The system will be down for regular maintenance from 8:00-10:00am PDT on April 3rd, 2024.
Analysis of fast-start movements using accelerometry and video tracking in the Great Sculpin (Myoxocephalus polyacanthocephalus)
View/ Open
Date
2011-09Author
Broell, Franziska
Noda, Takuji
Wright, Serena
Metadata
Show full item recordAbstract
While the use of accelerometers in the aquatic environment becomes an increasingly used tool in remotely observing animals; however, the data obtained from deploying accelerometers still needs better understanding. Observations gathered by studies using accelerometers are largely limited to the identification of simple behaviours such as resting and swimming, yet fine-scale movements such as feeding and escape responses are mostly undetected. In this experiment, we aim to establish a link between acceleration traces and fast-start movements in the Great Sculpin (M. polyacanthocephalus) by the analysis of acceleration data from accelerometers and a high-speed video camera. Feeding events, escape events and spontaneous movements were triggered and observed using a 100Hz recording accelerometer (Little Leonardo Ltd, Japan) and a high-speed video camera for n = 7 great sculpin. Kinematic comparison between acceleration obtained from accelerometers and high-speed video camera were performed using vector transformation, yet prove to be difficult due to differences in reference frames and different sources of error. To establish a link between behaviour and acceleration, statistical analysis shows that the signature of spontaneous events can be described by the variation of the magnitude of acceleration which is significantly lower in spontaneous events compared to fast-start movements. Most of this information is lost (50%) if the accelerometer sampling rate is lower than 30Hz. Furthermore, two parameters (the value of Amax the variation of acceleration in lateral and forward direction) allow us to differentiate between escape events and feeding events. These results are a valuable contribution to understanding acceleration data in the field and the issues associated with low sampling rates.