A simple method to quantify fish behavior by forming time-lapse images

In this paper, we propose a simple method for quantifying fish behavior by sampling frames from the video recording of swimming fish and combining them into a time-lapse composite image. Fish speed was measured by using a variant of particle image velocimetry (PIV) experiments method. To demonstrate this method, we perform two tests to determine whether juvenile rockfishes (Sebastes inermis) prefer blue LED (light emitting diode) or white LED. In the first test, the fishes are exposed to both lights simultaneously, and their attraction for light is judged by a statistical analysis of their swimming directions in the time-lapse image. In the second test, the fishes are exposed to each light separately, and their attraction for light is determined by their swimming speed toward each light. In both tests, we found that fishes exhibited a strong attraction for blue light. Analysis of swimming direction for 20 fishes showed that 12 fishes were attracted to the blue LED and 5 fishes for the white LED in the first test. By the second test, average swimming speed toward blue LED was 7.45 cm/s and 3.78 cm/s for white LED. It is thought that the composite image method is readily applicable to many fields.

► Time-lapse images is propose to quantify fish behavior.
► Fish speed was measured by using a variant of particle image velocimetry (PIV) experiments method.
► To demonstrate this method, two tests were preformed to determine whether juvenile rockfishes (Sebastes inermis) prefer blue LED (light emitting diode) or white LED.
► The blue LED attracted the fish more than the white LED.