JFS

Regular Article
ISSN 1226-8267
Journal of Forest Science Journal of Forest Science
Vol. 28, No. 3, pp. 152-157, August, 2012
http://dx.doi.org/10.7747/JFS.2012.28.3.152

Monitoring of Deforestation and Fragmentation

in Sarawak, Malaysia between 1990 and 2009
Using Landsat and SPOT Images
1, 1 1,2 3 1
Kamlisa Uni Kamlun *, Mia How Goh , Stephen Teo , Satoshi Tsuyuki and Mui-How Phua
1
School of International Tropical Forestry, Universiti Malaysia Sabah, Sabah, Malaysia
2
Forest Department Sarawak, Level 5, Wisma Sumber Alam, 92660 Kuching, Sarawak, Malaysia
3
Laboratory of Global Forest Environmental Studies, Department of Global Agricultural Sciences, Graduate School of Agricultural and Life Sciences,
The University of Tokyo, Bunkyo, Tokyo, Japan

Abstract
Sarawak is the largest state in Malaysia that covers 37.5% of the total land area. Multitemporal satellite images of
Landsat and SPOT were used to examine deforestation and forest fragmentation in Sarawak between 1990 and 2009.
Supervised classification with maximum likelihood classifier was used to classify the land cover types in Sarawak. The
overall accuracies of all classifications were more than 80%. Our results showed that forests were reduced at 0.62%
annually during the two decades. The peat swamp forest suffered a tremendous loss of almost 50% between 1990
and 2009 especially at coastal divisions due to intensified oil palm plantation development. Fragmentation analysis revealed
the loss of about 65% of the core area of intact forest during the change period. The core area of peat swamp forest
had almost completely disappeared during the two decades.

Key Words: land cover change, fragmentation, remote sensing, GIS, Borneo

Introduction referred to as patches (Turner et al. 2001). Fragmentation is

Deforestation is a long term or permanent loss of forest
cover, which is converted to other land use that is caused ei-
ther by natural or human causes (FAO 2007). It accounts
for 17% of total anthropogenic greenhouse gases (GHGs)
emission (IPCC 2007), with the major source from defor-
estation at an unprecedented rate in the tropics (Defries et
al. 2007). Deforestation not only increases the GHGs emis-
sions but also changes the landscape configuration that
leads to habitat fragmentation (Skole and Tucker 1993).
Habitat fragmentation is the breakup of a habitat or cover
type into smaller and less connected areas which commonly
associated with the fluxes of species, energy and matter af-
fecting the ecological conditions of the remaining forests
(Gasparri and Grau 2009). Continuation of fragmentation
will initiate threats to biodiversity of a natural ecosystem
(Abdullah and Nakagoshi 2006).
Deforestation is obvious to observe but associated frag-
mentation is a different dimension of forest changes that is
equally important to monitor. A combination of remote
sensing and Geographic Information System (GIS) pro-
vide a dynamic information system for land use planning
and developing national development policy (Sohel et al.
2011). Multitemporal monitoring of forest changes that are

Received: June 14, 2012. Revised: August 8, 2012. Accepted: August 9, 2012.

Corresponding author: Kamlisa Uni Kamlun

School of International Tropical Forestry, Universiti Malaysia Sabah, UMS Road, Kota Kinabalu, Sabah 88400, Malaysia
Tel: 6088320118, Fax: 6088320876, E-mail: kamlisa@ums.edu.my

linked to fragmentation analysis are invaluable techniques
of assessing the degree of threats to a forest ecosystem
(Sader et al. 2001). Landscape metrics have been used to
monitor the pattern of forest changes that is not readily visi-
ble to human eye or easily detectable by human analysis
(Frohn 1998). Riitters et al. (2000) developed a forest frag-
mentation model that classifies forest pixels into six catego-
ries based on forest density and connectivity. This model
has been useful to multitemporal monitoring of fragmenta-
tion of peat swamp forest in Sabah, Malaysia (Phua et al.
2008).
Sarawak is the largest state (12.3 million ha) in Malaysia
that occupies 37.5% of the total land. One of the driven
forces of land cover change is the dependency on state econ-
omy of forestry that contributes to 52% of state revenue
(Hansen 2005). Land cover change in Sarawak has been
monitored for recent decades with Landsat data, which
were affected by clouds and shadows. Expansion of agricul-
tural land especially for oil palm plantation was the main
cause of deforestation in recent years (Tsuyuki et al. 2011).
In this paper, we provide an update on land cover changes
in the recent two decades and monitor fragmentation of the
forest types that suffer highest area loss. We used satellite
images of Landsat and SPOT for monitoring deforestation
in Sarawak at a landscape scale between 1990 and 2009.
Based on the land cover classifications, patterns of forest
fragmentation were monitored using the forest fragmenta-
tion model of Riitters et al. (2000).
Materials and Methods
Study area
Sarawak is the largest state of Malaysia that is situated in
o o
Borneo Island. It is located between 0 50’ and 5 00’ latitude
o o
and 109 35’ and 115 40’ longitude. The total area of
2
Sarawak is 123,300 km and about 720 km from Peninsular
Malaysia. Sarawak is categorized as humid tropical region
with a wet season from October to February. The topography
is flat and undulating. Hilly area is found in the southern
part, near the border between Sarawak and Kalimantan.
There are five main natural forest types in Sarawak
namely; hill mixed Dipterocarp forest, peat swamp forest,
mangrove forest, kerangas forest (heath forest) and mon-
tane forest. Distributions of these forest types depend on al-
Kamlun et al.
titude, soil conditions and water regime. While there are
still large tracts of forests remain, the inland areas have been
modified by shifting cultivation activities. Recent agricul-
tural development especially oil palm plantation dominates
the coastal plains of Sarawak.
Data acquisition and pre-processing
Multi-temporal Landsat satellite images between 1990
and 2009 were the main data for monitoring deforestation
in Sarawak. A total of nine scenes were needed to cover the
whole area of Sarawak. Nine images of Landsat5-TM ac-
quired in 1990 were downloaded from USGS website.
Multiple Landsat7-ETM+ images in 2009 were down-
loaded for each of the nine scene areas for correcting the
missing line errors. The Landsat7-ETM images contained
5% to 40% cloud covers. For each scene area, the best-qual-
ity image with least cloud cover was placed on top while the
remaining images were used to fill the missing lines. A mi-
nority filter (3x3) followed by a median filter (5x5) was
used to eliminate the linear residues resulted from the miss-
ing line correction. In addition, three SPOT 4-HRVIR
images acquired in 2010 were used to fill in areas that were
heavily affected by clouds.
Deforestation rates and pattern
Post-classification comparison approach was used to
identify the land cover change in Sarawak between 1990
and 2009. Most of the land cover change studies adopt ei-
ther supervised or unsupervised classification in determin-
ing the spectrally distinctive land cover classes (Smits et al.
1999). In this study, all the preprocessed satellite images
were classified using supervised classification with max-
imum likelihood algorithm. Ten land cover classes were de-
fined; intact forest, degraded forest, peat swamp forest
(PSF), mangrove, rubber, oil palm, grassland, bareland,
water and cloud were classified. 5x5 majority filtering was
applied on the land cover classification maps to reduce the
salt and pepper effect. Digital Elevation Model (DEM) of
the Shuttle Radar Topography Mission (SRTM) data was
used to correct forest areas above an elevation of 60m that
were misclassified as PSF. The areas covered by clouds
were masked out from both images.
For land cover classification of 1990, reference data were
generated from historical land use maps with scales of
J For Sci 28(3), 152-157 153
Monitoring of Deforestation and Fragmentation in Sarawak, Malaysia between 1990 and 2009 Using Landsat and SPOT Images

1:50,000 and 1:25,000 from the Department of Survey and
Mapping, Malaysia. The maps covered areas of Nanga
Tamin, Tian, Sibu, Sarikei, Serantok and Kuching. We
generated 660 points using user defined point for assessing
the accuracy of the land cover classification of 1990. For
land cover classification of 2009, 450 points were collected
for accuracy assessment. Of these reference points, 105
points were collected in the field works in December 2009
and July 2010. The remaining ground truth points were
obtained from high-resolution quick-looks images in
Google Earth.
Annual deforestation rate can be calculated as follows
(Puyravaud 2003),
  
P  
 

  
Where A1 and A2 are the forest cover at time t1 and t2 re-
spectively (the unit indicates per year or percentage per
year).
Fragmentation analysis
Two key parameters in the Riitters model are forest den-
sity (Pf) and forest connectivity (Pff). Both parameters are
calculated for a focal pixel based on neighborhood analysis.
Pf within a moving window is calculated as follows,
Pf =
number of forest pixels
number of all non water pixels
Similarly, Pff is calculated from the neighborhood with
the following formula:
number of pixel pairs with both pixels forest
Pff =
number of pixel pairs with at least one pixel forest
Based on the two parameters, the fragmentation model
identifies six categories of fragmentation classes of interior,
undetermined, perforated, edge, transitional and patch.
The calculated Pf and Pff were converted to percentage with
a multiplication factor of 100. Using density and con-
iv. edge, if density＞60% and density＜connectivity
v. transitional, if density is between 40% and 60%
vi. patch, if density＜40%
Forest fragmentation classes can be differentiated using
different threshold values. The interior classis defined with
density above 90% and core pixels with density of 100%.
Any remaining pixels, although rare, is labelled as un-
determined (Fig. 1).
In this study we examined the fragmentation for intact
forest, peat swamp forest and mangrove forest. These forest
types were extracted from the land cover classifications of
1990 and 2009 for forest fragmentation analysis using the
Riitters model in SAGA-GIS. The connectivity values
were calculated with an immediate neighborhood, 3x3
moving windows.
Results and Discussion
Land cover classification accuracy assessment
Accuracies of land cover classifications in 1990 and 2009
were assessed using separate sets of reference points. The
overall accuracies of land cover classifications of 1990 and
2009 were 86% and 80% respectively. Land cover classi-
fication accuracy that is higher than 80% is considered ac-
curate and reliable (Swain and Davis 1978). We also calcu-
lated the kappa accuracy for each of the land cover
classifications. The Kappa accuracy was 0.84 and 0.75 for
land cover classification in 1990 and 2009 respectively. The
Core
Interior
100
Perforated
80
Edge
Density [%]

60
nectivity, each pixel can be assigned to one of six forest frag- Transitional
mentation classes. If we express the density and con- 40
nectivity ratios as percentages by multiplying it with factor
Patch
100, the fragmentation classes can be defined as follow, 20
i. interior, if density=100%
No Forest
ii. undetermined, if density＞60% and density=con- 0
nectivity 0 20 40 60 80 100
iii. perforated, if density＞60% and density＞con- Connectivity [%]

nectivity Fig. 1. Forest fragmentation categories.

154 Journal of Forest Science http://jofs.or.kr


Fig. 2. Land cover classification of 1990.
Fig. 3. Land cover classification of 2009.
missing lines in Landsat7-ETM+ data were likely the
cause of the lower accuracy in the land cover classification
of 2009.
Land cover change and deforestation
Fig. 2. shows that the intact forest, with more than 6 mil-
Kamlun et al.
Fig. 4. Coastal Mukah area shows conversion of peat swamp forest to other
land use from 1990 image satellite to 2009 image satellite.
lion ha, was the largest land cover type in 1990. Although it
remained as the largest land cover type in 2009, more than
550,000 ha were deforested (Fig. 3). Half a million ha of
peat swamp forest was cleared from 1990 to 2009 (Fig. 5).
The forests of Sarawak have been pressured by large-scale
expansion of oil palm development (Tsuyuki et al. 2011).
Our results showed that the forests (intact forest, degraded
forest, peat swamp forest and mangrove forest) were re-
duced at 0.62% annually during the two decades. The oil
palm was the fastest growing agricultural crop in Sarawak.
Oil palm plantation increased about 70% annually between
1990 and 2008 (Mohd Basri and Mohd Arif 2009). Our
results indicated an annual increment of 77.4% for the oil
palm plantation. The Sarawak state government had tar-
geted in 2007 to develop about 1 million ha of oil palm
plantation by 2010 (MPOB 2007). Our findings showed
that Sarawak’s oil palm plantation has reached about 1 mil-
lion ha in 2009. These recent oil palm plantation develop-
ments took place at the peat swamp forest, mainly located at
the coastal divisions of Sibu, Mukah, Bintulu and Miri
(Fig. 4).
In 1980, 1.6 million ha of peat swamp forest was re-
ported (Wong 1991). Our results indicated that only about
J For Sci 28(3), 152-157 155
Monitoring of Deforestation and Fragmentation in Sarawak, Malaysia between 1990 and 2009 Using Landsat and SPOT Images
Fig. 5. Land cover types of Sarawak in 1990 and 2009.
1 million ha of the peat swamp forest was found in 1990 and
almost half a million ha was cleared in the following two
decades. Oil palm planters prefer peatland for planting oil
palm because of the higher yield of production (Phillips
1998). In addition, planting oil palm on flat plain provides
better harvest management. The deforestation rate of peat
swamp forest between 1990 and 2009 was calculated at
2.6% per year. With Moderate-resolution Imaging
Spectoradiometer (MODIS) data (250 m), Miettinen et al.
(2011) found that peatlands in Sarawak have been drasti-
cally reduced with an annual rate of 8% between 2000 and
2010. The difference could be attributable to the differ-
ences in spatial resolution and change period.
Fragmentation of intact forest
Applying the Riitters model of fragmentation in
SAGA-GIS showed that the intact forest in Sarawak has
become more fragmented in 2009, compared to 1990. The
core area had declined more than 600,000 ha from year
1990 to 2009 (Fig. 6). Over the two decades, the patch
areas had reduced about half. Such rapid disappearance in-
dicates that the patches are converted to other land cover
types (Manier and Laven 2001).
For peat swamp forest, 95% of the core area was lost dur-
ing the two decades. There was a reduction of 58872 ha
from 1990 to 2009 for the patch category, which indicates
very high degree of fragmentation. Due to the high demand
of timber products, both within and outside of Malaysia,
overall timber production of Sarawak had been driven to its
peak capacity (Majid-Cooke 1999). Deforestation in
Sarawak is often associated with logging and reclassi-
156 Journal of Forest Science http://jofs.or.kr
Fig. 6. Forest fragmentation from 1990 to 2009 (ha).
fication of the state forest area to allow conversion to oil
palm plantations especially in peat swamp forest area
(Hansen 2005). The commercial loggers were identified as
the main culprits of major forest fragmentation in Sarawak
but not shifting cultivators (Nor Rasidah 2010).
The core class for mangrove forest was only 35ha in
1990 and completely lost in 2009. The producer and user
accuracies of the classification of mangrove forest in 2009
were relatively low. Mangrove is difficult to classify due to
spectral confusion with other types of wetlands (Ozesmi
and Bauer 2002). However, the trends of rapid defor-
estation and fragmentation were clear for the mangrove
forest.
Conclusions
Rapid deforestation in the tropics threatens biodiversity
and accelerates carbon emissions into the atmosphere. We
classified the land cover typesof Sarawak in 1990 and 2009
using Landsat and SPOT data for land cover change
analysis. The intact forest shrank more than half a million
ha during that period. The peat swamp forest lost half dur-
ing the change period. Overall, forests were cleared at
0.62% per year in the two decades mainly for oil palm plan-
tation, which increased more than 70% per year in the two
decades. The intact forest had become very fragmented.
The peat swamp forest had become critically fragmented
and the core area almost completely disappeared. It was at-
tributed to recent intensified oil palm development in the
peat swamp forest. The deforestation and fragmentation

during the two decades reflect the emphasis of agricultural
development over sustainable forestry in the largest state of
Malaysia. Similar trend is expected to continue unless more
forests are gazetted as permanent forest reserves.
Acknowledgements
We thank Sarawak Forestry Department for support and
cooperation. This study was supported from the Grant of
Asia Pacific Network for Global Change Research
(ARCP2010-02CMY-Phua) and Postgraduate Research
Grant of Universiti Malaysia Sabah (GPS0016).
References
Abdullah SA, Nakagoshi N. 2006. Forest fragmentation and its
correlation to human land use change in the state of Selangor,
peninsular Malaysia. Forest Ecology and Management 241:
39-48.
DeFries R, Achard F, Brown S, Herold M, Murdiyarso D,
Schlamadinger B, de Souza Jr C. 2007. Earth observations for
estimating greenhouse gas emissions from deforestation in de-
veloping countries. Environmental Science and Policy 10:
385-394.
FAO. 2007. Definitional issues related to reducing emissions from
deforestation in developing countries. Forest and Climate
Change Working Paper. Vol. 5.
Frohn RC. 1998. Remote Sensing for Landscape Ecology: New
Metrics for Monitoring, Modeling, and Assessment of Ecosy-
stems. Lewis Publishers, Florida.
Gasparri NI, Grau HR. 2009. Deforestation and fragmentation of
Chaco dry forest in NW Argentina (1972-2007). Forest Ecol
Manage 258: 913-921.
Hansen TS. 2005. Spatio-temporal aspect of land use and land cov-
er change in the niah catchment, Sarawak, Malaysia. Singapore
J Trop Geogr 26: 170-190.
IPCC. 2007. Climate Change 2007: Synthesis report. Intergovern-
mental Panel on Climate Change (IPCC). http://www.ipcc.ch.
Accessed 24 Dec 2011.
Majid-Cooke F. 1999. The challenge of sustainable forests: forest
resource policy in Malaysia, 1970-1995.
Malaysian Palm Oil Board (MPOB). 2007. Sarawak Sasar 1 Juta
Hektar 2010.
Malaysia Oil Palm Board (MPOB). 2007. Industri Sawit Tingkatkan
Sosioekonomi. Malaysia Palm Oil Board. http://mpob.gov.
my/html/05_pi/gfx_bs/bs66_01a.jpg. Accessed 18 Aug 2012
Manier DJ, Laven RD. 2001. Changes in landscape patterns asso-
ciated with the persistence of aspen (Populus tremuloides
Michx.) on the westernslope of the Rocky Mountains, Colorado.
Kamlun et al.
Forest Ecology and Management 167: 263-284.
Miettinen J, Shi C, Liew SC. 2011. Deforestation rates in insular
Southeast Asia between 2000 and 2010. Global Change Biology
17: 2261-2270.
Mohd Basri W, Mohd Arif S. 2009. Issues related to Production
Cost of Palm Oil in Malaysia. Oil Palm Industry Economic
Journal 9: 1-12.
Nor Rasidah H. 2010. Analyzing long-term landscape changes in a
borneon forest reserve using. Aerial photographs. Open Geogr J
3: 161-169.
Ozesmi SL, Bauer ME. 2002. Satellite remote sensing of wetlands.
Wetlands Ecology and Management 10: 381-402.
Phillips VD. 1998. Peatswamp ecology and sustainable develop-
ment in Borneo. Biodiversity and Conservation 7: 651-671.
Phua MH, Conrad O, Kamlisa UK, Fischer M, Böhner J. 2008.
In: Multitemporal Fragmentation Analysis of Peat Swamp
Forest in the Klias Peninsula, Sabah, Malaysia using GIS and
Remote Sensing Techniques (Böhner J, Blaschke T, Montanare-
lla L, eds). Hamburger BeiträgezurPhysischenGeographie und
Landschaftsökologie, Universität Hamburg Institutfür Geo-
graphie, Hamburg, pp 81-90.
Puyravaud JP. 2003. Standardizing the calculation of the annual
rate of deforestation. Forest Ecology and Management 177:
593-596.
Riitters K, Wickham J, O'Neill R, Jones B, Smith E. 2000.
Global-scale patterns of forest fragmentation. Conservation
Ecology 4: 3.
Sader SA, Hepinstall DJH, Coan M, Soza C. 2001. Forest change
monitoring of a remote biosphere reserve. Int J Rem Sens 22:
1937-1950.
Skole D, Tucker C. 1993. Tropical deforestation and habitat frag-
mentation in the Amazon: satellite data from 1978 to 1988.
Science 260: 1905-1910.
Smits PC, Dellepiane SG, Schowengerdt RA. 1999. Quality as-
sessment of image classification algorithm for land-cover map-
ping: review and a proposal for a cost- based approach. Int J
Rem Sens 20: 1461-1486.
Sohel MSI, Rana MP, Islam AZMZ, Akhter S. 2011. Monitoring
land-use changes by remote sensing and GIS techniques: case
study of barind tract, Bangladesh. Journal of Forest Science 27:
73-79.
Swain PH, Davis SM. 1978. Remote sensing: the quantitative
approach. McGraw-Hill, New York.
Tsuyuki S, Goh MH, Teo S, Kamlisa UK, Phua MH. 2011.
Monitoring deforestation in Sarawak, Malaysia using multi-
temporal landsat data. Kanto For Res J 62: 87-90.
Turner MG, Gardner RH, O'Neill RV. 2001. Landscape Ecology
in Theory and Practice: Pattern and Process. Springer Science,
New York.
Wong MH. 1991. The Distribution Characteristics and Agricul-
tural Utilization of Peat in Sarawak. Kuching, Malaysia.
Sarawak Agriculture Department, Kuching.
J For Sci 28(3), 152-157 157