Tropical Ecology International Society

https://doi.org/10.1007/s42965-022-00237-x for Tropical Ecology

RESEARCH ARTICLE

Identifying the potential global distribution and conservation

areas for Terminalia chebula, an important medicinal tree species
under changing climate scenario
B. R. Kailash1,2 · Bipin Charles3 · G. Ravikanth2 · Siddappa Setty2 · K. Kadirvelu4

Received: 4 November 2020 / Revised: 9 March 2022 / Accepted: 20 March 2022

© International Society for Tropical Ecology 2022

Abstract
Terminalia chebula Retz. (Combretaceae), commonly-known as chebulic myrobalan is one of the important Non-Timber
Forest Product (NTFP) species which is harvested for its fruits and galls. The species known as the “King of medicines” is
used widely in Ayurveda, Sidda, Unani, and traditional Chinese medicines for curing a wide variety of diseases in Asia and
Africa. Terminalia chebula is an important ingredient of Triphala (Ayurvedic medicine) along with Terminalia bellirica
and Phyllanthus emblica. The fruits of the tree also yields a dye which is used as an organic dye in the textile industries. In
recent years, there is an increasing demand for herbal remedies and organic dyes, resulting in extensive extraction of fruits
and galls from T. chebula. In this study, the major objective was to identify sites for the conservation of T. chebula and to
identify important environmental variables determining its distribution. Based on the existing species distribution records
(primary and secondary), along with a suite of climatic variables, the present and future distribution of the species were
predicted. The study identified ecological niches that are suitable for the cultivation of the species; the species occurs in
India, Pakistan, Sri Lanka, Cambodia, Myanmar, Vietnam, China, Laos, Thailand, Bhutan, Taiwan, Nepal and Bangladesh
under the current climatic scenario. Within India, our results suggest that the central and south India are highly suitable in
the current scenario. The mean annual temperature, temperature seasonality and isothermality seem to be the most important
variables determining the distribution of the species which is directly influenced by climate change. Overall, the study indi-
cated that under the future climate change scenarios the distribution of T. chebula is likely to decrease. The results indicate
that T. chebula is highly vulnerable to climate change. Considering the economic importance of the species, it is important
to understand how the species distribution will alter in the wake of climate change to develop effective conservation strate-
gies. The study also provides important environment variables that determine the species distribution which could aid in
identifying areas where the species could be cultivated.

Keywords Conservation · Conservation sites · Ecological niche models · Maximum entropy

Introduction
* B. R. Kailash
[email protected] Climate change is one of the major factors influencing the
distribution of plant species. The Intergovernmental Panel
1
Environmental Sciences, Bharathiar University, on Climate Change (IPCC) in its fifth annual report (2014)
Marudhamalai Road, Coimbatore, Tamil Nadu 641046, India has indicated that the mean annual temperature of the
2
Ashoka Trust for Research in Ecology & the Environment, earth’s surface increased by 0.85 °C over the past 130 years
Royal Enclave, Sriramapura, Jakkur Post, Bangalore 560064, (1880–2012). However, the mean temperature increased
Karnataka, India
3
by 0.72 °C between the years (1951–2012) and the mean
Institute for Biodiversity Conservation and Training, #5, 7th global surface temperature is predicted to increase by an
Main Road, Shankar Nagar, Bangalore 560096, Karnataka,
India additional 0.3–4.8 °C by the end of the twenty-first century
4 (IPCC 2013). Climate change has resulted in an increased
DRDO‑BU Centre for Life Sciences, Defence Research &
Development Organisation, Ministry of Defence, Bharathiar temperature and irregular rainfall patterns, both droughts,
University Campus, Coimbatore, Tamil Nadu 641046, India and floods, which has led to uncontrolled insect infestations

13
Vol.:(0123456789)
 Tropical Ecology

and dominance of invasive species, increased fire vulner- and Bhattacharyya 2007). In Myanmar, T. chebula sparsely
ability, and severely impacting biodiversity, as well as the occurs in degraded and timber harvested forests (Oo and Lee
livelihood of the people (Ravindranath et al. 2006; Mukho- 2007). In Sri Lanka, T. chebula occurs in mixed deciduous
padhyay 2009, Shresta 2019). With an average rise in tem- forests and scrubs (APFSOS II). Besides deciduous forests
peratures of about 2–3 °C globally, there is an increased in Cambodia, Laos, Southern Laos and northeastern Thai-
risk of upto 20–30 percent of species extinction (Warren land, T. chebula is grown in between paddy fields (Natuhara
et al., 2010; Stocker et al. 2014). Several studies have shown et al. 2012). In South-eastern China, T. chebula is present in
that climate change can cause significant change to ecologi- open deciduous and bamboo forests, along forest margins,
cal processes including change in the distributional range and cultivated near villages. It is distributed across India
of species (Priti et al. 2016; Subba et al., 2018; Rana et al. from Arunachal Pradesh to Punjab and from Tamil Nadu to
2021) as well as impacting the phenology of tree species Jammu & Kashmir. With the increasing demand for herbal
(Piao et al. 2019). Assessing and predicting the potential remedies and organic dyes, there is extensive extraction of
change in the distributional range of species is crucial for fruits and galls from the T. chebula from the various forests
developing optimum conservation and management plans. across India.
In recent years, ecological niche modelling (ENM, also In this study using distribution data, we have mapped
known as species distributional modelling, SDM) have been the geographical distribution of the species, predicted the
widely used to infer the environmental variables required possible sites of occurrence of T. chebula in India using
for the survival of a species. The ENMs use biophysical Ecological Niche Modelling (ENM). The major objective
methods and environmental variables taken from the known of the study was to identify sites for the conservation of
distribution areas of the species to predict their ecological this economically important medicinal tree species. We also
niche as well as their potential future distribution (Elith et al. intended to identify the environmental factors that are criti-
2009; Shivaprakash et al. 2013). The ENM is widely used cal for the distribution of the species both in the present as
for assessing the distribution of species (Sen et al. 2016a; well as in future climate change scenarios. Based on these,
Kumar et al. 2020), for assigning conservation areas (Joshi we intend to determine ecological niches that are suitable
et al. 2017; Sumangala et al. 2017), for predicting invasion for the cultivation of the species both in India as well as in
risk (Sen et al. 2016b) and in assessing impacts of climate other South Asian countries.
change on species distributions (Priti et al. 2016; Subba
et al., 2018; Rana et al. 2021).
Terminalia chebula Retz. (Combretaceae) is an impor- Materials and methods
tant plant used in traditional medicine not only in India and
China but also in South Asia and in Africa. The plant is Species occurrence records
widely used for curing multiple ailments such as digestive
disorders, urinary diseases, diabetes, skin diseases, heart Global occurrence data for T. chebula were collected by
diseases, irregular fevers, constipation, ulcers, vomiting, primary and secondary sources (See Table 1). Primary data
colic pain, and haemorrhoids (Muhammad et al. 2012; sources were collected by field survey’s using GPS, where
Nigam et al. 2020). The plant is considered as the “King the actual location of the species was recorded. Totally 49
of medicine” due to the presence of a large number of phy- primary data points were collected from the field. The sec-
tochemicals that are used for treating a variety of disorders ondary data were collected from multiple resources which
(Gupta 2012). The plant is extensively used in Indian tradi- include free and open access viz., 167 points from Global
tional medicine such as Ayurvedic, Sidda, Unani. Terminalia biodiversity information system (GBIF), 386 points from
chebula is an important ingredient of Ayurvedic medicine the India Bioresource Information network (www.​ibin.​gov.​
“Triphala” for curing many diseases. The fruits of T. chebula in), 50 points from the Indian Biodiversity Portal. Further,
have been used as a household remedy since time immemo- the distribution records were also obtained from floras (273
rial (Bag et al., 2013). This species has a profound impact points), from the Herbarium records collected from Central
on human health in the Asian tropics. In Southeast Asia, T. National Herbarium (CAL, 91 points), Botanical Survey of
chebula fruits are widely used in traditional medicine, dye- India (BSI), Calcutta, Northern Regional Centre, Dehradun
ing, and are also traded (Nigam et al. 2020). The dried galls (NRC, 10 points), Western Regional Centre, Pune (WRC, 15
of T. chebula are considered to have anti-aging properties, points) and Southern Regional Centre, Coimbatore (SRC, 51
are astringent, purgative, and boost immunity and are widely points). In all, up to 1092 records were collected.
used in Thai Lanna, Tibetan, and Chinese medicine. They The secondary resources often tend to have spatial sam-
are sold at markets across Southeast Asia (Manosroi et al. pling biases (Aiello‐Lammens et al. 2015; Sumangala et al.
2010). Terminalia chebula occurs in mixed deciduous forests 2017; Ahmad et al. 2019; Hamid et al. 2019). Addressing
and irregularly in teak forests in Malaysia (Chattopadhyay the effects of sampling bias remains an important issue. For

13
Tropical Ecology

Table 1  Distribution of Terminalia chebula in other Southeast Asian countries

Sl Countries Distribution References

1 Bangladesh Rare in Moulvibazar, Chittagong and Mymensingh districts of Kabir et al. (2014); Akbar (2020)
Bangladesh
2 Bhutan Goshing, Ngangla and Phangkhar Gewogs of Kheng region in Purna and Tenzi (2012); Phurpa et al (2017)
Bhutan. Deciduous forests
3 Cambodia Virachey National Park, Veun Sai-Siem Pang National Park, Siem, Philip (1999); Loveridge et al. (2017)
Pang Khang Lech Wildlife Sanctuary in Cambodia
Deciduous forests and cultivated
4 China Native in West Yunnan, cultivated in Fujian, Guangdong, Guangxi Flora of China
5 Indonesia Riau Province, Sumatra, Indonesia Catalogue of Life; Mahyar et al. (1991)
Cultivated
6 Laos Nakai-Nam Theun National Park, Laos.Dry deciduous forests and Philip (1999)
degraded savannas in northern Thailand and Laos
7 Malaysia Sarawak Rathinamoorthy and Thilagavathi (2014); Nicholson 1965
8 Myanmar Bago, Kachin, Mandalay and Yangon of Myanmar Rathinamoorthy and Thilagavathi (2014)
Deciduous forests
9 Nepal Rasuwa and Jhapa District, Nepal. Deciduous forests Annotated checklist of Flowering Plants of Nepal; Saru
et al 2020; Yadav et al 2010; Raut et al 2018
10 Pakistan North West Frontier Province, Pakistan and also cultivated Arshad et al. (2010); Flora of Pakistan
11 Sri Lanka Gal Oya National Park of Badulla and Wilpattu National Park of Forest Department Government of Sri Lanka (2009); San-
Anuradhapura and Puttalam districts jeewa et al 2013; Green (1990); Wilpattu National Park
Dry and open forests. Occur Naturally in Intermediate zones and
cultivated in agro-climatic zones
12 Taiwan Nantou Tsien, Chushan Town and Kaoshiung Hsien of Taiwan, Taiwan Plant names; Database of Native Plants of Taiwan
also cultivated
13 Thailand Chae-Son district, Lampang, Thailand Philip (1999); Wilart (2020)
Dry deciduous forests and degraded savannas
14 Vietnam Buon Ma Thuot, Dak Lak Province, Vietnam Philip (1999); Nguyen (2016)
Dry deciduous forests and degraded savannas Tropicos.org

many datasets of occurrence records (especially from muse- distribution of the species. Bioclimatic data used for mod-
ums and herbaria), geographic sampling bias is pervasive elling had a spatial resolution of 1 ­km2 (30 arc seconds).
(Hijmans et al. 2000; Reddy and Dávalos 2003; Graham The environmental variables often have a high correlation.
et al. 2004; Kadmon et al. 2004; Hijmans 2012). Ideally, Hence, we tested for collinearity by examining pairwise
spatial thinning of occurrence records is inevitable to sub- correlations between the variables. Among the pair of vari-
stantially reduce the effects of sampling bias whilst simulta- ables, those variables having Pearson’s correlation coeffi-
neously retaining the greatest amount of useful information. cient r > 0.7 were excluded from the analysis (Dormann et al.
The spatial thinning can be done manually; however, this is 2013; Sarma et al., 2018). The final set of variables used for
prohibitively time consuming for large datasets. Therefore, modelling T. chebula are Bio1, Bio2, Bio3, Bio4, Bio12,
we used the randomization approach, i.e., the ‘thin’ function Bio14, Bio15, Bio18 and Bio19 (Appendix S1).
in the spThin R package to obtain an occurrence dataset
with the maximum number of records for thinning distance Species distribution modelling
of 10 km. The output of this exercise was used to develop
a robust species distribution model for the target species. We used the maximum entropy implemented in Maxent
which simulates the potential geographical distribution of a
Climatic variables species using its current distribution as well as by employ-
ing various environmental data (Phillips et al. 2006, 2017).
Species Distribution Modelling (SDM) require a suite of the Maxent is the most favoured model as it is simple and quick
climatic representative dataset that influence distribution of to run, provides stable operation results, and allows predic-
species. Hence, ecologically meaningful variables i.e., bio- tion results to be tested (Ortega-Huerta et al. 2008). Max-
climatic variables i.e., Bioclim1—19 V1.4 (Hijmans et al. ent, requires the presence only data and background data
2005) were downloaded from WorldClim (Table 1, https://​ for modelling the ecological niches. The performance of the
www.​world​biocl​im.​org) and were used for modelling the model is robust and relatively better than other modelling

13
 Tropical Ecology

algorithms (Elith et al. 2006; Rahimian et al. 2019; Ghar- type as ‘cloglog’ and other features were default. We used 70
eghan et al. 2020) and is hardly influenced by small sample percent of the occurrence points for model calibration and 30
sizes (Pearson et al. 2007). We used maxent and ran the percent for model testing with independent validation as per
model for Current, and different Representative Concentra- Bohl et al. (2019). The variable importance was measured
tion Pathway (RCP) such as RCP 2.6, RCP 4.5, RCP 6.0 & using the Jack-knife test to determine the dominant climatic
RCP 8.5 for 2070. The year 2070 was taken for the analysis factors (Raman et al. 2020), and all other default settings
as the annual temperature increase in 2070 has been pre- based on Elith et al. (2010).
dicted to be over 7.5 °C across most of Asia (Xu et al. 2020).
The various RCP’s used in the study describe different plau- Model evaluation
sible climate futures, all of which are likely possible depend-
ing on the extent of greenhouse gas emissions. IPCC has The model was evaluated by Area under Receiver Operating
adopted these four different climate change scenarios under Curve (ROC) popularly known as AUC (Swets 1988). The
the various Representative Concentration Pathways (RCPs). AUC values vary from 0 to 1 where the value < 0.5 signi-
fies that the model is worse than random, 0.5–0.7 signifies
Model development and selection poor performance, 0.7–0.9 signifies moderate performance
and > 0.9 signifies high performance (Kumar et al., 2020).
Ranges of models were developed for mapping the climatic The final model obtained showed high performance value
niche of T. chebula. Further, to assess the model complex- indicating that the performance was good. The final pre-
ity, we additionally fit the MaxEnt models using the exist- dicted Model of habitat suitability were divided into four lev-
ing default settings, which enabled the implementation of els: unsuitable regions (0.00–0.25), poorly suitable regions
more complex feature classes such as (L-linear, Q-quadratic, (0.25–0.50), moderately suitable regions (0.50–0.75), and
H-Hinge) depending on the occurrence records. We used highly suitable regions (> 0.75).
LQH (= linear, quadratic, hinge), LQ (= linear, quadratic),
LQH, LQ and H (= hinge) and regularization multipliers
0.5, 1.0, 1.5, 2.0, 2.5, and 3.0. We used randomly generated Results
10,000 points and default settings to calibrate the models.
The outputs i.e., predictions and lambdas file were down- A total of 1092 was obtained from various sources. After
loaded (Radosavljevic and Anderson 2014) and used for spatial thinning, a total of 1002 occurrence records were
model selection in ENM Tools. Maxent also produces con- retained which showed uniform distribution across the sam-
tinuously varying, non-negative suitability scores for each pling landscape. The distribution of the species using both
cell in a specified geographic region. The ENM Tools uses primary and secondary data within India shows that the spe-
this suitability scores and occurrence points and estimates cies is widely distributed with higher densities in the West-
the likelihood of observing the occurrence records under a ern and Eastern Ghats (Fig. 1). These points were used in
selected model. Once observed, a suite of parameters cap- building a model for current and future distribution (Fig. 2).
tured in lambdas file counting any parameter with weightage After correlation analyses and jack-knife tests (Fig. 3), nine
having non-zero will be estimated and produces an output environmental variables were used to construct the mod-
with criterion-based model selection using Akaike Informa- els. The variables that contributed significantly to a model
tion Criterion (AIC), Akaike Information Criterion corrected for the current distribution for Asia are Bio1 (Annual Mean
for small sample sizes (AICc) and Bayesian information Cri- Temperature) having a contribution of 35.1%, Bio4 (Tem-
teria (BIC) (Warren et al. 2010). The AIC, AiCc and BIC are perature Seasonality) with 24.6% and Bio3 (Isothermality)
quality metrics that are generally used to ensure the accuracy with 10.9% contribution (Table 2). The result of the jack-
of the modelling outcome. These metrics help at selecting knife test of variable importance shows that the environmen-
the “best model” which is a key step in ensuring the model- tal variable with the highest gain, when used in isolation, is
ling outcome. To accomplish this we used a script with a Bio4 (Temperature seasonality), which appears to have the
relative path to the files (predictions and lambdas). Previ- most useful information by itself.
ous studies indicate that AICc outperforms BIC in selecting
models on simulated data and hence we selected the mod- Model evaluations and current distributions
els having low AIC value (Warren et al. 2010). The model
selected for final distribution modelling features LQHP The predicted distribution model had mean AUC of
with Regularization Multiplier 1.5 (Appendix S2). The final 0.884 ± 0.006, which indicate that the performance of the
model was calibrated using the following settings: number of model was good (Fig. 4). As per the predictions, T. chebula
iterations were set to 5000, and 10,000 random background occurs in India, Pakistan, Sri Lanka, Cambodia, Myanmar,
points, replicated type as “sub-sample”, 10 replicates, output Vietnam, China, Laos, Thailand, Bhutan, Taiwan, Nepal

13
Tropical Ecology

Fig. 1  Distribution of Termina-

lia chebula in India

and Bangladesh under the current climatic scenario. Under favourable for the species. The model predicts, Bhutan
the current climatic scenario, the MaxEnt predicted India, with moderate to highly favourable climatic conditions.
Bangladesh, Sri Lanka, Bhutan, Myanmar (Burma), Laos, The Bangladesh, Myanmar, Thailand, Cambodia, Viet-
Thailand, Cambodia, Vietnam and Pakistan as favourable nam, Laos, is largely moderately favourable and intermit-
for the sustenance of the T. chebula. In India, the Dec- tent with highly favourable conditions. The South Pakistan
can plateau, central India, Western Ghats, are extremely shows mixture of moderate and high favourable zones. The
favourable whilst the North, and Eastern parts are highly northern part of the Sri Lanka showed intermittent high,
favourable. The other parts of the country are moderately medium and excellent habitats.

13
 Tropical Ecology

Fig. 2  Distribution of Terminalia chebula under different projected climate change scenarios

Fig. 3  Results of Jackknife test,

using AUC on test data

Future projections Lanka, Thailand, Cambodia and Vietnam face significant

decline i.e., moderate to low habitats. Under the RCP
The distribution of T. chebula, across future projec- 4.5—2070, the Deccan plateau and Central Parts of the
tions, showed significant decline under all RCPs (Fig. 5, India shows significant loss of the habitat when compared
Table 3). The RCP 2.6—2070, RCP 4.5—2070, RCP to current climatic scenario and the RCP 2.6. The Bang-
6.0—2070 and RCP 8.5—2070 projections reveal a sig- ladesh, Bhutan, Sri Lanka, Myanmar, Thailand, Laos,
nificant decline in the suitable habitats in India, Pakistan, Cambodia and Vietnam face further significant loss when
Sri Lanka, Cambodia, Myanmar, Vietnam, China, Laos, compared to Current and RCP 2.6—2070 scenarios. The
Thailand, Bhutan, Taiwan, Nepal and Bangladesh. The climatically suitable habitat under RCP 6.0—2070 in the
MaxEnt model for RCP 2.6—2070 shows significant loss Deccan plateau, central parts of India and Sri Lanka expe-
in the climatically suitable habitats in the central and west- riences significant loss when compared to current scenario
ern parts of the India. The Bangladesh, Bhutan, Laos, Sri and RCP 2.6—2070 however, these regions experience

13
Tropical Ecology

Table 2  Contribution of environmental variables compare to Current, RCP 2.6, RCP 4.5 and RCP 6.0. In gen-
Variable Description Percent eral, the studied species T. chebula is highly vulnerable to
contribu- climate change in all the projected climate change scenarios.
tion

bio1 Annual mean temperature 35.1

bio4 Temperature seasonality 24.6
Discussion
bio3 Isothermality 10.9
A number of factors such as climate change, habitat dis-
bio15 Precipitation seasonality 8.8
turbance, overharvesting, invasive species and pollution
bio18 Precipitation of warmest quarter 5.8
are viewed as major threat to global biodiversity in general
bio12 Annual precipitation 5.4
and the medicinal plant species in particular (Hirch 2010).
bio19 Precipitation of coldest quarter 4
Climate change results in increased temperatures, disrupt-
bio2 Mean diurnal range 3.7
ing precipitation patterns and in increasing extreme climate
bio14 Precipitation of driest month 1.8
events. These factors are known to affect growth and repro-
duction, alter phenological patterns of trees as well as influ-
ence the distribution of species at a regional scale (Priti et al.
2016). The current measures used for the conservation of
medicinal plant species such as the establishment of medici-
nal plant conservation areas (MPCA’s) within the protected
areas network are based on the current distribution of spe-
cies. But these areas would face increasing challenges as the
effects of climate change increase (Sen et al. 2016a; Joshi
et al. 2017). In the light of climate change altering species
distributions and impacting native plant populations, using
niche modelling approach to map the potential distribution
of species can aid in the development of effective restoration
and conservation management plans (Adhikari et al. 2018).
The T. chebula has become endangered locally in sev-
eral regions of India (Jagdish 2014) and the status of this
Fig. 4  Receiver operating characteristic (ROC) curve for the same tree remains largely unknown. We used Maxent and opti-
data, again averaged over the replicate runs mized factors known to influence the prediction results of
the model. The performance of the ENM is significantly
influenced by the number of records as well as the vari-
increase in the moderately suitable habitat when compared ables selected. It is suggested that at least 100 point records
to RCP 4.5—2070. should be used in the Maxent model to ensure optimum
The other countries Vietnam, Sri Lanka, Bhutan, Bangla- results (Saatchi et al. 2011). The outcome of the model is
desh, Pakistan, Laos, and Cambodia too face significant loss also influenced by the input variables and the accuracy of the
in the moderate, highly suitable and excellent habitats when datapoints. In this study, we used a large data set (up to 1002

Fig. 5  Area change analysis 9,00,000.00

using MaxEnt in Asia 8,00,000.00
7,00,000.00
6,00,000.00
5,00,000.00
4,00,000.00
3,00,000.00
2,00,000.00
1,00,000.00
-
Current Rcp2.6 Rcp4.5 Rcp6.0 Rcp8.5

Not Suitable Low Suitable Moderately Suitable Highly Suitable

13
 Tropical Ecology

Table 3  Change in the area Range Description Current RCP2.6 RCP4.5 RCP6.0 RCP8.5
range under different climatic
scenario (in Square kilometre) 0.0–0.25 Not Suitable 7,15,207.00 7,54,495.00 8,11,714.00 7,85,952.00 8,16,547.00
0.25–0.5 Low Suitable 1,26,171.00 1,24,440.00 99,399.00 1,11,487.00 1,17,325.00
0.5–0.75 Moderately Suitable 95,253.00 60,164.00 46,600.00 47,117.00 25,835.00
0.75–1.0 Highly Suitable 36,856.00 34,388.00 15,774.00 28,931.00 13,780.00

records after filtering) to reduce any errors and also ensured phenological changes are likely to vary among species (Shel-
that the data points reflect the actual distribution of the spe- don 2019). Thus, though the study shows that the suitable
cies. We used the nine bioclimatic variables that contributed areas of T. chebula would vastly decrease in the dry tropical
significantly for modelling and mapping. Among the nine areas; the impact of climate change on the phenology of
variables, the Annual Mean Temperature, Isothermality and the species cannot be predicted. Most species in the trop-
Temperature Seasonality are the three variables that largely ics show variable capacity to adjust physiologically to tem-
determine the distribution of T. chebula. perature changes. Shifts in phenology could lead to tropic
Our data adequately covered the entire distributional mismatches between pollinators and dispersal agents and
range of the species across India, China, Pakistan, Sri Lanka, could have cascading effect on the species (Sheldon 2019).
Nepal and other south Asian countries. The predictive accu- Terminalia chebula is one of the tropical forest species
racy of the distributional models are estimated using the in high trade that has been listed as a priority species that
AUC (Lobo et al. 2008). The high AUC values in the present needs management intervention (Goraya and Ved 2017).
study for training and testing (0.884) indicate that the niche Considering that the suitable area for this species would
model has a good ability to differentiate between presence vastly decrease, various countries in which this tree is dis-
and absence areas for the species. AUC value (> 0.8) exhib- tributed should further strengthen the governance to provide
its that the model performance has been satisfactory and that adequate protection to T. chebula. Within India, both the
the prediction of the species distribution has been accom- RCP 4.5 as well as RCP 8.5 predict decrease in the overall
plished with higher discrimination of the input data (Lobo distribution especially in Southern India (Fig. 2). However,
et al. 2008). The Receiver Operating Characteristic (ROC) in the northern regions, more areas are likely to become suit-
curve summarizes the model performance over all condi- able areas especially in the state of Uttar Pradesh. In most of
tions. The niche model successfully predicted the distribu- Asia, it is predicted that climate change would have drastic
tion of the species in the Indian sub-continent. The species impact with temperatures set to increase by 7.5 °C (Xu et al.
has its core distribution in Southern and Central India, where 2020). This increase in temperatures especially in Central
the model predicts very high suitability. These areas could India could lead to shift in suitable habitats towards northern
be ear-marked for conservation of the species and population and more cooler latitudes. The results predicted here need
of the species in this region could be enriched. to be assessed in the background that a number of other
The overall distribution of T. chebula under different RCP variables have not been taken into account. For example,
scenarios shows a drastic decline indicating that it is highly we have not considered the possible change in the land-use
vulnerable to climate change. The mean annual tempera- categories. Any regions that are predicted to be highly suit-
ture, temperature seasonality and isothermality seem to be able here may not be actually available for the species if the
the most important variables determining the distribution of land is utilised for other purposes. However, our study has
the species which is directly influenced by climate change. identified areas both with India as well as across South Asia,
Lower degree of prediction in the future (Fig. 5, Table 3) regions that are likely to become highly suitable or excellent
also hints at the possibility of a niche shift for the species habitats in the future. These regions could be overlaid with
in future, which needs to be studied further. The study also other layers of information such as threat status, vegetation
found that there is a significant loss of suitable habitat in all type and levels of protection. Based on that, suitable areas
the countries in which T. chebula is distributed. can be considered for the conservation of the species.
The impacts of climate change depends on species vul- The study also provides geographical perspective of the
nerability to change, which is expected to be far greater in distribution of species across wide landscapes and habitats
the tropics (Deutsch et al. 2008). Tropical regions with less both in the present as well as under different scenarios in
than 1 °C of warming, are projected to experience extreme future. Based on this information, decision could be taken
conditions much earlier than rest of the world (Beaumont for conservation of the species as well as for collection
et al. 2011). Climate change is known to impact phenol- of germplasm of the species. Considering the economic
ogy including leafing, flowering, and fruiting in seasonally and medicinal potential of the species, it is inevitable that
dry tropical areas; however the magnitude and direction of large-scale cultivation of the species need to be taken up

13
Tropical Ecology

and establishing the arboreta with diverse germplasm gains a detail conservation action plan. Based on the niche maps,
importance. conservation, as well as cultivation of the species, can be
It is necessary to intensify the efforts to conserve (in- taken up.
situ and ex-situ conservation) and cultivate the T. chebula
to prevent it from further depletion. We also recommend Supplementary Information The online version contains supplemen-
tary material available at https://d​ oi.o​ rg/1​ 0.1​ 007/s​ 42965-0​ 22-0​ 0237-x.
in-situ conservation by enriching their densities in nature
reserves as well as ex-situ conservation through medicinal Acknowledgements We sincerely thank Dr. K.N. Ganeshaiah for
plant gardens, in botanical gardens and arboreta especially providing us the co-ordinates of T. chebula of entire Western Ghats,
in Southern and Central India, where the model predicts as Dr. M. Sanjappa for giving us access to BSI Herbariums at Kolkota,
the excellent niche for the species. These steps would also Dehradun, Pune and Coimbatore, we also thank Harisha, Vikram and
Arunima for providing co-ordinates of T. chebula from their field sites.
reduce the supply constraints as well as meet the industrial SS and GR acknowledge the support of DBT for undertaking the study
and the domestic demand. (BT/PR29859/FCB/125/23/2018). The authors thank the anonymous
The species also requires a stringent assessment of pop- reviewers for the detailed comments which has helped in vastly improv-
ulation size and the threatened status. The tropical region ing the manuscript.
within the country where the species is distributed also
Author contributions KBR, SS, GR & KK led the manuscript writing
offers diverse niches for the species and provides an oppor- with inputs from all the authors. KBR, gathered data, BC developed
tunity to assess the role of biotic factors that determine its methodology for modelling, KBR and BC performed analysis. GR
distribution (Dash et al. 2020). Efforts should also be made revised the manuscripts with inputs from all the authors. All the authors
among local farmers in various countries for their com- gave final approval for publication.
mercial cultivation. The study provides a useful means for
identifying species hotspots for prioritizing conservation.
The ecological niche maps generated in this study could be
overlaid with other factors such as demographic data, threat
status etc. so that a comprehensive map for the conserva-
tion of the species could be arrived at. Similarly, different
References
measures such as conservation status, risk factors, economic Adhikari D, Reshi DBK, Samant SS, Chettri A, Upadhaya K, Shah MA,
values, expert opinion and measure of species preference Singh PP, Tiwary R, Majumdar K, Pradhan A, Thakur ML, Salam
by the communities can be used as weights to evaluate the N, Zahoor Z, Mir MH, Kaloo ZA, Barik SK (2018) Inventory
worth of a given habitat for conservation and sustainable and characterization of new populations through ecological niche
modelling improve threat assessment. Curr Sci 114(3):519–531
management. Ahmad R, Khuroo AA, Hamid M et al (2019) Predicting inva-
sion potential and niche dynamics of Parthenium hysteropho-
rus (Congress grass) in India under projected climate change.
Conclusions Biodivers Conserv 28:2319–2344. https:// ​ d oi. ​ o rg/ ​ 1 0. ​ 1 007/​
s10531-​019-​01775-y
Aiello-Lammens MA, Boria RA, Radosavljevic A, Vilela B, Anderson
In this study, we used present-day occurrence data to pre- RP (2015) spThin: an R package for spatial thinning of species
dict the potentially suitable distribution area of T. chebula in occurrence records for use in ecological niche models. Ecography
Asia based on an optimized Maxent model under different 38:541–545
Akaike H (1974) A new look at the statistical model identification.
climate change scenarios. IEEE Trans Automat Contr 19:716–723
Our study indicated that the most suitable habitat for Akbar S (2020) Terminalia chebula Retz. (Combretaceae). In: Hand-
T. chebula is found to be in the Central and South India book of 200 Medicinal Plants. Springer, Cham., pp 1779–1793
(largely across Western Ghats). While there is not much Akerele O (1993) Nature’s medicinal bounty: don’t throw it away.
World Health Forum 14(4):390–395
change in the distribution of the species in RCP 2.6 sce- Annotated checklist of Flowering Plants of Nepal. http://​www.​eflor​as.​
nario, the RCP 4.5 scenario indicated that there would org/​flora​taxon.​aspx?​flora_​id=​110&​taxon_​id=​20001​4747
be decline in the suitable habitats in Central India and Aranya M, Pensak J, Toshihiro A, Worapaka M, Jiradey M (2010)
more suitable areas in northern India especially in the In vitro anti-aging activities of Terminalia chebula gall extract.
Pharma Biol 48(4):469–481. https://d​ oi.o​ rg/1​ 0.3​ 109/1​ 38802​ 0090​
state of Uttar Pradesh. Similarly, the RCP 8.5 scenario 35862​86
also indicated overall decrease in the habitat suitability of Arshad MA, Mir AK, Mushtaq A, Muhammad Z (2010) Herbal medi-
the species. The results also indicated that the predicted cines used to cure various ailments by the inhabitats of Abbot-
potential suitable distribution area of T. chebula of the tabad district, North West Frontier Province, Pakistan. Indian J
Tradit Knowl 91(1):175–183
present day matched the actual occurrence. Overall, the Bag A, Bhattacharyya SK, Chattopadhyay RR (2013) The development
future climate change scenarios indicated that the distribu- of Terminalia chebula Retz. (Combretaceae) in clinical research.
tion of T. chebula is likely to decrease. Future steps need Asian Pac J Trop Biomed 3(3): 244–252. https://​doi.​org/​10.​1016/​
to include data on demography, its threat status to develop S2221-​1691(13)​60059-3

13
 Tropical Ecology

Beaumont LJ, Pitman A, Perkins S, Zimmermann NE, Yoccoz NG, Hijmans RJ (2012) Cross-validation of species distribution models:
Thuiller W (2011) Impacts of climate change on the world’s removing spatial sorting bias and calibration with a null model.
most exceptional ecoregions. Proc Natl Ac Sci 108:2306–2311 Ecology 93:679–688
Bohl CL, Kass JM, Anderson RP (2019) A new null model approach Hijmans RJ et al (2000) Assessing the geographic representativeness of
to quantify performance and significance for ecological niche genebank collections: the case of Bolivian wild potatoes. Conserv
models of species distributions. J Biogeogr 46(6):1101–1111 Biol 14:1755–1765
Bonett DG, Wright TA (2000) Sample size requirements for estimat- Hirsch T (2010) Global Biodiversity Outlook 3; Secretariat of the Con-
ing Pearson, Kendall and Spearman correlations. Psychometrika vention on Biological Diversity: Montreal, QC, Canada.
65:23–28. https://​doi.​org/​10.​1007/​BF022​94183 Hirzel A, Guisan A (2002) Which is the optimal sampling strategy for
Catalogue of Life. https://​www.​catal​ogueo​flife.​org/​data/​taxon/​55G44 habitat suitability modelling. Ecol Modell 157:331–341
Chander J, Chauhan S. (2014). Current Status of Management of Joshi M, Charles B, Ravikanth G, Aravind NA (2017) Assigning con-
Harar ( Terminalia chebula Retz.) in Shivalik Hills. Journal of servation value and identifying hotspots of endemic rattan diver-
Krishi Vigyan. 3: 13. https://​doi.​org/​10.​5958/​2349-​4433.​2014.​ sity in the Western Ghats, India. Plant Divers 39:263–272
01241.0 Kadmon R et al (2004) Effect of roadside bias on the accuracy of
Climate Change and Non Timber Forest Produce Roles and Potential predictive maps produced by bioclimatic models. Ecol Appl
in Jharkhand, Chhattisgarh and Odisha. (2015–16) Indo Global 14:401–413
Social Service Society (IGSSS), pp 7–9 Kumar A, Kumar A, Dibyendu A, Ravikanth G, Purabi S, Khan L
Chattopadhyay RR, Bhattacharyya SK (2007) Species composition (2020) Ecological niche modelling for assessing potential distribu-
and stand structure of natural forest, timber-harvested forest and tion of Pterocarpus marsupium Roxb. in Ranchi, Eastern India.
degraded forest in the Bago Yoma region of Myanmar. J Korean Ecol Res 35(6):105–1105. https://​doi.​org/​10.​1111/​1440-​1703.​
For Soc 96:572–579 12176
Climate Change IPCC (2013) The physical science basis: contribu- Li R, Xu M, Wong MHG, Qiu S, Sheng Q, Li X et al (2015) Climate
tion of working group i to the fifth assessment report of the change-induced decline in bamboo habitats and species diver-
intergovernmental panel on climate change. Cambridge Uni- sity: implications for giant panda conservation. Divers Distrib
versity Press, Cambridge 21:379–391
Dash J, Behera MD, Jeganathan C, Jha CS, Sharma S, Lucas R, Lobo JM, Alberto JV, Raimundo R (2008) AUC: a mislead- ing meas-
Khuroo AA, Harris A, Atkinson PM, Boyd DS, Singh CP (2020) ure of the performance of predictive distribution models. Glob
India’s contribution to mitigating the impacts of climate change Ecol 17:145–151
through vegetation management. Trop Ecol 61(1):168–171 Loveridge R, Eang S, Eames JC (2018) Attracting white-shouldered
Database of Native Plants of Taiwan: http://​w ww.​h ast.​b iodiv.​t w/​ ibises to safe roosting sites in Siem Pang Kang Lech Wildlife
Taxon/​speci​esDet​ailE.​aspx?​speci​esID=​4547 Sanctuary, Cambodia. Cambodian J Nat Hist pp 58–62
Deutsch CA, Tewksbury JJ, Huey RB, Sheldon KS, Ghalambor CK lUCN Commission on National Parks and Protected Areas. lUCN
et al (2008) Impacts of climate warming on terrestrial ecto- Directory of South Asian. Digitized by the internet archive in
therms across latitude. Proc Natl Ac Sci 105(18):6668–6672 2010 with funding from UNEP-WCIVIC, Cambridge Protected
Dormann CF, Elith J, Bacher S, Buchmann C, Carl G (2013) Col- Areas, pp 39,207
linearity: a review of methods to deal with it and a simulation Mahyar UW, Burley JS, Gyllenhaal C, Soejarto DD (1991) Medicinal
study evaluating their performance. Echography 36:27–46 plants of Seberida (Riau Province, Sumatra, Indonesia). J Ethnop-
Elith J, Leathwick JR (2009) Species distribution models: ecological harmacol 31(2):217–237
explanation and prediction across space and time. Annu Rev Malhotra KC, Bhatacharya P (2010) Forest and Livelihood. Publ, CES,
Ecol Evol Syst 40:677–697 Hyderabad, p 246
Elith J, Graham CH, Anderson RP et al (2006) Novel methods Mark C, Vanderwel DW, Purves, (2014) How do disturbances and envi-
improve prediction of species distributions from occurrence ronmental heterogeneity affect the pace of forest distribution shifts
data. Ecography 29: 129–151 under climate change? Ecography 37(1):10–20
Fernandez-Going BM, Anacker BL, Harrison SP (2012) Tempo- Mohammad HK, Nur H, Mohammad MR, Mohammad AR, Jakia AK,
ral variability in California grasslands: soil type and species Nazia TH, Mohammad RQB, Sadia MM, Rownak J, Mohammed
functional traits mediate response to precipitation. Ecology R (2014) A survey of medicinal plants used by the Deb Barma
93(9):2104–2114 clan of the Tripura tribe of Moulvibazar district, Bangladesh. J
Flora of Pakistan. http://w
​ ww.e​ flora​ s.o​ rg/fl
​ orat​ axon.a​ spx?fl
​ ora_i​ d=5​ &​ Ethnobiol Ethnomedicine, pp 10–19
taxon_​id=​20001​4747 Muhammad S, Khan BA, Akhtar N, Mahmood T, Rasul A, Hussain I,
Ghareghan F, Ghanbarian G, Pourghasemi HR, Safaeian R (2020) Pre- Khan H, Badshah A (2012) The morphology, extractions, chemi-
diction of habitat suitability of Morina persica L. species using cal constituents and uses of Terminalia chebula: A review. J Med
artificial intelligence techniques. Ecol Indica 112:106096 Plants Res 6(33):4772–4775
Goraya GS, Ved DK (2017) Medicinal Plants in India: An Assessment Natuhara Y, Imanishi A, Kanzaki M et al (2012) Uses of trees in paddy
of their Demand and Supply. National Medicinal Plants Board, fields in Champasak Province, southern Lao PDR. Landsc Ecol
Ministry of AYUSH, Government of India, New Delhi and Indian Eng 8:115–122. https://​doi.​org/​10.​1007/​s11355-​011-​0172-1
Council of Forestry Research & Education, Dehradun. Nguyen QV, Nguyen VB, Eun JB, Wang SL, Nguyen DH, Tran TN,
Graham CH et al (2004) New developments in museum-based infor- Nguyen AD (2016) Anti-oxidant and antidiabetic effect of some
matics and applications in biodiversity analysis. Trends Ecol Evol medicinal plants belong to Terminalia species collected in Dak
19:497–503 Lak Province, Vietnam. Res Chem Intermed 42(6):5859–5871
Gupta PC (2012) Biological and pharmacological properties of Ter- Nicholson DI (1965) A study of virgin forest near Sandakan North Bor-
minalia chebula Retz. (Haritaki)-an overview. Int J Pharm Pharm neo. In Pmc. Symp. on ecological research in humid tropics veg-
Sci 4(3):62–68 etation. Kuching: UNESCO. Government of Sarawak, pp 67–86.
Hamid M, Khuroo AA, Charles B et al (2019) Impact of climate Nigam M, Mishra AP, Adhikari-Devkota A, Dirar AI, Hassan MM,
change on the distribution range and niche dynamics of Himalayan Adhikari A, Belwal T, Devkota HP (2020) Fruits of Terminalia
birch, a typical treeline species in Himalayas. Biodiver Conserv chebula Retz.: A review on traditional uses, bioactive chemical
28:2345–2370. https://​doi.​org/​10.​1007/​s10531-​018-​1641-8

13
Tropical Ecology

constituents and pharmacological activities. Phytother Res 34(10): Rathinamoorthy R, Thilagavathi G (2014) Department of fashion tech-
2518–2533. https://​doi.​org/​10.​1002/​ptr.​6702 nology, PSG college of technology, Coimbatore—641004, Tamil
Nishteswar K (2014) Depleting medicinal plant resources: a threat Nadu. India Int J Pharmtech Res 6(1):97–116
for survival of Ayurveda. AYU 35:349–350 Raut B, Khanal DP, Kharel A, Raut B (2018) Traditional Healing
Working Paper No. APFSOS II/WP/2009/29 (2009) Asia-Pacific Practice in Rajbanshi and Satar community of Jhapa, Nepal.
forestry sector outlook study II (APFSOS II). JMMIHS 4(1):103–116
Ortega-Huerta MA, Peterson AT (2008) Modeling ecological niches Reddy S, Dávalos L (2003) Geographical sampling bias and its
and predicting geographic distributions: a test of six presence- implications for conservation priorities in Africa. J Biogeogr
only methods. Rev Mexican Biodiversity 79:205–216 30: 1719–1727
Parmesan C, Yohe G (2003) A globally coherent fingerprint Rundel (1999) Forest Habitats and Flora in Laos PDR, Cambodia,
of climate change impacts across natural systems. Nature and Vietnam.
421(6918):37–42 Saatchi SS, Harris NL, Brown S, Lefsky M, Mitchard ET, Salas
Patrick D (2009) Forest Department Government of Sri Lanka, Sri W, Zutta BR, Buermann W, Lewis SL, Hagen S, Petrova S,
Lanka forestry outlook study. Asia-Pacific Forestry Sector Out- White L, Silman M, Morel A (2011) Benchmark map of forest
look Study APFSOS II:6–24 carbon stocks in tropical regions across three continents. Proc
Pearce W, Holmberg K, Hellsten I, Nerlich B (2014) Climate change Natl Acad Sci 108:9899–9904. https://​doi.​org/​10.​1073/​pnas.​
on twitter: topics, communities and conversations about the 10195​76108
2013 IPCC working group 1 report. PLoS ONE 9(4):e94785. Sanjeewa TABD, Pushpakumara DKNG, Sangakkara UR (2013)
https://​doi.​org/​10.​1371/​journ​al.​pone.​00947​85 Morphological Characterization of Terminalia chebula Retz. Sri
Philip W R (1999) Forest Habitats and Flora in Lao PDR, Cambodia, Lanka Trop Agri Res 25(1):127–132
and Vietnam. Conservation Priorities in Indochina - WWF Desk Sarma B, Baruah PS, Tanti (2018) Habitat distribution modeling for
Study. Prepared for World Wide Fund for Nature, pp 1–171 reintroduction and conservation of Aristolochia indica L.—a
Phillips SJ, Anderson RP, Schapire RE (2006) Maximum entropy threatened medicinal plant in Assam, India. J Threat Taxa
modelling of species geographic distributions. Ecol Modell 10(11):12531–12537. https://​doi.​org/​10.​11609/​jott.​3600.​10.​11.​
190:231–259 12531-​12537
Phillips SJ, Anderson RP, Dudík M, Schapire RE, Blair ME (2017) Saru G, Bipana B, Kabita S, Bijaya D, Deepak G (2020) Common
Opening the black box: an open-source release of Maxent. medicinal plants of Nepal: A review of Triphala: Harro (Termi-
Ecography 40: 887–893 nalia chebula), Barro (Terminalia bellirica), and Amala (Emblica
Phurpa W, Karma Y, Kinga J (2017) Pharmacological, ethnophar- officinalis). Asian J Pharmacogn 4(3):5–13
macological, and botanical evaluation of subtropical medicinal Sen S, Gode A, Ramanujam S, Ravikanth G, Aravind NA (2016a)
plants of Lower Kheng region in Bhutan. Integr Med Res, pp Modeling the impact of climate change on wild Piper nigrum
382–387 (Black Pepper) in Western Ghats, India using Ecological Niche
Piao S, Qiang L, Anping C, Ivan AJ, Yongshuo F, Junhu D, Lingli L, models. J Plant Res 129(6):1033–1040. https://​doi.​org/​10.​1007/​
Xu L, Miaogen S, Xiaolin Z (2019) Plant phenology and global s10265-​016-​0859-3
climate change: Current progresses and challenges. Glob Change Sen S, Shivaprakash KN, Aravind NA, Ravikanth G, Dayanandan S
Biol 25(6):1922–1940 (2016b) Ecological niche modeling for conservation planning of
Taiwan Plant names. www.​eflor​as.​org http://​www.​eflor​as.​org/​flora​ an endemic snail in the verge of becoming a pest in cardamom
taxon.​aspx?​flora_​id=​101&​taxon_​id=​20001​4747 plantations in the Western Ghats Biodiversity hot-spot. Ecol Evol
India Biodiversity Portal. Terminalia chebula Retz. . [online] India 6(18):6510–6523. https://​doi.​org/​10.​1002/​ece3.​2368
Biodiversity Portal, Species Page. https://​india​biodi​versi​ty.​org/​ Sheldon SK (2019) Climate change in the tropics: ecological and evo-
biodiv/​speci​es/​show/​31838 lutionary responses at low latitudes. Annu Rev Ecol Evol Syst
Priti H, Aravind NA, Uma Shaanker R, Ravikanth G (2016) Modeling 50(1):303–333
impacts of future climate on the distribution of Myristicaceae spe- Shivaprakash KN, Ravikanth G, Barve N, Ghazoul J, Ganeshaiah KN,
cies in the Western Ghats, India. Ecol Eng 89:14–23 Uma Shaanker R (2013) Do ecological niche model predictions
Purna BC, Kinley T (2012) State of Forest Genetic Resources of Bhu- reflect the adaptive landscape of species? A test using Myristica
tan, Country report. RNR Research and Development Centre, malabarica Lam, an endemic tree in the Western Ghats, India.
Yusipang, pp 1–68 Plos One 8(11):e82066. https://​doi.​org/​10.​1371/​journ​al.​pone.​
Rahimian BA, Salehi H, Pourghasemi HR, Blaschke T (2019). Pre- 00820​66
dicting habitat suitability and conserving Juniperus spp. habitat Shrestha S (2019) Effects of climate change in agricultural insect pest.
using SVM and maximum entropy machine learning techniques. Acta Sci Agric 3:74–80
Water 11(10):2049 Subba B, Sen S, Ravikanth G, Nobis MP (2018) Direct modelling of
Raman S, Shameer TT, Charles B et al (2020) Habitat suitability model limited migration improves projected distributions of Himalayan
of endangered Latidens salimalii and the probable consequences amphibians under climate change. Biol Conserv 227:352–360
of global warming. Trop Ecol 61:570-582. https://​doi.​org/​10.​ Sumangala RC, Rosario S, Charles B, Ganesh D, Ravikanth G (2017)
1007/​s42965-​020-​00114-5 Identifying conservation priority sites for Saraca asoca: an
Rana SK, Rana HK, Ghimire SK, Shrestha KK, Ranjitkar S (2017) important medicinal plant using ecological niche models. Ind
Predicting the impact of climate change on the distribution of two For 143(6):531–536
threatened Himalayan medicinal plants of Liliaceae in Nepal. J Swets JA (1988) Measuring the accuracy of diagnostic systems. Sci-
Mt Sci 14(3):558–570 ence 240:1285–1293
Rana SK, Rana HK, Luo D, Sun H. (2021). Estimating climate-induced Thaung NO, Don KL (2007) PHCOG REV.: plant review Terminalia
‘Nowhere to go’range shifts of the Himalayan Incarvillea Juss. chebula. Pharmacogn Rev 1(1):151–156
Using multi-model median ensemble species distribution models. Tropicos.org. Missouri Botanical Garden (2021) http://​www.​tropi​
Ecol Indic 121:107127 cos.​org © 2021 Missouri Botanical Garden—4344 Shaw Boule-
Rare and endangered economic plants of Bangladesh, Country Com- vard—Saint Louis, Missouri 63110. https://​www.​tropi​cos.​org/​
pass. http://​www.​fao.​org/3/​Y3660e/​y3660​e05.​htm name/​82000​10

13
 Tropical Ecology

Virginia HD et al (2001) Climate change and forest disturbances. Bio- Wilpattu National Park. http://​eastb​oundg ​roup.​com/​eastn​ews-​newsl​
science 51(9):723–733 etter/​2020/​August/​Cover%​20sto​ry.​html
Warren DL, Glor RE, Turelli M (2010) ENMTools a toolbox for Xu C, Kohler TA, Lenton TM, Svenning J-C, Scheffer M (2020)
comparative studies of environmental niche models. Ecography Future of the human climate niche. Proc Natl Acad Sci
33(3):607-611. https://d​ oi.o​ rg/1​ 0.1​ 111/j.1​ 600-0​ 587.2​ 009.0​ 6142.x 117(21):11350–11355
Wilart P, Saranya W, Phansuang D, Wipanoot B, Punchavee S, Ang- Yadav U, Hugo A, Boon EK, Yadav S, Shrestha KK (2010) Indigenous
khana C, Boonthawan W (2020) HPLC Determination of the Gal- use and bio-efficacy of medicinal plants in the Rasuwa District,
lic Acid and Chebulinic Acid Contents of Phyllanthus emblica Central Nepal. J Ethnobiol Ethnomed 6(3):1–10
Linn., Terminalia bellirica Roxb., Terminalia chebula Retz. and
Triphala Products from Chae Son district, Lampang, Thailand.
Am J Food Tech 8(3):87–98

13