Case study Malaysia: Spatial water quality assessment of Juru, Kuantan and Johor River Basins using environmetric techniques
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Journal of Survey in Fisheries Sciences 7(2) 19-40 2021
Case study Malaysia: Spatial water quality assessment of
Juru, Kuantan and Johor River Basins using environmetric
techniques
Masthurah A.1; Juahir H.2; Mohd Zanuri N.B.1*
Received: June 2020 Accepted: October 2020
Abstract
This study investigates spatial water quality assessment of selected river basins in the
three different states in Malaysia. Environmetric techniques namely, cluster analysis
(CA), principal component analysis (PCA), and discriminant analysis (DA), were
applied to study the spatial variations of the most significant water quality variables in
order to determine the origin of pollution sources on water quality data of Juru River
Basin, Kuantan River Basin and Johor River Basin. 13 water quality parameters were
initially selected and analyzed. Three spatial clusters were formed based on CA, and
these clusters were designated as high pollution source (HPS), medium pollution source
(MPS), and low pollution source (LPS) at the three river basins, respectively. Forward
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
and backward stepwise DA managed to discriminate water quality variables,
respectively from the original 13 variables. The result of this spatial analysis
assessment is supported by PCA (varimax functionality,) that was used to investigate
the origin of each water quality variable due to land use activities. Thus, this analysis
makes it possible to observe the significance of the pollutant sources which contribute
to river pollution. Five principal components (PCs) were obtained for all HPS, MPS
and LPS regions of all the three river basins, respectively. Pollution sources for the
three river basins were mainly originated from industrial waste, municipal waste,
domestics waste and also from agricultural runoffs. Finally, the environmetric
techniques analysis manage to provide convincing result on the spatial variation of
water quality in all the three studied river basins and this eventually will allow more
effective and efficient river quality management activities.
Keywords: Cluster analysis, Principal component analysis, Discriminant analysis
[ DOI: 10.18331/SFS2021.7.2.2 ]
1- Centre for Marine and Coastal Studies (CEMACS), Universiti Sains Malaysia, 11800
Gelugor Pulau Pinang.
2- East Coast Environmental Research Institute (ESERI), Universiti Sultan Zainal Abidin,
Kampung Gong Badak, 21300 Terengganu
*Corresponding author's Email: lailazanuri@usm.my
20 Masthurah et al., Case study Malaysia: Spatial water quality assessment of Juru, Kuantan and …
Introduction concerned due to the land use activities
Water is an important resource that is that affect river ecosystem (Razali,
necessary for all aspect of human and Syed Ismail, Awang, Praveena, &
ecosystem survival (Najah, El-Shafie, Zainal Abidin, 2018) This land use
Karim, & El-Shafie, 2013). Water is activities will gradually alter the types
widely used for irrigation, culturing of of pollutant loadings into the river
fishes and drinking. However in recent system (Juahir, Zain, Aris, Yusoff, &
years, the flow of river contains heavy Mokhtar, 2010). . Spatial analysis can
industrial effluent and municipal be conducted by using environmetric
sewage effluent with the advent of technique. Environmetric or also known
urbanization (Varunprasath & A. Daniel, as chemometric is one of the
2010). The general term water quality environmental analytical chemistry
is used to describe the condition of fields that utilize multivariate statistical
water characteristic including physical, approach for the data analysis (Einax,
chemical and biological characteristic Zwanziger, & Geiss, 1997; Simeonova,
of the water (Dogan, Sengorur, & Simeonov, & Andreev, 2003). It can be
Koklu, 2009). . Many rivers are considered to be the most appropriate
experiencing from deterioration quality analysis performance in order to
of its characteristic condition, which in prevent misinterpretation upon
turn affects people’s health, economy
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
analyzing a large environmental data set
and as well as the environment (Simeonov, Einax, Stanimirova, &
(Department of Environment (DOE), Kraft, 2002; Varmuza & Filzmoser,
2003). Surface water is one of the 2009) . Three common environmetric
environmental components that are analysis that usually perform in order to
most vulnerable to pollution impact classify wide range of data into groups
because this surface river water is the are the hierarchical agglomorative
place that received all of the waste that cluster analysis (HACA) and principal
are being disposed into the river by component analysis (PCA) which are
anthropogenic activities (Hamirdin, then furthered by pattern recognition
2000) In Malaysia, river is the main analysis namely, discriminant analysis
source of drinking water supplies. The (DA) (Adam, 1998) . The objectives of
contaminated river will result into a this study are (i) to evaluate spatial
limited quantity of clean water and thus variations in the river water quality data
will eventually increase the water of Juru, Kuantan and Johor river basins
treatment cost. using environmetric techniques and (ii)
[ DOI: 10.18331/SFS2021.7.2.2 ]
Spatial analysis is one of the to identify the pollution loadings
methods that usually performed for the variations due to land use and
purpose of evaluating and identifying anthropogenic activities in the three
the most significant water quality studied river basins.
parameters that supposed to be
Journal of Survey in Fisheries Sciences 7(2) 2021 21
Study areas and methods discharged into this river. The sources
Site description of pollutant which are mainly from the
Three river basins namely, Juru River sewerage network of Johor Bahru, Pasir
Basin, Kuantan River Basin and Johor Gudang, Ulu Tiram and Kota Tinggi
River Basin have been selected in this cities, the industrial wastewater from
study. many industries in the surrounding and
agricultural wastewater that contain
Johor River Basin fertilizers and pesticides are discharges
Johor which is the second largest state into Johor River (Ismail, 2009).
in Peninsula Malaysia drains a
catchment about 2636 km2 and the Juru River Basin
Johor River Basin become the main The Juru River Basin is about 75 km2 in
river in Johor that located at coordinate area which is originated from Bukit
1°27′00″N 104°01′00″E. This river Mertajam Hill (Lim & Kiu, 1995). This
flows through a north-south direction river basin is made of 2 main upstream
and empties into the Straits of Johor. which are western and eastern upstream.
Water quality of this river basin is The western upstream are Permatang
gradually decline due to the increasing Rawa River and Rambai River while
levels of various pollutants. The the eastern upstream are made of
contaminants eventually flow into Johor Kilang Ubi River and Pasir River.
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
River estuaries which are rich in Meanwhile, Juru River forms the
habitats that provide spawning and middle and downstream (Zali, Retnam,
feeding areas for fish and poultry & Juahir, 2011). Juru River Basin had
(Najah et al., 2013). The Johor River been identified as one of the polluted
Basin originates from Gunung Belumut river in Malaysia ('Department of
and Bukit Gemuruh in the north and Environmental' (DOE), 2008). This
flows to the south eastern part of Johor river has been undergoing extensive
and finally into the Straits of Johor. The monitoring on its water quality as this
maximum length and breadth of this river experience industrial activities
catchment are 80km and 40km, from nearby areas where by this river is
respectively. About 60% of the continuously become the strategic main
catchment area is undulating highland effluent discharges location from
rising to a height of 366m while the manufacturing industrial at Prai
remainders are lowland and swamps. Industrial Estate (Zali et al., 2011).
The highland in the north is mainly Types of industries that are operated
jungle. In the south, a major portion had along Juru River Basin are electronics,
[ DOI: 10.18331/SFS2021.7.2.2 ]
been cleared and planted with oil palm textiles, food processing, metal
and rubber. A great amount of products, rubber, chemical plants and
pollutants from various sources are transport equipment industries
22 Masthurah et al., Case study Malaysia: Spatial water quality assessment of Juru, Kuantan and …
(Alkarkhi, Ismail, Ahmed, & Easa, the period of 1990-2002. Since 1995,
2009) . There is also a vast harbour there was a degradation of forest areas
operation which involves petroleum due to logging activities that had been
based activities located at the estuary carried out nearby the basin and there
which may influence the river water was also an increasing of agricultural
quality during intertidal phase changes area especially in the area of Kenau
(Zali et al., 2011). Juru River Basin and River. Land use activities in the forest
its tributaries flow through urbanized of Kuantan River Basin affect the river
areas and are heavily polluted by water quality due to soil erosion and
domestic waste and discharges from pig
surface runoff (Fig. 1).
farms. According to the DOE report,
water quality of this river basin was
The Data
poor and it seems to be no improvement
For the purpose of this study, data of
over the years (Department of
river water quality from three river
Environmental (DOE), 1977; DOE-
basins, namely Juru River Basin,
USM, 1992).
Kuantan River Basin and Johor River
Basin which consist of a number of
Kuantan River Basin
monitoring stations were obtained from
On the other hand, Kuantan River Basin
Department of Environment (DOE),
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
is about 52,903 ha and located mainly
Ministry of Natural Resource and
in the forest reserve area. Kuantan
Environment of Malaysia. All the water
River Basin is an important water
quality data from each selected station
catchment area which is surrounded by
in this study were based on the
dipterocarp forest area that store variety
available data that had been recorded
of flora and fauna of tropical moist
from 2003-2007. Referring to the
forest. However, over the last decade,
sample site, 5 sites represent the Juru
the development of land for agricultural
sub-basin, namely, Juru, Kilang Ubi,
purposes and the production of natural
Pasir, Rambai, and Ara while 8 sites
resources such as timber had changed
represent Kuantan sub-basin, namely,
the entity coverage of this Kuantan
Belat, Kuantan, Galing Besar, Galing
River basin and indirectly affect the
Kecil, Pinang, Charu, Riau and Kenau.
water river quality of the basin. There is
On the other hand, 21 sites represent the
a clear land use change, particularly in
Johor subbasin namely, Layang,
the area of the origin forest reserve area
Serai(Hilir), Tiram, Tiram(Hulu), Bukit
of Kuantan River Basin. It is reported
[ DOI: 10.18331/SFS2021.7.2.2 ]
Besar, Semanger, Johor, Telor,
that more than 9000 ha of forest land
Berangan, Temon, Layau Kiri,
around the basin has been transformed
Semenchu, Chemangar, Lebam, Sening,
into bushes or agricultural land during
Journal of Survey in Fisheries Sciences 7(2) 2021 23
Santi, Anak Sg. Sayong, Sayong, Penggeli, Sebol and Linggiu.
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
Figure 1: Locations of sampling stations at the three river basins.
Due to the fact that some monitoring biological oxygen demand (BOD),
stations in these three studied river chemical oxygen demand (COD),
basins have missing data, only 13 suspended solid (SS), pH, ammoniacal
consistent parameters were analyzed nitrogen (NH3-N), dissolved solid (DS),
and examined among all the 30 river total solid (TS), nitrate (NO3-), chloride
water quality data available. A total of (Cl-), phosphate (PO43-), Escherichia
205 samples in Juru River Basins, 275 coli, and coliform were subjected for
samples in Kuantan River Basins and the environmetric techniques analysis
865 samples in Johor River Basins were by using XLSTAT2012 software. The
used for the analyses. For this study, all descriptive statistics of the 5 years
[ DOI: 10.18331/SFS2021.7.2.2 ]
the data obtained with 13 water quality measured data set of each river basin
parameters; dissolved oxygen (DO), are summarized in Table 1.
24 Masthurah et al., Case study Malaysia: Spatial water quality assessment of Juru, Kuantan and …
Table 1: Mean values of water quality measurement in Juru, Kuantan and Johor River Basins
(2003-2007).
Juru River Basin
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
Kuantan River Basin
[ DOI: 10.18331/SFS2021.7.2.2 ]
[ DOI: 10.18331/SFS2021.7.2.2 ] [ Downloaded from sifisheriessciences.com on 2021-12-14 ]
Johor River Basin
Journal of Survey in Fisheries Sciences 7(2) 2021
25
26 Masthurah et al., Case study Malaysia: Spatial water quality assessment of Juru, Kuantan and …
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
Methods of analysis cluster analysis groups based on the multivariate
CA operates on data sets and forms similarities among entities (McGarigal,
well-defined groups that actually do not Stafford, & Cushman, 2000) . In this
exist, but are assign to be clustered paper, hierarchical agglomerative
together due to the similar level (HACA) was employed in order to
characteristic that occupied by them. identify the group of river region site
(spatial). Ward’s method, using
[ DOI: 10.18331/SFS2021.7.2.2 ]
The objective of CA is to classify a
sample of entities into a smaller Euclidean distances to measure
numbers usually mutually exclusive similarity in HACA is the common
Journal of Survey in Fisheries Sciences 7(2) 2021 27
analysis used which aimed at Discriminant analysis (DA) refers to
identifying variables that have a high a couple of closely related procedures
rate of homogeneity level into a group that have similar objective in
based on the selection criteria set. discriminating among well-defined
HACA result is illustrated by a group of sampling entities based on a
dendrogram, presenting the clusters and suite of characteristic. This is contrast
their proximity (Juahir et al., 2010). to cluster analysis (CA), which attempt
Agglomerative techniques begin with to organize entities into classes or
each entity in a class of its own, then groups. Cluster analysis often serves as
fuse (agglomerate) the classes into a precursor to DA when prespecified
larger classes. These procedures are groups do not exist where artificial
well known and they are used widely in groups are created by CA, and then
ecological research (McGarigal et al., ecological differences among the newly
2000). In this paper, HACA was created groups are described using DA
conducted in order to determine the (McGarigal et al., 2000). In this paper,
classification of sampling sites (spatial) discriminant analysis was performed in
in the three river basins; Juru River order to examine whether group differ
Basin, Kuantan River Basin and Johor with regards to the mean of variable in
River Basin into groups. predicting the group membership.
In order to conduct this analysis,
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
Discriminant Analysis data from the three assigned region
Discriminant analysis specifies and groups in each of the three river basins
examines variables that are dominant or which had been obtained from CA were
well-discriminated among certain data selected. The discriminant analysis was
groups. Discriminant function (DF) for performed by using standard, forward
each group is created through this DA stepwise and backward stepwise modes.
technique (Johnson & Wichern, 2008) These three modes were performed in
by using below equation, Eq, 1: order to determine water quality
n variables that have high variations
F (Gi) = ki + ∑ wij Pij (1) according to their spatial distribution
j=1 among the three studied river basins. In
Where i is the number of groups (G) , ki the forward stepwise mode, variables
is the constant inherent to each group, n are included gradually starting from the
is the number of parameters used for the most significant variables until no
classification a set of data into a given significant changes are obtained. In
group and wj is the weight coefficient backward stepwise mode, the variables
[ DOI: 10.18331/SFS2021.7.2.2 ]
assigned by discriminant function are removed gradually starting with the
analysis (DFA) to a given parameter less significant variables until no
(pj). significant changes are obtained.
28 Masthurah et al., Case study Malaysia: Spatial water quality assessment of Juru, Kuantan and …
Principal component analysis this study, only the VF coefficient that
Principal component analysis gives have strong loadings (greater than 0.7)
information upon the most significant were being considered. Source
variables according to the spatial and identification of different pollutants was
temporal variations which distinguish made on the basis of different activities
the whole data set by excluding the less in the catchment area in light of
significant parameters with minimum previous literatures. In this paper, PCA
loss of original information (Kannel, was applied to the data set consist of 13
Lee, Kanel, & Khan, 2007; Singh, parameters for each region (HPS, MPS
Malik, Mohan, & Sinha, 2004; Singh,
and LPS) of the three studied river
Malik, & Sinha, 2005) The principal
basins. Calculations of input data
component (PCs) can be expressed as
matrices (variables × cases) for this
(Eq 2):
PCA were 13 × 224 for HPS region, 13
× 312 for MPS region and 13 × 807 for
zij=ai1x1j+aai2x2j+...+...aimxmj
LPS.
(2)
Where z is the component score, a is the Results and discussion
Determination of the sampling station
component loading, x is the measured
groups Cluster analysis
value of variable, i is the component
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
Under this section, water quality
number, j is the sample number, and m
parameters in each data set were
is the total number of variables.
examined by using cluster analysis.
The PCs generated by PCA are
Cluster analysis was performed towards
sometimes not readily interpreted;
all of the data sets in the three river
therefore, it is advisable to rotate the
basins in order to specify each
PCs by varimax rotation. Kim and
monitoring stations according to the
Mueller (1987) states that, the result of
level of their homogenous
the PCs after varimax rotation with the
characteristics. This analysis was
amount of eigenvalues more than 1 are
performed towards all of the water
considered significant for the purpose to quality data in order to assess spatial
acquire new groups of variables called variation among the sampling stations
varimax factors (VFs). The VF on each of the three river basins. The
coefficients which have correlation cluster analysis resulted into three
greater than 0.7 are considered as cluster of sampling stations in Juru,
―strong‖;0.5–0.69, as ―moderate‖; and Kuantan and Johor River Basins,
[ DOI: 10.18331/SFS2021.7.2.2 ]
0.30–0.49, as ―weak‖ significant factor respectively (Fig. 2).
loadings (Liu, Lin, & Kuo, 2003). InJournal of Survey in Fisheries Sciences 7(2) 2021 29
Juru River Basin
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
Kuantan River Basin
Johor River Basin
[ DOI: 10.18331/SFS2021.7.2.2 ]
Figure: 2 Dendrogram shows sampling stations that had been classified in each three river basins.
The cluster procedure formed three similar characteristics and natural
clusters or groups in a very convincing backgrounds. For example, in Juru
way as the sites in each group have River Basin, cluster 1 (2JR01, 2JR02,30 Masthurah et al., Case study Malaysia: Spatial water quality assessment of Juru, Kuantan and …
2JR03, 2JR04, 2JR05, 2JR06, 2JR07) three clusters with three different
represent the high pollution sources regions also generated in the other two
(HPS) region, cluster 2 (2JR10, 2JR11, river basins, Kuantan River Basin and
2JR12) represent the moderate pollution Johor River Basin. Table 2 represents
sources (MPS) region and cluster 3 the sampling stations that have been
(2JR08, 2JR09) represent the low grouped by HACA.
pollution sources (LPS) region. The
Table 2: Sampling stations that have been grouped into regions.
Regions River basins /sampling stations
Juru River Basin Kuantan River Basin Johor River Basin
HPS 2JR01,2JR02,2JR03, 4KN03,4KN04 3JH10,3JH46
2JR04,2JR05,2JR06,
2JR07
MPS 2JR10,2JR11,2JR12 4KN01,4KN02,4KN075, 3JH05,3JH06,3JH09,3JH18,3JH32,
4KN07 3JH35
LPS 2JR08,2JR09 4KN06,4KN08,4KN09, 3JH03,3JH07,3JH08,3JH1,3JH12,
4KN10,4KN11 3JH13,3JH15,3JH16,3JH19,3JH20,
3JH22,3JH25,3JH27,3JH28,3JH30,
3JH33,3JH36,3JH37,3JH40,3JH42,
3JH43,3JH44,3JH45,3JH47
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
The result of HACA technique implies (CA) technique is functional upon the
that rapid assessment of water quality classification of river water quality data
for the whole stations can be done by for optimum future sampling and
monitoring only one station in each monitoring strategies.
cluster that had been assigned by CA. Among the three river basins, Juru
This is because only one monitoring River Basin is proved to be the most
station is already enough to represent polluted river basin as 7 of its
the water quality data for the whole monitoring stations are clustered under
group members in each cluster group as high pollution sources (HPS) group.
every group have their own similar This is due to the fact this river basin
level of data quality characteristics. For receives heavy pollution loadings from
example, in Kuantan River Basin, only nearby urbanized area that densely
station 4KN03 in cluster 1 (HPS), populated by humans and multiple
station 4KN01 (MPS) in cluster 2 and types of industrial located along the
station 4KN06 (MPS) in cluster 3 need basin. In Kuantan River Basin, 2
to be monitored in order to represent the monitoring stations; 4KN03 at Galing
[ DOI: 10.18331/SFS2021.7.2.2 ]
water quality assessment of the whole Besar River and 4KN04 at Galing Kecil
Kuantan River Basins. The result of this River are clustered under high HPS
analysis proved that cluster analysis group. Galing Besar River is mainlyJournal of Survey in Fisheries Sciences 7(2) 2021 31
being polluted by sediment deposition Discriminant Analysis
and siltation that resulted from DA was performed in order to
anthropogenic activities. This river is determine water quality variables that
also overwhelmed with various have high variation according to their
contaminants entering it. The other spatial distribution which had been
monitoring stations are not heavily classified by CA into three main
polluted as Kuantan River Basin is clusters for each river basin. The
mainly located in the forested area. The combination of HPS region for
natural vegetation within this river Juru,Kuantan and Johor River Basin
basin act as filters that prevent the represent as the dependent variables
sediment deposition and uptake nitrate while the water quality parameters were
and phosphorus which is mainly the independent variables. The same
originate from fertilizer usage. There is goes for the combination of MPS and
also a little disturbance along the river LPS region for Juru, Kuantan and Johor
water within this basin which is not River Basin. DA was carried out via
resulted into a serious negative impact standard, forward stepwise and
to the river water quality such as Belat backward stepwise modes. In the
River where by it support mainly forward stepwise mode, variables are
residential areas with light industries included gradually beginning from the
with low water demand and the existing most significant variable until no
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
houses are served by water closets and significant changes are obtained. In
septic tanks, but there is no evidence of backward stepwise mode, variables are
serious water pollution problem removed gradually beginning with the
although the sullage water is discharged less significant variable until no
into surface drain (Hill, 1981). For significant changes are obtained. The
Johor River, station 3JH10 at Bukit accuracy of spatial classification all the
Besar River and station 3JH46 at three regions of the studied river basins
Sayong River are clustered under HPS were 79.33% (13 discriminant variables)
group. The major land use at the Johor for standard mode, 79.33.74% (9
River Basin is oil palm and other crops discriminant variables) for forward
plantations. There are many oil palm stepwise mode and 70.73%
plantation and FELDA land (10discriminant variables) for backward
development located in the surrounding stepwise mode. In forward stepwise
area of Bukit Besar River and Sayong mode, DO, NH3N, CI, pH, NO3-, BOD,
River which may influence to the river COD and PO43- were determined to be
water quality of the two rivers (Hamza, the significant variables which indicate
[ DOI: 10.18331/SFS2021.7.2.2 ]
2009). that all these nine parameters have high
variation upon their spatial distribution
Spatial variation of river water quality in the region of all the three studied32 Masthurah et al., Case study Malaysia: Spatial water quality assessment of Juru, Kuantan and …
river basins. For backward stepwise, E. shown in Table 3. Box and whiskers
coli exist as the tenth parameter that plot of some water quality parameters
have high distribution upon the spatial in five years periods (2003-2007) is
variation. The result of DA which had shown in Fig. 3.
been illustrated in a classification
matrix for each clustered region is
Table 3: Classification matrix for DA of spatial variations in the three studied river basins.
Sampling regions of 3% Regions assigned HPS LPS MPS
studied river basins Correct by DA
Standard DA mode (13 variables) HPS 75.34% 32 2 20
LPS 95.91% 0 635 13
MPS 39.23% 6 59 97
Total 79.55% 38 696 130
Forward stepwise mode ( 9 variables) HPS 77.58% 32 2 20
LPS 95.91% 0 634 14
MPS 39.23% 5 59 98
Total 79.70% 37 695 132
Backward stepwise mode (10 variables) HPS 75.34% 32 2 20
LPS 95.91% 0 634 14
MPS 39.23% 6 60 96
Total 79.33% 38 696 130
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
The Wilk’s Lambda test for standard computed p-value is lower than the
mode gave a Lambda value of 0.281 significance level of alpha=0.05, the
and pJournal of Survey in Fisheries Sciences 7(2) 2021 33
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
Figure 3: Box and whisker plots of some parameters separated by spatial DA associated with the
three river basins. The crosses are mean values, top and bottom of whiskers indicate
maximum and minimum values, respectively while horizontal lines of the boxes from top
to bottom indicate the third quartile, median, and first quartile, respectively.
Principal Component Analysis (PCA) of data set almost 81.9%, 78.8% and
Principal component analysis (PCA) 74.1% respectively. Varimax rotation
was performed on the data set for the that had been performed through this
purpose of identifying the source of PCA technique managed to obtain five
pollutant loading in each clustered varimax function (VF) in each HPS,
region among the three river basins. MPS and LPS region. Table 4 shows
Five PCs were obtained for each HPS, the details of five VFs obtained together
MPS and LPS regions, with the with the amount of variable loading and
[ DOI: 10.18331/SFS2021.7.2.2 ]
concerned amount of eigenvalues variance explained for every region
(larger than 1) sum up the total variance groups in the three river basins.34 Masthurah et al., Case study Malaysia: Spatial water quality assessment of Juru, Kuantan and …
Table 4: Result of VFs that consist of variables loading after varimax rotation for water quality
data in HPS, MPS and LPS regions of the three river basins
High Pollution Source (HPS) can be assumed to be originated from
For HPS region, among five VFs, VF1 point sources (PS) and non-point
accounts for 25.6% of the total variance, sources (NPS) (Ha & Bae, 2001)) as
which have strong positive loadings on these river basins receive a lot of
BOD, COD and SS. In this region, changes in the land development that
loading of BOD and COD are assumed also depends on the seasonal variation
in studied area. Cl- on the other hand, is
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
to be contributed by the direct
discharges from nearby pig farm which identified to be originated from the
are not equipped with proper sanitary mineral salt content in the river. VF3
treatment system (Lim & Kiu, 1995). It explaining15.4% of the total variance
is reported that Juru River flow through has strong loading on E. coli and
largely urbanized areas where it had coliform which indicate the
been polluted by domestic waste and microorganism parameters in the HPS
discharges from pig farms (Lim & Kiu, region of these three rivers. In Juru
1995). The present of COD in the river River for example, source of E. coli and
basins was assumed to come from the coliform are possibly originated from
anthropogenic activities that arise from Juru sewage pond located near the river
the nearby industrial areas that and also from nearby residential areas
discharge their industrial waste into as human settlements including
these three rivers. The strong loading on squatters along the river banks at Juru
SS is possibly originated from the high River are not equipped with proper
load of soil runoff and also from wood sanitary systems.VF4, explaining 10.8%
[ DOI: 10.18331/SFS2021.7.2.2 ]
industry (Zali et al., 2011) nearby these of the total variance and has high
three river basins. VF2, account 20.4% loading of NH3-N and PO43-. The NH3-
total variance with the positive loading N indicates that the HPS region of these
of three variables which are DS, TS three river basins experienced from
AND Cl. The two variables DS and TS pollution that caused by livestock wasteJournal of Survey in Fisheries Sciences 7(2) 2021 35
and as well as the agricultural and loading of DS and TS in this MPS
domestic sewage waste. For example, region are possibly due to extreme river
PO43- loads were mainly originated bank erosion that usually occur during
from agricultural runoff such as the storm flow which eventually cause
fertilizers at Juru River flow nearby the the bedload sediment enter the river
Prai Industrial Estate. VF5, explaining region (Bolstad & Swank, 1997; Hart,
9.5% of the total variance and has 2006). This assumption is reasonable
strong loading on NO3-. The loading of especially to the river water in Kuantan
NO3- is possibly due to the runoff from River Basin which is mainly polluted
agricultural land along the HPS region due to land development through
of these three river basins. This NO3- is agricultural, timber logging and forest
mainly originated from commonly used clearing activities. Strong positive
nitrogen and potassium fertilizers at the loading on NO3- is expected to originate
crop planted area of this HPS region. from the cultivation area (Vega, Pardo,
Barrado, & Debán, 1998) , where crops
Medium Pollution Source (MPS) are planted and the use of inorganic
For MPS, among five VFs, VF1 fertilizers such as ammonium nitrate is
accounts for 25% of the total variance rather frequent (Juahir et al., 2010) .
which include BOD, COD, NH3-N and NO3- may also arise from
PO43-. BOD and COD are among decomposition and degradation of
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
organic factors that assumed to be organic matters containing nitrogen
attributed from anthropogenic activities (USGS, 2007). The organic matters
such as farming and timber logging contained in the municipal waste
activities that take place along the river include urea and protein from the
basin. The presence of NH3-N in the wastewater discharges which enters this
river is due to the excessive runoff from MPS region of the three river basins.
the agricultural area nearby the basin VF3, explaining 14% of the total
regions (Sharip, Zaki, Shapai, Suratman, variance, has strong loading on E. coli
& Shaaban, 2014). The PO43- loading is and coliform which are related to
assumed to be originated from domestic waste and treatment plant
phosphate fertilizer that contain in soils from paper manufacturing industry,
from the agricultural farm area located rubber and palm oil refineries that
nearby the MPS region of these three located near the river (Qadir, Malik, &
river basins. VF2, explaining 23% of Husain, 2008).
the total variance, has strong loadings Normally, faecal contamination from
on DS, TS AND NO3-. Farming and human occurred when structural and
[ DOI: 10.18331/SFS2021.7.2.2 ]
construction were more frequent near technical flaws in the sewerage system
these river basins area and had resulted that causing the sewage to be flowed
into sediment deposited. Thus, the into the river which then leads to the36 Masthurah et al., Case study Malaysia: Spatial water quality assessment of Juru, Kuantan and …
present of E. coli and coliform.VF4, operation that is operated at Johor River
accounts the total variance 8.6%, area. The loading of Cl- is probably
showing loading on SS that can be comes from the mineral constituent in
attributed from high loads of soil and the water of this LPS region. VF2
waste disposal runoff. The last one is represent the total variance of 18.1%
VF5 which accounts 7.9% of the total and show the strong positive loading of
variance and include NO3- as the BOD and COD. The presence of these
positive strong loading variable. NO3- is BOD and COD in this LPS region of
expected to arise from vegetables farm, the three river basins is believed to be
oil palm and rubber plantation that are attributed from the influence of point
located along the MPS region of these source organic pollutants from
three river basins. The nitrate content in sewerage network of the cities located
river water is caused by agricultural nearby the river. VF3, explain 12.5% of
activity that is commonly associated the total variance and has strong
with the use of chemical fertilizer to loadings on E. coli and coliform that
facilitate the growth of trees. Thus, signify the contribution of domestic
when surface runoff occurs during rainy waste to this LPS region. VF4,
season, waste chemical fertilizer will explaining 9% of the total variance and
flow into these basins and caused has strong loading on pH and NO3-. The
increasing of NO3- content in the river. strong loading of pH is expected to
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
arise from several causes such as
Low Pollution Source (LPS) industrial effluent discharges and other
For LPS region, among five VFs, VF1 environmental factors. The decrease of
represent 26.7% of the total variance, pH range into acidic condition are
explaining strong loadings on DO, DS, mainly caused by the industrial effluent
TS and Cl-. The strong negative loading that release acidic discharges into the
on DO is cause by the presence of E. river while the significant increase in
coli in this LPS region of the three river pH level into alkaline condition are
basins which consumed large amount of possibly resulted from environmental
oxygen in order to undergo anaerobic factors such as the rapid algae growth
fermentation. The negative loading of which remove carbon dioxide from the
DO explained that the LPS region in water during the process of
-
these three river basins had been photosynthesis. The NO3 loading may
polluted by municipal waste, oxidation additionally derived from agricultural
ponds and animal husbandry. DS and area where inorganic nitrogen fertilizer
TS can be assumed as the sediment are in common use such as at vegetable
[ DOI: 10.18331/SFS2021.7.2.2 ]
accumulation result that happened due farm near the river. VF5, accounts for
to anthropogenic activities at these three 7.7% of total variance, showing strong
river basins such as sand mining loadings on NH3NL and PO43-. NH3-NJournal of Survey in Fisheries Sciences 7(2) 2021 37
indicates that the LPS region of the basins. Generally, this study had
three river basins experienced from showed the ability of environmetric
pollution that caused by livestock waste techniques for conducting the analysis
and as well as the agricultural and and interpretation of a large complex
domestic sewage waste while a large data set for water quality assessment
amount of PO43- loading is possibly and as well as the identification of
originated from the contamination of pollution sources. This analysis is also
fertilizer and pesticide discharges from useful upon investigating spatial
vegetables farm located nearby the variations of water quality as an effort
river basin. toward a more effective river basin
Environmetric analysis techniques management. Overall results obtained
managed to determine spatial variation from this study indicate that
among the three studied river basins anthropogenic activities have
namely, Juru River Basin, Kuantan significantly influence river water
River Basin and Johor River Basin. quality variations. If these activities are
Cluster analysis has successfully not controlled, they will consequently
classified the cluster region namely, generate a great pressure on the river
HPS, MPS and LPS in each of the three ecosystem and finally become severely
river basins. This classification enables polluted river with the loss of critical
the designation of sampling strategy habitat and overall decrease in the
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
which can reduce the number of quality of life that inhabit this river
sampling stations and the monitoring ecosystem.
cost as on one station in every cluster is
enough to represent the accurate rapid Acknowledgements
assessment of spatial water quality for The authors would like to express
the whole region among the group. whole gratitude and are grateful to
Discriminant analysis on the other hand thanks the Department of Environment
also gives encouraging results upon (DOE) Malaysia for providing the
discriminating the data of every secondary the data for this research
monitoring stations with discriminant project upon the process of completing
variables assigning high correctly this manuscript.
percentage of correlation matrix using
forward and backward stepwise modes. References
Principal component analysis that Adam, M.J., 1998. The Principles of
applied on the data set for each Multivariate Data Analysis. In P. R.
classified region had managed to Ashurst & M. J. Dennis (Eds.),
[ DOI: 10.18331/SFS2021.7.2.2 ]
Analytical Methods of Food
identify the pollutant loading variation Authentication. (pp. 350). London,
due to land use and anthropogenic UK: Blackie Academic &
activities in the three studied river Professional.38 Masthurah et al., Case study Malaysia: Spatial water quality assessment of Juru, Kuantan and …
Alkarkhi, A.F., Ismail, N., Ahmed, A. Ha, S. R., and Bae, M. S., 2001. Effects
and Easa, A., 2009. Analysis of of land use and municipal wastewater
heavy metal concentrations in treatment changes on stream water
sediments of selected estuaries of quality. Environ Monit Assess, 70(1-
Malaysia--a statistical assessment. 2), 135-151.
Environ Monit Assess, 153(1-4), 179- doi:10.1023/a:1010649705723
185. doi:10.1007/s10661-008-0347-x Hamirdin, I., 2000. Pengaruh aktiviti
Bolstad, P.V. and Swank, W.T., 1997. manusia terhadap kualiti air
Cumulative Impacts of landuse on permukaan di Lembangan Langat-
water quality in Southern Semenyih dan Linggi. In H. Mohd
Appalachian Waterhed 1. JAWRA Yusof, I. Nor Azizan, & Z. M.
Journal of the American Water Lukman (Eds.), Isu-isu pembangunan
Resources Association, 33(3), 519- di awal abad ke-21. Bangi, Selangor
533. doi:10.1111/j.1752- (Malaysia): Fakulti Sains
1688.1997.tb03529.x Pembangunan, Universiti Kebangsaan
Department of Environment (DOE). Malaysia.
1977. Report on Water Pollution Hamza, H., 2009. Water quality trend at
Control: Juru River Basin, the upper part of Johor river in
Department of Environment, Ministry relation to raindall and runoff pattern.
of Science, Technology and (Masters), Universiti Teknologi
Environment, Malaysia. Malaysia, Johor, Malaysia.
Department of Environment (DOE). Hart, H.M., 2006. Effect of land use on
2003. Water quality management in the total suspended solids and
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
Malaysia. DOE Documents, Kuala turbidity in the little river watershed,
Lumpur. Blount County, Tennesse. (Masters),
Department of Environment (DOE). University of Tennessee, Knoxville.
2008. River Water Quality Report Hill, A.R., 1981. Stream phosphorus
2008. exports from watersheds with
Department of Enviroment (DOE)- contrasting land uses in Southern
USM. 1992. Development of Water Antario. JAWRA Journal of the
Quality Criteria and Standards American Water Resources
Classification-Juru River, Vol VIII. Association, 17(4), 627-634.
Draft Final Report, Departmental of doi:10.1111/j.1752-
Environment, Malaysia, Innovation 1688.1981.tb01269.x
and Consultancy Centre, Universiti Ismail, M., 2009. Application of
Sains Malaysia. hydrodynamic water quality model in
Dogan, E., Sengorur, B., & Koklu, R. Sungai johor Estuary. (Masters),
2009. Modeling biological oxygen Universiti Teknologi Malaysia, Johor,
demand of the Melen River in Turkey Malaysia.
using an artificial neural network Johnson, R.A., & Wichern, D.W., 2008.
technique. Journal of Environmental Applied Multivariate Statistical
Management, 90(2),1229-1235. Analysis. United States: Pearson.
doi:10.1016/j.jenvman.2008.06.004 Juahir, H., Zain, S. M., Aris, A. Z.,
[ DOI: 10.18331/SFS2021.7.2.2 ]
Einax, J. W., Zwanziger, H.W. and Yusoff, M. K., & Mokhtar, M. B.
Geiss, S. 1997. Chemometrics in 2010. Spatial assessment of Langat
environmental analysis. Germany. river water quality usingJournal of Survey in Fisheries Sciences 7(2) 2021 39
chemometrics. Journal of Razali, A., Syed Ismail, S. N., Awang,
Environmental Monitoring, 12(1), S., Praveena, S. M., & Zainal
287-295. doi:10.1039/B907306J Abidin, E. 2018. Land use change in
Kannel, P. R., Lee, S., Kanel, S. R., & highland area and its impact on river
Khan, S.P. 2007. Chemometric water quality: a review of case studies
application in classification and in Malaysia. Ecological Processes,
assessment of monitoring locations of 7(1), 19. doi:10.1186/s13717-018-
an urban river system. Analytica 0126-8
Chimica Acta, 582(2), 390-399. Sharip, Z., Zaki, A. T. A., Shapai, M.
doi:10.1016/j.aca.2006.09.006 A. H. M., Suratman, S., & Shaaban,
Kim, J. O., & Mueller, C. W. 1987. A. J. 2014. Lakes of Malaysia: Water
Introduction to factor analysis: What quality, eutrophication and
it is and How to do it (Quantitative management. Lakes & Reservoirs:
application in sosial sciences). United Science, Policy and Management for
State: SAGE UNIVERSITY PAPER. Sustainable Use, 19(2), 130-141.
Lim, P.E., & Kiu, M.Y. 1995. doi:10.1111/lre.12059
Determination and speciation of Simeonov, V., Einax, J., Stanimirova,
heavy metals in sediments of the Juru I., & Kraft, J. 2002. Environmetric
River, Penang, Malaysia. modeling and interpretation of river
Environmental Monitoring and water monitoring data. Analytical and
Assessment, 35(2), 85-95. Bioanalytical Chemistry, 374(5), 898-
doi:10.1007/BF00633708 905. doi:10.1007/s00216-002-1559-5
Liu, C. W., Lin, K. H., & Kuo, Y. M. Simeonova, P., Simeonov, V., &
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
2003. Application of factor analysis Andreev, G. 2003. Water quality
in the assessment of groundwater study of the Struma river basin,
quality in a blackfoot disease area in Bulgaria (1989–1998). Central
Taiwan. Sci Total Environ, 313(1-3), European Journal of Chemistry, 1(2),
77-89. doi:10.1016/s0048- 121-136. doi:10.2478/BF02479264
9697(02)00683-6 Singh, K. P., Malik, A., Mohan, D., &
McGarigal, K., Stafford, S., & Sinha, S. 2004. Multivariate
Cushman, S. 2000. Multivariate statistical techniques for the
Statistics for Wildlife and Ecology evaluation of spatial and temporal
Research New York, NY: Springer variations in water quality of Gomti
New York. River (India)—a case study. Water
Najah, A., El-Shafie, A., Karim, O. A., Research, 38(18), 3980-3992.
& El-Shafie, A. H. 2013. Application doi:10.1016/j.watres.2004.06.011
of artificial neural networks for water Singh, K. P., Malik, A., & Sinha, S.
quality prediction. Neural Computing 2005. Water quality assessment and
and Applications, 22(1), 187-201. apportionment of pollution sources of
doi:10.1007/s00521-012-0940-3 Gomti river (India) using multivariate
Qadir, A., Malik, R. N., & Husain, S. statistical techniques—a case study.
Z. 2008. Spatio-temporal variations in Analytica Chimica Acta, 538(1), 355-
water quality of Nullah Aik-tributary 374. doi:10.1016/j.aca.2005.02.006
[ DOI: 10.18331/SFS2021.7.2.2 ]
of the river Chenab, Pakistan. United States Geological Survey
Environ Monit Assess, 140(1-3), 43- (USGS). 2007. Water quality in the
59. doi:10.1007/s10661-007-9846-4 upper Anacostia River, Maryland:40 Masthurah et al., Case study Malaysia: Spatial water quality assessment of Juru, Kuantan and …
Continous and discreate monitoring Vega, M., Pardo, R., Barrado, E., &
with simulation to estimate Debán, L. 1998. Assessment of
concentrations and yields. Retrieved seasonal and polluting effects on the
from Virginia, United States: quality of river water by exploratory
Varmuza, K., & Filzmoser, P. 2009. data analysis. Water Research, 32(12),
Introduction to Multivariate 3581-3592. doi:10.1016/S0043-
Statistical Analysis in Chemometrics. 1354(98)00138-9
Varunprasath, K., & A. Daniel, N. Zali, M. A., Retnam, A., & Juahir, H.
2010. Physico-Chemical Parameters 2011. Spatial Characterization of
of River Bhavani in Three Stations, Water Quality Using Principal
Tamilnadu, India. Iranian (Iranica) Component Analysis Approach at
Journal of Energy & Environment, Juru River Basin, Malaysia.
1(4).
[ Downloaded from sifisheriessciences.com on 2021-12-14 ]
[ DOI: 10.18331/SFS2021.7.2.2 ]
Powered by TCPDF (www.tcpdf.org)You can also read
