different sites exposed to mine wastes (Xiao and Ji 2007). In fact, several studies
have demonstrated that landscape metrics are notable factors that illustrate the
association between land uses and water quality and its effect on the hydrological
processes (Shen et al. 2014).
Landscape impact on water characteristics varies over seasons (Pratt and Chang
2012), and is mainly reliant on scale (Zhou et al. 2012). Correlations of landscape
metrics with water quality parameters are shown in Table 2.2. SHMN is considered
as an indicator of the complexity of land use shapes. This index is found to be
positively correlated with DO (p < 0.01) during the normal season and showed a
significant negative correlation with BOD, turbidity, and ammoniacal nitrogen,
which implies that more complexity in land use shapes that may result in a better
water quality (Lee et al. 2009). SHDI reflects the number of land uses in the area and
the proportion of change in land shapes. The presence of numerous different land use
types in a certain area leads to higher SHDI values (McGarigal et al. 2002). This
metric is positively correlated with almost all the variables and negatively with DO,
which signifies that water quality tends to decline with high intersperse of different
land use types as well as high spatial occupancy (Lee et al. 2009). Furthermore, this
consequently alters water purification functions due to the constant changes in the
natural environment (Xia et al. 2012). CONTAG percentage echoes the aggregated
proportion or dispersion of land use patterns that are present in a certain area, which
tends to be high when there is a low level of land use dispersion (Bu et al. 2014).
During both seasons, CONTAG correlated negatively with almost all of the variables (significantly with E.C, turbidity, pH, temperature, DO, and TSS), this would
indicate that water deterioration typically happens in landscapes that have a high
level of dispersion and are greatly fragmented and uneven (Uuemaa et al. 2005).
Likewise, ED and PD metrics are considered as pointers to the unevenness and
fragmentation of a landscape (Fichera et al. 2012). McGarigal and Marks (1995)
observed that PD and ED values would rise, while LPI value would drop when there
are numerous small spots and patches of land covers in the area. These metrics are
positively correlated with most water quality parameters in the normal season. This
could mean that large populace concentration in areas near water bodies would
relatively increase pollution levels (Shen et al. 2015). In addition, PD & ED are
positively correlated with nutrients in the wet season, which indicates problems with
runoff and soil erosion (Griffith 2002).
Since it has been reported that forests and natural areas contribute to better water
characteristics (Tong and Chen 2002), it is more probable that point source pollution
in the area is degrading the water quality more than non-point sources. LPI gives an
indication to the largest cover or use inside a certain area (Ding et al. 2016). Forests
are the dominant landscape in this case study area and most of the water quality
variables are negatively correlated with this metric. This result is compatible with the
findings of Lee et al. (2009)), and indicates that extensive forests cover could
enhance water quality to some degree (Shen et al. 2014). Moreover, this reinforces
the previous suggestion that non-point sources had a lesser effect on the variables
than point source pollution. AI &COHE echo the physical and structural linkage and
connectedness of a particular land cover or use inside an area, and consequently,
32
N. R. Jamil and Z. N. Shehab
have demonstrated that landscape metrics are notable factors that illustrate the
association between land uses and water quality and its effect on the hydrological
processes (Shen et al. 2014).
Landscape impact on water characteristics varies over seasons (Pratt and Chang
2012), and is mainly reliant on scale (Zhou et al. 2012). Correlations of landscape
metrics with water quality parameters are shown in Table 2.2. SHMN is considered
as an indicator of the complexity of land use shapes. This index is found to be
positively correlated with DO (p < 0.01) during the normal season and showed a
significant negative correlation with BOD, turbidity, and ammoniacal nitrogen,
which implies that more complexity in land use shapes that may result in a better
water quality (Lee et al. 2009). SHDI reflects the number of land uses in the area and
the proportion of change in land shapes. The presence of numerous different land use
types in a certain area leads to higher SHDI values (McGarigal et al. 2002). This
metric is positively correlated with almost all the variables and negatively with DO,
which signifies that water quality tends to decline with high intersperse of different
land use types as well as high spatial occupancy (Lee et al. 2009). Furthermore, this
consequently alters water purification functions due to the constant changes in the
natural environment (Xia et al. 2012). CONTAG percentage echoes the aggregated
proportion or dispersion of land use patterns that are present in a certain area, which
tends to be high when there is a low level of land use dispersion (Bu et al. 2014).
During both seasons, CONTAG correlated negatively with almost all of the variables (significantly with E.C, turbidity, pH, temperature, DO, and TSS), this would
indicate that water deterioration typically happens in landscapes that have a high
level of dispersion and are greatly fragmented and uneven (Uuemaa et al. 2005).
Likewise, ED and PD metrics are considered as pointers to the unevenness and
fragmentation of a landscape (Fichera et al. 2012). McGarigal and Marks (1995)
observed that PD and ED values would rise, while LPI value would drop when there
are numerous small spots and patches of land covers in the area. These metrics are
positively correlated with most water quality parameters in the normal season. This
could mean that large populace concentration in areas near water bodies would
relatively increase pollution levels (Shen et al. 2015). In addition, PD & ED are
positively correlated with nutrients in the wet season, which indicates problems with
runoff and soil erosion (Griffith 2002).
Since it has been reported that forests and natural areas contribute to better water
characteristics (Tong and Chen 2002), it is more probable that point source pollution
in the area is degrading the water quality more than non-point sources. LPI gives an
indication to the largest cover or use inside a certain area (Ding et al. 2016). Forests
are the dominant landscape in this case study area and most of the water quality
variables are negatively correlated with this metric. This result is compatible with the
findings of Lee et al. (2009)), and indicates that extensive forests cover could
enhance water quality to some degree (Shen et al. 2014). Moreover, this reinforces
the previous suggestion that non-point sources had a lesser effect on the variables
than point source pollution. AI &COHE echo the physical and structural linkage and
connectedness of a particular land cover or use inside an area, and consequently,
32
N. R. Jamil and Z. N. Shehab