Skip to main content
Dryad

Nesting chinstrap penguins accrue large quantities of sleep through seconds-long microsleeps

Cite this dataset

Libourel, Paul-Antoine; Lee, Won Young (2023). Nesting chinstrap penguins accrue large quantities of sleep through seconds-long microsleeps [Dataset]. Dryad. https://doi.org/10.5061/dryad.000000087

Abstract

Microsleeps, the seconds-long interruptions of wakefulness by eye closure and sleep-related brain activity, are dangerous when driving and might be too short to provide the restorative functions of sleep. If microsleeps fulfill sleep functions, then animals faced with a continuous need for vigilance might resort to this sleep strategy. We investigated electroencephalographically-defined sleep in wild chinstrap penguins at sea and while nesting in Antarctica constantly exposed to an egg predator and aggression from other penguins. The penguins nodded off >10,000 times per day, engaging in bouts of bihemispheric and unihemispheric slow-wave sleep lasting on average only 4 s, but resulting in the accumulation of over 11 h of sleep for each hemisphere. The investment in microsleeps by successfully breeding penguins suggests that the benefits of sleep can accrue incrementally. This repository contains raw electrophysiological data (Electroencephalography and electromygraphy) with identified accelerometry, ambiant temperature and depth. We also add the different ethograms and Slow waves sleep episod detected for each cerebral hemisphere.

README

Nesting chinstrap penguins accrue large quantities of sleep via seconds-long microsleeps

This dataset containts all the electrophysiological data (4 EEG, 1 EMG) with accelerometry, video, behavioral scoring, GPS, depth, weather data and any metadata related to the experiment.

The dataset was collected in December 2019 on Barton Peninsula (62°14.30 S, 58°46.50 W), King George Island, Antarctica on Chinstrap penguins (Pygoscelis antarctica).

We investigated electroencephalographically-defined sleep in wild chinstrap penguins at sea and while nesting in Antarctica constantly exposed to an egg predator and aggression from other penguins.

The data show that the penguins nodded off >10,000 times per day, engaging in bouts of bihemispheric and unihemispheric slow-wave sleep lasting on average only 4 s, but resulting in the accumulation of over 11 h of sleep for each hemisphere.

The investment in microsleeps by successfully breeding penguins, suggests that the benefits of sleep can accrue incrementally.

This repository contains the data and code for our paper:

Libourel P.-A.; Lee W. Y.; Achin I.; Chung H.; Kim J.; Massot B.; Rattenborg N.C., Nesting chinstrap penguins accrue large quantities of sleep throught seconds-long microsleeps.

Methods description and detailled results are available in the article supplementary materials.

To run the script the filename file and folder organisation should be keep as it.

Description of the data and file structure

Additional Information (folder)

nest_distance (foder)

the csv table with GPS data for each individual (W#) are stored in the W#_gps.csv

Column descriptions:

Date_Time - Date & time (format: 'YYYY-MM-DD HH:MM:SS') UTC for penguin GPS location

Decimal_Longitude - Decimal degree for longitude of penguin GPS point

Decimal_Latitude -Decimal degree for latitude of penguin GPS point

POINT_X - X coordinate value for penguin GPS point in Antarctic Polar Stereographic projection EPSG:3031

POINT_Y - X coordinate value for penguin GPS point in Antarctic Polar Stereographic projection EPSG:3031

distance_from_nest - Shortest distance from nest to penguin GPS point inferred from Haversine formula

Script to use : ImportLocationTimesXls.m

individual information chinstrap sleep project All Data.xlsx

this file contains all individual information collected for experimental and control birds

Empty cell mean that no data are available for that particular condition (animal not recaptured, not weigthed, information not collected ...)

Column descriptions:

id - Animal number W# for xperimental birds and C# for control birds

capture time - date and Time in local time (UTC-3), format dd/mm/yyy hh:mm:ss for a single event, yyyy/mm/dd hh:mm:ss AM/PM - hh:mm:ss AM/PM for long event (start - end time)

sex - sex

weight at capture (kg) - weight at capture in kg

release time - release date and time format: dd/mm/yyyy hh:mm:ss (UTC-3)

recapture time - recapture date and time format: dd/mm/yyyy hh:mm:ss (UTC-3)

weight at recapture (kg) - weight at recapture in kg

recapture site - recapture site , could be : coast / nest during incubation / standing next to the nest

distance from the colony boundary (m) - nest distance from the colony boundary (m)

bill length (mm) - bill length (mm)

bill depth (mm) - bill depth (mm)

RFID number - RFID number

breeding success (no. of chicks) - breeding success (nb. of chicks)

Nest Hill position - Nest position related to the hill. Could be: Shore/ Up to the hill / down to the hill

EEG Reference - EEG reference position and whether or not the screw was going through the skull, on the cranial suture or not. The cell precide the localisation of the reference electrodes

EMG - EMG type, could wires cut at the tip or wires with 1mm of the wire exposed. On one animal there is a doubt if the electrodes are in the muscles (writen muscle ??)

EEG Signal Quality - EEG Signal Quality based on visual observations

EMG Signal Quality - EMG Signal Quality based on visual observations

sleep logger recording start date - Sleep Logger starting date and time in local time (UTC -3)

Theorical recording end date - Sleep Logger theorical ending date and time in local time (UTC -3)

recording end date - Real end recording date and time in local time (UTC -3)

Nb Bin Files - Number of 24 hour files recorded by the sleep logger

EEG Signal Quality - Global visual appreciation of the EEG signal quality

EMG Signal Quality - Global visual appreciation of the EMG signal quality

Script to use : ImportCaptureAndReleaseTimeXls.m

Temperature depth recording Offset and drift correction Chinstrap_tdr_offset_ephy.xlsx

this file contains time events from referenced from the electrophysiological data and the corresponding event (offset in sec) from the tdr file

this is used to evaluate the drift and offset between electrophysiological data and temperature depth recording recording (TDR).

Each line is an event found in both files type.

Column descriptions

Animal - Animal number W#

Time (min) - Time elasped in min from the first sample recorded corresponding to the event

offset (s) - offset in sec on the TDR file for the same event

Script to use: ImportTDRCsv.m

Behavioral and video information model chinstrap location.xlsx :

This file contains date and time for direct behavioral observations and video recordings

Empty cell mean that no data are available for that particular condition (no video for that particular time)

Column descriptions.

Animal ID - Animal number W#

date - date and Time in local time (UTC-3), format dd/mm/yyyy hh:mm:ss for a single event, yyyy/mm/dd hh:mm:ss AM/PM - hh:mm:ss AM/PM for long event (start - end time)

State - direct observation : 0 = not seen at the nest, 1 = incubating, 2 = standing near the nest, 3 = observed near the shore, 4 = start to dive into the sea, 5 = arrive at the seashore

Video - 1 if a video is available at that time

Script to use: ImportLocationTimesXls.m

Weather information

The wind speed and ambiant temperature in the colony during the experiment could be obtain here: https://kpdc.kopri.re.kr/live-data/weather

Raw Data

all individual raw data are stored in a folder named with the animal ID W#, which contains one folder tdr containing temperatiure depth recording data and on data folder with video and electrophysiology data

  • ### temperature depth recording (tdr)

this folder contains a CSV file tabuated file with the temperature (°C), and depth (m), recorded from a logger fixed at the leg of the birds.

W#_TD.csv

Empty cell mean that no data are available for that particular condition (Temperatures are often missing because the sampling rates are different between temperature and depth)

Column descriptions.

Date_Time - date and Time in (local time UTC -3), format yyyy-mm-dd hh:mm:ss

Temperature - ambiant temperatue at the leg level in °C recorded every minutes

Depth - Depth recorded in meters from the surface recorded every second

Depth_Filt - Depth recording filtered in meters

Script to use: ImportTDRCsv.m

data

this folder contains:

  • Header files ('.exp' files which are xml coded: W#.exp & W5_Activity-Inactivity.Exp)

  • raw electrophysiological data (bin folder)

  • land/sea scoring (folder Hypno_W#)

  • behavioral activity scoring (folder Hypno_W#_Activity-Inactivity)

  • Video (video folder)

  • Sleep events (W#_SWS_Events 2-12.mat)

Header files

the two exp files W#.exp & W#_Activity-Inactivity.expare xml files.

they contain the following information

fields descriptions

Animal/Name Animal ID W#

Animal/Acquisition/SamplingRate : the acquisition sampling rate in Hz for all the bin files

Animal/Acquisition/NbChan : number of channel in the bin files

Animal/Acquisition/AcquisitionType : type of encoding

Animal/Acquisition/HeaderOffset : number of bytes before in the header before the data (ususally 0)

Animal/Acquisition/Files/File/FileName : list of the bin files related to the each indivual recording

Animal/Acquisition/Files/File/TStart : absolute time start format: yyyy-mm-ddThh:mm:ss:fff (fff are the 3 digit of the milliseconds)

Animal/Acquisition/Files/File/Duration : file duration in seconds

Animal/Acquisition/Channels/Channel/Id : channel number

Animal/Acquisition/Channels/Channel/Name : channel name, AL: EEG Hyperpallium anterior Left; AR: EEG Hyperpallium anterior right: PL: EEG Hyperpallium posterior left; PR: EEG Hyperpallium posterior right

Animal/Acquisition/Channels/Channel/Offset : channel offset in the unit of the channel

Animal/Acquisition/Channels/Channel/Gain : channel Gain

Animal/Acquisition/Channels/Channel/AcquisitionRangeMax : channel range in the unit of the channel

Animal/Videos/Video/Files/File/FileName : list of the video files related to the each indivual recording

Animal/Videos/Video/Files/File/TStart : absolute time start format: yyyy-mm-ddThh:mm:ss:fff (fff are the 3 digit of the milliseconds)

Animal/Videos/Video/Files/File/Duration : video file duration in seconds

Animal/Videos/Video/Files/File/SamplingRate : video sampling rate (frames per second)

Animal/Hypnogram/Files/File/FileName : file name of the .H file which contains scorig states, the file a coded in uint16 sampled at 1Hz

Animal/Hypnogram/Files/File/TStart :absolute time start format: yyyy-mm-ddThh:mm:ss:fff (fff are the 3 digit of the milliseconds)

Animal/Hypnogram/Files/File/Duration : file duration in seconds

Animal/Hypnogram/SamplingRate duration : of each epoch scored, this is not realted to the sampling in the file which is always 1 Hz

Animal/Hypnogram/Statuses : States definition of .H files

Animal/Hypnogram/Statuses/Status/Level : value coded in the H file

Animal/Hypnogram/Statuses/Status/Label : Label of the corresponding state

Animal/Hypnogram/Statuses/Status/Key : Key used to score

Animal/Hypnogram/Statuses/Status/Color : Color used for the corresponding state, coded in decimale see function dec2ARGB.m

Script to use: loadEXP.m

this script can open the '.exp' file into matlab

bin

Bin File description

This folder contain all the binary files ('.bin') corresponding to each individual recording

They contain the EEG, EMG, and accelerometry data

the files are named with their starting date time coded as follow: PREFIX_yyyy-mm-dd_hh-mm-ss-fff.bin

the binary file are interlaced 16 bit files, they could be uint16 or int16 see corresponding exp file. the number of channels is also indicated in th eexp file.

Each sample is coded on 16bits and the absolute value could be obtain with the gain, offset and range specified in the exp file the function ADC2Real.m script could be used to convert them.

Script to use: ReadChanBinFile.m / ADC2Real.m

Hypno_W#

Folder description

This folder contain one .H file per electrophysiological binary file.

We coded whether the animal is on land or at sea based. these files where used as a reference to later score automatically the activity and inactivity state

they are nammed with their respective absolute starting time

the H file are uint16 coded with one channel sampled at 1Hz. the H file contained three state code

0 for undeterminded, 15 when the animal is on land, 75 when the animal is at sea.

Hypno_W#_Activity-Inactivity

H file description

This folder contain one .H file per electrophysiological binary file.

they are nammed with their respective absolute starting time

the H file are uint16 coded with one channel sampled at 1Hz.

These files contain the animal location and behavior. This ethogram was obtained from accelerometry data and depth recording.

They contain the following states:

0 for undeterminded, 10 On land Active Lying Down 15 On land Inactive Lying Down 40 On land Active Standing 45 On land Inactive Standing

70 At Sea Active at the surface 71 At sea Active UnderWater 75 At sea Inactive at the surface 76 At sea Inactive UnderWater

video

folder description

This folder contain all the Videos related to an individual recording they are encoded in MPEG4 with the avi container

Sleep events

the matlab file W#_SWS_Events 2-12.mat contain the sleep event detected from the clustering algorythm presented in our related article.

these event were detected from the instant power inthe 2-12 Hz spectral band, in the left and right hemisphere

File description

BandLimDen - 1x2 double matrix containing the limit of the denominator used for the sleep event detection could be empty if no deniminator was used

BandLimNum - 1x2 double matrix containing the limit of the numerator used for the sleep event detection usually [2 12], which correspond to the 2-12Hz frequency band used for sleep detection in our article.

Data - Sleep event

Settings.Chan 1x2 double matrix containing the channels number (from the exp file) used for the left and right hemispheres is the value, there is two value if a differential measurement is used (this is not the case in this dataset)

Event.Deb 1xn double matrix with the begining of all events coded in seconds since the first sample of the first binary file listed in the exp file recorded.

Event.End 1xn double matrix with the ending of all events coded in seconds since the first sample of the first binary file listed in the exp file recorded.

Event.Type 1xn double matrix containing the hemisphere in which the event was detected 0 left hemisphere, 1 right hemisphere

Sharing Information

This is a section for linking to other ways to access the data, and for linking to sources the data is derived from, if any.

Links to other publicly accessible locations of the data:

Code/Software

The analysis pipeline was writen in matlab R2020 (associated with signal processing and image processing toolbox and chronux, http://chronux.org/)). Our scripts are provided with our exptoolbox build to read and process easily our data.

Methods

See the associated code repository and manuscript for additional information on the methods for data collection and processing.

Usage notes

The electrophysiology data could be opened on any scientific software that could opened interlaced 16bits binary data.

The video could be run with any videoplayer.

the CSV and xls file could be opened with adapted software

our .exp file could be opened with a text editor like notepad++

The sleep event are provided in a mat file which could be opened in matlab (Mathworks) 

The analysis pipeline was writen in matlab R2020 (associated with signal processing and image processing toolbox and chronux, http://chronux.org/). Our scripts are provided with our exptoolbox build to read and process easily our data.