Modelling Environmental Driving Factors of Macroinvertebrate Distributions in Mountainous Rivers Using Partial Least Squares Structural Equation Modelling: A Case Study from Lai Chi Wo, Hong Kong

The interaction between macroinvertebrates and environmental factors has long been a question of great interest across a wide range of fields. However, our understanding of the impacts of environmental factors on macroinvertebrates remains vague. Moreover, the study of the distribution drivers of macroinvertebrates in mountainous rivers has been rarely undertaken, primarily due to the sampling difficulties and the inherent complexity of the terrestrial ecosystems in these regions. This study applied a partial least squares structural equation modelling (PLS-SEM) to investigate the interactions between specific environmental factors, namely, hydrometeorology, water temperature, nutrient levels and substrate composition, and their effects on macroinvertebrates. Five stations along the Lai Chi Wo River in the Northwest New Territories of Hong Kong, China, were selected for data monitoring. The model results reveal site-specific environmental influences on macroinvertebrates. The furthest downstream station, close to the sea, is vulnerable to seawater intrusion during the dry season, indicating a hydrometeorological-driven model as freshwater macroinvertebrates are intolerant to seawater. Two stations, one downstream and the other upstream, are occupied by rocks that stir river water and increase dissolved oxygen, forming a nutrient-driven model. The remaining two stations, located at small ponds next to man-made weirs, experienced accumulated sedimentation and debris, which are conducive to macroinvertebrates, resulting in a substrate-driven model. Furthermore, this study demonstrated that PLS-SEM provides a more precise understanding of the direct and indirect effects of environmental factors on macroinvertebrates compared to linear fitting and principal component analysis methods.