Skip to main content
NCBI home page
Ecology and Evolution logoLink to Ecology and Evolution
. 2024 Jul 14;14(7):e11710. doi: 10.1002/ece3.11710

Evaluating the vulnerability of Tetracentron sinense habitats to climate‐induced latitudinal shifts

Yuanjie Gan 1,2,3, Lijun Cheng 1,2,3, Junfeng Tang 1,2,3, Hongyan Han 1,2,3, Xiaohong Gan 1,2,3,✉
PMCID: PMC11246751  PMID: 39005881

Abstract

Exploring the changing process of the geographical distribution pattern of Tetracentron sinense Oliv. and its main influencing factors since the last interglacial period can provide a scientific basis for the effective protection and management of the species. The MaxEnt model was used to construct the potential distribution areas of T. sinense in different periods such as the last interglacial (LIG), the last glacial maximum (LGM), the mid‐Holocene (MID), and the current and future (2050s and 2070s). On the premise of discussing the influence of dominant environmental factors on its distribution model, the suitable area changes of T. sinense under different ecological climate situations were quantitatively analyzed. (1) The AUC and TSS values predicted by the optimized model were 0.959 and 0.835, respectively, indicating a good predictive effect by the MaxEnt model; the potential suitable areas for T. sinense in the current period are mainly located in Southwest China, which are wider compared to the actual habitats. (2) Jackknife testing showed that the lowest temperature in the coldest month (Bio6), elevation (Elev), seasonal variation coefficient of temperature (Bio4), and surface calcium carbonate content (T‐CACO3) are the dominant environmental factors affecting the distribution of T. sinense. (3) From the last interglacial period to the current period, the total suitable area of T. sinense showed a decreasing trend; the distribution points of T. sinense populations in mid‐Holocene period may be the origin of the postglacial population, and Southwest China may be its glacial biological refuge. (4) Compared with the current period, the total suitable area ranges of T. sinense in China in the 2050s and 2070s decreased, and the centroid location of its total fitness area all migrated to the northwest, with the largest migration distance in 2070s under the SSPs 7.0 climate scenario. Temperature was the principal factor influencing the geographical distribution of T. sinense. With global warming, the range of T. sinense suitable areas will show a shrinking trend, with a shift toward higher‐latitude regions. Ex situ conservation measures could be taken to preserve its germplasm resources.

Keywords: climate change, impact factors, MaxEnt, potential suitable distribution area, Tetracentron sinense

Temperature was the most important factor affecting the distribution of Tetracentron sinense. With global warming, the range of T. sinense suitable areas will show a shrinking trend. Ex situ conservation measures could be taken to preserve its germplasm resources.

graphic file with name ECE3-14-e11710-g004.jpg

1. INTRODUCTION

The geographical distribution of plant populations is influenced by both the biological characteristics of the plant species and their environment (Ge et al., 2012; Thuiller et al., 2005). Primarily, the climate within large‐scale regions serves as the principal determinant affecting species distribution. Changes in climate consequently alter species' responses and selection to climate and habitat (Ma et al., 2013). In recent times, exacerbated by climate change and human interference, species' habitats have suffered severe degradation. This degradation is particularly notable in the significant reduction of suitable growth areas for endangered plants, leading to diminished resources for wild species (Lenoir et al., 2008; Liu et al., 2015). Consequently, investigating the impacts of changing climates on species distribution patterns is crucial for understanding historical and future changes in species range and can furnish a scientific foundation for conserving germplasm resources in endangered plants (Li et al., 2021).

Species distribution models (SDMs) rely on various environmental variables such as climate and soil, closely related to the real growth and distribution of species. These models can predict potentially suitable distribution areas of species using specific algorithms, thereby elucidating the predominant environmental factors influencing their distribution and exploring the ecological requirements of species (Araújo & Peterson, 2012). Among the myriad of models, SDMs encompass 19 different methodologies, including the rule‐set genetic algorithm model (GARP), maximum entropy model (MaxEnt), and ecological factor analysis models (ENFA) (Phillips & Dudik, 2008). The MaxEnt model stands out due to its relative maturity, ease of operation, and high prediction accuracy (Hao et al., 2020). It can infer and predict from incomplete known information, making it widely applicable in studying the introduction and cultivation of relict plants, horticultural tree species, and invasive plants (Elith et al., 2006).

Tetracentron sinense Oliv., a Tertiary relict plant, represents the sole surviving species in the Tetracentron genus of the Trochodendraceae family (Fan et al., 2021). This species holds significant importance in discussing the systematic evolution of angiosperm plants. Unfortunately, due to its ornamental, furniture, and medicinal value, T. sinense has been subjected to extensive exploitation by humans, resulting in poor regeneration of its natural populations (Lu et al., 2020; Pang, 2018; Wang et al., 2023). Therefore, it has been listed as a National Second‐level Key Protected Wild Plant in China (https://www.gov.cn/zhengce/zhengceku/2021‐09/09/content_5636409.htm). The preservation of germplasm resources has garnered considerable attention from researchers (Duan et al., 2020; Zhang et al., 2019).

Fossil records indicate that Tetracentron Oliv. was once widely distributed across Europe, North America, and East Asia during the Pleistocene era (Rix & Crane, 2007). A phylogeographical analysis based on the chloroplast genome suggested that the current geographical distribution pattern of T. sinense may have been shaped by Quaternary climate oscillations, with Southwest China potentially serving as a biological refuge during glacial periods (Sun et al., 2014). Han et al. (2017) discovered that the phenotypic variation of T. sinense is closely correlated with the corresponding environmental factors such as the mean annual sunshine duration, mean temperature in July, and annual mean precipitation. Li (2021) held that low altitude was more conducive to the reproduction, regeneration, and development of T. sinense natural populations. In addition, litter depth as well as the soil zinc and calcium content importantly impacted the small‐scale distribution of this species (Zhang et al., 2024). However, the specific influence of these environmental factors on the geographic distribution of T. sinense remains ambiguous. How will the distribution pattern of T. sinense evolve in the context of past and future climate changes? What are the primary environmental factors constraining its geographical distribution? And how do these factors influence its distribution? These questions remain unanswered, impeding the effective protection and management of T. sinense germplasm resources.

Utilizing the MaxEnt model and ArcGIS spatial analysis technology, this study examines potentially suitable areas for T. sinense across historical periods (the last interglacial period, the last glacial maximum, and the middle Holocene) as well as current and future periods (2050s and 2070s). The objectives of this study are to (1) analyze the dynamic changes in potentially suitable areas, (2) investigate the main environmental factors driving changes in the distribution pattern of T. sinense, and (3) suggest a scientific basis for the effective protection and management of T. sinense.

2. MATERIALS AND METHODS

2.1. Data collection and preprocessing of distribution point data

A total of 200 distribution records of T. sinense were sourced from various data platforms, including Flora Reipublicae Popularis Sinicae (http://iplant.cn/), Plant Photo Bank of China, and the Global Biodiversity Information Facility (GBIF, http://www.gbif.org). Additionally, our research group conducted field investigations, yielding 121 distribution data points. Consequently, a comprehensive dataset comprising 321 distribution records of T. sinense was compiled (Figure 1).

FIGURE 1.

FIGURE 1

Open in a new tab

Images of individuals and inflorescences of Tetracentron sinense in their natural habitats.

The coordinate points of T. sinense, denoted by latitude and longitude, were converted into decimal numbers compatible with ARCGIS using standard formulas. The resulting table contained three columns: space, longitude, and latitude, and was stored in CSV format. Subsequently, the distribution data in CSV format were imported into ARCGIS 10.2. By utilizing the “Data/Display XY data” and “Data/Export data” functions, a vector file representing the distribution points of T. sinense was generated. The projection utilized the WGS1984 geographic coordinate system, with each grid layer retaining only one distribution point. This process ensured an accuracy of 2.5 m, facilitated by ENMTTools software. Following the removal of duplicate entries and the mitigation of sampling deviations, a final set of 233 distribution points was selected (Figure 2).

FIGURE 2.

FIGURE 2

Open in a new tab

Distribution points of Tetracentron sinense after data cleaning.

2.2. Environmental variables

We gathered environmental variables associated with bioclimatic, soil, and topographic factors as potential predictors of species distribution (Table 1). Nineteen climate variables spanning the last glacial, last glacial maximum, mid‐Holocene, current periods, and future scenarios were sourced from the World Climate Database (http://worldclim.org) (Gao & Qin, 2018). Future scenarios included low‐concentration emissions (SSPs1‐2.6) and high‐concentration emissions (SSPs5‐8.5) of greenhouse gases. Sixteen soil variables pertaining to the soil surface were acquired from the Chinese Soil Dataset within the Harmonized World Soil Database (HWSD, http://www.fao.org/faostat/en/#data), while elevation data were obtained from the same source. The spatial resolution was set at 2.5 m. Map data were represented in SHP format based on a 1:1‐million‐scale Chinese map obtained from the National Center for Basic Geographic Information.

TABLE 1.

Environmental variables used in the study.

Type Variables Description Contribution/%
Bioclimatic variables Bio1 Annual mean temperature 0.1
Bio2 Mean diurnal range 12.5
Bio3 Isothermality 5.8
Bio4 Temperature seasonality 1.5
Bio5 Max temperature 3.1
Bio6 Min temperature of coldest month 14.6
Bio7 Temperature annual range 10.4
Bio8 Mean temperature of wettest 0
Bio9 Mean temperature of driest quarter 0
Bio10 Mean temperature of warmest quarter 1.7
Bio11 Mean temperature of coldest quarter 14.9
Bio12 Annual precipitation 3.3
Bio13 Precipitation of wettest month 0.6
Bio14 Precipitation of driest month 0
Bio15 Precipitation seasonality 1.1
Bio16 Precipitation of wettest quarter 1.4
Bio17 Precipitation of driest quarter 0.8
Bio18 Precipitation of warmest quarter 1.3
Bio19 Precipitation of coldest quarter 1.7
Soil variable T_GRAVEL Topsoil gravel content 3.4
T_SAND Topsoil sand fraction 0
T_SILT Topsoil silt fraction 4.5
T_CLAY Topsoil clay fraction 0.1
T_USDA_TEX_CLASS Topsoil USDA texture classification 0.8
T_REF_BULK_DENSITY Topsoil reference bulk density 0.3
T_OC Topsoil organic carbon 0.1
T_PH_H2O Topsoil pH (H2O) 0.3
T‐ESP Topsoil sodicity (ESP) 0
T_CEC_CLAY Topsoil CEC (clay) 0.8
T_BS Topsoil base saturation 2.2
T_TEB Topsoil TEB 0.2
T_CACO3 Topsoil calcium carbonate 4.7
T_CASO4 Topsoil gypsum 0
T_ECE Topsoil salinity (Elco) 0
T_CEC_SOIL Topsoil CEC (soil) 0.1
Topography Elev Elevation 7.5
Open in a new tab

To mitigate multicollinearity and potential model overfitting (Graham, 2003), we conducted Spearman's correlation analysis within ArcGIS (Yang et al., 2013) to examine relationships among environmental factors. Variables demonstrating a correlation coefficient |r| ≥ .8 were considered highly correlated, and the less influential factor was excluded from subsequent analysis. Consequently, a total of 16 environmental factors were retained for calculation and analysis within the MaxEnt model.

Climate and elevation variable data were clipped according to the vectograph of a 1:1‐million‐scale Chinese administrative map and then converted to ASC format using ArcGIS software. Soil variable data were integrated by importing the China soil file and HWSD DATA file into ArcGIS software. Subsequently, the grid layer comprising the 16 soil variables within the MU_GLOBAL layer was extracted and converted into ASC format. Finally, all environmental variable layers underwent batch processing using ArcGIS software, resulting in environment layers with non‐overlapping extents.

2.3. Model building, optimization, and evaluation

In accordance with the latest model optimization methodology proposed by Cobos et al. (2019), adjustments were made to the parameters of the MaxEnt model to assess the degree of alignment between the distribution points of T. sinense and the model, as determined by the corrected Akaike information criterion (AICc) (Yu et al., 2019). Optimal model parameters, indicated by the lowest AICc value, were selected for utilization within the MaxEnt software (Philips et al., 2017).

The distribution data of T. sinense and its corresponding environmental variables spanning the study period were inputted into the MaxEnt model. In this investigation, 25% of the 232 distribution sites of T. sinense were earmarked for model evaluation, while the remaining 75% constituted the training set for constructing a response curve. Ten replicates were generated for each training partition, and the resultant outcomes were averaged. Model results were generated in both logistic and ASC format files, with a multifeature combination based on optimization outcomes (Moreno et al., 2011).

To calibrate the model and assess its robustness, the prediction accuracy of the MaxEnt model was evaluated using the area under the curve (AUC) and the true skill statistic (TSS) (Xu et al., 2019). The AUC values ranged between 0 and 1, with higher values indicating higher prediction accuracy. The AUC values were classified into five categories as follows: unqualified (0.5–0.6), poor (0.6–0.7), average (0.7–0.8), good (0.8–0.9), and excellent (0.9–1.0) (Fithian et al., 2015). The TSS values ranged between −1 and 1, with values closer to 1 indicating better prediction performances. A TSS value between 0.6 and 1 suggests a good prediction capability (Allouche & Kadmon, 2006).

2.4. Classification of fitness levels and area statistics

The average output data from each period, following 10 simulations in the MaxEnt model, were imported into ArcGIS software. These data were then converted into raster layers and subsequently reclassified based on the distributed probability (p) values. Employing natural breakpoint classification, the distribution area of T. sinense was categorized into four distinct levels: non‐suitable area (p < .2), low suitable zone (.2 ≤ p < .4), intermediate suitable zone (.4 ≤ p < .6), and high suitable zone (p ≥ .6). The reclassified layers were processed using an ArcGIS grid table to determine the area encompassed by each level (Zhuang et al., 2018).

2.5. Screening and threshold analysis of dominant climate factors

We conducted an analysis to identify the primary environmental factors influencing the distribution of T. sinense. This analysis relied on assessing the permutation importance (PI) and permutation contribution (PC) of environmental variables within the prediction results (Mbari, 2020).

2.6. Spatial pattern change

The distribution probability of T. sinense underwent reclassification, and a grid calculator was employed to delineate the distribution change layer of T. sinense between historical or future periods and current climate scenarios. Based on the grid values, four types of distribution area changes in T. sinense were redefined: retention (3), additions (2), losses (1), and non‐suitable (0). The suitable area for each type was determined by calculating the proportion of each value (Wang et al., 2024).

3. RESULTS

3.1. Species distribution model and its accuracy

Utilizing 233 geographic distribution data points and 36 environmental variables, we employed the MaxEnt model to simulate and predict potentially suitable areas for T. sinense. Optimization using the Emavel data package revealed that when the feature combination (FC) was set to LP and the regularization multiplier (RM) to 0.9, an AICc value of 0 indicated optimal predictive performance. Hence, FC = LP and RM = 0.9 were chosen as the final parameter settings (Table 2).

TABLE 2.

Model parameter.

Model evaluation Feature combination Regularization multiplier Value of delta Akaike information criterion corrected 5% Training omission
Default LQHPT 1 259.485332 0.103
Optimized LP 0.9 0 0.034
Open in a new tab

In this study, MaxEnt was executed with optimal model parameter settings and subjected to 10‐fold cross‐validation, resulting in an average AUC and TSS of 0.959 and 0.835, respectively (Figure 3). These values indicated that the MaxEnt model could be applied to simulate and predict the distribution of suitable areas for T. sinense.

FIGURE 3.

FIGURE 3

Open in a new tab

Model evaluation results of AUC value and TSS value.

3.2. Importance of environmental variables affecting T. sinense distribution

The influence of 16 environmental factors on distribution was evaluated through a jackknife test (Figure 4). Notably, the contribution rates (PC) of Bio6 (44.60%), T_CACO3 (10.29%), and Bio9 (8.90%) ranked highest, with a cumulative PC of 63.79%. Similarly, the permutation importance (PI) of Bio6 (34.18%), Bio4 (25.70%), and elevation (21.18%) were among the top three, with a cumulative PI of 81.06%. This analysis revealed that the primary environmental factors shaping the current geographical distribution of T. sinense include bioclimatic variables (Bio4 and Bio6), soil variables (T_CACO3), and topography (Elevation) (Table 3).

FIGURE 4.

FIGURE 4

Open in a new tab

The jackknife test result for the environmental factors.

TABLE 3.

Main parameters of environmental factors.

Variable PC PI RTGw RTGo TGw TGo AUCw AUCo
bio6 44.6012 34.1752 2.1863 1.2083 2.2081 1.2511 0.9592 0.8887
T_CACO3 10.2886 2.4837 2.1824 0.268 2.2161 0.2534 0.9592 0.6611
bio9 8.8969 0.5763 2.1975 0.9028 2.218 0.9267 0.9595 0.8435
Elev 7.1365 21.1835 2.1492 0.6041 2.167 0.6384 0.9573 0.8059
bio2 6.8383 0.036 2.1981 0.8624 2.2208 0.9352 0.9596 0.8285
bio10 5.0883 0.0595 2.1982 0.3042 2.2196 0.3437 0.9595 0.723
T_GRAVEL 3.4221 0.1516 2.1925 0.1493 2.2205 0.1051 0.9597 0.6468
bio7 3.0458 0.1643 2.1981 1.1432 2.218 1.2907 0.9595 0.8902
bio4 2.9022 25.6958 2.1867 0.9363 2.2065 1.0171 0.9592 0.8546
T_SILT 2.288 1.5817 2.1681 0.3146 2.1914 0.352 0.9581 0.7978
bio5 1.4892 0.0014 2.1982 0.4146 2.2185 0.4884 0.9595 0.7767
bio15 1.2819 1.5014 2.1853 0.5794 2.2109 0.6629 0.9595 0.7961
bio3 1.1263 6.6682 2.1787 0.031 2.2096 0.0282 0.9593 0.5629
T_BS 0.7965 0.1273 2.1921 0.025 2.216 0.0211 0.9595 0.599
bio19 0.5863 3.4403 2.1792 0.6257 2.1864 0.7266 0.9581 0.8264
bio17 0.2119 2.1538 2.1901 0.535 2.2044 0.6404 0.9589 0.8139
Open in a new tab

Environmental factor response curves further elucidated the relationship between the probability of T. sinense occurrence and environmental variables (Figure 5). Generally, when the probability exceeded 0.5, corresponding environmental factor values favored T. sinense growth (Li et al., 2022). Based on these curves, suitable environmental factor ranges for T. sinense growth were determined as follows: 407.4 to 731.3 (Bio4), −8 to 0.61°C (Bio6), 0% to 0.17% (T_CACO3), and 1355.8 to 3508 m (elevation).

FIGURE 5.

FIGURE 5

Open in a new tab

Response curves of dominant environmental factors.

3.3. Suitable distribution areas for T. sinense under current climatic conditions

Currently, T. sinense finds suitable habitat across approximately 7.24% of China's total land area. Central Sichuan, Northwestern Yunnan, Western Guizhou, and Southwestern Hubei emerge as primary distribution hubs, with suitable areas radiating outward in distinct patterns (Figure 6). Notably, highly suitable areas connect with moderately suitable regions, while less suitable areas extend from them. Compared to the current distribution, new suitable areas, such as Zhejiang, Anhui, Fujian, Henan, Guangdong, and Taiwan, have emerged along coastal regions. Expansion of T. sinense distribution is observed in provinces such as Chongqing, Guizhou, and Hunan, while the species' range remains largely unchanged in other regions.

FIGURE 6.

FIGURE 6

Open in a new tab

Suitable distribution of Tetracentron sinense in China under current climatic conditions.

3.4. Prediction of suitable areas of T. sinense in historical and future climates

This study analyzed six periods to predict potential distribution areas for T. sinense. Optimal distribution areas varied across periods, predominantly favoring Southwest China.

The total suitable area for T. sinense has exhibited fluctuations from the last interglacial period to the present (Figure 7). Initially, it stood at 758,258 km2, decreasing to 731,469.3 km2, then rising to 750,132.8 km2 before sharply declining to 675,550.8 km2. Similar trends were observed in high, medium, and low suitability areas.

FIGURE 7.

FIGURE 7

Open in a new tab

Prediction of potentially suitable areas of Tetracentron sinense in historical period. Note: LGM, last glacial maximum; LIG, last interglacial; MID, Mid‐Holocene; 50s‐126 ~ 50s‐585 were SSPs1‐2.6, SSPs2‐4.5, SSPs3‐7.0, and SSPs5‐8.5 of 2050s, respectively; and 70s‐126 ~ 70s‐585 were SSPs1‐2.6, SSPs2‐4.5, SSPs3‐7.0, and SSPs5‐8.5 of 2070s, respectively. The same as below.

Under various climate scenarios for the 2050s and 2070s, significant changes occurred in suitable areas due to climate shifts (Figure 8). In the 2050s, the climate scenario yielding the largest suitable area was SSPs2‐4.5, which was 60,955.2 km2 less than the current scenario. By 2070, SSPs1‐2.6 exhibited the largest suitable area, 39,032.8 km2 less than the current scenario. Across all grades, the future suitable habitat area for T. sinense is expected to decrease compared to the current area.

FIGURE 8.

FIGURE 8

Open in a new tab

Prediction of potentially suitable areas of Tetracentron sinense in future period.

3.5. Changes in spatial pattern of potential suitable areas of T. sinense

Compared to the current period, the suitable area for T. sinense decreased from the last interglacial period to the mid‐Holocene, followed by an increase (Figure 9). During the mid‐Holocene, an additional area of approximately 119,693.5 km2 emerged, constituting 17.72% of the total area. These additions were mainly concentrated in Central Yunnan and Southern Gansu. Conversely, during the last glacial period, the cold climate led to a significant reduction in suitable habitats, resulting in a loss of 42,620.2 km2, or a 6.3% decrease (Table 4). Losses occurred predominantly in fragmented areas at the junction of Sichuan, Yunnan, Gansu, and Shaanxi provinces. Retention areas during the last glacial maximum were primarily situated in Southwest China (Figure 10).

FIGURE 9.

FIGURE 9

Open in a new tab

Comparison of suitable distribution areas in different periods.

TABLE 4.

Spatial changes in suitable areas of Tetracentron sinense under different climate change scenarios.

Period Area/km2 Area change rate/%
Addition Loss Retention Change Increasing rate Attrition rate Retention rate Total
LIG 111,055 28,374.16 647,234 82,680.79 16.44 4.20 95.81 108.05
LGM 98,509.7 42,620.18 632,988 55,889.52 14.58 6.31 93.70 101.97
MID 119,693.5 29,653.95 645,954.2 90,039.55 17.72 4.39 95.62 108.95
50s‐126 32,884.35 121,586.2 553,949 −88,701.9 4.87 18.00 82.00 68.87
50s‐245 77,403.53 138,390.2 537,160.6 −60,986.6 11.46 20.49 79.51 70.49
50s‐370 61,542.51 159,033.9 516,485.6 −97,491.4 9.11 23.54 76.45 62.02
50s‐585 47,145.94 203,047.5 472,503.4 −155,902 6.98 30.06 69.94 46.87
70s‐126 48,964.43 88,028.21 587,522.6 −39,063.8 7.25 13.03 86.97 81.19
70s‐245 99,629.48 145,428.4 530,038.2 −45,798.9 14.75 21.53 78.46 71.68
70s‐370 78,531.53 243,828.3 431,704.5 −165,297 11.62 36.09 63.90 39.44
70s‐585 113,049.2 196,851.1 478,699.7 −83,801.9 16.73 29.14 70.86 58.46
Open in a new tab

FIGURE 10.

FIGURE 10

Open in a new tab

Spatial pattern changes of Tetracentron sinense under historical climate change scenarios.

The expansion rate for 2050 was lower than that for 2070 across eight different climate scenarios with similar concentrations (Figure 11). In the 2050s, the additional area for T. sinense initially expanded, then contracted with increasing greenhouse gas emissions. Conversely, by the 2070s, the additional area exhibited an upward trend with increased emissions. Except for the SSPs2.6 scenario, the loss rate in the 2070s was significantly higher than that in the 2050s (Table 4). Overall, under future climate scenarios, fragmentation of potential distribution areas suitable for T. sinense is expected to increase. Loss areas will concentrate in the southern region of the current suitable area, while added areas will primarily occur in the north. These areas require close monitoring for potential pattern changes.

FIGURE 11.

FIGURE 11

Open in a new tab

Spatial pattern changes of Tetracentron sinense under future climate change scenarios.

3.6. Core distributional shifts

Throughout history, the centroid of T. sinense has exhibited notable fluctuations (Figure 12). From the last interglacial period to the last glacial maximum, the center of mass shifted southwest by 49,638 m. Subsequently, from the last glacial period to the mid‐Holocene, the centroid moved northeast by 36,430 m. Continuing into the present era, the center of mass further migrated northeast by 18,472.6 m.

FIGURE 12.

FIGURE 12

Open in a new tab

The core distributional shifts of suitable habitat under different climate scenarios for Tetracentron sinense.

Projection analysis indicates a northward shift in the centroid of the suitable area for T. sinense by 2050 and 2070, under future climate change scenarios (Figure 12). With increasing greenhouse gas emissions, the spatial distribution of potentially suitable areas undergoes more pronounced alterations over greater distances. Notably, the migration distance of T. sinense in the 2070s surpasses that of the 2050s under low and medium emission concentrations, except for the center of the 2050s‐SSPs8.5. The most extensive migration distance occurs under the SSPs7.0–2070s climate scenario.

4. DISCUSSION

4.1. Environmental factors affecting T. sinense distribution

Results from the jackknife test and analysis of the main parameter table underscore that the potential distribution of T. sinense is influenced by four key environmental factors: Bio4, Bio6, T_CACO3, and elevation. Elevation plays a significant role in species distribution by indirectly affecting temperature and precipitation (Clark & Husband, 2007; Ma, Guo, et al., 2021; Ma, Lu, et al., 2021). While topsoil calcium carbonate does impose some restrictions on plant distribution, its impact appears less pronounced (Zhang et al., 2024). Therefore, temperature emerges as the primary factor shaping the geographical distribution of T. sinense, a finding corroborated by previous studies. Li et al. (2021), for instance, employed the concept of space–time substitution to investigate the influence of altitude on the reproductive characteristics of T. sinense, proposing that temperature variations impact species fitness and could prompt migration to higher altitudes with rising temperatures. Similarly, Chen et al. (2023) through correlation analysis of chronological and meteorological factors identified air temperature during specific periods as a key influencer during T. sinense growth stages. In general, temperature fluctuations can indirectly affect the physiological and metabolic activities of T. sinense by influencing their leaf phenotype, respiration, photosynthesis, and water absorption capacity. Consequently, these effects would influence T. sinense survival, reproduction, and growth, ultimately impacting importantly its distribution (Han et al., 2017; Li et al., 2020).

The permutation importance (PI) and jackknife tests further highlight the critical role of the minimum temperature of the coldest month (bio6) in shaping the potential geographical distribution of T. sinense. This finding aligns with observations in other species such as Quercus mongolica (Yin et al., 2013), Santalum album (Hu et al., 2014), Gymnocarpos przewalskii (Zhao et al., 2020), and Thuja sutchuenensis (Ma, Guo, et al., 2021; Ma, Lu, et al., 2021). Chen's research also revealed a significant negative correlation between T. sinense growth and the lowest temperatures in November. Consequently, the minimum temperature during the coldest month emerges as a pivotal factor constraining northward expansion of T. sinense. Low temperatures not only hinder seed germination and morphological development but also pose challenges to the species' cold resistance, ultimately impeding its normal growth and development in Northern China.

4.2. Changes in the potential geographical distribution of T. sinense

Our investigation revealed fluctuations in the total suitable area of T. sinense, which decreased from 758,258 km2 to 731,469.3 km2 from the last interglacial period to the last glacial maximum. During this period, suitable habitats contracted, primarily concentrating in the central part of Southwest China. The mountainous terrain in this region acted as a barrier against cold air, mitigating extreme climate fluctuations and providing stable conditions conducive to species survival (Li, 2017; Stewart et al., 2010). Furthermore, the absence of geographical barriers facilitated migration to this area, establishing Southwest China as a critical refuge for the tertiary relict T. sinense, consistent with the results from Sun et al. (2014). The refuges can buffer ice‐age climates, help T. sinense through harsh environmental conditions, and provide a source for the species to re‐enter other habitats (Chen, Hill, et al., 2011; Chen, Kang, et al., 2011; Liang, 2020). Subsequently, from the last glacial maximum to the mid‐Holocene, the total suitable area expanded to 750,132.8 km2, with suitable habitat extending outward from the Sichuan Basin and Yunnan–Guizhou Plateau. This expansion correlates with the warmer and wetter global climate during the mid‐Holocene, aligning with the hydrothermal conditions favorable for T. sinense growth. Consequently, the population of T. sinense exhibited significant glacial contraction and post‐glacial expansion, consistent with findings for other species such as Thuja sutchuenensis (Qin et al., 2017), Davidia involucrata (Ye et al., 2021), and Ulmus elongata (Zhang et al., 2021).

From the mid‐Holocene to our current time, the total suitable area for T. sinense reduced by 74,582.0 km2, potentially attributed to the intensified human activities such as increased contemporary energy consumption, amplified carbon dioxide emissions, deforestation, and agricultural land expansion (Ye et al., 2021), resulting in T. sinense habitat fragmentation. However, we observed that the currently predicted suitable habitat area exceeds the actual habitat area, a discrepancy potentially attributable to several factors. First, the modeling dataset predominantly comprises data from earlier periods, with scant recent distribution data. Second, the native habitat of T. sinense wild populations is mainly concentrated in the southwestern region of China, marked by diverse climate conditions and vegetation types. Hence, the probability of undiscovered distribution points within these intricate ecosystems is notably high.

In the future, global climate warming is anticipated to substantially impact suitable habitats for T. sinense, resulting in a significant distribution shift. The extent of this shift varies depending on emission scenarios, with the largest loss area observed in the SSPs8.5 scenario and the smallest in the SSPs2.6 scenario. This disparity is attributed to temperature surpassing the threshold required for optimal T. sinense growth in the SSPs8.5 scenario. Consequently, temperature increase emerges as a primary driver of future reductions in suitable distribution areas. Therefore, the reduction of suitable habitat areas for T. sinense in the future will increase its risk of extinction, posing challenges to the continued survival and reproduction of its natural populations.

The centroid analysis revealed that the future distribution of T. sinense is shifting northwestward, toward higher latitudes, a trend consistent with that for various other species (Chen, Hill, et al., 2011; Chen, Kang, et al., 2011). This migration pattern could lead to the original habitat loss of the species coupled with population size reduction. Moreover, as T. sinense enters new habitats, it will encounter intense interspecific competition with new coexisting species. Its limited adaptation and weaker photosynthetic capacities compared to those of coexisting species would likely represent a disadvantage during interspecific competitions, thereby hindering its ability to survive and reproduce in new habitats (Yang, 2023).

4.3. Protection and management strategy

Our findings underscore a concerning trend: under future climate conditions, the rate of decline in T. sinense populations is projected to surpass the rate of expansion, posing a significant risk of extinction. Urgent action is therefore warranted to address this potential survival crisis. Identification of the primary potential distribution regions of T. sinense enables the strategic establishment of nature reserves in these areas, crucial for safeguarding the natural habitats of wild T. sinense populations.

Currently, the main potential distribution regions of T. sinense have been identified, enabling the designation of nature reserves in these areas to bolster the safeguarding of the natural habitats of wild T. sinense. The suitable habitat for T. sinense would expectably shift toward higher‐latitude regions in the future. Therefore, moderate northwest movement beyond the existing predicted areas is imperative for nature reserve designation. Given the habitat loss, it is imperative to consider the impacts on associated species when devising ex situ conservation strategies. Moreover, proactive measures tailored to the growth requirements of T. sinense should be implemented in newly identified areas to mitigate potential disruptions. Preserving retention areas can serve as secure sanctuaries, allowing trees to adapt to climate change. Thus, it is essential to bolster the protection and management efforts in these critical zones.

4.4. Limitations and future perspectives

Although the MaxEnt model is widely used in species distribution prediction and demonstrates good performance in the case of incomplete data and small sample sizes, its accuracy might be potentially compromised when few species distribution points or inaccurate environmental variables are available (Phillips et al., 2006). Multiple studies described that constructing small ensemble models (ESMs) based on generalized linear models (GLM), gradient boosting machines (GBM), and MaxEnt could yield modeling results superior to those of individual species distribution models, effectively overcoming the limitations of single models in the modeling of rare species (Breiner et al., 2015). Furthermore, if the MaxEnt model only considers abiotic factors, the resulting measurements might also be subject to certain degree of error. Therefore, generalized linear models (GLM) and gradient boosting machines (GBM) could be combined in the future, taking into account various factors such as biotic factors as well as intra‐species and inter‐species relationships (Zhou et al., 2023), and even certain anthropogenic factors (Feeley & Silman, 2010) to enhance the accuracy of species distribution prediction and better elucidate the changing geographical distribution patterns of T. sinense.

Temperature has been identified as the primary factor determining T. sinense distribution, marking the first step toward understanding its response to global warming. Future research should focus on how temperature impacts T. sinense population dynamics, growth, reproduction, adaptation, and ecosystem function under different climatic conditions. Supporting long‐term monitoring and ecosystem dynamics research is crucial for understanding the ecological and evolutionary effects of temperature and developing climate change adaptation strategies for T. sinense.

5. CONCLUSIONS

This study employed the MaxEnt model to delineate the suitable distribution areas of T. sinense across different temporal periods. Our analysis identified four key environmental factors—Bio4, Bio6, T_CACO3, and elevation—as primary determinants shaping the potential distribution of T. sinense. Specifically, temperature emerged as the principal factor influencing the geographical distribution of T. sinense.

The dynamic changes in suitable areas for T. sinense during historical periods adhered to the pattern of “glacial contraction and postglacial expansion.” However, in the context of global warming, our findings reveal a concerning trend of decreasing suitability for T. sinense, with a shift toward higher‐latitude regions.

These results furnish a robust scientific foundation for guiding the management, conservation, and judicious site selection strategies aimed at preserving T. sinense populations in the face of ongoing environmental changes. Furthermore, they provide novel insights into the future changes of rare species facing climate change, thereby offering theoretical guidance for endangered plant conservation and references for biodiversity conservation.

AUTHOR CONTRIBUTIONS

Yuanjie Gan: Conceptualization (equal); data curation (equal); methodology (equal); software (equal); writing – original draft (equal). Lijun Cheng: Data curation (equal); software (equal); visualization (equal). Junfeng Tang: Methodology (equal); software (equal). Hongyan Han: Data curation (equal); software (equal); visualization (equal). Xiaohong Gan: Data curation (equal); project administration (equal); supervision (equal); visualization (equal); writing – review and editing (equal).

FUNDING INFORMATION

This study was supported by the National Natural Science Foundation of China (No. 32070371), the Innovation Team Funds of China West Normal University (No. KCXTD2022‐4), the Sichuan Meigu Dafengding National Nature Reserve (No. mgdfd2022‐13), Sichuan Micang Mountain National Nature Reserve (No. N5108212022000043), and Natural Science Foundation of Sichuan Province (No. 23NSFSC1272).

CONFLICT OF INTEREST STATEMENT

The authors declare no competing interests.

ACKNOWLEDGEMENTS

We would like to thank Editage (www.editage.cn) for English language editing.

Gan, Y. , Cheng, L. , Tang, J. , Han, H. , & Gan, X. (2024). Evaluating the vulnerability of Tetracentron sinense habitats to climate‐induced latitudinal shifts. Ecology and Evolution, 14, e11710. 10.1002/ece3.11710

DATA AVAILABILITY STATEMENT

Datasets used in this study are available online from the Dryad Digital Repository (https://datadryad.org/stash/share/osQnAcuL0yh0p3g4z78YVAhXCRIoPeEDpRev5Xjc_jE).

REFERENCES

  1. Allouche, O. , & Kadmon, T. R. (2006). Assessing the accuracy of species distribution models: Prevalence, kappa and the true skill statistic (TSS). Journal of Applied Ecology, 43, 1223–1232. 10.1111/j.1365-2664.2006.01214.x [DOI] [Google Scholar]
  2. Araújo, M. B. , & Peterson, A. T. (2012). Uses and misuses of bioclimatic envelope modeling. Ecology, 93(7), 1527–1539. 10.1890/11-1930.1 [DOI] [PubMed] [Google Scholar]
  3. Breiner, F. T. , Antoine, G. , Bergamini, A. , & Nobis, M. P. (2015). Overcoming limitations of modelling rare species by using ensembles of small models. Methods in Ecology and Evolution, 6(10), 1210–1218. 10.1111/2041-210X.12403 [DOI] [Google Scholar]
  4. Chen, D. M. , Kang, H. Z. , & Liu, C. J. (2011). An overview on the potential quaternary glacial Refugia of plants in China mainland. Bulletin of Botanical Research, 31(5), 623–632. [Google Scholar]
  5. Chen, I. C. , Hill, K. J. , Ohlemüller, R. , Roy, B. D. , & Thomas, D. C. (2011). Rapid range shifts of species associated with high levels of climate warming. Science, 333(6045), 1024–1026. 10.1126/science.1206432 [DOI] [PubMed] [Google Scholar]
  6. Chen, R. , Mao, W. L. , Li, W. Y. , Han, H. Y. , Zhang, X. M. , & Gan, X. H. (2023). The photosynthetic eco‐physiological adaptability of the endangered plant Tetracentron sinense to different habitats and altitudes. Biologia Plantarum, 67, 54–66. 10.32615/bp.2023.005 [DOI] [Google Scholar]
  7. Clark, M. J. , & Husband, B. C. (2007). Plasticity and timing of flower closure in response to pollination in Chamerion angustifolium (Onagraceae). International Journal of Plant Sciences, 168(5), 619–625. 10.1086/513486 [DOI] [Google Scholar]
  8. Cobos, M. E. , Peterson, A. T. , Barve, N. , & Osorio‐Olvera, L. (2019). Kuenm: An R package for detailed development of ecological niche models using Maxent. PeerJ, 7, e6281. 10.7717/peerj.6281 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Duan, F. , Li, S. , & Gan, X. H. (2020). Diversity analysis on leaf phenotypic traits of the endangered plant Tetracentron sinense . Subtropical Plant Science, 49(1), 32–40. [Google Scholar]
  10. Elith, J. , Graham, C. H. , Anderson, R. P. , Dudik, M. , Ferrier, S. , Guisan, A. , Hijmans, R. J. , Huettmann, F. , Leathwick, J. R. , Lehmann, A. , Li, J. , Lohmann, L. G. , Loiselle, B. A. , Manion, G. , Moritz, C. , Nakamura, M. , Nakazawa, Y. , Overton, J. M. C. , Peterson, A. T. , … Zimmermann, N. E. (2006). Novel methods improve prediction of Species' distributions from occurrence data. Ecography, 29(2), 129–151. https://www.jstor.org/stable/3683475 [Google Scholar]
  11. Fan, W. Q. , Li, W. Y. , Zhang, X. M. , & Gan, X. H. (2021). Photosynthetic physiological characteristics of Tetracentron sinense Oliv in different DBH classes and the factors restricting regeneration. Journal of Plant Growth Regulation, 41(5), 1943–1952. 10.1007/S00344-021-10421-3 [DOI] [Google Scholar]
  12. Feeley, K. J. , & Silman, M. R. (2010). Land‐use and climate change effects on population size and extinction risk of Andean plants. Global Change Biology, 16(12), 3215–3222. 10.1111/j.1365-2486.2010.02197.x [DOI] [Google Scholar]
  13. Fithian, W. , Elith, J. , Hastie, T. , & Keith, D. A. (2015). Bias correction in species distribution models: Pooling survey and collection data for multiple species. Methods in Ecology and Evolution, 6(4), 424–438. 10.1111/2041-210X.12242 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gao, Y. , & Qin, H. (2018). Discussion on the application of MaxEnt ecologic niche model in the prediction of Acer rubrum introduction area. Chinese Landscape Architecture, 34(4), 89–93. [Google Scholar]
  15. Ge, X. J. , Hsu, T. W. , Hung, K. H. , Lin, C. J. , Huang, C. C. , Huang, C. C. , Chiang, Y. C. , & Chiang, T. Y. (2012). Inferring multiple refugia and phylogeographical patterns in Pinus massoniana based on nucleotide sequence variation and DNA finger‐printing. PLoS One, 7(8), e43717. 10.1371/journal.pone.0043717 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Graham, M. H. (2003). Confronting multicollinearity in ecological multiple regression. Ecology, 84(11), 2809–2815. https://www.jstor.org/stable/3449952 [Google Scholar]
  17. Han, H. Y. , Li, S. , Gan, X. H. , & Zhang, X. M. (2017). Phenotypic diversity in natural populations of an endangered plant Tetracentron sinense . Botanical Sciences, 95(2), 283–294. 10.17129/botsci.84 [DOI] [Google Scholar]
  18. Hao, T. X. , Elith, J. , Lahoz‐Monfort, J. J. , & Guillera‐Arroita, G. (2020). Testing whether ensemble modelling is advantageous for maximising predictive performance of species distribution models. Ecography, 43(4), 549–558. 10.1111/ecog.04890 [DOI] [Google Scholar]
  19. Hu, X. , Wu, F. C. , Guo, W. , & Liu, N. (2014). Identification of potential cultivation region for Santalum album in China by the MaxEnt ecologic niche model. Scientia Silvae Sinica, 50(5), 27–33. [Google Scholar]
  20. Lenoir, J. , Gégout, J. C. , Marquet, P. A. , Ruffray, P. , & Brisse, H. (2008). A significant upward shift in plant species optimum elevation during the 20th century. Science, 320(5884), 1768–1771. 10.1126/science.1156831 [DOI] [PubMed] [Google Scholar]
  21. Li, G. Q. , Rogers, P. C. , & Huang, J. H. (2021). Black locust (Robinia pseudoacacia L.) range shifts in China: Application of a global model in climate change futures. Climate Change Ecology, 2, 100036. 10.1016/J.ECOCHG.2021.100036 [DOI] [Google Scholar]
  22. Li, S. (2017). Study on genetic diversity of an endangered plant Tetracentron sinense Oliv . China West Normal University.
  23. Li, Y. (2021). The effect of altitude on the reproductive characteristics of an endangered plant Tetracentron sinense Oliv . China West Normal University. 12, 62. https://link.cnki.net/doi/10.27859/d.cnki.gxhsf.2021.000344
  24. Li, Y. , Li, S. , Lu, X. H. , Wang, Q. Q. , Han, H. Y. , Zhang, X. M. , Ma, Y. H. , & Gan, X. H. (2020). Leaf phenotypic variation of endangered plant Tetracentron sinense Oliv. And influence of geographical and climatic factors. Journal of Forestry Research, 32(2), 1–14. 10.1007/s11676-020-01124-8 [DOI] [Google Scholar]
  25. Li, Y. X. , Shao, W. H. , Huang, S. Q. , Zhang, Y. Z. , Fang, H. F. , & Jiang, J. M. (2022). Prediction of suitable habitats for Sapindus delavayi based on the MaxEnt model. Forests, 13(10), 1611. 10.3390/F13101611 [DOI] [Google Scholar]
  26. Liang, L. Y. (2020). The influence of quaternary glacial period on flora and vegetation in China . China Place Name, 07, 51+53.
  27. Liu, H. , Ren, H. , Liu, Q. , Wen, X. Y. , Maunder, M. , & Gao, J. Y. (2015). Translocation of threatened plants as a conservation measure in China. Conservation Biology, 29(6), 1537–1551. 10.1111/cobi.12585 [DOI] [PubMed] [Google Scholar]
  28. Lu, X. H. , Xu, N. , Chen, Y. , Li, Y. , & Gan, X. H. (2020). Effects of light intensity and ground cover on seedling regeneration of Tetracentron sinense Oliv. Journal of Plant Growth Regulation, 40(2), 736–748. 10.1007/s00344-020-10137-w [DOI] [Google Scholar]
  29. Ma, F. Q. , Guo, Q. S. , Qin, A. L. , Jian, Z. J. , & Huang, Y. J. (2021). Survival and growth of reintroduced Thuja sutchuenensis seedlings in relation to environmental factors. Scientia Silvae Sinica, 57(11), 1–12. [Google Scholar]
  30. Ma, Y. N. , Lu, X. L. , Li, K. W. , Wang, C. Y. , Guna, A. , & Zhang, J. Q. (2021). Prediction of potential geographical distribution patterns of Actinidia arguta under different climate scenarios. Sustainability, 13(6), 3526. 10.3390/SU13063526 [DOI] [Google Scholar]
  31. Ma, Y. P. , Chen, G. , Grumbine, R. E. , Dao, Z. L. , Sun, W. B. , & Guo, H. J. (2013). Conserving plant species with extremely small populations (PSESP) in China. Biodiversity and Conservation, 22(3), 803–809. 10.1007/s10531-013-0434-3 [DOI] [Google Scholar]
  32. Mbari, J. N. (2020). Using species distribution models to predict the potential distribution regions of Adansonia in Madagascar. Chinese Academy of Sciences, 1, 72. 10.27603/d.cnki.gkhzs.2020.000021 [DOI] [Google Scholar]
  33. Moreno, R. , Zamora, R. , Molina, J. R. , Vasquez, A. , & Herrera, M. Á. (2011). Predictive modeling of microhabitats for endemic birds in south Chilean temperate forests using maximum entropy (Maxent). Ecological Informatics, 6(6), 364–370. 10.1016/j.ecoinf.2011.07.003 [DOI] [Google Scholar]
  34. Pang, Y. F. (2018). Seedling raising technology of Tetracentron sinense in Xiaolongshan forest area. Contemporary Horticulture, 18, 49. 10.14051/j.cnki.xdyy.2018.18.036 [DOI] [Google Scholar]
  35. Philips, S. J. , Anderson, R. P. , Dudík, M. , Schapire, R. E. , & Blair, M. E. (2017). Opening the black box: An open‐source release of Maxent. Ecography, 40(7), 887–893. 10.1111/ecog.03049 [DOI] [Google Scholar]
  36. Phillips, S. J. , Anderson, R. P. , & Schapire, R. E. (2006). Maximum entropy modeling of species geographic distributions. Ecological Modelling, 190(3–4), 231–259. 10.1016/j.ecolmodel.2005.03.026 [DOI] [Google Scholar]
  37. Phillips, S. J. , & Dudik, M. (2008). Modeling of species distributions with Maxent: New extensions and a comprehensive evaluation. Ecography, 31(2), 161–175. 10.1111/J.0906-7590.2008.5203.X [DOI] [Google Scholar]
  38. Qin, A. L. , Liu, B. , Guo, Q. S. , Bussmann, R. W. , Ma, F. Q. , Jian, Z. J. , Xu, G. X. , & Pei, S. X. (2017). Maxent modeling for predicting impacts of climate change on the potential distribution of Thuja sutchuenensis Franch., an extremely endangered conifer from southwestern China. Global Ecology and Conservation, 10, 139–146. 10.1016/j.gecco.2017.02.004 [DOI] [Google Scholar]
  39. Rix, M. , & Crane, P. (2007). Tetracentron sinense: Tetracentraceae. Curtis's Botanical Magazine, 24(3), 168–173. 10.1111/j.1467-8748.2007.00581.x [DOI] [Google Scholar]
  40. Stewart, J. R. , Lister, A. M. , Barnes, L. , & Dalén, L. (2010). Refugia revisited: Individualistic responses of species in space and time. Proceedings. Biological Sciences, 277(1682), 661–671. 10.1098/rspb.2009.1272 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Sun, Y. X. , Moore, M. J. , Yue, L. L. , Feng, T. , Chu, H. J. , Chen, S. T. , Ji, Y. H. , Wang, H. C. , & Li, J. Q. (2014). Chloroplast phylogeography of the east Asian Arcto‐tertiary relict Tetracentron sinense (Trochodendraceae). Journal of Biogeography, 41(9), 1721–1732. 10.1111/jbi.12323 [DOI] [Google Scholar]
  42. Thuiller, W. , Lavorel, S. , Araújo, M. B. , Sykes, M. T. , & Prentice, I. C. (2005). Climate change threats to plant diversity in Europe. Proceedings of the National Academy of Sciences of the United States of America, 102(23), 8245–8250. 10.1073/pnas.0409902102 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Wang, T. X. , Li, W. T. , Cui, H. X. , Song, Y. R. , Liu, C. Y. , Yan, Q. , Wu, Y. X. , Jia, Y. H. , Fang, L. Z. , & Qi, L. H. (2024). Predicting the potential habitat distribution of relict plant Davidia involucrata in China based on the MaxEnt model. Forests, 15(2), 272. 10.3390/F15020272 [DOI] [Google Scholar]
  44. Wang, X. , Duan, F. , Zhang, H. , Han, H. Y. , & Gan, X. H. (2023). Fine‐scale spatial genetic structure of the endangered plant Tetracentron sinense Oliv. (Trochodendraceae) in Leigong Mountain. Global Ecology and Conservation, 41, e02382. doi: 10.1016/J.GECCO.2023.E02382 [DOI] [Google Scholar]
  45. Xu, D. , Zhuo, Z. , Wang, R. , Ye, M. , & Pu, B. (2019). Modeling the distribution of Zanthoxylum armatum in China with MaxEnt modeling. Global Ecology and Conservation, 19, e00691. 10.1016/j.gecco.2019.e00691 [DOI] [Google Scholar]
  46. Yang, X. Q. , Kushwaha, S. P. S. , Saran, S. , Xu, J. C. , & Roy, P. S. (2013). Maxent modeling for predicting the potential distribution of medicinal plant, Justicia adhatoda L. in lesser Himalayan foothills. Ecological Engineering, 51, 83–87. 10.1016/j.ecoleng.2012.12.004 [DOI] [Google Scholar]
  47. Yang, Z. Regional adaptability of Tetracentron sinense Oliv. Based on functional traits and phylogeny . (2023). China West Normal University, 02, 68. https://link.cnki.net/doi/10.27859/d.cnki.gxhsf.2023.000559
  48. Ye, L. Q. , Zhang, W. H. , Ye, X. Z. , Liu, Y. P. , Zhang, G. F. , Liu, B. , & Ruan, S. N. (2021). Prediction of potential distribution area and analysis of dominant environmental variables of Davidia involucrate based on Maxent. Journal of Sichuan Agricultural University, 39(5), 604–612. 10.16036/j.issn.1000-2650.2021.05.006 [DOI] [Google Scholar]
  49. Yin, X. J. , Zhou, G. S. , Sui, X. H. , He, Q. J. , & Li, R. P. (2013). Potential geographical distribution of Quercus wutaishanica Forest and its dominant factors. Scientia Silvae Sinica, 49(8), 10–14. [Google Scholar]
  50. Yu, Y. Z. , Zhang, M. H. , Du, H. R. , Li, Q. , Zhang, L. B. , & Mu, W. J. (2019). Optimized MAXENT model in simulating distribution of suitable habitat of moose. Journal of Northeast Forestry University, 47(10), 81–84+95. 10.13759/j.cnki.dlxb.2019.10.016 [DOI] [Google Scholar]
  51. Zhang, H. , Deng, H. P. , & Gan, X. H. (2024). The spatial pattern of the tertiary relict plant Tetracentron sinense Oliver and its influencing factors. Forests, 15(1), 110. 10.3390/f15010110 [DOI] [Google Scholar]
  52. Zhang, H. , Wang, D. , Duan, F. , Li, S. , & Gan, X. H. (2019). Construction strategy of core collection based on leaf phenotypic traits of Tetracentron sinense . Forest Research, 32(2), 166–173. 10.13275/j.cnki.lykxyj.2019.02.024 [DOI] [Google Scholar]
  53. Zhang, M. Z. , Ye, X. Z. , Li, J. H. , Liu, Y. P. , Chen, S. P. , & Liu, B. (2021). Prediction of potential suitable area of Ulmus elongate in China under climate change scenarios. Chinese Journal of Ecology, 40(12), 3822–3835. 10.13292/j.1000-4890.202112.018 [DOI] [Google Scholar]
  54. Zhao, Z. F. , Wei, H. Y. , Guo, Y. L. , Luan, W. F. , & Zhao, Z. B. (2020). Impact of climate change on the suitable habitat distribution of Gymnocarpos przewalskii, a relict plant. Journal of Desert Research, 40(2), 125–133. [Google Scholar]
  55. Zhou, Y. R. , Lu, X. , & Zhang, G. F. (2023). Potentially differential impacts on niche overlap between Chinese endangered Zelkova schneideriana and its associated tree species under climate change. Frontiers in Ecology and Evolution, 11, 1–18. 10.3389/FEVO.2023.1218149 [DOI] [Google Scholar]
  56. Zhuang, H. F. , Zhang, Y. B. , Wang, W. , Ren, Y. H. , Liu, F. Z. , Du, J. H. , & Zhou, Y. (2018). Optimized hot spot analysis for probability of species distribution under different spatial scales based on MaxEnt model: Manglietia insignis case. Biodiversity Science, 26(9), 931–940. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Datasets used in this study are available online from the Dryad Digital Repository (https://datadryad.org/stash/share/osQnAcuL0yh0p3g4z78YVAhXCRIoPeEDpRev5Xjc_jE).

Articles from Ecology and Evolution are provided here courtesy of Wiley

RESOURCES