Predicting potentially suitable Bletilla striata habitats in China under future climate change scenarios using the optimized MaxEnt model

doi:10.21203/rs.3.rs-5216981/v1

Download PDF

Article

Predicting potentially suitable Bletilla striata habitats in China under future climate change scenarios using the optimized MaxEnt model

https://doi.org/10.21203/rs.3.rs-5216981/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Bletilla striata, an important traditional Chinese medicine resource, holds high medicinal and ornamental value. However, unscientific habitat selection for its cultivation has led to low yields and poor quality as medicinal materials. The optimized MaxEnt model is a powerful tool for analyzing the potential impacts of environmental factors on species distribution and predicting habitat changes under climate change. It offers great significance for the protection and development of B. striata in China. Based on 269 B. striata distribution records in China and 15 major environmental factors, this study simulated the distribution patterns of potentially suitable B. striata habitats under four different climate change scenarios (SSP2.6, SSP4.5, SSP7.0, and SSP8.5) and three time periods (the current period, 2050s, and 2070s). The analysis was conducted using the MaxEnt model which exhibited high predictive accuracy and minimal overfitting. Solar radiation, annual temperature range, mean diurnal range and vapor pressure were revealed as the dominant factors affecting B. striata distribution, and their thresholds were ≤ 16265.39 kJ/m²·d^− 1, ≤ 39.7℃, ≤ 12.6℃, and ≤ 2.9 kPa, respectively. The results showed that the total potentially suitable B. striata habitats in China were 30.07×10⁵ km² under current climate conditions, mainly distributed in 14 provinces or regions in southern China. Under future climate change conditions, the predicted potentially suitable B. striata habitats will decrease significantly over time, and the centroid of the predicted potentially suitable habitats at all levels will shift northward. The research results can guide future B. striata resource conservation, variety selection, and cultivation.

Earth and environmental sciences/Climate sciences/Climate change/Climate change impacts

Biological sciences/Plant sciences/Plant ecology

potential suitable area

climate change

environmental variable

MaxEnt model

distribution centroid

shared socio-economic path

1. The optimized MaxEnt model has high discriminability and heterogeneity, generating a potential distribution map of

2. The core suitable habitat regions of were located in Sichuan, Chongqing, Hubei, Hunan, and Guizhou provinces.

3. habitats will decrease significantly over time, and the centroid of the predicted potentially suitable habitats at all levels will shift northward.

4. In the process of artificial planting management, the adverse effects of environmental factors on growth can be reversed by regulating the water or light during key months.

Included in the List of National Key Protected Wild Plants in 2021, B. striata is an important traditional Chinese medicine resource found in forests, roadside grasslands, and rock crevices at 100 to 3200 m altitudes. B. striata is mainly cultivated in Sichuan, Guizhou, Shaanxi, Anhui, Zhejiang, Hubei, Hunan, Guangdong, and Guangxi ^[1–2]. B. striata is noted for its properties of astringency, hemostasis, swelling reduction, and muscle regeneration in ancient Chinese medical texts such as the Shennong Bencao Jing and the Compendium of Materia Medica. B. striata polysaccharides exhibit various pharmacological activities, such as hemostasis, gastric mucosal protection, anti-tumor effects, anti-fibrosis, wound healing promotion, and plasma replacement ^[3–4]. The B. striata gum extracted from the tubers of B. striata plants also serves as an auxiliary material in various medicines, including insoluble drug carriers, topical hydrogels, wound dressings, targeted drugs, and biological scaffolds ^[5–6]. In cosmetics, B. striata gum is a non-toxic and harmless plant-based additive that can moisturize the skin and delay its aging ^[7]. With the recent increase in industrial production of relevant products, the demand for B. striata has been increasing year by year.

However, the recent increase in extreme weather conditions due to climate changes has intensified the heat index and stress, creating a severe crisis in the agricultural sector worldwide ^[8–9]. This environmental pressure limits B. striata growth and development and disturbs its different metabolic and physiological activities. Meanwhile, the sensitivity of B. striata to high temperatures leads to increased pests and diseases, decreased adaptability, and decreased yield ^[10–11].

Species distribution models (SDMs) empirically quantify the existing environmental niches of species and predict the underlying species distribution by correlating the distribution data to a set of environmental variables ^[12–13]. Among them, MaxEnt is a common SDM to model species distribution with presence-only species distribution data. It calculates the probability distribution of the maximum entropy based on constraints and remains relatively robust even with small sample sizes ^[14–15]. Thus, it has been widely used to predict species distribution in recent years ^[16–18]. Previous studies have assessed the potential impact of climate change on species to support rare and endangered species conservation ^[19–21]. Scholars also assessed species invasion, tested evolutionary hypotheses, and predicted disease transmission risks ^[22–25]. However, studies also showed that complex species-environment relationships may accurately fit a certain dataset but not others ^[26–28]. The MaxEnt model performs poorly without model setting adjustments and model optimization ^[29–30]. Numerous studies have demonstrated the significantly improved predictive performance of optimized MaxEnt and its more accurate and reasonable species distribution maps ^{[23, 31]}.

Through simulation analysis, potentially suitable B. striata habitats in China can be predicted and categorized into different suitability levels, providing scientific references for its habitat selection and protection ^[32–35].A clear understanding of geographical distribution is foundational and instrumental for effective B. striata protection and management. Therefore, this research aims to (1) use an optimized MaxEnt to predict the distribution of currently suitable B. striata habitats, (2) determine the distribution of important environmental variables and the limitations of ecological characteristics, and (3) predict future changes in the potentially suitable B. striata habitats. By analyzing the distribution trend under future climate changes, this study provides a reasonable reference and decision-making basis for the long-term management and protection of B. striata.

2.1. B. striata distribution data sources

Detailed distribution information in the form of latitude and longitude data was acquired from the literature and various sources, including the Global Biodiversity Information Facility (GBIF, https://www.gbif.org), Chinese Virtual Herbarium (CVH, http://v5.cvh.org.cn/), Chinese Flora Herbarium (CFH, https://www.cfh.ac.cn/), and National Specimen Information Infrastructure (NSII, http://www.nsii.org.cn/), and specimen coordinates were derived in Google Earth. To prevent spatial autocorrelation of species distribution data and match climate data at 2.5 arcminutes resolution, only one data point was retained within a diameter of 5 km ^[12]. Finally, 269 origin coordinates of B. striata were obtained to construct the maximum entropy model (Fig. 1).

2.2. Environmental data sources and screening

The 24 bioenvironmental climate data in this study are among the most important environmental factors affecting the potential species distribution and are often employed to construct niche models ^[36]. The current and future climate data (2021–2040) were collected from WorldClim (http://www.worldclim.org). Among them, the future climate data were based on the BCC-CSM2-MR Climate model for China's geographical environment in CMIP6. The included sustainable development path was limited to 2°C (SSP2.6), the moderate development path was limited to 3°C (SSP4.5), the local development path was limited to 4.1°C (SSP7.0), and the conventional development path was limited to 5°C (SSP8.5) ^[37]. The slope and slope direction data were extracted from the digital elevation data of the world climate data using ArcGIS. The surface solar radiation was obtained from the National Tibetan Plateau Data Center (http://data.tpdc.ac.cn). When simulating species geographic distribution, the MaxEnt model should be used first to avoid over-fitting model prediction results caused by the multicollinearity of various environmental variables. The importance and contribution rate of the environmental variables should be tested through Jackknife, and factors with a contribution rate ≥ 0.1% should be considered. Factors with a correlation coefficient | r | ≥ 0.8 and a low contribution rate are removed ^[38]. Then, the distribution data and environment variables were imported into ArcGIS, the sample tool within the extraction tool was used for sampling, and the dot and tape layers were extracted into Excel tables. Then, the correlation between B. striata distribution and climate factor variables was analyzed in SPSS, and environmental variables with a correlation coefficient above 0.95 were retained. Finally, ENMTools was used to conduct Pearson correlation analysis on the environmental variables, and Pearson correlation heat maps were plotted in R (Fig. 2). Variables with a correlation coefficient above 0.8 but a small contribution to the prediction were excluded ^[39]. First, the Jackknife method in MaxEnt was used to evaluate the contribution rate of the 24 environmental factors to the distribution of potential B. striata habitats, and Spearman correlation analysis was performed on these factors in SPSS (Fig. 1) Second, efforts were made to avoid over-fitting. Specifically, factors with a contribution rate ≥ 0.1% were selected, and those with a correlation coefficient | r | ≥ 0.8 but a low contribution rate were removed (Fig. 2). Finally, 15 factors with high influence on the distribution of potential B. striata habitats were identified and selected for modeling. These include climate factors (bio1, bio2, bio3, bio7, bio8, bio10, bio12, bio15, bio17, and bio18), altitude (ALT), solar radiation (SRAD), VAPR, SLOPE, and ASPECT (Table 1).

Table 1

Potential environmental factors affecting the distribution of *B. striata.*
Environment variable	Interpretation	Environment variable	Interpretation
bio_1	Annual mean temperature (℃)	bio_13	Precipitation of wettest month (mm)
bio_2	Mean diurnal range[Mean of monthly (maxtemp-mintemp)] (℃)	bio_14	Precipitation of driest month (mm)
bio_3	Isothermality (bio_2/bio_7) (×100)	bio_15	Precipitation seasonality (Coefficient of variation) (mm)
bio_4	Temperature seasonality (standarddeviation×100)	bio_16	Precipitation of wettest quarter (mm)
bio_5	Maxtemperature of warmest month (℃)	bio_17	Precipitation of driestquarter (mm)
bio_6	Mintemperature of coldestmonth (℃)	bio_18	Precipitation of warmestquarter (mm)
bio_7	Temperature annual range (bio_5-bio_6) (℃)	bio_19	Precipitation of coldestquarter (mm)
bio_8	Meantemperature of wettestquarter (℃)	SRAD	Solarradiation (J/m²·d)
bio_9	Meantemperature of driestquarter (℃)	VAPR	Vaporpressure (Pa)
bio_10	Meantemperature of warmestquarter (℃)	ALT	Altitude (m)
bio_11	Meantemperature of coldestquarter (℃)	SLOPE	Slope (°)
bio_12	Annualprecipitation (mm)	ASPECT	Aspect

2.3. Maximum entropy model establishment, optimization, and evaluation

The "kuenm" package in R 3.6.3 was used to optimize the MaxEnt model parameters. The regularization multiplier (RM) was from 0.5 to 4 (increment 0.5), and the feature classes (FC) adopt the following combinations: L, LQ, H, LQH, LQHP, and LQHPT (L: linear, Q: quadratic, H: hinged, P: Product, T: threshold) ^[40].The "kuenm" package was used to test the above parameter combinations. The model's fit and complexity were evaluated based on the delta AICc value of the Akaike Information Criterion, and the parameter combination with the smallest delta AICc was selected for model construction ^[41]. Furthermore, the area under the Receiver Operating Characteristic (ROC) curve (AUC value) was used to evaluate and verify the accuracy of model results ^[42]. An AUC value below 0.6 indicates model failure, a 0.6 to 0.7 AUC is poor, a 0.7 to 0.8 AUC is fair, a 0.8 to 0.9 AUC is good, and an AUC above 0.9 is excellent ^[37]. Finally, the B. striata occurrence point data and the optimized environmental factors were imported into MaxEnt, and the optimized model parameter combination was repeated 10 times to produce the prediction results of the current potential B. striata distribution. ROC test was performed, and the contribution rates of the environmental factors were obtained using the Jackknife method. The prediction results were imported into ArcGIS 10.8 for reclassification, and the threshold of training sensitivity plus specificity was considered a more robust method for species classification ^[43]. The training sensitivity plus specificity threshold of the model prediction results after 10 iterations was 0.19 ± 0.06, and the potential B. striata habitats were classified into four fitness grades: non-fitness areas (0 to 0.2), low fitness areas (0.2 to 0.4), medium fitness areas (0.4 to 0.6), and high fitness areas (0.6 to 1). Finally, ArcGIS 10.8 was used to visualize the prediction results.

2.4. Optimized MaxEnt modeling

MaxEnt fits a model with a series of feature classes and controls the model complexity via a regularization multiplier that imposes a penalty for additional constraints ^[14]. The feature classes and regularization multiplier greatly influence the accuracy of the MaxEnt model ^[44]. Thus, we selected the most appropriate model by evaluating different combinations of feature classes and regularization multipliers ^[30]. All occurrence data were randomly split into a training set (75% of the occurrence data) and a test set (the remaining 25%) ^[45]. In total, 527 candidate models were created by combining 16 regularization multiplier settings (1.1 to 1.9 at a 0.1 interval and 0.5 to 4 at a 0.5 interval) and all 31 combinations of five feature class settings (l, q, p, t, h) with the "kuenm" package in R. Then, we evaluated the 527 candidate models' performance (Appendix C) based on significance (partial ROC, with 100 iterations and 50% of the data for bootstrapping), omission rates (E = 5%), and model complexity (AICc) ^[46]. The final MaxEnt model was built in MaxEnt 3.4.1 with the selected parameters; the combinations of the feature class were qpt, and the regularization multiplier was 1.9. To reduce the uncertainty of the MaxEnt model, 10 iterations of cross-validation were used to build the model, and the final results were the average of the 10 iterations (Fig. 3).

2.5. Potential B. striata distribution changes and centroid shift based on future CMIP6

The ASCⅡ files generated by the MaxEnt model were imported into ArcGIS 10.8 and converted into raster data to calculate the changes in B. striata and distribution areas of different fitness grades under different scenarios in the future and current times ^[47]. The SDM toolkit of ArcGIS 10.8 was used to compare the changes in the area and scope of the suitable habitats under the future and current scenarios, allowing for the identification of the expansion and contraction areas and the geographical locations in the future. Taking 0.2 as the threshold, the data was converted into binary data, and the centroid of the suitable habitats and its change direction were calculated using the SDM toolbox ^[48].

3.1. Model optimization and accuracy evaluation

The MaxEnt model based on the 269 distribution sites and the selected environmental variables was used to simulate and predict the potentially suitable B. striata habitats. According to the model optimization results, a p-value (pval_pROC) of 0 indicates statistical significance. The omission rate of 0.0333% is below the 5% threshold and the default value of the MaxEnt model. The W_AICc of 1 is below 2, and its corresponding RM and FC are 0.1 and L. The ROC curve has an area of 0.5 to 1.0. The prediction reliability of the model is generally considered low when the AUC value is between 0.5 and 0.7. When the AUC is between 0.7 and 0.9, the prediction reliability of the model is medium. When the AUC exceeds 0.9, the model has the highest predictive reliability. Our AUC of 0.89 indicates a high predictive reliability of the model ^[49] (Fig. 4).

3.2. Currently suitable B. striata habitat distribution

The potential B. striata habitats are mainly located in Southwest China and parts of South China. High suitability areas are mainly in Guizhou and most parts of Chongqing, southeast Sichuan, northeast Yunnan, Hubei, Hunan, Guangxi, and other regions. The high suitability area is about 94,200 km², accounting for 31.33% of the total suitable habitats. The medium suitability area is about 121,300 km², accounting for 40.34% of the total suitable habitats. The low suitability area is about 85,200 km², accounting for 28.33% of the total suitable habitats (Fig. 5).

3.3. Major environmental factors affecting B. striata resource distribution

Among the 15 environmental factors used for modeling, the contribution rate of climate factors SRAD, bio7, bio2, and VAPR reached 91.6%, making them the leading environmental factors affecting potentially suitable B. striata habitats. The average contribution rate of SRAD was 45.5%, and the average contribution rate of bio7 was 39.4%, making them the two main environmental factors. The average contribution rates of the remaining environmental factors were below 5%, making them general environmental factors. Regarding the permutation importance of climate factors among the 15 environmental factors, SRAD, bio7, bio2, bio15, and bio18 accounted for 85.2% of the importance of species distribution. Among them, bio7 contributed the most at 48.3%. The average contribution rate of temperature variable factors (bio1, bio2, bio3, bio7, bio8, and bio10) (61.7%) was significantly higher than that of precipitation variable factors (bio12, bio15, bio17, and bio18) (19.5%) (Table 2). The thresholds of the dominant environmental factors were calculated. Specifically, the threshold of solar radiation was 9292.4 to 16265.39 kJ/m^− 2·d^− 1, that of the annual temperature range (bio7) was 8.76℃ to 39.7℃, that of the mean diurnal range (bio2) was 3.81℃ to 12.6℃, and that of the vapor pressure (VAPR) was 0 to 2.9 kPa (Fig. 6).

Table 2

Analysis of the contribution rate and importance of the distribution of *B. Striata* potential habitats.
Variable	Percent contribution (%)	Permutation importance(%)
SRAD	45.5	14
bio7	39.4	48.2
bio2	3.4	8.4
VAPR	3.3	0.6
ALT	3	3.7
bio1	1.5	3.8
bio15	1.2	6.7
SLOPE	0.6	0
bio12	0.4	4.5
bio3	0.4	0.9
bio18	0.4	7.9
ASPECT	0.4	0.4
bio17	0.2	0.4
bio10	0.2	0.2
bio8	0.1	0.1

3.4. The area and suitability level changes of B. striata habitats under future climate changes (CMIP6 scenario)

Under the four shared social and economic paths, the variation trends of B. striata habitats and suitability levels were basically the same, while the predicted area of high suitability habitats showed obvious declines under different climate scenarios, with a decline amplitude of over 60%. Under the climate scenarios of SSP4.5 and SSP8.5, the area of high-suitability B. striata habitats in 2050 decreased significantly, reaching 95.22% and 93.52%, respectively. Under the climate scenarios of SSP2.6 and SSP8.5, the area of high-suitability B. striata habitats decreased significantly in 2070, reaching 94.27% and 88.32%, respectively. The predicted area of medium-suitability B. striata habitats also decreased to a certain extent, and the area of medium-suitability B. striata habitats declined the most, reaching 75.93%, in 2070 under the climate scenario of SSP2.6 (Table 3, Fig. 7). These results showed that the growth of SSP8.5 was most restricted under the conventional development path with the growth in high temperatures. In addition to the SSP2.6 condition, the area of suitable B. striata habitats decreased linearly, and the other climatic conditions showed fluctuating declines.

Table 3

Suitable area of *B. Striata* under different climate scenarios in the present, 2050 and 2070.
Climate Scenarios	Highly (×10⁵ km²)	Medium (×10⁵km²)	Low (×10⁵km²)	Total (×10⁵km²)
Current	9.42	12.13	8.52	30.07
2050-SSP2.6	1.98	6.9	18.08	26.96
2050-SSP4.5	0.45	8.58	17.94	26.97
2050-SSP7.0	2.4	6.12	17.73	26.25
2050-SSP8.5	0.61	5.1	18.34	24.05
2070-SSP2.6	0.54	3.09	16.52	20.15
2070-SSP4.5	3.46	8.61	15.87	27.94
2070-SSP7.0	3.61	6.73	17.65	27.99
2070-SSP8.5	1.1	5.1	15.8	22

3.5. The change range and centroid transfer characteristics of B. striata habitats under future

climate changes (CMIP6 scenario)

Under the four different climate scenarios, the suitable B. striata habitats in 2050 and 2070 showed significant decreasing trends compared with the current, and the decreases were mainly in Guizhou, Chongqing, northeast Yunnan, and southeast Sichuan. Over time, the centroid gradually shifted to higher latitudes (Fig. 8). In the SSP4.5 scenario, the suitable B. striata habitats showed a trend of decreasing first and then increasing over time. The centroid transfer direction also presented irregular changes from eastward to westward to eastward again.

4.1. Model predictive performance

B. striata is a rare plant genus valued for its medicinal and ornamental purposes. Due to the extremely low germination rate of sexually propagated seeds under natural conditions, climate change, and human exploitation, wild B. striata resources are being depleted. A detailed understanding of its distribution is the prerequisite for understanding its rational distribution and utilization in the ecosystem. This study conducts a detailed analysis of global potential B. striata habitats under current and future climate conditions based on the MaxEnt model, whose good performance ensures accurate prediction of species distribution in relation to local climate conditions. Key environmental factors affecting B. striata distribution include solar radiation, annual temperature range, mean diurnal range, and vapor pressure. Under various climate change scenarios, the model predicts the potential expansion of suitable B. striata habitats in different regions during the 2050s and 2070s. The results provide a theoretical basis for formulating practical B. striata cultivation and management measures. MaxEnt proves capable of performing and predicting independently, as opposed to an integrated approach that combines many commonly used and highly regarded algorithms to identify important areas for medicinal plant conservation ^[50]. These findings do not necessarily mean that MaxEnt is a better technique than other methods, and there are still cases where it is less appropriate. The accuracy of SDMs is largely influenced by the occurrence record, the study area, and the choice of environmental factors ^[51]. However, MaxEnt should still be considered one of the most reliable and available techniques when modeling species distributions from incomplete data ^[52–54]. We believe it could be effectively applied to identify issues in important species protection areas. This study only used bioclimatic variables as predictors, while species distributions may also be influenced by other factors (biological factors, soil, evolution, dispersal capacity, etc.) ^[55]. However, as with most plants worldwide, ecological information related to B. striata does not exist. Despite these inevitable uncertainties in model predictions, SDMs remain useful macro-ecological tools for exploring the dynamic relationship between species distribution and climate conditions ^[56].

4.2. Potential distribution area and ecological characteristics of B. striata

Previous studies have shown that B. striata in China is mainly distributed south of the Qinling-Huaihe Line, and its endemic provinces are Yunnan, Hubei, Sichuan, Hunan, Jiangxi, and Zhejiang ^[57]. The minimum temperature in April and October, the annual temperature range, and the average precipitation in November are the most important meteorological factors affecting the potential distribution of B. striata. Our results are similar to those of previous research in that the most suitable habitats are in Yunnan, Sichuan, Chongqing, Guizhou, Hunan, Hubei, Shaanxi, and other places, and the most important meteorological factors affecting its potential distribution are SRAD, bio7, bio2, and VAPR. However, the predicted results of contiguously suitable habitats did not include Guizhou, Shaanxi, and other places, which is inconsistent with the actual situation, indicating the ability of the model in this study to identify suitable habitats. Moreover, the contribution rate of the newly introduced factor SRAD accounts for 45.5%, which also shows the importance of illumination. These results also showed that vapor pressure significantly affected B. striata distribution. Vapor pressure plays an important role in the relationship between plants and water resources, and increased vapor pressure causes the closure of plant stomata, reduces water loss, and avoids adverse water tension conditions within the xylem ^[58–59]. However, little is currently known about the effect of vapor pressure on herb performance, and future controlled quantitative studies may address this issue.

4.3. Effects of climate change on future B. striata distribution

Obviously, the increased temperature greatly hinders B. striata growth, and in all scenarios, the area of high-suitability B. striata habitats has contracted to different degrees (over 67.8%). Under the RCP4.5 scenario in 2050, the high-suitability B. striata habitats show the largest reduction at 95.5%. On a large spatial scale, the future distribution of B. striata has undergone great changes, and the optimal habitat area is stable at below 20.0%. According to the Blue Book on Climate Change of China 2023, China will continue warming in the future, with an increasing climate risk index and the likelihood of extreme weather disasters, posing great challenges to the survival of B. striata. Environmental temperature, humidity, and seasonal changes will have a series of cascading effects on plant growth. Considering the influence of these factors, we believe that the actual B. striata distribution will face the risk of further shrinking and fragmentation in the future, and its living environment will not be optimistic. Therefore, establishing nature reserves and national forest parks is an important countermeasure, and those species on the brink of extinction urgently require strengthened human intervention to restore their populations.

4.4. B. striata habitat selection and protection strategies

Effective conservation of rare and endangered plants, such as B. striata, is essential. Based on their survival status and population distribution, we believe that the following protection policies and measures should be adopted to enhance protection efforts. (1) Strengthened efforts should be devoted to constructing nature reserves and protecting mountains and forests. Based on the situation of natural and biological resources, efforts should be made to upgrade certain nature reserves. Provincial and county-level nature reserves should be expanded when conditions permit, and new nature reserves should be established to protect rich biodiversity and special biological populations. Special attention should be paid to special and minimal populations. Moreover, the wild B. striata populations have been endangered by human exploitation, and their habitats must be prioritized in the conservation plans. (2) Research on B. striata germplasm resource collection and breeding should be strengthened. Rare and endangered plants are important germplasm genetic resources, and breaking reproductive barriers helps species escape survival crises. Priority should be given to the research on B. striata germplasm resource conservation and breeding excellent varieties, and efficient B. striata breeding technologies and high-quality seeds and seedlings should be explored. (3) Manual interventions should be conducted to expand B. striata planting bases. New planting bases can be established in suitable areas under the forest, and the existing bases can be improved using shade nets and intercropping with tall crops to address adverse factors such as high temperatures. These measures could promote the sustainable development of B. striata resources.

Under future climate change scenarios, the area of suitable B. striata habitats is significantly reduced, and their centroids shift northward. SRAD, bio7, bio2, and VAPR are the dominant environmental factors affecting B. striata distribution in China, and SRAD and bio7 are the key factors. The predicted results are generally consistent with the current B. striata distribution in China. However, due to anthropogenic factors and habitat loss, fragmentation, degradation, and severe damage, viable B. striata habitats may be further reduced and fragmented. Conditions for B. striata are expected to be tougher, and measures should be taken to enhance its conservation, utilization, and development. These include strengthening nature reserves and forest protection, promoting B. striata breeding and germplasm resource collection research, expanding B. striata planting bases, and improving B. striata resource management.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Author Contribution

M. L.: Writing—original draft, Writing—review and editing. P. Y.: Methodology, Investigation. L. Y.: Methodology, Project administration. Z. Z.: Data curation, Visualization, Supervision. H. L.: Writing—review and editing. M. W.: Conceptualization, Funding acquisition, Supervision.

Acknowledgements

This research was supported by the project of the Central Guidance on Local Science and Technology Development Fund of Guizhou Province (Grant No.[2023]007), the National Key Research and Development Program of China (Grant No. 2021YFD1601002), the Science and Technology Program of Guizhou Province (grant No. ZK[2024] General 554) and the Modern Agricultural Technology System of Traditional Chinese Medicine in Guizhou Province (GZYCCYJSTX-04). We thank Mingkai Wu for valuable advises on terminology.

Data Availability

All data used and analyzed during the current study are available from the corresponding author on reasonable request.

Chen, J. et al. Integrative analyses of transcriptome and metabolome shed light on the regulation of secondary metabolites in pseudobulbs of two Bletilla striata (Thunb.) Reichb. f. varieties. J. Appl. Res. Med. Aromatic Plants. 20, 100280. https://doi.org/10.1016/j.jarmap.2020.100280 (2021).
Zhang, M., Luo, D., Fang, H. L., Zhao, W. & Zheng, Y. Effect of light quality on the growth and main chemical composition of Bletilla striata. J. Plant Physiol. 272, 153690. https://doi.org/10.1016/j.jplph.2022.153690 (2022).
Kong, W. H., Xu, J. B. & Cui, Q. Research progress in chemical components and pharmacological actions of Bletilla striata and extraction technology of its polysaccharide. Inform. TCM. 38, 69–78 (2021).
Li, X. L., Zhang, X. G. & Yin, S. P. Polysaccharides of radix bletillae inhibit inflammatory reaction and oxidative stress in rats with ulcerative colitis. Basic. Clin. Med. 40 (2), 224–228 (2020).
Zhang, R. R. et al. Research Progress of Natural High Polymer Bletilla striata Polysaccharide in the Medical Field. Adv. Clin. Med. 10, 2926–2933 (2020).
Liu, X. W. et al. Experimental study of the Bletilla striata glucomanan composited scaffolds by orthogonal experiment. JOURNAL OF SHANDONG UNIVERSITY (HEALTH SCIENCES). 52(3), 40–44 (2014).
Zhu, S. M., Chen, H. Y. & Fan, Z. D. Research Progress on Dosage Forms of Bletilla Striata Gum. Chin. J. Mod. Appl. Pharm. 36, 3130–3135 (2019).
Ahmad, H. et al. Impact of pre-anthesis drought stress on physiology, yield-related traits, and drought-responsive genes in green super rice. Front. Genet. 13, 832542. https://doi.org/10.3389/fgene.2022.832542 (2022).
Zafar, S. A. et al. Genome wide analysis of heat shock transcription factor (HSF) family in chickpea and itscomparison with Arabidopsis. Plant. Omics. 9 (2), 136–141 (2016).
Hasanuzzaman, M., Nahar, K., Alam, M. M. & Roychowdhury, R. M. Fujita Physiological, biochemical, and molecular mechanisms of heat stress tolerance in plants. Int. J. Mol. Sci. 14 (5), 9643–9684 (2013).
Bakhtavar, M. A., Afzal, I., Basra, S. M. A., Ahmad, A. H., Noor, M. A. & Physiological strategies to improve the performance of spring maize (Zea mays L.) planted under early and optimum sowing conditions. PLoS ONE. 10, e0124441. https://doi.org/10.1371/journal.pone.0124441 (2015).
Elith, J. & Leathwick, J. R. Species distribution models: ecological explanation and prediction across space and time. Annu. Rev. Ecol. Evol. Syst. 40 (1), 677–697 (2009).
Guisan, A., Thuiller, W. & Zimmermann, N. E. Habitat Suitability and Distribution Models: With Applications in R (Cambridge University, 2017).
Phillips, S. J., Anderson, R. P. & Schapire, R. E. Maximum entropy modeling of species geographic distributions. Ecol. Model. 190 (3–4), 231–259 (2006).
Guisan, A., Zimmermann, N. E., Graham, C., Phillips, S. J. & Peterson, A. What matters for predicting the occurrences of trees: techniques, data, or species' characteristics? Ecol. Monogr. 77 (4), 615–630 (2007).
Qin, A. et al. Maxent modeling for predicting impacts of climate change on the potential distribution of thuja sutchuenensis franch., an extremely endangered conifer from southwestern China. Glob Ecol. Conserv. 10, 139–146 (2017).
Alatawi, A. S., Gilbert, F. & Reader, T. Modelling terrestrial reptile species richness, distributions and habitat suitability in Saudi Arabia. J. Arid Environ. 178, 104153. https://doi.org/10.1016/j.jaridenv.2020.104153 (2020).
Holder, A. M., Markarian, A., Doyle, J. M. & Olson, J. R. Predicting geographic distributions of fishes in remote stream networks using maximum entropy modeling and landscape characterizations. Ecol. Model. 433, 109231. https://doi.org/10.1016/j. ecolmodel.2020.109231 (2020).
Guo, Y. et al. Prediction of the potential geographic distribution of the ectomycorrhizal mushroom tricholoma matsutake under multiple climate change scenarios. Sci. Rep. UK. 7, 46221. https://doi.org/10.1038/srep46221 (2017).
Guo, Y., Li, X., Zhao, Z. & Nawaz, Z. Predicting the impacts of climate change, soils and vegetation types on the geographic distribution of polyporus umbellatus in China. Sci. Total Environ. 648, 1–11 (2019).
Nneji, L. M. et al. Species distribution modelling predicts habitat suitability and reduction of suitable habitat under future climatic scenario for Sclerophrys perreti: a critically endangered Nigerian endemic toad. Afr. J. Ecol. 58 (3), 481–491 (2020).
Li, R. Protecting rare and endangered species under climate change on the Qinghai Plateau, China. Ecol. Evol. 9, 427–436 (2019).
Bowen, A. K. M., Stevens, M. H. H. & Temperature topography, soil characteristics, and NDVI drive habitat preferences of a shade-tolerant invasive grass. Ecol. Evol. 10 (19), 10785–10797 (2020).
Lin, H., Gu, K., Li, W. & Zhao, Y. Integrating coalescent-based species delimitation with ecological niche modeling delimited two species within the Stewartia sinensis complex (Theaceae). J. Syst. Evol. 60 (5), 1037–1048 (2021).
Escobar, L. E. et al. Declining prevalence of disease vectors under climate change. Sci. Rep. UK. 6, 39150. https://doi.org/10.1038/srep39150 (2016).
Heikkinen, R., Marmion, M. & Luoto, M. Does the interpolation accuracy of species distribution models come at the expense of transferability? Ecography. 35(3), 276–288 (2011).
Wenger, S. & Olden, J. Assessing transferability of ecological models: an underappreciated aspect of statistical validation. Methods Ecol. Evol. 3 (2), 260–267 (2012).
Čengić, M. et al. On the importance of predictor choice, modelling technique, and number of pseudoabsences for bioclimatic envelope model performance. Ecol. Evol. 10, 12307–12317 (2020).
Radosavljevic, A. & Anderson, R. P. Making better Maxent models of species distributions: complexity, overfitting and evaluation. J. Biogeogr. 41 (4), 629–643 (2014).
Morales, N. S., Fernández, I. C. & Baca-González, V. MaxEnt’s parameter configuration and small samples: are we paying attention to recommendations?A systematic review. Peer J. 5, e3093. https://doi.org/10.7717/peerj.3093 (2017).
Freeman, B., Jiménez-García, D., Barca, B. & Grainger, M. Using remotely sensed and climate data to predict the current and potential future geographic distribution of a bird at multiple scales: the case of Agelastes meleagrides, a western African forest endemic. Avian Res. 10, 1–9 (2019).
Hu, S. P. & He, L. W. Analysis of suitable distribution areas of Fargesia denudata in Baishuijiang National Nature Reserve using MaxEnt model and ArcGIS. Chin. J. Ecol. 39 (6), 2115–2122 (2020).
Yang, Q. J. & Li, R. Predicting the potential suitable habitats of Alsophila spinulosa and their changes. Chin. J. Appl. Ecol. 32 (2), 538–548 (2021).
Che, L., Cao, B., Bai, C. K., Wang, J. J. & Zhang, L. L. Predictive distribution and habitat suitability assessment of Notholirion bulbuliferum based on MaxEnt and ArcGIS. Chin. J. Ecol. 33 (6), 1623–1628 (2014).
Liu, X. et al. Study on growth suitability for Coptis chinensis based on ecological factors analysis by Maxent and ArcGIS model. China J. Chin. Materia Med. 41, 3186–3193 (2016).
Dan, L. W. & Seifert, S. N. Ecological niche modeling in Maxent: the importance of model complexitly and the performanee of model selection criteria. Ecol. Appl. 21 (2), 335–342 (2011).
Walden-Schreiner, C., Leung, Y. F., Kuhn, T., Newburger, T. & Tsai, W. L. Environmental and managerial factors associated with pack stock distribution in high elevation meadows: case study from Yosemite National Park. J. Environ. Manage. 193, 52–63 (2017).
Zhu, G. P. & Qiao, H. J. Effect of the MaxEnt model^,s complexity on the prediction of species potential distributions. Biodivers. Sci. 24 (10), 1189–1196 (2016).
Dormann, C. F. et al. Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography. 36 (1), 27–46 (2013).
Zhang, H., Zhao, H. X. & Wang, H. Potential geographical distribution of populus euphratica in China under future climate change scenarios based on Maxent model. Acta Ecol. Sin. 40 (18), 6552–6563 (2020).
Waren, D. L. & Seilert, S. N. Ecological niche modeling in Maxent: the importanee of model complexity and the perfomance of model selection eriteria. Ecol. Appl. 21 (2), 335–342 (2011).
Ahmad, R., Khuroo, A. A., Hamid, M., Charles, B. & Rashid, I. Predicting invasion potential and niche dynamics of Parthenium hysterophorus (Congress grass) in India under projected climate change. Biodivers. Conserv. 28, 2319–2344 (2019).
Meyer, A. L. S., Pie, M. R. & Passos, F. C. Assessing the exposure of lion tamarins (Leontopithecus spp.) to future climate change. Am. joumal Primatol. 76 (6), 551–562 (2014).
Phillips, S. J. & Dudík, M. Modeling of species distributions with MAXENT: new exten sions and a comprehensive evaluation. Ecography. 31 (2), 161–175 (2008).
Amiri, M., Tarkesh, M. & Shafiezadeh, M. Modelling the biological invasion of Prosopis juliflora using geostatistical-based bioclimatic variables under climate change in arid zones of southwestern Iran. J. Arid Land. 14 (2), 203–224 (2022).
Cobos, M. E., Peterson, A., Barve, N. & Osorio-Olvera, L. Kuenm: an R package for detailed development of ecological niche models using Maxent. PeerJ. 7, e6281. https://doi.org/10.7717/peerj.6281 (2019).
Zhao, G. H. et al. Analysis of the distribution pattern of Chinese Ziziphus jujuba under climate change based on optimized biomod2 and MaxEnt models. Ecol. Indic. 132, 108256. https://doi.org/10.1016/j.ecolind.2021.108256 (2021).
Cong, M. Y., Xu, Y. Y., Tang, L. Y., Yang, W. J. & Jian, M. F. Predicting the dynamic distribution of Sphagnum bogs in China under climate change since the last interglacial period. PLoS ONE. 15 (4), e0230969. https://doi.org/10.1371/journal.pone.0230969 (2020).
Hosmer, D. W. Jr, Lemeshow, S. & Sturdivant, R. X. Applied logistic regression (Wiley, 2013).
Guillera-Arroita, G., Lahoz-Monfort, J. & Elith, J. MaxEnt is not a presence absence method: a comment on Thibaud et al. Methods Ecol. Evol. 5 (11), 1192–1197 (2014).
Commander, C. J. C., Barnett, L. A. K., Ward, E. J., Anderson, S. C. & Essington, T. E. The shadow model: how and why small choices in spatially explicit species distribution models affect predictions. Peer J. 10, e12783. https://doi.org/10.7717/peerj.12783 (2022).
Abdelaala, M., Foisa, M., Fenua, G. & Bacchettaa, G. Using MaxEnt modeling to predict the potential distribution of the endemic plant Rosa arabica Crép. Egypt. Ecol. Inf. 50, 68–75 (2019).
Fois, M., Cuena-Lombraña, A., Fenu, G. & Bacchetta, G. Using species distribution models at local scale to guide the search of poorly known species: review, methodological issues and future directions. Ecol. Model. 385, 124–132 (2018).
Kaky, E. & Gilbert, F. Assessment of the extinction risks of medicinal plants in Egypt under climate change by integrating species distribution models and IUCN Red List criteria. J. Arid Environ. 170, 103988. https://doi.org/10.1016/j.jaridenv.2019 (2019). 05.016.
Urban, M. C. et al. Improving the forecast for biodiversity under climate change. Science. 353 (6304), aad8466. https://doi.org/10.1126/science.aad8466 (2016).
Vasconcelos, T. S., Rodríguez, M. Á. & Hawkins, B. A. Species distribution modelling as a macroecological tool: a case study using New World amphibians. Ecography. 35 (6), 539–548 (2012).
Gong, Y., Jing, P. F., Wei, Y. K., Huang, W. C. & Cui, L. J. Potential Distribution of Bletilla striata (Orchidaceae) in China and Its Climate Characteristics. Plant. Divers. Resour. 36 (2), 237–244 (2014).
Appleby, R. F. & Davies, W. J. A possible evaporation site in the guard cell wall and the influence of leaf structure on the humidity response by stomata of woody plants. Oecologia. 56, 30–40 (1983).
Grossiord, C. et al. Plant responses to rising vapor pressure deficit. New. Phytol. 226 (6), 1550–1566 (2020).

No competing interests reported.

Download PDF

Reviewers agreed at journal
15 Nov, 2024
Reviewers agreed at journal
14 Nov, 2024
Reviewers invited by journal
14 Nov, 2024
Editor assigned by journal
14 Nov, 2024
Editor invited by journal
23 Oct, 2024
Submission checks completed at journal
22 Oct, 2024
First submitted to journal
07 Oct, 2024

You are reading this latest preprint version

Predicting potentially suitable Bletilla striata habitats in China under future climate change scenarios using the optimized MaxEnt model

Status:

Version 1

Abstract

Figures

Highlights

1. Introduction

2. Materials and methods

2.1. B. striata distribution data sources

2.2. Environmental data sources and screening

2.3. Maximum entropy model establishment, optimization, and evaluation

2.4. Optimized MaxEnt modeling

2.5. Potential B. striata distribution changes and centroid shift based on future CMIP6

3. Results and analysis

3.1. Model optimization and accuracy evaluation

3.2. Currently suitable B. striata habitat distribution

3.3. Major environmental factors affecting B. striata resource distribution

3.5. The change range and centroid transfer characteristics of B. striata habitats under future

4. Discussion

4.1. Model predictive performance

4.2. Potential distribution area and ecological characteristics of B. striata

4.3. Effects of climate change on future B. striata distribution

4.4. B. striata habitat selection and protection strategies

5. Conclusion

Declarations

Declaration of Competing Interest

Author Contribution

Acknowledgements

Data Availability

References

Additional Declarations

Status:

Version 1