See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/227242806

Mangroves of Godavari – Analysis Through Remote Sensing Approach

Article in Wetlands Ecology and Management · February 2006

DOI: 10.1007/s11273-005-2175-x

CITATIONS READS

63 1,876

4 authors:

Ramasamy Ramasubramanian Gnanappazham Lakshmanan

M S Swamininathan Research Foundation Indian Institute of Space Science and Technology
35 PUBLICATIONS 318 CITATIONS 40 PUBLICATIONS 525 CITATIONS

SEE PROFILE SEE PROFILE

Ravishankar Thupalli Navamuniyammal Muthu

Indpendent International Consultant Anna University, Chennai
15 PUBLICATIONS 273 CITATIONS 7 PUBLICATIONS 214 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Ramasamy Ramasubramanian on 05 June 2015.

The user has requested enhancement of the downloaded file.

Wetlands Ecology and Management (2006) 14: 29–37
Springer 2006
DOI 10.1007/s11273-005-2175-x

Mangroves of Godavari – analysis through remote sensing approach

R. Ramasubramanian, L. Gnanappazham*, T. Ravishankar and M. Navamuniyammal

M.S. Swaminathan Research Foundation, Taramani, Chennai 600 113, India; *Author for correspondence
(e-mail: [email protected]; [email protected]; phone: +91-044-2254-1698; 2254-1229; fax: +91-
044-2254-1319)

Received 15 March 2004; accepted in revised form 14 February 2005

Key words: Accretion and erosion, Geomorphology, Mangrove ecosystem, Mudflats, Remote sensing,
Satellite images, Shoreline changes

Abstract

The expansion of agriculture and aquaculture farms in the coastal areas has led to conversion of mangroves
in the recent past. The extent of mangroves has also changed due to the erosion of mangroves along the
coast and accretion near river mouths, leading to the formation of new mangrove areas. This study has
been undertaken in the mangroves of the Godavari estuary, Andhra Pradesh, India to understand the
changes in the extent of mangroves, namely accreted mangroves, erosion due to wave action and river water
flow during floods, and changes due to forest restoration between 1986 and 2001, through remote sensing.
The geomorphological changes due to river water flow in and around the mangroves have also been
analysed. The changes in the vegetation due to forest restoration and natural regeneration are appreciable,
while the changes in the area due to erosion and accretion are more or less equal. An analysis of the remote
sensing images of 1986 and 2001 reveal that the mangroves outside the forest boundary have been con-
verted to aquaculture. The sand spit of Hope Island has changed with time and has grown nearly 2.6 km
between 1937 and 2001.

Introduction ties in their physiological characteristics and

structural adaptations. Mangrove areas are asso-
The mangrove ecosystem is one of the most pro- ciated with high leaf production and leaf fall and
ductive ecosystems. Increase in land use for agri- rapid decomposition of the litter. Mangrove for-
culture and aquaculture, especially near ests perform multiple ecological functions: they
mangroves, has led to the conversion of this produce woody trees, provide habitat and detritus
important ecosystem. Mangroves are plant com- food for fish and shellfish and act as spawning
munities in tropical and subtropical regions, which ground for a variety of fishes, prawns and shellf-
live in inter tidal areas inundated by tides. Man- ishes. They harbour a variety of valuable fauna,
grove plants live in muddy and wet saline soils. including migratory birds. They also protect the
They are trees associated with shrubs, growing coasts from storms and tidal waves and help in the
both on the sheltered coast and inland, lining the formation of land by trapping sediments brought
bank of estuaries and rivers. They are an ecologi- by the river run-off. A significant percentage of the
cal group of evergreen plant species, belonging to coastal community, especially the fishermen, is
different families, but possessing marked similari- traditionally dependent on mangroves for their
30

domestic needs such as firewood and timber, apart second largest river in India, with a total catchment
from collecting fishes for their livelihood. Though area of 314,685 sqkm and a total length of
the mangroves provide a large amount of 1465 km. Two-thirds of the catchment flows into
resources, they are being increasingly cleared for the Bay of Bengal. The Godavari River is con-
other landuse activities such as settlements, agri- nected to Kakinada Bay by two major canals,
culture and shrimp farming. namely the Corangi and the Gaderu. There are
Remote sensing is used as a tool for monitoring numerous other small canals branching from these
the changes, especially in forests, because the hilly canals, supplying tidal water to the mangroves. The
or swampy terrain is inaccessible and vast in area. Godavari River flows into the sea through two
It provides relatively accurate information mouths, one near Bhairavapalem in the north and
regarding the status of vegetation in the forest and the other near Kottapalem in the south (Figure 1).
is cost-effective and time saving. Geographic
Information System (GIS) and remote sensing
tools are being extensively used to understand the Materials and Methods
changes in mangrove areas, for purposes of plan-
ning and management. Mangroves have been The remote sensing data was obtained from
mapped and monitored for better management National Remote Sensing Agency (NRSA) and
through aerial photographs or satellite data, as in interpreted with the base line information derived
Florida (Ramsey and Jensen 1996), Guiana (Sery from Survey of India (SOI) toposheets. To moni-
et al. 1995), Kenya (Brakel 1984), Senegal, Goa tor the geomorphological and shoreline changes,
and Cauvery in India, Cilacap in Java, Thailand baseline information from the Survey of India
(Silapathong and Blasco 1992) and Ecuador maps of 1937–1938 (65 L) and 1975–1976 (65 L/1,
(Klemas and Bartlett 1975; Terchunian et al. 2, 5 and 6) were used. The remote sensing data of
1986). Remote sensing technique is a useful tool to 1986 and 2001 were mapped for wetland features
assess and monitor the effectiveness of mangrove such as dense mangroves, degraded mangroves,
restoration and conservation programmes where mudflat, beach, lagoon and water bodies and the
physical monitoring is difficult (Selvam et al. changes were analysed. The classification was
2003). verified with intense ground truth survey.
Mangrove areas are commonly characterized by As the remote sensing digital data have geo-
rapid changes, i.e. conversion of the mangroves for graphical distortions, they were geographically
other land uses such as shrimp farming, agricul- projected to real earth coordinates system, using
ture and industry. Natural changes such as accre- Survey of India toposheets (1:50,000 scale). The
tion of new land area and erosion along the coast ground control points like road/road intersection,
and rivers are also common in this ecosystem. road/railway line intersection, road/canal inter-
Information about these changes is crucial for section and sharp bends, found both in remote
developing management policies. Remote sensing sensing data and the reference toposheets, were
allows quantitative and qualitative assessments of used for rectification. The field points were col-
ground conditions over large and inaccessible lected using GARMIN hand held Global Posi-
geographical areas. This study describes the tioning System (GPS). The satellite data obtained
changes in the mangroves and associated landuse from National Remote Sensing Agency were used
of the Godavari estuary between 1986 and 2001. for this study (Table 1).
After the rectification process, the standard
False Color Composite (FCC) image was gener-
Study area ated with green, red and near infrared bands and
enhanced through the histogram equalization
The Godavari mangroves are located between the method to clearly distinguish the land use and land
latitudes 1630¢ and 17 N¢ and the longitudes cover features. Onscreen digitization procedure on
8210¢ and 8225¢ E in East Godavari District, ERDAS Imagine was carried out to delineate
Andhra Pradesh, India. The mean annual rainfall mangrove and associated land cover categories.
in this region is 110 cm and the annual mean The visual interpretation keys used to identify the
temperature is 28 C. The Godavari River is the coastal features are given in Table 2.
 31

Figure 1. Study area – Godavari mangrove wetland, India.

Ground verification of 72 points was done for decided, based on its area coverage and also
the land used and land cover classes. The num- the longevity towards its correctness, based on
ber of verification points of each class was ground knowledge. The error matrix of the
32

Table 1. Details of data used in this study. River, which includes large areas of water bodies.
The digital data (2001) shows that dense
Year Data Purpose
mangroves occupy only 16,406 ha; degraded
1937 Survey of India Shoreline changes mangroves occupy 3355 ha; mudflats and water
Toposheet (1:2,50,000) bodies cover the rest of the area (Table 4). The
1975 Survey of India Base data, shoreline and
interpretation of the satellite data and the ground
Toposheet (1:50,000) geomorphological changes
1986 Landsat 5 TM Mangroves mapping and truth information revealed that the density of
monitoring mangrove cover is more on the northern side
1998 IRS 1C LISS III Mangroves mapping and (Coringa Wild Life Sanctuary) of the Godavari
monitoring River than on the southern side (non-sanctuary
2001 IRS 1D LISS III Mangroves mapping and
area). This may be due to the fact that the north-
monitoring
ern areas receive fresh water for a prolonged
period from agricultural run-off. The agriculture
ground verification of 2001 data is given in run-off through the creeks of Corangi and Gaderu
Table 3. canals brings dissolved nutrients such as nitrates
and phosphates to the mangroves. This area is also
Results and discussion protected from high velocity winds and high-
energy water currents by the sand spit which is
Changes in mangrove area due to restoration about 18 km long. The mangrove vegetation
found in the periphery near the villages is degraded
The Godavari mangroves occupy an area of about due to anthropogenic pressures. The restoration of
33,150 ha in the deltaic region of the Godavari mangrove vegetation in the degraded area is

Table 2. Identification of coastal features.

Area Colour Shape and size Texture Association

Wetland without vegetation

Mudflat with water Dark bluish gray to Irregular Smooth Estuaries, inland
brownish gray wetland and coastal
wetlands
Mudflat without water Medium gray to bluish gray Irregular Smooth Estuaries, inland
wetland and coastal
wetlands
Sandy mudflat Very light yellowish gray Irregular Smooth to River beds
moderately rough
Tidal flats Dark gray or bluish gray The shape of the river Smooth River course or
bank or coast coastline near the
reach of tidal water
Saline area Pure white to bluish white Patchy or irregular Smooth Coastal or river
course or mudflat
Salt pan Pure white when dried or Regular rectangular Smooth Coastal area
black when filled with water or square
Aquaculture Dark blue or black Regular rectangular Smooth Just outside the
forest reserve
boundary
Water bodies
Deep or clear water Dark blue/black – Smooth –
Shallow water Light blue/cyan – Smooth –
Turbid water Grayish blue or greenish blue – Smooth –
Vegetated wetland area
Dense Mangrove Bright red – Smooth Estuary and backwater
Sparse and dwarf Pinkish red to brownish red – Moderately rough Estuary and backwater
mangrove to smooth
Marshy vegetation Dark to light brownish red – Moderately rough Along with mudflat
to smooth and wetland
 33

Table 3. Error matrix of ground truth verification points of classified map of Godavari mangroves.

Classified Actual

Dense Sparse Degraded Sand Mudflat/tidal Coastal Aqua-culture Salt Total

Mangrove Mangrove area flat plantation pan

Dense 10 0 0 0 0 0 0 0 10
Mangrove
Sparse Mangrove 0 18 0 0 0 0 0 0 18
Degraded area 0 3 12 0 0 0 0 0 15
Sand 0 0 0 3 0 0 0 0 3
Mudflat/Tidal flat 0 0 0 0 5 0 0 0 5
Coastal plantation 0 0 0 0 0 8 0 0 8
Aquaculture 0 0 0 0 0 0 8 2 10
Salt pan 0 0 0 0 0 0 0 3 3
Total 10 21 12 3 5 8 8 5 72

Table 4. Mangrove and nearby landuse areas (ha) in 1986 and Table 5. Changes in mangrove vegetation between 1986 and
2001. 2001.

Description Area in ha. Description Changes in

area between
1986 2001 1986 and 2001
(ha)
Dense mangroves 16,038 16,406
Degraded area 5708 3355 Restoration of mangroves 594
Shrimp farms – 7310 Natural regeneration in the degraded areas 2318
Salt pan 1868 685 Natural regeneration in the accreted areas 586
Total increase 3498
Degradation due to Aquaculture/Salt pans 658
Erosion 806
clearly noticed in the 2001 remote sensing data Degradation due to other causes – coupe 1666
(Figure 2). The degraded area of 5708 ha, seen in felling, grazing and felling
1986, was reduced to 3355 ha in 2001 (Table 4). Total loss 3130
This reduction was due to the restoration of de- Net increase 368
graded mangroves by the Forest Department, the
Government of Andhra Pradesh and the mangrove
restoration programme of M.S. Swaminathan 1986 and 2001 (Table 5). The accretion was due to
Research Foundation. Selvam et al. (2003) studied sedimentation of alluvium from the runoff in the
the impact of the mangrove rehabilitation pro- Godavari River through the Corangi and Gaderu
gramme in Pichavaram, using remote sensing, in canals. Erosion reduced the mangroves by 806 ha
which the mangrove cover was seen to have in the same period. Mangrove vegetation along the
increased from 325 ha to 618 ha and the degraded southern bank of the Nilarevu River has been
area reduced from 375 ha to 60 ha between 1986 eroded due to floods during the southwest mon-
and 2002. The earlier system of coup felling was soons. The river had become wider in 2001 than it
found to be one of the causes for the degradation of was in 1986. In 1986 the river span was about
the mangroves while the geographically elevated 1.2 km while in 2001 it had become 2.6 km due to
area, formed by sediment deposition during floods erosion. The mangroves along the coast on the
along the rivers and canals, was another. northern side of the Kottapalem River mouth have
also been eroded; the small mangrove islands seen
in the 1986 image are not seen in the 2001 image.
Changes due to accretion and erosion The vegetation in the accreted areas near the
mouth of the Corangi River and the Gaderu
The increase in the extent of mangroves due to River in the Kakinada Bay is distinct. Porteresia
newly accreted areas was about 586 ha between coarctata is found along with seedlings of
34

Avicennia alba and Sonneratia apetala Pure 15.6 km during 1937–1938 and in 2001 it was
stands of Avicennia alba are seen near the 18.2 km (Figure 3; Table 6). There has been a shift
Kakinada Bay. Next to this, towards the land, is in the sand bar towards the west, which can be
Sonneratia apetala. From here onwards, the seen in the 2001 image, which has resulted in the
vegetation is found to be mixed, with small patches loss of mangrove vegetation. The river course has
of pure stands of Excoecaria agallocha, Aegiceras also changed much near the mouth. The small
corniculatum, Lumnitzera racemosa and Avicennia sand spit on the southern side of the Nilarevu
marina. Suaeda maritima and S. nudiflora are seen in River mouth has also grown significantly as seen in
the degraded areas. Rhizophora apiculata and the map of 2001 (Figure 2). The river discharge
R. mucronata are found along the creeks on the into the sea, which was towards the east in 1986,
seaward side. has shifted towards the northeast in 2001.
Reddy and Prasad (1982) studied the geomor-
phological changes of Kakinada Bay, the sand spit
Loss of mangrove forests due to aquaculture farms and mangrove waterways of Coringa River. The
initial formation of a small sand spit dates back to
Analysis of the remote sensing data of 2001 1864, which extended to a length of about 16 km
reveals that an area of 658 ha of mangrove vege- by 1968. The survey charts during the period 1848
tation has been converted into aquaculture farms and 1971 revealed that till 1889 the river dis-
outside the forest reserve boundary (Table 5). The charged a major portion of water directly into
rapid growth in shrimp farming, which began in Kakinada Bay. From 1893 to 1929 the river dis-
the Godavari delta during the early 1990s, has charge took place through the Godavari in the
resulted in landuse conversion of agricultural south (Kottapalem mouth). At present, the dis-
fields and mangrove vegetation into shrimp farms. charge is mainly through the mouth near Bhai-
These shrimp farms are located in the revenue ravapalem on the northern side (Ranga Rao et al.
lands and in the private lands abutting the man- 2003). The ground verification of the satellite
groves. In East Godavari district alone the area of images of 1986, 1996 and 2001 also confirm that
shrimp farms has increased from 2006 ha in 1989 the discharge through the mouth near Bhairavap-
to 19,239 ha in 1999. About 14% of the aqua- alem is very high. The southern part of Kakinada
culture farms have been constructed on mangrove Bay is very shallow, with a depth of less than 2 m.
lands outside the forest reserves. The rate of The Corangi mangrove region, including the
conversion of mangroves into shrimp ponds creeks and channels, is also found to be shallow
between 1997 and 1999 increased substantially, near the Bay, with depths varying between 1 and
suggesting that suitable fallow/agriculture lands 3 m. During low tide, large areas of mud flats are
are not available along the coast (Andhra Pradesh exposed in Kakinada Bay.
Remote Sensing Application Centre (APRSAC)
1999).
Conclusion

Geomorphology and shoreline changes The Godavari mangroves and the coastline have
undergone drastic changes within a short period
The base maps of Survey of India toposheets of of about 60 years due to both natural and
1937–1938, 1975–1976 and the remote sensing data
of 1986 and 2001 were used to assess the changes Table 6. Increase in the length of Hope Island spit.
in the shoreline. Appreciable changes are seen in
both the sanctuary and non-sanctuary areas. Year Length of Hope Island spit (km)
Erosion of the coastline can be seen along the Bay 1937–1938 15.60
of Bengal on the northern side of the Godavari 1975–1976 16.33
River, i.e., from the Godavari River mouth to the 1986 16.93
tip of Hope Island. Elongation and enlargement of 1996 17.75
1998 18.00
Hope Island in the north and northwest directions 2001 18.20
is also seen. The length of the sand spit was
 35

Figure 2. Maps showing changes in the area of mangroves between 1986 and 2001.

anthropogenic causes. On the one hand, it shows area is not very much, since the area of accreted
a positive impact on mangrove vegetation as and restored mangroves is almost equal to the
seen by the increase in the mangrove cover in area of degradation, landuse conversion and
accreted areas, as also restoration and conser- erosion. But the drastic change in the island spit
vation activities. On the other hand, the man- needs to be studied in detail for the management
groves facing the sea have vanished due to of mangroves and other coastal developmental
coastal erosion. The overall change in the mangrove activities.
36

Figure 3. Map showing the shoreline changes in Kakinada Bay and Hope Island.
 37

Acknowledgements Ramsey E.W. and Jensen J.R. 1996. Remote sensing of man-
grove wetlands: relating canopy spectra to site-specific data.
Photogramm. Eng. Rem. Sens. 62: 939–948.
The authors thank Prof. M. S. Swaminathan, Ranga Rao V., Reddy B.S.R., Raman A.V. and Ramana
Chairman, Dr. M. Velayutham, Executive Murthy A.V. 2003. Oceanographic features of the Bay –
Director and Dr. V. Selvam, Programme Direc- Mangrove waterways of Coringa, East coast of India. Proc.
tor, M. S. Swaminathan Research Foundation, AP Akademi Sci. 7: 135–142.
for their constant guidance and encouragement. Reddy B.S.R. and Prasad K.V.S.R. 1982. The sand spit near
Kakinada – further studies. Indian J Ear. Sci. 9: 167–173.
The authors wish to thank India-Canada Envi- Selvam V., Ravichandran K.K., Gnanappazham L. and
ronment Facility (ICEF), New Delhi for financial Navamuniyammal M. 2003. Assessment of community based
support. restoration of Pichavaram mangrove wetland using remote
sensing data. Curr. Sci. 85: 794–798.
Sery F., Ducrot D., Mougin E. and Fromard F. 1995. Mapping
on mangrove forest of French Guyana using multisource
References data. In Proceedings of the IGARSS ’95, IEEE 95CH35770,
Florence, Italy, 10–14 July 1995 2: 1122–1124.
Andhra Pradesh Remote Sensing Application Centre (APR- Silapathong Ch. and Blasco F. 1992. The application of geo-
SAC) 1999. Environmental management and monitoring of graphic information systems to mangrove forest manage-
shrimp culture project, East Godavari District, Andhra ment: Khlung, Thailand. Asian Pac. Rem. Sens. J. 5: 97–104.
Pradesh – land use/land cover. Hyderabad, India. Terchunian A., Klemas V. and Asegovia M. 1986. Mangrove
Brakel W.H. 1984. Seasonal dynamics of suspended sediment mapping in Ecuador: the impact of shrimp pond construc-
plumes from Tano and Sabaki rivers, Kenya; analysis of tion. Environ. Manage. 10: 345–350.
coastal imagery. Remote Sensing Envt. 18: 165–173.
Klemas V. and Bartlett D.S. 1975. Coastal zone classification
from satellite imagery. Photogramm. Eng. Rem. Sens. 41: 499–513.

View publication stats