se 1699 ms

7/9/01

7:31 AM

Page 387

BEHAVIOR OF DAIRY COWS IN RESPONSE TO

DIFFERENT BARN COOLING SYSTEMS
E. Frazzi, L. Calamari, F. Calegari, L. Stefanini
ABSTRACT. Forty-two dairy cows were subdivided in three groups of 14 cows each. Each pen of cows was provided with a
feeding area, a freestall area, and an external paddock. The study was conducted in an experimental barn located in
the Po Valley of Italy. The cows were observed during summer 1995 for milk yield and quality and for their behavior
(movements and residence time in the different areas of each pen) under different microclimatic conditions (temperature,
relative humidity, light exposure, air speed). The feeding and freestall areas were arranged with: (1) fans in the first pen;
(2) fans and misting in the second pen; and 3) the third pen was used as control. During the summer months (from June to
September) microclimatic conditions, milk yield, and cow position (monitored with video cameras and automatic still
cameras placed in different areas of the barn) were recorded. During the hottest period, lower milk yield reductions were
recorded for the cows housed in pens with fans or with fans and misting. This result was consistent with the behavior of
the cows inside experimental pens that spent longer periods of resting time in the feeding and freestall areas, and showed
overall behavior similar to that observed during a cooler period. Some cows also lay in the wet dirty areas of the pens or
crowded in a well-ventilated place. The results of this study confirm the usefulness of cooling systems for cows in a warm
climate.
Keywords. Animal housing, Dairy barns, Cooling systems, Ventilation.

eep body temperature is usually maintained at a

stable level in homeothermic animals. When the
ambient temperature rises or falls from optimum,
the animal reacts with different physiological
and behavioral means to prevent body temperature from
diverging more than a small amount from the optimal setpoint. The physiological responses of the cow to heat stress
are to (1) increase heat loss through evaporation (sweating
and panting), and (2) reduce the heat generated by
maintenance (15-20% under heat stressing conditions) and
due to production (specifically heat generation linked to
digestive activity). The reduction of metabolic processes
can lead to typical consequences of heat stress including
hormonal changes, lower fertility, and lower milk yield
(Johnson, 1987). Some milk characteristics also degrade. In
particular, decreases in cheese-making qualities, such as
titratable acidity and coagulation properties are observed
(Cappa et al., 1989; Vazhapilly et al., 1992).
Several behavior modifications in response to heat
stressing conditions have been observed in dairy cattle. In
freestall barns, cows stay inside during the hottest hours of
the day to obtain shelter from intense solar radiation.
During the night, cows go outside. During warm summer

Article has been reviewed and approved for publication by the

Structures & Environment Division of ASAE.
The authors are Ermes Frazzi, Associate Professor, Agricultural
Engineering Institute, Faculty of Agricultural Sciences, Luigi Calamari,
Associate Professor, Institute of Animal Production, and Ferdinando
Calegari, Researcher, Agricultural Engineering Institute, Faculty of
Agricultural Sciences, Universit Cattolica del Sacro Cuore, Piacenza,
Italy; and Luigi Stefanini, Research Farm, Vittorio Tadini, Gariga di
Podenzano, Piacenza, Italy. Corresponding author: E. Frazzi, Universit
Cattolica del Sacro Cuore, Istituto di Genio rurale, Via Emilia PSE 84,
29100 Piacenza, Italy, phone: 39.523.599241, fax: 39.523.599924, e-mail:
<genrur@pc.unicatt.it>.

weather, cattle tend to abandon freestalls to lie outside the

barn (Arave and Albright, 1981), especially at night.
During the day, cows often lie down in dark damp
passages rather than in the freestalls (Frazzi and Calamari,
1993). Under heat stressing conditions, the cows spend a
considerable part of the day standing rather than lying
down (Igono et al., 1987).
The percentage of cows resting or ruminating in the
standing position has been observed to increase linearly as
temperature increased (Shultz, 1984). By standing, cows
maximize evaporation from body surfaces and benefit from
convection due to wind. In hot weather, the percentage of
cows drinking water or lingering around the trough without
drinking was observed to be greater for unshaded animals
than for shaded animals (Shultz, 1984). Cows under heat
stressing conditions reduce dry matter intake and decrease
the frequency of eating activity during the day and increase
such activities during the evening, nighttime, and early
morning hours (Schneider et al., 1988).
Another very important point concerns cows behavior
changes when the surroundings undergo a change. To know
how animals react in these circumstances is quite important
in order to correctly design and operate both the barn and
the ventilation and cooling systems. Research on dairy barn
environmental modification under heat stress conditions
has demonstrated the positive effect on cow performance
of microclimatic interventions such as forced ventilation
(Calamari et al., 1994; Frazzi et al., 1997), ventilation with
misting and ventilation with sprinkling (Bucklin et al.,
1991; Lin et al., 1998; Turner, 1998), and cooling (Frazzi et
al., 1998a). However, none of these studies investigated
animal behavior to show animal preferences. To answer
these questions of cow behavior, a trial was carried out to
evaluate the behavior of dairy cows in summer in relation
to the environmental conditioning system (fans with or

Transactions of the ASAE

VOL. 43(2): 387-394

2000 American Society of Agricultural Engineers 0001-2351 / 00 / 4302-387

387

se 1699 ms

7/9/01

7:31 AM

Page 388

without misting), physiological parameters, milk yield, and

milk traits.

MATERIAL AND METHODS

A study was carried out with 42 Italian Friesian cows
raised in an experimental freestall barn located in the Po
Valley (10 km south of Piacenza, lat. 4988N, long.

942E). The barn had freestalls in the rest area and the
building was partially open. The largest side (exposed to
the west) was completely open t o an unshaded paddock,
while the other was half closed by a masonry wall. The
barn housed a total of 100 lactating cows.
The 42 cows used for the trial were in an intermediate
phase of lactation. Average days in milk were 145 with an

Figure 1Barn layout.

388

TRANSACTIONS OF THE ASAE

se 1699 ms

7/9/01

7:31 AM

Page 389

average milk yield of 34 kg/cow/day at the beginning of

the trial. The animals were divided into three groups of 14
cows each. Cow selection was based on days in lactation,
calving number, and milk yield. The three groups were
housed in three pens with different environmental
conditioning systems (fans or fans with misting). The first
group was used as a control (C), the second group was
placed in a pen with fans (V), and the third group was
placed in the pen with fans plus misting (VS).
The V and VS areas were equipped with axial flow fans
(0.735 kW; 125 cm diameter; 37 500 m3 /h maximum
airflow rate) installed in the east side along the feed
passage (fig. 1). The fans were spaced 3 m apart on either
side of the drive-through feed alley and positioned to
provide airflow in the direction of the prevailing winds.
The fans were mounted at a height of approximately 2.5 m
and angled downward at about 10 from vertical. The
variable speed fans were thermostatically controlled and
were switched on at 25C and reached maximum flow rate
at 28C. A set of two misters (VS) was mounted on each
fan. Each mister had a delivery rate of 18 L/h at 750 kPa.
The misters were placed in front of each fan (1 m and 6 m
from the fan) at a height of approximately 2.5 m. The
misters operated on an 8-min cycle activated at 27C
(6 min of misting and ventilation followed by 2 min of
ventilation alone).
The study was started at the beginning of June and was
concluded at the beginning of September. The
environmental conditioning system was switched on at the
end of June and was switched off at the beginning of
August. During the trial, data were collected dealing with:
Microclimatic parameters (temperature, relative
humidity, and air speed) inside and outside the barn.
The microclimatic parameters inside the barn were
measured with electronic probes located at cow

height in different areas within the three pens and

connected to a data logger programmed to record
every 10 min.
Cow behavior with use of video cameras and
automatic photo cameras placed in different areas of
the barn such as feeding areas, resting areas, and
external paddocks. Observations were made during
four days before the hottest period and eight days
during the hottest period.
Physiological parameters (rectal temperature and
breathing rate) were measured weekly in the
afternoon (1630 h).
Milk yield and some milk traits, from the afternoon
milking, after physiological measurements were
recorded. A representative sample was collected from
the afternoon milking (1400 h) and analyzed within
1 h for pH, titratable acidity, and rheological

Figure 2Maximum daily air temperature and minimum relative

humidity for the control group.

Figure 3Hourly average temperatures measured inside the barn during hotter period (21-27 July 1995).
VOL. 43(2): 387-394

389

se 1699 ms

7/9/01

7:31 AM

Page 390

parameters. Milk yield from the morning milking

(0230 h) was also measured.
The physiological parameters, milk yield, and milk
characteristics results were statistically analyzed with the
GLM (General Linear Model) procedure of SAS (1988)
using covariance analysis with the preliminary period
values as covariates. The factors used were environmental
conditions, period, cows within environmental conditions

and the interaction between environmental conditions and

period. There were three levels, natural ventilation (C),
fans (V), fans and misting (VS), of environmental
conditions. The three levels of the period factor were:
preliminary period (B) before 1 July, with ventilation and
cooling systems switched off; experimental period (D), in
the period of maximum heat stress (21 and 27 July) with
the environmental conditioning system switched on; and

Figure 4Hourly average relative humidities measured inside the barn during hotter period (21-27 July 1995).

Figure 5Standing cows in the paddock during hotter period in summer 1995.
390

TRANSACTIONS OF THE ASAE

se 1699 ms

7/9/01

7:31 AM

Page 391

post-experimental period (A after 15 August with

ventilation and cooling system switched off). The results
relative to cow behavior were analyzed with a chi square
analysis.

RESULTS
Air speed values measured at a height of 2.5 m in the
pen with fans and in the pen with fans and misting (fig. 1)
ranged from 0.5 to 2.5 m/s with a mean value of 1.7 m/s.
At cow level, the air speed dropped to a mean value near
0.8 m/s with values higher than 1 m/s in the feeding area
and in the freestalls near to the feeding area. Our previous
trials indicated that significant reduction of heat stress was

obtained with air speeds of 0.8 to 1.0 m/s that are

obtainable with relatively simple and inexpensive
ventilation equipment (Frazzi et al., 1998b). Beyond this
limit there were further improvements, but the cooling
effect was reduced (Frazzi et al., 1998b).
The behavior of maximum temperature and minimum
humidity inside the barn for the control group (fig. 2),
measured on the days when cow physiological parameters
were measured and milk samples were collected, indicated
that the microclimatic conditions during the summer
season studied were not extremely hot. The last 10 days of
July was the hottest period. During this period (from 21
July to 27 July), mean daily maximum temperature reached
at 1700 h, was 31.5C and the daily minimum temperature

Figure 6Standing cows as a percentage of total cows in rest area during hotter (top) and cooler period (bottom) in summer 1995.
VOL. 43(2): 387-394

391

se 1699 ms

7/9/01

7:31 AM

Page 392

reached at 0600 h was 22.2C (fig. 3). The behavior of

microclimatic conditions during this hottest period was
similar in all the pens except for the pen equipped with
fans and misters which showed a lower increase of
temperature during the day (maximum value of 30.9 with
respect to 31.7C for other two pens). Small differences
were also observed for relative humidity during this hottest
period (fig. 4). The minimum value, reached at 1700 h, was
57% in the pen with fans and misters with respect to 54%
for the control. In all pens, the relative humidity reached
values above 90% at 0600 h, when the temperature reached
minimum values.
The results from observations of cow behavior
confirmed that the cows in all treatments stayed inside the
barn during the hours of the day when solar radiation was
high. The temperature was lower in the barn, but the
humidity was higher and the air movement was very low in
the pen without fans. It seemed that the solar radiation load
dominated as compared to other microclimatic parameters.
Our results showed that if solar radiation exceeded
500 W/m 2 the cows preferred to stay inside the barn both
in the morning and in the afternoon. In the afternoon, a
small percentage of animals preferred to stay outside the
barn for a period of time (fig. 5). This was probably due to
a less favorable microclimate in the building.
The presence of fans, with or without evaporative
cooling, modified cow behavior from that observed in the
control group with increased animal presence in the
paddock during the hotter hours of the day (fig. 5) when
fans were used. Around 10 to 15% of cows in the paddock
were observed in the area with misters during the hotter
hours from 1000 h to 1900 h. Cows left the barn and stayed
in the paddock for a few minutes (maximum 10-15 min)
and then returned inside the barn. The fact that the cows,
after being wetted, went out in the paddock to dry in the
sun could have negative implications, because water
evaporation from the skin, instead of removing body heat,
occurs using solar energy.
This phenomenon, that for this study was limited, was
much more evident in a previous test conducted during the
summer 1994; one of the three sections was treated only
with showers without forced ventilation. The number of
animals present in the paddock during the day was larger,
and performances were not improved. In these examples,
the paddocks presence could counteract the results
obtained inside the barn. This point must be attentively
considered.
Cows with white-colored coats stayed in the paddock
longer. Other studies (Goodwin et al., 1997) have shown a
preference by white-coated cows for unshaded areas
compared to areas below an iron roof.
The cows that were cooled by fans only stayed inside
the barn in the morning until the first hours of the
afternoon and then crowded in the paddock in the late
afternoon. Ventilation without misting when air
temperature and humidity were high was avoided by the
animals and they preferred outside conditions in the late
afternoon.
It is known that cows under heat stressing conditions
spend a considerable part of the day standing rather than
lying down. The data shown in figure 6 confirm this, with a
higher number of cows in standing position during the
hotter period as compared to the cooler period. Animals in
392

the feeding area were not included in the count of standing

cows. These results indicate that the higher percentage of
standing animals is due to thermal stress (fig. 6).
The different behavior of cows in the areas with
different environmental conditioning systems was
interesting. The differences, often statistically significant,
indicated a higher number (especially in the afternoon) of
standing animals in the pen without fans with respect to
those in conditioned pens (table 1). In the pen with fans
and misting, the number of standing animals was lower
with respect to the other areas. The higher time spent
standing for C groups cows can be explained by the cows
greater difficulty of maintaining body temperature. In this
group, during the hotter period mean rectal temperature in
the afternoon was 39.5C compared to values lower than
39C (P < 0.05) of the other two groups (table 2) and mean
breathing rate was 94/min with respect to 78/min of V
group and 70/min of VS group (P < 0.05).
Similar results were found from the data concerning the
cows lying in the freestalls (fig. 7). The presence of cows
in the freestalls was lower in control group with respect to
the others, both in the cooler and hotter period (table 1). In
the hotter period, a reduction of lying cows in the freestalls
was observed in VS and V group. In the C group, the
Table 1. Standing cows in rest area and lying cows in freestalls during cooler period
(Pre) and hotter period (Hot); cows raised in pen without (C) or with fans (V)
in control group (C), or fans and misters (VS)
Hour of the Day
0900-1100

1200-1400

Pre

Pre

Hot

49
55
57

54
50
53

Hot
38
40AB
51 * 36A
35 * 49B

Cows lying
in freestalls
(%)

V
VS
C

Standing in
Rest area
(%)

B15
V
21AB AC25
25
B37
VS AB32 * 14A
31
A13 * 24B
A14 * 28
C

1500-1700

1800-2000

Pre
Hot
A54 * 26A
B61 * 41B
AB33
31AB

Pre
Hot
A67
20A
B49 * 31B
AB21
22AB

23AB
17A
29B

21
23
24

B10
13A
AB23
14A
A9 * 21B

* = P < 0.05: differences between Pre and Hot.

AB = P < 0.05: differences between groups.

Table 2. Least squares means (covariance analysis) of selected physiological

and milk parameters of dairy cows belonging to the control (C), fans (V),
and fans plus misting (VS) group, in the preliminary (Pre), hotter
(Hot), and after hotter (Post) periods*
Parameter

Period

VS

MSE

Rectal temperature (C)

Pre
Hot
Post

A 38.63
B 39.49 c
A 38.52

A 38.63
B 38.94 b
A 38.56

38.63
38.6 a
38.59

0.08

Breathing rate (n/min)

Pre
Hot
Post

B 59
C 94 c
A 37

B 54
C 78 b
A 37

B 53
C 70 a
A 42

71.74

Milk yield (kg/d)

Pre
Hot
Post

B 34.1
A 29.0
A 28.1

B 33.8
A 29.6
A 29.8

B 33.9
A 29.4
A 28.7

6.13

Milk acidity (SH/50mL)

Pre
Hot
Post

3.46
3.35
3.38

3.47
3.42
3.47

3.46
3.46
3.40

0.02

Milk rheologycal parameters

Clotting time (min)

Pre
Hot
Post

A 19.4
B 24.1
A 19.2

B 19.0
B 22.7
A19.0

A 19.2
B 21.9
A 20.6

8.66

Curd firmness (mm)

Pre
Hot
Post

B 21.1
A 10.4
B 22.1

B 22.2
A 14.4
B 23.3

B 22.1
A 15.0
B 21.6

46.74

* Measurements were performed in the afternoon. The means on the rows

without common superscripts (right) are significantly different for P < 0.05.
The means on the columns with common superscripts (left) are significantly
different for P < 0.05.

TRANSACTIONS OF THE ASAE

se 1699 ms

7/9/01

7:31 AM

Page 393

Figure 7Lying cows in freestall as a percentage of total cows during hotter (top) and cooler period (bottom) in summer 1995.

values were unchanged in the afternoon, but during the

morning an increase of cows lying in the freestalls was
observed. This behavior could be due to the higher time
spent during the hottest hours of the day standing rather
than lying down for C groups cows. This also indicates
that during the morning, after feed distribution, the cows of
C group left the feed area quicker and stayed in the
freestalls. This behavior can reduce dry matter intake
which lowers metabolic heat production but has negative
effects on the satisfaction of cows requirements for
maintenance and production and on milk yield and quality.
During the hottest period, the smallest milk yield
reduction and the best milk characteristics were recorded
for the cows housed in the pens with fans and with fans and
VOL. 43(2): 387-394

misting (table 2). In particular, the cheesemaking properties

were positively influenced. Clotting time (lower values
indicate better characteristics) and especially curd firmness
(higher values indicate better characteristics) worsened to a
lesser extent during hottest period in V and VS with respect
to C group. These results are similar to previous trials
(Calamari et al., 1994; Frazzi et al., 1997). The results are
consistent with observations of cow behavior. Cows spent
longer periods of resting time in the feeding area and in the
freestalls, and showed overall behavior similar to that of
cooler periods (table 3). Some cows also lay in the wet
dirty areas of the pens or crowded in the well-ventilated
portion of the areas. Our results confirm the usefulness of

393

se 1699 ms

7/9/01

7:31 AM

Page 394

Table 3. Presence of dairy cows in different areas (percentage) for

control (C), fans (V), and fans plus misting (VS) group in the
afternoon hours (from 12 pm to 18 pm) during hottest period
Parameter (%)

VS

Cows in the freestalls

Cows in the passage areas
Cows in the paddock
Cows feeding

40 ab*
26
17 a
17

37 a
20
26 b
17

46 b
21
17 a
16

* a, b = P < 0.15: differences between groups.

using an adequate environmental modification system

when constructing barns in a warm climate.

CONCLUSIONS
The data on the behavior of milking cows under heat
stressing conditions demonstrated a correspondence
between performance and behavior of cows in areas of a
barn with different environmental conditioning systems.
The presence of misting induced the animals to use a
paddock in the daytime, despite a strong solar radiation
load. This reduced the concentration of animals inside the
barn with an improvement of the microclimatic conditions.
The best results (rectal temperature, breathing rate, milk
yield, and milk characteristics) were observed in the hottest
period for cows housed in pens with environmental
conditioning systems (fans with or without misting). These
results were consistent with the behavior of the cows in
experimental pens. Cows spent longer periods of resting
time in the feeding and freestall areas and showed overall
behavior similar to that observed during cooler periods.
Our results confirm the usefulness of installing adequate
equipment for environmental modification in barns in a
warm climate.
ACKNOWLEDGMENT. This research was supported by a
grant from the Italian M.U.R.S.T. 40% and from EmiliaRomagna Region (with the operational coordination of
C.R.P.A.)

REFERENCES
Arave, C. W., and J. L. Albright. 1981. Cattle behavior. J. Dairy
Sci. 64(6): 1318-1329.
Bucklin, R. A., L. W. Turner, D. K. Beede, D. R. Bray, and R. W.
Hemken. 1991. Methods to relieve heat stress for dairy cows in hot
humid climates. Applied Engineering in Agriculture 7(2): 241-247.
Calamari, L., M. G. Maianti, V. Cappa, and E. Frazzi. 1994. The
influence of the air speed on yield and milk characteristics in
dairy cows during summer. In Proc. Int. Conf. on Agricultural
Engineering, Milano, Italy, 1: 214-215. Milano, Italy: AGENG.

394

Cappa, V., P. Vazhapilly, M. G. Maianti, R. Lombardelli, and E.

Frazzi. 1989. Effetti delle variazioni ambientali (microclima)
sulle performances di vacche da latte [Effect of environmental
variations (microclimate) on the performance of dairy cows].
Sci. Tecn. Latt.-Cas. 40(2): 98-115.
Frazzi, E., and L. Calamari. 1993. Influenza dei recinti esterni
sullo stress da caldo nelle stalle libere per vacche da latte. Atti
V Convegno nazionale A.I.G.R. Il ruolo dellIngegneria per
lagricoltura del 2000 (The influence of outside enclosures on
heat stress in free barns for dairy cattle. In Proc. Nat. Congress
AIGR The role of rural engineering on agriculture of the 21st
century). Maratea (PZ), 7-11 Giugno, 207-213. Europa
Editrice, PZ, Italy.
Frazzi, E., L. Calamari, F. Calegari, and V. Cappa. 1997. The
aeration with and without misting: Effects on heat stress in
dairy cows. In Proc. 5th Int. Symp. Livestock Environment
V, Bloomington, Minn., 29-31 May, 2: 907-914. St. Joseph,
Mich.: ASAE.
Frazzi, E., L. Calamari, and F. Calegari. 1998a. Different systems
of air conditioning for dairy cows housing in Mediterranean
climate. In Proc. 13th Int. Congress on Agricultural
Engineering, Rabat, Maroc, 2-6 February, 2: 219-225. Rabat,
Morocco: ANAFID.
______. 1998b. Dairy cows heat stress index including air speed
parameter. Riv. di Ing. Agr. 2: 91-96.
Goodwin, P. J., J. B. Gaughan, T. A. Schoorl, B. A. Young, and A.
Hall. 1997. Shade type selection by Holstein-Friesian dairy
cows. In Proc. 5th Int. Symp. Livestock Environment V,
Bloomington, Minn., 29-31 May, 2: 915-917. St. Joseph,
Mich.: ASAE.
Igono, M. O., H. D. Johnson, B. J. Steevens, G. F. Krause, and M. D.
Shanklin. 1987. Physiological, productive, and economic benefits
of shade, spray, and fan system versus shade for Holstein cows
during summer heat. J. Dairy Sci. 70(5): 1069-1079.
Johnson, H. D. 1987. Bioclimates effect on growth reproduction
and milk production, Ch. 3: 35-52. In Bioclimatology and the
Adaptation of Livestock, ed. H. D. Johnson. Amsterdam, The
Netherlands: Elsevier.
Lin, J. C., B. R. Moss, J. L. Koon, C. A. Flood, R. C. Smith III, K.
A. Cummins, and D. A. Coleman. 1998. Comparison of various
fan, sprinkler, and mister systems in reducing heat stress in
dairy cows. Applied Engineering in Agriculture 14(2): 177-182.
SAS. 1988. SAS/STAT Users Guide, Rel. 6.03 Ed. Cary, N.C.
Schneider, P. L., D. K. Beede, and C. J. Wilcox. 1988.
Nycterohemeral patterns of acid-base status, mineral
concentrations and digestive function of lactating cows in natural
or chamber stress environments. J. Animal Sci. 66(1): 112-125.
Shultz, T. A. 1984. Weather and shade effects on cow corral
activities. J. Dairy Sci. 67(4): 868-873.
Turner, L. W. 1998. Fan and high pressure mist (fog) system
performance for cooling lactating cows. In Proc. 4th Int. Dairy
Housing Conf., St Louis, Mo., 28-30 January, pp. 201-208. St.
Joseph, Mich.: ASAE.
Vazhapilly, P., L. Calamari, E. Frazzi, A. Azzoni, and V. Cappa.
1992. Effect of heat stress on cheese making quality of milk
from dairy cows raised in Po valley. In Proc. CIGR
Environmental and Energy Aspects of Livestock Housing,
Polanica (Wroclaw), Poland, 19-28. Polanica, Poland: CIGR.

TRANSACTIONS OF THE ASAE