The Influence of the Recording Time in Modelling the Swimming Behaviour of the Freshwater Inbenthic Copepod Bryocamptus pygmaeus

Abstract

The analysis of copepod behaviour gained an increasing impetus over the past decade thanks to the advent of computer-assisted video analysis tools. Since the automated tracking consists in detecting the animal’s position frame by frame and improving signals corrupted by strong background noise, a crucial role is played by the length of the video recording. The aim of this study is to: (i) assess whether the recording time influences the analysis of a suite of movement descriptive parameters; (ii) understand if the recording time influences the outcome of the statistical analyses when hypotheses on the effect of toxicants/chemicals on the freshwater invertebrate behaviour are tested. We investigated trajectory parameters commonly used in behavioural studies—swimming speed, percentage of activity and trajectory convex hull—derived from the trajectories described by the inbenthic–interstitial freshwater copepod Bryocamptus pygmaeus exposed to a sub-lethal concentration of diclofenac. The analyses presented in this work indicate that the recording time did not influence the outcome of the results for the swimming speed and the percentage of activity. For the trajectory convex hull area, our results showed that a recording session lasting at least 3 min provided robust results. However, further investigations are needed to disentangle the role of concurrent factors, such as the behavioural analysis of multiple individuals simultaneously, whether they are of the same or opposite sex and the implications on sexual behaviour, competition for resources and predation.

Quantitative comparison of 2D and 3D monitoring dimensions in fish behavior analysis

“…The data recording was conducted over a period of 10 days, with two sections per day (8:00–10:00 a.m. and p.m.). Four replicates of monitoring experiments were conducted, each consisting of a group of four fish and lasting for 5 min (Audira et al, 2019; Di Cicco et al, 2022; Torisawa et al, 2011). As a result, there were a total of 320 trajectories (2 sessions 4 replicates 4 fish 10 days = 320 trajectories) obtained.…”