2.3. Statistical Methods
According to situations and objectives, multivariate statistical
techniques are used extensively in ecological and conservation studies
to explain the environmental and land-use changes in riparian zones, for
example, the algorithms found in Bombino et al. (2019), McIntosh (2015),
and Zema et al. (2018). In this study, we used techniques including the
Kruskal-Wallis non-parametric alternative to the analysis of variance
(ANOVA), principal component analysis (PCA),
Pearson correlation, and cluster
analysis (CA). The Kruskal-Wallis test is a standard procedure used in
scientific and non-scientific disciplines to analyze differences between
means (McIntosh, 2015). The level of significance was established at
both p < 0.01 and p < 0.05. All the
transects were considered spatially independent in terms of group and
indicator indexing for the rural, rural–urban transitional, and urban
areas. PCA (factor analysis) is used in ecological and conservation
studies to extract key elements, factors, and indicators from
multidimensional data. This process is effective in clustering the
indicators by creating diverse groups. Pearson correlation is a
statistical metric used to determine linear relationships and measures
the strength and direction between two random variables or indicators
(Zhou et al., 2016). This method is used in data classification and
analysis for various indices in scientific and non-scientific research
(Pavanello et al., 2015; Zhou et al., 2016). CA is used to group similar
or dissimilar characteristics among diverse random variables to
establish parallel functionality (Bombino et al., 2019; Zema et al.,
2018). In our study, the agglomerative hierarchical cluster (AHC) method
was used to determine CA. Origin release 2021 (Northampton, MA, USA) was
used for statistical analyses and graphing.