Skip to main content

Predictive distribution modeling of Swertia bimaculata in Darjeeling-Sikkim Eastern Himalaya using MaxEnt: current and future scenarios

Abstract

Background

As global temperatures continue to rise, species distribution modeling is a suitable tool for identifying rare and endangered species most at risk of extinction, along with tracking shifting geographical range.

Methods

The present study investigates the potential distribution of Swertia bimaculata in the Darjeeling-Sikkim region of Eastern Himalaya in current and future climate scenarios of GFDL-CM3 (Geophysical Fluid Dynamics Laboratory-Climate Model 3) for the year 2050 and year 2070 through MaxEnt presence data modeling. Two sets of variables were used for modeling current scenario. The models were evaluated using AUC (area under the curve) values and TSS (true skill statistic).

Results

Habitat assessment of the species shows low and sporadic distribution within the study area. A significant decrease is observed in the possible range of the species in the future climate scenario with the habitat decreasing from 869.48 to 0 km2. Resultant maps from the modeling process show significant upward shifting of the species range along the altitudinal gradient. Still, results should be taken with caution given the low number of occurrences used in the modeling.

Conclusions

The results thus highlight the vulnerability of the species towards extinction in the near future.

Introduction

The Intergovernmental Panel on Climate Change (IPCC) fifth assessment report (2014) reiterates that recent climate change caused by heightened concentrations of greenhouse gasses generated by anthropogenic activity consequently impacted both natural and human systems. The occurrences of climate change are easily observable from the increasing average global temperatures (Fischer and Knutti 2015), disappearances of mountainous glaciers (Roe et al. 2017), melting polar ice caps (Chen et al. 2006), increasing sea levels (Nerem et al. 2018), and desertification (Guo et al. 2017) to name a few. One of the impacts of climate change is human-driven biodiversity loss predicting the extinction of an overall 7.9% of total species on the planet (Urban 2015). Four representative concentration pathways (RCPs) have been described based on the mitigation scenarios fulfilled (IPCC 2014). The current climatic trend affects mountainous regions across the world including the Himalayas (Hamid et al. 2019), where several plant and animal species have had a notable upward altitudinal shift (Dullinger et al. 2012). Changes in climatic patterns have further complicated downstream effect on species conservation. This highlights the importance of predictive tools for quicker assessment of priority species.

Species distribution modeling (SDM) is a tool used for designing effective conservation strategies by generating predictive maps of the range of a species in a given area (Kumar and Stohlgren 2009). SDMs are often used to prioritize areas for conservation (Chunco et al. 2013) and assess the immediate and future impact of climate change under different scenarios (Ranjitkar et al. 2016). The main principle based on which most SDMs function is the correlation of presence and sometimes absence data, with different environmental variables to create a predictive map of the species range (Pearson et al. 2006). SDMs essentially estimate the geographic distribution of a species, both actual and potential which is done primarily by characterizing environmental conditions most suitable for the species and then, identifying where these environments are distributed across the geographic space on a spatial scale (Pearson 2007).

MaxEnt is one such SDM program widely used for distribution modeling both internationally, such as China (Yi et al. 2016), New-Caledonia (Kumar and Stohlgren 2009), and the USA (Mingyang et al. 2008), and also in the Indian context (Sarma et al. 2018, Rawat et al. 2017). MaxEnt has been used previously in the Himalayas for modeling several species such as Justicia adhatoda (Yang et al. 2013), Rhododendrons (Kumar 2012), and Berberis aristata (Ray et al. 2011). MaxEnt is based on the principle of maximum entropy, a machine-learning technique that uses presence (occurrence of a species) and background environmental data. The information provided by the environment and the presence records of a species limit the probable distribution of a species (Pearson 2007). It can also predict the hypothetical habitat of a species in a future scenario. Such predictions have assessed possible habitat for invasive species like Arundina graminifolia (Kolanowska and Konowalik 2014) and endangered species like Thuja sutchuenensis (Qin et al. 2017). As the expected impact on biodiversity could be high, it is thus necessary to understand the exact effect of climate change over a set period of years at the species level.

The Eastern Himalaya is one of the extensions of the Himalaya biodiversity hotspot (CEPF 2020) in the Indian subcontinent with an estimated 5800 species of plants (Pande and Arora 2014). The Darjeeling and Sikkim Himalayas, a contiguous part of the Indian Eastern Himalaya (Starkel and Sarkar 2014), harbors 14 genera under the family Gentianaceae with about 70 species (Grierson and Long 1999). Swertia bimaculata is one of the few species under genus Swertia (Gentianaceae) found in the Darjeeling-Sikkim Himalayas (Grierson and Long 1999). Worldwide, this species has been recorded primarily in East Asian and Southeast Asian countries like Japan, China, Republic of Korea, Bhutan, and India. One of the factors negatively impacting S. bimaculata population is habitat degradation (Kanade and John 2018, Pal et al. 2016). Only seventeen occurrences have been recorded previously in India (GBIF 2020). While far less in demand than the medicinal plant Swertia chirayita (Pandey et al. 2012), the value of S. bimaculata as an adulterant has resulted in the decline of the population of the species making it less common in Darjeeling-Sikkim Himalaya. Das et al. (2013) report that S. bimaculata is a species with valuable bioactive compounds, with a rapidly declining population. Hence, it is crucial to understand the distribution pattern and assess the current and future spatial range of the species for conservation measures.

Materials and methods

Location

The present study area focuses on the Darjeeling-Sikkim part of Eastern Himalaya (Fig. 1). The contiguous part of the Eastern Himalaya, the exclusively mountainous region of Darjeeling-Sikkim Himalaya, extends from 26° 31′ to 28° 10′ N Latitude and 87° 59′ to 88° 58′ E Longitude covering a total area of about 9020 km2. The region is composed of two transverse ranges Dongkya in the east and Singalila in the west with the countries Nepal, Bangladesh, Bhutan, and China bordering the region that occupies a 100-km-long segment of the highest mountain barrier on the globe (Starkel and Sarkar 2014). The altitudinal range varies from ca. 132 m at Sukna in Darjeeling to 8598 masl at Sikkim, the summit of the third highest peak of the world, Mount Kanchenjungha. A combination of factors like geographic location and climate makes the vegetation of Darjeeling-Sikkim Himalaya a meeting ground for the Indo-Malaysian and Indo-Chinese tropical lowland flora, the Sino-Himalayan east Asiatic, and the Western Himalayan flora comprising about 9000 species with a high percentage of endemic plant species (Das 1995). Due to the altitudinal variation, the area exhibits a characteristic monsoon climate, with wet summer and dry winter caused by the direct exposure to the moisture-laden southwest monsoon flowing upwards from the Bay of Bengal that lies at proximity. This region is also a part of the designated Himalayan biodiversity hotspot as selected by Conservation International and the home to many endemic species (Conservation International 2005).

Fig. 1
figure1

Map of study area (shaded portion indicates protected areas)

Swertia bimaculata

The genus Swertia of the family Gentianaceae is represented by nearly 150 species worldwide (WFO 2019) of which 40 species are distributed in India and 15 within the Eastern Himalaya (Samaddar et al. 2014). Taxonomically, S. bimaculata is identified by its ovate leaves with three distinct nerves in its vegetative state, flowers with black spots at the apex of each corolla lobe, and two usually distinct green orbicular glands in the middle of each corolla lobe (Grierson and Long 1999, Pandey et al. 2012). In India, 17 occurrences of the species have been recorded previously, with only two recorded from the eastern region (GBIF 2020).

Population assessment

The phytosociological assessment of the habitat for S. bimaculata was carried out with the help of quadrat sampling method. The quadrat size of 2 × 2 m was placed in the habitat where the S. bimaculata showed its occurrence at eleven occurrence points. The data collected for the associated species was pooled to compute and estimate frequency, density, abundance, and dominance using the following equations (Curtis and McIntosh 1950, Phillips 1959).

$$ \mathrm{Frequency}\ \left(\%\right)=\frac{\mathrm{Number}\ \mathrm{of}\ \mathrm{quadrats}\ \mathrm{in}\ \mathrm{which}\ \mathrm{the}\ \mathrm{species}\ \mathrm{occured}}{\mathrm{Total}\ \mathrm{number}\ \mathrm{of}\ \mathrm{quadrats}\ \mathrm{studied}}\times 100 $$
$$ \mathrm{Density}=\frac{\mathrm{Total}\ \mathrm{number}\ \mathrm{of}\ \mathrm{in}\mathrm{dividuals}\ \mathrm{of}\ \mathrm{a}\ \mathrm{species}\ \mathrm{in}\ \mathrm{a}\mathrm{ll}\ \mathrm{quadrats}}{\mathrm{Total}\ \mathrm{number}\ \mathrm{of}\ \mathrm{quadrats}\ \mathrm{studied}} $$
$$ \mathrm{Abundance}=\frac{\mathrm{Total}\ \mathrm{number}\ \mathrm{of}\ \mathrm{in}\mathrm{dividuals}\ \mathrm{of}\ \mathrm{a}\ \mathrm{species}\ \mathrm{in}\ \mathrm{a}\mathrm{ll}\ \mathrm{quadrats}}{\mathrm{Total}\ \mathrm{number}\ \mathrm{of}\ \mathrm{quadrats}\ \mathrm{in}\ \mathrm{which}\ \mathrm{the}\ \mathrm{species}\ \mathrm{occured}} $$

The relative values were summed up using IVI =  ∑ [RF + RD + RA] to extract the importance value index (IVI) for each taxa associated in the habitat. Several different diversity indices for the associated species in the habitat of S. bimaculata were estimated using PAST version 3.24, namely, Shannon index H ′  =  −  ∑ [(ni/N) ln(ni/N)] (Shannon and Weaver 1963); Richness index \( \boldsymbol{D}=S/\sqrt{N} \) (Menhinick 1964); Evenness index J = H ′ / ln S (Pielou 1966); and Index of dominance CD =  ∑ (ni/N) (Simpson 1949). The spatial distribution pattern of the species was also studied (Whitford 1949).

Species occurrence data

The occurrence data was collected for S. bimaculata through field investigation, and the coordinate data for the species was recorded using Garmin eTrexH. The population count for S. bimaculata was done with habitat assessment through studying the associated species in the ecological niche. Occurrence points for Sikkim Himalayas were also consulted (Das et al. 2013). Simultaneously, the coordinates were taken in the field for climate change modeling, which was used to determine the IUCN threat status of the species in the study area using GeoCAT (Bachman et al. 2011). During modeling, however, the two coordinates obtained from GBIF (GBIF 2020) were excluded as while they were within India, they were outside the study area. The coordinates obtained from field visit and literature review were combined and used and are presented in Table 1.

Table 1 Species occurrence data

Environmental variables

Environmental variables (Table 2) for the current and future climate condition were sourced from the WorldClim database (Hijmans et al. 2005). The current climate data were obtained from the WorldClim database version 2.0 at ~1 km2 (30 arc second) resolution. In case of future climate change predictions, the nineteen bioclimatic variables were obtained for RCP 2.6, RCP 4.5, and RCP 8.5 for the year 2050 and the year 2070 based on the global climatic model GFDL-CM3 (Griffies et al. 2011, Chaturvedi et al. 2012). The data obtained were first clipped and then converted to the ASCII file format by using QGIS 3.4 Madeira. The elevation data were obtained from the Global Multi Resolution Terrain Elevation Data in 1-km2 resolution (Danielson and Gesch 2011), and the data was used to generate slope and aspect layers in ASCII format in QGIS 3.4 Madeira.

Table 2 Environmental variables used for species distribution modeling in MaxEnt

Modeling procedure

The modeling was done using MaxEnt algorithm version 3.4.1 (Phillips et al. 2006). First, a complete model was run with all 22 bioclimatic variables with default settings applied except for using linear, quadratic, and hinge features as the number of data points was 16, along with 10 percentile training presence threshold rule and replicated 10 times. ENM Tools 1.3 (Warren et al. 2010) was used to perform multi-collinearity test to eliminate highly correlated variables, i.e., any correlation where the value of r was greater than 0.9 (Jueterbock et al. 2016), and the second model was run using nine bioclimatic variables. Both models were generated by setting the random test percentage to 30% (Thapa et al. 2018). For future climate predictions, the model was run using the nine uncorrelated bioclimatic variables and the occurrence points first calibrated with the current environment layers and then projected on future climate layers for RCP 2.6, RCP 4.5, and RCP 8.5 for both years 2050 and 2070. The variables used for each model are given in Table 3. The quality of all models was assessed using the average AUC (area under the curve) values and are recorded in Table 5. Models were also evaluated by calculating true skill statistic (TSS) (Allouche et al. 2006). By default, MaxEnt performs multivariate environmental similarity surface (MESS) analysis, which was also used (Elith et al. 2010).

Table 3 Species distribution models

Mapping

Mapping was prepared in the QGIS 3.4 Madeira software, and for each of the generated maps for each model, the corresponding 10th percentile training presence was set as the threshold in order to remove areas with low probability of occurrence (Young et al. 2011). Area was calculated from the generated maps using the QGIS 3.4 Madeira software.

Results

Population assessment

The surveyed habitat revealed that the population of S. bimaculata (Fig. 2) is quite low with only 36 individuals (n=36) counted across eleven suitable habitats in the Darjeeling region. The lowest count of individual S. bimaculata was one whereas the maximum count within a quadrat was observed to be nine individuals. The estimated density was 3.27 with abundance score of 3.3. The abundance to frequency ratio was found to be 0.03, indicating random distribution of the taxa across the habitat.

Fig. 2
figure2

Swertia bimaculata in its natural habitat

The phytosociological assessment of the habitat of S. bimaculata showed the occurrence of tree species like Acer campbellii, Acer sikkimense, Acer thomsonii, Cryptomeria japonica, Eriobotrya dubia, Eriobotrya petiolata, Eurya acuminata, Exbucklandia populnea, Macaranga denticulata, Neolitsea impressa, and species of Symplocos. The estimation of different phytosociological parameters in recognized habitat of S. bimaculata revealed 45 species of shrubs and herbs that belonged to 38 genera and 27 families with a total of 183 individuals in eleven recognized sites from Darjeeling Himalaya. The family Rosaceae showed highest number with 8 species under 4 genera followed by Compositae and Urticaceae with 3 species each. Acanthaceae, Araliaceae, Balsaminaceae, Hydrangeaceae, Primulaceae, and a fern family Pteridaceae showed 2 species each within the habitat.

The total density calculated for the shrubs and herbs was 16.64 with highest density of 1.45 expressed by Melissa axillaris and Oplismenus compositus. The total abundance of taxa associated with S. bimaculata in its niche was estimated to be 134.6 with highest abundance expressed by Melissa axillaris. The score for importance value index for the shrubs and herb taxa within the habitat ranged between 2.85 and 22.19 with highest score for Melissa axillaris and lowest for Helwingia himalaica, Lysimachia japonica, Maesa chisia, Osbeckia melastomacea, Rubus splendidissimus, Smilax aspericaulis, and Viburnum mullaha as recorded in Table 4.

Table 4 Phytosociological status of the associated species in S. bimaculata habitat

The estimation for the diversity indices showed that the Shannon index was 3.52, the richness index was estimated as 3.32, the evenness index as 0.92, and the Simpson’s index of dominance was recorded to be 0.96 for the associated species within the habitat.

Species distribution modeling

The predicted distribution of S. bimaculata for the complete model (all 22 variables) and the uncorrelated model (9 uncorrelated variables) is shown (Fig. 3).

Fig. 3
figure3

Potential distribution of Swertia bimaculata under current climate: a complete and b uncorrelated models

Both of these models performed well in MaxEnt, since for each, the AUC values were higher than 0.9, with the mean AUC values being 0.982 and 0.980 for the complete and the uncorrelated model respectively (Table 5). The TSS value was approximately 0.78 and 0.64 for the complete and uncorrelated model respectively (Table 5). For each of the complete and uncorrelated model, the spatial distribution of the species in the study area is 3.92% (451.89 km2) and 7.55% (869.48 km2) respectively (Table 5). In the complete model, the percentage of contribution was highest for BIO13 (precipitation of wettest month) at 39.1%. In the uncorrelated model, BIO18 (precipitation of warmest quarter) showed the highest contribution at 40.2% (Table 5).

Table 5 Prediction accuracy of Swertia bimaculata distribution models

Figure 4 depicts the ROC curves along with the AUC value. Overall, both the complete and uncorrelated models show very similar curves and AUC values. Figure 5 illustrates the Jackknife of the complete and the uncorrelated model.

Fig. 4
figure4

ROC curve of Swertia bimaculata under current climate: a complete and b uncorrelated

Fig. 5
figure5

Jackknife of regularized training gain for Swertia bimaculata under current climate: a complete model and b uncorrelated model

The potential spatial distribution of each generated model for future distribution in RCP 2.6, RCP 4.5, and RCP 8.5 in the year 2050 and year 2070 are shown in Fig. 6. The percentage of the study area where S. bimaculata occurs after applying the threshold is recorded in Table 5. For each of the six future predictions, the AUC value ranged from 0.982 to 0.985 making all six models well-performing. The TSS value ranged from approximately 0.63 to 0.79. Among the future models, the probable area of occurrence for S. bimaculata is highest for the year 2050 with 2.07% (238.38 km2) in RCP 4.5 and the lowest estimated area is 0 (0 km2) for the year 2070 in RCP 8.5 (Fig. 6).

Fig. 6
figure6

Potential distribution of Swertia bimaculata under future climate: RCP 2.6, a year 2050 and b year 2070; RCP 4.5, c year 2050 and d year 2070; RCP 8.5, e year 2050 and f year 2070

Figure 7 illustrates the ROC curves of the six future models with the graphs being very similar. The corresponding AUC values are given in Table 5. Figure 8 shows the jackknife of all future models computed. Here, elevation (elev) is the most influential variable while the coldest quarter precipitation (bio_19) is the most influential bioclimatic variable. Figure 9 shows the result of MESS analysis of the species for all future models. Figure 10 shows the response of the species to the different bioclimatic variables.

Fig. 7
figure7

ROC curve of Swertia bimaculata under future climate: RCP 2.6, a year 2050 and b year 2070; RCP 4.5, c year 2050 and d year 2070; RCP 8.5, e year 2050 and f year 2070

Fig. 8
figure8

Jackknife of regularized training gain of Swertia bimaculata under future climate: a RCP 2.6 in year 2050, b RCP 2.6 in year 2070, c RCP 4.5 in year 2050, d RCP 4.5 in year 2070, e RCP 8.5 in year 2050, and f RCP 8.5 in year 2070 using only the nine uncorrelated variables

Fig. 9
figure9

MESS analysis of Swertia bimaculata under future climate: a RCP 2.6 in year 2050, b RCP 2.6 in year 2070, c RCP 4.5 in year 2050, d RCP 4.5 in year 2070, e RCP 8.5 in year 2050, and f RCP 8.5 in year 2070

Fig. 10
figure10

Response curves of Swertia bimaculata to variables: a asp (aspect), b elev (elevation), c slope (slope), d bio03 (isothermality), e bio07 (temperature annual range), f bio14 (precipitation of driest month), g bio15 (precipitation of seasonality), h bio18 (precipitation of warmest quarter), and i bio19 (precipitation of coldest quarter)

Discussion

The present study explores both the habitat assessment of S. bimaculata and its spatial distribution in the present and future climate change scenarios. Only 36 individual plants were recorded with abundance to frequency ratio of 0.05 indicating random distribution in the study area. Since the individual count of S. bimaculata was observed to be low in number, the species was modeled under the present and projected under the predicted future climate conditions in Darjeeling-Sikkim Himalaya. The species most dominant in the habitat of S. bimaculata were Melissa axillaris and a grass species Oplismenus compositus.

As S. bimaculata is considered endangered by area of occupancy in India (Bachman et al. 2011), the MaxEnt algorithm was used to assess habitat suitability of S. bimaculata using two different suites of environmental variables. Such species distribution modeling is guided by both spatial coordinates and ecological variables in MaxEnt (Phillips and Dudík 2008). Preceding studies involving species distribution modeling in India has explored medicinally significant species such as Berberis aristata (Ray et al. 2011), Justicia adhatoda (Yang et al. 2013), and Brucea mollis (Borthakur et al. 2018); endangered species such as Gymnocladus assamicus (Menon et al. 2010), Ilex khasiana (Adhikari et al. 2012), and Neottia cordata (Tsiftsis et al. 2019); invasive species such as Hyptis suaveolens (Padalia et al. 2014); and the bamboo Yushania maling (Srivastava et al. 2018) along with tree species like Fagus sylvatica (Castaño-Santamaría et al. 2019). In the present study, all the models had AUC values higher than 0.9. Similar AUC values are reported for Vincetoxicum arnottianum (Khanum et al. 2013), Hyptis suaveolens (Padalia et al. 2014), Scutellaria baicalensis (Zhang et al. 2016), and Oxytenanthera abyssinica (Gebrewahid et al. 2020). TSS value is an alternative measure of model performance (Allouche et al. 2006). The TSS values ranged from 0.6 to 0.8 for all models, indicating that the models are not very robust.

The nine variables used in the uncorrelated model were utilized for future climate change projections of RCP 2.6, RCP 4.5, and RCP 8.5 for both the year 2050 and the year 2070. It should be noted that the threshold used to calculate area was 10th percentile training presence, which removes areas with very low probability of species occurrence. In RCP 8.5, no mitigation strategies are applied, and the temperature is expected to rise to 2.0 °C above the pre-industrial level in the year 2050 and 3.7 °C in the year 2070 (IPCC 2014). In such a scenario, suitable habitat disappears completely in the year 2070 (Table 5). In RCP 4.5, moderate mitigation strategies are adopted, and the temperature increases to 1.4 °C in the year 2050 and 1.8 °C in the year 2070 (IPCC 2014). The percentage of area for the year 2050 is 2.07%, and in the year 2070, the area is 1.28%. It is possible that due to moderate mitigation, warming results in opening up of previously colder regions. In RCP 2.6, all expected climate change alleviation goals have been fulfilled, and temperature increases by 1.0°C above pre-industrial levels in the year 2050 and does not further increase in the year 2070 (IPCC 2014). The percentage of area in 2050 for RCP 2.6 is 0.73%, and it decreases to 0.03% in the year 2070. The smaller increase of global temperature may mean that areas of higher elevations remain unsuitable while other bioclimatic variables are influenced enough to decrease suitable habitat area. Overall, a similar trend is observed across all three RCPs wherein suitable habitat decreases with time. It was observed that the largest area for the year 2050 and the year 2070 was for RCP 4.5. This could be due to the warming climate, which lets the species spread towards higher altitudes, but more elevated temperatures of RCP 8.5 result in a restricted range for the species. Studies conducted by Parolo and Rossi (2008) and Matteodo et al. (2013) demonstrate that changing climate pushes plants towards higher elevations and latitudes. Similarly, S. bimaculata migrates northwards.

Overall, elevation had the most influence on species distribution with it having the highest training gain among all the variables both by itself and by its absence regardless of the combination of variables used. In MESS analysis of future models, the similarity of the future and the current bioclimatic variables are compared. Darker areas show that future data is outside the range of the current data. The species response curves demonstrate the response of S. bimaculata to each variable, particularly altitude (peaking at 2000 m AMSL), temperature (between 18 and 34 °C) and precipitation. Climate change therefore perceptibly affects the spatial range of S. bimaculata in the future with the range first decreasing in the year 2050 and then further declining in the year 2070 (Remya et al. 2015).

Samaddar et al. (2013) reports the presence of high concentrations of bioactive compounds such as swertiamarin and amarogentin, which makes this species further vulnerable to exploitation. The species of Swertia are traditionally precious in this part of the Himalaya, and S. chirayita, a critically endangered plant of the Himalaya, has numerous medicinal properties including anti-cancerous potential (Joshi and Dhawan 2005, Saha et al. 2004). Yonzone (2017) reports the use of S. bimaculata mainly as a substitute for S. chirayita. A decoction of either fresh or dried plant mass S. chirayita is orally administered to treat common ailments. The collection of S. chirayita has extensively reduced the population of the species, and therefore, as its substituent, S. bimaculata, locally known as “Bhaley chirauto,” is used widely in this part of the Himalaya. Thus, overharvesting coupled with rapid urban development in the Darjeeling-Sikkim Himalaya makes this species a vulnerable one. Due to such activity, the population of S. bimaculata has reduced extensively with sporadic distribution and decline in population in natural vegetation. A decrease in the spatial range of a species increases the potential risk of local extinction (Thomas et al. 2004). Climate change is another hazard to this species as it can lead to further shrinkage of the potential habitat. Human-induced climate change is directly affecting the biological species from genes to ecosystems (Hoffmann et al. 2019). Active conservation efforts and management strategies are utmost necessary at this point to preserve the species in situ as the population of the species is seemingly declining at an alarming rate.

While MaxEnt is an efficient tool for modeling endangered species (Gebrewahid et al. 2020), the accuracy is limited by the methodology and data applied during modeling. For instance, the collection procedure of the occurrence data alongside the low number of occurrence points can make the modeling approach biased. Thus, it is necessary to use correct and reliable habitat models to reduce the biasness (Sobek-Swant et al. 2012). However, the models generated in this study provide an insight to the possible distribution of Swertia bimaculata in present and future.

Conclusion

In the present study, the MaxEnt algorithm predicted the presence of suitable habitat for S. bimaculata in both present and future scenarios. With the changing climate depending on the climate mitigation strategies, the adopted area of suitable habitat for S. bimaculata continues to decrease. Coupled with its sporadic distribution, low population count, and its use in traditional medicine, this species would require aggressive conservation strategies in order to prevent its probable extinction in the future. Such conservation strategies would require collaboration between the requisite government agencies, forest departments, and research institutes including active participation of the locals with a variety of measures employed.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Abbreviations

IPCC:

Intergovernmental panel on climate change

RCP:

Representative concentration pathway

SDM:

Species distribution modeling

MaxEnt:

Maximum entropy

GFDL-CM3:

Geophysical fluid dynamics laboratory-climate model 3

ROC:

Receiver operating characteristic curve

AUC:

Area under the curve

MESS:

Multivariate environmental similarity surface

TSS:

True skill statistic

GBIF:

Global Biodiversity Information Facility

References

  1. Adhikari D, Barik SK, Upadhaya K (2012) Habitat distribution modeling for reintroduction of Ilex khasiana Purk, a critically endangered tree species of northeastern India. Ecol Eng 40:37–43. https://doi.org/https://doi.org/10.1016/j.ecoleng.2011.12.004

  2. Allouche O, Tsoar A, Kadmon R (2006) Assessing the accuracy of species distribution models: prevalence, kappa and the true skill statistic (TSS). J Appl Ecol 43(6):1223–1232. https://besjournals.onlinelibrary.wiley.com/doi/full/https://doi.org/10.1111/j.1365-2664.2006.01214.x

  3. Bachman S, Moat J, Hill AW, De La Torre J, Scott B (2011) Supporting red list threat assessments with GeoCAT: geospatial conservation assessment tool. ZooKeys 150(SI):117–126. https://doi.org/10.3897/zookeys.150.2109

  4. Borthakur SK, Baruah PS, Deka K, Das P, Sarma B, Adhikari D, Tanti B (2018) Habitat distribution modelling for improving conservation status of Brucea mollis Wall. ex Kurz. – an endangered potential medicinal plant of Northeast India. J Nat Conserv 43:104–110. https://doi.org/10.1016/j.jnc.2018.02.010

  5. Castaño-Santamaría J, López-Sánchez CA, Obeso JR, Barrio-Anta M (2019) Modelling and mapping beech forest distribution and site productivity under different climate change scenarios in the Cantabrian Range (North-western Spain). Forest Ecol Manage 450:117488. https://doi.org/10.1016/j.foreco.2019.117488

  6. CEPF (2020) Critical Ecosystem Partnership Fund: biodiversity hotspots. https://www.cepf.net/our-work/biodiversity-hotspots

  7. Chaturvedi RK, Joshi J, Jayaraman M, Bala G, Ravindranath NH (2012) Multi-model climate change projections for India under representative concentration pathways. Curr Sci 103(7):791–802

    Google Scholar 

  8. Chen JL, Wilson CR, Tapley BD (2006) Satellite gravity measurements confirm accelerated melting of Greenland ice sheet. Science 313(5795):1958–1960. https://doi.org/. https://doi.org/10.1126/science.1129007

    CAS  Article  Google Scholar 

  9. Chunco AJ, Phimmachak S, Sivongxay N, Stuart BL (2013) Predicting environmental suitability for a rare and threatened species (Lao Newt, Laotriton laoensis) using validated species distribution models. PLoS One 8:e59853. https://doi.org/10.1371/journal.pone.0059853

  10. Conservation International (2005) Biodiversity hotspots: Himalaya. http://www.biodiversityhotspots.org/xp/Hotspots/himalaya/. Accessed 10 Oct 2019

  11. Curtis JT, McIntosh RP (1950) The interrelations of certain analytic and synthesis phytosociological characters. Ecology 31:434–455. https://doi.org/10.2307/1931497

  12. Danielson JJ, Gesch DB (2011) Global multi-resolution terrain elevation data 2010 (GMTED2010). US Geological Survey, USA https://topotools.cr.usgs.gov/gmted_viewer/viewer.htm. Accessed 10 March 2020

    Book  Google Scholar 

  13. Das A (1995) Diversity of angiospermic flora of Darjeeling hills. In: Pandey AK (ed) Taxonomy and biodiversity. CBS Publishers and Distributers, Delhi, pp 118–127

    Google Scholar 

  14. Das J, Thapa S, Pradhan D, Thorat SS, Talukdar NC (2013) Intra-specific genetic diversity, phytochemical analysis and antioxidant activities of a potential Himalayan Swertia (Swertia bimaculata Hook. F. & Thomas.). Ind Crops Prod 49:341–347. https://doi.org/10.1016/j.indcrop.2013.05.017

  15. Dullinger S, Gattringer A, Thuiller W, Moser D, Zimmermann NE, Guisan A, Willner W, Plutzar C, Leitner M, Mang T, Caccianiga M, Dirnböck T, Ertl S, Fischer A, Lenoir J, Svenning J, Psomas A, Schmatz D R, Silc U, Vittoz P, Hülber K (2012) Extinction debt of high-mountain plants under twenty-first-century climate change. Nat Clim Change 2:619–622. https://doi.org/10.1038/nclimate1514

  16. Elith J, Kearney M, Phillips S (2010) The art of modelling range-shifting species. Methods Ecol Evol 1(4):330–342. https://doi.org/10.1111/j.2041-210X.2010.00036.x

    Article  Google Scholar 

  17. Fischer EM, Knutti R (2015) Anthropogenic contribution to global occurrence of heavy-precipitation and high-temperature extremes. Nat Clim Change 5:560–564. https://doi.org/10.1038/nclimate2617

  18. GBIF (2020) GBIF Occurrence Download. https://www.gbif.org/. Accessed 28 Apr 2020

  19. Gebrewahid Y, Abrehe S, Meresa E, Eyasu G, Abay K, Gebreab G, Kidanamariam K, Adissu G, Abreha G, Darcha G (2020) Current and future predicting potential areas of Oxytenanthera abyssinica (A. Richard) using MaxEnt model under climate change in Northern Ethiopia. Ecol Process 9:6. https://doi.org/10.1186/s13717-019-0210-8

  20. Grierson AJC, Long DG (1999) Flora of Bhutan: volume II part 2. The Royal Botanic Garden of Edinburgh & The Royal Government of Bhutan, United Kingdom & Bhutan

    Google Scholar 

  21. Griffies SM, Winton M, Donner LJ, Horowitz LW, Downes SM, Farneti R, Gnanadesikan A, Hurlin WJ, Lee H, Liang Z, Palter JB, Samuels BL, Wittenberg AT, Wyman BL, Yin J, Zadeh N (2011) The GFDL CM3 coupled climate model: characteristics of the ocean and sea ice simulations. J Climate 24:3520–3544. https://doi.org/10.1175/2011JCLI3964.1

  22. Guo Q, Fu B, Shi P, Cudahy T, Zhang J, Xu H (2017) Satellite monitoring the spatial-temporal dynamics of desertification in response to climate change and human activities across the Ordos Plateau, China. Remote Sens 9:525. https://doi.org/10.3390/rs9060525

  23. Hamid M, Khuroo AA, Charles B, Ahmad R, Singh CP, Aravind NA (2019) Impact of climate change on the distribution range and niche dynamics of Himalayan birch, a typical treeline species in Himalayas. Biodivers Conserv 28:2345–2370. https://doi.org/10.1007/s10531-018-1641-8

  24. Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A (2005) Very high resolution interpolated climate surfaces for global land areas. Int J Climatol 25:1965–1978. https://doi.org/10.1002/joc.1276, 15

  25. Hoffmann AA, Rymer PD, Byrne M, Ruthrof KX, Whinam J, McGeoch M, Bergstrom DM, Guerin GR, Sparrow B, Joseph L, Hill SJ, Andrew NR, Camac J, Bell N, Riegler M, Gardner JL, Williams SE (2019) Impacts of recent climate change on terrestrial flora and fauna: some emerging Australian examples. Austral Ecol 44:3–27. https://doi.org/10.1111/aec.12674

  26. IPCC (2014) Climate Change 2014: synthesis report. In: Pachauri RK, Meyer LA (eds) Contribution of working groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change Core Writing Team. IPCC, Switzerland

    Google Scholar 

  27. Joshi P, Dhawan V (2005) Swertia chirayita - an overview. Curr Sci 89:635–640

  28. Jueterbock A, Smolina I, Coyer JA, Hoarau G (2016) The fate of the Arctic seaweed Fucus distichus under climate change: an ecological niche modeling approach. Ecol Evol 6:1712–1724. https://doi.org/10.1002/ece3.2001

  29. Kanade R, John R (2018) Topographical influence on recent deforestation and degradation in the Sikkim Himalaya in India: implications for conservation of East Himalayan broadleaf forest. Appl Geogr 92:85–93. https://doi.org/10.1016/j.apgeog.2018.02.004

  30. Khanum R, Mumtaz AS, Kumar S (2013) Predicting impacts of climate change on medicinal asclepiads of Pakistan using Maxent modeling. Acta Oecol 49:23–31. https://doi.org/10.1016/j.actao.2013.02.007

  31. Kolanowska M, Konowalik K (2014) Niche conservatism and future changes in the potential area coverage of Arundina graminifolia, an invasive orchid species from Southeast Asia. Biotropica 46:157–165. https://doi.org/10.1111/btp.12089

  32. Kumar P (2012) Assessment of impact of climate change on Rhododendrons in Sikkim Himalayas using Maxent modelling: limitations and challenges. Biodivers Conserv 21:1251–1266. https://doi.org/10.1007/s10531-012-0279-1

  33. Kumar S, Stohlgren TJ (2009) Maxent modeling for predicting suitable habitat for threatened and endangered tree Canacomyrica monticola in New Caledonia. J Ecol Nat Environ 1:94–98

    Google Scholar 

  34. Matteodo M, Wipf S, Stöckli V, Rixen C, Vittoz P (2013) Elevation gradient of successful plant traits for colonizing alpine summits under climate change. Environ Res Lett 8(2):024043. https://doi.org/10.1088/1748-9326/8/2/024043

    Article  Google Scholar 

  35. Menhinick EFA (1964) Comparison of some species diversity indices applied to samples of field insects. Ecology 45:858–868

    Article  Google Scholar 

  36. 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 11:175–181. https://doi.org/10.3354/esr00275

  37. Mingyang L, Yunwei J, Kumar S, Stohlgren TJ (2008) Modeling potential habitats for alien species Dreissena polymorpha in continental USA. Acta Ecol Sin 28:4253–4258. https://doi.org/10.1016/S1872-2032(08)60080-3

  38. Nerem RS, Beckley BD, Fasullo JT, Hamlington BD, Masters D, Mitchum GT (2018) Climate-change–driven accelerated sea-level rise detected in the altimeter era. Proc Natl Acad Sci 115:2022–2025

    CAS  Article  Google Scholar 

  39. Padalia H, Srivastava V, Kushwaha SPS (2014) Modeling potential invasion range of alien invasive species, Hyptis suaveolens (L.) Poit. in India: comparison of MaxEnt and GARP. Ecol Inform 22:36–43

    Article  Google Scholar 

  40. Pal R, Biswas SS, Mondal B, Pramanik MK (2016) Landslides and floods in the Tista Basin (Darjeeling and Jalpaiguri Districts): historical evidence, causes and consequences. J Ind Geophys Union 20:209–215

    Google Scholar 

  41. Pande HK, Arora S (2014) India’s Fifth National Report to the Convention on Biological Diversity. Ministry of Environment and Forestry, Government of India, India

    Google Scholar 

  42. Pandey DK, Basu S, Jha TB (2012) Screening of different East Himalayan species and populations of Swertia L. based on exomorphology and mangiferin content. Asian Pac J Trop Biomed 2:S1450–S1456. https://doi.org/10.1016/S2221-1691(12)60436-5

  43. Parolo G, Rossi G (2008) Upward migration of vascular plants following a climate warming trend in the Alps. Basic Appl Ecol 9:100–107. https://doi.org/10.1016/j.baae.2007.01.005

  44. Pearson RG (2007) Species’ distribution modeling for conservation educators and practitioners. Synth Am Mus Nat Hist 50:54–89

    Google Scholar 

  45. Pearson RG, Thuiller W, Araújo MB, Martinez-Meyer E, Brotons L, McClean C, Miles L, Segurado P, Dawson TP, Lees DC (2006) Model-based uncertainty in species range prediction. J Biogeogr 33:1704–1711. https://doi.org/10.1111/j.1365-2699.2006.01460.x

  46. Phillips EA (1959) Methods of vegetation study. Holt, Reinhart and Winston Co., United States of America

    Google Scholar 

  47. Phillips SJ, Anderson RP, Schapire RE (2006) Maximum entropy modeling of species geographic distributions. Ecol Model 190:231–259. https://doi.org/10.1016/j.ecolmodel.2005.03.026

  48. Phillips SJ, Dudík M (2008) Modeling of species distributions with Maxent: new extensions and a comprehensive evaluation. Ecography 31(2):161–175. https://doi.org/10.1111/j.0906-7590.2008.5203.x

    Article  Google Scholar 

  49. Pielou EC (1966) The measurement of diversity in different types of biological collections. J Theor Biol 13:131–144. https://doi.org/10.1016/0022-5193(66)90013-0

    Article  Google Scholar 

  50. Qin A, Liu B, Guo Q, Bussmann RW, Ma F, Jian Z, Pei S (2017) Maxent modeling for predicting impacts of climate change on the potential distribution of Thuja sutchuenensis Franch., an extremely endangered conifer from southwestern China. Glob Ecol Conserv 10:139–146. https://doi.org/10.1016/j.egecco.2017.02.004

  51. Ranjitkar S, Sujakhu NM, Lu Y, Wang Q, Wang M, He J, Mortimer PE, Xu J, Kindt R, Zomer RJ (2016) Climate modelling for agroforestry species selection in Yunnan Province, China. Environ Model Softw 75:263–272. https://doi.org/10.1016/j.envsoft.2015.10.027

  52. Rawat N, Kandpal K, Purohit S, Singh G, Pant D (2017) Predicting potential habitat distribution of Rauwolfia serpentina an important medicinal plant using Maxent modeling in Doon Valley, Uttarakhand State, India. Int J Adv Remote Sens GIS 6(1):2267–2273. https://doi.org/10.23953/cloud.ijarsg.288

  53. Ray R, Gururaja KV, Ramchandra TV (2011) Predictive distribution modeling for rare Himalayan medicinal plant Berberis aristata DC. J Environ Biol 32:725–730

  54. Remya K, Ramachandran A, Jayakumar S (2015) Predicting the current and future suitable habitat distribution of Myristica dactyloides Gaertn. using MaxEnt model in the Eastern Ghats, India. Ecol Eng 82:184–188. https://doi.org/10.1016/j.ecoleng.2015.04.053

  55. Roe GH, Baker MB, Herla F (2017) Centennial glacier retreat as categorical evidence of regional climate change. Nat Geoscience 10(2):95–99. https://doi.org/10.1038/ngeo2863

    CAS  Article  Google Scholar 

  56. Saha P, Mandal S, Das A, Das PC, Das S (2004) Evaluation of the anticarcinogenic activity of Swertia chirata Buch. Ham, an Indian medicinal plant, on DMBA-induced mouse skin carcinogenesis model. Phytother Res 18:373–378

  57. Samaddar T, Chaubey B, Jha S, Jha TB (2013) Determination of swertiamarin and amarogentin content and evaluation of antibacterial activity in Eastern Himalayan species of Swertia L. Pharmacognosy Commun 3:64–70

    Article  Google Scholar 

  58. Samaddar T, Jha S, Jha TB (2014) Indian Swertia from Eastern Himalaya: strategies of conservation and biotechnological improvements. In: Rybczyński JJ, Davey MR, Mikuła A (eds) The Gentianaceae-volume 1: characterization and ecology. Springer, Berlin, pp 279–301. https://doi.org/10.1007/978-3-642-54010-3_11

  59. Sarma B, Baruah PS, Tanti B (2018) Habitat distribution modeling for reintroduction and conservation of Aristolochia indica L. - a threatened medicinal plant in Assam, India. J Threatened Taxa 10:12531–12537. https://doi.org/10.11609/jott.3600.10.11.12531-12537

  60. Simpson EH (1949) Measurement of diversity. Nature 163:188

    Article  Google Scholar 

  61. Sobek-Swant S, Kluza DA, Cuddington K, Lyons DB (2012) Potential distribution of emerald ash borer: what can we learn from ecological niche models using Maxent and GARP? Forest Ecol Manage 281:23–31. https://doi.org/10.1016/j.foreco.2012.06.017

  62. Srivastava V, Griess VC, Padalia H (2018) Mapping invasion potential using ensemble modelling. A case study on Yushania maling in the Darjeeling Himalayas. Ecol Model 385:35–44. https://doi.org/10.1016/j.ecolmodel.2018.07.001

  63. Starkel L, Sarkar S (2014) The Sikkim-Darjeeling Himalaya: landforms, evolutionary history and present-day processes. In: Kale VS (ed) Landscapes and landforms of India. Springer, Germany, pp 157–164. https://doi.org/10.1007/978-94-017-8029-2_15

    Google Scholar 

  64. Thapa S, Chitale V, Rijal SJ, Bisht N, Shrestha BB (2018) Understanding the dynamics in distribution of invasive alien plant species under predicted climate change in Western Himalaya. PLoS One 13:e0195752. https://doi.org/10.1371/journal.pone.0195752

  65. Thomas CD, Cameron A, Green RE, Bakkenes M, Beaumont LJ, Collingham YC, Erasmus BFN, Ferreira de Siqueira M, Grainger A, Hannah L, Hughes L, Huntley B, van Jaarsveld AS, Midgley GF, Miles L, Ortega-Huerta MA, Townsend Peterson AT, Phillips OL, Williams SE (2004) Extinction risk from climate change. Nature 427:145–148. https://doi.org/10.1038/nature02121

  66. Tsiftsis S, Djordjeviņć V, Tsiripidis I (2019) Neottia cordata (Orchidaceae) at its southernmost distribution border in Europe: threat status and effectiveness of Natura 2000 Network for its conservation. J Nat Conserv 48:27–35. https://doi.org/10.1016/j.jnc.2019.01.006

  67. Urban MC (2015) Accelerating extinction risk from climate change. Science 348:571–573. https://doi.org/10.1126/science.aaa4984

  68. Warren DL, Glor RE, Turelli M (2010) Enmtools: a toolbox for comparative studies of environmental niche models, evaluation. Ecography 31:161–175. https://doi.org/10.1111/j.1600-0587.2009.06142.x

  69. Shannon CE, Weaver W (1963) The mathematical theory of communication. University of Illinois Press, USA

  70. WFO (2019) World Flora Online. Published on the Internet. http://www.worldfloraonline.org. Accessed 24 Oct 2019

  71. Whitford PB (1949) Distribution of woodland plants in relation to succession and clonal growth. Ecology 30(2):199–208. https://doi.org/10.2307/1931186

    Article  Google Scholar 

  72. Yang XQ, Kushwaha SPS, Saran S, Xu J, Roy PS (2013) Maxent modeling for predicting the potential distribution of medicinal plant, Justicia adhatoda L. in Lesser Himalayan foothills. Ecol Eng 51:83–87. https://doi.org/10.1016/j.ecoleng.2012.12.004

  73. Yi YJ, Cheng X, Yang ZF, Zhang SH (2016) Maxent modeling for predicting the potential distribution of endangered medicinal plant (H. riparia Lour) in Yunnan, China. Ecol Eng 92:260–269. https://doi.org/10.1016/j.ecoleng.2016.04.010

  74. Yonzone R (2017) The genus Swertia L. (Gentianaceae): an important ethnomedicinal plant of Darjeeling Himalaya of West Bengal, India. Trends Biosci 10:9045–9049

    Google Scholar 

  75. Young N, Carter L, Evangelista P (2011) A MaxEnt model v3.3.3e tutorial (ArcGIS v10). Fort Collins, Colorado

  76. Zhang L, Cao B, Bai C, Li G, Mao M (2016) Predicting suitable cultivation regions of medicinal plants with Maxent modeling and fuzzy logics: a case study of Scutellaria baicalensis in China. Environ Earth Sci 75:361. https://doi.org/10.1007/s12665-015-5133-9

Download references

Acknowledgements

The authors are sincerely thankful to the Department of Forests, Government of West Bengal, India, for all the necessary permissions. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Funding

Not applicable

Author information

Affiliations

Authors

Contributions

DB and SM have equal contribution. The authors read and approved the final manuscript.

Corresponding author

Correspondence to Saurav Moktan.

Ethics declarations

Ethics approval and consent to participate

Not applicable

Consent for publication

Not applicable

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Boral, D., Moktan, S. Predictive distribution modeling of Swertia bimaculata in Darjeeling-Sikkim Eastern Himalaya using MaxEnt: current and future scenarios. Ecol Process 10, 26 (2021). https://doi.org/10.1186/s13717-021-00294-5

Download citation

Keywords

  • Distribution modeling
  • Swertia bimaculata
  • MaxEnt
  • Eastern Himalaya
\