1. Introduction
Greenhouse gas (GHG) emissions have led to global warming, which has now become a global issue [1]. It was projected that, by the end of the 21st century (2081–2100), relative to 1986–2005 with the reason of increasing GHG emission during the industrial age, the global average surface temperature is likely to increase by 0.3–1.7 °C under the RCP 2.6 scenario, 1.1–2.6 °C under the RCP 4.5 scenario, and could be 1.4–3.1 °C under the RCP 6.0 scenario and 2.6–4.8 °C under the RCP 8.5 scenario [2]. Changing climatic conditions have had a profound impact on the distribution range of species, placing the global ecological environment under inconceivable pressure [3,4], determining the viability and conservation potential of various species [5], and altering species’ dominance, survival, migration, and community structure [6]. Consequently, some species will either shift their distributions or be extinguished as a result of further climate change in the future [7].
The ecological niche was defined by Grinnell [8] in 1917 and Elton [9] in 1927, respectively. Hutchinson [10] proposed the hypervolume niche and defined it as the set of all variables that allow species to exist indefinitely in 1957. There are numerous established models, such as Bioclim [11], DOMAIN [12], GARP [13], and MaxEnt [14], to simulate ecological requirements and distribution areas of species based on niche theory. Among these models, MaxEnt has been widely used owing to its advantages over other approaches, for instance, stable operation result, high simulation precision, and so on [15]. These features stand out in cases involving incomplete known geographical distribution data and the relevant environment variables that affect distributions [16,17].
The Juniperus rigida Sieb. et Zucc., Cupressaceae plant with the characteristics of hardiness, drought resistance, and soil and water conservation, is endemic to East Asia. The species is mainly distributed in cold temperate regions of the northern hemisphere, North and Northeastern China, Korea, and Japan. It is the key protected plant and almost endangered species now in China [18]. It is climate warming and its biological characteristics that shrink and fragment its distributions [19]. Therefore, it is meaningful to figure out the impact of climate change on its distribution in terms of formulating effective conservation strategies for J. rigida.
Here, we used MaxEnt to project the potential distribution of Juniperus rigida Sieb. et Zucc. under current and future climate scenarios and to analyze the change of potential distribution in three main distributions across China. The specific objectives are (1) to project the potential distribution areas of J. rigida across China, (2) to assess the important variables affecting the distribution of J. rigida, and (3) to analyze the rules of the area change and the centroid shift in the three main distributions.
2. Materials and Methods
2.1. Collection and Processing of Species Distribution Data
The current occurrence records of Juniperus rigida Sieb. et Zucc. throughout China derived from the Global Biodiversity Information Facility (GBIF,
2.2. Collection and Selection Climate, Soil, and Topographical Data
The current period (1950–2000) bioclimatic variables with 30 arc seconds (approximately 1 km2) spatial resolution were downloaded from the WorldClim-Global Climate Database (
Future period (2050s and 2070s) bioclimatic variables with the same geographical accuracy and variable names with current bioclimatic variables were extrapolated by a global circulation model. There were four representative concentration pathways (RCPs) established in the 5th report of the IPCC (Intergovernmental Panel on Climate Change) [22], which means the future climate change was diverse and uncertain. RCP 2.6 was a lower greenhouse gas emission scenario and was paid widespread attention because it could achieve the goal of no more than 2 °C of global warming by 2100 relative to pre-industrial levels [23,24]. Holding the increase in the global average temperature below 2 °C above pre-industrial levels and striving to limit the increase to 1.5 °C above pre-industrial levels is the common goal for all countries in the world by reducing greenhouse gas emissions [25]. RCP 8.5 was a robust GHG emission scenario without additional efforts to constrain emissions, which posed a high risk of abrupt and irreversible regional-scale change in the composition, structure, and function of ecosystems [22]. Therefore, we chose RCP 2.6 and RCP 8.5 under BCC-CSM1-1 climate change modeling, an efficient global climate tool for the simulation of future climatic conditions that performs perfectly for the simulation of precipitation and temperature in the Chinese region [26,27]. There were two periods for model projection; the 2050s and 2070s represent 2041–2060 and 2061–2080, respectively.
Insolation duration and relative humidity were provided by the China Meteorological Data Service Center (CMDC,
All the data were processed with the same spatial resolution as the climate data. Many of the environmental variables that are spatially correlated can result in model over-fitting [16]. We used the Pearson correlation coefficient to examine the multi-collinearity of variables and removed one or more highly correlated environmental factors if p > 0.8. Ultimately, we retained 22 factors for modeling (Table 1) and converted the data to the ASCII format.
2.3. Model Establishment and Accuracy Test
The maximum entropy model (MaxEnt, version 3.4.1) is available freely online (
The fitted model precision was evaluated based on the area under the receiver operating characteristic curve (AUC-ROC), with the criterion of values for AUC > 0.9 representing more adequate fit [30] by 10 fold cross-validation. The receiver operating characteristic curve (ROC) graphically displayed the performance of the model [31]. The output was in cloglog format with the probability (0~1) of the presence of J. rigida in each grid of 30 arc-seconds resolution. The maximum training sensitivity plus specificity criterion we utilized had been recognized as one of the best threshold (TH) selection methods on account of the optimized specificity and sensitivity using training data [32]. The maximum number of background points and iterations were set to 10,000 and 3000, respectively, which ensured that the model had adequate time to converge, while the other parameters defaulted according to the MaxEnt Tutorial [33].
2.4. Potential Habitat Assessment
We classified the distribution into 4 levels, unsuitable (<TH), general (TH~0.5), moderate (0.5~0.8), and core (0.8~1), and illustrated them in pictures, according to the suitable indices for each climate scenario provided by MaxEnt.
The distribution changes in the future climate scenarios were in three patterns, new, disappeared, and unchanged, compared with current habitat. To illustrate the change in future scenarios, we replaced the continuous probability value with the discrete binary and recalculated the grid value using the following formula. The grids with a probability above the threshold were assigned to 1 representing suitability habitats, and the others to 0 representing unsuitability habitats.
It can be calculated using the formula following.
X = GCC × 2 + GCF, X∈(0, 1, 2, 3)(1)
where GCC was the grid code under the current climate, GCF was the grid code under the future climate. When X = 0 represented the unsuitable distribution under current and future climate, X = 1 represented the new distribution under future climate, X = 2 represented the disappeared distribution under future climate, and X = 3 represented the unchanged distribution.2.5. The Shift of the Core Distribution Centroid
We assumed the core distribution area of Juniperus rigida Sieb. et Zucc. to a single centroid point that could be calculated with the help of the SDM tool-box, a type of python-based GIS software. The core distribution is the geospatially highly suitable area for J. rigida. The magnitude and direction of distribution area change were analyzed more clearly by comparing the centroid point position under the current and future climate.
3. Results
3.1. Model Evaluation and Important Variables
The MaxEnt model for Juniperus rigida Sieb. et Zucc. generated excellent evaluations under the given parameter setting conditions, with the average test AUC value of 0.928 ± 0.025 for the 10 times replicate runs. Making a comprehensive analysis of the percent contribution and permutation importance of all predictive variables provided by the MaxEnt result, those variables, slope (Slo), altitude (Alt), mean temperature of driest quarter (MTDq), and temperature annual range (TAR), were the domain environmental factors with higher percent of contribution relative to other variables. Furthermore, annual mean temperature (AMT), annual precipitation (AP), temperature seasonality (TS), and precipitation of wettest month (PWm) made a greater contribution to the permutation importance. The cumulative contributions of these factors hit a value as high as 76.8% (Table 1).
3.2. The Potential Distribution of J. rigida under the Current Climate
The suitability was classified into four levels and was shown in Figure 2A. The simulated suitable habitat under the current climate scenario was mainly distributed in three regions, the Xinjiang region, the Loess-Inner Mongolian Plateau region, and the Changbai Mountain region, which were illustrated in Figure 2B–D, respectively.
The national ecological suitable area was 86.87 × 104 km2 and the core suitable area was 10.13 × 104 km2 accounting for 11.66% of the total potential habitat (Table 2). The suitable distribution of the Loess-Inner Mongolia plateau region was larger than other two regions, reaching 61.39 km2. The core ecological suitable area in this region accounted for 13.04% of the total suitable area, and the percentage was higher than the national average.
The Xingjiang region had the smallest ecological suitable area, with the area size of 5.65 km2. The proportions of the moderate and core suitable habitat were 19.25% and 3.89% respectively, which were significantly lower than the national average.
3.3. Diversified Changes of Suitable Habitat Areas among Different Regions under Future Climate
Under the future climate scenario, the variation trends of the suitable habitat area and the pattern of increase–decrease were not exactly the same among the three regions. However, by the period of the 2070s, the area of the newly increased and disappeared suitable habitat in the three regions gradually decreased compared with that in the previous period.
As far as the Loess-Inner Mongolian plateau region was concerned, in the 2050s relative to the current distribution, the area of the potential habitat increased to 74.04 × 104 and 72.10 × 104 km2 (Table 3). The new areas were mainly distributed around the unchanged area, while the disappeared habitat is mainly in the west and northeast of the area under the two climate scenarios (Figure 3(A1,A3)). By the 2070s, the total habitat area decreased to 68.38 × 104 and 58.43 × 104 km2, respectively (Table 3), and the disappeared and new regions will be distributed almost randomly around the suitable regions relative to the 2050s (Figure 3(A2,A4)).
For Xinjiang region, the new habitat area was in the northwest of Xinjiang in 2050s under the two scenarios (Figure 3(B1,B3)), amounting to 4.01 × 104 and 3.42 × 104 km2 (Tabble 3), and the area of the ecological suitable raised 2.67 × 104 km2 with the percentage of 47.16 under RCP 2.6 in 2050s relative to current distribution, 1.89 × 104 km2 for RCP 8.5. The area of ecological suitable loss was 1.35 × 104 and 1.53 × 104 km2, respectively, projected for the south of the Tianshan mountains (Figure 3(B1,B3)). By the 2070s, the total habitat area went down to 7.90 × 104 and 5.6 × 104 km2, and the disappeared and new regions appeared in a sporadic pattern throughout the entire distribution area (Figure 3(B2,B4)).
For the Changbai mountain region, the areas of the total, new, and disappeared ecological habitat were 13.81 × 104, 2.06 × 104, and 2.89 × 104 km2, respectively, in the 2050s under RCP 2.6 and 13.23 × 104, 1.94 × 104, and 3.52 × 104 km2 for RCP 8.5 (Table 3). The northern region disappeared seriously, while the central region gained more area (Figure 3(C1,C3)). By the 2070s, the total suitable area remained down continuously to 13.23 × 104 and 10.20 × 104 km2 relative to the 2050s (Table 3, Figure 3(C2,C4)).
3.4. Varied Spatial Shifts of Core Distribution in China under the Future Climate Scenario
In the Loess-Inner Mongolian plateau region, the geographical centroid of the ecological habitat of Juniperus rigida Sieb. et Zucc. was located at 113.85° E and 40.16° N where the elevation was 1518 m, migrated northeastward about 18.99 km to the position of 113.91° E and 40.33° N in the 2050s, and then shifted southwestward about 15.95 km to 113.73° E and 40.29° N in the 2070s under the RCP 2.6 scenario. There were the same pattern under RCP 8.5 to RCP 2.6. The centroid moved northeastward about 35.62 km and went southwestward 59.79 km. The altitude of the geographical centroid for core distribution decreased and then increased, from 1518 m down to 1373 m and then up to 1464 m under RCP 2.6, and from 1518 m fell to 1373 m and rose to 1618 m under RCP 8.5 (Table 4, Figure 4(A1,A2)).
Talk about the Xinjiang region, the geographical centroid of the core habitat of Juniperus rigida Sieb. et Zucc., moved northwestward under the two scenarios on the whole. The center started at the position of 89.17° E and 49.36° N and stopped at 85.03° E, 44.49° N for RCP 2.6, 84.91° E, 44.35° N for RCP 8.5. Similarly, the elevation of the geographical centroid for core area showed an escalating trend and then task back under the two scenarios. The initial altitude was 2138 m, and in the 2050s descended to 2004 and 2099 m, respectively, afterwards ascending to 2070 m for RCP 2.6 and 2104 m for RCP 8.5 (Table 4, Figure 4(B1,B2)).
There was a difference in the Changbai mountain region compared with the other two regions in terms of migration direction and elevation change. The geographical centroid migrated southwesterly without changing direction, and the elevation raised continuously over the future two periods under the two scenarios. It started at 127.44° E, 42.24° N and then went through 125.27° E, 41.21° N, finally ending at 125.01° E, 41.18° N under RCP 2.6 with the distance of about 227.47 km and positioned at 124.65° E, 40.06° N with about 260.32 km for RCP 8.5. The altitude ran upward from 450 to 477 and 484 respectively (Table 4, Figure 4(C1,C2)).
4. Discussion
4.1. Suitable Distribution in Current Climate Scenario
In this study, we calibrated the MaxEnt model and projected the distribution regions of J. rigida on the national scale under the current and future climate scenario base on the climate, soil, and topographical variables. Overall, the model showed a perfect performance under all the climate scenarios [34], with the AUC 0.928 ± 0.025.
According to the estimated result, the distribution regions of Juniperus rigida Sieb. et Zucc. concentrated on three regions under current and future climate scenarios, which were renamed Xinjiang region, Loess-Inner Mongolian plateau region, and Changbai Mountain region from west to east (Figure 2). The result was in general agreement with the previous occurrence recorded in nine provinces, including Heilongjiang, Jinlin, Liaoning, Inner Mongolian, Shaanxi, Shanxi, Gansu, Ningxia, and Hebei [35]. There was ca. 86.87 × 104 km2 area for the projected suitable habitat of J. rigida in China, and the largest area was located in the Loess-Inner Mongolian plateau region, whose area was ca. 61.39 × 104 km2 with 70.59% of the national ecological suitable area (Table 2). There was a region that we should emphasize, the Xingjiang region. Although there were only three natural occurrence records ever in this region, we projected about 5.65 × 104 km2 suitable distribution in this region. In the background of global climate warming, the arid climate and less rainfall coincided with its physiological characteristics [35] and provided inhabitable conditions for sprouting and growth of J. rigida in this region [36]. Furthermore, the Changbai Mountain region was also the major distribution area, which can be verified by the previous studies on species distribution to some degree [37].
4.2. Ecological Variables Influencing the Spatial Distribution Pattern of J. rigida
The geographical spatial distribution of plants is the result of long-term interaction between plants and the environment. Nevertheless, environmental factors do not affect the geographical spatial distribution equally. There were four identified variables that played a significant role in influencing the distribution pattern among all the factors that characterized the ecological habitat of Juniperus rigida Sieb. et Zucc. by running the model. The properties of J. rigida, light-loving and drought-resistant [35], made the specific requirements for slope (Slo) and altitude (Alt) in different distribution regions. The distribution elevation was 300~600 m in the Changbai mountain region, which was the lowest distribution area, and was ca. 1100~1600 m in Yinshan mountain, 1100~1300 m for the Erdos plateau, and 2000~2500 m for Helan mountain [35]. The complex topography was responsible for the diverse climate on a regional scale, and the slope affected the light intensity, evaporation, soil nutrients, and available water content to some extent.
The two factors associated to the temperature, mean temperature of the driest quarter (MTDq) and temperature annual range (TAR), also played important roles in affecting the distribution. It is stated that temperature was the dominant factor in the progress of the vegetative growth [38], flower bud differentiation [39], and seed dormancy and germination [40]. Temperature could germinate seeds and shortened the dormancy period caused by the encased in hard seed coat and oil [41]. The optimum germination temperature for Juniperus rigida Sieb. et Zucc. seeds was 15~25 °C, and the light could inspire its germination rate [19]. Global warming was accompanied by rising precipitation [42]. The effect of precipitation on the geographic distribution should not be ignored when we focus on the effect of temperature. We projected that the precipitation of the wettest month (PWm) and annual precipitation (AP) were also important factors in terms of percent contribution and permutation importance of all variables.
4.3. Prospective Change on Distribution in the Face of Climate Warming
With global warming, it was reported that species will either migrate [43] or change their physiology [44] to resist climate change. There may be a main trend of species migrating to high latitude or high elevation [45,46]. The mean altitude of the core suitable area gradually increased in Chanbai mountain regions, first up and then down in other two regions, over time, which may be a strategy for Juniperus rigida Sieb. et Zucc. to adapt to the changing environment (Table 3). In this respect, our results were completely consistent with those of previous studies [47]. The migration paths were almost the same in the same study area under the two different climate scenarios; nevertheless, they moved in different directions in different regions (Figure 4). The trajectory of core area centroid was triangular in the Loess-Inner Mongolian Plateau Region, first northeast and then southwest, and almost linear in the Xinjiang Region and Changbai Mountain region, northwest and southwest direction, respectively (Figurre 4). This phenomenon was mainly related to the complexity of climate change. The centroid of Changbai Mountain distribution migrated to low latitude that was different from the previous work [45,46]. The impact of global climate change on China was more obvious than other countries, and the temperature in Northeast of China rose faster than that in other regions [48]. All of those could clarify why the distribution centroid in Changbai Mountain migrated to lower latitudes rather than higher ones. In any region, the centroid shifted further under RCP 8.5 to RCP 2.6 scenario. RCP 8.5, a very high GHG emission scenario, led to robust global surface warming [22,23,24,25], affecting more prominently the distribution of J. rigida in China to RCP 2.6.
In addition to the migration of suitable areas, climate change also affected the area of suitable distribution. On the side, it is also affected the adaptability of species to climate change [49]. We gained the similar achievement of previous studies that showed that the suitable habitat range will increase as global warming intensity proceeds under the low concentration emissions pathway (RCP 2.6) [50], but excessive temperature raised might restrict plants’ diffusion under the powerful GHG emissions pathway (RCP 8.5) [51]. The suitable area increased and then decreased in the Xinjiang region and the Loess-Inner Mongolian plateau region and decreased continuously in Chanbai Mountain region in both scenarios. This may be caused by the more significant effect of climate warming and precipitation in the northeast of China than other regions [48,52]. Comparing the effects of climate change and precipitation on China [53], the change in temperature and precipitation were marked in the following areas where the area of disappeared suitable habitat were severe, including the north part of Chanbai Mountain and the south part of Taihang Mountain and Lvliang Mountain. On the contrary, the adjacent regions of Hebei, Inner Mongolia, and Shaanxi province were flooded with new distributions in the 2050s and 2070s, in which there were mild changea in temperature and precipitation under the stringent concentration emissions pathway (RCP 2.6) and the high greenhouse emission scenario (RCP 8.5) [48].
5. Conclusions
There, we projected the current and future distribution areas of Juniperus rigida Sieb. et Zucc. in China using the MaxEnt model under the climatic scenario of RCP 2.6 and RCP 8.5 for two periods, the 2050s and 2070s. These results indicated that the temperature-dependent variables affected the suitable habitat more than the factors associated with the precipitation. The suitable area increased first and then decreased on the whole, and the centroid of the core habitat migrated to high elevation. These insights were remarkable in the Changbai Mountain region under the same scenario and were more obvious under RCP 8.5 in any region. Though we gained meaningful results that will provide a more effective and evidence-based conservation strategy, they were not comprehensive, and other factors should be taken into account too, for instance, adaptability to climate change, diffusivity, and so on. However, these were unfamiliar to us in this paper.
Author Contributions
Conceptualization, D.L. and Z.L.; methodology, Z.L.; software, Z.L.; validation, D.L. and Z.L.; formal analysis, Z.L.; investigation, D.L. and Z.L.; resources, Z.L.; data curation, Z.L.; writing—original draft preparation, Z.L.; writing—review and editing, D.L.; visualization, Z.L.; supervision, D.L.; project administration, D.L.; funding acquisition, D.L. Both authors have read and agreed to the published version of the manuscript.
Funding
This work was supported by the program from the National Natural Science Foundation of China (No. 31570655).
Data Availability Statement
The bioclimatic variables are available from the WorldClim-Global Climate Database (
Conflicts of Interest
The authors declare no conflict of interest.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Figures and Tables
Enlarge this image.Figure 2. Potential distribution area of Juniperus rigida Sieb. et Zucc. under the current climate. The potential distribution of China was illustrated in (A), and (B), (C) and (D) shown the distribution of the Loess-Inner Mongolian plateau region, Xinjiang region and Changbai Mountain region, respectively.
Enlarge this image.Figure 3. The potential distribution range shift in the extent of Juniperus rigida Sieb. et Zucc. under the low concentration emissions pathway (RCP2.6) in three regions. The changes in the 2050s to current showed in (A1), (B1), and (C1) under RCP2.6, and in (A3), (B3), and (C3) under RCP 8.5. (A2), (B2), and (C2) illustrated the changes in 2070s relative to 2050s under RCP 2.6, (A4), (B4), and (C4) for RCP 8.5.
Enlarge this image.Figure 4. The core distributional shift under future climate scenarios for Juniperus rigida Sieb. et Zucc. in three main distribution areas. (A1) illustrated the shift path in the Loess-Inner Mongolian region—full figure, (A2) illustrated the shift path in the Loess-Inner Mongolian region—detail figure, (B1) for the Xinjiang region—full figure, (B2) for the Xinjiang region—detail figure, (C1) for the Changbai Mountain region—full figure, (C2) for the Changbai Mountain region—detail figure.
Percentage contributions and permutation importance of the environmental variables.
| Abbreviation | Variables Description | Percent Contribution (%) | Permutation Importance |
|---|---|---|---|
| Slo | Slope | 16.8763 | 7.3039 |
| Alt | Altitude | 15.7851 | 29.669 |
| MTDq | Mean Temperature of Driest Quarter | 12.8316 | 1.4061 |
| TAR | Temperature Annual Range | 11.6773 | 11.4124 |
| PWm | Precipitation of Wettest Month | 6.6254 | 0.7003 |
| SpH | Subsoil pH | 5.3958 | 1.2187 |
| PS | Precipitation Seasonality | 4.9589 | 3.4087 |
| RH | Relative Humidity | 3.9251 | 2.8779 |
| TS | Temperature Seasonality | 3.8679 | 5.0902 |
| Iso | Isothermality | 3.8106 | 9.43 |
| Ins | Insolation Hour | 3.3007 | 3.063 |
| Asp | Aspect | 2.1047 | 2.1337 |
| TpH | Topsoil pH | 1.9548 | 1.8882 |
| MDR | Mean Diurnal Range | 1.8851 | 3.2606 |
| AMT | Annual Mean Temperature | 1.4083 | 8.0216 |
| TTEX | Topsoil Texture | 1.2019 | 0.9042 |
| AWC | Available Water Content | 1.113 | 0.9102 |
| AP | Annual Precipitation | 0.6486 | 4.1459 |
| PCq | Precipitation of Coldest Quarter | 0.2826 | 1.3119 |
| SOC | Subsoil organic content | 0.1834 | 0.6614 |
| MTWm | Max Temperature of Warmest Month | 0.1526 | 1.1502 |
| TECE | Topsoil Electric Conductivity | 0.0105 | 0.0318 |
Summary of the area and proportion of various suitable distribution areas of J. rigida in China and three regions.
| Regions | General Distribution | Moderate Distribution | Core Distribution | Total | |||
|---|---|---|---|---|---|---|---|
| Area |
Percentage |
Area |
Percentage |
Area |
Percentage |
Area |
|
| China | 49.59 | 57.08 | 27.15 | 31.25 | 10.13 | 11.66 | 86.87 |
| Changbai Mountain region | 6.53 | 44.58 | 6.32 | 43.12 | 1.80 | 12.29 | 14.65 |
| Loess-Inner Mongolian |
34.95 | 56.93 | 18.43 | 30.02 | 8.01 | 13.04 | 61.39 |
| Xinjiang Region | 4.06 | 71.82 | 1.39 | 19.25 | 0.20 | 3.89 | 5.65 |
Area variation of the suitable habitat in the three kind regions under the future climate scenarios relative to current.
| Scenario | Type | Xinjiang Region |
Loess-Inner Mongolian |
Changbai Mountain |
|||
|---|---|---|---|---|---|---|---|
| 2050s | 2070s | 2050s | 2070s | 2050s | 2070s | ||
| Current | Total | 5.651 | 61.394 | 14.646 | |||
| RCP 2.6 | Disappeared | −1.348 | −1.329 | −5.789 | −7.253 | −2.887 | −2.891 |
| New | 4.012 | 3.580 | 18.434 | 14.237 | 2.057 | 2.055 | |
| Unchanged | 4.303 | 4.322 | 55.604 | 54.140 | 11.756 | 11.753 | |
| Total | 8.316 | 7.902 | 74.039 | 68.377 | 13.813 | 13.808 | |
| RCP 8.5 | Disappeared | −1.525 | −2.415 | −6.094 | −12.693 | −3.352 | −5.244 |
| New | 3.416 | 2.362 | 16.799 | 9.729 | 1.943 | 0.806 | |
| Unchanged | 4.126 | 3.236 | 55.300 | 48.701 | 11.284 | 9.392 | |
| Total | 7.542 | 5.598 | 72.099 | 58.430 | 13.227 | 10.198 | |
The geographic coordinates of the core areas and the average elevation of the suitable regions. The distances represented the migration distance relative to the previous time point. The altitude means the average elevation of the core areas.
| Regions | Scenarios | Periods | Longitude | Latitude | Altitude (m) | Distance (km) |
|---|---|---|---|---|---|---|
| Changbai Mountain |
Current | 127.44° E | 42.24° N | 450 | —— | |
| RCP 2.6 | 2050s | 125.27° E | 41.21° N | 470 | 206.2 | |
| 2070s | 125.01° E | 41.18° N | 477 | 21.27 | ||
| RCP 8.5 | 2050s | 125.03° E | 41.10° N | 465 | 228.97 | |
| 2070s | 124.65° E | 41.06° N | 484 | 31.35 | ||
| Loess-Inner Mongolian |
Current | 113.85° E | 40.16° N | 1518 | —— | |
| RCP 2.6 | 2050s | 113.91° E | 40.33° N | 1373 | 18.99 | |
| 2070s | 113.73° E | 40.29° N | 1464 | 15.95 | ||
| RCP 8.5 | 2050s | 114.14° E | 40.41° N | 1373 | 35.62 | |
| 2070s | 113.41° E | 40.37° N | 1618 | 59.79 | ||
| Xinjiang Region | Current | 87.17° E | 43.76° N | 2138 | —— | |
| RCP 2.6 | 2050s | 85.03 ° E | 44.49° N | 2004 | 182.4 | |
| 2070s | 84.91° E | 44.35° N | 2070 | 18.15 | ||
| RCP 8.5 | 2050s | 85.76° E | 44.50° N | 2099 | 134.78 | |
| 2070s | 84.97° E | 44.51° N | 2104 | 60.28 | ||
© 2021 by the authors.
Details
; Li, Dengwu 1 1 College of Forestry, Northwest A & F University, Yangling, Xianyang 712100, China;