Data collection
My data were collected at the International Coastal Research Center of the Atmosphere and Ocean Research Institute, The University of Tokyo, from June to August 2024.
Five loggerhead sea turtles were used for this study. For each individual, I attached two electrodes (negative, and positive/earth) on the ventral side of the turtle to record the heart rate data. These electrodes were attached to an ECG logger that was glued on the turtle’s carapace, and that also recorded acceleration and temperature.
I had video recordings days from all the individuals in which I had heart rate and acceleration data. From those days, I recorded when the turtle was active and inactive (activity data) and counted the breaths per minute.
Data analysis
Since I did not have video recordings for the 28 hours I was studying, I calculated a threshold for acceleration that would allow me to differentiate the activity status (active vs inactive), based on the activity and acceleration data when video was not available. Also, I used the breaths per minute data (from the video available days) alongside heart rate, acceleration, temperature, and daytime to create an algorithm that would estimate the average breathing frequency.
Essentially, for my 28 hours, I studied:
- Heart rate and acceleration data patterns
- Inferred activity status (active vs inactive) based on videos and acceleration data
- Inferred breathing frequency based on videos, heart rate, acceleration, temperature, and time of day data.