impacts of climate change on medicinal plants in vietnam an assessment

1 THAI NGUYEN UNIVERSITY UNIVERSITY OF AGRICULTURAL AND FORESTRY VU THU TRA IMPACTS OF CLIMATE CHANGE ON MEDICINAL PLANTS IN VIETNAM: AN ASSESSMENT Study Mode: Full-time Major: Environme

INTRODUCTION

Research rationale

The variety and distribution of therapeutic plants in Vietnam are significantly impacted by climate change (Applequist et al., 2020) Many plants species' habitat appropriateness is changing due to changes in patterns of rainfall and temperature, which is changing their range and distribution As a result, many medicinal plants are becoming increasingly rare, threatened, or endangered Additionally, changes in the timing and duration of seasons can affect the flowering, fruiting, and other phenological events, leading to a mismatch between pollinators and plants, and reducing the reproductive success of plant populations Climate change can also exacerbate the stress of medicinal plants, making them more susceptible to pests, diseases, and other biotic and abiotic factors These impacts can have serious consequences on the availability and quality of medicinal plant resources, as well as the traditional knowledge and practices associated with their use Hence, it is critical to determine how vulnerable medicinal plants are to climate change, come up with practical conservation plans, and encourage sustainable usage and management techniques In this study, a total of 9 species of precious medicinal plants with many therapeutic effects were mentioned, including: (1) Coscinium fenestratum (Gaertn.) Colebr., (2) Morinda officinalis FC How, (3) Anoectochilus setaceus

Blume, (4) Stephania japonica (Thunb.) Miers, (5) Berberis julianae CKSchneid, (6) Panax vietnamensis Ha & Grushv, (7) Cupressus torulosa D Don, (8) Taxus wallichiana

Zucc and (9) Panax bipinnatifidus Seem The selection of these 9 medicinal plants for the research was based on several important factors They are known to have significant medicinal value and are commonly used in traditional medicine practices in Vietnam They represent a diverse range of plant families and exhibit a variety of medicinal properties, making them of great interest for further study On the other hand, the plants selection criteria are based on their ecological and conservation significance Since many of them are unique to Vietnam or have small geographic ranges, they are especially sensitive to the effects of climate change We can learn more about the possible impacts of climate change on botanical medicines at the species and ecosystem levels by researching these species Furthermore, the selected plants also have relevance in terms of cultural importance and economic value They are widely used in traditional medicine practices and play a vital role in local communities For the purpose of maintaining traditional knowledge and guaranteeing the ongoing availability of medicinal resources for future generations, it is crucial to comprehend the possible effects of climate change on these plants Lastly, the selected plants may have limited existing research on their response to climate change By concentrating on particular species, the study seeks to close information gaps and explain how climate change may affect their distribution

Because of such great medicinal uses, the above medicinal plants are always actively hunted, leading to a situation where demand exceeds supply, thereby pushing up their prices, and at the same time pushing them into the endangered group threatened in the IUCN Red List Besides, they also face the risk of changing land use area Land is exploited and used by people for many purposes such as: land for agricultural, forestry and fishery production, residential land, land for construction of infrastructure, land for production, business and services (Nguyen Van Hieu, 2018) Climate change has also greatly affected tree species (Applequist et al., 2020), specifically: increased temperature affects the ability of plants to grow, leading to changes in yield and output; at the same time, increasing temperature causes water resources to decrease, leading to a decrease in cultivated area; melting ice causes many lands to be encroached on and flooded; changes in climatic conditions will cause biodiversity loss and ecological imbalance; In addition, there will be an increase in extreme, irregular weather events (IPCC, 2007; Stern, 2009) The aforementioned factors have had a significant impact on the species' distribution area

For the purpose of creating efficient conservation and management plans to guarantee the long-term survival of these significant species, assessing the effects of climate change on medicinal plants is essential Thus, this thesis uses the MaxEnt program, a species distribution model (SDM), to evaluate the effects of climate change on a variety of medicinal plants in Vietnam (Truong et al., 2017) This is an effective method for forecasting the distribution of medicinal plants under various climatic conditions The distribution of these medicinal plants under both present and future climatic scenarios will be modeled using a variety of environmental factors, including temperature and precipitation The findings of this study will aid in identifying the species at risk to climate change and guide the creation of conservation plans to save these significant species of medicinal plants

This study is crucial to clarifying the possible effects of climate change on the number and distribution of medicinal plants in Vietnam as well as for creating methods that mitigate these effects.

Research’s objectives

The research project intends to assess the present and future effects of climate change on a variety of important medicinal plants in Vietnam To estimate the potential distribution of medicinal plants under the existing and predicted climatic conditions, the Species Distribution Model (SDM) will be utilized.

Research questions and hypotheses

1) Research questions a What is the current distribution of 9 medicinal plant species in Vietnam? b What will be the effects of climate change on 9 medicinal plants species’ suitable habitat in Vietnam from 2050 to 2080?

2) Hypotheses a The current distribution of the 9 medicinal plant species in Vietnam is influenced by various environmental factors, specifically temperature and precipitation b The effects of climate change on the suitable habitat of the 9 medicinal plant species in Vietnam from 2050 to 2080 will result in significant shifts and changes in their distribution patterns These changes may include range expansions, contractions, or shifts to higher elevations or latitudes.

Limitations

MaxEnt (Maximun Entropy) modelling is a tool specialized for modeling species distribution, including medicinal plants While this software can be useful for assessing the probable impact of climate change on therapeutic plants, it is crucial to be aware of its limitations

MaxEnt software is reliant on the availability and quality of input data This means that the accuracy of the predictions generated by the software relies on the accuracy of the data that is fed into it If the data is incomplete or inaccurate, the predictions generated by the software may be unreliable or misleading

MaxEnt software assumes that species distribution is driven by environmental variables While this assumption may be true in some cases, it does not take into account other important factors that can influence the distribution of medicinal plants For instance, human activities such as logging and deforestation can impact the distribution of medicinal plants, and the software may not be able to model the effects of these activities

Additionally, the program also makes the assumption that species and their environments are in balance Yet, this is often not the case in reality, particularly in the case of invasive species or changing environments due to climate change This means that the predictions generated by the software may not reflect the current or future distribution of medicinal plants

Finally, it is important to note that the outputs generated by MaxEnt software should be interpreted with caution, as they represent potential distributions rather than actual distributions This means that the software can be a useful tool for identifying areas where proposing threats to medicinal plants due to climate change, but it should not be the only tool used in evaluating the effects of climate change on therapeutical plants

In conclusion, while MaxEnt software can be a valuable tool for assessing the potential effects of climate change on medicinal plants, it is vital to acknowledge its limitations and use it in conjunction with other methods to ensure that the outcomes are reliable and accurate.

Definitions

In the context of this thesis, it is necessary to establish a firm grasp of key terms and concepts linked to the analysis of climate change impacts on medicinal plants in Vietnam

(1) MaxEnt: The maximum entropy rule is used by this program's machine learning algorithm to forecast the possible distribution of species depending on a variety of environmental factors The method initially calculates the probability density of each environmental variable in the regions where the species is known to appear using a collection of georeferenced occurrence data for the species The program then calculates the probability density of each environmental variable over the whole research region using a set of background points that are typical of the study area The computer then integrates these probability densities to produce a model that, depending on environmental factors, calculates the species' likelihood of occurring at any location within the research region

By reducing the discrepancy between the estimated and recorded probability densities for the species occurrence data, the model is optimized The generated model may be used to forecast the species' potential range under various environmental conditions Because it can handle complicated interactions between environmental factors and provide predictions for uncommon or poorly sampled species based on little occurrence data, MaxEnt is an effective tool for developing SDMs (Phillips et al., 2006)

(2) Species Distribution Models (SDMs): Based on known occurrences or presence data, these models are used to forecast the possible distribution of a species in a specific habitat or geographical area SDMs connect environmental factors, such as temperature and precipitation, to the presence of a species using statistical and mathematical methodologies MaxEnt software is an example of an SDM (Phillips et al., 2006)

(3) Ecological Niche Model (ENM): This kind of SDM simulates the environmental conditions necessary for a species to thrive and procreate The ENM is based on the idea that each species has a specific set of environmental conditions in which it can thrive, and by modeling these conditions, we can forecast the distribution of the species MaxEnt is an example of an ENM (Phillips et al., 2006)

(4) Area Under the Curve (AUC): This is a measurement of the accuracy of a model that predicts the occurrence or absence of a species in a specific environment AUC varies between 0 and 1, with higher numbers suggesting more accuracy A measurement of 0.5 indicates a model that is no better than random chance (Phillips, 2005)

(5) Representative Concentration Pathway (RCP): This is a set of scenarios used by climate scientists to project future greenhouse gas concentrations and resulting climate change The RCPs are named based on their predicted radiative force (the difference between incoming and outgoing radiation in the atmosphere) in the year 2100 (Van et al., 2011) Specifically, RCP2 represents a lower emission scenario, indicating relatively moderate greenhouse gas emissions and a more gradual increase in global temperatures It is based on the idea that substantial efforts will be undertaken to cut greenhouse gas emissions and slow down climate change On the other hand, RCP8 represents a higher emission scenario, implying a continuation of high greenhouse gas emissions throughout the 21 st century It projects more significant increases in global temperatures and more severe climate impacts This scenario assumes limited efforts to reduce emissions and a higher dependency on fossil fuels

(6) Receiver Operating Characteristic (ROC): This is a curve that plots the true positive rate (i.e., the proportion of actual species occurrences correctly predicted by the model) against the false positive rate (i.e., the proportion of non-occurrences mistakenly predicted as occurrences) at various threshold values AUC is a measure of the accuracy of the model in predicting species occurrences or absences (Phillips, 2005)

(7) Jackknife of regularized training gain: This statistical technique is employed by MaxEnt to assess the effectiveness and prognostication of species distribution models With this approach, each data point is methodically eliminated one at a time, and the model is then re-run to see how much each item contributes to the model's overall performance The jackknife process is used to choose potentially important data points and to quantify the variability of the model's performance The regularized training gain measures how well the model matches the training data The jackknife of regularized training gain, which combines these two methods, offers a thorough evaluation of the model's predictive ability and can aid in locating regions of ambiguity or potential sources of bias (Phillips et al., 2006)

(8) Marginal response curves: This is used to show how the spread of a species and environmental factors are related Marginal response curves demonstrate how, while maintaining all other factors constant, a species' likelihood of occurrence changes when the value of a single environmental variable changes The functional link between the species and the environment may be visually explored using marginal response curves, which plot the expected likelihood of occurrence versus the value of an environmental variable This can be helpful for comprehending a species' ecological needs, figuring out the environmental constraints on its range, and forecasting how future environmental changes would influence its habitat (Phillips et al., 2006)

(1) Taxonomy: Taxonomy is a scientific discipline focused on the classification, nomenclature, and identification of organisms It involves organizing species into a hierarchical system that reflects their evolutionary relationships and helps in understanding the diversity of life

(2) Phenological Events: Phenological events are recurring life cycle stages in plants and animals that are influenced by seasonal changes These events encompass activities such as flowering, fruiting, migration, and hibernation, providing insights into the biological responses to environmental cues

(3) Climatic Environmental Variables: Climatic environmental variables are quantifiable aspects of the climate system, encompassing factors like temperature, precipitation, humidity, and solar radiation They play a pivotal role in shaping the ecological dynamics and distributions of species

(4) Species Occurrences: Species occurrences refer to the documented presence of particular species within specific geographic areas This information aids in mapping the distribution and understanding the geographic range of various organisms

(5) Habitat Suitability: Habitat suitability signifies the degree to which a specific environment provides the requisite conditions for a particular species to thrive and reproduce It considers various ecological factors, including food availability, shelter, and climate suitability

LITERATURE REVIEW

Climate Change

Globally acknowledged issue, climate change has recently attracted more attention It deals with long-term changes in temperature patterns, amounts of precipitation, and general weather patterns that are predominantly attributable to human activity According to the scientific community, greenhouse gas emissions from human activity are the main cause of climate change (Change, IPOC 2007)

The quantities of greenhouse gases in the atmosphere have dramatically grown as a result of human activity, particularly the combustion of fossil fuels like natural gas, petroleum, and coal (Pimm et al., 1995) The greenhouse effect is a phenomena wherein certain gases, such as carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O), retain heat inside the Earth's atmosphere and cause an increase in global temperatures According to the Intergovernmental Panel on Climate Change, the concentrations of these greenhouse gases have risen to previously unheard-of heights since the beginning of the industrial era (Assessment ME, 2005) In addition, industrial operations, land use changes, and deforestation all contribute to greenhouse gas emissions (Foley et al., 2005)

Climate change has far-reaching consequences for natural systems and ecosystems It affects various aspects of the environment, including temperature patterns, precipitation regimes, sea levels, and extreme weather events (Change, IPOC 2007) Sea levels are rising as a result of melting glaciers and polar ice caps brought on by an increase in global temperatures Due to increasing hazards of floods, erosion, and salinization of freshwater supplies, coastal locations are particularly susceptible to this issue Changes in weather patterns have also been observed, leading to alterations in rainfall distribution and intensification of droughts in some regions These changes can have significant implications for agriculture, water resources, and food security (Stern et al., 2007)

Moreover, altered climatic conditions can impact biodiversity, leading to species range shifts, changes in phenology, and increased extinction risks (Thomas et al., 2004)

Combating climate change necessitates the use of both mitigation and adaptation strategies By switching to renewable energy sources, improving the conservation of energy, and enacting sustainable land-use techniques, mitigation aims to lower greenhouse gas emissions Global warming is intended to be kept far below 2 degrees Celsius above pre-industrial levels, and efforts are being made to keep the temperature rise to 1.5 degrees Celsius (UNFCCC, 2015) Building resilience and readiness to deal with the consequences of climate change that are already being felt are the main goals of adaptation methods This includes measures such as improving disaster management systems, implementing sustainable agricultural practices, and enhancing infrastructure resilience

There are several issues associated with the complicated and diverse problem of climate change to our planet's ecosystems and human societies The scientific literature emphasizes how urgent it is to combat climate change using both adaptation and mitigation measures To reduce greenhouse gas emissions, safeguard vulnerable people, and conserve the planet's ecosystems for future generations, governments, scientists, and society as a whole must cooperate.

Species Occurrences

Understanding the factors influencing species occurrence is fundamental to ecology and conservation biology Numerous elements, including abiotic and biotic variables, have an impact on the existence of species Abiotic factors, such as climate, topography, and soil characteristics, play a crucial role in determining species distributions (Phillips, 2008; Pearson, 2010; Figueiredo et al., 2018) Climatic factors, including temperature, precipitation, and seasonality, have been widely recognized as key drivers of species occurrence patterns These variables shape the availability of resources, influence physiological constraints, and impact species' adaptations to specific environments (Phillips, 2008) Biotic factors, including interactions with other species, competition, predation, and mutualism, also influence species occurrence (Anderson, 2017) The presence or absence of specific species in an environment is significantly influenced by species interactions and ecological linkages For instance, competition for resources can limit the occurrence of certain species, while mutualistic interactions may facilitate coexistence (Palmer et al., 2003)

Researchers employ various methods and approaches to assess species occurrence patterns Field surveys, including presence-absence sampling and occupancy modelling, are commonly used to determine species presence or absence in specific locations (MacKenzie et al., 2004) Remote sensing and satellite imagery are valuable tools for mapping species distributions over larger spatial scales, providing valuable insights into habitat suitability and land cover changes SDMs have gained prominence in recent years for predicting species occurrence and potential habitat suitability In order to forecast species distributions based on observed occurrences and environmental factors, SDMs use environmental variables such climatic data, land cover, and terrain (Phillips et al., 2006) These models enable the assessment of prospective effects of climate change and habitat loss on species occurrence and can offer insightful information about present and future species distributions

Effective biodiversity conservation and management depend on an understanding of the variables that affect species occurrence The evaluation of species occurrence patterns is made easier by a variety of techniques and strategies, including field surveys, remote sensing, and species distribution modeling Further research is needed to delve into the complex interactions and mechanisms that drive species occurrence, especially in the context of changing environmental conditions and human-induced disturbances By integrating multidisciplinary approaches and considering the cumulative effects of multiple factors, we can improve our understanding of species occurrence patterns and implement more targeted conservation strategies to protect and sustainably manage biodiversity.

Research regarding Climate change impacts on species Occurrences

Several studies have demonstrated how species ranges and occurrences are affected by climate change Warming temperatures and altered precipitation patterns have been linked to fluctuations in species distribution and phenology For instance, the observation of increasing global temperatures indicates a notable influence on plant phenology, particularly in terms of altering the temporal patterns of flowering events (Wolf et al., 2017) Revamping in temperature and rainfall can also influence the availability of suitable habitats for different species Research has shown that certain plant species are experiencing range contractions or expansions in response to changing climate conditions (Truong et al., 2017) These changes have the potential to reshape community structures and impact overall ecosystem functioning

Researchers have utilized various methods and approaches to assess climate change impacts on species occurrences SDM is commonly employed to predict possible shifts in species distributions based on climate projections (Forster et al., 2013) These models incorporate climate data, species occurrence records, and environmental variables to estimate changes in species ranges and habitat suitability Understanding the effects of global warming on species occurrences has benefited significantly by research studies using SDMs, particularly those using the Maxent program In order to create prediction models that evaluate prospective changes in species distributions and habitat suitability, this research have used climatic data, species occurrence records, and environmental factors Through the application of Maxent, researchers have gained insights into the changing ranges of various taxa, including plants, animals, and marine organisms, as well as their responses to altered climatic conditions These investigations have improved our knowledge of how shifting climatic patterns affect species distributions, which, in the context of climate change, can help direct conservation and management measures

In conclusion, changing climatic conditions have the potential to influence species distributions, community dynamics, and ecological interactions Species distribution modelling has provided valuable insights into these impacts and the mechanisms underlying species responses To fully comprehend the intricacies of species reactions to climate change, including interactions with other environmental change agents and potential thresholds or tipping points, further study is required Such information is essential for efficient conservation and management methods that work to protect biodiversity worldwide and lessen the negative effects of global warming on species occurrences.

METHODS

Study Specie

Plant species are selected based on 3 criteria:

1 It is a rare and valuable medicinal plant species

2 Being endangered or about to be threatened in Vietnam, according to (1) Government decree No 06/2019/NĐ-CP on Management of endangered, precious and rare species of forest fauna and flora and Observation of convention on international trade in endangered species of wild fauna and flora and (2) Vietnam Red List

3 To ensure the reliability of the model, the assessed species must have a record of the occurrence of at least 10 sites in the world

Table 2.1 provides a comprehensive compilation of important information about the selected plant species that were included in the modeling process This serves as a valuable reference for understanding the characteristics and attributes of each species within the context of this study The table includes several columns to facilitate a comprehensive overview of each species

• Scientific Name: the scientific or botanical name of the plant species, ensuring accuracy and precision in identification

• Familia: the taxonomic family to which each species belongs, providing insights into its classification and relationships with other plants

• Life Form: the overall pattern of development or existence of a species of plant, such as a tree, shrub, herb, or vine This knowledge aids comprehension of the ecological and physical aspects of the species

• Medical Uses: the various medicinal applications and traditional uses associated with each plant species This information underscores the importance of these plants in traditional medicine practices and their potential significance in healthcare and well-being

• Distribution: insights into the native or natural distribution range of each species, indicating the geographic areas where they are commonly found This information is crucial for understanding the baseline distribution patterns and potential shifts in suitable habitats under changing climatic conditions

• Conservation Status in Vietnam Red List: the evaluation made by the Vietnam Red List regarding the conservation status of each plants This classification reflects the degree of vulnerability or threat faced by the species within the context of the country's conservation efforts

Table 3.1 List of medicinal plant species used in the model

No Science name Familia Life form Medical uses Distribution

Fevers, Diabetes, Celiac disease, Snake bites

Gia Lai, Kon Tum, Lam Dong and southern provinces

Kidney deficiency; Weak tendons and bones

Cao Bang, Bac Ninh, Son La, Lam Dong

Blood circulation, antibacterial Cure bronchitis, hepatitis, Improve health, Reduce nervous breakdown

Cao Bang, Yen Bai, Hoa Binh, Lao Cai, Tuyen Quang, Lai Chau

Vertigo, headache, and sleep disturbances

Ninh Binh, Hoa Binh, Mekong Delta

Antipyretic, liver disease and other digestive ailments, e.g

Sapa - Lao Cai Endangered digestive disorders or lack of appetite

6 Panax vietnamensis Araliaceae Perennial Anti-fatigue and life saving

Kon Tum, Quang Nam, Lam Dong Endangered

7 Cupressus torulosa Cupressaceae Evergreen conifer

Stomach pain, diabetes, inflammation, toothache, laryngitis and as contraceptive

Cao Bang, Tuyen Quang, Lang Son

8 Taxus wallichiana Taxaceae Evergreen conifer

Common cold, cough, fever, and pain

Anaemia, general debility, aphrodisiac, sterility treatment

Species occurrences

One of the crucial inputs to the model for predicting species distribution is data on species occurrence that has been observed The website for the Global Biodiversity Information Base (http://www.gbif.org/) is where these emerging data (Table 3.1) were found The observatory for CSP is called the Global Biodiversity Information Facility (GBIF) Governments support this worldwide platform for open access to data It enables everyone to have access to knowledge provided over the world wide web regarding the majority of living forms on Earth The duplicate data is sorted out in Excel once the data has been downloaded.

Environmental variables

Modeling regions suited for nine medicinal plant species using a mix of climate and climate forecasting technologies Eleven temperature measures and eight metrics for precipitation were obtained from the WorldClim database (Table 2.2) (Hijmans et al., 2005) for the time period 1950–2000 (http://www.worldclim.org) Additionally, certain WorldClim data on climate are strongly connected and may result in precise forecasts

(Hijmans et al., 2005; Mbatudde et al., 2012) To regulate the connections between all of the WorldClim climate data for the Southeast Asia area, the Pearson correlation was performed To increase the effectiveness and precision of the prediction model, a variable with a significant correlation was eliminated

Table 3.2 The list of climatic environmental variables used in the model

Bio 2 Mean Diurnal Range (Mean of monthly

Bio 4 Temperature Seasonality (standard deviation ×100) °C * 100

Bio 5 Max Temperature of Warmest Month °C * 10

Bio 6 Min Temperature of Coldest Month °C * 10

Bio 7 Temperature Annual Range (BIO5-

Bio 8 Mean Temperature of Wettest Quarter °C * 10

Bio 9 Mean Temperature of Driest Quarter °C * 10

Bio 10 Mean Temperature of Warmest Quarter °C * 10

Bio 11 Mean Temperature of Coldest Quarter °C * 10

Bio 12 Annual Precipitation mm (millimeter)

Bio 13 Precipitation of Wettest Month mm (millimeter)

Bio 14 Precipitation of Driest Month mm (millimeter)

Bio 15 Precipitation Seasonality (Coefficient of

Bio 16 Precipitation of Wettest Quarter mm (millimeter)

Bio 17 Precipitation of Driest Quarter mm (millimeter)

Bio 18 Precipitation of Warmest Quarter mm (millimeter)

Bio 19 Precipitation of Coldest Quarter mm (millimeter)

Future climate Scenarios

For this investigation, Hadley Model Matching Center, version 3 (HADCM3) climate forecasts were employed According to Jaramillo et al (2011), when compared to other models, this model is said to yield good average results, and it is frequently employed in ecological research (IPCC, 2007) The study's future projections were created for the years

2050 to 2080 The CGIAR Research Program on Climate Change, Agriculture, and Food Security provides future statistics (http://www.ccafs-climate.org/downcaling_delta/).

Species Distribution Modelling

MaxEnt software (version 3.4.1) was utilized to predict the accuracy to the habitat (Phillips et al., 2006) MaxEnt is one of the most widely used species distribution modeling tools due to its accuracy in prediction and simplicity of usage (Elith et al., 2006; Phillips and Dudk, 2008) Low input requirements for species data, flexibility in handling environmental data, including continuous variables and categorical variables, and the capacity to fit complex responses to environmental variables are just a few of the characteristics that make MaxEnt well suited for modeling the distribution of species (Phillips et al., 2006) MaxEnt is a recommended prediction model since it is notable that it is less susceptible to small sample sizes (Wisz et al., 2008)

Running the model multiple times, often referred to as replicate runs, is a common practice in species distribution modeling to account for the inherent uncertainty and variability in the modeling process Each replicate run involves slightly different randomization or partitioning of the data, leading to variations in model outputs

There are several reasons why replicate run is beneficial:

• Robustness assessment: By running the model multiple times, you can assess the stability and robustness of the results If the model consistently produces similar outputs across the replicate runs, it suggests that the findings are reliable and not heavily influenced by random variations

• Uncertainty estimation: The replicate runs allow for the estimation of uncertainty associated with model outputs The variations among the replicate runs provide an indication of the range of potential outcomes, helping to quantify the uncertainty surrounding the model predictions

The standard deviation is a statistical measure that quantifies the amount of variation or dispersion in a dataset In the context of the AUC result from replicate runs, the standard deviation of the AUC values provides an estimate of the variability in model performance across the different runs

A low standard deviation suggests that the model outputs are relatively consistent, indicating high stability and confidence in the results On the other hand, a high standard deviation indicates more variability among the replicate runs, suggesting greater uncertainty in the model predictions

By considering both the mean AUC value and the standard deviation, you can assess the average model performance as well as the variability around that performance It helps in understanding the reliability and robustness of the model and provides insights into the uncertainty associated with the predicted habitat suitability for the studied species.

Model validation

AUC is a widely used statistic for evaluating the effectiveness and prognostication precision of SDMs The ROC curve, a visual depiction of the correlation between true positive rate and false positive rate, is the source from which AUC is produced

AUC has a value range of 0 to 1, with higher numbers signifying greater model performance An AUC value of 1 denotes a perfect model with 100% discrimination, whereas an AUC value of 0.5 denotes a model that performs no better than random chance

To assess models based on AUC, certain criteria are often applied These criteria can vary depending on the specific study or context, but some common thresholds are:

• AUC < 0.1: This suggests a model with poor performance, worse than random chance It indicates that the model's predictions are in the opposite direction of the actual occurrences

• 0.1 ≤ AUC < 0.4: This range suggests a model with low discrimination and relatively weak predictive power The model's performance is considered to be marginally better than random chance

• 0.4 ≤ AUC < 0.6: This range indicates a model with moderate to high discrimination The model's performance is considered acceptable, with a reasonable ability to distinguish between presence and absence points

• AUC ≥ 0.6: This suggests a model with excellent discrimination and high predictive power The model's performance is considered very good, indicating a strong ability to accurately predict species distribution

By using AUC, researchers can rigorously evaluate the performance of SDMs, assess their predictive accuracy, and determine the reliability of the models in capturing the distribution patterns of medicinal flora species under current and future climatic conditions

MaxEnt's logistic output is a map that indexes environmental appropriateness using values from 0 (inappropriate) to 1 (optimal) In order to conduct additional analysis, the MaxEnt findings were imported into ArcGIS 10.4.1 and divided into four categories: unsuitable (0.10), low (0.11-0.40), medium (0.41-0.60), and high potential (0.61) environments Finally, MaxEnt's predicted maps for the present and the future were connected to the various elevation classes.

RESULTS

Model performance

Table 4.1 and figure 4.1 show the exact AUC for each species’ distribution model Overall, all the AUCs are higher than 0.9 The values indicate that the models are in

“excellent” conditions (Phillips et al., 2006) and indicate that the models are very reliable and can reflect their circulation under current and future climate scenarios The standard deviations of the AUCs are less than 0.06, suggest that there are relatively low variability or dispersion in the model performance across the replicate runs This indicates a high level of consistency and stability in the model outputs

Species Occurrences in the world

Average test AUC for the replicate runs

Figure 4.1 Area under the ROC curve (AUC) of 9 species

A - Coscinium fenestratum; B - Morinda officinalis; C - Anoectochilus setaceus; D - Stephania japonica; E - Berberis julianae; F - Panax vietnamensis; G - Cupressus torulosa; H - Taxus wallichiana; I - Panax bipinnatifidus

Importance of environment variables

After running each models, the results of the top 3 highest contributing environment factors for each species are shown in Table 4.2 and Figure 4.2

1 Coscinium fenestratum: The distribution of Coscinium fenestratum is significantly influenced by Bio 4 (40.2%) and Bio 12 (25.4%), indicating the importance of temperature and precipitation-related factors

2 Morinda officinalis: The occurrence of Morinda officinalis is notably influenced by Bio 2 (26.2%) and Bio 18 (18.6%), which highlight the significance of temperature and precipitation patterns in shaping its distribution

3 Anoectochilus setaceus: Although there is no specific contribution rate mentioned in the table, other environmental variables may have an impact on the distribution of

4 Stephania japonica: The distribution of Stephania japonica is predominantly influenced by Bio 16 (32.4%) and Bio 18 (22.2%), indicating the significance of temperature and precipitation-related factors

5 Berberis julianae: Berberis julianae exhibits a strong dependence on Bio 14

(35.1%), emphasizing the importance of temperature-related factors in its distribution

6 Panax vietnamensis: Panax vietnamensis shows a significant dependence on Bio

18 (58.3%), highlighting the crucial role of precipitation-related factors in shaping its distribution

7 Cupressus torulosa: The occurrence of Cupressus torulosa is strongly influenced by Bio 7 (12.5%), indicating the importance of temperature-related variables

8 Taxus wallichiana: The distribution of Taxus wallichiana is notably affected by Bio 18 (56.6%) and Bio 11 (16%), indicating the critical role of precipitation and temperature variables in its occurrence

9 Panax bipinnatifidus: Panax bipinnatifidus exhibits a substantial reliance on Bio

18 (39%), emphasizing the significance of precipitation-related factors in its distribution.

Figure 4.2 Jackknife of regularized training gain of 9 species

A - Coscinium fenestratum; B - Morinda officinalis; C - Anoectochilus setaceus; D - Stephania japonica; E - Berberis julianae; F - Panax vietnamensis; G - Cupressus torulosa; H - Taxus wallichiana; I - Panax bipinnatifidus

Table 4.2 Contribution rates of environmental variables in current climate condition

Current habitat suitability

The response curves (Figure 4.3.1.a) demonstrate that: The most suitable areas (0.9– 1) for the distribution of C fenestratum are places with Seasonal Temperatures (Bio 4) below 0 o C, Precipitation of Wettest Month (Bio 13) is over 600 mm and the coldest quarter (Bio 19) is over 1500 mm The ENM predicts that under current environmental conditions, the highly suitable habitat area for C fenestratum is 431.24 km 2 , including the Southern Mekong Delta provinces of Vietnam (Figure 4.3.1.b)

Figure 4.3.1.a Marginal response curves of Coscinium fenestratum under current climate conditions

Figure 4.3.1.b Predicted current distribution map of Coscinium fenestratum

The response curves (Figure 4.3.2.a) demonstrate that M officinalis is most suitable in habitats with a Mean Day Range (Bio 2) of about 2-9 o C, and Mean Temperature of Warmest Quarter (Bio 10) is 25-40 o C and Annual Precipitation (Bio 12) is over 1000 mm The ENM predicts that under current environmental conditions, the highly suitable habitat for M officinalis is 46471.12 km 2 , including the Northeast and Red River Delta regions of the North Central provinces of Vietnam (Figure 4.3.2.b)

Figure 4.3.2.a Marginal response curve of Morinda officinalis under current climate conditions

Figure 4.3.2.b Predicted current distribution map of Morinda officinalis

The response curves (Figure 4.3.3.a.) demonstrate that A setaceus is most suitable in habitats with Annual Precipitation (Bio 12) above 2000mm, Rainiest Month (Bio 13) over 400mm, humid quarter wettest (Bio 16) over 1000 mm, warmest quarter (Bio 18) over 500 mm The ENM predicts that under current environmental conditions, the most likely habitat area for A setaceus is 112374.40 km 2 , including the Central Coastal Plain, Tay Nguyen Plateau, Mekong Delta and some Northern provinces of Vietnam (Figure 4.3.3.b)

Figure 4.3.3.a Marginal response curves of Anoectochilus setaceus under current climate conditions

Figure 4.3.3.b Predicted current distribution map of Anoectochilus setaceus

The response curves (Figure 4.3.4.a) demonstrate that S japonica is most suitable in habitats with moderate rainfall such as Precipitation of Wettest Month (Bio 13) above 300 mm, Precipitation of Wettest Quarter (Bio 16) above 800 mm and Precipitation of Warmest Quarter (Bio 18) above 600 mm The ENM predicts that under current environmental conditions, the highly suitable habitat area for S japonica is 203195.66 km 2 , including the North, North Central, Mekong Plateau as well as Mekong Delta regions of Vietnam (Figure 4.3.4.b)

Figure 4.3.4.a Marginal response curves of Stephania japonica under current climate conditions

Figure 4.3.4.b Predicted current distribution map of Stephania japonica

The response curves (Figure 4.3.5.a) demonstrate that B julianae is most suitable in thermal habitats with the Temperature Annual Range (Bio 7) is below 3 o C, Mean Temperature of Wettest Quarter (Bio) 8) is above 10 o C, Precipitation Seasonality (Bio 15) is less than 45 mm, Precipitation of Driest Month (Bio 14) is over 30 mm The ENM predicts that under current environmental conditions, the low-suitability habitat area for B julianae is 6088.68 km 2 , including the Hoang Lien Son mountain range (Figure 4.3.5.b)

Figure 4.3.5.a Marginal response curves of Berberis julianae under current climate conditions

Figure 4.3.5.b Predicted current distribution map of Berberis julianae

Please note that there are yellow areas in the Northern mountainous provinces

The response curves (Figure 4.3.6.a) show that the most suitable areas (0.9–1) for the distribution of P vietnamensis are those with high Annual Precipitation (Bio 12) (above

5000 mm), Precipitation of Wettest Month (Bio 13) (above 750 mm), Precipitation of Wettest Quarter (Bio 16) (above 1800 mm) and Precipitation of Warmest Quarter (Bio 18) (above 1500 mm) The ENM predicts that under current environmental conditions, the highly suitable habitat area for P vietnamensis is 262275.38 km 2 , including the North, North Central as well as Mekong Delta provinces of Vietnam (Figure 4.3.6.b)

Figure 4.3.6.a Marginal response curves of Panax vietnamensis under current climate conditions

Figure 4.3.6.b Predicted current distribution map of Panax vietnamensis

The response curves (Figure 4.3.7.a) show that the most suitable regions (0.9–1) for the distribution of C torulosa are those where the Mean Annual Temperature (Bio 1) is above 20 o C, Precipitation Seasonality (Bio 15) above 100 mm, Temperature Annual Range (Bio 7) below 10 o C The ENM predicts that under current environmental conditions, the highly suitable habitat area for C torulosa is 38975.78 km 2 , including the mountainous Northwest regions of Vietnam (Figure 4.3.7.b)

Figure 4.3.7.a Marginal response curves of Cupressus torulosa at present climatic conditions

Figure 4.3.7.b Predicted current distribution map of Cupressus torulosa

The response curves (Figure 4.3.8.a) show that the most suitable areas (0.9–1) for the distribution of T wallichiana are those with Annual Precipitation (Bio 12) above 3200 mm and Temperature Seasonality (Bio 4) below 0 o C The ENM predicts that under current environmental conditions, the highly suitable habitat area for T wallichiana is 12,177.36 km 2 , including the Hoang Lien Son mountain range and some northern provinces bordering China of Vietnam (Figure 4.3.8.b)

Figure 4.3.8.a Marginal response curves of Taxus wallichiana at present climate conditions

Figure 4.3.8.b Predicted current distribution map of Taxus wallichiana

The response curves (Figure 4.3.9.a) show that the most suitable areas (0.9–1) for the distribution of P bipinnatifidus are those with Annual Precipitation (Bio 12) above 1600 mm, Precipitation of Warmest Quarter (Bio 18) above 1700 mm and the Max Temperature of Warmest Month (Bio 5) below 10 o C The ENM predicts that under current environmental conditions, the highly suitable habitat area for P bipinnatifidus is 8450.23 km 2 , including the Hoang Lien Son mountain range and some northern provinces bordering China of Vietnam (Figure 4.3.9.b)

Figure 4.3.9.a Marginal response curves of Panax bipinnatifidus under current climate conditions

Figure 4.3.9.b Predicted current distribution map of Panax bipinnatifidus

Future potential distribution

A warming climate with the fewest ozone-depleting substances will cause a progressively higher range of suitable habitats south of Vietnam (Figure 4.4.1.b, RCP 2

2050, RCP 2 2080), from 431.24 km 2 (present) to almost nothing(RCP 2 2050) and then to 4424.15 km2(RCP 2 2080) (Figure 4.4.1.a) In the RCP 8 scenario, the geological range of highly suitable habitats will gradually increase to 5147.73 km 2 by 2050 and decrease to 3700.58 km 2 by 2080 (Figure 4.4.1.a) The coverage distribution maps have demonstrated that the area of high suitability habitat shifts southward (Figure 4.4.1.b, RCP 8 2050, RCP

U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h

Present RCP 2 2050 RCP 2 2080 RCP 8 2050 RCP 8 2080

Figure 4.4.1.a The forecasted suitable area for Coscinium fenestratum in Vietnam under different climate scenarios

Figure 4.4.1.b Habitat map with appropriate levels - Coscinium fenestratum in

Vietnam under current and future climate conditions

A – Present; B – RCP 2 2050; C - RCP 2 2080; D – RCP 8 2050; E – RCP 8 2080

According to Table 4.4.1, Temperature Seasonality (Bio 4) dropped significantly from 40.2% in current time to 5.1%-17.1% in future Temperature Annual Range (Bio 7) surpassed Annual Precipitation (Bio 12) and took the highest position, contributing a range between 46.5% and 54.4% In summary, temperature-related variables add up more to the distribution of C fenestratum in the future, with the mean value of 72.6%

Table 4.4.1 Contribution rates of environmental variables - Coscinium fenestratum

Temperature Seasonality (Bio 4) 9.9 11.6 5.1 17.1 Temperature Annual Range (Bio 7) 46.6 52.8 54.4 46.5 Annual Precipitation (Bio 12) 25.3 16.3 23 25.8

A warming climate with the fewest ozone-depleting substances will promote a gradual elevation of suitable habitat towards western Vietnam (Figure 4.4.2.b, RCP 2 2050, RCP 2 2080), from 46471.12 km 2 (present) to 58506.33 km 2 (RCP 2 2050) and then 72957.19 km 2 (RCP 2 2080) (Figure 4.4.2.a) In a warming climate with the greatest amount of ozone-depleting substances, the geological range of highly suitable habitats will gradually increase to 84637.78 km 2 by 2050 and decrease to 84193.30 km 2 by 2080 (Figure 4.4.2.a) The coverage distribution maps demonstrated that the area of the highly suitable habitat shifted towards northwest (Figure 4.4.2.b RCP 8 2050, RCP 8 2080)

Figure 4.4.2.a The forecasted suitable area for Morinda officinalis in Vietnam under different climate scenarios

U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h

Present RCP 2 2050 RCP 2 2080 RCP 8 2050 RCP 8 2080

Figure 4.4.2.b Habitat map with appropriate levels - Morinda officinalis in Vietnam under current and future climate conditions

A – Present; B – RCP 2 2050; C - RCP 2 2080; D – RCP 8 2050; E – RCP 8 2080

According to Table 4.4.2, Mean Diurnal Range (Bio 2) still remain the top position, with 26.2% in current time and 20.1%-26.4% in future Precipitation of Warmest Quarter (Bio 18) and Mean Temperature of Warmest Quarter (Bio 10) surpassed Bio 3 and Bio 6 for the second and third position, respectively However, it is important to note that Bio 18 only perform well in RCP2 scenarios In summary, temperature-related variables add up more to the distribution of M officinalis in the future, with the mean value of 69.825%

Table 4.4.2 Contribution rate of environmental variables - Morinda officinalis

Min Temperature of Coldest Month (Bio 6) 7.4 6.4 5.1 0.1

Mean Temperature of Warmest Quarter (Bio 10) 17.1 17.6 15.1 6.1

Precipitation of Driest Month (Bio 14) 6.1 2.2 4.1 3.2

Precipitation of Warmest Quarter (Bio 18) 22.8 11.2 2.7 5.5

A warming climate with the fewest ozone-depleting substances will gradually reduce highly suitable habitats (Figure 4.4.3.b RCP 2 2050, RCP 2 2080), from 112374.40 km 2 (current) to 73804.81 km 2 (RCP2 2050) and then 72150.92 km 2 (RCP 2 2080) (Figure 4.4.3.a) In a warming climate with the greatest amount of ozone-depleting substances, the geological range of highly suitable habitats will gradually decrease to 34731.67 km 2 by

2050 and increase to 36408.51 km 2 by 2080 (Figure 4.4.3.a)

Figure 4.4.3.a The forecasted suitable area for Anoectochilus setaceus in Vietnam under different climate scenarios

U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h

Present RCP 2 2050 RCP 2 2080 RCP 8 2050 RCP 8 2080

Figure 4.4.3.b Habitat map with appropriate levels - Anoectochilus setaceus in Vietnam under current and future climate conditions

A – Present; B – RCP 2 2050; C - RCP 2 2080; D – RCP 8 2050; E – RCP 8 2080

According to Table 4.4.3, there has been no changes in the top 3 positions Precipitation of Wettest Quarter (Bio 16) still remain the top position, followed by Mean Temperature of Coldest Quarter (Bio 11) and Temperature Annual Range (Bio 7) at the second and third position, respectively In summary, precipitation-related variables add up more to the distribution of A setaceus in the future, with the mean value of 55.975%

Table 4.4.3 Contribution rate of environmental variables - Anoectochilus setaceus

Min Temperature of Coldest Month (Bio 6) 0.7 4.4 0.9 25.7

Temperature Annual Range (Bio 7) 7.7 9 7.2 7.4 Mean Temperature of Coldest Quarter (Bio 11) 31.6 34.2 31.9 8.7

Precipitation of Wettest Month (Bio 13) 1.0 - 1.1 -

Precipitation of Wettest Quarter (Bio 16) 54.4 49.5 57.1 54.5

Precipitation of Warmest Quarter (Bio 18) 1.0 0.4 1.1 1.4

A warming climate with the fewest ozone-depleting substances will promote a gradual elevation of suitable habitat north of Vietnam (Figure 4.4.4.b RCP 2 2050, RCP 2 2080), from 203195.66 km 2 (present) to 204565.43 km 2 (RCP 2 2050) and then 211305.03 km 2 (RCP 2 2080) (Figure 4.4.4.a) In a warming climate with the greatest amount of ozone-depleting substances, the geological range of highly suitable habitats will decrease to 198962.88 km 2 by 2050 and to 175229.57 km 2 by 2080 (Figure 4.4.4.a)

Figure 4.4.4.a The forecasted suitable area for Stephania japonica in Vietnam under different climate scenarios

U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h

Present RCP 2 2050 RCP 2 2080 RCP 8 2050 RCP 8 2080

Figure 4.4.4.b Habitat map with appropriate levels - Stephania japonica in Vietnam under current and future climate conditions

A – Present; B – RCP 2 2050; C - RCP 2 2080; D – RCP 8 2050; E – RCP 8 2080

According to Table 4.4.4, there has been a slight change in the position of the top 3 contributors Precipitation of Warmest Quarter (Bio 18) surpassed Precipitation of Wettest Quarter (Bio 16) for the highest position, while Precipitation of Wettest Month (Bio 13) climbed its way up to the third rank In summary, precipitation-related variables add up more to the distribution of S japonica in the future, with the mean value of 79.15%

Table 4.4.4 Contribution rates of environmental variables - Stephania japonica

Precipitation of Wettest Month (Bio 13) 12.1 0.1 41.7 4.5

Precipitation of Wettest Quarter (Bio 16) 26.9 34.8 0.1 36.5 Precipitation of Warmest Quarter (Bio 18) 30.1 26.8 24.7 27.3

A warming climate with the fewest ozone-depleting substances will gradually reduce suitable habitat (Figure 4.4.5.b RCP 2 2050, RCP 2 2080), from 6088.68 km 2 (current) to 630.55 km 2 (RCP 2 2050) and then 2811.61 km 2 (RCP 2 2080) (Figure 4.4.5.a) In a warming climate with the greatest amount of ozone-depleting substances, the geological range of highly suitable habitats will decrease to 1612.54 km 2 by 2050 and to 919.98 km 2 by 2080 (Figure 4.4.5.a)

Figure 4.4.5.a The forecasted suitable area for Berberis julianae in Vietnam under different climate scenarios

U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h U ns ui ta bl e L ow M ed iu m H ig h

Present RCP 2 2050 RCP 2 2080 RCP 8 2050 RCP 8 2080

Figure 4.4.5.b Habitat map with appropriate levels - Berberis julianae in Vietnam under current and future climate conditions

A – Present; B – RCP 2 2050; C - RCP 2 2080; D – RCP 8 2050; E – RCP 8 2080

According to Table 4.4.5, there has been no changes in the top 3 positions Precipitation of Driest Month (Bio 14) still remain the top position, followed by Min Temperature of Coldest Month (Bio 6) and Mean Temperature of Coldest Quarter (Bio 11) at the second and third position, respectively In summary, temperature-related variables add up more to the distribution of B julianae in the future, with the mean value of 59.575%

Table 4.4.5 Contribution rates of environmental variables - Berberis julianae

Min Temperature of Coldest Month (Bio 6) 25.1 22.9 18.2 17.8 Mean Temperature of Coldest Quarter (Bio 11) 14.6 13.9 12.3 13.2 Precipitation of Driest Month (Bio 14) 37.3 39.7 37.3 33.9

DISCUSSION AND CONCLUSION

Discussion

The outcomes offer important new information on the effects of climate change on the number and distribution of medicinal plants in Vietnam The distribution of medicinal plants under various climatic scenarios may now be predicted thanks to the use of MaxEnt software and SDMs The results indicate that many species may experience range changes and losses in appropriate habitats, which would likely have a substantial influence on the distribution and abundance of therapeutic flora in Vietnam

The study revealed several medicinal plant species, including Berberis julianae and Anoectochilus setaceus, that are at high risk of climate change

On the other hand, the study also identified several medicinal plant species that may benefit from climate change For example, Morinda officinalis and Taxus wallichiana are likely to thrive under future climate scenarios This suggests that these species may become more widely available for medicinal use in the future, potentially providing new opportunities for traditional medicine practices

Overall, this study provides valuable insights into the potential influences of climate change on medicinal plants in Vietnam The findings highlight the need for effective conservation strategies to protect the most vulnerable species and ensure their continued availability for traditional medicine practices Conservation strategies for medicinal plant species in Vietnam must consider the unique cultural, ecological, and economic factors that influence their use and conservation Effective conservation strategies may include measures such as habitat restoration and protection, sustainable harvesting practices, and ex-situ conservation efforts, namely seed banks and botanical gardens The involvement of local communities in conservation efforts is also critical for the success of these strategies Engaging with local people and practitioners of traditional medicine may assist in ensuring that conservation efforts are successful and culturally acceptable while also revealing important information on the usage and conservation of medicinal plants The development of alternative medicinal plant sources, such as plantations or cultivation programs, may together help to reduce pressure on wild populations and ensure the continued availability of these species for traditional medicine practices

Furthermore, the identification of species that may benefit from climate change provides new opportunities for medicinal plant use and highlights the potential for adaptation to changing environmental conditions For example, the increased potential distribution of Morinda officinalis and Taxus wallichiana under future climate scenarios could result in expanding the cultivation of these species in new areas or under different environmental conditions, thereby diversifying their production and increasing their availability for medicinal purposes Additionally, the study findings could inspire research into the potential medicinal properties of other plant species that may be expanding their ranges under future climate scenarios This could lead to the identification of new medicinal plant species that could be used to develop new medicines or supplement existing ones Ultimately, the identification of species that may benefit from climate change provides an opportunity to explore new ways of using and managing medicinal plants in the face of changing environmental conditions and can help to ensure their continued availability for traditional medicine practices

The output AUC of the nine study-created models, with a minimum value of 0.945 and a maximum value of 0.994 (Table 3.1), demonstrates good predictive capacity The AUC statistic, which is frequently employed in SDMs, is used to assess a model's capacity to distinguish between presence and absence points for a species

The lowest-performing model in the collection nonetheless appears to have a significant discriminating power in correctly predicting the suitable habitat for therapeutic plants, according to the minimum AUC value of 0.945 On the other hand, the highest- performing model has an excellent capacity to distinguish between habitat regions that are appropriate and unsuitable for the species, according to the maximum AUC value of 0.994

These high AUC values suggest that the models have captured the important environmental factors influencing the distribution of the medicinal plant species accurately

It indicates that the models can effectively predict the possible distribution of the species under various environmental scenarios However, it's important to note that although AUC is a valuable measure of model performance, it's recommended to assess other evaluation metrics and consider the ecological context and limitations of the specific study system to have a comprehensive understanding of the model's accuracy and reliability

The limitations associated with the use of MaxEnt software and SDMs must be considered when interpreting these results For instance, the quality and resolution of the environmental data utilized has a significant impact on how accurate these models are In addition, the models use the assumption that species and their environments are always in balance, which may not always be the case, particularly in climates that are changing quickly

The research findings indicate that different climatic factors play significant roles in controlling the distribution of medicinal plant species in Vietnam (Table 4.2.1) Specifically, rainfall appears to be a key determinant for the distribution of Anoectochilus setaceus, Stephania japonica, Panax vietnamensis, Taxus wallichiana, and Panax bipinnatifidus These species show a high percentage (ranging from 63.2% to 80.9%) of their distribution influenced by rainfall patterns This suggests that these plants are particularly sensitive to precipitation levels, and alterations in rainfall patterns due to climate change can have a substantial impact on their suitable habitat availability

On the other hand, temperature emerges as a significant determinant for the distribution of Coscinium fenestratum, Morinda officinalis, Berberis julianae, and Cupressus torulosa These species exhibit a notable percentage (ranging from 63.3% to

82%) of their distribution influenced by temperature This indicates that variations in temperature, including average temperatures and temperature ranges, have a strong influence on the habitat suitability for these floras

These findings highlight the importance of considering specific climatic factors such as rainfall and temperature when studying the distribution patterns of medicinal plant species Understanding the climatic determinants of their distribution can aid in how vulnerable they are to climate change and developing effective conservation strategies to mitigate potential impacts on their availability and survival

5.1.3 Habitat Suitability and Climate Change Impacts

Overall, the study identifies a general trend of range reductions for several medicinal plant species under future climate scenarios This suggests that the suitable habitats for these species may shrink, potentially leading to localized extinctions or significant population declines Conversely, there are also instances where certain species exhibit an expansion of suitable habitats, indicating the potential for range shifts in response to changing environmental settings These contrasting trends highlight the dynamic nature of species responses to climate change and the importance of considering species-specific characteristics and ecological requirements in understanding their potential distributional changes

Among the studied species, specific plants and regions in Vietnam are distinguished as being especially vulnerable to climate change For instance, the species Morinda officinalis, Stephania japonica, Panax vietnamensis, Taxus wallichiana, and Panax bipinnatifidus are found to be highly vulnerable, with significant range reductions projected under future climate scenarios These species are concentrated in certain regions of Vietnam, such as the mountainous areas or specific ecosystems, where they are highly dependent on specific climatic conditions for their survival and persistence

Understanding the general trends of species increase or decrease and the identification of vulnerable species and regions are crucial for compelling conservation planning and management These findings highlight the urgency of implementing conservation measures to protect and sustainably manage the most vulnerable medicinal plant species It is essential to consider the ecological requirements and adaptive capacities of these species to progress robust conservation policies that can diminish the adverse effects of climate change and ensure the continued availability and accessibility of these valuable plant resources for traditional medicine practices and the overall well-being of communities relying on their medicinal properties

The effects of climate change on plant distribution are a subject of increasing concern in ecological research As Earth experiences rapid shifts in climatic conditions, it is becoming evident that these changes can have both positive and negative consequences for species distribution patterns While some species may experience range contractions and reduced suitable habitat under future climate scenarios, others may exhibit range expansions and colonization of new areas The outcome largely depends on various factors, including species-specific characteristics, their ecological requirements, and the specific climatic conditions projected for different regions a Coscinium fenestratum:

Conclusion

This study was conducted to predict the distributions of nine medicinally important species of medicinal plants in Vietnam using the SDM, which can help ensure wild and effectively plan viable or developed field transformation actions

1) Species whose distribution is largely influenced by rainfall: Anoectochilus setaceus (74.5%), Stephania japonica (80.9%), Panax vietnamensis (63.2%), Taxus wallichiana

Species whose distribution is largely influenced by temperature: Coscinium fenestratum (64%), Morinda officinalis (70.8%), Berberis julianae (63.3%) and Cupressus torulosa (82%)

2) Under the influence of climate change with two different future scenarios, RCP 2 and RCP 8, the suitable habitat area of each tree species also changes in different directions (Table 4.2)

3) The habitat maps depicting the specific distribution patterns of each species have provided valuable insights into the shifting trends of suitable habitats By analyzing these maps, we can discern the direction in which the suitable habitats for each species are projected to move under changing climatic conditions This knowledge becomes instrumental in formulating targeted strategies for the cultivation, conservation, and protection of these medicinal plant species

1 Abdelaal, M., Fois, M., Fenu, G., & Bacchetta, G (2019) Using MaxEnt modeling to predict the potential distribution of the endemic plant Rosa arabica Crép in Egypt Ecological informatics, 50, 68-75

2 Anderson, R P (2017) When and how should biotic interactions be considered in models of species niches and distributions? Journal of Biogeography, 44(1), 8-17

3 Anh, B T., Tu, T V., & Trang, N T M (2022) Some morphological, biological and ecological characteristics of Hoang Lien gai (Berberis julianae CK Schneid) in Hoang Lien National Park, Lao Cai Journal of Agricultural Science and Technology Hue University of Agriculture and Forestry, 6(2), 3067-3074

4 Applequist, W L., Brinckmann, J A., Cunningham, A B., Hart, R E., Heinrich, M., Katerere, D R., & Van Andel, T (2020) Scientistsʼ warning on climate change and medicinal plants Planta medica, 86(01), 10-18

5 Austin, MP (2002) Spatial prediction of species distribution: an interface between ecological theory and statistical modeling Ecological modeling, 157 (2-3), 101-

6 Assessment, ME (2005) Ecosystems and human well-being: synthesis Island Washington, DC

7 Baldwin, RA (2009) Use of maximum entropy modeling in wildlife research Entropy, 11 (4), 854-866

8 Balmford, A., & Bond, W (2005) Trends in the state of nature and their implications for human well-being Ecology Letters, 8 (11), 1218-1234

9 Change, IPOC (2007) Climate change 2007: The physical science basis

10 Donoghue, MJ, & Edwards, EJ (2014) Biome shifts and niche evolution in plants Annual Review of Ecology, Evolution, and Systematics, 45, 547-572

11 Elith*, J., H Graham*, C., P Anderson, R., Dudík, M., Ferrier, S., Guisan, A., & E Zimmermann, N (2006) Novel methods improve the prediction of species distributions from occurrence data Ecography, 29 (2), 129-151

12 Forester, B R., DeChaine, E G., & Bunn, A G (2013) Integrating ensemble species distribution modelling and statistical phylogeography to inform projections of climate change impacts on species distributions Diversity and Distributions, 19(12), 1480-

13 Figueiredo, FO, Zuquim, G., Tuomisto, H., Moulatlet, GM, Balslev, H., & Costa, FR (2018) Beyond climate control on species range: The importance of soil data to predict the distribution of Amazonian plant species Journal of Biogeography, 45 (1), 190-

14 Foley, JA, DeFries, R., Asner, GP, Barford, C., Bonan, G., Carpenter, SR,

& Snyder, PK (2005) Global consequences of land use science, 309 (5734), 570-574

15 Guisan, A., & Thuiller, W (2005) Predicting species distribution: offering more than simple habitat models Ecology Letters, 8 (9), 993-1009

16 Guisan, A., and Zimmermann, NE (2000) Predictive habitat distribution models in ecology Ecol Model 135, 147–186

17 Hawkins, BA, Field, R., Cornell, HV, Currie, DJ, Guégan, JF, Kaufman, DM, & Turner, JR (2003) Energy, water, and broad-scale geographic patterns of species richness Ecology, 84 (12), 3105-3117

18 Hijmans, RJ, Cameron, SE, Parra, JL, Jones, PG, & Jarvis, A (2005) Very high resolution interpolated climate surfaces for global land areas International Journal of Climatology: A Journal of the Royal Meteorological Society, 25 (15), 1965-1978

19 Hoi, Q V., Tien, T V., Trieu, L N., Duy, N V., Chac, L D., Chinh, H V.,

& Thinh, B B (2021) Use of inter simple sequence repeat (ISSR) markers to assess the genetic diversity of Panax bipinnatifidus Seem Collected From Northern Vietnam Russian Journal of Genetics, 57, 341-347

20 Hong-Qun, L., Li-Gang, X., & Xie-Ping, S (2022) Predicting the potential distribution of Taxus wallichiana var mairei under climate change in China using Maxent modeling Pakistan J Bot, 54, 1305-1310

21 Kovács, P., Csaba, G., Pállinger, É., & Czaker, R (2007) Effects of taxol treatment on the microtubular system and mitochondria of Tetrahymena Cell biology international, 31 (7), 724-732

22 Le Tran, Q., Tezuka, Y., Ueda, J Y., Nguyen, N T., Maruyama, Y., Begum, K., & Kadota, S (2003) In vitro antiplasmodial activity of antimalarial medicinal plants used in Vietnamese traditional medicine Journal of Ethnopharmacology, 86(2-3), 249-

23 MacKenzie, D I., Bailey, L L., & Nichols, J D (2004) Investigating species co‐occurrence patterns when species are detected imperfectly Journal of Animal

24 Menon, S., Choudhury, BI, Khan, ML, & Peterson, AT (2010) Ecological niche modeling and local knowledge predict new populations of Gymnocladus assamicus a critically endangered tree species Endangered Species Research, 11 (2), 175-181

25 Merow, C., Smith, MJ, & Silander Jr, JA (2013) A practical guide to MaxEnt for modeling species' distributions: what it does, and why inputs and settings matter

26 Ministry of Agriculture and Rural Development (2017) Inter-Ministerial Circulation No 04/2017/TT-BNNPTNT

27 Molyneux, N., Soares, I., & Neto, F (2013) WorldClim and DIVA software for modeling current and future climates at a 5 km resolution Combating Climate Change, 325-342

28 Nguyen, H T., Nguyen, T D., Le, D H., Nobleza, N., & Trinh, Q P (2019) Diversity of plant-parasitic nematodes on medicinal plants in melinh station for biodiversity, Vinh Phuc Province, Vietnam Academia Journal of Biology, 41(3), 19-24

29 Nguyen, H T., Vu-Huynh, K L., Nguyen, H M., Le, H T., Le, T H V., Park, J H., & Nguyen, M D (2021) Evaluation of the Saponin Content in Panax vietnamensis Acclimatized to Lam Dong Province by HPLC– UV/CAD Molecules, 26(17), 5373

30 Nguyen, T H., Phan, K L., Nguyen, D T L., Thomas, P I., Farjon, A., Averyanov, L., & Regalado Jr, J (2005) Vietnam conifers: conservation status review

2004 Hanoi: Fauna and Flora International Vietnam Programme

31 Nguyen, T S., Xia, N H., Van Chu, T., & Van Sam, H (2019) Ethnobotanical study on medicinal plants in traditional markets of Son La province, Vietnam Forest Soc, 3(2), 171-92

32 Nisar, M., Khan, I., Simjee, SU, Gilani, AH, & Perveen, H (2008) Anticonvulsant, analgesic and antipyretic activities of Taxus wallichiana Zucc Journal of

33 Palmer, T M., Stanton, M L., & Young, T P (2003) Competition and coexistence: exploring mechanisms that restrict and maintain diversity within mutualist guilds the american naturalist, 162(S4), S63-S79

34 Paz, A., Reginato, M., Michelangeli, FA, Goldenberg, R., Caddah, MK, Aguirre-Santoro, J., & Carnaval, A (2020) Predicting Patterns of Plant Diversity and Endemism in the Tropics Using Remote Sensing Data: A Study Case from the Brazilian Atlantic Forest Print Remote Sensing of Plant Biodiversity (pp 255-266) Springer, Cham

35 Pearson, RG (2010) Species' distribution modeling for conservation educators and practitioners Lessons Conserv 3, 54–89

36 Pimm, SL, Russell, GJ, Gittleman, JL, & Brooks, TM (1995) The future of biodiversity Science, 269 (5222), 347-350

37 Phillips, S J (2005) A brief tutorial on Maxent AT&T Research, 190(4),

38 Phillips, SJ, Anderson, RP, & Schapire, RE (2006) Maximum entropy modeling of species geographic distributions Ecological modeling, 190 (3-4), 231-259

39 Phillips, SJ, & Dudik, M (2008) Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation Ecography, 31 (2), 161-175

40 Pownitha, K V., Hulegaru Nagaraja, P B., Charles, B., Vasudeva, R., Aravind, N A., & Ravikanth, G (2022) Ecological niche modelling to identify suitable sites for cultivation of two important medicinal lianas of the Western Ghats, India Tropical

41 Radosavljevic, A., & Anderson, RP (2014) Making better MaxEnt models of species distributions: complexity, overfitting, and evaluation Journal of Biogeography,

42 Scarpelli, CM (2018) Out of the Ashes: Examining Forest Biodiversity Conservation Efforts in Post-Conflict Vietnam and Sierra Leone (Doctoral dissertation, University of Oregon)

43 Stern, N., & Stern, NH (2007) The economics of climate change: the Stern review Cambridge University Press

44 Swets, JA (1988) Measuring the accuracy of diagnostic systems Science,

45 Thomas, CD, Cameron, A., Green, RE, Bakkenes, M., Beaumont, LJ, Collingham, YC, & Williams, SE (2004) Extinction risk from climate change Nature,

46 Center for Forest Resources and Environment - Forest Inventory and Planning Institute (2010) Project report "Investigation and assessment of conservation status of endangered and rare forest plant species on the list of Decree 32/2006/ND-CP by ecological region"

47 Truong, TT, Hardy, GESJ, & Andrew, ME (2017) Contemporary remotely sensed data products refine invasive plants risk mapping in data-poor regions Frontiers in plant science, 8, 770

48 UNFCCC Adoption of the Paris Agreement Report No FCCC/CP/2015/L.9/Rev.1, http://unfccc.int/resource/docs/2015/cop21/eng/l09r01.pdf

49 Van Vuuren, D P., Edmonds, J., Kainuma, M., Riahi, K., Thomson, A., Hibbard, K., & Rose, S K (2011) The representative concentration pathways: an overview Climatic change, 109, 5-31.