Physical habitat modeling for river macroinvertebrate communities

Area of Science:

• River ecology
• Benthic macroinvertebrate community assessment
• Habitat modeling

Background:

• Current river habitat models often overlook macroinvertebrate communities, focusing instead on individual species.
• This limits understanding of community-level ecological needs and overall river functionality.

Purpose of the Study:

• To develop and validate a novel approach for modeling river macroinvertebrate community habitat.
• To extend existing habitat assessment methodologies to a broader ecological target.

Main Methods:

• Utilized the Flow-T index combined with Random Forest (RF) regression for mesohabitat scale modeling.
• Calibrated and validated the model using field data and 2D hydrodynamic simulations across three Italian rivers.
• Identified key mesohabitat descriptors influencing macroinvertebrate communities.

Main Results:

• The RF model identified 12 critical mesohabitat descriptors (water depth, flow velocity, substrate, connectivity).
• Achieved good predictive accuracy with cross-validation R² of 0.71 and test R² of 0.63.
• Demonstrated reliable agreement between simulated and experimental data.

Conclusions:

• The developed model accurately assesses macroinvertebrate habitat and predicts responses to flow and morphological changes.
• Successfully extended the MesoHABSIM methodology for fish habitat assessment to macroinvertebrate communities.
• Offers valuable applications for ecological flow design in diverse riverine systems.