3.3 The Response of Stress and Riparian Health Indexing and
Sub-Indexing
Pearson correlation revealed the inter-indicator relationships between
stress indicators and RHIs, allowing us to understand riparian zone
changes within the TGDR under land-use variations. Although PCA provided
information on the key indicators responsible for the total variation in
riparian zones with significant factor groups, Pearson correlation was
performed for all RHIs and stress indicators, and valuable information
was obtained independently for the key stress indicators (7) and RHIs
(14). Pearson correlation relationships are described below.
With regards to the stress indicators, associations were continuously
significant at p < 0.01 (|r | ≤
0.933), and both positive (r ≤ 0.933) and negative (r ≤ ̵̶
0.584) correlations were found between pressure indicators (Table 2).
The highest correlations were generally observed in urban transects,
whereas lower correlations were detected in rural transects
(Supplementary A, Tables A.1–A.5).
The associations between RHIs within different geographical locations
were widely significant at p < 0.01 or p< 0.05 (|r | ≤ 0.989), except for
regeneration. Positive correlations were found in most of the situations
for H (r ≤ 0.624), PC (r ≤ 0.989), R (r ≤ 0.455),
Er (r ≤ 0.963), and Ex (r ≤ 0.935) (Table 2). The
transects from urban areas showed the highest correlation, whereas those
from rural–urban transitional areas showed a lower correlation.
Stronger relationships were formed from PC, Ex, and Er indicators,
whereas a relatively lower correlation was observed from those of H and
R (Supplementary A, Tables A.1–A.5).
During the last phase of the Pearson correlation, the correlations
between pressure and condition indicators for the different geographical
locations were determined. These correlations were significant both atp < 0.01 and p < 0.05 (r =
-0.731 – 0.529) between the indicators of pressure and the indicators
of H, PC, R, Er, and Ex (Table 2). The highest comparative correlation
strength was generally observed in urban transects, whereas the lowest
strength was primarily found in rural–urban transitional transects. The
results showed that pressure indicators correlated with habitat
(r = 0.207 – 0.624), plant cover (r = -0.658 – 0.436),
regeneration (r = -0.693 – 0.377), erosion (r = -0.731 –
0.583), and exotics (r = 0.168 – 0.529) (Supplementary A, Tables
A.1–A.5).
[Table 2 to be inserted about here]
Heat maps were developed separately for stress and riparian health
indices as well as sub-indices to indicate their correlation strength
via their colors (Figure 7). Results showed that indices and sub-indices
exposed unique features for different land-uses. The correlations were
relatively positive in rural areas, where conditions correlated with
plant cover (0.852**), habitat (0.646**), erosion (0.597**), and
regeneration (0.574**). Comparably, correlation strengths were moderate
(both positive and negative) in rural–urban transitional areas. Exotics
correlated negatively with regeneration (-0.483**), plant cover
(-0.469**), erosion (-0.332**), and condition (-0.273*). In contrast,
condition showed a relatively positive correlation with plant cover
(0.781**), erosion (0.587**), and habitat (0.468**). Urban zones
displayed a markedly strong positive (in condition and plant cover) and
negative (in exotic and pressure) correlation among other indices and
sub-indices.
[Figure 7 to be inserted about here]