Data from: Ultradian rhythms of activity in a wild subterranean rodent
Data files
Sep 20, 2024 version files 3.84 MB
-
Finn_activity_rhythm_analysis.R
25.51 KB
-
Highveld_metadata.csv
1.47 KB
-
Highveld_molerat_activity.csv
2.99 MB
-
README.md
3.52 KB
-
soiltemp.csv
819.66 KB
Abstract
Many animals adapt their activity patterns to the best environmental conditions using diel rhythms. African mole-rats (family Bathyergidae) are among the mammals that have become models for studying how these rhythms can be entrained by different cues (light or temperature) in experimental laboratory studies. However, it is unclear whether they exhibit similar circadian rhythms in their natural lightless, subterranean environment. In this study, we used biologging to investigate the activity rhythms of wild, highveld mole-rats. We show that their activity cycle exhibited an ultradian rhythm with a length between 4 - 8 hours. On an individual level, mole-rats displayed about 5 activity bouts per day, occurring at various times during the day and night. On a population level, activity peaked in the afternoon, coinciding with the peak in ambient temperature. Our research suggests that wild subterranean mammals, which experience constant darkness and reduced environmental variation, are unlikely to show clear circadian rhythmicity. Instead, activity periods are distributed over several bouts throughout the day and night and is likely synchronized to daily temperature cycles. We propose that ultradian rhythms may be more common than previously thought and discuss how physiological processes may generate differences in periodicity between laboratory and wild populations.
README
Dataset description for Ultradian rhythms of activity in a wild subterranean rodent
Kyle T. Finn, Otto Brede, Nigel C. Bennett, Markus Zöttl
Data collection methods:
Data was collected using accelerometer devices attached to highveld mole-rats (Cryptomys pretoriae). Accelerometer devices were set to record at 25hz and were deployed for up to 14 days (mean duration 12.95 +/- 2.2 days). Data was downloaded from the accelerometer devices using the program X Manager provided by Technosmart (https://www.technosmart.eu/). Temperature data (soiltemp.csv) was recorded at the study site every 10 minutes with ibuttons buried in the soil at 10cm and 30cm. Temperature (degree C) was also recorded in the sun at 2m height.
Data pre-processing:
As a result of the high frequency and long recording time of the accelerometer data, the raw data files for each individual could exceed 20 million rows with a file size of 1.8GB. To allow for public sharing of data we have processed the raw accelerometer data for each individual. We first calculated overall dynamic body acceleration (ODBA) in 2-second windows. Then we took the mean ODBA for each individual in 10-minute windows. We combined the results of all individuals into one csv file (Highveld_molerat_activity.csv). Example R code used to process the accelerometer data is provided in Finn_activity_rhythm_analysis.r on lines 22-46.
Description of the files:
R script we used for analysis is provided in Finn_activity_rhythm_analysis.r
We used R 4.3.0 for analysis. The R script relies upon the following packages (version in parenthesis): tidyverse (2.0.0), TTR (0.24.4), scales (1.3.0), behavr (0.3.2), ggetho (0.3.7), zeitgebr (0.3.5), glmmTMB (1.1.8), lubridate (1.9.3), hms (1.1.3), sleepr (0.3.0), patchwork (1.2.0), performance (0.10.9), and ggeffects (1.7.0). Data for the analysis is provided in 3 csv files, described below. Columns and data descriptions are provided below each. Dates are in YYYY-MM-DD format, times are HH:MM:SS.
- Highveld_molerat_activity.csv a. AnimalID - character; the individual's unique identification code. The characters before the M/F represents the animal's group. The M or F designates male or female. And the number after sex is the animal's original capture order. b. Timestamp - character; date and time of the 10-minute window for activity c. Time - character; just the time from Timestamp d. t - numeric; the number of seconds which have passed from midnight on the day the device was turned on. Note that the first 24 hours of recording were dropped for all individuals (86400 seconds in a day) e. ODBA - numeric; mean ODBA within a 10 minute window, calculated from ODBA values in 2 second intervals.
- highveld_metadata.csv" a. AnimalID - character; the individual's unique identification code. Same as for Highveld_molerat_activity.csv b. Sex - character; animal's sex c. Group - character; animal's group\ d. CaptureDate - character; date of capture e. Status - character; animal's reproductive status. N = non-reproductive, R = reproductive f. OpenDate - character; date when the accelerometer device was activated g. LiftTime - character; time of release on OpenDate and start of actual data recording.
- soiltemp.csv a. DateTime - character; date and time when the temperature was recorded. b. Temp - numeric; temperature c. Depth - character; depth were temperature was recorded: 10cm, 30cm, Air
Methods
Data was collected using accelerometer devices attached to highveld mole-rats (Cryptomys pretoriae). Accelerometer devices were set to record at 25hz and were deployed for up to 14 days (mean duration 12.95 +/- 2.2 days). Data was downloaded from the accelerometer devices using the program X Manager provided by Technosmart (https://www.technosmart.eu/). As a result of the high frequency, accelerometer data files for each individual were over 1GB in size, and even reducing the data to mean values every 2 seconds results in a 1.7GB file with over 11.5 million rows. To allow public sharing of data we have shared the mean overall dynamic body acceleration (ODBA) values in 10 minute windows. Temperature data (soiltemp.csv) was recorded at the study site every 10 minutes with ibuttons buried in the soil and in full sun at 2m high.