Automatic Image Processing to Determine the Community Size Structure of Riverine Macroinvertebrates

Area of Science:

• Ecology
• Environmental Science
• Bioassessment

Background:

• Body size is a key functional trait used as a bioindicator for ecological perturbations.
• Community size structure reflects environmental changes but manual measurement of small organisms is time-consuming.
• Benthic macroinvertebrates are crucial bioindicators for freshwater ecosystem health.

Purpose of the Study:

• To develop and validate a protocol for semi-automatically measuring the body size of preserved river macroinvertebrates.
• To expedite the estimation of community size structure for rapid bioassessment.
• To adapt existing scanning methodologies for freshwater macroinvertebrate analysis.

Main Methods:

• A protocol was developed based on scanning subsamples of macroinvertebrates and processing digitized images.
• Artificial intelligence was used to create a learning set for distinguishing macroinvertebrates from detritus and artifacts.
• The protocol involves scanning, image processing, AI-driven classification, and size structure depiction.

Main Results:

• The study presents a validated protocol for semi-automatic body size measurement of river macroinvertebrates.
• The developed method successfully individualizes and measures macroinvertebrates from scanned images.
• Calibration results and recommendations for further improvements are provided.

Conclusions:

• The scanning system is effective for automatic body size measurement of river macroinvertebrates.
• Depicting the size spectrum of macroinvertebrate communities is a valuable tool for rapid bioassessment.
• This protocol offers a faster alternative to manual measurements for assessing freshwater ecosystems.