Three-dimensional vision-driven assessment of lambda-cyhalothrin-induced alterations in zebrafish swimming behavior

Area of Science:

• Environmental Toxicology
• Behavioral Ecology
• Aquatic Toxicology

Background:

• Fish swimming behavior is a sensitive indicator of environmental contaminant impacts.
• Previous studies often lack detailed 3D analysis of group swimming patterns.
• Understanding pesticide effects on fish behavior is crucial for ecological risk assessment.

Purpose of the Study:

• To develop and validate a 3D vision-driven pipeline for analyzing group-level fish swimming behavior.
• To quantify the effects of lambda-cyhalothrin (LCH) on zebrafish swimming behavior and spatial distribution.
• To differentiate the impacts of different lambda-cyhalothrin formulations (CS and EC) on fish behavior.

Main Methods:

• Utilized a dual-view imaging and refraction-corrected 3D reconstruction workflow.
• Employed improved object detection (YOLOv10n) and 3D multi-object tracking (MOTA, IDF1).
• Quantified behavioral features like posture, orientation, movement paths, and volumetric spatial distribution.

Main Results:

• The 3D analysis pipeline achieved high accuracy (mAP@0.5: 99.3%, MOTA: 93.9%, IDF1: 96.8%).
• Exposure to 0.2 μg/L LCH for 9 hours significantly affected zebrafish posture, orientation, and movement paths.
• Different LCH formulations (CS vs. EC) induced distinct changes in fish group spatial distribution and movement patterns.

Conclusions:

• The developed 3D methodology provides robust analysis of fish schooling behavior.
• Lambda-cyhalothrin poses significant risks to fish, altering their swimming behavior and spatial ecology.
• This approach offers new insights and methodological support for ecological risk assessments of aquatic contaminants.