Quantifying fish posture and fine-scale movement in flowing water: an enriching approach for ecohydraulic experimentation

The study of fish movement behavior has been greatly advanced by the development of automated visual tracking, which can provide quantitative data on movement and body posture. Software tools enable researchers to explore various ecological questions but fundamentally rely on achieving good visual contrast between the moving organism and its background. Many laboratory studies expose fish to highly artificial environments with rectangular channels, flat concrete floors, and glass walls. However, imaging fish remains technically challenging, and the absence of standardized guidelines often makes automated tracking a time-consuming process of trial and error. Clear recommendations are lacking that could help reduce imaging difficulties and improve tracking efficiency. Here, we address challenges related to tracking fish through the air-water interface while replicating natural light and flow conditions to enrich ecohydraulics studies. We demonstrate how bottom-up infrared illumination allows for high-contrast imaging of fish position and posture, and how structural features such as depressions and cobbles can mimic the complexity of natural riverbeds. Finally, we discuss challenges posed by surface waves and light refraction and provide a framework for designing flume experiments that improve tracking accuracy and thereby enhance the study of fish movement under ecologically relevant flow conditions.