Home Articles Evaluating the change (1968-2001) in landscape pattern and analyzing disturbance in Baratang...

Evaluating the change (1968-2001) in landscape pattern and analyzing disturbance in Baratang Forest Division (Andaman Islands), South East Asia

Dr Nidhi Nagabhatla
Post Doctoral Scientist- Landscape Ecology/ Remote Sensing
International Water Management Institute (IWMI)
Colombo, Sri Lanka
[email protected]

P.S Roy
National Remote Sensing Agency,
Balanagar, Hyderabad-500 037, India

Rajendra Jagdale
S&T Park, University of Pune,
Pune – 411 007, Maharastra, India.

As the 21st century unfolds, accompanied is the range of long term environmental problems forecasting immense consequences for the economic well being of nature. The present approach draws attention towards evaluating the change in landscape pattern and analyzing disturbance in Baratang Islands caused as a result of human activity triggering long-term detrimental effects like climate change, global warming, variation in rainfall pattern, temperature fluctuation and ecosystem imbalance. Advancement in spatial technology has opened new vistas to understand the fundamentals of the complex interrelations between human activity and natural forces. Ecological effects of land use change are tools for anticipating and evaluating the complex interactions between humans and ecological systems. Changes in vegetation cover over time are important spatial data to assist in understanding the change in natural processes and the influence of human contributions, or its subsequent impact, on ecological condition. Natural ecosystems of the Andaman Islands are under severe pressure due to population explosion, forest fragmentation and other natural causes. The forest vegetation with high species diversity is noteworthy but due to human intervention the forest area has remarkably shrunk and ecological balance splintered. Estimating the change in landuse/ landcover over time assist in understanding the landscape dynamics and the influence of human contributions on ecological complexity. The use of two time period data i.e. 1986 and 2001 facilitated in preparation of vegetation cover map and the change matrix analysis was done to find out the change in each type. The change detection analysis reveals that forest vegetation complimented with high species diversity is noteworthy but due to human interventions the forest area has remarkably shrunk. Change detection analysis also highlights the change in landscape pattern and spotlights two major issues of concern viz., the conversion of forest into agriculture and secondly the transformation of primary forest into secondary formations. It also reveals that significant correlation exist between change in natural boundaries of vegetation communities and the causal agents viz., population pressure, extraction, development, and plantation activities.

Humans have altered the natural environment far as evidences provide us (Thomas, 1956). Increasing population growth coupled with unsustainable resource use is a major factor responsible for change in landscape pattern. Initially ecological approaches started with study on biogeographical patterns and were mostly of taxonomic nature. Ecological science plays an important role in broadly understanding the structure and function of nature. The branch has evolved a science of human sustenance and sustainability (Singh, 2004). The increasing interest in ecosystems spatial dynamics has led the need for new quantitative methods capable of analyzing patterns, determining the importance of spatial processes and developing landscape models (Turner and Gardner, 1991). Williams, (1993) advocated that land use/land cover change acts as a surrogate for the terrain interest points. The information generated on landscape change and configuration facilitates to answer key management questions and help to analyze global ecological and environmental change (Lunetta, 1998). Blaschke et al., (2001) has emphasized the need to measure landscape dynamics quantitatively by incorporating landscape pattern and change from satellite image analysis. Change detection is the measure of landscape-temporal evolution involving comparison of a pair of images to identify areas that have distinctly different brightness values. New images representing change are created by taking the difference between images. Singh, (1989) accounted many change detection procedures including principal component analysis and comparison of classified images. Roy et al., (1996) explains that basic premise of change detection is that spectral signature change commensurate with change in land cover. The information on landscape dynamics highlight specific threats to vegetation and anticipate some of the future threats.

Andaman Islands placed under a biogeographical separate zone (10A) (Rodgers and Panwar, 1988) are known to be emergent remnants of a tertiary mountain chain, the Andaman ridge (Ahmad, 1982).These winds swept, isolated, fragile, oceanic islands have volcanic foundations and are characterized by change of sea levels. Rich species diversity is largely attributed to soil variability, climate and colonization of plant species from widely different territories. Luxuriant vegetation shields the island environments from the full amplitude of continental climatic fluctuations. In recent times, the island ecology has undergone serious changes owing to various causes. Tourism has adversely affected sea turtle nesting beaches and the overall ecology (Andrews et al., 2002). Clearing vegetation for cultivation has exposed the soil to high-intensity rainfall resulting in has led to erosion. In turn, the sediments have smothered the mangrove and coral ecosystems (Sirur, 1999, In Andrews and Sankaran, 2002). Kothari et al., (1989), explained the vast encroachments and extensive honeycombing of the forests in Saddle Peak NP (North Andaman).

Land use change and extraction practices offers a potential measures for studying its dynamic. Tracing the historical records it was noted that till the middle of 20th century no large intrusion was made into Baratang Islands apart from movement of Jarwas while Ritchie’s archipelago was totally free from any disturbance activities. The colonization started with the penal settlement in 1970 wherein, forest clearance activities resulted in vegetation cover change, starting the first phase of forest degradation. The road construction in 1980’s added to the worse. The developmental activities, increased settlements, agriculture expansion land and the passing of ATR has resulted in increasing the buffer zone of disturbance. However, the remote islands still harbor intact patches of evergreen forest with well-defined storeys and the canopy density of more than sixty percent. Bay islands have undergone an observable change after independence. The changes pertaining to developmental activities, administrative infrastructure, tribal upliftment and economic sustainability have been somewhere responsible for change in the vegetation pattern in these islands. In order to conserve the left intact patches of the tropical forest in the isles, regular monitoring of the landscape is of paramount importance. A comprehensive databases mentioning the causes and the statistics of change will help us to evaluate the change in vegetation cover in the region. Sharma, Naik and Pandian (1988) in their study have visually interpreted major vegetation types in the division using Ariel photographs (acquired by the forest division in 1968). The present study however has made an effort to compare the change in the major vegetation classes using temporal data (1986 & 2000). The vegetation classes have been reframed and the area normalized according to the previous study. Analysis was opted for major communities in the region.

In the territorial division of Baratang Islands three out of the twenty-eight are inhabited viz. Baratang, Havelock and Neil. These islands were subjected to maximum disturbance and anthropogenic pressure in the last few decades. In the present study these islands were separately analyzed for change in forest and non-forest area. Monitoring change over time using multi-temporal imagery is an important application of spatial technology. Collins and Woodcock (1996) advocated methods for monitoring vegetation change ranging from intensive field sampling with plot inventories to extensive analysis of remotely sensed data. While aerial photography can detect change over relatively small areas, satellite imagery has proven cost effective for large regions. Change detection involves comparison of a pair of images to identify areas that have distinctly different brightness values. The signature trajectories are tracked over time and combined with site-specific rule-base techniques to determine the vegetation trend. In the present attempt, change detection technique is applied to depict information of landscape level changes in vegetation extent in last one and half decades (using two-time period Landsat TM data (1986 & 2001)). Visual interpretation approach was used to identify the changes in the natural boundaries of the vegetation communities and to derive a cover change map (Fig.1). Image elements are significantly taken into consideration during the process. Areas of classified change are related to cover type by GIS overlay analysis by intersecting change classes with cover types and quantifying the change.

Fig 1 Analyzing landuse cover change using temporal data sets

In Baratang Islands the change is the vegetation cover in prominent on the western coast while southern and eastern coast has less human intrusion. A net decrease in primary forest cover (including evergreen, semi-evergreen, deciduous, mangrove, littoral and bamboo), which is attributable to increased population and consequent higher pressure on the forests, accounts to 2.34 per cent. Mangrove forest of the region is one of the most important coastal ecosystems. The status of mangrove forest in 1986 was 113.78 sq km while in 2002 it is analyzed as 112.27 sq km, which shows a very insignificant change (0.58 sq km of the area has been converted to degraded mangrove due to coastal activities and tidal fluctuations)(Table 1& Fig 2&3). The non-forest classes in the islands viz. settlements and agriculture has shown in increase in last two decades. The area under each type depicts that tropical moist deciduous forests occupy a large area at present compared to previous years. The total area under primary evergreen forest in 1986 was 129.9 4.sq km (22.36 %) whose status has reduced by 8.7 sq km (Table 2).

Fig 2: Vegetation cover derived in Baratang Island using two time period data

Fig 3 : Change in the vegetation cover is prominent on the western coast in Baratang Islands Vegetation classesArea (2001) Area (1986)Change in Area sq km )% of the total area
Giant evergreen forest 29.13 32.73 -3.60 -0.64
Andaman evergreen forest 89.24 94.21 -4.97 -0.88
Secondary evergreen forest 36.89 27.81 +9.08 +1.60
Semi -evergreen forest 157.89 171.38 -13.49 -2.38
Moist deciduous forest 67.75 55.38 +12.37 +2.18
Bamboo 2.81 3.51 -0.70 -0.12
Mangrove 112.27 113.78 -1.51 -0.27
Littoral 1.99 3.58 -1.59 -0.28
Teak 13.91 20.76 -6.85 -1.21
Padauk 1.82 1.22 +0.60 +0.11
Degraded forest 1.80 0.71 +1.09 +0.19
Degraded mangrove 1.67 1.09 +0.58 +0.10
Mudflats 10.35 9.42 +0.93 +0.16
Sandy beaches 1.51 1.35 +0.16 +0.03
Coconut 5.61 3.93 +1.68 +0.30
Agriculture 31.91 25.69 +6.22 +1.10

Table 1 : Vegetation cover types in Baratang Islands estimated using Landsat of 1986 and 2001

It can also be inferred that the loss in primary forest can be compensated with the increase in secondary formations. Relatively more conversion of teak plantation into semi-naturalized deciduous forest was noticed in the central portion of the island. Havelock Island has undergone noticeable changes in last one and a half decade. The evergreen forest had undergone a decrease of 3.46 %. The mangrove forest area has reduced from a total of 12.51 sq km in 1986 to 11.77 sq. km in 2001. Neil Island has undergone outstanding changes in the evergreen forest cover. Extraction activities coupled with increasing population ahs led to the conversion of evergreen cover to semi-evergreen. At places encroached land has been completely cleared for expansion in agricultural activities (increase by 5.93 %). The littoral forest on the northern coast of the islands has also show slight decrease (0.15 sq km) (Table 3 ).

Change from 1986 to 2001Area in (sq km )% of the total area
Evergreen forest to Secondary forest 8.57 1.51
Evergreen forest to degraded forest 0.52 0.09
Semi evergreen forest to moist deciduous 5.42 0.96
Semi evergreen forest to degraded 0.58 0.10
Semi evergreen forest to agriculture 6.22 1.10
Teak to mixed Teak -Padauk plantation 0.60 0.11
Semi evergreen forest to coconut plantation 0.27 0.05
Mangrove to degraded mangrove 0.58 0.10
Mangrove to mudflats 0.93 0.16
Littoral to coconut plantation 1.43 0.25
Teak to Moist deciduous 6.22 1.10
Littoral to increase in sandy beaches 0.16 0.03
Bamboo to Moist deciduous 0.70 0.12

Table 2: Vegetation cover transformation in Baratang Islands from 1986 to 2001

Change in primary forest (1968-2001): The analysis indicates that area has undergone crucial changes in the primary forest pattern. The plantation area showed remarkable increase between 1968 and 1986, but not much from 1986 to 2001. The scientific approach in management practices after 1970’s have emphasized on mangrove conservation and regeneration therefore the area of this eco-fragile formation shows stability in occurrence. It was also observed that evergreen and semi-evergreen forest had been major forest type in the region before the rise in settlement and agriculture activities. Due to progression in plantation practices large stretches for primary forest was cleared from between 1970 -1980, majority of it in the main Baratang island resulting in shrinkage of evergreen forest. Extraction activities for the fulfilment of timber requirement have created gaps leading to increment in the percentage of secondary forest in the region mostly after 1980’s.During the last four decades many fold increase in area under agriculture and settlements was observed (Table 4 & 5, Fig 4 & 5).

Fig 4 Change in primary forest and plantation pattern in Baratang islands Baratang IslandVegetation classesArea (2001)Area (1986)Change Status% of the total areaHavelock IslandNiel Island
Evergreen forest 48.11 58.47 -10.36 -4.98
Secondary evergreen forest 26.19 16.84 +9.35 +4.50
Semi-evergreen forest 43.98 53.93 -9.95 -4.78
Moist deciduous forest 50.15 39.16 +10.99 +5.28
Mangrove 46.73 48.47 -1.74 -0.84
Bamboo 0.53 0.83 -0.3 -0.14
Littoral 0.52 0.85 -0.33 -0.16
Teak 15.18 19.43 -4.25 -2.04
Padauk 1.03 0.74 +0.29 +0.14
Coconut 0.38 0.16 +0.22 +0.11
Degraded forest 1.14 0.41 +0.73 +0.35
Degraded mangrove 0.91 0.36 +0.55 +0.26
Mudflats 6.23 4.98 +1.25 +0.60
Agriculture 17.29 13.84 +3.45 +1.66
Evergreen forest 26.38 30.22 -3.84 -3.46
Secondary evergreen forest 12.21 9.09 +3.12 +2.81
Semi-evergreen forest 42.27 45.29 -2.99 -2.70
Moist deciduous forest 0.86 0.41 +0.45 +0.40
Mangrove 11.77 12.51 -0.74 -0.67
Littoral 1.31 2.59 -1.28 -1.15
Degraded forest 0.82 0.29 +0.53 +0.48
Degraded mangrove 0.58 0.26 +0.32 +0.29
Mudflats 1.39 0.83 +0.56 +0.50
Coconut 5.15 3.11 +2.04 +1.84
Agriculture 8.38 6.54 +1.84 +1.66
Evergreen forest 3.59 4.78 -1.19 -8.02
Semi-evergreen forest 3.25 2.96 + 0.29 +1.95
Mangrove 0.65 0.91 -0.26 -1.75
Littoral 0.21 0.36 -0.15 -1.01
Mudflats 0.36 0.13 +0.23 +1.55
Coconut 0.52 0.32 +0.20 +1.35
Agriculture 6.26 5.38 +0.88 +5.93

Table 3 : Change in the vegetation pattern of three major inhabited islands (1986-2001)

YearsRate of forest loss sq km/year)Rate of increase in population (persons/year)
1968-1986 0.66 465.63
1986-2001 0.27 213.07

Table 4 Rate of change of forest cover and anthropogenic influence

1968-2001Major Vegetation Types *Arial Photography Satellite data
Satellite data
Evergreen forest 132.43 106.39 102.37
Secondary evergreen 16.11 23.81
Semi-evergreen forest 166.11 159.72 152.24
Moist deciduous forest 83.36 89.45 90.42
Mangrove 106.69 104.33 103.09
Plantations (Teak, Padauk, Coconut) 9.13 13.51 16.32
Agriculture 9.63 17.85 19.69

Table 5 Change in vegetation pattern of Baratang Islands from 1968-2001

The change detection analysis provides insight to ecosystem functioning and stability and puts forth the landuse impacts. The present approach is based on analyzing the change in landscape pattern using onscreen visual classification approach. In the present study vegetation cover change was analyzed using two-time period (1986 & 2001) Landsat TM data. It significantly highlight change in landscape pattern in Baratang Islands, as forest cover under major vegetation classes (evergreen, semi-evergreen, moist deciduous, littoral and mangroves) in 1986 was ascertained as 495.16 sq km while in 2001 it reduced to 471.93 sq km. All major primary forest types had undergone a change pertaining to the growing population pressure and developmental activities in the region. The lack of scientific approach in management practices has been one of the causes for increase in deciduous cover and conversion of evergreen to semi-evergreen type. Teak plantation in Baratang Island was not a successful approach but has led to permanent change in the identity of the primary forest, which used to exist there.

Fig 5 Change in vegetation pattern in Baratang Islands

The human interventions and extraction activities are the major causes of increase in secondary evergreen forests (which increased from 27.81 sq km (1986) to 36.89 sq km (2001)). The expansion of agriculture land from 25.69 sq km (1986) to 31.91 (2001)) sheds light on population accommodation at the cost of forest resources. The mangrove vegetation indicates positive sign of regeneration and conservation. The study reveals that Baratang, Havelock and Neil Islands have undergone considerable modification in primary forest cover (Fig 5.1). The settlers have spread their domain in all directions marginalizing the indigenous people (the Jarawas). Roy et al., (1991) attempted to monitor land cover transformation during 1968-1986 in Baratang Islands concluded that change in spatial vegetation has occurred due to increased anthropogenic as well as natural disturbances. The scientific approach in management practices after 1970’s emphasized on mangrove conservation and regeneration therefore the area under mangrove forest shows stability in occurrence. The visual classification approach followed to analyze the change depict a gradual transition from natural to artificial cultivated systems initiating environmental degradation. About ten percent decrease in forest cover was analyzed. The conversion of primary types into secondary formations and increase in degraded forest is accounted to clear felling of forests for timber besides; stone mining has altered the natural landscape to a great extent. The changes have brought in impacts like, fragmentation, loss of biodiversity and degradation of rich biodiversity sites.

Information related to change in existing vegetation pattern outlines successional changes taking place in Baratang Islands in last decades highlighting the main causes of disturbance in the region being anthropogenic influence. The outputs generated will aid in creating attribute information of the composite strata. The study sets to establish peculiarities of island biota, particularly remote islands, reflecting their ecological significance. The basic important conclusion drawn from this comparative account of two different studies was that a regular monitoring and mapping of the landscape structure is the prerequisite step for any conservation programme.

Acknowledgment: We wish to thank Department of Space and Department of Biotechnology, Government of India for financial support for the study.


  • Ahmad E. 1982, India, coastal morphology, The Encyclopedia of Earth Sciences, (ed) Schwartz M.L, The Encyclopedia of Beach and Coastal Environments, 15, 481-484
  • Andrews Harry.V and Sankaran Vasumathi (eds); 2002 Sustainable Management of Protected Areas in the Andaman and Nicobar Islands. And ANET, IIPA, and FFI Uthra Print Communications, New Delhi.
  • Blaschke,T.,Conradi, M., Lang, S. 2001, Multi-scale image analysis for ecological monitoring of heterogeneous, small structured landscapes. Proceedings of SPIE, Toulouse, 35-44
  • Collins B. and Woodcock C.E., 1996, An Assessment of Several Linear Change Detection Techniques for Mapping Forest Mortality Using Multitemporal Landsat TM Data. Remote Sensing of Environment, 56, No. 1, 66-67
  • Kothari, A., Pande P., Singh S, and Variava D. 1989. Management of National Parks and Sanctuaries in India, A Status Report, Environmental Studies Division, Indian Institute of Public Administration, New Delhi
  • Lunetta R.S and Elvidge.C.D (eds), 1998, Remote Sensing Change Detection: Environmental Monitoring Methods and Applications”. Ann Arbor Press, 318
  • Rodgers, W. A. and Panwar, S. H., 1988. Biogeographical classification of India. New Forest, Dehra Dun, India.
  • Roy P.S., Tomar S, Pujar G, Saran S,. Maslekar A.R and Rikimaru A., 1996. Forest density stratification in Tropical Forests -A case study in South Andaman Islands (India). Paper presented in 2nd International Tropical Timber Organisation/Japan Overseas Forestry Consultant Association, Bangkok, (Thailand), Aug. 1-3, 1996.
  • Singh. J. S, 2004, Ecology and the Environment: Issues and Vision for India, Presidential Address, Indian Science Congress
  • Thomas, William L. 1956: Man’s Role in Changing the Face of the Earth, Chicago, London
  • Turner, M.G., and R.H. Gardner (eds.). 1991. Quantitative methods in landscape ecology. Springer-Verlag, New York.
  • Williams M.R. 1993, Land-use changes and limnological conditions in the central Amazon floodplain using remote sensing and field data. Speaker at the Fourth Annual Walker Branch Watershed Research Symposium, Oak Ridge, TN.
  • Sinha, P.C. 1998, Protected Area Networks. Anmol Publications. 290