RUHUNA JOURNAL OF SCIENCE Vol 12 (2): 144 -154, December 2021 eISSN: 2536-8400 © Faculty of Science http://doi.org/10.4038/rjs.v12i2.108 University of Ruhuna © Faculty of Science, University of Ruhuna Sri Lanka 144 Determination of the distribution of Calotropis gigantea (L.) in Sri Lanka using MaxEnt modelling technique W.P.S.N. Wijeweera1*, K.A.D.W. Senaratne2, K. Dhileepan2 and M.P.K.S.K. de Silva1 1Department of Zoology, University of Ruhuna, Matara, Sri Lanka 2Biosecurity Queensland, Department of Agriculture and Fisheries, Queensland, Australia *Correspondence: surendi87nisha@gmail.com; ORCID: https://orcid.org/0000-0003-1359-0188 Received: 5th May 2021, Revised: 03rd November 2021, Accepted: 21st December 2021 Abstract Calotropis gigantea is a drought-resistant, salt-tolerant, native plant in Sri Lanka with ayurvedic medicinal values. The plant is used for fiber, fodder and fuel, as well as a fertilizer. Despite its benefits, C. gigantea has become an emerging problem in countries where it has been introduced because of its invasiveness. Although C. gigantea is widely distributed in Sri Lanka, precise information on its distribution is lacking. Therefore, the present study was aimed at determining the distribution of C. gigantea in Sri Lanka. Field surveys were conducted in 120 sites covering all provinces in Sri Lanka from December 2014 to June 2015 to record the occurrence of C. gigantea. C. gigantea was distributed in all provinces except the Central province. It was more widespread along coastal regions, but its occurrence was low in the Western and Sabaragamuwa provinces. MaxEnt modelling predicted that the entire coast of Northern, North-Central and Eastern provinces contain the highest probability of C. gigantea distribution whereas the low probability was in North-Western, Western, Southern, Uva, Central, and Sabaragamuwa provinces. No occurrence probability was predicted in certain regions of Southern, Sabaragamuwa, Uva, and Central provinces of Sri Lanka. The study provides information on the current and potential distribution range of C. gigantea in Sri Lanka. Keywords: Calotropis gigantea, MaxEnt modelling, plant distribution 1 Introduction The spatial distribution of a species is the arrangement of that species across the earth’s surface. No species in the world is adapted to live in all environmental conditions of the earth. Their spatial distribution is limited by biotic and abiotic factors (Pidwirny 2018). Climatic factors act as major limiting factors for the distribution of plants. The microclimate is the climatic condition that prevails in localized regions closer to the surface of the earth. It consists of environmental variables including temperature, moisture content, wind speed, and light (Naiman et https://rjs.ruh.ac.lk/index.php/rjs/index https://rjs.ruh.ac.lk/index.php/rjs/index http://doi.org/10.4038/rjs.v12i2.108 https://creativecommons.org/licenses/by-nc/4.0/ mailto:surendi87nisha@gmail.com W.P.S.N. Wijeweera et al. Distribution of Calotropis gigantea in Sri Lanka Ruhuna Journal of Science Vol 12 (2): 144 -154, December 2021 145 al. 2005) and is closely bound with the habitat and influence on plants on a fine scale (Bramer et al. 2018). Climatic differences create significant changes in the vegetation and form climatic zonation (Azarkhavarani et al. 2015). The close relationship of climate and vegetation is used to define and analyze bio-climatic zones using different vegetation and climatic maps (Azarkhavarani et al. 2015). Later, the use of computerized data related to climate and vegetation has led to species distribution modelling techniques (Barbet-Massin et al. 2018). Primarily, species distribution models are developed to predict the distribution of species and secondarily to study the functional association of species with their living environment (Austin 2002). Since 2006, the MaxEnt software package is widely used for modelling the species distribution covering more than 1000 publications. MaxEnt program uses presence- only data (species recorded locations) and data of environmental predictors as input data of the program (Merow et al. 2013). Modelling techniques have been applied to study the distribution range of Calotropis procera, an invasive plant in Australia. The driving factors for its distribution had not been identified and assuming “climatic conditions” as a driving factor on the plant distribution, MaxEnt modelling has been applied to identify the current and potential distribution of C. procera. According to the model prediction, the distribution of plants is best explained through climatic variables and human disturbances (Menge et al. 2016). C. gigantea is native to India, China, Sri Lanka, and Malaysia (Dhileepan 2014) and is also distributed in Afghanistan, Algeria, Burkina Faso, Cameroon, Ghana, Guinea-Bissau, and Iran (Kumar et al. 2013). It is considered a medicinal plant having Ayurvedic value in its native range (Kumar and Kumar 2015). Different plant parts of C. gigantea are used to cure a wide range of diseases including bronchial asthma, cholera, convulsions, pneumonia, ringworm infection, smallpox infection, toothache, epilepsy, skin diseases, and epilepsy (Kumar and Kumar 2015). In addition, C. gigantea is used as fodder, fiber source, fuel and febrifuge (Kumar et al. 2013). However, in certain countries, the plant is considered as an invasive species. It is recorded as an exotic- invasive plant in Australia, the Virgin Islands of the United States, Mexico, and Brazil (Kumar et al. 2013). They prefer to grow on the abandoned over-cultivated areas, over-grazed grounds, roadsides and lagoon edges (Kumar et al. 2013). C. gigantea is commonly found in Sri Lanka, but there are no published records on its distribution in Sri Lanka and the habitats they occupy. Therefore, the objectives of the present study are to collect occurrence data of C. gigantea in Sri Lanka, to identify habitats and habitat characteristics of C. gigantea, to prepare a map of its distribution concerning selected environmental variables by using MaxEnt modelling technique and to study the density of the plant in different regions of Sri Lanka. W.P.S.N. Wijeweera et al. Distribution of Calotropis gigantea in Sri Lanka Ruhuna Journal of Science Vol 12 (2): 144 -154, December 2021 146 2 Material and Methods 2.1 Collecting occurrence data of C. gigantea A field survey on the distribution of C. gigantea was conducted from December 2014 to June 2015 covering 120 sampling sites representing nine provinces of Sri Lanka (Figure 1). Each field visit covered eight (8) sampling sites on average. Sampling was done only once for each site. The roadside sampling was done as Calotropis spp. tend to grow closer to road-edges and for easy accessibility of sampling (Sharma et al. 2010). Roadside sampling sites were selected randomly maintaining equal distances on the main road at 30-minute intervals while travelling on a vehicle with a speed of 50 Km per hour. If a new site with C. gigantea plants was not observed after 30 minutes, travelling continued until observing a site with C. gigantea plants. In every sampling site, C. gigantea distribution (GPS coordinates) occurrence data were recorded. 2.2 Mapping the distribution of C. gigantea in Sri Lanka The bioclimatic variables were downloaded from the Worldclim Global climate data website (https://www.worldclim.org/). Bioclimatic variables used in the present study were derived from the monthly temperature and rainfall values. Biologically meaningful variables derived from the monthly temperature and rainfall (e.g., mean annual temperature, annual precipitation, annual range in temperature and precipitation) and extreme or limiting environmental factors (e.g. temperature of the coldest and warmest month, and precipitation of the wet and dry quarters) are used to represent annual trends. The above data is in the form of layers in a grid format covering the global land area. They are in the latitude/longitude coordinate reference system. This data is available in ESRI grid (raster) format, Geo TIFF format and, Generic grid (Raster format). It is also available at 4 different spatial resolutions; from 30 seconds (0.93 x 0.93 = 0.86 km2 at the equator) to 2.5, 5 and 10 minutes (18.6 x 18.6 = 344 km2 at the equator). Environmental variables such as temperature and rainfall should be included in raster Arc/Info ASCII Grid format. The maximum entropy model (MaxEnt) software and DIVA-GIS software were used. In Worldclim, the globe is divided into 60 squares which refer as tiles. Each tile consists of an enormous data package including climatic data of different regions of the World. The global climatic data is available in 1 km2 spatial resolution approximately (Fick and Hijimans 2017). As original data of WorldClim is available in 30 seconds spatial resolution, it was selected for the study. For this study, bioclimatic data of WorldClim related to 28 tiles (India - Sri Lanka region) was downloaded from http://www.worldclim.org/tiles.php and in addition, the Generic grid (Raster) format was used. For the calculations of MaxEnt, two file types are needed. They are Comma Separated Values (.csv) and ESRI ASCII GIS (.asc) (Young et al. 2011). However, in Worldclim, the data is not in the .asc format and it W.P.S.N. Wijeweera et al. Distribution of Calotropis gigantea in Sri Lanka Ruhuna Journal of Science Vol 12 (2): 144 -154, December 2021 147 does not support MaxEnt. DIVA-GIS software was used to convert downloaded bioclimatic data of WorldClim into .asc format and mapping of C. gigantea distribution of Sri Lanka was done (Figure 1). 2.3 Model suitability and validation The suitability of the model was determined by the Area Under the Curve /AUC. It represents the ratio of sensitivity vs. specificity of C. gigantea. AUC makes a comparison between the performances of one model with another. In addition, AUC is useful to evaluate multiple models of MaxEnt. ACU contained a possible value range 0 to 1, and values above 0.5 represent the higher predictive power while values less than 0.5 represent lower model performance. In other words, if the value of AUC is closer to 1 means the mode is extremely appropriate for the predicted distribution while a value close to 0 means the model is not suitable (Young et al. 2011). Model validation is necessary to assess the model's overall performance and application potential (Uden et al. 2015). Independent data or data which is not used for model training is required for the model validation (Uden et al. 2015). If the data set is sufficient, observations are taken as randomly or spatially subset into training and testing data sets. As there were sufficient observations (n=120) in this study, 20 per cent of data was set aside for testing the model (Uden et al. 2015). 2.4 Habitats of C. gigantea in Sri Lanka The patterns of land use indicate the changes in the plant habitats. As a result, land- use patterns greatly influence plant distribution (Honnay et al. 1999). To study the land-use pattern, habitats of C. gigantea were recorded during the field study. Habitats were categorized as roadsides without disturbances, roadsides with wastelands, roadsides with abandoned lands, roadsides near to seashore, roadsides adjacent to a cemetery, roadsides closer to reservoirs and roadsides with continuous anthropogenic activities such as railway tracks, road construction sites and cattle grazing lands. 2.5 Plant density of C. gigantea To calculate plant density, a randomly selected (3m ×3m) plant patch of C. gigantea in each site was observed. C. gigantea plants in randomly selected 9 m2 areas were counted and recorded. The recorded values were ranked as High density (H), Moderate density (M) and, Low density (L). These ranks were given according to the number of plants that were present in 9 m2 of selected C. gigantea patch (1 to 2 plants: low density, 3 to 4 plants: moderate density, more than 4 plants: high density). Percentage of C. gigantea plant density was calculated as, W.P.S.N. Wijeweera et al. Distribution of Calotropis gigantea in Sri Lanka Ruhuna Journal of Science Vol 12 (2): 144 -154, December 2021 148 C. gigantea plant density % = 3 Results 3.1 Predicted distribution of C. gigantea in Sri Lanka Fig 1. Sampling sites and predicted probability of Calotropis gigantea distribution in Sri Lanka using MaxEnt modelling The probability of C. gigantea distribution according to the environmental variables of the MaxEnt modelling technique is shown in Figure 1. In Figure 1, warmer colours (red and orange) show the areas with a predicted high probability of C. gigantea distribution including the whole coastal belt of the country, Northern, North-Central, and Eastern provinces of Sri Lanka. In addition, a high probability of C. gigantea is predicted in certain regions of Southern and North-Western provinces. The low Number of sampling sites having the same density rank of the selected district Number of C. gigantea sampling sites per selected district ×100% W.P.S.N. Wijeweera et al. Distribution of Calotropis gigantea in Sri Lanka Ruhuna Journal of Science Vol 12 (2): 144 -154, December 2021 149 probability of C. gigantea distribution (yellow and green) is predicted in regions in North-Western, Western, Southern, Uva, Central and, Sabaragamuwa provinces of the country. According to the prediction of the model, there is no probability of occurrence of C. gigantea (white) in certain regions of Southern, Sabaragamuwa, Uva and, Central provinces of Sri Lanka. 3.2 Evaluation of quality of MaxEnt model AUC is used to predict the accuracy of the model. It determines whether the probability of a present location is ranked higher than a random background location or not, and Reddy et al. (2015) described the ranking system in values of AUC as 0.50-0.60 (fail), 0.60-0.70 (poor), 0.70-0.80 (fair), 0.80-0.90 (good), and 0.90-1.0 (excellent). Fig 2. AUC curve of sensitivity versus specificity for Calotropis gigantea In the present study, the AUC was obtained based on the potential climatic factors which affect the distribution of C. gigantea in Sri Lanka. The AUC values were 0.971 and 0.973, for training and test data, respectively (Figure 2). It indicates the constructed model to be appropriate with an ‘excellent’ predictive accuracy. Therefore, it is suitable to make predictions on the geographic distribution of C. gigantea in Sri Lanka. By entering ‘20’ in the settings of ‘random test percentage’, the program randomly set aside 20% of the sample records for testing. The analysis utilizes a threshold to make a binary prediction with suitable conditions predicted above the threshold level (suitable) and below the threshold level (unsuitable). Figure 3 indicates the omission rate and predicted area as a function of the cumulative threshold. The calculation of the omission rate was done on the training records as well as the test records (80% and 20% of the presence records, respectively). W.P.S.N. Wijeweera et al. Distribution of Calotropis gigantea in Sri Lanka Ruhuna Journal of Science Vol 12 (2): 144 -154, December 2021 150 According to Reddy et al. (2015), the omission rate should be close to the predicted omission. Fig 3. Graph of omission and predicted area for Calotropis gigantea Figure 3 shows how testing and training omission and predicted area for C. gigantea vary with respect to the cumulative threshold. The omission on test samples has high compatibility with the predicted omission rate. In some situations, the test omission line lies well below the predicted omission line. On the other hand, in some situations, the test omission line lies well above the predicted omission line. Such conditions appear due to the dependency of test and training data, as they are derived from the same spatially auto-correlated presence data. This denotes that the MaxEnt model is significantly better than random in the binomial test of omission and predicted area curve. 3.3 Habitats of C. gigantea in Sri Lanka The percentage occurrence of C. gigantea according to habitat category is given in Table 1. The majority (66.9%) of C. gigantea plants were located on either side of the roads with undisturbed soil. During the survey, 12.1 % of C. gigantea plant sites were recorded in dumped lands closer to roadsides while 7.3% of them were recorded at sea-shore closer to roadsides. In the dry zone, C. gigantea plants were recorded closer to reservoirs such as tanks, lakes and, estuaries. It was also recorded in cultivations such as paddy and coconut. Only 4% of C. gigantea habitats were associated with roads where continuous anthological activities prevail. W.P.S.N. Wijeweera et al. Distribution of Calotropis gigantea in Sri Lanka Ruhuna Journal of Science Vol 12 (2): 144 -154, December 2021 151 Table 1. Percentage of occurrence of C. gigantea in Sri Lanka according to habitat category. Habitat category % Occurrence of C. gigantea according to habitat Roadsides without disturbances 66.9 Dumped lands near roadsides 12.1 Seashore near roadsides 7.3 Cultivations near roadsides 6 Reservoirs close roadsides 4.8 Roadsides with continuous anthropogenic activities 4.0 Abandoned lands close to roadsides 2.4 Cemetery at roadsides 2.4 3.4. Plant density of C. gigantea The percentage of plant density according to the province is recorded in Table 2. C. gigantea plants are highly preferred to grow in low plant densities in Northern (73.68%) and Eastern (72.0%) provinces. A higher percentage of moderate plant density was recorded in Western (33.33%) and North-Western (31.25%) provinces. In the Uva province, C. gigantea plants tend to grow as high-density mass with a peak (50%) percentage. In contrast, Western and Northern provinces have no sites with high C. gigantea plant density. The plant was absent in Central province. Table 2: Percentage Calotropis gigantea plant density with respect to provinces in Sri Lanka. Province Low plant density % Moderate plant density % High plant density % Northern 73.68 26.32 0 Eastern 72.0 4.0 24.0 Western 66.67 33.33 0 North- Central 57.89 15.72 26.32 Southern 50.0 19.44 30.56 North- Western 37.5 31.25 31.25 Uva 30.0 20.0 50.0 Central 0 0 0 Sabaragamuwa NE NE NE NE- Not estimated due to less coverage area of Sabaragamuwa Province 4 Discussion There are many records on roadside surveys on invasive species (Baard and Kraaij 2019). It may be due to two reasons; invasive species tend to grow on roadsides and easy accessibility for sampling. Roads act as corridors that facilitate the distribution of invasive species to introduced areas. The development of road networks and frequent road constructions further facilitate the range expansion of invasive species. Therefore, road edges are ideal habitats for invasive plants which facilitate their dispersal to different geographical regions (Sharma et al. 2010). W.P.S.N. Wijeweera et al. Distribution of Calotropis gigantea in Sri Lanka Ruhuna Journal of Science Vol 12 (2): 144 -154, December 2021 152 According to Sharma et al. (2010), anthropogenic activities on roads pave the way for the invasion of Calotropis procera into introduced areas. The study also describes that road usage, vehicle smoke, and vehicle gust further facilitate seed dispersal of C. procera, and that invasion from urban areas to rural regions is possible via road systems (Sharma et al. 2010). In the present study, C. gigantea is in a broad range of habitats including undisturbed roadsides, cemeteries, abandoned lands, dumped lands, seashore and, cultivations. Quazi et al. (2013) also mentions that Calotropis spp. tend to grow on sand dunes closer to estuaries as well as over-grazed grasslands. Kumar et al. (2013) also record that, it appears as a weed in over-grazed lands where there is no competition from grasses. However, if there is grass, C. gigantea may not act as a dominant plant in open areas which is well studied for C. procera in Australia. According to Menge et al. (2017), C. procera is a poor competitor with respect to the native grass (Mitchell grass) and fails to invade grasslands. It also prefers to grow in desert regions due to its xerophytic nature such as milky latex in leaves, highly branched root system and waxy, thick leaves (Kumar et al. 2013). Low population densities limit the reproductive output of the plant. Pollinator limitation is a leading factor for low population densities of Calotropis spp. (Menge et al. 2017). In the present study, low population densities are present in the Western, Eastern and, Northern provinces on a large scale where there is a smaller number of pollinators. Personal observations reveal that highly urbanized Western Province contains a low number of pollinators which may be the reason for low plant density. Eastern and Northern regions are under road construction after the civil war and as a result, C. gigantea populations have been cut down and cleared. It may lead to low population densities in Northern and Eastern regions. As Calotropis spp. are salt and drought tolerant and resistant to low rainfall (300- 400 mm) (Kumar et al. 2013), they are abundant in places with similar climatic conditions such as in coastal regions of Southern, Northern, Eastern and, North- Central provinces of Sri Lanka. The observations (Table 1) are compatible with MaxEnt output (Figure 1) indicating climatic conditions highly affecting the C. gigantea distribution. As an example, MaxEnt predicts a low probability of C. gigantea in the Western province which belongs to the wet zone of the country. A similar observation is recorded during field visits also. In contrast, a higher percentage occurrence of C. gigantea is recorded in the Eastern, North-Central, Northern and Southern provinces of Sri Lanka. The MaxEnt model also predicts the same result indicating climatic suitability for C. gigantea growth prevailing in these areas. However, the model predicts a high probability of C. gigantea distribution in the coastal belt of Puttalam to Mannar which is not observed during field visits (Figure 1). The fragmentation of C. gigantea distribution from Puttalam to Mannar may be due to human activities that occurred during road construction projects after the civil war prevailed in these regions although there are ideal environmental conditions available for C. gigantea plants. Maxent model also predicted, a low probability of C. gigantea from Induruwa to Colombo, as it belongs to the wet zone of the country. On W.P.S.N. Wijeweera et al. Distribution of Calotropis gigantea in Sri Lanka Ruhuna Journal of Science Vol 12 (2): 144 -154, December 2021 153 the other hand, field observations reveal that clearance of the C. gigantea coverage occurs due to high urbanization. The combined effect of lack of climatic suitability and urbanization may have led to the disappearance of C. gigantea from Induruwa to Colombo. AUC values indicate that MaxEnt produced significantly accurate results. The sensitivity versus 1-specificity graph indicated that the MaxEnt model got an excellent predictive accuracy (mean AUC = 0.972) concerning the relationship between the distribution of C. gigantea and the selected environmental variables. The results show that the MaxEnt model can be used to study the climatic suitability for the distribution of C. gigantea in Sri Lanka. It acts as a tool to understand the potential distribution of C. gigantea in Sri Lanka. Most of the observed results in Table 2 are compatible with the predicted distributions in Figure 1. Field observations revealed high plant occurrence in the Northern, North- Central and, Eastern provinces of the country which is compatible with the model prediction. Similarly, the model predicts a low probability of C. gigantea distribution in large land areas of Uva, North-Western, Western and Central provinces of Sri Lanka which is observed during field visits also (Table 2). In addition, the model predicts a low probability of C. gigantea distribution in Sabaragamuwa province where the percentage of occurrence is not evaluated due to lack of land coverage during field visits (Table 2). Therefore, the MaxEnt model facilitated the prediction of species distributions where there is a lack of occurrence data. 5 Conclusions The present study provides a detailed map of C. gigantea as well as detailed information on C. gigantea plant distribution, plant density and, habitats of the plant in Sri Lanka. The knowledge of the distribution of C. gigantea is important as it has medicinal value. On the other hand, the association of environmental factors for its distribution is greatly important to control its invasiveness in the introduced range. The present study would be the first of its kind in Sri Lanka using MaxEnt to evaluate climatic suitability on C. gigantea. MaxEnt modelling predicted the distribution of the plant within the whole country concerning environmental variables. According to the MaxEnt, climatic factors highly influence on the distribution of C. gigantea. In addition, anthropological activities also play a considerable role in C. gigantea distribution. The study also provides occurrence data of C. gigantea in Sri Lanka which will be greatly important for further studies on its distribution in Sri Lanka as well as in the World. Acknowledgements Two anonymous reviewers are acknowledged for valuable comments on the initial draft of the manuscript. W.P.S.N. Wijeweera et al. Distribution of Calotropis gigantea in Sri Lanka Ruhuna Journal of Science Vol 12 (2): 144 -154, December 2021 154 References Austin M. 2002. Spatial prediction of species distribution: an interface between ecological theory and statistical modelling. Ecological Modelling 157: 101-118. Azarkhavarani SF, Rahimi M, Bernardi M. 2015. The most important climatic factors affecting distribution of Zygophyllum atriplicoides in semi-arid region of Iran (Case Study: Isfahan Province). Desert 20 (2): 145-156. Barbet-Massin M, Rome Q, Villemant C, Courchamp F. 2018. Can species distribution models really predict the expansion of invasive species? Plos ONE 13(3): e0193085. Baard JA, Kraaij T. 2019. Use of a rapid roadside survey to detect potentially invasive plant species along the Garden Route, South Africa. Koedoe 61(1): 1-10. Bramer I, Anderson BJ, Bennie J, Bladon AJ, De Frenne, P, Hemming D, Hill RA, Kearney MR, Korner C, Korstjens AH, Lenoir J, Maclean IMD, Marsh CD, Marecroft MD, Ohlemuller R, Slater HD, Saggitt AJ, Zellweger F, Gillingham PK. 2018. Advances in monitoring and modelling climate at ecologically relevant scales. Advances in Ecological Research 58: 101-161. Dhileepan K. 2014. Prospects for the classical biological control of Calotropis procera (Apocynaceae) using coevolved insects. Biocontrol Science and Technology 24(9): 977-998. Fick SE, Hijmans RJ. 2017. WorldClim 2: new 1-km spatial resolution climate surfaces for global land area. International Journal of Climatology, 37 (12). Honnay O, Hermy M, Coppin, P. 1999. Impact of habitat quality on forest plant species colonization. Forest Ecology and Management 115 (2-3): 157-170. Kumar D, Kumar S. 2015. Calotropis gigantea (L.) Dryand – A review update. Indian Journal of Research in Pharmacy and Biotechnology 3(3): 218-230. Kumar P, Suresh E, Kalavath S. 2013. Review on a potential herb Calotropis gigantea (L.) R. Br. Scholars Academic Journal of Pharmacy 2(2): 135-143. Menge EO, Bellairs SM, Lawes MJ. 2017. Disturbance-dependent invasion of the woody weed, Calotropis procera, in Australian rangelands. The Rangeland Journal 39: 201-211. Menge EO, Stobo-Wilson A, Oliveira SL, Lawes MJ. 2016. The potential distribution of the woody weed Calotropis procera (Aiton) W.T. Aiton (Asclepiadaceae) in Australia. The Rangeland Journal 38(1): 35-46. Merow C, Smith MJ, Silander JA. 2013. A practical guide to MaxEnt for modelling species’ distributions: what it does, and why inputs and settings matter. Ecography 36(10): 1058-1069. Naiman RJ, Decamps H, McClain ME, Likens GE. 2005. Biotic Functions of Riparia, California, 1st edition, Elsevier academic press, California, 1-488pp. Pidwirny M. 2018. Spatial distribution of species and Ecosystems. British Columbia, Our planet Earth publishing Quazi S, Mathur K, Arora S. 2013. Calotropis procera: An overview of its Phytochemistry and pharmacology. Indian Journal of Drugs 1(2): 63-69. Reddy M, Begum H, Sunil N, Pandravada S, Sivara N. 2015. Mapping the climate suitability using MaxEnt modelling approach for Ceylon Spinach (Basellaalba L.) cultivation in India. The journal of agricultural sciences 10(2): 87-97. Sharma GP, Kumar M, Raghubanshi AS. 2010. Urbanization and road-use determines Calotropis procera distribution in the eastern Indo-Gangetic plain, India. Ambio 39(2): 194–197. Uden DR, Allen CR, Angeler DG, Corral L, Fricke, KA. 2015. Adaptive invasive species distribution models: a framework for modelling incipient invasions. Biological Invasions 17 (10): 831-2850. Young N, Carter L, Evangelista P. 2011. A MaxEnt Model v3.3.3e Tutorial (ArcGIS v10). Colarado: Colarado State University.