approach to illustrate landscape patterns and measure spatial heterogeneity is via
landscape metric analysis (Yu et al. 2019). These metrics function as indicators of
structural composition and spatial configuration of landscapes. The highest values of
CONTAG (86.34%), AI (99.2%), COHE (99.74), SHMN (1.77), and LPI (93.75%)
and the lowest values of SHDI (0.27), PD (0.052/100 ha), and ED (1.92 m/ha) were
all recorded in zone 7. While the highest values of SHDI (1.25), ED (12.95 m/ha),
and PD (0.28/100 ha) were recorded in zone 5. Whereas, zone 1 had displayed the
lowest values of ENNMN (488.4 m), SHMN (1.55), and LPI (40.1%). There were
clear differences with correlations between these metrics and water quality parameters in both seasons, which highlight the sensitivity toward seasonality and hydrological changes.
By using cluster analysis, these zones are further categorized into 22 sampling
sites based upon pollution load, i.e., signified slight pollution, moderate pollution,
and high pollution load to analyze the seasonal variations (Fig. 2.4). During the
normal season, cluster 1 included sites 4, 13, and 14. These sites are located at the
brinks or on the perimeter of Bentong River and are mainly surrounded with forest
cover and limited population. Naturally, they should be much less polluted than the
other sites. Cluster 2 comprised of sites 3, 6, 7, 9, 11, and 22; and cluster 3 covered
the rest of the sites. Clusters 2 and 3 are somewhat similar but vary in pollution
severity; both are represented by mostly residential and agricultural areas. Still, the
sites in cluster 3 are located in areas that contain more industrial activities, infrastructure, and business practices besides the residential and agricultural parts. Therefore, these sites are subjected to more wastewater discharge and much more
contaminated runoff.
During the wet season, cluster 1 included sites 1, 2, 3, 5, 6, 7, 8, 9, 10, 12, 14, 15,
and 21. While cluster 2 contained sites 4, 13, and 22. Interestingly, most of the
sampling sites in cluster 1 from the normal season shifted to cluster 2 in the wet
season, which is significantly different. The reason could be attributed to runoff
resulting from rainfalls, which carries various pollutants and compounds from
nearby and surrounding areas, especially agricultural lands, and deposit them in
waterways. Cluster 3 comprised sites 11, 16, 17, 18, 19, and 20, which are primarily
positioned in zone 4. These sites, particularly sites 16 and 17, are located downstream of the study area, thus the generated runoff passing through various areas
would ultimately worsen the water quality downstream as cited by (Absalon and
Matysik (2007)). Additionally, Ravichandran (2003)) stated that the confluence of
rivers sometimes leads to water quality deterioration due to different water chemical
composition and the alterations in hydrology.
28
N. R. Jamil and Z. N. Shehab
landscape metric analysis (Yu et al. 2019). These metrics function as indicators of
structural composition and spatial configuration of landscapes. The highest values of
CONTAG (86.34%), AI (99.2%), COHE (99.74), SHMN (1.77), and LPI (93.75%)
and the lowest values of SHDI (0.27), PD (0.052/100 ha), and ED (1.92 m/ha) were
all recorded in zone 7. While the highest values of SHDI (1.25), ED (12.95 m/ha),
and PD (0.28/100 ha) were recorded in zone 5. Whereas, zone 1 had displayed the
lowest values of ENNMN (488.4 m), SHMN (1.55), and LPI (40.1%). There were
clear differences with correlations between these metrics and water quality parameters in both seasons, which highlight the sensitivity toward seasonality and hydrological changes.
By using cluster analysis, these zones are further categorized into 22 sampling
sites based upon pollution load, i.e., signified slight pollution, moderate pollution,
and high pollution load to analyze the seasonal variations (Fig. 2.4). During the
normal season, cluster 1 included sites 4, 13, and 14. These sites are located at the
brinks or on the perimeter of Bentong River and are mainly surrounded with forest
cover and limited population. Naturally, they should be much less polluted than the
other sites. Cluster 2 comprised of sites 3, 6, 7, 9, 11, and 22; and cluster 3 covered
the rest of the sites. Clusters 2 and 3 are somewhat similar but vary in pollution
severity; both are represented by mostly residential and agricultural areas. Still, the
sites in cluster 3 are located in areas that contain more industrial activities, infrastructure, and business practices besides the residential and agricultural parts. Therefore, these sites are subjected to more wastewater discharge and much more
contaminated runoff.
During the wet season, cluster 1 included sites 1, 2, 3, 5, 6, 7, 8, 9, 10, 12, 14, 15,
and 21. While cluster 2 contained sites 4, 13, and 22. Interestingly, most of the
sampling sites in cluster 1 from the normal season shifted to cluster 2 in the wet
season, which is significantly different. The reason could be attributed to runoff
resulting from rainfalls, which carries various pollutants and compounds from
nearby and surrounding areas, especially agricultural lands, and deposit them in
waterways. Cluster 3 comprised sites 11, 16, 17, 18, 19, and 20, which are primarily
positioned in zone 4. These sites, particularly sites 16 and 17, are located downstream of the study area, thus the generated runoff passing through various areas
would ultimately worsen the water quality downstream as cited by (Absalon and
Matysik (2007)). Additionally, Ravichandran (2003)) stated that the confluence of
rivers sometimes leads to water quality deterioration due to different water chemical
composition and the alterations in hydrology.
28
N. R. Jamil and Z. N. Shehab