RUB » Bibliotheksportal

Sina M. Langer, Martin Horstmann, Linda C. Weiss, Winfried Junke, Stefan Herlitze, Ralph Tollrian

• Analyzing movement is essential for understanding complex behavioral interactions. Up to date, there are different movement analysis methods. Previous studies working with 2D movement analysis systems provided first insights into this field. 2D systems can capture only two of three spatial dimensions and thus allow analyses of movements on surfaces. 3D systems include all three spatial dimensions. Thus, 3D movement analysis is essential for analyzing movements in air or in water. Especially during the last years, 3D movement analysis has progressed a lot thanks to technological advances. Unfortunately, these technological advanced systems are often very expensive and handling flexibility is limited. Therefore, we implemented a customizable 3D movement analysis system for aquatic organisms. The system is composed of an experimental arena, backgrounds with different contrasts, three visible light diode spots, one infrared diode spot for analysis in the dark, and two Raspberry Pi 3 Model B V 1.2 equipped with cameras without infrared filters. We assembled a user-friendly system that is controlled directly via a graphical user interface (GUI). It not only delivers raw data but also runs detailed analyses of common behavioral parameters. The accuracy of those analyses can be enhanced by the integrated manual error-correction method that allows to interrupt the automatic track analysis and to correct detection errors by hand. In our case study, we show the spectrum of opportunities of the system by determining the swimming behavior of the freshwater crustacean Daphnia magna. We successfully recorded its swimming behavior under visible light and infrared light (simulating darkness for the daphniids) conditions and analyzed swimming velocity, motion patterns, depth selection, and tendency of clustering. In our case study, we show that using the integrated manual error-correction method leads to a clear accuracy improvement. Our movement analysis system is not restricted to aquaria but can be adopted to movements in cages and even open areas of different sizes both in water or air, which emphasizes its potential. Together with a user manual, the open-source MATLAB algorithm and the recommended hardware components, the here presented system enables semi-automatic 3D movement analysis, which can be adapted to the individual needs.