Biodiversity hot-spot modeling and temporal analysis of ... · In 1997 Harini Nagendra from Center for Ecological Sciences, Bangalore and Madhav Gadgil from Biodiversity unite, Bangalore

INTERNATIONAL JOURNAL OF ENVIRONMENTAL SCIENCES Volume 4, No 5, 2014

© Copyright by the authors - Licensee IPA- Under Creative Commons license 3.0

Research article ISSN 0976 – 4402

Received on December 2013 Published on March 2014 772

Biodiversity hot-spot modeling and temporal analysis of Meghalaya using

Remote sensing technique Krishnanjan Pakrasi1, Arya V.S 2, Sudhakar S.3

1- Student (M.Tech Geoinformatics), Department of Environmental Science and Engineering

Guru Jambheshwar Univeristy of Science and Technology (Hisar, Haryana).

2- Senior Scientist-SG, Haryana Space Applications Centre (HARSAC)(Department of

Science and Technology, Government of Haryana).

3- Director of North Eastern Space Applications Centre, Umiam, Meghalaya, India Govt. of

India, Department of space.

[email protected]

doi: 10.6088/ijes.2014040404518

ABSTRACT

Biodiversity is variation of leaving creatures within an ecosystem, biome or an entire planet.

Health of an ecosystem can be measured through this variation. Remote sensing helps to

manipulate and analyze the image data produced by these remote sensors. The primary goal

of remote sensing is not only the obtain knowledge, but also its application of gained

knowledge. Satellites provide real-time global coverage of the Earth. Biophysical spectral

modeling techniques allow stratifying vegetation types based on the canopy closure (Roy et

al., 1996). Through this such approach mapping and monitoring of forest condition and

degradation processes can be done. This study has been taken up Meghalaya, state of India.

The flora of Meghalaya is the richest in India. A wide variety of timber species, medicinal

plants, and economically important plants are reported from this region. Due to lack of

systemic approach it is being destroyed by weed, firewood, timber, jhum cultivation. Large

scale over exploration has led to decrease in population of various orchids found here. Here

in this study LISS-III data are used.

Keywords: Classification, NDVI, Patchiness, Porosity, Interspersion, Shape index.

1. Introduction

The North Eastern Region (NER) of India combined of the states of Arunachal Pradesh,

Assam, Manipur, Nagaland, Mizoram, Sikkim and Tripura. The region stretches between 210

50’ and 29034’N latitude and 85034’ and 97050’ E longitude. The region has a population of

39 m and geographical area of 26.2 million hectare. On an average, the NE region receives

about 2450 mm of rainfall. The Cherrapunji-Mawsynram range receives rainfall as high as

11,500 mm, annually. Great variation in temperature regime can be fiend too in this region.

The temperature varies from 150C to 320C in summer and 00C to 260C in winter. At the

confluence of the Indo-Malayan, Indo-Chinese and Indian biogeographical realms, the NE

region is unique in providing a profusion of habitats, which features diverse biota with a high

level of endemism. The region is also the residence of approximately 225 tribes in India, out

of 450 in the country. North eastern region has been in focus for its high biodiversity and this

region has been a priority for leading conservation agencies of the world.

At global level many works has done on Biodiversity using Remote sensing and GIS. P.S

Roy et al. (1999) worked on Biodiversity Characterization at landscape level using

Biodiversity hot-spot modeling and temporal analysis of Meghalaya using Remote sensing technique

Krishnanjan Pakrasi et al, International Journal of Environmental Sciences Volume 4, No.5, 2014

773

Geospatial modeling Technique. Department of Space and Department of Biotechnology did

joint project (2002) on Biodiversity Characterization at Landscape Level in North-East India

using Satellite Remote Sensing and Geographic Information System. A.K. Skidmore, B.O.

Oindo, M.Y Said from ITC Netherland did Biodiversity Assessment by Remote sensing in

2002. “Remote sensing for biodiversity science and conservation” was published by Woody

Turner from NASA office of Earth Science, Sacha Spector from Center for biodiversity and

conservation, Ned Gardiner from Earth science division NASA in 2003. “Source book on

Remote sensing and Biodiversity indicators” was prepared in 2007 by jointly with NASA-

NGO Biodiversity and UNEP-WCMC. In 1997 Harini Nagendra from Center for Ecological

Sciences, Bangalore and Madhav Gadgil from Biodiversity unite, Bangalore studied on

Remote sensing as a Tool for Estimating Biodiversity. D.M. Debinski et al. (1998) worked

on Remote sensing and GIS based model for Habitats and Biodiversity in the Greater

Yellowstone Ecosystem. In 2003 B. B. Salem published his paper on Application of GIS to

biodiversity monitoring in Journal of Arid Environments.

2. Study area

The study “Modeling on Biodiversity hot-spot modeling and temporal analysis of Meghalaya

using Remote sensing technique” has been taken up Meghalaya state of India is a state in

north-eastern India. Geographically the area lies between coordinates 2500’00” N to

26010’00” N and 89045’00” E to 92045’00” E. Area of Meghalaya is about 8,700 sq mi

(22,720 km²). The population numbered 2,175,000 in 2000. The state is bounded on the north

by Assam and by Bangladesh on the south. The capital is Shillong also known as the

Scotland of the East, which has a population of 260,000. It was previously part of Assam, but

on 21 January 1972, the districts of Khasi, Garo and Jaintia hills became the new state of

Meghalaya. The area is rich in a wide range of flora and fauna. The vegetation is very

interesting having a mixture of Asiatic and Indian Peninsular elements. Meghalaya occupies a

unique position in the North Eastern Himalayas. More than 90 per cent of the forested areas

fall in private or community lands, the richness of biodiversity in these areas outside the state

owned lands cannot be regarded as permanent repositories. The chances of them being put

under shifting cultivation are inherently predominant (Singh, 2001). This state has two

national parks and three wildlife sanctuaries, the national parks covers an area of 267.48 km2

approx. These protected areas owned by the state forest department are diverse and support a

large number of species, which are important with respect to biodiversity.

Figure 1: Map showing study area (Meghalaya)



774

3. Objectives of the study

1) Preparation of vegetation cover type and land use maps using temporal satellite remote

sensing data to identify Vegetation changes over Meghalaya between 1997 and 2007.

2) Monitoring the biological rich area of Meghalaya. This study is carried out using IRS 1C-

LISS III images of Year: 1997 and 2007. ASTER DEM, Forest map of Meghalaya (Forest

and Environment Department of Meghalaya), SOI topo-sheets on 1:250,000, Rainfall and

Temperature data from Indian Meteorological Department have been used as Ancillary data.

4. Material and methods

This study is carried out using IRS 1C-LISS III images of Year: 1997 and 2007. ASTER

DEM, Forest map of Meghalaya (Forest and Environment Department of Meghalaya), SOI

topo-sheets on 1:250,000, Rainfall and Temperature data from Indian Meteorological

Department have been used as Ancillary data.

Methodology used to carry out this study is divided mainly into two parts i.e. Vegetation /

Land Cover Mapping and Monitoring Biological rich areas. The brief descriptions of the

methodology along with Flow Charts are given in figure 2 and 3.

4.1 Methodology for Vegetation / Land Cover Map:

Satellite images are geo-rectified from 1:250000 SOI topo-sheets and mosaic together.

Mosaic file is further subseted by administrative vector boundary (figure. 1) of Meghalaya.

NDVI values are extracted and classified. Then main subseted image have classified by

hybrid classification technique and classified NDVI image used as vegetation mask (figure.

4) based on knowledge based approach. Extracted administrative boundaries, communication

and settlement layers are superimposed on classified vegetation layers.

Figure 2: Methodology for Vegetation Land Cover Map

Methodology for Monitoring Biological rich areas



775

Figure 3: Methodology for Monitoring Biological rich areas

The classified vegetation and land cover images are further classified in forest and non forest

areas. Extracted Bio rich areas from satellite images (work done previously by NESAC) have

used as mask. After this landscape analysis like patchiness, porosity, and interspersion has

done. Patchiness is a measure of the density of patches of all types (Romme, 1982). The

vegetation/land cover type map was reclassified into two classes’ viz., forest and non-forest.

A grid cell of n x n is convoluted through the entire spatial layer.

(1)

Where N is the number of boundaries between adjacent cells, and Di is dissimilarity value for

the ith boundary between adjacent cells. Count of density or number of patches of patches

within single land cover type. Lower porosity values indicate less interaction among the

landscape elements and higher porosity values indicate higher interaction among the

landscape elements and heterogeneity resulting in high fragmented landscape.

(2)

Where Cpi is the number of closed patches of ith cover class.

Interspersion is the measure of spatial intermixing of land use/ land cover. It is calculated by

the number of surrounding grid cells that differ from the central cell. It shows the

homogeneity of the landscape and landscape diversity. For analyzing interspersion 3X3 grid

has been moved over the classified images.

(Lyon 1983) (3)

Where SFi is shape factor.

Calculations of disturbance patches are calculated by shape index value through object base

approach. The smoother the border of an image object represents lower its shape index. It is



776

calculated from the Border Length feature of the image object divided by four times the

square root of its area. Smoother shape represents less disturbed patch and irregular shape

indicate high disturbed patch.

(4)

Where bv is the image object border length and is the border of square with area.

Figure 4: Derived Vegetation mask

Figure 5: Showing the vegetation type of 1997



777

Figure 6: Showing the vegetation type of 2007

Figure 7: Fragmentation 1997



778

Figure 8: Fragmentation 2007

Figure 9: Homogeneity 1997



779

Figure 10: Homogeneity 2007

Figure 11: Porosity 1997



780

Figure 12: Porosity 2007

Figure 13: Snowing disturbance patches in 1997



781

Figure 14: Snowing disturbance patches in 2007

Table 1: shows the changes between 1997 and 2007

Class Area in ha (2007) Area in ha (1997) Changes (Area in ha)

Semievergreen 479250.67 665553.13 186302.46

Evergreen 312999.18 428611.72 115612.54

Sal 12899.3 10002.73 -2896.56

Bamboo 13329.82 25661.72 12331.9

Mixed pine 116019.77 104281.25 -11738.52

Pine 26867.28 52501.56 25634.28

Degraded forest 136816.31 50974.61 -85841.7

Agriculture 64532.85 28810.16 -35722.7

Abandoned Jhum 737562.74 461928.13 -275634.62

Grassland 329641.81 390532.42 60890.61

Water body 8696.12 13433.98 4737.87

Table 2:Fragmentation 1997 Table 3: Fragmentation 2007

Category Area in ha Category Area in ha

High 5812.25 High 117258.2

Medium 7080.1 Medium 514193

Low 5176.47 Low 757079.7

Table 4: Homogeneity 1997 Table 5: Homogeneity 2007


Very high 557112.1 Very high 664682.2

High 516328.5 High 617746.8



782

Low 316877.7 Low 410197.9

Very Low 557145 Very low 856440.2

Table 6: Porosity 1997 Table 7: Porosity 2007


High 360139.8 High 441422

Medium 235832 Medium 298583.1

Low 354931.3 Low 854694.9

Table 8: Disturbance area in Biological richness

Area in sq km

Patch Category Year 1997 Year 2007 Area change (Sq Km)

Highly disturbed 132.039 223.54 91.5

Very highly disturbed 101.83 152.2 50.37

less disturbed 51.89 213.51 161.62

Very less disturbed 82.81 85.47 2.65

5. Result and conclusion

Forests of Meghalaya have been classified into Sub-Tropical-Evergreen, Semi-evergreen,

Grasslands, Sal forest, Bamboo forest, Mixed pine forest, Pine forest, Degraded forest,

Agriculture, Abandoned Jhum, Water body. From the comparative study between 1997 and

2007 (figure.5 and 6) it is found that 325246 ha forest areas are reduced within ten years. In

the other hand Jhum cultivation is increased about 275634 ha. Following table no.1 is shown

the changes.

The parameters like fragmentation, porosity, and interspersion have been used to understand

the disturbance regime and biologically rich areas. Forest fragmentation is one of the greatest

threats to biodiversity in forests, especially in the tropics. The effect of fragmentation on the

flora and fauna of a forest patch depends on a) the size of the patch, and b) its degree of

isolation. From the compare study between 1997 and 2007 (figure. 7 and 8) it can be seen

that total fragmentation is highly increased (table 2 and 3). Forests under the private land

parcels are fragmented more. Re bhoi, East khasi hills, jaiantia hills are more affected by

forest fragmentation. Human land uses, like shifting cultivation, tend to expand over time in

this area.

Interspersion represents the spatial intermixing or homogeneity (Figure 9 and 10) among the

landscape features. Higher value of interspersion means dispersal ability of the central class

will be low or reduced or in other words the influence of resistance by neighbors will be

much which may lead to the extinction of the central class.

Porosity shows the overall clue of to the change of species isolation between 1997 and 2007.

In the maps (figure 11 and figure 12) Lower porosity value indicates the lower interaction

among landscape elements, homogeneity and low fragmented habitats in the forest class and

higher porosity value indicates the higher interaction among landscape elements,

heterogeneous and high fragmented habitats in the forest class. During ten years period the

porosity is increased in the state (table 6 and 7).



783

It is previously mentioned that category of disturbance patchs in biorich area is calculated by

shape index value. These parameters were derived using detailed object based analysis. The

disturbance type has been derived using masking and classification technique. For the better

analysis patch categories are divided into highly disturbed, very highly disturbed, less

disturbed, very less disturbed. Very highly disturbed area represents man made features,

highly disturbed area represents Jhum cultivation and agricultural areas, less disturbed area

represents natural vegetation cover changes, and very less disturbed area represents degraded

forest, grassland etc ( Figure 13 and 14). Bellow table 8 is shows the changes between 1997

and 2007.

From the above study it is clearly seen that Biodiversity of Meghalaya is under great threat

due to impact of human activities. Old practice of shifting cultivation, over exploitation of

forests and mining activities are responsible for loss of biodiversity. Permanent agriculture in

the higher plateaus has also resulted in the loss of forest cover. From the study of this project

it is found that in 1997 there were 57.47% forest cover in Meghalaya but in after ten years, in

2007 forest cover reduced at 42.92%. It is clearly seen that the jhum cultivation is increased

about 12% during ten years. It is also observed that fragmentation of landscape has been one

of the major causes of biodiversity loss. Disturbance patch in biorich area in 1997 was 368.56

km2 and in 2007 it increased almost double at 674.72 km2 which affects the Biologicaly rich

areas of Meghalaya. The species like Diospyros undulata, Nymphaea pygmaea and Luvunga

scandens are thought to be locally extinct (Khan et al., 1997). Important bio-rich sites

identified in Meghalaya are Nokrek Biosphere reserve and Balphakram National Park. So it is

recommended that immediate action must be taken to control human impact on forest

resources especially in biorich areas.

Acknowledgement

I am very thankful to Mrs. Kasturi Chakroborty Scientist D and Dr. K.K. Sharma form North

Eastern Space Applications Centre. I am very grateful to them for providing timely

suggestions and sorting out implementation difficulties. Also thanks to the research fellows

of this organization. I give special thanks to Dr. S.Sudhakar Director of the same organization

that he gives me his valuable time to guide me and teach about scenarios of Biodiversity in

Meghalaya and Dr. V.S. Arya Scientist SG from Haryana Space Applications Centre who

inspire me to write this paper. Thanks to all the members of North Eastern Space

Applications Centre for cooperation with me every time and provide me logistical supports.

6. References

1. Report: Biodiversity Characterization at Landscape Level in North-East India using

Satellite Remote Sensing and Geographic Information System (2002), Department of

Space and Department of Biotechnology. ISBN - 81-901418-0-5, pp 313.

2. Report: State of the Environment Report, Meghalaya, (2005), Department of

Environment and Forests, Government of Meghalaya, pp 76.

3. Report: State of Forest Report 2005, Forest Survey of Inda, Dehra Dun, pp 87.

4. Annual Report 2009 – 10, ICAR Research Complex for NEH Region, Dr. S. V.

Ngachan, Dr. N. S. Azad Thakur, Dr. Anupam Mishra, Dr. A. Pattanayak, pp 252.



784

5. Report: Sourcebook on Remote Sensing and Biodiversity Indicators (2007): NASA-

NGO Biodiversity Working Group and the World Conservation Monitoring Centre of

the United Nations Environment Programme, ISBN: 92-9225-072-8, pp 203.

6. Technical Report: The impacts of climate change on biodiversity in Caribbean

islands, Caribbean Natural Resources Institute. Report No.386, pp 28.

7. Policy Notes for Parliamentarians July-August (2010), The Human Impact of Climate

Change in India Advancing political leadership. Research and Coordination: Freya

Seath and Pallavi Agarwal, pp 94.

8. Antenucci J.C., Brown K., Croswell P.L. and Kevany M. (1991), Geographic

Information Systems: A Guide to the Technology, Van Nostrand Reinhold, New

York.

9. Burnett M.B., August J. Brown and Killingbeck K.T., (1998). The influence of

geomorphological heterogeneity on biodiversity: I. A patch-scale perspective.

Conservation Biology, 12, pp 363-370.

10. Bujarbarua P and Baruah (2009), Vulnerability of fragile forest ecosystem of North

East India in context with the global climate change: an ecological projection, Earth

and Environmental Science, 6, doi:10.1088/1755-1307/6/7/072016.

11. Das P.J., (2004), Rainfall Regime of Northeast India: A Hydrometeorological Study

with Special Emphasis on the Brahmaputra Basin. Unpublished Ph.D. Thesis, Gauhati

University (accessed October 25, 2009).

12. Debinski D. M. and Kindscher K., (1999), A remote sensing and GIS-based model of

habitats and biodiversity in the Greater Yellowstone Ecosystem. International journal

of Remote sensing, 20(17), pp 3281- 3291.

13. Delcourt H.R. and Delcourt P.A., (1988), Quaternary Landscape ecology: relevant

scales in space and time, Landscape Ecology, 2, pp 23-44.

14. Hannah L and Lohse D, (1993), Annual Report, Conservation Institutions,

Washington D.C.

15. Lidicker W.Z.J., (1995), The landscape concept: something old, something new. In:

Lidicker, W.Z. Jr. (ed.) Landscape Approaches in Mammalian Ecology and

Conservation, University of Minnesota Press, Minneapolis, pp 3- 19.

16. MacKinnon J. and MacKinnon K., (1986), Review of the Protected Areas System in

the Indo-Malayan Realm, International Union for the Conservation of Nature and

Natural Resources, Gland, Switzerland and Cambridge, U.K, 284.

17. Majumdar K., Tiwari K.C., Tiwari, Bhattacharjee and Nair A.P, (1979), Some

folklore medicines from Assam and Meghalaya, J. Crud. Drug. Res. 172, pp 275-281.



785

18. Mittermeier R.A, (1988), Primate diversity and the tropical forest: Case studies from

Brazil and Madagascar, and the importance of the megadiversity countries, In: E.O.

Wilson (ed.), Biodiversity. National Academy Press, Washington, D.C.

19. Miller R.I, (1994), Mapping the Diversity of Nature. Chapman and Hall, London.

20. Nagendra H and Gadgil M, (1997), Remote Sensing as a tool for estimating

biodiversity, Journal of Spacecraft Technology, 7 (2), pp 1-10.

21. Nayar M.P. and Shastry A.R.K. (eds.), (1987), Red Data Book of Indian Plants, 1,

Botanical Survey of India, Calcutta, pp 367.

22. Nicholas, David, (1998), Keyword bibliography of ethnoarcheology and related

topics, Ver 3.1.

23. Peters Debra P.C., Sarah C. Goslee, (2001) Landscape Diversity, In S.A. Levin.

editor(s), Encyclopedia of Biodiversity, Academic Press, New York, pp 645-658.

24. Roy P.S. and Tomar S., (2000), Biodiversity characterization at landscape level using

geospatial-modeling technique, Biological Conservation, 95(1), pp 95-109.

25. Romme W.H., (1982), Fire and landscape diversity in subalpine forests of

Yellowstone National Park, Ecological Monographs, 52(2), pp 199-221.

26. Salem B. B, (2003), Application of GIS to biodiversity monitoring, Journal of Arid

Environments, 54, 91–114.

27. Singh B, (2001), Comments on the report, Biodiversity Characterisation at landscape

level using Remote Sensing and GIS in Meghalaya, Report of DOS-DBT Phase-I,

Personal communication, D.O. No. MFG/85/2/93 dated 20/9/2001.

28. Singh H.B. and R.K. Arora, (1978), Wild Edible Plants of India (1st Ed.), ICAR

Publication, New Delhi, p 88.

29. Tomlin C.D., (1990), Geographic Information Systems and Cartographic Modelling,

Prentice-Hall, New York.

30. Tripathi RS, Pandey HN and Tiwari BK., (1996), State of Environment of

Meghalaya. North –Eastern Hill University, Shillong -22.

31. Turner W, Spector S, Gardiner N, Fladeland M, Sterling E and Steininger M 2003:

Remote sensing for biodiversity science and conservation. TRENDS in Ecology and

Evolution,18(6), pp 306-314.

Biodiversity hot-spot modeling and temporal analysis of ... · In 1997 Harini Nagendra from Center for Ecological Sciences, Bangalore and Madhav Gadgil from Biodiversity unite, Bangalore

Documents