RESEARCH ARTICLE
Fi s -nigh e ec on sleep ime in dai y cows
Emma Te nman
1
*, Ma i Pas ell
2
, Lau a Ha
¨nninen
3
, Sig id Agena
¨s
4
, Pe P. Nielsen
5
1Depa men o Animal Science, Aa hus Uni e si y, Tjele, Denma k, 2Na u al Resou ces Ins i u e Finland
(Luke), G een Technology, Helsinki, Finland, 3Depa men o P oduc ion Animal Medicine, Uni e si y o
Helsinki, Finland and Resea ch Cen e o Animal Wel a e, Helsinki, Finland, 4Depa men o Animal
Nu i ion and Managemen , Swedish Uni e si y o Ag icul u al Sciences, Uppsala, Sweden, 5Depa men o
Ve e ina y and Animal Sciences, Uni e si y o Copenhagen, F ede iksbe g C, Denma k
*emma. e [email protected]
Abs ac
In human sleep s udies, he p obabili y o discom o om he elec odes and he change in
en i onmen usually esul s in i s -nigh eco dings being disca ded. Sleep eco dings om
he i s nigh in human subjec s o en di e in amoun o REM ( apid eye mo emen ) sleep
and he o e all sleep a chi ec u e. This s udy in es iga ed whe he eco dings o sleep
s a es in dai y cows also show a i s -nigh e ec . Non-in asi e elec ophysiological eco d-
ings we e ca ied ou on nine cows o he Swedish Red b eed du ing h ee consecu i e 24-
hou pe iods ( eco ding days 1–3). O e all, cows spen 12.9 ±1.4 hou s awake, 8.2 ±1
hou s umina ing, 57.2 ±20.3 min d owsing, 44.1 ±20.2 min in REM sleep and 64.3 ±38.1
min in NREM (non- apid eye mo emen ) sleep (mean ±SD) and he e we e no signi ican
di e ences be ween eco ding days in o al du a ion o any o he sleep and awake s a es.
Howe e , he bou s o REM sleep and umina ion we e longe , and he awake bou s we e
sho e , a nigh ime compa ed o day ime, ega dless o eco ding day. The awake bou s
also showed an in e ac ion e ec wi h longe bou s a day ime du ing day 1 compa ed o
day ime on day 3. Da a on sleep and awake s a es eco ded in adul dai y cows du ing h ee
consecu i e 24-h pe iods showed g ea a ia ion in sleep ime be ween cows, bu o al ime
o each s a e was no signi ican ly a ec ed by eco ding day. Fu he and mo e de ailed
s udies o how sleep a chi ec u e is a ec ed by eco ding day is necessa y o ully comp e-
hend he i s -nigh e ec in dai y cows.
In oduc ion
Mode n esea ch on sleep elies on non-in asi e eco dings o he elec ophysiology o he
b ain, commonly using skin adhesi e elec odes. A e ac s du ing he i s nigh o elec o-
physiological eco dings ha e been desc ibed in s udies on humans and occu du ing sleep
eco dings in bo h he home en i onmen and sleep labo a o ies [1,2,3]. Ea ly wo k by
Rech scha en and Ve done [1] showed ha o al sleep ime did no di e be ween eco ding
nigh s, bu apid eye mo emen (REM) sleep dec eased and numbe o awakenings inc eased
in human subjec s du ing he i s nigh compa ed wi h he ou h nigh o eco ding in a
sleep labo a o y. These indings we e con i med by Le Bon [3], who obse ed no di e ence in
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0195593 Ap il 13, 2018 1 / 10
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN ACCESS
Ci a ion: Te nman E, Pas ell M, Ha¨nninen L,
Agena¨s S, Nielsen PP (2018) Fi s -nigh e ec on
sleep ime in dai y cows. PLoS ONE 13(4):
e0195593. h ps://doi.o g/10.1371/jou nal.
pone.0195593
Edi o : Juan J Loo , Uni e si y o Illinois, UNITED
STATES
Recei ed: No embe 20, 2017
Accep ed: Ma ch 25, 2018
Published: Ap il 13, 2018
Copy igh : ©2018 Te nman e al. This is an open
access a icle dis ibu ed unde he e ms o he
C ea i e Commons A ibu ion License, which
pe mi s un es ic ed use, dis ibu ion, and
ep oduc ion in any medium, p o ided he o iginal
au ho and sou ce a e c edi ed.
Da a A ailabili y S a emen : All ele an da a a e
wi hin he pape and i s Suppo ing In o ma ion
iles.
Funding: This s udy was unded by he Swedish
Resea ch Council Fo mas and was gi en g an
numbe 2011-1094, h p://p oj. o mas.se/de ail.
asp?a endeid=25678&x=250&y=20&sp ak=
1& edo isning=0. The unde s had no ole in s udy
design, da a collec ion and analysis, decision o
publish, o p epa a ion o he manusc ip .
Compe ing in e es s: The au ho s ha e decla ed
ha no compe ing in e es s exis .
o al sleep ime, bu epo ed ha REM sleep did no e u n o a s eady s a e un il he ou h
nigh o eco ding. These dis u bances a e known as he i s -nigh e ec , and a e p obably
due o discom o om he elec odes and he change o en i onmen . As a consequence, i s -
nigh eco dings a e o en disca ded in human sleep s udies [4,5]. When es ima ing a e age
o al lying ime in dai y cows on he d le el, accu acy changed li le be ween he 3- and 5-day
sampling pe iod [6].
Changes in en i onmen ha e also been shown o a ec beha iou and sleep dis ibu ion in
cows. Fo example, mo ing cows be ween g oups a ec lying ime and eeding beha iou on
he day o eg ouping bu no subsequen days [7]. Sleep eco dings ha e been ca ied ou in
adul dai y cows du ing 24-hou pe iods using pe manen ly implan ed b ain elec odes [8].
The cows in Ruckebusch’s [8] s udy showed non- apid eye mo emen (NREM) sleep o
3.25h, REM sleep o 45 min and hey engaged in d owsing, desc ibed as an in e media e s a e
wi h a educed senso y h eshold bu no as low as o sleep, o 7.5h pe 24h. Mo eo e hey
spen 12.5h awake, and i migh be possible o some animals o anced in o NREM sleep du -
ing umina ion [8]. Acco ding o ha s udy and Ruckebusch [9], cows sleep bo h du ing he
day and a nigh when on pas u e, bu only a nigh when indoo s in a ie-s all en i onmen .
Ruckebusch [9] concluded ha cows b ough in om pas u e o a ie-s all en i onmen need
3-6 days o es ablish hei indoo sleep pa e n. Howe e , i hey we e s anchioned in he ie-
s all using a headlock ga e, 14-21 days o adap a ion we e necessa y o es ablish hei indoo
sleep pa e n when b ough in om pas u e. In hose ea ly s udies, i ook longe o he cows
o adap o he eco ding equipmen because o he mo e in asi e me hod used o collec ing
elec ophysiological da a. This echnique equi ed he animals o be e he ed o e en s an-
chioned du ing eco ding. Thus, he lying beha iou and sleep pa e n o he cows migh ha e
been dis u bed. Whe he he eco ding equipmen and moni o ing o he cows du ing he
eco ding i sel in e e ed wi h he sleep o he cows was no s udied.
A non-in asi e me hod o eco ding sleep, allowing he animals o mo e eely in a pen
du ing eco dings, has ecen ly been de eloped o cal es and adul cows [10,11].
The aim o he p esen s udy was o in es iga e whe he he non-in asi e eco ding o sleep
and awake s a es in dai y cows [11] cause “ i s -nigh e ec s” simila o wha has been epo ed
om sleep esea ch in humans.
Ma e ial and me hods
Sleep eco dings we e conduc ed on nine cows o he Swedish Red b eed du ing h ee conse-
cu i e 24-h pe iods, based on indings o a ia ion in lying ime by on Keyse lingk e al. [6],
a he Swedish Uni e si y o Ag icul u al Sciences, Uppsala, Sweden. Da a eco ding was suc-
cess ul o e hese h ee days, bu was in e up ed o abou 20 min each mo ning when da a
had o be ans e ed om he eco ding uni o he compu e . The expe imen al design and
handling o he animals du ing he expe imen was app o ed by he Uppsala Local E hics
Commi ee (C313/10, 2010 Uppsala, Sweden). The eco dings we e conduc ed du ing win e ,
when he empe a u e in he expe imen al acili y was 11 ±2˚C.
Animals and housing
The expe imen was designed o elimina e ac o s ha could dis u b he cows. Two days be o e
he eco ding pe iod, he cows we e mo ed om a loose housing sys em o indi idual pens (3
m x 3 m) in a quie oom wi h no o he cows excep hose pa icipa ing in he s udy. The pens
we e equipped wi h headlocks o e he he cows a milking, a eeding ough o concen a e
and silage, and a p essu e al e wa e bowl. The cows had ad libi um access o wa e and silage
du ing he eco ding pe iod. Silage e usals we e emo ed and esh silage was gi en be o e
Fi s -nigh e ec in dai y cows
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0195593 Ap il 13, 2018 2 / 10
mo ning milking, a lunch ime and be o e a e noon milking. Concen a e was p o ided o
mee he indi idual equi emen s o each cow acco ding o hei milk p oduc ion [12]. Bed-
ding consis ed o wood sha ings and was eplaced daily. Ligh ing was all a i icial and ligh
hou s we e 05.00-20.00 h, while a nigh a dim nigh -ligh was p o ided (app oxima ely 30
lux). The cows we e milked wice daily, a 06:00 h and 17:00 h. A e age days in milk was
203 ±20 and milk yield was 27 ±4 li es pe day. One cow was p imipa ous and he o he s
we e mul ipa ous, in lac a ion numbe 2-4.
Elec ophysiological eco dings
In he indi idual pens he cows we e i ed wi h an udde holde and a ex ile hal e . A mini-
mum o 10 h be o e he eco ding s a ed, hey we e sha ed a he elec ode a achmen si es o
ensu e su icien conduc ion.
Elec odes (ø3 cm) (Unilec , Unomedical L d, Bi ke ød, Denma k) we e secu ed o he
skin using issue adhesi e (Ve Bond™, 3M, S . Paul, MN, USA) and posi ioned as desc ibed by
Te nman e al. [11] on he mo ning o eco ding day 1 (Fig 1). Fou EEG elec odes we e
placed e ically wo by wo in lines d awn om he middle o he eye bulb up o he c anial
pa o he ho n base in a squa e on he la a ea o he cow’s o ehead. The e e ence elec ode
was placed in he middle o he squa e and he g ound elec ode was placed caudally o he
ho n base. The elec odes we e some imes ubbed o agains he in e io by he cow, and was
hen apidly eplaced by new elec odes. Elec odes o measu ing eye mo emen s we e posi-
ioned abo e and below one eye, and wo elec odes o measu ing neck muscle ac i i y we e
ixed symme ically on each side o he neck on ce ical pa o he apezius muscle. All elec-
odes we e connec ed o a mobile eco ding de ice (Embla Ti anium, Embla Sys ems Inc.,
B oom ield, CO, USA) using snap-on cables (Embla Sys ems Inc., B oom ield, CO, USA). The
cables, mobile eco ding de ice and a coun e -weigh we e ixed o he udde holde on he
back o he cow. Cables we e bunched wi h plas ic s aps and ixed o he udde holde o
educe ension on he elec odes when he cow was mo ing and o educe he isk o cables
i i a ing he cow. The impedance o he elec ode connec ions was checked and alues
be ween 0.2 and 5 kΩwe e deemed accep able. Two o h ee cows in neighbou ing pens we e
eco ded in pa allel du ing each eco ding pe iod.
Elec ophysiological da a we e eco ded by elec oencephalog aphy (EEG) o b ain ac i -
i y, elec ooculog aphy (EOG) o eye mo emen s and elec omyog aphy (EMG) o muscle
ac i i y. The da a we e sampled a 256 Hz, as desc ibed by Te nman e al. [11]. The eco ding
pe iod s a ed du ing mo ning milking a 06 h on day 1 and las ed o 72 h. Du ing he whole
72-h eco ding pe iod, he cows we e closely moni o ed o ensu e hei well-being and o
check ha he elec odes we e in place and ha he equipmen was wo king p ope ly. The da a
we e s o ed on he eco ding de ice un il he end o each 24-h pe iod and hen downloaded
on o a compu e . This was done du ing mo ning milking. Da a downloading ook 20 min and
a new eco ding was s a ed immedia ely a e he p e ious se o da a had been downloaded.
Analysis o elec ophysiological da a
The EEG, EOG and EMG da a we e analysed isually using digi al sleep eco ding, moni o ing
and analysing so wa e (RemLogic 2.0.1, Embla Sys ems, B oom ield, CO, USA) and sco ed in
30-s in e als o he igilance s a es: awake, d owsing, NREM sleep and REM sleep [11].
Be o e sco ing, he signals we e il e ed using he ollowing indi idual h esholds: EEG 0.3–30
Hz; EMG >10 Hz and EOG 0.15–15 Hz. The signals we e displayed and analysed in 30-second
in e als and gi en a ixed scale; 50 μV o EEG and EMG and 500 μV o EOG. F om each
eco ding session, one good quali y ace o he EEG and EMG, and bo h EOG aces we e
Fi s -nigh e ec in dai y cows
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0195593 Ap il 13, 2018 3 / 10
selec ed o isual analysis. Powe spec um was calcula ed using as Fou ie ans o m (FFT)
wi h sample size o 2048 poin s and 50% o e lap and was displayed in he equencies 10–30
Hz and he powe 0–1 μV
2
/Hz
The igilance s a es (awake, NREM sleep and REM sleep) we e sco ed acco ding o he
Te nman e al. [11]. REM sleep was sco ed when he EEG da a displayed a desynch onized
pa e n wi h a ying high and low ampli ude, combined wi h e y low muscle one. NREM
sleep was cha ac e ized by low equency wa es (ampli udes usually less han 15 μV) and was
sco ed when he powe spec um showed less han 0.2 μV
2
/Hz in he highe equencies (10–
30 Hz) and EMG ac i i y was educed. Da a was sco ed as d owsing when he EEG signal dis-
played a synch onized pa e n and he ac i i y in he highe equencies (10–30 Hz) o he
powe spec um was be ween 0.2 and 1.0 μV
2
/Hz, based on Ruckebusch’s [8] indings on
Fig 1. Elec ode placemen and equipmen o non-in asi e EEG eco dings in dai y cows. The snap-on elec ode
wi es we e led h ough ab ic ubes and connec ed o he eco ding de ice. A coun e weigh is a ached on he back o
he cow, opposi e he de ice.
h ps://doi.o g/10.1371/jou nal.pone.0195593.g001
Fi s -nigh e ec in dai y cows
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0195593 Ap il 13, 2018 4 / 10
cows, and wi h e e ence o Te nman e al. [11]. EMG signal du ing d owsing was educed bu
could a y due o mo emen s.
S a is ical analysis
To al ime, numbe o bou s and bou du a ion pe 24-h pe iod we e calcula ed o all sco ed
igilance s a es. In addi ion, hese pa ame e s we e calcula ed o day ime (05.00-20.00 h) and
nigh ime (20.00-05.00 h) sepa a ely. A bou was de ined as a pe iod o a gi en igilance s a e
sepa a ed by a leas one 30s epoch o a di e en igilance s a e.
Di e ences be ween eco ding days (24-h pe iods) and ime o he day we e analysed using
he PROC MIXED p ocedu e in SAS (SAS 9.3, SAS Ins i u e Inc., Ca y, NC, USA). Reco ding
day, ime o he day and eco ding day × ime o he day we e included as ixed e ec s in he
model and cow was included as a andom e ec . Values decla ed signi ican a P0.05. Pos -
hoc means sepa a ion o signi ican main e ec s was done using a T- es , da a p esen ed a e
LSM ±SE, i no hing else s a ed. No mali y and equali y o a iance we e checked by isual
inspec ion o he esiduals, and du a ion o awake bou s pe eco ding day was log10 ans-
o med due o non-no mal dis ibu ion.
Resul s
Success ul eco dings we e ob ained du ing all 24-h pe iods. Typical elec ophysiological
cu es o sleep and awake s a es and umina ion a e shown in Fig 2.
On a e age du ing he h ee consecu i e eco ding days, cows spen 12.9 ±1.4 hou s
awake, 8.2 ±1 hou s umina ing, 44.1 ±20.2 min in REM sleep and 64.3 ±38.1 min in NREM
sleep (mean ±SD). Cows also spen 57.2 ±20.3 min (mean ±SD) in he in e media e s a e
d owsing. O e all, cows we e awake mo e du ing day ime han a nigh (P <0.001; Table 1);
howe e , he e we e no e ec o eco ding day o he in e ac ion o eco ding day × ime o
he day. Cows showed mo e umina ion, REM and NREM sleep a nigh as compa ed o day-
ime (P <0.001; Table 1) bu he e we e no di e ences be ween eco ding days o o he
in e ac ion o eco ding day × ime o day.
Fig 2. Elec ophysiological eco dings om one cow du ing non- apid eye mo emen (NREM) sleep (15s), apid eye
mo emen (REM) sleep (15s), awake (15s) and umina ion (30s). The a e ac s in REM sleep da a a e he ypical muscle
wi ches occu ing du ing REM sleep. Eye mo emen s eco ded using elec ooculog aphy (EOG), b ain ac i i y using
elec oencephalog aphy (EEG), and neck muscle one using elec omyog aphy (EMG). Swallowing and e up ion o a new bolus
is enci cled in he Rumina ion image.
h ps://doi.o g/10.1371/jou nal.pone.0195593.g002
Fi s -nigh e ec in dai y cows
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0195593 Ap il 13, 2018 5 / 10

Awake bou s we e longe du ing day ime compa ed o nigh (P <0.001; Table 2), and
he e was an in e ac ion e ec o eco ding day × ime o he day wi h longe awake bou s du -
ing day ime on eco ding 1 compa ed o day ime on eco ding day 3 (11 min s 9 min;
P<0.05). Bo h REM sleep bou s and umina ion bou s we e longe du ing nigh compa ed o
day ime, bu he e we e no di e ences be ween eco ding days and no in e ac ion e ec o
eco ding day × ime o day.
The pe cen age o NREM sleep bou s we e highe a nigh compa ed o day ime (P <0.001;
Table 3), bu he e we e no e ec o ime o he day o eco ding day. Pe cen age o REM sleep
bou s did no di e be ween eco ding days o ime o day, and did no show any in e ac ion
e ec o eco ding day × ime o day (Table 3). Howe e , he pe cen age o awake and
Table 1. Pe cen o o al du a ion (min) o igilance s a es/day s nigh spli in Time o Day (TOD) and Reco ding Day (RecDay). Resul s shown as LSMeans ±SE.
Time o day (TOD) P- alues
Vigilance s a e RecDay Day Nigh RecDay ×TOD TOD RecDay DF
NREM sleep % o o al ime (min ±SEM in b acke s) 1 2.5 ±1.1 (21.0 ±7.3) 6.0 ±1.1 (36.0 ±7.3) 0.6060 <0.001 0.1147 38
2 3.4 ±1.1 (28.8 ±7.9) 8.4 ±1.1 (49.3 ±7.3)
3 2.8 ±1.1 (24.8 ±7.9) 6.6 ±1.1 (37.5 ±7.3)
REM sleep % o o al ime (min ±SEM in b acke s) 1 2.4 ±0.7 (20.1 ±4.9) 4.5 ±0.7 (26.6 ±4.6) 0.5235 <0.001 0.1563 39
2 2.6 ±0.7 (22.2 ±4.6) 5.0 ±0.7 (29.3 ±4.6)
3 2.1 ±0.7 (18.3 ±4.6) 3.1 ±0.7 (18.1 ±4.6)
Awake % o o al ime (min ±SEM in b acke s) 1 65.0 ±2.5 (555.6 ±9.5) 36.5 ±2.5 (211.8 ±9.5) 0.2037 <0.001 0.1560 40
2 62.9 ±2.5 (532.6 ±9.5) 37.4 ±2.5 (215.7 ±9.5)
3 64.4 ±2.5 (564.4 ±19.5) 43.9 ±2.5 (245.6 ±9.5)
Rumina ion % o o al ime (min ±SEM in b acke s) 1 27.5 ±1.6 (233.5 ±14.2) 47.6 ±1.6 (277.2 ±14.2) 0.1700 <0.001 0.2379 40
2 27.9±1.6 (236.3 ±14.2) 43.2 ±1.6 (250.6 ±14.2)
3 27.7 ±1.6 (243.7 ±14.2) 42.5 ±1.6 (239.1 ±14.2)
Resul s shown as LSMeans ±SEM. P- alues indica e signi ican di e ences o he in e ac ion e ec o Reco ding Day ×Time o Day (RecDay×TOD), be ween day- and
nigh ime (TOD) and be ween eco ding days (RecDay). DF = deg ees o eedom o each calcula ion.
h ps://doi.o g/10.1371/jou nal.pone.0195593. 001
Table 2. Bou du a ion (min) o igilance s a es spli in Time o Day (TOD) and Reco ding Day (RecDay). Resul s shown as LSMeans ±SE.
Time o he day P- alues
Vigilance s a e RecDay Day Nigh RecDay ×TOD TOD Recday DF
NREM sleep 1 4.4 ±0.6 4.0 ±0.5 0.4322 0.6782 0.9967 388
2 4.0 ±0.6 4.4 ±0.5
3 3.9 ±0.6 4.4 ±0.5
REM sleep 1 4.0 ±0.4 4.6 ±0.4 0.3585 <0.05 0.7913 254
2 4.4 ±0.4 4.6 ±0.3
3 4.0 ±0.4 5.2 ±0.4
Awake Log10 (s), (back- ans o med in b acke s, min) 1 6.5 ±0.2 (11.5) 5.7 ±0.1 (5.2) <0.05 <0.001 0.4781 1011
2 6.4 ±0.1 (9.8) 5.6 ±0.1 (4.7)
3 6.2 ±0.1 (8.6) 6.0 ±0.2 (6.5)
Rumina ion 1 30.5 ±2.0 32.4 ±1.9 0.7850 <0.05 0.2579 426
2 26.3 ±1.9 30.9 ±1.9
3 29.3 ±2.0 33.0 ±2.1
Resul s shown as LSMeans ±SEM. P- alues indica e signi ican di e ences o he in e ac ion e ec o Reco ding Day and Time o Day (RecDay×TOD), be ween day-
and nigh ime (TOD) and be ween eco ding days (RecDay). DF = deg ees o eedom o each calcula ion.
h ps://doi.o g/10.1371/jou nal.pone.0195593. 002
Fi s -nigh e ec in dai y cows
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0195593 Ap il 13, 2018 6 / 10
umina ion bou s showed an in e ac ion e ec wi h mo e bou s occu ing du ing day ime on
eco ding day 3 compa ed o day ime on eco ding day 1 (P <0.05 and 0.01, espec i ely),
du ing nigh ime on eco ding day 1 compa ed o nigh ime on eco ding day 3 (P <0.05
and 0.01, espec i ely), and day ime on eco ding day 3 compa ed o nigh ime on eco ding
day 3 (P <0.05 and 0.01, espec i ely; Table 3). In addi ion, o umina ion he pe cen age o
bou s we e highe also o nigh - ime on day 1 compa ed o day ime on eco ding day 1
(P <0.01; Table 3).
Discussion
This s udy examined whe he o al du a ion and he a chi ec u e o REM sleep and awake in
dai y cows di e be ween eco ding days, o be ween he nigh s o each eco ding day espec-
i ely, using a non-in asi e me hod eco ding b ain and muscle ac i i y and eye mo emen s.
We ound ha cows we e awake mo e du ing day ime compa ed o nigh ime and ha each
bou was longe du ing day ime compa ed o a nigh . This diu nal e ec o awake is mos
likely due o he managemen ou ines as hese a e known o signi ican ly a ec dai y cow
beha iou [13,14]; in he p esen s udy all eeding and mos cleaning o he pen occu ed du -
ing day ime. Fu he mo e, he in e ac ion be ween eco ding day and ime o he day, wi h
longe awake bou s du ing day ime on day 1 compa ed o day 3, could indica e ha cows we e
a ec ed by he equipmen and/o housing si ua ion. The eco dings we e always s a ed du ing
mo ning milking on eco ding day 1, and cows would po en ially ha e been mo e a ec ed by
he equipmen du ing day ime on day 1. The i s -nigh e ec in human subjec s includes a
p olonged la ency o all asleep [4,2] and an inc eased numbe o awakenings on he i s
nigh compa ed o nigh 2-4 [1]. Howe e , in ou s udy he e was no e ec on numbe o
awake bou s be ween he eco ding days, sugges ing ha he cha ac e o a possible i s -nigh
e ec in cows migh di e om wha has been shown in human subjec s.
Al e a ion o REM sleep ime and a chi ec u e is an addi ional aspec o he i s -nigh e ec
in human subjec s [15,3]. Cows in he p esen s udy displayed se e al REM sleep bou s pe
eco ding day indica ing a polyphasic sleep pa e n. This ype o sleep a chi ec u e is common
o mos mammals and has also been shown in sleep s udies on dai y cows [8] and dai y cal es
Table 3. Pe cen age o numbe o bou s pe igilance s a e/day s nigh spli in Time o Day (TOD) and Reco ding Day (RecDay). Resul s shown as LSMeans ±SE.
Time o he day P- alues
Vigilance s a e RecDay Day Nigh RecDay ×TOD TOD Recday DF
NREM % o o al numbe o bou s (ac ual numbe in b acke s) 1 35.8 ±5.9 (5 ±1) 64.2 ±5.9 (9 ±1) 0.6013 <0.001 1.0000 38
2 33.5 ±5.9 (7 ±1) 66.5 ±5.9 (11 ±1)
3 39.4 ±5.9 (6 ±1) 60.6 ±5.9 (8 ±1)
REM % o o al numbe o bou s (ac ual numbe in b acke s) 1 44.9 ±6.9 (5 ±1) 55.1 ±6.9 (6 ±1) 0.1565 0.5669 1.0000 39
2 44.0 ±6.9 (5 ±1) 56.0 ±6.9 (6 ±1)
3 56.2 ±6.9 (5 ±1) 43.8 ±6.9 (3 ±1)
AWAKE % o o al numbe o bou s (ac ual numbe in b acke s) 1 45.5 ±3.6 (15 ±2) 54.5 ±3.6 (18 ±2) <0.05 0.8808 1.0000 40
2 48.3 ±3.6 (20 ±2) 51.7 ±3.6 (21 ±2)
3 55.5 ±3.6 (21±2) 44.5 ±3.6 (16 ±2)
RUMINATION % o o al numbe o bou s (ac ual numbe in b acke s) 1 46.6 ±2.0 (7 ±0.5) 53.4 ±2.0 (8 ±0.5) <0.01 0.4424 1.0000 40
2 52.1 ±2.0 (9 ±0.5) 47.9 ±2.0 (8 ±0.5)
3 53.2 ±2.0 (8 ±0.5) 46.8 ±2.0 (7 ±0.5)
Resul s shown as LSMeans ±SEM. P- alues indica e signi ican di e ences o he in e ac ion e ec o Reco ding Day and Time o Day (RecDay×TOD), be ween day-
and nigh ime (TOD) and be ween eco ding days (RecDay). DF = deg ees o eedom o each calcula ion.
h ps://doi.o g/10.1371/jou nal.pone.0195593. 003
Fi s -nigh e ec in dai y cows
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0195593 Ap il 13, 2018 7 / 10
[10]. The only e ec on REM sleep in he p esen s udy was he ime o he day, wi h longe
REM sleep bou s du ing nigh ime compa ed o day ime.
Rumina ion was sco ed as a dis inc igilance s a e, since elec ophysiological da a du ing
umina ion a e domina ed by a e ac s om he chewing [11]. Al hough ne e ully examined,
i was p oposed by Bell and I abishi [16] and by Ruckebusch [8] ha uminan s can engage in
he igilance s a es awake and NREM sleep bu no REM sleep [17] du ing umina ion. Simi-
la ly o Ruckebusch [8], we ound ha cows spen abou 8 hou s pe day umina ing and he e
we e no di e ences be ween eco ding days. As o he REM sleep bou leng hs, umina ion
bou s we e longe du ing nigh - ime compa ed o day ime, bu he e was no hing ha could
indica e a “ i s -nigh e ec ”, such as e ec o eco ding day o an in e ac ion be ween eco d-
ing day × ime o day. Howe e , i migh sugges ha also umina ion bou s a e a ec ed by
managemen ou ines, such as eeding, milking and cleaning, and ha cows engage in longe
bou s o beha iou s p e e ably pe o med while lying down when he e is li le ac i i y in he
ba n. The dec ease in numbe o umina ion bou s on day ime du ing eco ding day 1 com-
pa ed o day ime du ing eco ding day 2 migh indica e ha cows we e no com o able du -
ing he i s day o eco ding as compa ed o he second day o eco ding. Howe e , he o al
ime cows spen umina ing showed no in e ac ion e ec o eco ding day and ime o day
and nei he did leng h o each bou , why his is belie ed o be o less impo ance.
To al NREM sleep du a ion pe day in he p esen s udy was signi ican ly sho e , jus o e
an hou , han wha has been p esen ed by Ruckebusch [8], abou 3 hou s. In he la e s udy,
s age o lac a ion, p egnancy s a us and age a e no epo ed, bu hese ac o s ha e been
shown o al e sleep ime in o he species [18,19]. Ruckebusch [8] also epo ed ha cows e-
quen ly d owse du ing umina ion and ha cows may all asleep o sho pe iods du ing
umina ion. Using he non-in asi e eco ding echnique, igilance s a e du ing umina ion
canno be es ima ed due o he muscle a e ac s om chewing, hus we migh ha e unde es i-
ma ed he amoun o NREM sleep he cows had in ou s udy. Cows showed mo e NREM sleep
a nigh compa ed o day ime, which con i ms indings by Ruckebusch [8]. Ruckebusch [9]
concluded ha when cows a e housed indoo s hei sleep is es ic ed o nigh ime, his was
howe e no he case in he p esen s udy. As men ioned p e iously, i is possible ha he cows
s udied by Ruckebusch [8,9,17] p e e ed o es and sleep a e s a wo king hou s, when
he e was less dis u bance in hei immedia e en i onmen .
The a ia ion be ween cows o o al du a ion, bou leng hs and numbe o bou s in awake
and sleep s a es was la ge making i di icul o es ablish a pa e n o he i s -nigh e ec wi h
his limi ed sample size. Howe e , he e seem o be a wi hin cow consis ency indica ing ha
using he same indi iduals o egis e sleep unde di e en condi ions, like s age o lac a ion o
housing sys em, is mo e impo an han he numbe o days eco ded. Mo e s udies wi h a
la ge numbe o animals a e needed o es ablish he de ailed i s -nigh e ec s o sleep eco d-
ings in dai y cows, bu o o al amoun o sleep and awake s a es he e seem o be no di e -
ence be ween eco ding days. This is in ag eemen wi h s udies in human subjec s whe e o al
sleep ime also was una ec ed by day o eco ding [2,3]. Type o eco ding me hod should be
aken in o accoun as i may a ec he o al du a ion and he a chi ec u e o sleep since cows
may be mo e o less dis u bed by he equipmen . A minimally in asi e me hod wi h wi eless
da a ans e o he polyg aphic da a would pe haps dis u b he cows less.
Conclusion
Da a on sleep and awake s a es eco ded in adul dai y cows using a non-in asi e elec ophysi-
ological me hod applied du ing h ee consecu i e 24-h pe iods showed g ea a ia ion in sleep
ime be ween cows, bu he a ia ion be ween days in o al ime o any o he eco ded sleep
Fi s -nigh e ec in dai y cows
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0195593 Ap il 13, 2018 8 / 10
and awake s a es was no as p onounced. This indica es ha e en s eco ded du ing he i s
24-h pe iod can be ega ded as ep esen a i e i he o al amoun a he han he a chi ec u e
o he sleep and awake s a es a e o p ima y in e es . Fu he and mo e de ailed s udies, wi h a
la ge sample size, o how sleep a chi ec u e is a ec ed by eco ding day is necessa y o ully
comp ehend he i s -nigh e ec in dai y cows.
Suppo ing in o ma ion
S1 Table. Raw da a om he s udy. Du a ion is shown in seconds, abb e ia ions o e en s
a e W = awake, N1 = d owsing, N2 = NREM sleep, N3 = REM sleep and R = Rumina ion, and
RecDay co esponds o eco ding day.
(XLSX)
Au ho Con ibu ions
Da a cu a ion: Emma Te nman, Lau a Ha¨nninen, Sig id Agena¨s, Pe P. Nielsen.
Fo mal analysis: Emma Te nman, Ma i Pas ell, Lau a Ha¨nninen, Sig id Agena¨s, Pe P.
Nielsen.
Me hodology: Emma Te nman, Lau a Ha¨nninen, Sig id Agena¨s, Pe P. Nielsen.
P ojec adminis a ion: Emma Te nman.
Resou ces: Sig id Agena¨s, Pe P. Nielsen.
So wa e: Ma i Pas ell, Lau a Ha¨nninen, Pe P. Nielsen.
Supe ision: Lau a Ha¨nninen, Sig id Agena¨s.
Valida ion: Ma i Pas ell.
W i ing – o iginal d a : Emma Te nman, Ma i Pas ell, Lau a Ha¨nninen, Sig id Agena¨s, Pe
P. Nielsen.
W i ing – e iew & edi ing: Emma Te nman, Ma i Pas ell, Lau a Ha¨nninen, Sig id Agena¨s,
Pe P. Nielsen.
Re e ences
1. Rech scha en A, Ve done P. Amoun o d eaming: E ec o incen i e, adap a ion o labo a o y, and
indi idual di e ences. Pe cep Mo Skills. 1964; 19: 947–958. h ps://doi.o g/10.2466/pms.1964.19.3.
947 PMID: 14238245
2. Wauquie A, an Sweden B, Ke kho GA, Kamphuisen HAC. Ambula o y i s nigh sleep e ec eco d-
ing in he elde ly. Beha B ain Res. 1991; 42: 7–11. PMID: 2029347
3. Le Bon O, S ane L, Ho mann G, D amaix M, San Sebas ian I, Mu phy e al. The i s -nigh e ec may
las mo e han one nigh . J Psychia Res. 2001; 35: 165–172. PMID: 11461712
4. Agnew HW, Webb WB, Williams RL. The i s nigh e ec : An EEG s udy o sleep. Psychophysiology.
1966; 2: 263–266. PMID: 5903579
5. Kis A, Szakada
´ S, Simo P, Gombos F, Ho a
´ h K, Bo
´dizs R. Objec i e and subjec i e componen s o
he i s -nigh e ec in young nigh ma e su e e s and heal hy pa icipan s. Beha Sleep Med. 2013; 12
(6): 469–480. h ps://doi.o g/10.1080/15402002.2013.829062 PMID: 24294972
6. on Keyse lingk MAG, Olenick D, Wea y DM. Acu e Beha io al E ec s o Reg ouping Dai y Cows. J
Dai y Sci 2008: 91: 1011–1016. h ps://doi.o g/10.3168/jds.2007-0532 PMID: 18292257
7. I o K, Wea y DM, on Keyse lingk MAG. Lying beha io : Assessing wi hin- and be ween he d a ia ion
in ee-s all-housed dai y cows. J Dai y Sci 2009: 92: 4412–4420. h ps://doi.o g/10.3168/jds.2009-
2235 PMID: 19700701
8. Ruckebusch Y. Rele ance o d owsiness in ci cadian cycle o a m animals. Anim Beha . 1972; 20:
637–643. PMID: 4661312
Fi s -nigh e ec in dai y cows
PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0195593 Ap il 13, 2018 9 / 10