![Hourly variations in feed (as-fed basis) and free water intake of lactating dairy cows [least squared mean (LSM) ± SEM]. Values are expressed as % of the total daily intake. Cows were fed at 0800 and 1800 h and milked at 0530 and 1530 h twice per day.](https://www.researchgate.net/profile/Aira-Maye-Serviento/publication/381084567/figure/fig2/AS:11431281253887332@1718962981995/Hourly-variations-in-feed-as-fed-basis-and-free-water-intake-of-lactating-dairy-cows_Q320.jpg)
Content uploaded by Aira Maye Serviento
Author content
All content in this area was uploaded by Aira Maye Serviento on Jun 21, 2024
Content may be subject to copyright.
Modeling the effect of ambient temperature on reticulorumen
temperature, and drinking and eating behaviors of late-lactation dairy
cows during colder seasons
A.M. Serviento
a,1
,T.He
a,b,1
,X.Ma
a
, S.E. Räisänen
a
, M. Niu
a,
⇑
a
Animal Nutrition Group, Institute of Agricultural Sciences, Department of Environmental Systems Science, ETH Zürich, 8092 Zürich, Switzerland
b
College of Animal Science and Technology, China Agricultural University, Beijing 100193, China
article info
Article history:
Received 24 January 2024
Revised 24 May 2024
Accepted 24 May 2024
Available online 31 May 2024
Keywords:
Feed intake
Generalized additive mixed model
Thermal environment
Thermoregulation
Water intake
abstract
Dairy cows may suffer thermal stress during the colder seasons especially due to their open-air housing
systems. Free water temperature (FWT) and feed temperature (FT) are dependent on ambient tempera-
ture (AT) and can be critical for maintaining body and reticulorumen temperature (RT) in cold conditions.
The objective of this study was to determine the effects of FWT and FT on RT fluctuations, and of AT on RT
and drinking and eating behaviors in late-lactation cows during cold exposure. Data were collected from
16 multiparous lactating cows for four 6-d periods during the autumn and winter seasons. The cows
(224 ± 36 days in milk; mean ± SD) had an average milk yield (MY) of 24.8 ± 4.97 kg/d and RT of 38.8
4 ± 0.163 °C. Daily average AT ranged from 4.38 to 17.25 °C. The effects of the temperature and amount
of the ingested water or feed on RT change and recovery time, and the effect of the daily AT on RT, feed
and water intake, and drinking, eating, and rumination behaviors were analyzed using the generalized
additive mixed model framework. Reticulorumen temperature change and recovery time were affected
by FWT (+0.0596 °C/°C and 1.27 min/°C, respectively), but not by FT. The amount of the ingested free
water and feed affected RT change (0.108 °C/kg drink size and 0.150 °C/kg meal size, respectively),
and RT recovery time (+2.13 min/kg drink size and + 3.71 min/kg meal size, respectively). Colder AT
decreased RT by 0.0151 °C/°C between 9.91 and 17.25 °C AT. Cows increased DM intake (DMI)by
0.365 kg/d per 1 °C drop in AT below 10.63 °C, but with no increase in MY. In fact, MY:DMI decreased
by 0.0106/°C as AT dropped from 17.25 to 4.38 °C. Free water intake (FWI) was reduced by 0.0856
FWI:DMI/°C as AT decreased from 17.25 to 8.27 °C. Cold exposure influenced animal behavior with fewer
drink and meal bouts (0.432 and 0.290 bouts/d, respectively), larger drink sizes (+0.100 kg/bout), and
shorter rumination time (5.31 min/d) per 1 °C decrease in AT from 17.25 °C to 8.77, 12.53, 4.38, and
10.32 °C, respectively. In conclusion, exposure to low AT increased feed intake, reduced water intake,
and changes in eating, drinking and rumination behaviors of dairy cows in late lactation. Additionally,
the consequences of cold exposure on cows may be aggravated by ingestion of feed and free water at
temperatures lower than the body, potentially impacting feed efficiency due to the extra energetic cost
of thermoregulation.
Ó2024 The Author(s). Published by Elsevier B.V. on behalf of The Animal Consortium. This is an open
access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Implications
Colder seasons in temperate climates often reach ambient tem-
peratures lower than the body temperature maintained by cows.
The present study demonstrates decreased reticulorumen temper-
ature, increased feed intake, decreased water intake, and decreased
feed efficiency of late lactation dairy cows as ambient temperature
decreased. Ingestion of coldfree water and feed also leads to
greater fluctuations in reticulorumen temperature, partly explain-
ing the decreased feed efficiency that was likely due to the higher
energy cost of thermoregulation. These findings provide a potential
guide in practices to mitigate thermal stress, optimizing cow health
and productivity in cold climates, benefiting the livestock industry.
Introduction
Ruminants, including dairy cows, tightly regulate their body
temperature in a wide range of ambient temperature (AT)
https://doi.org/10.1016/j.animal.2024.101209
1751-7311/Ó2024 The Author(s). Published by Elsevier B.V. on behalf of The Animal Consortium.
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
⇑
Corresponding author.
E-mail address: mutian.niu@usys.ethz.ch (M. Niu).
1
These authors contributed equally to this work.
Animal 18 (2024) 101209
Contents lists available at ScienceDirect
Animal
The international journal of animal biosciences
(Collier and Gebremedhin, 2015). Ideally, AT should be lower than
the animal’s body temperature to allow body heat loss. However,
when AT is too low such as during cold exposure, the body may
lose more heat than it produces, leading to a risk of hypothermia.
Modern dairy cows, due to genetic selection for higher milk pro-
duction, have high metabolic activity especially during lactation.
Therefore, they are typically known to be more resistant to cold
temperature and to have a relatively low thermoneutral zone that
ranges from 5 to 20 °C(NASEM, 2021). As a result, there are several
studies related to the effects of heat stress on lactating dairy cows
(Kadzere et al., 2002; Tao et al., 2020) and notably fewer studies
have investigated responses of cows in cold conditions (Young,
1983; Brouc
ˇek et al., 1991). However, temperate climates during
colder seasons can often fall below their thermoneutral zone, and
far below the standard body temperature they maintain which typ-
ically varies between 38–39 °C(Ammer et al., 2016). Therefore,
cold exposure might also be a problem especially for dairy cows,
especially with the increasingly extreme global weather conditions
(Cohen et al., 2021) and with cows being exposed to outdoor con-
ditions due to the open or semi-open housing of most dairy cow
production systems.
Rumen temperature is essential for maintaining optimal rumen
conditions for microbial fermentation (Linville et al., 2017) and can
serve as a reliable indicator of core body temperature in cows
(Ammer et al., 2016). The use of radiotelemetric ruminal boluses
allows reticulorumen temperature (RT) measurements that show
the variability of the animal’s heat exchange with its environment
throughout the day. Research has already demonstrated that RT
fluctuations are affected by a variety of factors such as AT and
behavioral activities, i.e., drinking, eating, and lying (Bewley
et al., 2008b; Liang et al., 2013; Rutherford et al., 2019). Drinking
activities are reported to cause distinct RT changes (Ipema et al.,
2008; Rutherford et al., 2019), with studies indicating that both
the quantity and temperature of water during drinking events
can impact RT fluctuations (Bewley et al., 2008b; Petersen et al.,
2016; Cantor et al., 2018). However, the immediate effects of eat-
ing events on RT remain underexplored.
Given that the rumen is directly in physical contact with
ingested feed and water, ingesta characteristics (i.e., amount and
temperature) can likely affect the extent of RT fluctuations. More-
over, free water temperature (FWT) and feed temperature (FT),
which are intrinsically linked to AT, could play vital roles in regu-
lating RT, especially in open barn systems during colder seasons
(e.g., autumn and winter) when AT, FWT, and FT are all substan-
tially lower than RT. These factors are important as they can have
direct implications on drinking and eating behaviors which are
well-known thermoregulatory responses to AT changes
(Bernabucci et al., 2010). In this context, we hypothesized that
ambient temperatures far below body temperature, their corre-
sponding influence on FWT and FT, and the ensuing drinking and
eating events collectively decrease RT and affect the eating and
drinking behaviors of lactating dairy cows. The objectives of the
current study were to determine the effects of FWT and FT on RT
fluctuations, and to evaluate how AT affects RT, drinking and eating
behaviors, and feed efficiency in late-lactation dairy cows during
colder seasons. Preliminary results of the present study were pre-
viously presented in abstract form (He et al., 2023).
Material and methods
Animal experiment, measurements, and data collection
The experiment was conducted from September to December of
2022 at the AgroVet-Strickhof Dairy Farm (Lindau, Switzerland).
This is an observational study using data from another experiment
designed to achieve a different set of objectives (Ma et al., 2023).
The present study focuses on the analysis of RT and animal behav-
ioral parameters, such as drinking, eating, and rumination, in rela-
tion to AT. The experiment from which the data were taken
included 16 multiparous lactating dairy cows [24.8 ± 4.97 kg/d of
milk yield (MY), 224 ± 36 days in milk, 727 ± 46.6 kg BW;
mean ± SD] arranged in a split-plot design, with breed as the main
plot (eight Brown Swiss and eight Holstein Friesian). Within each
subplot, the cows were used in a 4 4 Latin Square design with
four dietary treatments and four experimental periods. The four
diets allocated to the cows as total mixed ration (TMR) had similar
nutrient composition (16.0% CP, 43.0% NDF, 3.98% EE, and 22.2%
starch). All cows had ad libitum access to feed, provided at 110%
of the previous day’s intake, and to drinking water. Cows were
individually fed twice daily at 0800 and 1800 h at a 60:40 ratio.
Refusals were collected daily at 0630 h before fresh feed delivery
in the morning. Cows were milked twice daily at 0530 and 1530 h.
Data were collected for a total of 24 d (6 d per period, with each
period spaced 18 d apart) when the animals were in a tie-stall
barn. Feed intake (FI) and free water intake (FWI) of individual
cows were continually measured using calibrated built-in floor
scales (every 10 s; PFA575 Mettler Toledo, Greifensee, Switzerland)
and using flow meters (every 15 min; GWF technology, Kauf-
beuren, Germany), respectively. Additional manual daily record-
ings were made for feed offerings, refusals, FWI (via flow meter
readings), and MY. Radiotelemetric ruminal boluses (smaXtec
bolus, Freienbach, Switzerland) were used to measure RT every
10 min throughout the experiment. RumiWatch (Itin + Hoch
GmbH, Fütterungstechnik, Liestal, Switzerland) noseband halters
were used to evaluate rumination behavior during the data collec-
tion periods (Zehner et al., 2017). Ambient temperature (AT) and
relative humidity (RH) in the tie-stall barn were recorded at 5-
min intervals (Neatmo smart system, Boulogne-Billancourt,
France). As the source of drinking water was similar for all cows
and a limited number of temperature probes, 3 temperature probes
were randomly placed in different drinking bowls during each data
collection period to measure the drinking water temperature (ev-
ery 15 min; iButtonLink Technology, Whitewater, WI., USA).
Calculations and data processing
Temperatures
The temperature-humidity index was computed using the fol-
lowing formula: THI = 0.8 AT + RH (AT – 14.4) + 46.4, with
AT in °C and RH as a proportion (Mader et al., 2006; Nienaber
and Hahn, 2007). Although several hours of drinking water tem-
perature data were missing due to technical problems of the tem-
perature probes, we observed strong positive correlation between
FWT and AT (r = 0.89). Therefore, FWT was calculated from its rela-
tionship with AT using the resulting equation: FWT = 0.12612 + 0.
8998 AT. Feed temperature was assumed to be equal to AT. For
RT to have a similar resolution with AT, FWT, and FT (in 5-min
interval) and to align with the FWI measurements (in 15-min
interval), RT of each 5 min was considered to be the average
between two consecutive measurements.
Eating and drinking behaviors
An experimental day was defined as the 24-h period from 0701
to 0700 h of the next day. Daily FI was the difference between
offered and refused feed, and daily FWI was summed from flow
meter readings. Feed intake was divided into DM and in-feed water
components. Total water intake (TWI) was calculated as the sum of
FWI and in-feed water intake. In addition, FWI, TWI, and MY were
also expressed per kg DM intake or DMI (FWI: DMI, TWI: DMI, and
MY: DMI, respectively), and hourly FWI and FI were calculated as
percentage of the daily total based on the activity of the previous
A.M. Serviento, T. He, X. Ma et al. Animal 18 (2024) 101209
2
hour. For example, the reported intake at 0900 h was aggregated
from 0801 to 0900 h.
Eating behavior parameters were calculated from continuous
weight measurements of the feeding plates on as-fed basis. Two
consecutive feeding visits separated by a time interval not exceed-
ing a certain meal criterion were counted as the same meal. To
identify this criterion, time intervals ranging from 1 to 60 min were
tested that resulted in different average numbers of meal bouts per
day. Similar to the approach of Dado and Allen (1993), pointwise
slopes (derivatives) of the tested time intervals were calculated
to determine the accepted minimum interbout interval. The meal
criterion of this study was determined to be 24 min. Using this
meal criterion, daily eating behavior parameters were calculated
similarly to Niu et al. (2017) including meal bout frequency (n/
d), meal size (kg/bout; amount of FI in kg per meal bout), meal bout
length (min/meal; time interval from the start to the end of a bout),
meal bout interval (min/bout; time interval from the end of a bout
to the start of the consecutive bout), total eating time (min/d; meal
bout frequency meal bout duration), and eating rate (kg/min;
meal size/meal bout length).
Drinking behavior was analyzed by considering a single 15-min
log with a recorded FWI as one drinking bout. Key parameters cal-
culated daily included drink bout frequency (n/d), drink size (kg/
bout), and drink bout interval (min/bout). For rumination behavior,
data were processed into 10-min intervals using RumiWatch con-
verter software (version 0.7.4.13). Due to the rhythmic nature and
extended duration of rumination activities, two or more consecu-
tive 10-min recordings with a recoded rumination activity were
considered to belong to one rumination bout. Based on this crite-
rion, daily rumination parameters were calculated, including rumi-
nation bout frequency (n/d), rumination bout length (min/bout),
rumination bout interval (min/bout), and total rumination time
(min/d).
Reticulorumen temperature features related to ingestion events
Fluctuations of RT were analyzed in relation to ingestion events,
encompassing both individual drinking and eating instances (col-
lectively termed as ingestion events) and were identified as previ-
ously described in the context of drinking and eating (meal) bouts.
Specific details captured for each event included timing, and
respective amount and temperature of free water or feed con-
sumed. Fig. 1 visually illustrates the RT parameters derived from
these ingestion events. For each event, continuous RT measure-
ments were taken starting from 15 min prior to drinking (predrink)
or 5 min prior to eating (pre-eat) until the occurrence of the next
consecutive event. From this continuous data, three critical points
were identified for each event: pre-event RT (Point A; Fig. 1), poste-
vent minimum RT (Point B), and recovery RT (Point C; the moment
when RT recovered to within ± 0.10 °C of the pre-event level). The
RT dynamics were then characterized using two specific parame-
ters: (1) RT change, defined as the temperature difference between
Point B and Point A (B – A), and, (2) RT recovery time, defined as the
duration required to recover from Point B to Point C.
Drinking and eating events were processed separately and fil-
tered to reduce confounding effects related to quantity or temper-
ature of the ingesta, as detailed in the filtration flowchart in Fig. 2.
Out of the initially identified 3 025 drinking (with > 1 kg drink size)
and 3 338 eating events, 50 and 62% were excluded due to missing
or unidentified Points A, B, and/or C. Such exclusions were linked to
undetected RT drop or failure to recover RT before the next event.
Further filtration was conducted for overlapping events, specifi-
cally, 61% of drinking events coinciding with a meal or eating activ-
ity, and 44% of eating events coinciding with a drinking activity.
Additionally, 11% of drinking events occurring as part of two or
more consecutive drink bouts were also excluded. From these
remaining observations, events that contained outliers (values
with less than or greater than 1.5 times the interquartile range)
of drink or meal size, FWT or FT, RT change and RT recovery time
parameters were excluded. Ultimately, 490 drinking and 618 eat-
ing events were retained representing 16 and 19%, respectively,
of the total observations identified.
Modeling the effect of ambient temperature in dairy cows
Reticulorumen temperature fluctuations trigged by ingestion events
The two response variables describing RT dynamics, namely RT
change and RT recovery time, were analyzed using a generalized
additive mixed model framework using gam procedure of mgcv
package (Wood, 2011) in R. The model was developed as shown:
Y¼b
0
þsTðÞþsSðÞþC
i
þe
where Ydenotes the response variable, b
0
is the intercept, sTðÞ
denotes the fixed smooth term of free water or feed temperature,
sðSÞdenotes the fixed smooth term of drink or meal size, and C
i
denotes the random effect of cow (i = 1–16), and edenotes the ran-
dom residual error. The smooth term used was thin plate regression
splines, which are lowrank isotropic smoothers (Wood, 2017). The
restricted maximum likelihood method was used for the estimation
of smoothing parameters. A P-value of < 0.05 was considered statis-
tically significant.
Eating and drinking behaviors
To describe intraday variations in drinking and eating behav-
iors, hourly FWI and FI (expressed as % of total daily intake) were
analyzed using a mixed effect model using the lmer (Bates et al.,
2015) procedure of R statistical language (R Core Team, 2023;R
version 4.3.0) with time of day, period, diet, and breed as the fixed
effects and cow as random effect. Data points with studentized
residuals outside of ± 3 were considered outliers and were
removed from the analysis (< 5% and < 3% data points removed
for FWI and FI, respectively).
For modeling the effect of AT, the dataset for eating, drinking,
and rumination behavioral parameters was on the cow-day basis,
as different days had variable AT. Similarly, the random effect of
cow was included in the model as different days. Incomplete
records caused by technical issues such as lost connection or bat-
tery problem were removed. Specifically, rumination data from d
1 of each data collection period were considered as adaptation
and excluded from further analyses. As a result, a total of 384,
367, 367, 234, and 211 observations were included for analyzing
daily DMI, FWI, TWI, drinking, eating, and rumination behaviors,
respectively. The summary statistics of related variables are
detailed in Table 1. All the response variables except for FWI,
TWI, and MY were fitted to a generalized additive mixed model
using gam procedure in R. The model was developed as shown,
Y¼b
0
þsATðÞþD
i
þB
j
þP
k
þC
l
þe
where Ydenotes the response variable, b
0
is the intercept, sATðÞ
denotes the fixed smooth term of the average ambient temperature
of the day (AT), D
i
denotes fixed effect of the diet (i = 1–4), B
j
denotes the fixed effect of breed (j = 1 and 2), P
k
denotes the random
effect of period (k = 1–4), C
l
denotes the random effect of cow (l = 1–
16), and edenotes the random residual error. For FWI, TWI, and MY,
a similar generalized additive mixed model was used with an addi-
tional smooth term of the DMI, s DMIðÞ. The smooth term and
parameter estimation method used were similar as described
above.
Quantifying the significant changes in response variables
From the fitted generalized additive mixed models, the AT
range with significant changes was identified for each response
variable as described by Larsen et al. (2020). In brief, the pointwise
A.M. Serviento, T. He, X. Ma et al. Animal 18 (2024) 101209
3
Fig. 1. Illustration of the dairy cow reticulorumen temperature (RT) change and RT recovery time calculations.
Fig. 2. Data filtering flowchart depicting a number of ingestion events identified and used in the final analysis of event-based reticulorumen temperature changes in dairy
cows. IQR = interquartile range. Obs. = observations.
A.M. Serviento, T. He, X. Ma et al. Animal 18 (2024) 101209
4
derivatives (i.e., the slopes) of the fitted means were determined
based on the partial effects of AT and the 95% confidence intervals
of these slopes were calculated using the derivatives function of
the gratia package (Simpson, 2023) in R. For parameters relating
to intake and behavior, the AT range showing significant change
was defined as areas where the simultaneous confidence interval
of the derivative (slope) did not include zero. If the response vari-
able within the significant AT range had a relatively linear response
to AT, all fitted mean values of the response variable within the
identified AT range were analyzed using a linear model to estimate
the marginal effect of AT. For the drinking and eating events, a sim-
ilar statistical procedure was conducted to extract the ranges of the
free water or feed temperature and the drink or meal size with sig-
nificant effect on the RT parameters, and to estimate their marginal
effects on the RT parameters.
Results
Eating and drinking behaviors
The daily average AT and RH during the study are presented in
Fig. 3, and the summary statistics for AT, RH, and THI are outlined
in Table 1. The daily average AT during the study ranged from 4.38
to 17.3 °C. The RH remained relatively stable throughout the exper-
iment, registering at 73.3 ± 2.02%, and the daily average THI was
55.0 ± 6.10. The statistics summarizing eating and drinking behav-
iors are detailed in Table 1. Cows in the present study had, on aver-
age (mean ± SD), a DMI of 24.7 ± 3.52 kg/d and a TWI of 111 ± 17.
9 kg/d (with 77% from free water and 23% from the water compo-
nent of feed). Frequencies of drink and meal bouts were similar, at
11.8 ± 3.91 and 11.8 ± 2.20 bouts per day, respectively, and so were
the drinking and meal intervals (97.8 ± 43.59 and 93.1 ± 19.73 min/
bout, respectively). Drink sizes were larger than meal sizes (7.97
± 2.787 kg/drink bout vs 4.00 ± 1.088 kg as-fed/meal bout). Rumi-
nation bout frequency was 15.0 ± 2.60 per day with an average
interval of 63.4 ± 11.71 min between bouts. The cows spent an
average of 330 ± 83.5 min eating and 476 ± 84.9 ruminating per
day, with each bout having an average duration of 28.9 ± 8.60 mi
n and 32.5 ± 6.30 min, respectively. Time spent drinking was not
measurable in this experiment due to the limited resolution of
the water intake recording system (one log every 15 min is the
sum of the water intake recorded after the previous consecutive
log).
The hourly variations in drinking and eating behaviors are dis-
played in Fig. 4. The cows consumed around 70% of their total daily
FI between 0800 and 1900 h, with 2 main peaks of eating activity
at 0900 and 1900 h. These peaks corresponded with the morning
and afternoon fresh feed delivery, accounting for 11.9 and 9.0% of
daily FI, respectively. The highest peak of drinking activity
occurred at 0900 h after morning feeding, where the cows drank
8.0% of their daily FWI. Similar to FI, the majority of the total
FWI (61%) was consumed during the day between 0800 and
1900 h. Increased drinking activities were also observed after
morning and afternoon milkings with 2.7% of their daily FWI
observed at 0600 h and 6.1% at 1600 h, respectively.
Individual drinking and eating events
The summary statistics for drinking and eating events and the
corresponding RT parameters included in the final analysis can
be found in Table 2. In the observations utilized, the ingested
FWT ranged from 2.31 to 19.0 °C with an average (mean ± SD), of
12.9 ± 3.60 °C, while the ingested FT spanned 0.85–20.10 °C, aver-
aging 12.0 ± 4.32 °C. Drink sizes ranged from 2.00 to 20.0 kg, aver-
aging 8.60 ± 4.211 kg, and meal sizes ranged from 0.35 to 7.27 kg,
with an average of 2.68 ± 1.549 kg. Reticulorumen temperature
Table 1
Summary statistics of daily climatic parameters, reticulorumen temperature, intake, and behavior of the dairy cows
1
.
Item n
2
Mean Median Min Max SD
Climatic parameters
Ambient temperature, °C 24 12.5 13.0 4.38 17.3 3.99
Relative humidity, % 24 73.3 72.8 69.2 78.1 2.02
Temperature-humidity index 24 55.0 55.7 42.7 62.3 6.10
Reticulorumen temperature, °C 384 38.84 38.83 38.47 39.41 0.163
Milk yield, kg/d 377 24.8 25.0 25.0 39.1 4.968
Intake
Water intake, kg/d
From feed 384 25.4 25.3 15.6 38.5 4.15
Free water 367 85.8 86.0 44.0 130 16.43
Total 367 111 111 66.1 161 17.9
DM intake, kg/d 384 24.7 24.6 13.9 37.9 3.52
Drinking parameters
Bouts, n/d 367 11.8 11.0 4.00 23.0 3.91
Size, kg/bout 367 7.97 7.36 4.00 20.3 2.787
Interval, min/d 367 97.8 86.0 25.5 268 43.59
Meal parameters
Bouts, n/d 234 11.8 11.0 7.00 19.0 2.20
Size, kg/bout 234 4.00 3.97 1.78 7.41 1.088
Length, min/bout 234 28.9 28.0 13.3 69.4 8.60
Interval, min/bout 234 93.1 91.5 46.5 167 19.73
Total eating time, min/d 234 330 321 167 647 83.5
Eating rate, kg/min 234 0.147 0.143 0.0925 0.234 0.0294
Rumination parameters
Bouts, n/d 211 15.0 15.0 8.00 24.0 2.60
Length, min/bout 211 32.5 32.7 15.1 48.2 6.30
Interval, min/bout 211 63.4 61.3 37.5 122 11.71
Total rumination time, min/d 211 476 489 265 639 84.9
1
Except for climatic parameters, all measurements were from 16 lactating dairy cows, four non-consecutive 6-d data collection periods during the fall and winter seasons.
The values reported are summarized from the daily average production and behavioral parameters of the cows.
2
Number of observations for respective variables.
A.M. Serviento, T. He, X. Ma et al. Animal 18 (2024) 101209
5
change averaged 1.74 ± 1.182 °C after a drinking event with a
range of 0.110 to 5.23 °C, and averaged 1.01 ± 1.073 °C after
an eating event with a range of 0.100 to 4.23 °C. Upon reaching
the minimum RT after an ingestion event (Point B; Fig. 1), it took an
average of 44.9 ± 35.50 min for the RT to recover and to reach pre-
drink level and 34.5 ± 25.55 min to reach pre-eat level. The shortest
and longest durations of RT recovery time were 10 and 160 min,
respectively, for the drinking events, and 5 and 130 min, respec-
tively, for the eating events.
Effects of ambient temperature and ingestion events on rumen
temperature
The influence of AT on the daily RT is presented in Fig. 5a, with
detailed information (i.e., pointwise slopes of the fitted model, the
identified AT range with a significant effect on RT, and the esti-
mated marginal effect of AT within this range) provided in
Table S1. Additionally, Figs. 6 and 7 illustrate the impact of individ-
ual drinking and eating events on RT fluctuations (RT change and
RT recovery time), with corresponding details in Tables S2 and
S3, respectively.
The daily RT declined approximately 0.0151 °C for every 1 °C
decrease in AT within the range of 9.91–17.25 °C(Fig. 5a;
P< 0.01). Regarding drinking events, both the quantity (drink size)
and temperature (FWT) influenced the RT change and recovery
time (P< 0.01). Consuming colder water resulted in a greater RT
drop by 0.0596 °C(P< 0.01; Fig. 6a) and in a longer RT recovery
time by 1.27 min (P< 0.01; Fig. 6b) for every 1 °C decrease in the
FWT. Each additional kg increase in drink size also increased the
magnitude of RT drop by 0.108 °C(P< 0.01; Fig. 6c) and the recov-
ery time by 2.13 min (P< 0.01; Fig. 6d). While the effect of FWT on
RT change was consistent across the entire AT range studied (2.31–
18.98 °C FWT), its impact on RT recovery time was only significant
between 12.34 and 14.52 °C AT. The effect of drink size on both RT
change and recovery time was observed between 2.00–12.0 kg and
2.00–14.0 kg, respectively, but the responses became more vari-
able with larger drink sizes.
In contrast to FWT, FT only tended to affect RT change (P= 0.07;
Fig. 7a) and did not affect RT recovery time (P= 0.32; Fig. 7b). How-
Fig. 3. Daily (mean ± SD) ambient temperature (green line) and relative humidity (yellow line) of the dairy cattle barn during the four non-consecutive data collection periods
conducted 18 d apart.
Fig. 4. Hourly variations in feed (as-fed basis) and free water intake of lactating dairy cows [least squared mean (LSM) ± SEM]. Values are expressed as % of the total daily
intake. Cows were fed at 0800 and 1800 h and milked at 0530 and 1530 h twice per day.
A.M. Serviento, T. He, X. Ma et al. Animal 18 (2024) 101209
6
ever, the meal size influenced both RT change (0.150 °C/kg FI;
P< 0.01; Fig. 7c) and RT recovery time (+3.71 min/kg FI;
P< 0.01; Fig. 7d). The effect of meal size on RT change was consis-
tent across the entire meal size range (0.35–7.27 kg), but its impact
on RT recovery time was observed in a narrower AT range and
exhibited a wider confidence interval when meal size exceeded
4.63 kg.
Effect of ambient temperature on intake and behavior
The effects of the AT on daily DMI, FWI, and TWI are illustrated
in Fig. 5,onMYinFig. 8, and on eating, drinking, and rumination
behaviors in Fig. 9, with corresponding details in Tables S1, S4,
and S5, respectively. Cows consumed more feed in colder temper-
atures, particularly below 10.63 °C AT, with DMI increasing by
approximately 0.365 kg/d for each 1 °C reduction in AT (P< 0.01;
Fig. 5b). The pointwise slopes revealed that this change in DMI
was greater with decreasing AT (e.g., 0.435 vs 0.201 kg DM/d per
1°C at 4.38 °C and 10.57 °C AT, respectively; Table S1). In the
AT range of 4.38–10.63 °C, neither meal bouts nor sizes were
affected, as demonstrated by a wide confidence interval of the
smooth function (Fig. 9c to 9d). However, cows had fewer meal
bouts in colder environment within the 12.53–17.25 °C AT range
(0.290 meal bouts/d per 1°C AT; P< 0.05; Fig. 9c). This increased
DMI did not lead to an improvement in MY, as the cows had lower
MY per DMI (0.0106 MY: DMI; P< 0.01) for every 1 °C decrease in
AT (Fig. 8). Cold ambient temperatures negatively impacted water
intake (Fig. 5c to 5f), with cows drinking less when AT was lower,
as evidenced by a decrease in FWI and TWI by 1.97 kg/d (P< 0.01)
and 2.44 kg/d (P< 0.01), respectively, per unit decrease in the AT
range from 17.25 to 8.94 or 8.72 and °C, respectively. Similar
effects were observed when water intake was expressed per DMI,
with cows drinking less per kg DMI by 0.0862 FWI: DMI
(P< 0.01) and 0.0873 kg TWI: DMI per unit decrease in the AT
range of 8.28 or 7.49–17.25 °C, respectively. While FWI and TWI
still continued to decrease below the aforementioned AT ranges,
the effect of AT was less significant as indicated by a wider confi-
dence interval. This negative impact of colder temperatures on
FWI was accompanied with fewer drinking bouts by 0.432 bout/d
per 1°CAT(P< 0.01) between 8.77 and 17.25 °CAT(Fig. 9a). Con-
versely, drink sizes increased by 0.100 kg/drink bout per 1°CAT
(P= 0.02) across the entire AT range tested (4.38–17.25 °C AT)
(Fig. 9b). Regarding rumination behavior, AT had no significant
effect on the number of rumination bouts (P= 0.57; Fig. 9e), but
the total rumination time was influenced by AT (P<0.01) espe-
cially within the AT range of 10.32–17.25 °C, where it decreased
by 5.31 min/day for each 1 °C drop in AT.
Discussion
This study aimed to assess the effects of AT on RT, and on drink-
ing and feeding behaviors of late-lactation dairy cows when the
ambient temperature was considerably lower than the body.
Although both AT and THI are valid indicators of an animal’s per-
ceived thermal environment (Bohmanova et al., 2007; Dikmen
and Hansen, 2009), the indoor location of the experimental barn
ensured a relatively constant RH. In addition, the temperatures of
ingested feed and water exhibited a close relationship to AT. There-
fore, AT was primarily used to depict the thermal environment,
especially since the results were similar whether AT or THI was
applied. The daily AT range in the present study (4.38–17.25 °C)
was still within the generally accepted thermoneutral zone for
dairy cows of around 5–20 °CAT(NASEM, 2021). However, the
AT range in this study was consistent with the objectives, being
at least 21–34 °C colder than the body temperature maintained
by dairy cows.
Effect of ambient temperature on reticulorumen temperature
The mean RT in this study (38.84 °C) was similar to previous
studies (AlZahal et al., 2008; Rutherford et al., 2019). However, this
value was lower when compared to mean RT calculated without
values associated with drinking activities (39.28–40.14 °C) as
reported by other studies (Bewley et al., 2008a; Liang et al.,
2013). It is well-documented that drinking activities can cause dis-
tinct fluctuations of RT (Ipema et al., 2008; Rutherford et al., 2019),
leading to significant decreases that may reach up to 8.5 °C(Dracy
and Kurtenbach, 1968; Bewley et al., 2008b). Our observations of
RT change following a drinking event agrees with previous findings
(Bhattacharya and Warner, 1968; Bewley et al., 2008b; Cantor
et al., 2018), with RT declining and then gradually recovering. Even
though there are no studies to our knowledge about the RT changes
immediately after eating events, the pattern of RT fluctuations
observed in this study was similar whether the cows were drinking
or eating. In previous studies with controlled water intake and
temperature, RT recovery times related to drinking activities varied
greatly from 20 to 103 min (Cunningham et al., 1964; Cantor et al.,
2018) and even up to an extrapolated value of 210 min (Bewley
Table 2
Summary statistics of ingestion events and of reticulorumen temperature parameters of the dairy cows
1
.
Item n
2
Mean Median Min Max SD
Drinking events
Free water temperature
3
,°C 490 12.9 13.5 2.31 19.0 3.60
Drink size
4
, kg 490 8.60 8.00 2.00 20.0 4.211
RT change
5
,°C 490 1.74 1.68 0.110 5.23 1.182
RT recovery time
6
, min 490 44.9 35.0 10.0 160.0 35.50
Eating events
Feed temperature
7
,°C 618 12.0 12.6 0.85 20.1 4.32
Meal size
8
, kg 618 2.68 2.38 0.35 7.27 1.549
RT change, °C 618 1.01 0.378 0.100 4.32 1.073
RT recovery time, min 618 34.5 25.0 5.000 130.0 25.55
1
Independent drinking and eating events were identified from 16 lactating dairy cows (eight Brown Swiss and eight Holstein Friesian) observed for four non-consecutive
6-d data collection periods during the autumn and winter seasons.
2
Number of observations included in the analyses.
3
Temperature of the ingested water at the time of a specific drinking event.
4
Amount of water ingested at a specific drinking event.
5
Difference between pre-event reticulorumen temperature (RT) and the postevent minimum RT.
6
Time elapsed from the time of the postevent minimum RT until the time at which the RT recovered to within ± 0.10 °C of the pre-event RT.
7
Temperature of the ingested feed at the time of a specific eating event.
8
Amount of feed ingested at a specific eating event.
A.M. Serviento, T. He, X. Ma et al. Animal 18 (2024) 101209
7
et al., 2008b). Our data, ranging from 10 to 160 min, follow this
trend. Discrepancies in these findings may be due to differences
in water intake recording systems and recovery time definitions.
In our study, we calculated RT recovery time from the postevent
minimum RT, not from the start of drinking event, owing to our
measurement system’s constraints in identifying the exact time
of drinking. Another distinction in our study was the uncontrolled
variables such as water temperature and intake. We included
observations with various free water and feed temperatures, and
drink and meal sizes, throughout the day in the analyses for each
event due to the animals’ unrestricted access to water and feed.
Therefore, we had a different pre-event or baseline RT for each
ingestion event unlike in previously mentioned studies with pre-
determined baseline RT value (> 38 °C). Nevertheless, our study
successfully demonstrated that both drinking and eating events
influence RT fluctuations, as reported in earlier research
(Bhattacharya and Warner, 1968; Bewley et al., 2008b; Cantor
et al., 2018).
Meanwhile, effects on RT changes and recovery times are influ-
enced by the temperature and amount of the ingesta of the drink-
ing and eating events. This effect was observed with FWT wherein
colder drinking water resulted in greater change in RT and longer
recovery time, which is consistent with earlier studies (Dracy
and Kurtenbach, 1968; Bewley et al., 2008b; Cantor et al., 2018).
Although eating events still caused a decrease in RT, feed temper-
ature per se did not affect RT recovery time and only tended to
affect RT change. This might be explained by the faster passage rate
of liquid fractions with large amounts of drinking water bypassing
the rumen (Garza et al., 1990; Zorrilla-Rios et al., 1990). Tempera-
tures of ingested free water possibly remained unchanged from
before ingestion until coming in direct contact with the reticuloru-
men. This was in contrast to feed that was chewed and swallowed,
Fig. 5. Effect of ambient temperature (AT, °C) on (a) reticulorumen temperature (°C); (b) DM intake (DMI, kg/d); (c) free water intake (FWI, kg/d); (d) FWI:DMI (kg FWI/kg
DMI); (e) total water intake (TWI, kg/d); and (f) TWI:DMI (kg TWI/kg DMI) of dairy cows as fitted in a generalized additive mixed model. The shaded regions represent the 95%
confidence interval for each point of the smooth. Temperatures between the two red horizontal lines represent the AT range wherein the model’s estimated pointwise slope is
significantly different from zero (P< 0.05).
A.M. Serviento, T. He, X. Ma et al. Animal 18 (2024) 101209
8
and was likely already warmer than the feed’s pre-ingestion tem-
perature before reaching the reticulorumen. Therefore, FT may
have been all similar upon reaching the reticulorumen regardless
of its temperature before ingestion, partly explaining the lack of
FT effect on RT parameters.
Meanwhile, for the amount of ingesta, both drink and meal
sizes positively affected the degree of RT drop and the length of
RT recovery time. While there are no studies to our knowledge
to compare the results related to the effect of meal sizes on RT,
our results regarding drink size agree with previous studies
(Bewley et al., 2008b; Cantor et al., 2018) wherein larger drink
sizes cause more substantial drops in RT. However, the effect of
the amount of ingesta on RT changes may also depend on their
temperature. As Bewley et al. (2008b) observed, large quantities
of water intake only have minimal impact on RT drop and recov-
ery time when provided at 38.9 °C. Indeed, there would be zero
net heat exchange between the reticulorumen and the ingesta
if they have similar temperatures. Previous studies (Petersen
et al., 2016; Grossi et al., 2021) have already demonstrated how
cows with access to heated drinking water during winter had
fewer instances of RT dropping below 37–38 °Ccomparedto
those with access to only cold or ambient temperature water.
This indicates that only when feed or drinking water is colder
than body temperature will it cause significant disturbances to
the RT. This was the case in our study as FWT ranged from 2.31
to 19.0 °C,andFTfrom0.85to20.1°C (i.e., around 18–38 °C
colder than the RT). This confirms our hypothesis that colder sea-
sons can be detrimental in maintaining a stable rumen environ-
ment caused by ingestion of free water and feed that are colder
than the body.
Previous studies have associated prolonged exposure to cold
and reduced RT with decreased rumen functions (Christopherson,
1976; Brod et al., 1982; Kennedy, 1985). An in vitro study also
demonstrated that decreasing RT from 39 °Cto35°C led to a
reduction in microbial diversity and richness, resulting in lower
DM digestibility and methane production (Duarte et al., 2017).
However, we did not observe an impact of daily RT on nutrient
digestibility or on the methane production of cows in the present
study (Fig. S1). As previously discussed, the effects of drinking
and eating events on RT were transient, and RT consistently strived
to recover and maintain a stable temperature. Aside from the cur-
rent study not being designed to evaluate the effect of AT or RT on
nutrient digestibility specifically, the AT range along with the asso-
ciated FWT and FT may not have been cold enough to illicit notable
effects on rumen functions. Nevertheless, our results suggest that
providing warmer or heated drinking water might be a way to
minimize disruptions in the reticulorumen environment during
colder seasons.
Ambient temperature not only contributed to the fluctuations
in RT but also had an overall effect on daily RT. In this study, the
daily RT in cows decreased with colder temperatures, at a rate of
0.0151 °C per degree of AT, within the range of 9.91–17.25 °C AT.
The effect of AT on RT was less pronounced below this range, which
can be partially attributed to the observed increase in DMI below
10.63 °C that could have resulted in greater heat production in
the rumen, thereby offsetting the effect of AT on RT. This is in con-
trast with findings by Liang et al. (2013), who reported stable daily
RT in cows from 5 to around 23 °C AT. However, in their study, RT
data potentially influenced by drinking events were excluded from
analysis, explaining the discrepancy with the present study. While
Fig. 6. Effect of free water temperature (FWT, °C; a and b) and of drink size (kg; c and d) on reticulorumen temperature (RT) change (°C; a and c) and RT recovery time (min; b
and d) of dairy cows as fitted in a generalized additive mixed model. The shaded regions represent the 95% confidence interval for each point of the smooth. Temperatures or
drink sizes between the two red vertical lines represent the range of FWT or drink size wherein the model’s estimated pointwise slope is significantly different from zero
(P< 0.05). Abbreviation: FWI = free water intake.
A.M. Serviento, T. He, X. Ma et al. Animal 18 (2024) 101209
9
many studies have suggested removing drinking activities from
daily RT analysis due to its impact on RT fluctuations (Ipema
et al., 2008; Bewley et al., 2008a), it was important to include these
drinking and eating events given the objective of the current study
and the close relationships between AT, FWT, and FT. Even when
considering only drinking bouts, these events already accounted
for at least 9 h per day of data with the cows in this study having
12 drinking bouts/d with an average RT recovery time of 45 min
for each event. Thus, removing them would ignore the discussed
effects of drinking and eating activities on RT when FWT and FT
are below body temperature.
Effect of ambient temperature on feed and water intake
Cows in this study were affected by their thermal environment
especially when the AT dropped below 10.63 °C (< 52 THI). Even
though still above the accepted lower critical zone of the ther-
moneutral zone, the cows increased their DMI by 0.365 kg/d per
degree of AT drop (equivalent to 0.24 kg/d increase per unit
decrease in THI). This value is greater than the relative change
reported by NRC (1981) which was + 0.20 kg/d between 5 and
10 °C (or + 0.04 kg/d per 1°C AT). However, this reported value
was an average across several days and may have underestimated
Fig. 7. Effect of feed temperature (FT, °C; a and b) and of meal size (kg; c and d) on reticulorumen temperature (RT) change (°C; a and c) and RT recovery time (min; b and d) of
dairy cows as fitted in a generalized additive mixed model. The shaded regions represent the 95% confidence interval for each point of the smooth. Temperatures or drink
sizes between the two red vertical lines represent the range of FT or meal size wherein the model’s estimated pointwise slope is significantly different from zero (P< 0.05).
Abbreviation: FI = feed intake.
Fig. 8. Effect of ambient temperature (AT, °C) on (a) milk yield (kg/d) and (b) on feed efficiency [milk yield (MY): DM intake (DMI)] of dairy cows as fitted in a generalized
additive mixed model. The shaded regions represent the 95% confidence interval for each point of the smooth. Ambient temperatures between the two red horizontal lines
represent the AT range wherein the model’s estimated pointwise slope is significantly different from zero (P< 0.05).
A.M. Serviento, T. He, X. Ma et al. Animal 18 (2024) 101209
10
the effect of daily AT changes. Indeed, the effect of AT on DMI in
our study was greater at colder AT (+0.435 vs + 0.201 kg DM/d
per 1°C AT at 4.38 °C and 10.57 °C AT, respectively; Table S1).
The increased DMI as a consequence of cold exposure may be
related to previous reports of increased ruminal passage rate
and/or to a greater maintenance energy requirement to maintain
body temperature in cold conditions (Christopherson and
Kennedy, 1983; Young, 1983; Fox and Tylutki, 1998). It also agrees
with the long-established concept that changes in feed intake are
thermoregulatory responses, depending on the animal’s need for
heat in order to maintain a stable body temperature (Brobeck,
1997). The increased energy needs with lower AT is supported by
the decrease in feed efficiency of the cows in the present study
by 0.0106 MY:DMI per 1°C AT. Based on these results, a cow
producing 25 kg/d of milk would need to eat around 2.65 kg more
DM/d when AT is at 5 °C compared to when it is at 15 °C. Other
studies have reported similar results on lower productivity despite
increased feed intake during colder conditions (Brouc
ˇek et al.,
1991; Kang et al., 2016; Wang et al., 2023). Our findings suggest
that, for the cows in this study, AT below 10 °C may already be
at the lower critical zone of their thermoneutral zone as their
increased energy intake is already being used for increasing heat
production and maintaining body temperature homeostasis, rather
than for milk production. However, it must be noted that the cows
in the present study were already in late lactation and had rela-
tively low production level (average of 224 days in milk and of
24.8 kg MY/d, respectively). This can indicate a relatively low
metabolic heat production and might explain their need for higher
DMI at only 10 °C AT.
Water intake of the cows in the present study was negatively
affected by colder temperatures. The FWI change of 1.97 kg/d
per degree decrease in AT is slightly greater than values reported
Fig. 9. Effect of the ambient temperature (AT, °C) on (a) drink bout frequency (n/d); (b) drink size (kg/drink bout); (c) meal bout frequency (n/d); (d) meal size (kg/meal bout);
(e) rumination bout frequency (n/d); and (f) total rumination time (min/d) of dairy cows as fitted in a generalized additive mixed model. The shaded regions represent the 95%
confidence interval for each point of the smooth. Ambient temperatures between the two red horizontal lines represent the AT range wherein the model’s estimated
pointwise slope is significantly different from zero (P< 0.05).
A.M. Serviento, T. He, X. Ma et al. Animal 18 (2024) 101209
11
in other studies ranging from 0.76 to 1.52 kg/d FWI (Stockdale
and King, 1983; Meyer et al., 2004; Appuhamy et al., 2016). How-
ever, the effect of AT on FWI might be confounded with its effect on
DMI (Appuhamy et al., 2016). When expressed per kg DMI, the
cows in this study had an average of 3.54 FWI:DMI in line with val-
ues reported by NRC (1981). Similar to the trend on FWI, FWI:DMI
decreased with colder AT, which can be associated with changes in
the body’s evaporative losses (Khelil-Arfa et al., 2014). Since the
main drinking motivation of animals is the need to maintain body
water homeostasis (McKinley and Johnson, 2004; McKinley et al.,
2009), a reduced water intake suggests limited evaporative heat
losses which is a classic response in cold conditions in order to con-
serve more body heat (Diesel et al., 1990). The AT had a weaker
effect on both free and total water intake < 8 °C (less steep slope),
which might be explained by the previously discussed increasing
DMI when AT was below 10.63 °C. The cows may have been pro-
ducing more body heat from the thermic effect of feed and thus
no longer needed to limit their evaporative losses as much even
though AT was still decreasing.
Ingested water below body temperature can serve as a heat
sink, with cows losing theoretically 4.184 kJ per kg of ingested
water that is 1 °C lower than the body temperature (specific heat
of water). This increased energy requirement caused by consuming
ambient temperature water during cold seasons can thereby par-
tially explain the increased DMI in colder environments. This
agrees with Bhattacharya and Warner (1968), who reported a
24% increase in feed intake when rumen was infused with 5 vs
30 °C water, and with Petersen et al. (2016) on how consumption
of water at temperatures less than the body temperature can
increase energy needs of cattle. Therefore, the negative effect of
cold environment on FWI may not only be related to the daily AT
per se but also to the temperature of drinking water provided to
animals. While this cannot be confirmed since drinking water tem-
perature was not controlled in the present study, other studies
have reported that provision of heated water (>30 vs < 15 °C)
increased FWI in cattle (Osborne et al., 2002; Petersen et al.,
2016). Provision of heated water during winter has also been
reported to improve ruminal pH and temperature stability
(Petersen et al., 2016; Grossi et al., 2021), which can be beneficial
to the cows, especially given the reported negative impact of RT
fluctuations on rumen microbial activity (Duarte et al., 2017),
and the increased energy costs related to the consumption of cold
water. Nevertheless, our results show that the detrimental effects
of low AT are intrinsically linked to the low temperatures of
ingested free water and feed during colder seasons. This can fur-
ther increase the energy needs of the animal during colder seasons,
with potential consequences on feed efficiency.
Effect of ambient temperature on animal behavior
The cows in our study had an average of 11.8 meal bouts/d, sim-
ilar to previous reports for cows housed in a tie-stall barn having
10–13 meal bouts/d (Dado and Allen,1993; Niu et al., 2017). Ambi-
ent temperature increased the number of meal bouts of cows when
AT was between 12.53 and 17.25 °C, while meal size seemed unaf-
fected by AT. While DMI increased at AT lower than 10.63 °C, this
effect was not as clearly associated with the feeding behavior of
cows in our study. This discrepancy might be due to fewer obser-
vations in the feeding behavior data compared with DMI data.
Additionally, using average daily AT may not capture the immedi-
ate effect of AT on feeding behavior of cows since AT is not con-
stant throughout the day. Looking at individual eating events
(Fig. S2b), it is evident that cows had larger meal sizes when FT
(also based on AT at the time of eating) was lower. Therefore, cows
might eat less frequently but with larger meal sizes as ambient and
feed temperatures decrease. Previous studies reported more fre-
quent meal bouts and smaller meal sizes with higher AT (Shiao
et al., 2011; Eslamizad et al., 2015), which aligns with our findings
(>12 °C AT). Studies on the effect of cold exposure on feeding
behavior are limited, but some have shown increased time spent
eating in colder temperatures (Fu et al., 2022; Wang et al., 2023).
Further research might be needed to elucidate the effect of cold
temperatures on feeding behavior.
Regarding drinking behavior, the previously discussed decline
in FWI in response to the decreasing AT corresponded with fewer
drinking bouts but larger drink sizes in our study. The effect of
lower AT on increasing drink size was clearly observed on both a
daily level (Fig. 8b) and an individual event level (Fig. S2a). This
trend is similar to the pattern of larger meal sizes when FT was
colder. As shown in the current and in previous studies (Nocek
and Braund, 1985; Andersson, 1987), drinking activities are highly
correlated with eating activities. A behavioral study reported that
most eating bouts were followed by a drinking bout within
15 min, and half of the cows drank water in between their eating
bouts (Miller and Wood-Gush, 1991). This is consistent with our
results, where 44% of eating activities coincided with a drinking
bout. Moreover, meal sizes of dairy cows have also been reported
to decrease when water intake was restricted (Burgos et al.,
2001), further demonstrating the positive relationship between
free water and feed intake even on a meal level. This increase in
drink size with a larger meal size may be due to a reduced ruminal
osmolality which can stimulate water intake, or to increased water
needs for facilitating ingestion via saliva which dairy cows excrete
at large quantities during feeding and/or chewing (Bailey and
Balch, 1961; Allen, 1997; Beauchemin et al., 2008).
In the present study, rumination time decreased with colder
temperatures but plateaued when the AT was below 10.32 °C, pos-
sibly related to the negative effect of AT on water intake which has
been reported to negatively influence rumination (Gordon, 1965).
A review on the link between rumen function and feeding behavior
found that ingestion of larger meals (as observed in cows when FT
is lower) and a decrease in meal frequency could negatively impact
rumination time and salivary secretion (Gonzalez-Rivas et al.,
2018). While the current study did not aim to determine the rela-
tionships between these parameters, the results still demonstrated
that AT below 17 °C, along with their subsequent effects on FWT
and FT, can induce changes in drinking, eating, and even rumina-
tion behaviors of dairy cows.
Conclusion
Cold exposure can illicit thermoregulatory responses such as
increased feed intake, reduced water intake, and alterations in
feeding, drinking and rumination behaviors of late-lactation dairy
cows with an average production of 25 kg MY/d. The present
study also demonstrated how drinking and eating events con-
tribute to RT fluctuations, specifically to the magnitude of RT
change and length of RT recovery time. Furthermore, our data sug-
gest that the effect of cold exposure may be partially exacerbated
by the ingestion of feed and free water at temperatures signifi-
cantly lower than the body. These effects may have subsequent
consequences on feed efficiency, likely due to the additional ener-
getic demands placed on the animal.
Supplementary material
Supplementary material to this article can be found online at
https://doi.org/10.1016/j.animal.2024.101209.
A.M. Serviento, T. He, X. Ma et al. Animal 18 (2024) 101209
12
Ethics approval
The study was approved by the Cantonal Veterinary Office of
Zürich (license no. ZH207/2021), conforming to the Swiss legisla-
tion on animal experimentation.
Data and model availability statement
None of the data were deposited in an official repository. Data
and models are available from the authors upon request.
Declaration of Generative AI and AI-assisted technologies in the
writing process
During the preparation of this work the author(s) did not use
any AI and AI-assisted technologies.
Author ORCIDs
Aira Maye Serviento: https://orcid.org/0000-0002-2166-0791.
Tengfei He: https://orcid.org/0000-0001-5643-2062.
Xiaoqi Ma: https://orcid.org/0000-0002-1652-9035.
Susanna Räisänen: https://orcid.org/0000-0001-9199-7026.
Mutian Niu: https://orcid.org/0000-0003-4484-4611.
CRediT authorship contribution statement
A.M. Serviento: Writing – review & editing, Writing – original
draft, Methodology, Formal analysis, Data curation, Conceptualiza-
tion. T. He: Writing – original draft, Investigation, Formal analysis,
Data curation, Conceptualization. X. Ma: Writing – review & edit-
ing, Investigation, Formal analysis, Data curation. S.E. Räisänen:
Writing – review & editing, Investigation, Conceptualization. M.
Niu: Writing – review & editing, Methodology, Funding acquisition,
Conceptualization.
Declaration of interest
None.
Acknowledgements
The authors would like to thank the barn staff and the research
team of AgroVet-Strickhof (Lindau, Switzerland) for their support,
technical assistance and care of the experimental cows. We would
also like to acknowledge the members of the Animal Nutrition
group at ETH Zürich (Zürich, Switzerland) for their assistance in
sample collections, and the lab technicians, led by Carmen Kunz,
for lab analyses.
Financial support statement
This research received no specific grant from any funding
agency, commercial or not-for-profit section.
References
Allen, M.S., 1997. Relationship between fermentation acid production in the rumen
and the requirement for physically effective fiber. Journal of Dairy Science 80,
1447–1462. https://doi.org/10.3168/jds.S0022-0302(97)76074-0.
AlZahal, O., Kebreab, E., France, J., Froetschel, M., McBride, B.W., 2008. Ruminal
temperature may aid in the detection of subacute ruminal acidosis. Journal of
Dairy Science 91, 202–207. https://doi.org/10.3168/jds.2007-0535.
Ammer, S., Lambertz, C., Gauly, M., 2016. Comparison of different measuring
methods for body temperature in lactating cows under different climatic
conditions. Journal of Dairy Research 83, 165–172. https://doi.org/10.1017/
S0022029916000182.
Andersson, M., 1987. Effects of number and location of water bowls and social rank
on drinking behaviour and performance of loose-housed dairy cows. Applied
Animal Behaviour Science 17, 19–31. https://doi.org/10.1016/0168-1591(87)
90004-9.
Appuhamy, J.A.D.R.N., Judy, J.V., Kebreab, E., Kononoff, P.J., 2016. Prediction of
drinking water intake by dairy cows. Journal of Dairy Science 99, 7191–7205.
https://doi.org/10.3168/jds.2016-10950.
Bailey, C.B., Balch, C.C., 1961. Saliva secretion and its relation to feeding in cattle: 1.
the composition and rate of secretion of parotid saliva in a small steer. British
Journal of Nutrition 15, 371–382. https://doi.org/10.1079/BJN19610047.
Bates, D., Mächler, M., Bolker, B., Walker, S., 2015. Fitting linear mixed-effects
models using lme4. Journal of Statistical Software 67, 1–48. https://doi.org/
10.18637/jss.v067.i01.
Beauchemin, K.A., Eriksen, L., Nørgaard, P., Rode, L.M., 2008. Short communication:
salivary secretion during meals in lactating dairy cattle. Journal of Dairy Science
91, 2077–2081. https://doi.org/10.3168/jds.2007-0726.
Bernabucci, U., Lacetera, N., Baumgard, L.H., Rhoads, R.P., Ronchi, B., Nardone, A.,
2010. Metabolic and hormonal acclimation to heat stress in domesticated
ruminants. Animal 4, 1167–1183. https://doi.org/10.1017/
S175173111000090X.
Bewley, J.M., Einstein, M.E., Grott, M.W., Schutz, M.M., 2008a. Comparison of
reticular and rectal core body temperatures in lactating dairy cows. Journal of
Dairy Science 91, 4661–4672. https://doi.org/10.3168/jds.2007-0835.
Bewley, J.M., Grott, M.W., Einstein, M.E., Schutz, M.M., 2008b. Impact of intake
water temperatures on reticular temperatures of lactating dairy cows. Journal
of Dairy Science 91, 3880–3887. https://doi.org/10.3168/jds.2008-1159.
Bhattacharya, A.N., Warner, R.G., 1968. Influence of varying rumen temperature on
central cooling or warming and on regulation of voluntary feed intake in dairy
cattle. Journal of Dairy Science 51, 1481–1489. https://doi.org/10.3168/jds.
S0022-0302(68)87216-9.
Bohmanova, J., Misztal, I., Cole, J.B., 2007. Temperature-humidity indices as
indicators of milk production losses due to heat stress. Journal of Dairy
Science 90, 1947–1956. https://doi.org/10.3168/jds.2006-513.
Brobeck, J.R., 1997. Food intake as a mechanism of temperature regulation. Obesity
Research 5, 641–645. https://doi.org/10.1002/j.1550-8528.1997.tb00589.x.
Brod, D.L., Bolsen, K.K., Brent, B.E., 1982. Effect of water temperature in rumen
temperature, digestion and rumen fermentation in sheep. Journal of Animal
Science 54, 179–182. https://doi.org/10.2527/jas1982.541179x.
Brouc
ˇek, J., Letkovic
ˇová, M., Kovalc
ˇuj, K., 1991. Estimation of cold stress effect on
dairy cows. International Journal of Biometeorology 35, 29–32. https://doi.org/
10.1007/BF01040960.
Burgos, M.S., Senn, M., Sutter, F., Kreuzer, M., Langhans, W., 2001. Effect of water
restriction on feeding and metabolism in dairy cows. American Journal of
Physiology-Regulatory, Integrative and Comparative Physiology 280, R418–
R427. https://doi.org/10.1152/ajpregu.2001.280.2.R418.
Cantor, M.C., Costa, J.H.C., Bewley, J.M., 2018. Impact of observed and controlled
water intake on reticulorumen temperature in lactating dairy cattle. Animals 8,
194. https://doi.org/10.3390/ani8110194.
Christopherson, R.J., 1976. Effects of prolonged cold and the outdoor winter
environment on apparent digestibility in sheep and cattle. Canadian Journal of
Animal Science 56, 201–212. https://doi.org/10.4141/cjas76-024.
Christopherson, R.J., Kennedy, P.M., 1983. Effect of the thermal environment on
digestion in ruminants. Canadian Journal of Animal Science 63, 477–496.
https://doi.org/10.4141/cjas83-058.
Cohen, J., Agel, L., Barlow, M., Garfinkel, C.I., White, I., 2021. Linking Arctic variability
and change with extreme winter weather in the United States. Science 373,
1116–1121. https://doi.org/10.1126/science.abi9167.
Collier, R.J., Gebremedhin, K.G., 2015. Thermal biology of domestic animals. Annual
Review of Animal Biosciences 3, 513–532. https://doi.org/10.1146/annurev-
animal-022114-110659.
Cunningham, M.D., Martz, F.A., Merilan, C.P., 1964. Effect of drinking-water
temperature upon ruminant digestion, intraruminal temperature, and water
consumption of nonlactating dairy cows. Journal of Dairy Science 47, 382–385.
https://doi.org/10.3168/jds.S0022-0302(64)88671-9.
Dado, R.G., Allen, M.S., 1993. Continuous computer acquisition of feed and water
intakes, chewing, reticular motility, and ruminal pH of cattle. Journal of Dairy
Science 76, 1589–1600. https://doi.org/10.3168/jds.S0022-0302(93)77492-5.
Diesel, D.A., Tucker, A., Robertshaw, D., 1990. Cold-induced changes in breathing
pattern as a strategy to reduce respiratory heat loss. Journal of Applied
Physiology 69, 1946–1952. https://doi.org/10.1152/jappl.1990.69.6.1946.
Dikmen, S., Hansen, P.J., 2009. Is the temperature-humidity index the best indicator
of heat stress in lactating dairy cows in a subtropical environment? Journal of
Dairy Science 92, 109–116. https://doi.org/10.3168/jds.2008-1370.
Dracy, A.E., Kurtenbach, A.J., 1968. Temperature change within the rumen, crop
area, and rectal area when liquid of various temperatures was fed to calves.
Journal of Dairy Science 51, 1787–1790. https://doi.org/10.3168/jds.S0022-
0302(68)87278-9.
Duarte, A.C., Holman, D.B., Alexander, T.W., Kiri, K., Breves, G., Chaves, A.V., 2017.
Incubation temperature, but not Pequi oil supplementation, affects methane
production, and the ruminal microbiota in a rumen simulation technique
(Rusitec) system. Frontiers in Microbiology 8, 1076. https://doi.org/10.3389/
fmicb.2017.01076.
Eslamizad, M., Lamp, O., Derno, M., Kuhla, B., 2015. The control of short-term feed
intake by metabolic oxidation in late-pregnant and early lactating dairy cows
A.M. Serviento, T. He, X. Ma et al. Animal 18 (2024) 101209
13
exposed to high ambient temperatures. Physiology & Behavior 145, 64–70.
https://doi.org/10.1016/j.physbeh.2015.03.044.
Fox, D.G., Tylutki, T.P., 1998. Accounting for the effects of environment on the
nutrient requirements of dairy cattle. Journal of Dairy Science 81, 3085–3095.
https://doi.org/10.3168/jds.S0022-0302(98)75873-4.
Fu, X., Zhang, Y., Zhang, Y.G., Yin, Y.L., Yan, S.C., Zhao, Y.Z., Shen, W.Z., 2022. Research
and application of a new multilevel fuzzy comprehensive evaluation method for
cold stress in dairy cows. Journal of Dairy Science 105, 9137–9161. https://doi.
org/10.3168/jds.2022-21828.
Garza, J.D., Zorrilla-Rios, J., Owens, F.N., 1990. Ruminal water evasion and steady
state. Animal Science Research Report MP-129, 114–121.
Gonzalez-Rivas, P.A., Sullivan, M., Cottrell, J.J., Leury, B.J., Gaughan, J.B., Dunshea, F.
R., 2018. Effect of feeding slowly fermentable grains on productive variables and
amelioration of heat stress in lactating dairy cows in a sub-tropical summer.
Tropical Animal Health and Production 50, 1763–1769. https://doi.org/10.1007/
s11250-018-1616-5.
Gordon, J.G., 1965. The effect of water deprivation upon the rumination behaviour
of housed sheep. The Journal of Agricultural Science 64, 31–35. https://doi.org/
10.1017/S0021859600010443.
Grossi, S., Rossi, L., Dell’Anno, M., Biffani, S., Rossi, C.A.S., 2021. Effects of heated
drinking water on the growth performance and rumen functionality of fattening
charolaise beef cattle in winter. Animals: an Open Access Journal from MDPI 11,
2218. https://doi.org/10.3390/ani11082218.
He, T., Serviento, A.M., Ma, X., Räisänen, S.E., Niu, M. 2023. The effects of water
temperature and water intake on rumen temperature of dairy cows. Journal
of Dairy Science 106 (Suppl. 1), 355. (Abstr.).
Ipema, A.H., Goense, D., Hogewerf, P.H., Houwers, H.W.J., van Roest, H., 2008. Pilot
study to monitor body temperature of dairy cows with a rumen bolus.
Computers and Electronics in Agriculture 64, 49–52. https://doi.org/10.1016/
j.compag.2008.05.009.
Kadzere, C.T., Murphy, M.R., Silanikove, N., Maltz, E., 2002. Heat stress in lactating
dairy cows: a review. Livestock Production Science 77, 59–91. https://doi.org/
10.1016/S0301-6226(01)00330-X.
Kang, H.J., Lee, I.K., Piao, M.Y., Gu, M.J., Yun, C.H., Kim, H.J., Kim, K.H., Baik, M., 2016.
Effects of ambient temperature on growth performance, blood metabolites, and
immune cell populations in korean cattle steers. Asian-Australasian Journal of
Animal Sciences 29, 436–443. https://doi.org/10.5713/ajas.15.0937.
Kennedy, P.M., 1985. Influences of cold exposure on digestion of organic matter,
rates of passage of digesta in the gastrointestinal tract, and feeding and
rumination behaviour in sheep given four forage diets in the chopped, or
ground and pelleted form. British Journal of Nutrition 53, 159–173. https://doi.
org/10.1079/BJN19850020.
Khelil-Arfa, H., Faverdin, P., Boudon, A., 2014. Effect of ambient temperature and
sodium bicarbonate supplementation on water and electrolyte balances in dry
and lactating Holstein cows. Journal of Dairy Science 97, 2305–2318. https://
doi.org/10.3168/jds.2013-7079.
Larsen, B., Bourque, J., Moore, T.M., Adebimpe, A., Calkins, M.E., Elliott, M.A., Gur, R.
C., Gur, R.E., Moberg, P.J., Roalf, D.R., Ruparel, K., Turetsky, B.I., Vandekar, S.N.,
Wolf, D.H., Shinohara, R.T., Satterthwaite, T.D., 2020. Longitudinal development
of brain iron is linked to cognition in youth. The Journal of Neuroscience 40,
1810–1818. https://doi.org/10.1523/JNEUROSCI.2434-19.2020.
Liang, D., Wood, C.L., McQuerry, K.J., Ray, D.L., Clark, J.D., Bewley, J.M., 2013.
Influence of breed, milk production, season, and ambient temperature on dairy
cow reticulorumen temperature. Journal of Dairy Science 96, 5072–5081.
https://doi.org/10.3168/jds.2012-6537.
Linville, K., Casper, D.P., Osorio, J.S., 2017. 377 In vitro analysis of rumen microbial
fermentation at different temperatures. Journal of Animal Science 95, 182–183.
https://doi.org/10.2527/asasmw.2017.377.
Ma, X., Räisänen, S.E., He, T., Islam, M.Z., Li, Y., Peng, R., Reichenbach, M., Sun, X.,
Wang, K., Yang, S., Zeng, Z., Müller, I., Niu, M., 2023. Combined effects of 3-
nitrooxypropanol (3-NOP, Bovaer10) and whole cottonseed on production and
enteric methane emissions of dairy cows. Journal of Dairy Science 106 (Suppl 1),
355.
Mader, T.L., Davis, M.S., Brown-Brandl, T., 2006. Environmental factors influencing
heat stress in feedlot cattle. Journal of Animal Science 84, 712–719. https://doi.
org/10.2527/2006.843712x.
McKinley, M.J., Johnson, A.K., 2004. The physiological regulation of thirst and fluid
intake. Physiology 19, 1–6. https://doi.org/10.1152/nips.01470.2003.
McKinley, M.J., Weissenborn, F., Mathai, M.L., 2009. Drinking-induced
thermoregulatory panting in rehydrated sheep: influences of oropharyngeal/
esophageal signals, core temperature, and thirst satiety. American Journal of
Physiology. Regulatory, Integrative and Comparative Physiology 296, R1881–
R1888. https://doi.org/10.1152/ajpregu.90890.2008.
Meyer, U., Everinghoff, M., Gädeken, D., Flachowsky, G., 2004. Investigations on the
water intake of lactating dairy cows. Livestock Production Science 90, 117–121.
https://doi.org/10.1016/j.livprodsci.2004.03.005.
Miller, K., Wood-Gush, D.G.M., 1991. Some effects of housing on the social
behaviour of dairy cows. Animal Science 53, 271–278. https://doi.org/
10.1017/S0003356100020262.
NASEM. 2021. National Academies of Sciences, Engineering, and Medicine. 2021.
Nutrient Requirements of Dairy Cattle: Eighth Revised Edition. Washington, DC:
The National Academies Press. National Academies of Sciences, Engineering,
and Medicine. 2021. Nutrient Requirements of Dairy Cattle: Eighth Revised
Edition. The National Academies Press, Washington, DC, USA. https://doi.org/10.
17226/25806.
Nienaber, J.A., Hahn, G.L., 2007. Livestock production system management
responses to thermal challenges. International Journal of Biometeorology 52,
149–157. https://doi.org/10.1007/s00484-007-0103-x.
Niu, M., Ying, Y., Bartell, P.A., Harvatine, K.J., 2017. The effects of feeding rations that
differ in fiber and fermentable starch within a day on milk production and the
daily rhythm of feed intake and plasma hormones and metabolites in dairy
cows. Journal of Dairy Science 100, 187–198. https://doi.org/10.3168/jds.2016-
11129.
Nocek, J.E., Braund, D.G., 1985. Effect of feeding frequency on diurnal dry matter and
water consumption, liquid dilution rate, and milk yield in first lactation. Journal
of Dairy Science 68, 2238–2247. https://doi.org/10.3168/jds.S0022-0302(85)
81096-1.
NRC, 1981. Effect of environment on nutrient requirements of domestic animals.
National Academy Press, National Academies Press, Washington, DC, USA.
Osborne, V.R., Hacker, R.R., McBride, B.W., 2002. Effects of heated drinking water on
the production responses of lactating Holstein and Jersey cows. Canadian
Journal of Animal Science 82, 267–273. https://doi.org/10.4141/A01-055.
Petersen, M.K., Muscha, J.M., Mulliniks, J.T., Roberts, A.J., 2016. Water temperature
impacts water consumption by range cattle in winter. Journal of Animal Science
94, 4297–4306. https://doi.org/10.2527/jas.2015-0155.
R Core Team, 2023. R: A language and environment for statistical computing. R
Foundation for Statistical Computing, Vienna, Austria.
Rutherford, N.H., Gordon, A.W., Lively, F.O., Arnott, G., 2019. The effect of behaviour
and diet on the rumen temperature of holstein bulls. Animals: an Open Access
Journal from MDPI 9, 1000. https://doi.org/10.3390/ani9111000.
Shiao, T.F., Chen, J.C., Yang, D.W., Lee, S.N., Lee, C.F., Cheng, W.T.K., 2011. Feasibility
assessment of a tunnel-ventilated, water-padded barn on alleviation of heat
stress for lactating Holstein cows in a humid area. Journal of Dairy Science 94,
5393–5404. https://doi.org/10.3168/jds.2010-3730.
Simpson, G. 2023._gratia: Graceful ggplot-Based Graphics and Other Functions for
GAMs Fitted using mgcv_. R package version 0.8.1. Retrieved on 11 July 2023,
from https://gavinsimpson.github.io/gratia/.
Stockdale, C.R., King, K.R., 1983. A note on some of the factors that affect the water
consumption of lactating dairy cows at pasture. Animal Science 36, 303–306.
https://doi.org/10.1017/S1357729800001375.
Tao, S., Orellana Rivas, R.M., Marins, T.N., Chen, Y.-C., Gao, J., Bernard, J.K., 2020.
Impact of heat stress on lactational performance of dairy cows. Theriogenology
150, 437–444. https://doi.org/10.1016/j.theriogenology.2020.02.048.
Wang, S., Li, Q., Peng, J., Niu, H., 2023. Effects of long-term cold stress on growth
performance, behavior, physiological parameters, and energy metabolism in
growing beef cattle. Animals 13, 1619. https://doi.org/10.3390/ani13101619.
Wood, S.N., 2011. Fast stable restricted maximum likelihood and marginal
likelihood estimation of semiparametric generalized linear models. Journal of
the Royal Statistical Society. Series B (statistical Methodology) 73, 3–36.
Wood, S.N. 2017. Generalized Additive Models: An Introduction with R, 2nd edition.
Chapman and Hall/CRC, New York, NY, USA. https://doi.org/10.1201/
9781315370279.
Young, B.A., 1983. Ruminant cold stress: effect on production. Journal of Animal
Science 57, 1601–1607. https://doi.org/10.2527/jas1983.5761601x.
Zehner, N., Umstätter, C., Niederhauser, J.J., Schick, M., 2017. System specification
and validation of a noseband pressure sensor for measurement of ruminating
and eating behavior in stable-fed cows. Computers and Electronics in
Agriculture 136, 31–41. https://doi.org/10.1016/j.compag.2017.02.021.
Zorrilla-Rios, J., Garza, J.D., Owens, F.N., 1990. Fate of drinking water in ruminants:
simultaneous comparison of two methods to estimate ruminal evasion. Animal
Science Research Report 1990, 167–169.
A.M. Serviento, T. He, X. Ma et al. Animal 18 (2024) 101209
14
