of Ormosia microphylla under different climatic scenarios. Sci. Rep. 2025, 15, 2607. [Google Scholar] [CrossRef]
  8. Wang, Z.J.; Gai, Z.S.; Yang, T.G.; Zhai, J.T.; Wu, Z.H.; Guo, X.F.; Chen, X.X.; Sun, J.H.; Jiao, P.P.; Li, Z.J. Prediction of The Distribution Area and Environmental Variables of Populus euphratica and Populus pruinosa in China Based on the optimized Maxent Model. Appl. Ecol. Environ. Res. 2023, 21, 3223–3241. [Google Scholar] [CrossRef]
  9. Yan, Q.; Zhang, X.; Li, M.; Zhang, G.; Huang, L.-F. Global climate change promotes aboveground growth and high-latitude species distribution in medicinal plants: Insights from the genus Panax. J. Clean. Prod. 2024, 478, 143863. [Google Scholar] [CrossRef]
  10. Yao, W.; Wang, Z.; Fan, Y.; Liu, D.; Ding, Z.; Zhou, Y.; Hu, S.; Zhang, W.; Ou, J. Prediction of Potential Habitat Distributions and Climate Change Impacts on the Rare Species Woonyoungia septentrionalis (Magnoliaceae) in China Based on MaxEnt. Plants 2024, 14, 86. [Google Scholar] [CrossRef] [PubMed]
  11. Yuan, Q.; Zhang, J.; Yao, Z.; Zhou, Q.; Liu, P.; Liu, W.; Liu, H. Prediction of potential distributions of Morina kokonorica and Morina chinensis in China. Ecol. Evol. 2024, 14, e11121. [Google Scholar] [CrossRef]
  12. Chawengkul, P.; Tiwutanon, P.; Sanevas, N.; Kraichak, E. Predicting the Future Distribution of Leucobryum aduncum under Climate Change. Diversity 2024, 16, 125. [Google Scholar] [CrossRef]
  13. Bao, X.; Zhou, P.; Zhang, M.; Fang, Y.; Zhang, Q. MaxEnt-Based Habitat Suitability Assessment for Vaccinium mandarinorum: Exploring Industrial Cultivation Opportunities. Forests 2024, 15, 2254. [Google Scholar] [CrossRef]
  14. Shangguan, T.; Zhang, F. On Synecological Features and biomass of Ostryopsis davidiana Bush-Wood in Yunding Mountain, Shanxi Province. J. Shanxi Univ. (Nat. Sci. Ed.) 1989, 12, 347–352. [Google Scholar] [CrossRef]
  15. Hou, J.; Li, B. Cultivation techniques of Ostryopsis davidiana, an excellent plant for desert control. For. By-Prod. Spec. China 2009, 24, 56–57. [Google Scholar] [CrossRef]
  16. Yan, W.; Han, X.; Bai, S.; Shao, D. Research on Drought Resistance Mechanism to Different Mycorrhizal Seedlings of Ostyopsis davidiana. Sci. Silvae Sin. 2006, 42, 73–76. [Google Scholar] [CrossRef]
  17. Xue, L.; Zhao, H.; Yao, Q.; Zhang, W.; Yan, W. Studies on the Tissue Culture and Mycorrhizal Technology of Ostryopsis davidiana. Chin. Agric. Sci. Bull. 2012, 28, 17–21. [Google Scholar] [CrossRef]
  18. Yao, Z.; Shi, Y.; Ma, W.; Zhao, L.; Juhua; Haoren, T.; Su, C.; Zhang, X.; Liang, C. Community characteristics of Ostryopsis davidiana alliance in Nei Mongol, China. Chin. J. Plant Ecol. 2022, 46, 1342–1349. [Google Scholar] [CrossRef]
  19. Bai, Z. Survey and Ecological Evaluation of Ostryopsis davidiana Community in Liaoning Qinglong River National Nature Reserve. Spec. Econ. Anim. Plant 2023, 26, 166–168. [Google Scholar] [CrossRef]
  20. Elith, J.; Leathwick, J.R. Species Distribution Models: Ecological Explanation and Prediction Across Space and Time. Annu. Rev. Ecol. Evol. Syst. 2009, 40, 677–697. [Google Scholar] [CrossRef]
  21. Guo, Y.; Zhao, F.; Qiao, H.; Wang, R.; Wei, H.; Wang, L.; Gu, W.; Li, X. Challenges and Development Trend of Species Distribution Model. Adv. Earth Sci. 2020, 35, 1292–1305. [Google Scholar] [CrossRef]
  22. Whitford, A.M.; Shipley, B.R.; McGuire, J.L. The influence of the number and distribution of background points in presence-background species distribution models. Ecol. Model. 2024, 488, 110604. [Google Scholar] [CrossRef]
  23. Thuiller, W.; Lafourcade, B.; Engler, R.; Araujo, M.B. BIOMOD—A platform for ensemble forecasting of species distributions. Ecography 2009, 32, 369–373. [Google Scholar] [CrossRef]
  24. Hao, T.; Elith, J.; Guillera-Arroita, G.; Lahoz-Monfort, J.J. A review of evidence about use and performance of species distribution modelling ensembles like BIOMOD. Divers. Distrib. 2019, 25, 839–852. [Google Scholar] [CrossRef]
  25. Cai, H.; Zhang, G. Predicting the Potential Distribution of Rare and Endangered Emmenopterys henryi in China Under Climate Change. Ecol. Evol. 2024, 14, e70403. [Google Scholar] [CrossRef]
  26. Cong, Y.; Gu, Y.; Wang, W.J.; Wang, L.; Xue, Z.; Chen, Y.; Jin, Y.; Xu, J.; Li, M.-H.; He, H.S.; et al. The interaction between temperature and precipitation on the potential distribution range of Betula ermanii in the alpine treeline ecotone on the Changbai Mountain. For. Ecosyst. 2024, 11, 100166. [Google Scholar] [CrossRef]
  27. Igawa, T.K.; de Toledo, P.M.; Anjos, L.J.S. Climate change could reduce and spatially reconfigure cocoa cultivation in the Brazilian Amazon by 2050. PLoS ONE 2022, 17, e0262729. [Google Scholar] [CrossRef]
  28. Park, C.; Byeon, J.; Park, J.; Lee, S.; Park, B.; Kim, J.; Woo, H. Impact of Climate Change on Species Distribution of Picea jezoensis Carrire in Baekdudaegan Using Ensemble Modeling. Sens. Mater. 2022, 34, 4629–4637. [Google Scholar] [CrossRef]
  29. Yang, M.; Zhao, H.; Xian, X.; Qi, Y.; Li, Q.; Guo, J.; Chen, L.; Liu, W. Reconstructed Global Invasion and Spatio-Temporal Distribution Pattern Dynamics of Sorghum halepense under Climate and Land-Use Change. Plants 2023, 12, 3128. [Google Scholar] [CrossRef]
  30. Gu, R.; Wei, S.; Li, J.; Zheng, S.; Li, Z.; Liu, G.; Fan, S. Predicting the impacts of climate change on the geographic distribution of moso bamboo in China based on biomod2 model. Eur. J. For. Res. 2024, 143, 1499–1512. [Google Scholar] [CrossRef]
  31. Zhao, G.; Cui, X.; Sun, J.; Li, T.; Wang, Q.; Ye, X.; Fan, B. Analysis of the distribution pattern of Chinese Ziziphus jujuba under climate change based on optimized biomod2 and MaxEnt models. Ecol. Indic. 2021, 132, 108256. [Google Scholar] [CrossRef]
  32. Zhang, H.; Zheng, S.; Huang, T.; Liu, J.; Yue, J. Estimation of Potential Suitable Habitats for the Relict Plant Euptelea pleiosperma in China via Comparison of Three Niche Models. Sustainability 2023, 15, 11035. [Google Scholar] [CrossRef]
  33. Ma, D.; Lun, X.; Li, C.; Zhou, R.; Zhao, Z.; Wang, J.; Zhang, Q.; Liu, Q. Predicting the Potential Global Distribution of Amblyomma americanum (Acari: Ixodidae) under Near Current and Future Climatic Conditions, Using the Maximum Entropy Model. Biology 2021, 10, 1057. [Google Scholar] [CrossRef]
  34. Zhang, X.; Zhao, J.; Wang, M.; Li, Z.; Lin, S.; Chen, H. Potential distribution prediction of Amaranthus palmeri S. Watson in China under current and future climate scenarios. Ecol. Evol. 2022, 12, e9505. [Google Scholar] [CrossRef]
  35. He, X.; Gao, J.; Zhu, L.; Hao, R.; Huang, L.; Zhu, J.; Cheng, L.; Zhou, J. Response of Potential Geographic Distribution of Amygdalus mongolica to Climate Change. Bull. Bot. Res. 2024, 44, 180–191. [Google Scholar] [CrossRef]
  36. Allouche, O.; Tsoar, A.; Kadmon, R. Assessing the accuracy of species distribution models: Prevalence, kappa and the true skill statistic (TSS). J. Appl. Ecol. 2006, 43, 1223–1232. [Google Scholar] [CrossRef]
  37. Wang, Y.; Liu, H.; Xu, J.; Yu, S.; Huang, Y.; Zhang, Y.; He, X.; Chen, W. Prediction of suitable planting areas of Rubia cordifolia in China based on a species distribution model and analysis of specific secondary metabolites. Ind. Crops Prod. 2023, 206, 117651. [Google Scholar] [CrossRef]
  38. Luo, M.; Wang, H.; Lyu, Z. Evaluating the performance of species distribution models Biomod2 and MaxEnt using the giant panda distribution data. Chin. J. Appl. Ecol. 2017, 28, 4001–4006. [Google Scholar] [CrossRef]
  39. Zhan, P.; Wang, F.; Xia, P.; Zhao, G.; Wei, M.; Wei, F.; Han, R. Assessment of suitable cultivation region for Panax notoginseng under different climatic conditions using MaxEnt model and high-performance liquid chromatography in China. Ind. Crops Prod. 2022, 176, 114416. [Google Scholar] [CrossRef]
  40. Guga, S.; Xu, J.; Riao, D.; Li, k.; Han, A.; Zhang, J. Combining MaxEnt model and landscape pattern theory for analyzing interdecadal variation of sugarcane climate suitability in Guangxi, China. Ecol. Indic. 2021, 131, 108152. [Google Scholar] [CrossRef]
  41. Li, R.; Hu, X.; Li, Q.; Liu, L.; He, Y.; Chen, C. Gap analysis of Firmiana danxiaensis, a rare tree species endemic to southern China. Ecol. Indic. 2024, 158, 111606. [Google Scholar] [CrossRef]
  42. Zhu, G.; Qiao, H. Effect of the Maxent model’s complexity on the prediction of species potential distributions. Biodivers. Sci. 2016, 24, 1189–1196. [Google Scholar] [CrossRef]
  43. Zhang, C.; Guo, Y.; Li, M. Review of Development and Application of Artificial Neural Network Models. Comput. Eng. Appl. 2021, 57, 57–69. [Google Scholar] [CrossRef]
  44. Yang, Q.; Wu, W.; Liu, H. Prediction of spatial distribution of soil available iron in a typical hilly farmland using terrain attributes and random forest model. Chin. J. Eco-Agric. 2018, 26, 422–431. [Google Scholar] [CrossRef]
  45. Kuemmerlen, M.; Schmalz, B.; Guse, B.; Cai, Q.; Fohrer, N.; Jähnig, S.C. Integrating catchment properties in small scale species distribution models of stream macroinvertebrates. Ecol. Model. 2014, 277, 77–86. [Google Scholar] [CrossRef]
  46. Araújo, M.B.; Peterson, A.T. Uses and misuses of bioclimatic envelope modeling. Ecology 2012, 93, 1527–1539. [Google Scholar] [CrossRef]
  47. Thuiller, W. Editorial commentary on ‘BIOMOD-optimizing predictions of species distributions and projecting potential future shifts under global change. Glob. Change Biol. 2014, 20, 3591–3592. [Google Scholar] [CrossRef]
  48. Xie, C.; Liu, C.; Wang, H.; Liu, D.; Jim, C.Y. Distribution pattern of large old Ginkgo biloba in China under climate change scenarios. Ecol. Evol. 2024, 14, e11367. [Google Scholar] [CrossRef]
  49. Gaal, M.; Moriondo, M.; Bindi, M. Modelling the Impact of Climate Change on the Hungarian Wine Regions Using Random Forest. Appl. Ecol. Environ. Res. 2012, 10, 121–140. [Google Scholar] [CrossRef]
  50. Araújo, M.B.; Guisan, A. Five (or so) challenges for species distribution modelling. J. Biogeogr. 2006, 33, 1677–1688. [Google Scholar] [CrossRef]
  51. Li, H.; Guo, J.; Wang, Y.; Wang, W.; Jia, Q.; Wan, H.; Li, F.Y. Boundary migration between zonal vegetation types in Inner Mongolia over the past two decades. Catena 2024, 246, 108354. [Google Scholar] [CrossRef]
  52. Wang, N.; He, X.; Zhao, F.; Wang, D.; Jiao, J. Soil seed bank in different vegetation types in the Loess Plateau region and its role in vegetation restoration. Restor. Ecol. 2020, 28, A5–A12. [Google Scholar] [CrossRef]
  53. Wen, G.; Ye, X.; Lai, W.; Shi, C.; Huang, Q.; Ye, L.; Zhang, G. Dynamic analysis of mixed forest species under climate change scenarios. Ecol. Indic. 2021, 133, 108350. [Google Scholar] [CrossRef]
  54. Bellard, C.; Bertelsmeier, C.; Leadley, P.; Thuiller, W.; Courchamp, F. Impacts of climate change on the future of biodiversity. Ecol. Lett. 2012, 15, 365–377. [Google Scholar] [CrossRef]
  55. Silva, T.; Whitley, B.S.; Biersma, E.M.; Abermann, J.; Raundrup, K.; de Vere, N.; Høye, T.T.; Schöner, W. Bio-climatic factors drive spectral vegetation changes in Greenland. EGUsphere 2024, 2024, 1–31. [Google Scholar] [CrossRef]
  56. Zhao, Y.; Cao, H.; Xu, W.; Chen, G.; Lian, J.; Du, Y.; Ma, K. Contributions of precipitation and temperature to the large scale geographic distribution of fleshy-fruited plant species: Growth form matters. Sci. Rep. 2018, 8, 17017. [Google Scholar] [CrossRef]
  57. Ma, Z.; Liu, H.; Mi, Z.; Zhang, Z.; Wang, Y.; Xu, W.; Jiang, L.; He, J.-S. Climate warming reduces the temporal stability of plant community biomass production. Nat. Commun. 2017, 8, 15378. [Google Scholar] [CrossRef]
  58. Zhang, C.; Lei, S.; Wu, H.; Liao, L.; Wang, X.; Zhang, L.; Liu, G.; Wang, G.; Fang, L.; Song, Z. Simplified microbial network reduced microbial structure stability and soil functionality in alpine grassland along a natural aridity gradient. Soil Biol. Biochem. 2024, 191, 109366. [Google Scholar] [CrossRef]
  59. Xi, X.; Tian, Y.; He, Y. Photosynthetic Characteristics of Ostryopsis davidiana Decaisne and Relationship with Main Influencing Factor. J. Art Coll. Inn. Mong. Univ. 2012, 9, 81–86. [Google Scholar] [CrossRef]
  60. Bai, S.; Bai, Y.e.; Fang, L.; Liu, Y. Mycorrhiza of Cenococcum geophilum (Fr.) Formed on Ostryopsis daidiana and Mycorrhizal Affection on the Growth of Ostryopsis davidiana. Sci. Silvae Sin. 2004, 40, 194–196. [Google Scholar] [CrossRef]
  61. Maharjan, S.K.; Sterck, F.J.; Raes, N.; Zhao, Y.; Poorter, L. Climate change induced elevational range shifts of Himalayan tree species. Biotropica 2023, 55, 53–69. [Google Scholar] [CrossRef]
  62. Shao, D.; Han, R.; Ning, X.; Bai, S.; Fang, L.; Ren, Q. The Study on Soil Enzyme Activity in the Pinus tabulaeformis and Ostryopsis davidiana Rhizosphere in Daqingshan Mountain, Inner Mongolia. J. Arid Land Resour. Environ. 2008, 22, 190–193. [Google Scholar] [CrossRef]
  63. Zheng, L.; Zhang, Y.; Lu, C.; Zhang, W.; Tan, B.; Jiang, L.; Zhang, Y.; An, C. Response of forest belt on the south slope of Tianshan Mountains in China to global warming during 1990–2020. Front. Earth Sci. 2024, 18, 831–848. [Google Scholar] [CrossRef]
  64. Varol, T.; Canturk, U.; Cetin, M.; Ozel, H.B.; Sevik, H. Impacts of climate change scenarios on European ash tree (Fraxinus excelsior L.) in Turkey. For. Ecol. Manag. 2021, 491, 119199. [Google Scholar] [CrossRef]
  65. Zhang, J.; Huang, S.; Li, J.; Liao, L.; Jiang, X.; Xu, Y.; Yu, X.; Wu, L.; Zhao, L.; Fu, J.; et al. “Two zones and three centers” distribution and suitable areas shift of an evergreen oak in subtropical China under climate scenarios. Ecol. Evol. 2024, 14, e70300. [Google Scholar] [CrossRef]
  66. Liu, L.; Li, F.; Hai, L.; Sa, R.; Gao, M.; Wang, Z.; Tie, N. Priority Conservation Area of Quercus mongolica Under Climate Change: Application of an Ensemble Modeling. Sustainability 2024, 16, 9816. [Google Scholar] [CrossRef]
  67. Zhang, H.; Wei, Y.; Yue, J.; Wang, Z.; Zou, H.; Ji, X.; Zhang, S.; Liu, Z. Prediction of Potential Suitable Areas and Priority Protection for Cupressus gigantea on the Tibetan Plateau. Plants 2024, 13, 896. [Google Scholar] [CrossRef]
  68. Li, C.; Luo, G.; Yue, C.; Zhang, L.; Duan, Y.; Liu, Y.; Yang, S.; Wang, Z.; Chen, P. Distribution patterns and potential suitable habitat prediction of Ceracris kiangsu (Orthoptera: Arcypteridae) under climate change-a case study of China and Southeast Asia. Sci. Rep. 2024, 14, 20580. [Google Scholar] [CrossRef]
  69. Zhang, W.; Wang, X.; Shen, S.; Zhao, Y.; Hao, S.; Jiang, J.; Zhang, D. Analyzing the distribution patterns and dynamic niche of Magnolia grandiflora L. in the United States and China in response to climate change. Front. Plant Sci. 2024, 15, 1440610. [Google Scholar] [CrossRef]
  70. Liu, Y.; Li, Y.; Wang, R.; Guo, L.; Ji, Y.; Chen, Y.; Hao, L.; Lin, K. Impacts of Human Activity and Climate Change on the Suitable Habitats for Xanthium spinosum in China. Plants 2025, 14, 306. [Google Scholar] [CrossRef]
  71. Wilson, M.C.; Chen, X.-Y.; Corlett, R.T.; Didham, R.K.; Ding, P.; Holt, R.D.; Holyoak, M.; Hu, G.; Hughes, A.C.; Jiang, L.; et al. Habitat fragmentation and biodiversity conservation: Key findings and future challenges. Landsc. Ecol. 2016, 31, 219–227. [Google Scholar] [CrossRef]
  72. Yuan, R.; Zhang, N.; Zhang, Q. The impact of habitat loss and fragmentation on biodiversity in global protected areas. Sci. Total Environ. 2024, 931, 173004. [Google Scholar] [CrossRef]
  73. Lemordant, L.; Gentine, P. Vegetation Response to Rising CO2 Impacts Extreme Temperatures. Geophys. Res. Lett. 2019, 46, 1383–1392. [Google Scholar] [CrossRef]
  74. Yu, X.; Luo, T.; Zhou, H. Distribution of carabid beetles among 40-year-old regenerating plantations and 100-year-old naturally regenerated forests in Southwestern China. For. Ecol. Manag. 2008, 255, 2617–2625. [Google Scholar] [CrossRef]
  75. Chhetri, P.K.; Gaddis, K.D.; Cairns, D.M. Predicting the Suitable Habitat of Treeline Species in the Nepalese Himalayas Under Climate Change. Mt. Res. Dev. 2018, 38, 153–163. [Google Scholar] [CrossRef]
  76. Rawat, B.; Gairola, S.; Tewari, L.M.; Rawal, R.S. Long-term forest vegetation dynamics in Nanda Devi Biosphere Reserve, Indian west Himalaya: Evidence from repeat studies on compositional patterns. Environ. Monit. Assess. 2021, 193, 459. [Google Scholar] [CrossRef]
  77. Praticò, S.; Solano, F.; Di Fazio, S.; Modica, G. Machine Learning Classification of Mediterranean Forest Habitats in Google Earth Engine Based on Seasonal Sentinel-2 Time-Series and Input Image Composition Optimisation. Remote Sens. 2021, 13, 586. [Google Scholar] [CrossRef]
  78. Zhang, K.; Yao, L.; Meng, J.; Tao, J. Maxent modeling for predicting the potential geographical distribution of two peony species under climate change. Sci. Total Environ. 2018, 634, 1326–1334. [Google Scholar] [CrossRef]
Sustainability 17 06332 g001
Figure 1. Distribution point data of Ostryopsis davidiana.
Figure 1. Distribution point data of Ostryopsis davidiana.
Sustainability 17 06332 g001
Sustainability 17 06332 g002
Figure 2. The suitable habitat of Ostryopsis davidiana under current environmental conditions.
Figure 2. The suitable habitat of Ostryopsis davidiana under current environmental conditions.
Sustainability 17 06332 g002
Sustainability 17 06332 g003
Figure 3. Main environmental variables’ contribution rates.
Figure 3. Main environmental variables’ contribution rates.
Sustainability 17 06332 g003
Sustainability 17 06332 g004
Figure 4. Response curves of main environmental variables (A–C). The red dashed line represents the value of the variable with probability 0.5 in the response curve.
Figure 4. Response curves of main environmental variables (A–C). The red dashed line represents the value of the variable with probability 0.5 in the response curve.
Sustainability 17 06332 g004
Sustainability 17 06332 g005
Figure 5. Suitable habitats of Ostryopsis davidiana under the SSP126 climate scenario.
Figure 5. Suitable habitats of Ostryopsis davidiana under the SSP126 climate scenario.
Sustainability 17 06332 g005
Sustainability 17 06332 g006
Figure 6. Suitable habitats of Ostryopsis davidiana under the SSP370 climate scenario.
Figure 6. Suitable habitats of Ostryopsis davidiana under the SSP370 climate scenario.
Sustainability 17 06332 g006
Sustainability 17 06332 g007
Figure 7. Suitable habitats of Ostryopsis davidiana under the SSP585 climate scenario.
Figure 7. Suitable habitats of Ostryopsis davidiana under the SSP585 climate scenario.
Sustainability 17 06332 g007
Sustainability 17 06332 g008
Figure 8. Changes in suitable habitats of Ostryopsis davidiana under future climatic conditions.
Figure 8. Changes in suitable habitats of Ostryopsis davidiana under future climatic conditions.
Sustainability 17 06332 g008
Sustainability 17 06332 g009
Figure 9. Centroid changes in the total suitable habitat of Ostryopsis davidiana in future climate scenarios.
Figure 9. Centroid changes in the total suitable habitat of Ostryopsis davidiana in future climate scenarios.
Sustainability 17 06332 g009
Sustainability 17 06332 g010
Figure 10. Average elevation changes in the total suitable habitat of Ostryopsis davidiana in future climate scenarios.
Figure 10. Average elevation changes in the total suitable habitat of Ostryopsis davidiana in future climate scenarios.
Sustainability 17 06332 g010
Table 1. Evaluation indices of single and ensemble predictive models in “biomod2”.
Table 1. Evaluation indices of single and ensemble predictive models in “biomod2”.
ModelsAUCTSSKAPPA
GLM0.9720.9070.9
GBM0.9670.9130.909
GAM0.9760.9160.912
CTA0.9650.9380.934
ANN0.9720.890.885
SRE0.8220.6440.66
FDA0.970.8960.89
MARS0.9730.9080.9
RF0.9780.9170.908
MAXENT.10.9580.8470.845
MAXENT.20.9820.8970.891
XGBOOST0.9820.9190.904
myBiomodEM0.9890.940.928
Table 2. Environmental variables involved in modeling.
Table 2. Environmental variables involved in modeling.
Types of VariablesEnvironmental Variables (Code)UnitsContribution Rate (%)
Climatic factorsMean diurnal range (bio2)°C3.84
Isothermality (bio3 = (bio1/bio7) × 100)-24.31
Maximum temperature of warmest month (bio5)°C14.45
Mean temperature of coldest quarter (bio11)°C10.07
Precipitation of the driest month (bio14)mm3.23
Variation of precipitation seasonality (bio15)C of V11.22
Precipitation of wettest quarter (bio16)mm25.69
Terrain factorsSlope°1.92
Aspect-0.15
Soil factorsTop soil pH-2.1
Human factorsHuman footprint-3.02
Table 3. The habitat fragmentation of Ostryopsis davidiana under different climate scenarios.
Table 3. The habitat fragmentation of Ostryopsis davidiana under different climate scenarios.
Climate ScenariosNPAI
Current24894.04
SSP126_2041–206053687.54
SSP126_2061–208066887.74
SSP126_2081–210077681.05
SSP370_2041–206072984.56
SSP370_2061–208090775.27
SSP370_2081–210098076.42
SSP585_2041–206086981.97
SSP585_2061–208097975.21
SSP585_2081–2100101571.17
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Zhang, H.; Cui, X.; Zhang, Y.; Wang, Z.; Liu, Z. Global Warming-Driven Changes in the Suitable Habitat of Ostryopsis davidiana (Betulaceae) Shrubs. Sustainability 2025, 17, 6332. https://doi.org/10.3390/su17146332

AMA Style

Zhang H, Cui X, Zhang Y, Wang Z, Liu Z. Global Warming-Driven Changes in the Suitable Habitat of Ostryopsis davidiana (Betulaceae) Shrubs. Sustainability. 2025; 17(14):6332. https://doi.org/10.3390/su17146332

Chicago/Turabian Style

Zhang, Huayong, Xinxing Cui, Yihe Zhang, Zhongyu Wang, and Zhao Liu. 2025. "Global Warming-Driven Changes in the Suitable Habitat of Ostryopsis davidiana (Betulaceae) Shrubs" Sustainability 17, no. 14: 6332. https://doi.org/10.3390/su17146332

APA Style

Zhang, H., Cui, X., Zhang, Y., Wang, Z., & Liu, Z. (2025). Global Warming-Driven Changes in the Suitable Habitat of Ostryopsis davidiana (Betulaceae) Shrubs. Sustainability, 17(14), 6332. https://doi.org/10.3390/su17146332

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop