J Coast Conserv (2012) 16:1–12

DOI 10.1007/s11852-011-0162-3

Identification of hot spots and well managed areas

of Pichavaram mangrove using Landsat TM
and Resourcesat—1 LISS IV: an example of coastal
resource conservation along Tamil Nadu Coast, India
T. Srinivasa Kumar & R. S. Mahendra &
Shailesh Nayak & K. Radhakrishnan & Kali C. Sahu

Received: 25 September 2009 / Revised: 18 July 2011 / Accepted: 21 August 2011 / Published online: 3 September 2011
# Springer Science+Business Media B.V. 2011

Abstract The present work is a multi-temporal satellite their temporal variations, it is possible to get a reasonably
based study on the spatial dynamic of an important coastal accurate picture of the extent and condition of the
habitat, the Pichavaram mangrove ecosystem, over a period mangrove ecosystem. The total area of the Pichavaram
of 15 years. The Pichavaram mangrove forest near mangrove showed a net increase of 2.51 km2 within a span
Chidambaram, South India is the second largest mangrove of 15 years (1991 to 2006). The hot spots that are at a risk
forest in the world. Unsupervised classification, the of being degraded, and on the other hand, the mangrove
Iterative Self Organising Data Analysis Technique (ISO- areas that are well managed are identified using Geographical
DATA), has been used to classify the mangrove cover into Information System (GIS) tools for the restoration and
the open and dense classes. The status of the classes has conservation measures.
been monitored using Landsat TM of 1991, 2001, and
Resourcesat–1 LISS IV of 2006 satellite data. The study Keywords False colour composite . Well managed areas .
demonstrated that by classifying mangrove ecosystem into Spatial variability analysis . Inter-class variability analysis .
just the 3 classes using remote sensing data and by studying Coast . Geographical information system

Abbreviations
T. Srinivasa Kumar (*) : R. S. Mahendra ERDAS Earth Resources Data Analysis Systems
Indian National Centre for Ocean Information Services (INCOIS), ESRI Environmental Systems Research
Hyderabad 500 055, India
e-mail: [email protected]
Institute
FCC False Colour Composite
R. S. Mahendra GIS Geographical Information System
e-mail: [email protected] GOES Global Observatory for Ecosystem Services
IRS-P6 Indian Remote Sensing Satellite
S. Nayak
(Resource Satellite 1)
Ministry of Earth Sciences (MoES),
New Delhi 110003, India ISODATA Iterative Self Organising Data Analysis
e-mail: [email protected] Technique
Landsat MSS Land satellite Multispectral Scanner
K. Radhakrishnan
System
Vikram Sarabhai Space Centre (VSSC),
Thiruvananthapuram 695022, India Landsat TM Land satellite Thematic Mapper
e-mail: [email protected] LISS Linear Imaging Self-Scanner
NRSA National Remote Sensing Agency
K. C. Sahu
Resourcesat Resource satellite
Department of Marine Science, Berhampur University,
Berhampur 760 007, India RGB Red Green Blue
e-mail: [email protected] SPOT Satellite Pour l’Observation de la Terre
2 T. Srinivasa Kumar et al.

Introduction International organizations and government agencies in

several countries are, for this reason, urgently implementing
Mangroves are the salt-tolerant evergreen forests found mapping and monitoring programs to measure the extent of
along sheltered coastlines, shallow-water lagoons, estuaries, the decline of these important ecosystems (Linneweber and
rivers or deltas in one hundred and twenty four tropical and de Lacerda 2002; Barbier et al. 2004).
subtropical countries, mainly growing on soft substrates
(FAO 2007). Mangrove ecology and management have been
described by many authors over time and the literature on this Status of global mangroves
subject is substantial. Mangroves are one of the world’ s most
productive vegetation communities. By virtue of these habitats The global mangrove coverage has been variously estimated
located in coastal areas, their high productivity and the services at 10 million ha (Bunt 1992), 14–15 million ha (Schwamborn
they offer, they are subjected to ecological pressure due to and Saint-Paul 1996), 24 million ha (Twilley et al. 1992) and
natural processes and human interventions and such pressure 18 million ha (Spalding 1997). FAO (2007) estimated that
causes changes in these critical habitats leading to deterioration the world mangrove total was at 15.6 million ha in 1980,
and/or loss of these areas over the years. Mangrove forests while more recent estimates range from 12 to 20 million ha.
once covered 75% of the coastlines of tropical and sub-tropical According to the available data, some 15.2 million ha of
countries (Spalding et al. 1997). Today less than 50% remain, mangroves are estimated to exist worldwide as of 2005,
and of this remaining forest, over 50% is degraded. This is down from 18.8 million ha in 1980. The most extensive
leading to decline of fisheries, degradation of clean water mangrove area is found in Asia, followed by Africa and
supplies and salinisation of coastal soils, erosion, and land North and Central America. Five countries, viz., Indonesia,
subsidence, as well as the release of carbon dioxide into the Australia, Brazil, Nigeria and Mexico together account for
atmosphere. In fact, mangrove forests fix more carbon per 48% of the total global area, and 65% of the total
unit area than phytoplankton in tropical oceans. Hence there mangrove area is found in just ten countries. Asia,
is a need to protect these productive critical habitats. They which according to FAO (2006) is the region with the
need to be regularly monitored for their extent, distribution lowest forest cover in terms of percentage of land area, has
and health (FAO 2007). the largest extent of mangroves (approximately 6 million
Availability of up-to-date information on the status and ha), and five of the ten countries with the largest extent of
conditions of this important ecosystem is critical and mangroves worldwide are found in this region.
essential for managing mangrove resources in a sustainable Mangrove forests in many parts of the world are declining
manner. Remote sensing could play an important and at an alarming rate, possibly even more rapidly than inland
effective role in the assessment and monitoring of man- tropical forests, and much of what remains is in degraded
grove forest cover dynamics. While remote-sensing data condition (Wilkie and Fortune 2003). The remaining man-
analysis does not replace field inventory, it provides grove forests are under immense pressure from clear cutting,
supplementary information quickly and efficiently. The encroachment, hydrological alterations, chemical spills, and
use of remotely sensed data offers many advantages climate change (McKee 2005). The forests themselves are
including synoptic coverage, availability of low-cost or dynamic in nature and are undergoing constant changes due to
free satellite data, availability of historical satellite data, and both natural and anthropogenic forces (Giri et al. 2007).
repeated coverage. In addition, recent advances in the
hardware and software used for processing a large volume of
satellite data has helped increase the usefulness of remotely Status of Indian mangroves
sensed data. Moreover, it is extremely difficult to get into vast
swamps of mangrove forests, and conducting field inventory According to a status report of the Government of India
is time consuming and costly (Giri et al. 2007). publication, the total area of the mangroves in India was
The ecological value of these tropical mangroves is reckoned at about 6,740 km2. This covered about 7% of the
acknowledged in many respects, for example, in (i) protecting world mangroves and 8% of the Indian coastline. But the
the coastline from tidal waves and storm surges, (ii) acting as recent Indian remote sensing data shows that the total area
biological filters in polluted coastal areas, (iii) supporting of the mangroves has decreased to 4,474 km2. Recent data
aquatic food chains, and (iv)shielding a large number of available with the Forest Survey of India, Dehra Dun shows
juvenile aquatic organisms (Barbier et al. 2004). Unfortunate- an extent of 4,827 km2 mangrove areas in India. Out of the
ly, mangroves in general are in serious decline owing to the total area, 57% of the mangroves are found on the east
expansion of human settlements, the boom in commercial coast, 23% on the west coast and the remaining 20% on the
aquaculture, the impact of tidal waves and storm surges, etc. Bay Islands (Andaman and Nicobar). Of the total,
(Linneweber and de Lacerda 2002; Barbier et al. 2004). 276,455 ha (5%) are found in the east coast of India: West
Hot spots and well managed areas of Pichavaram mangrove 3

Bengal 212,500 ha, Orissa 21,500 ha, Andhra Pradesh were principally used for management practices. Remote
39,700 ha and Tamil Nadu 2,755 ha. (Selvam et al. 2003; sensing has been extensively used as a tool to study and
Ranjit Daniels et al. 2006). The mangroves in the Tamil map mangroves (Baltzera et al. 2004; Bird et al. 2004;
Nadu state contribute to only 0.5% of the country’ s Fromard et al. 2004; Kovacs et al. 2004; Kovacs and
mangroves. Data available at the Department of Environment, Flores-Verdugo 2005; Nayak 2002). Today, with high
Tamil Nadu suggests that the Nagapattinam district has the spatial resolution imagery, such mapping has become an
most extensive mangrove forests of more than 1,000 ha, easily manageable and cost-effective task. Remote sensing
followed by Cuddalore 700 ha, Tanjore 400 ha, Tuticorin images of mangroves are for the most part reliable, accurate
100 ha and Ramanathapuram 100 ha. (Kovacs et al. 2004), and collected consistently in space
Pichavaram mangroves form the major part of the and time. Furthermore, the information can be obtained
mangroves in Cuddalore district. This area was declared as very quickly, almost in near real time. Other benefits of
the reserve forest in 1987, total area of the forest is 1,470 ha using remote sensing for mangrove mapping are that
including 50 small islands. The Pichavaram mangrove covers data can be collected in a non-destructive manner since
an area of about 11 km2, of which 50% is covered by forest, it is not necessary to move around in the mangrove to
40% by water-ways and the remaining filled by sand-flats obtain data. Large, remote, and inaccessible areas can
and mud-flats (Krishnamurthy and Prince Jayaseelan 1983). therefore, be easily mapped and monitored. Without
On-screen visual interpretation technique has been used by doubt, the remote sensing technique is a serious
Selvam et al. (2003) in conjunction with extensive field alternative to the traditional field monitoring for large-scale
validation to generate the wetland maps of the Pichavaram tropical mangrove management (Blasco et al. 1998).
mangrove forest area that indicated an improvement in In India, a number of studies conducted in the
mangrove cover (dense and young) from 3.25 km2 to Sundarbans have begun to develop and apply remote
6.18 km2 between 1986 and 2002 due to restoration measures. sensing techniques mainly for mapping purposes (Dwivedi
Kathiresan and Ramanathan (1997), Kathiresan (2000), et al. 1999; Nayak et al. 2001). Dense, degraded, mangrove
and Selvam et al. (2003) reported 13 species of true plantation areas and reclaimed areas can be distinguished
mangrove species, beside 73 sp. of other vegetation, which using Indian Remote Sensing (IRS) Linear Imaging Self-
include 24 tree, 21 shrub, 28 herb, seven climber, three scanning Sensor (LISS) III data (Nayak et al. 1996, Nayak
parasite and three creeper species in Pichavaram. Avicennia and Bahugunam 2001). Mangroves are characterized by the
marina is the most abundant species, followed by Rhizophora presence of particular dominant communities/species,
apiculata, R. mucronata, Bruguiera cylindrica, Excoecaria physico-chemical environment, extent and frequency of
agallocha, Ceriops decandra, A. officinalis, Aegiceras inundation by tidal waves, salinity, and soil types. It is
corniculatum, Rhizophora annamalayana, Acanthus ilicifolius possible to identify major mangrove communities such as
and Lumnitzera racemosa. Xylocarpus granatum and Avicennia sp.; Rhizophora forest, Sonneratia sp., Phoenix
Sonneratia apetala are rare. Kandelia candel is extinct in sp. and mangrove scrub using LISS III and PAN data on the
the study area (Muniyandi 1986). Kathiresan (1995) discov- Indian sub-continent. The variation in pigment content, leaf
ered new species, Rhizophora annamalayana, and described structure and water content in mangroves leaves along with
it, as a hybrid between R. apiculata and R. mucronata. the associated features and location were exploited to
During 1970s, there were no aquaculture ponds in the identify dominant mangrove types.
Pichavaram area, whereas in 1984 it increased to 3.99 km2
and in 1996 to 6.99 km2 (Kathiresan 2005). Furthermore, the
anthropogenic (domestic and industrial) discharges from the Gaps addressed in the present study
nearby densely populated area contribute significantly to the
ecological vulnerability of this ecosystem (Prasad 2005; While, some of the above works proved the effectiveness of
Ranjan et al. 2008a, b). The decrease in fresh water input remote sensing for studying the mangroves, there has been
or the interruption of flows caused by the construction no significant study so far that recommends a robust,
of dams and diversion of water for irrigation may objective methodology for their routine monitoring and
increase soil salinity, damaging the surrounding mangroves, as management. The present study is an attempt in this
had occurred in Singapore and Pakistan. direction, wherein remote sensing and GIS tools have been
used to study the spatial dynamic of the Pichavaram
mangrove ecosystem, an important coastal habitat of India
Remote sensing study of mangrove over a period of 15 years. Digital Image-processing and
GIS techniques have been used to classify the mangrove
Landsat MSS, Landsat Thematic Mapper (TM), and System cover into two classes (open and dense mangrove) and
Pour l’Observation de la Terre (SPOT) multi-spectral data the status of the classes has been monitored using
4 T. Srinivasa Kumar et al.

Pichavaram
Tamil Nadu
INDIA

Tamil Nadu

79°45'0"E 79°50'0"E

ry

11°30'0"N
11°30'0"N

ua
st

Parangipettai
rE
la

l
Ve

B A
Y
O F

Kille
B E

Pichavaram
N G

Mangrove Area 11°25'0"N

11°25'0"N

A L

T S Pettai

C ole r
on
e
st
ua
ry km
0 1.25 2.5 5

79°45'0"E 79°50'0"E

Fig. 1 Study area for coastal habitat: Pichavaram

Hot spots and well managed areas of Pichavaram mangrove 5

Table 1 Fact sheet

of Pichavaram 1 Location Lat. 11°23′–11°30′ N
mangroves Long. 79°45′–79°50′ E
2 Mangrove species 13 (Rare species: Avicennia & Rhizophara)
3 Boundaries East: Bay of Bengal
West: Uppanar Estuary
North: Vellar Estuary
South: Coleroon Estuary
4 State Tamil Nadu, South India
5 Climate Sub-humid with very warm summer (< 30°C)
6 Annual average rainfall 1,310 mm
7 Area Extent 1,470 ha (including islands)
8 Tidal characteristics Micro, Semi diurnal, Range: 15–100 cm
9 No. of rivers and rivulets 4,440
10 Salinity 10–34 ppt
11 No. of islands 51
12 No. of waterways 428 (few are navigable)
13 Significance of Pichavaram Defense against December 26, 2004
Mangrove Stands monstrous tsunami wave
14 Depth 0.3–3 m
15 River/tide dominated River dominated
16 Fishermen population 3,000
17 Coastal Hamlets 18
18 Bird species 177

multi-date satellite data. Attempt has also been made to latitudes and meridians of 79°45′–79°50′ E longitudes
identify hot spots and well managed areas using change (Fig. 1) at about 200 km south of Chennai (Madras) in
detection tools. This methodology can be used for Tamil Nadu State. The whole stretch of mangrove contains
routine monitoring of the mangrove cover and plan about 51 small and large islands with their sizes ranging
restoration measures, wherever required. from 10 m2 to 2 km2. The total area of mangrove forest is
about 1,470 ha (including islands). The Pichavaram
mangroves are considered among the healthiest mangrove
Study area occurrence in the world. It supports the existence of many
rare varieties of economically important shell and finfishes.
The Pichavaram Mangrove Forest near Chidambaram, The mangrove habitat is a home to 177 species of local and
South India is the second largest mangrove forest in the migratory birds. It serves as a good example of the
world. It is one of the unique eco-tourism spots in south degradation of a mangrove ecosystem due to aquaculture
India. It is located between the parallels of 11°23′–11°30′ N practices (Ranjan et al. 2008a). On the other hand, it is also

Table 2 Data set used

for the study Satellite Sensor Spatial Wavelength Date of Source
Resolution (μm) Acquisition

Landsat TM 30 m 0.52–0.60 29/01/1991 Open source data from

0.76–0.90 www.landsat.org
1.55–1.75
Landsat ETM 30 m 0.52–0.60 28/10/2000 Open source data from
0.76–0.90 www.landsat.org
1.55–1.75
Resurcesat-1 LISS-IV 5.8 m 0.52–0.59 22/07/2006 NRSC
0.62–0.68
0.77–0.86
6 T. Srinivasa Kumar et al.

one of those rare mangrove forests which have actually September during which the salinity goes high between 35
increased by 90% between 1986 and 2002 (Selvam et al. and 45 ppt (Selvam 2003).
2003). The tides in this area are semi-diurnal and vary in A study conducted by Kathiresan and Rajendran (2005)
amplitude from about 15 to 100 cm in different regions aftermath the tsunami of 26 December 2004 indicates that
during different seasons, reaching a maximum during the coastal mangrove forests can act as effective shield
monsoon and post-monsoon and a minimum during against tsunami. It is therefore an urgent call to initiate
summer (Muniyandi 1986). The salinity is maximum during conservation and restoration measures detecting the hot
summer and minimum in monsoon months. It ranges from 10 spot areas in the mangrove forest.
(December) to 34 gl−1 (April) in water and from 11.2
(November) to 31.5 gkg−1 (May) in sediments (Kathiresan et
al. 1996). The detailed at a glance are given in Table 1. The Data set used for the study
ultimate source of salinity in the mangrove habitat is the
neritic water, ingression of which is higher during summer Multi-temporal cloud-free satellite data from Landsat
than during monsoon. During monsoon, flood water TM and Enhanced Thematic Mapper (ETM) pertaining
dominates limiting the influence of seawater. The high to 1991 and 2000 were downloaded from the open
temperature of summer increases evaporation, resulting in source domain of www.landsat.org. Satellite data from
high salinity. The Pichavaram mangrove wetland receives Resourcesat-1 (IRS-P6) LISS IV sensor pertaining to 2006
freshwater mostly during the northeast monsoon from October was procured from National Remote Sensing Centre
to November. Fair weather season extends from February to (NRSC), Hyderabd. A total of three data sets were used
in the current study corresponding to the years 1991, 2000
and 2006 (Table 2).
Input
(Satellite Data)
Methodology
Pre processing

Geo-correction Ortho-rectified Landsat TM data acquired in 1991 and 2000

are downloaded from www.landsat.org website.
Ortho-rectified Resourcesat-1 (IRS P6) LISS-IV ortho-rectified data of
Radiometric Correction Reference Image 2006 was obtained from the NRSC. All these data sets were
imported into the Earth Resources Data Analysis Systems
Spatial normalization and (ERDAS) Imagine 9.1 digital image processing software.
Extraction of Study Area IRS P6 LISS-IV digital data of July 22, 2006 was geo-
referenced for polynomial order 2 using Landsat TM data
Processing acquired in 2000 as the reference. Since the digital data do
not have the real earth coordinates, they were geometrically
Unsupervised Classification corrected using reference image by taking common feature
Ancillary
Information
on both the image such as road-road interaction, road-rail
interaction, canal-road interaction etc., using ERDAS
Contextual Editing Imagine 9.1 digital image processing software. A union of
the outer boundary of the Mangrove cover digitized from
the satellite images of all the three periods is taken as the
extent of the study area. This is to ensure that new
Output
(Mangrove Map)
mangrove areas that would have come up outside/periphery
of the traditional mangrove zone represented by the
reference/base imagery (Landsat TM 1991) are not omitted
Post Processing (GIS Overlay)
during the subsequent analysis. Study area was then
Change Detection extracted from all the three datasets.
False Colour Composite (FCC) image of the Pichavaram
mangrove was generated with the band combinations of 5,
Identification of Hotspots 4, 3 and 3, 2, 1 in Red Green Blue (RGB) in TM and LISS
IV data respectively. The band selection has been done
based on the published works (Selvam et al., 2003; Long
Fig. 2 Methodology for classification of mangroves and Skewes 1996; Green et al., 1998; Chauhan and
Hot spots and well managed areas of Pichavaram mangrove 7

False Colour Composites

Landsat TM Landsat TM IRS LISS-IV

DOA: 29/01/1991 DOA: 28/01/2000 DOA: 22/07/2006

True Colour Composites

1991 20 00 2006

Mangrove Maps
1991 20 00 200 6

Legend
Dense Mangrove

Kilometers Open Mangrove

0 0.5 1 2 3 4
Coastline

Fig. 3 FCC and spatio-temporal variations of Pichavaram mangrove

Dwivedi, 2007) as well as the spectral signatures in areas from the study site. The radiance conversion
different bands were collected from the known mangrove technique is a necessary process to be performed on the
8 T. Srinivasa Kumar et al.

Table 3 Spatio-temporal variations in Pichavaram Mangroves

Mangrove Class Area (km2) Change in area (km2) Rate of change (km2/y)

1991 2000 2006 1991 to 2000 2000 to 2006 1991 to 2006 1991 to 2000 2000 to 2006 1991 to 2006

Dense Mangrove 2.76 3.17 5 0.41 1.83 2.24 0.04 0.31 0.14
Open Mangrove 1.8 3.38 2.06 1.58 −1.32 0.26 0.16 −0.22 0.02
Total area 4.56 6.55 7.06 1.99 0.51 2.5 0.20 0.09 0.16

time series satellite data (Lunetta, 1999; Chauhan and and classify further within mangroves based on the spectral
Dwivedi, 2007; Singh, 1989) in order to normalize the signatures.
varying energy (Jensen, 1985) recorded from different In order to carryout GIS analysis composites of the
sensors and bands using the known conversion factors mangrove images were converted to Environmental Sys-
provided in the metadata files of the individual scenes. tems Research Institute (ESRI) shape files by raster to
Further the LISS-IV data of spatial resolution 5.8 m has vector conversion techniques. These shape files were
been resample to 30 m in order to be spatially comparable analysed using GIS package ESRI Arc Map. The geo-
with TM and ETM. Histogram equalization method of graphic area for the individual feature classes were
image enhancement technique was used to make the classes estimated by area calculation technique. The geographic
spectrally more distinct for on-screen visual interpretation. area of the dense and open mangrove classes recorded in
Analysis of the FCC indicated that spectral signatures of the different period was extracted from the feature attribute
mangrove vegetation classes are distinctly separable from table. The new shape file has been obtained by intersecting
other classes (water, mud, barren land, aquaculture, sand, overlay technique using the shape files comprising the
etc.,) the only exception being non-mangrove vegetation mangrove classes pertaining to different periods. The
such as agriculture and plantations. However, these have attributes of intersected file have been used to estimate
been easily isolated by contextual editing taking into the spatio-temporal changes of the Pichavaram Mangroves
consideration their unique spectral/spatial properties. during 1991–2006.
The classification involved two steps. In the first step,
Iterative Self Organising Data Analysis Technique (ISO-
DATA) clustering was performed on individual images to Results and discussion
segment them into twenty classes each depending upon the
spectral signatures in green, red and infra-red bands. Using The changes that occurred in both the categories from 1991
contextual editing, the number of classes has been reduced to 2006 are shown in Fig. 3. The statistics of spatio-
initially to mangrove and non-mangrove classes. In the temporal variation for dense and open mangrove classes for
second step, the Mangrove Class was further classified into the study period is given in Table 3, Figs. 4 & 5. From the
dense and open Mangrove Classes. Methodology adopted analysis of the classified images the following results are
for classification is shown in Fig. 2. We have adopted two derived. The total spatial extent of Pichavaram mangrove
step classification processes to separate mangrove as whole during 1991 was about 4.56 km2. Dense mangrove

2.5
5
Dense Mangrove
4.5 2 Open Mangrove

4
Change in area (sq. km)

Dense Mangrove 1.5

Open Mangrove
3.5
1
Area (sq. km)

2.5 0.5

2
0
1.5 1991-2000 2000-2006 1991-2006
-0.5
1

0.5 -1

0
1991 2000 2006 -1.5
Year Period (year)

Fig. 4 Bar-chart is representing the area of mangrove classes during Fig. 5 Bar-chart is representing the spatial change of mangrove
1991, 2000 and 2006 classes during 1991–2000, 2000–2006 and 1990–2006
Hot spots and well managed areas of Pichavaram mangrove 9

Fig. 6 Field photos

depicting the open
mangroves (a) were
developed by preventive
measures taken from Forest
Department by trenching (b)
and planting mangroves
(c) result in well developed
mangrove (d). (Courtesy:
NWCMP, MoEF report of
October 2008)

contributed 2.76 km2 and open mangrove contributed increased by 0.41 km2 with a very low rate of 0.04 km2/y.
1.80 km2. During 2000 spatial extent of the Pichavaram Whereas open mangrove cover reduced in the spatial extent
mangrove were recorded 3.17 km2 areas by dense mangrove of 1.32 km2 with a reduction rate of 0.22 km2/y. Overall the
and 3.38 km2 area of open mangrove out of total mangrove open mangrove category increased during 1991 to 2006 with
cover 6.55 km2. Where as during 2006 total spatial extent of an area of 0.26 km2 at a very low rate 0.02 km2/y.
the Pichavaram mangrove was about 7.06 km2. This
comprises an area of 5.0 km2 dense mangrove and
2.06 km2 of open mangrove. The remote sensing measure- Spatial variability analysis
ments from 1991 to 2006 reveal there is a net increase in
mangrove spread area of 2.50 km2. The change analysis Kathiresan (2000) reported that within the last century, there
indicates that the increase in the dense mangrove cover from has been a decline in 75% of the mangrove cover in Pichavaram
1991 to 2000 and till 2006. However the maximum rate of and of the existing 25%; only 10% has dense vegetation, while
increase in dense mangrove cover was observed during the remaining 90% of the area has been degrading.
2000–2006 with an area of 1.83 km2 having a rate of The current study, however, indicates an increasing trend
0.31 km2/y. During the period from 1991 to 2006 dense in the spatial extent of the mangrove cover of the region.
mangroves have been increased to 2.24 km2 with a rate During the entire period under study, it was observed that
0.14 km2/y. During 1991 to 2000 dense mangrove has been the total area of Pichavaram mangrove gradually increased

Table 4 Inter-class
conversion statistics within Change Class Area in km2
Pichavaram Mangroves
1991 to 2000 2000 to 2006 1991 to 2006

Dense Mangrove to Sparse Mangrove 0.43 0.30 0.23

Sparse Mangrove to Dense Mangrove 0.48 1.23 0.70
Dense Mangrove to Non-Mangrove 0.07 0.22 0.21
Sparse Mangrove to Non-Mangrove 0.59 1.07 0.77
Non-Mangrove to Sparse Mangrove 2.22 0.91 1.74
Non-Mangrove to Dense Mangrove 0.43 0.42 1.28
10 T. Srinivasa Kumar et al.

Fig. 7 Hot spots and 79°46'0"E 79°48'0"E

well managed areas

11°28'0"N

11°28'0"N
Kille
11°26'0"N Bay of Bengal

11°26'0"N
Pichavaram
Mangroves

T S Pettai
11°24'0"N

11°24'0"N
Legend
Hot Spot 0 0.5 1 2
Well Managed km

79°46'0"E 79°48'0"E

from 4.56 km2 in 1991 to 6.55 km2 in 2000 and to not reachable earlier (ICMAM 2000). This resulted in
7.06 km2 in 2006. This amounts to net increase of 2.50 km2 recruiting of young mangroves into this area. Young
within a span of 15 years. A similar improvement is seen in mangrove class picked up in the study of Selvam et al.
the Dense Mangrove Class that increased from 2.76 km2 in (2003) is under open mangrove in our classification.
1991 to 3.17 km2 in 2000, and 5.00 km2 in 2006. This Selvam et al. (2003) in their studies, classified young
increase is due to conversion of open mangrove areas into mangroves stand in Pichavaram area into two groups, i.e.
dense mangroves. In the case of Open Mangroves, it was above three-year-old plantation and below 3 year old
seen that the area was 1.80 km2 in 1991 that increased to plantation. The reduction in the open mangrove class
3.38 km2 in 2000 and gradually decreased to 2.06 km2 in between 2000 and 2006 is due to the young mangroves
2006. The increase in area from 1991 to 2000 is attributed above three-year-old having grown up to dense mangroves
to the efforts of the State Forest Department by digging which is normally attained in about 7 year old trees. This is
trenches to flush the tidal waters into the areas where it was being verified from the existing records of Tamil Nadu
Hot spots and well managed areas of Pichavaram mangrove 11

Forest Department and village mangrove councils engaged mangroves. The study demonstrated that by classifying
in restoration and conservation measures. The field photos mangrove ecosystems into just 3 classes using remote
showing different phases of the mangrove development is sensing data and by studying their temporal variations, it is
provided in the Fig. 6. FCC for the years 1991, 2000 and possible to get a reasonably accurate picture of the extent
2006 along with the corresponding classified images of the and condition of the mangrove ecosystem. Techniques used
Pichavaram mangrove is shown in Fig. 3. The increase in in the study such as ISODATA clustering and change
the open mangrove during the 1991 is due to the detection analysis have proven to be objective and provide
conservation measure and plantation of the new mangrove. reasonably accurate information that will be useful for
The part of open mangroves during the 1991 converted into coastal managers for better management of mangroves.
dense mangrove during 2000 and similarly 2000 to 2006. It Using this technique, it is possible to identify more than
was observed that the maximum mangrove recruitment was three signatures within the mangrove class that might
observed during 1991–2006 led to the increase in the open correspond to different mangrove communities. However,
mangrove. The maximum restorations of the mangroves it was not attempted as part of the current study since
were observed during 2000–2006 result in increased dense special emphasis was laid on arriving at simple techniques
mangrove. This study highlights two phases of the of the that could be useful for routine monitoring. Arriving at
mangrove management such as recurrent phase (1991– community-level maps would require specific and more
2000) and restoration phase (2000–2006). However the complicated techniques in conjunction with extensive
study reveals the consistent increase in the dense mangrove ground-truthing that is time consuming.
during 1991–2000. Contrary to some of the earlier studies, it is evident from
the study that the total area of the Pichavaram mangrove
showed a net increase of 2.51 km2 within a span of 15 years
Inter-class variability analysis (1991 to 2006). This is due to the identification of causative
factors of mangrove degradation and subsequent conserva-
The statistics of spatio-temporal variation within the three tional measures taken up by the State Forest Department.
classes namely, Dense Mangrove, Sparse Mangrove and The current research also pin points areas of well managed
Non-Mangrove are given in the Table 4. Hot spots and well mangroves as well as hot-spots that are at a risk of being
managed mangrove zones in the Pichavaram mangroves degraded and need immediate attention.
derived from inter-class variability analysis between 1991
and 2006 are shown in Fig. 7. Acknowledgments The authors are thankful to Dr. P. S. Goel,
former Secretary, Ministry of Earth Sciences, for his encouragement.
From the analysis of inter-class conversions between
The authors would like to thank Global Observatory for Ecosystem
1991 and 2006 (Table 4) the following observations are Services (GOES), Michigan State University for the Landsat data
made. An area of 3.02 km2 was converted from Non- available on their Website. This is INCOIS contribution number 77.
Mangrove to Mangrove areas. In addition, an area of
0.70 km2 was converted from Sparse Mangrove to Dense
Mangrove. This is an indication of well managed mangrove References
areas. This is due to the developmental measures taken up
by the State Forest Department. An area of 0.98 km2 was Baltzera F, Allison M, Fromard F (2004) Material exchange between
converted from mangrove to non-mangrove. In addition, the continental shelf and mangrove-fringed coasts with special
reference to the Amazon- Guianas coast. Mar Geol 208:115–126
0.23 km2 was converted from dense mangrove to sparse Barbier EB, Cox M, Sarntisart I (2004) Household use of mangrove
mangrove. These areas are to be identified as hotspots and mangrove conservation decisions. In: Barbier EB, Suthawan
wherein immediate interventions are required to identify S (eds) Shrimp farming and mangrove loss in Thailand. Edward
problems and take up measures to reclaim the mangrove Elgar, London, pp 115–130
Bird M, Chua S, Fifield LK, The TS, Lai J (2004) Evolution of the
ecosystem. Since these areas are traditionally mangrove Sungei Buloh–Kranji mangrove coast, Singapore. Appl Geogr
habitats, if serious measures are taken they can be 24:181–198
reclaimed back as healthy mangrove ecosystem with Blasco F, Gauquelin T, Rasolofoharinoro M, Denis J, Aizpuru M,
minimal efforts. Using this information will help in Caldairou V (1998) Recent advances in mangrove studies using
remote sensing data. Marine and Freshwater Research 49(4):287–
prioritizing areas of immediate concern. 296
Bunt JS (1992) The ecology of mangrove and related ecosystems. In:
Jaccarini V, Martens E (eds) Kluwer, Dordrecht, 1992, 266 pp.,
Conclusions Proceedings of the International Symposium held at Mombasa,
Kenya, 24–30 September 1990, Aquat Bot 49(2–3), August
1994, pp 198–199
Remote Sensing and GIS have been found to be extremely Chauhan HB, Dwivedi RM (2007) Inter sensor comparison between
useful for understanding the coastal habitats, especially RESOURCESAT LISS III, LISS IV and AWiFS with reference to
12 T. Srinivasa Kumar et al.

coastal landuse/landcover studies. Int J App Earth Obs Geoinfo CD (eds) Environmental monitoring methods and applications.
10(2):181–185 Taylor & Francis, London, pp 318
Dwivedi RS, Rao BRM, Bhattacharya S (1999) Mapping wetlands of McKee KL (2005) Global change impacts on mangrove ecosystems.
the Sundarban delta and its environs using ERS-1 SAR data. Int J USGS Fact Sheet 2004–3125. Available via http://www.nwrc.
Remote Sens 20:2235–2247 usgs.gov/factshts/2004-3125.pdf. Accessed on April 10, 2009
FAO (2006) Global First Global Resource Assessment 2005-Progress Muniyandi K (1986) Studies on mangroves of Pichavaram (southeast
towards sustainable management. FAO Forestry Paper 147, FAO coast of India). Ph.D Thesis, Annamalai University, Parangipet-
Rome tai, India: 215 p
FAO (2007) The World’s Mangrove 1980–2005, A thematic study Nayak S (2002) Use of satellite data in coastal mapping. Indian
prepared in the frame work of the Global Forest Resources Cartographer, CMMC-01, 147–156
Assessment, 2005. FAO Forestry Paper 153, FAO Rome Nayak S, Bahugunam A (2001) Application of remote sensing data to
Fromard F, Vega C, Proisy C (2004) Half a century of dynamic coastal monitor mangroves and other coastal vegetation of India. Ind J
change affecting mangrove shorelines of French Guiana. A case Mar Sci 30(4):195–213
study based on remote sensing data analyses and field surveys. Nayak S, Chauhan P, Chauhan HB, Bahuguna A, Nath-Narendra A
Mar Geol 208:265–280 (1996) IRS-1 C application for coastal zone management. Curr
Giri C, Pengra B, Zhu Z, Singh A, Tieszen LL (2007) Monitoring Sci 70(7):614–618
mangrove forest dynamics of the Sundarbans in Bangladesh and Nayak S, Sarangi RK, Rajawat AS (2001) Application of IRS P4
India using multi-temporal satellite data from 1973 to 2000. OCM data to study the impact of cyclone on coastal environment
Estuar Coast Shelf Sci 73:91–100 of Orissa. Curr Sci 80:1208–1213
Green EP, Clark CD, Mumby PJ, Edwards AJ, Ellis AC (1998) Prasad MBK (2005) Nutrient dynamics in Pichavaram mangroves,
Remote sensing techniques for mangrove mapping. Int J Remote southeast coast of India. Ph D Thesis, Jawaharlal Nehru
Sens 19(5):935–956 University, New Delhi, India
ICMAM (2000) ICMAM Report on “Critical habitat Information Ranjan RK, Al R, Singh G, Chidambaram S (2008a) Assessment of
System using GIS for Pichavaram Mangrove South India. metal enrichments in tsunamigenic sediments of Pichavaram
Available via http://www.icmam.gov.in. Accessed on April 13, mangroves, southeast coast of India. Environ Monit Assess
2009 147:389–411. doi:10.1007/s10661-007-0128-y
Jensen JR (1985) Introductory Digital Image Processing. Prentice Hall Ranjan RK, Ramanathan AL, Singh G (2008b) Evaluation of
PTR Upper Saddle River, New Jersey geochemical impact of tsunami on Pichavaram mangrove
Kathiresan K (1995) Rhizophora annamalayana: a new species of ecosystem, southeast coast of India. Environ Geol 55:687–697.
mangroves. Envir Ecol 13:240–241 doi:10.1007/s00254-007-1019-9
Kathiresan K (2000) A review of studies on Pichavaram mangrove, Ranjit Daniels RJ, Jayshree Vencatesan R, Krishnamani SK, Dutta C,
southeast India. Hydrobiol 30:185–205 Sivasubramanian R, Sivakkumar A, Manimekalan, KM, Sanjib D
Kathiresan K (2005) Distribution of mangroves. In: Kathiresan K, (2006) The ground beneath the water. Post–tsunami Impact
Khan SA (eds) UNU-INWEH-UNESCO assessment of wildlife and their habitats in India. In: Kaul R and
Kathiresan K, Rajendran N (2005) Coastal mangrove forests mitigated Menon V (eds) The main land
tsunami. Estuar Coast Shelf Sci 65:601–606 Schwamborn R, Saint-Paul U (1996) Mangroves - Forgotten Forests?
Kathiresan K, Ramanathan T (1997) Medicinal plants of Parangipettai Natural Resources and Development 43/44:13–36
coast. PA: CASMB. Annamalai University, India, p 79 Selvam V (2003) Environmental classification of mangrove wetlands
Kathiresan K, Rajendran N, Thangadurai G (1996) Growth of of India. Curr Sci 84(6):757–765
mangrove seedlings in intertidal area of Vellar estuary southeast Selvam V, Ravichandran KK, Gnanappazham L, Navamuniyammal-
coast of India. Ind J Mar Sci 25:240–243 Taramani M (2003) Assessment of community-based restoration
Kovacs JM, Flores-Verdugo F (2005) Mapping mangrove leaf area of Pichavaram mangrove wetland using remote sensing data.
index at the species level using IKONOS and LAI-2000 sensors Curr Sci 85(6):795–797
for the Agua Brava Lagoon, Mexican Pacific. Estuar Coast Shelf Singh A (1989) Review article—digital change detection techniques
Sci 62:377–384 using remotely sensed data. Int J Remote Sens 10(6):989–1003
Kovacs JM, Flores-Verdugo F, Wang J, Aspden LP (2004) Estimating Spalding M (1997) The global distribution and status of mangrove
leaf area index of a degraded mangrove forest using high spatial ecosystems. International News Letter of Coastal Management-
resolution satellite data. Aquat Bot 80:13–22 Intercoast Network, Special edition 1:20–21
Krishnamurthy K, Prince-Jeyaseelan MJ (1983) The Pichavaram Spalding MD, Blasco F, Field CD (1997) World Mangrove Atlas. Int
(India) mangrove ecosystem. Int J Ecol Envir Sci 9:79–85 Soc Mangrove Ecosys, Okinawa
Linneweber V, de Lacerda LD (2002) Mangrove ecosystems: function Twilley RR, Chen R, Hargis T (1992) Carbon sinks in mangroves and
and management. Springer, Berlin their implication to carbon budget of tropical ecosystems. Water,
Long BG, Skewes TD (1996) A Technique for Mapping Mangroves Air and Soil Poll 64:265–288
with Landsat TM Satellite Data and Geographic Information Wilkie ML, Fortune S (2003) Status and trends of mangrove extent
System. Estuar Coast Shelf Sci 43:373–381 worldwide. Available via http://www.fao.org/documents/
Lunetta RS (1999) Remote sensing change detection; environmental show_cdr.asp?Url_file¼/docrep/007/j1533e/J1533E41.htm.
monitoring methods and applications. In: Lunetta RS, Elvidge Assessed on April 10, 2009
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.