Content uploaded by Nicolai Livshin
Author content
All content in this area was uploaded by Nicolai Livshin on Sep 10, 2022
Content may be subject to copyright.
NUTRITION, FEEDING, AND CALVES
Regularity
of
Dairy Cow Feeding Behavior
with Computer-Controlled Feeders192
N. LIVSHIN, E. MALTZ,
and
Y.
EDAN3
Institute
of
Agricultural Engineering
Agricultural Research Organization
The
Volcani Center
Bet Dagan, Israel
50250
ABSTRACT
The regularity with which cows visit
a computerized concentrate feeding sta-
tion was investigated under two fixed-
time feeding routines with four and six
equal diurnal feeding intervals (feeding
windows). The experiment was con-
ducted using an Israeli commercial dairy
herd, and space was sufficient for cow
traffic related to feeding. In both trials, a
regular diurnal pattern of feeding in ac-
cordance with programmed feed availa-
bility could
be
observed within feeding
windows. Concentrates were consumed
in the same feeding window in which
they were allotted in
95
to
97%
of the
cases. Patterns
of
feeding events and
visits to a feeding station within a feed-
ing window were identical for both feed-
ing routines investigated. Distributions
of feeding intervals for individual cows
point to
a
definite feeding order
in
the
herd rather than a random sequence of
the cows feeding within a feeding win-
dow. Number and timing of visits to the
feeding station were quickly reoriented
to the new system of feeding windows.
Cow behavior was based on an under-
standing of and adaptation to the pro-
posed feeding routines, rather than being
Received December 29, 1993.
Accepted June 28, 1994.
‘An earlier, abridged version of this paper
was
presented at the Third International Dairy Housing Confer-
ence
Daily
Systems for the
21st
Century, Orlando,
FL,
February 2-5, 1994.
2Contribution from the Agricultural Research Organi-
zation, The Volcani Center, Bet Dagan 50250, Israel,
Number 1250-E, 1993 series.
3Present address: Department
of
Industrial Engineering
and Management, Ben-Gurion University
of
the Negev,
Beer-Sheva, Israel.
a result of periodic visits to feeding sta-
tion, regardless of the feeding program,
thus indicating ability to control this be-
havior.
(Key
words:
dairy, feeding systems, be-
havior)
Abbreviation
key:
FR
=
feeding routine,
FW
=
feedmg window,
ICA
=
individual concen-
trate allotment.
INTRODUCTION
Individual concentrate allotment
OCA)
with
an identification system and computerized
feeding station is an accepted method for feed-
ing dairy cows according to their energy
re-
quirements. The potential for improved
management with this technology is substantial
and is enhanced by prospects of automatic
milking
(2).
To
ensure more uniform consump-
tion of concentrates, ICA feeding stations may
be
used with a variety of variable-time
or
fixed-time feeding routines
(FR).
In variable-
time FR, the amount of feed
(or
concentrate)
available for the cow at any given time de-
pends on the interval from the previous feed-
ing. Hence, the diurnal number of feeding
opportunities (whether used by a cow or not)
may be very large, depending on the amount of
programmed minimal concentrate allotment
per visit. In fixed-time FR, each feeding inter-
val,
or
“feeding window”
(FW),
is the same for
all cows in the herd, and a fixed part of the
daily ICA is available for each cow. The diur-
nal number of FW defines the number of
feeding opportunities proposed by this type of
FR
but does not exclude several feedings by a
cow in one
FW
or
nonattendance by a cow
during a
FW.
The ICA feeding station requires
behavioral adaptation by the cows because of
the absence of visual information about the
presence of food and the limited number of
feeding devices. Results of earlier studies indi-
1995
J
Dairy
Sci 78:296-304
296
FEEDING
REGULARITY
AND
FEEDING
ROUTINE
297
cated that cows were able to adapt to this type
of equipment (1, 11); variability was great
among cows in number of daily visits to the
feeding station (1, 4, 7); and visiting activity
depended on biological (age, weight, and pari-
ty), psychological (temperament and social
be-
havior), and technical (number and layout
of
feeding devices) factors (1, 9, 11).
Behavioral responses of cows to various
types and designs of ICA
FR
are
not fully
understood, especially the capability of cows
to recognize and to respond to
FR
design. It is
known that the mean frequency of visits to a
feeding station correlates positively with the
frequency of ICA allotments
(7,
12). Some
observations indicate cows
are
capable of
recognizing
FR
design. At least some cows
grasp the periodic character of feeding station
operation and the feeding interval duration,
as
manifested by cows crowding around a feeding
station before the
start
of
a
FW
under fixed-
time FR (7, 12). Cows also may relate re-
warded visits of other cows with the availabil-
ity of the next feed allotment (12). Neverthe-
less, one study
[p.
247, (12)] concluded that
visiting behavior of cows was based on ran-
dom visits to feeding stations rather than an
understanding of the
FR
and that cows “did
not respond to the variation over the 24-h
period in their chances
of
obtaining concen-
trates” and “chose a strategy of paying regular
visits to the feeding station, because
the
cost
for such visits is low and the reward is suffi-
ciently
high”.
The regularity and predictability of cow
response
to
a specific ICA FR, which also
would characterize the ability of cows
to
adapt
to the ICA FR, still has not been thoroughly
examined but may be seen
as
part of the
ethological information needed to improve
farm operation efficiency and well-being of
farm animals (3). In particular, for automatic
milking systems based on rewarding with con-
centrate, predictable regularity of visits is
highly significant for maintaining regular milk-
ing intervals. But ensuring feeding regularity,
in
itself, may be of importance for the cow’s
well-being because dairy cows “respond more
than any other class of livestock to regularity
and systematic feeding. Therefore, whatever
routine of feeding, watering, etc. is adopted,
should be followed regularly every day” [p.
290,
(6)].
Furthermore,
an
understanding of the
FR
that is required for the proper adaptation of
cow behavior is the main component of a
predictable feeding environment and thus be-
comes an important factor for the cow’s wel-
fare
(8,
10).
The general objective of this work was to
investigate the regularity of feeding behavior
for a fixed-time
FR.
This system has the
potential to introduce regular feeding cycles
for the entire herd and to promote feeding
regularity by triggering mechanisms of group
behavior. The previous studies with this FR (4,
11, 12, 13) were conducted with cows housed
in free stalls. We assumed that-1) under the
corral housing conditions typical in Israel,
which do not impede cow traffic, waiting, or
regrouping related to feeding station usage,
and,
2)
with a specific design of FR that is
comprehensible to the cows-feeding behavior
should
be
consistent with feeding opportunities
proposed by the FR, rather than being only a
result of casual feeding attempts by individual
The specific goals of this work were 1) to
investigate regularity of feeding behavior
of
the herd and of individual cows in relation to
feeding opportunities from different fixed-time
FR
(measured
as
regularity of consumption of
available ICA within the same
FW)
and time
patterns of feed consumption within
FW;
and
2) to characterize the behavioral background
behind feeding patterns and determine whether
they are a result of casual visits of individual
cows or whether they indicate the herd’s adap-
tation to programmed
FW
times under condi-
tions of limited access to a feeding station.
The regularity of usage of feeding opportu-
nities may also show whether the FR design
can
be
an effective instrument to control the
frequency and timing of feeding events.
cows.
MATERIALS
AND
METHODS
The experiment was performed on a com-
mercial farm with about
50
milking Israeli
Holstein cows in 1993. The cows were housed
in a corral-type shed with a common feeding
bunk (Figure 1). An ICA feeding station, with
three
self-feeders at the center of the covered
feeding area, allowed undisturbed feeding at
the bunk and was visible to the cows from
anywhere in the dairy. The layout of the facili-
ties allowed
free
cow traffic and grouping
Journal of
Dairy
Science
Vol.
78,
No.
2,
1995
298
LIVSHIN
ET
AL.
each
FW
(successive visits by a cow at an
interval <1 min were counted as one visit). 2)
The completeness of usage of feeding opportu-
nities allowed by the
FR,
defined as the com-
plete consumption of
FW
ICA in the same
FW
[the total number of these feeding opportuni-
ties for the herd was defined
as
(number of
days of observation)
x
(number of diurnal
FW)
x
(number of cows)]. 3) Feeding intervals be-
tween rewarded visits in successive
FW
for the
individual cows were defined as differences
between feeding times in these
FW:
feeding interval
=
mk
-
mk-1
-
U
I
ROOFED YARD
m-
UNROOFED
YARD
DRY
RI
1
cowsr
Figure
1.
Cow
house layout: self-feeders
(dark
rectan-
gle). water troughs (dot-filled rectangle), fence (solid line),
border
of
roofed area (dotted line).
around the feeding station. Concentrates were
fed only through the feeding station. The mean
ICA was approximately 10 kg/d and ranged
from 3 to 18 kg/d according to milk produc-
tion. Concentrates were delivered at a rate of
300
g/min. The self-feeders (Fulwood; Elles-
mere, Shropshire, England) were served by
software (‘Afifeed’; Zaham, Afikim, Israel)
used to monitor visiting behaviors of time,
duration, and allotment of concentrates of each
cow’s visit to each self-feeder. The cows were
milked twice daily at 0500 to 0530 h and 1700
to 1730 h. At about 0715
h,
hay (alfalfa and
vetch) was distributed, and the cows were
locked at the feeding bunk for veterinary in-
spection until about 0800 h. At about 1400 and
2130 h, corn silage was distributed.
The feeding stations were operated to dis-
pense the daily ration in four equal 6-h
FW
(opened at 2400,0600, 1200, and 1800 h)
as
a
standard routine for 2 yr. During
this
period,
fresh cows adapted to the ICA feeding without
need of any special training. Daily ICA were
equally distributed among
FW,
and uncon-
sumed ICA was transferred to the following
FW.
After four
48-h
measurements of time,
duration, and concentrates intake
of
each visit
to the feeding station within 30 d
(trial
l), the
FR was changed to six equal 4-h
FW
(opened
at 2400, 0400, 0800, 1200, 1600, and 2000 h)
with no FR-related changes in daily allowance
of concentrates. Ten days, without any special
training, were allowed for the adaptation, after
which 72- and 96-h measurements were taken
(trial 2).
The following variables were analyzed.
1)
Number of rewarded and unrewarded visits in
where
mk
=
feeding time in
FW
k. If feedings
were interrupted, cow feeding time for the
FW
was computed
as
the weighted mean for all
rewarded visits in this
FW:
mk
=
&&kgik)
where t&
=
time of feeding i
(i
=
,
1,2,
,
.)
in the
FW
k, and gk
=
concentrates consumption
during the feeding i in the
FW
k. 4) Deviations
of the relative feeding time (from the
FW
start)
of the cow in current
FW
from that time in
previous
FW.
These deviations were calculated
for each feeding interval, as previously de-
fined,
as
differences between this feeding in-
terval and
FW
duration.
5)
Consumption
of
concentrates by the herd within a
FW.
6) Feed-
ing station workload in a FW (time of re-
warded visits, time of unrewarded visits, and
idle time).
To
assess the reliability of recorded data on
ICA allotments, several observation periods
were conducted prior to the data collection to
evaluate amount of concentrates that were left
unconsumed after rewarded visits. Cows that
were not interrupted consumed all allotted con-
centrates during their visits. As a rule, no
significant amount of concentrates remained
unconsumed after interrupted (by butting from
the
rear)
feedings. We concluded that data on
ICA dispensing, recorded by the computer,
could be used as ICA consumption.
Because of the routine changes in the herd
during the study, data related to individual
cows were analyzed and compared for the 42
cows that were present in both trials.
The analysis
of
the influence of FR design
on milk production was not an objective of this
Journal
of
Dairy
Science
Vol.
78,
No.
2,
1995
FEEDING REGULARITY
AND
FEEDING ROUTINE
299
arts
(13 in trial
1
and 62 in trial 2) were a
result of some cows not attending the feeding
station during some of the
FW.
The mean
number of rewarded visits per
FW
was 1.43
and 1.25 per cow in trials
1
and 2, respectively
Feeding windows
'O0.O
1764
IOo.'
(P
<
.01).
Nevertheless, in both trials, approxi-
attendance
1163b 98.9 17ma 96.5
mately 98% of
FW
attendance by a cow
resulted in complete consumption of
this
FW
consumption
1135 96.5 1674 94.9
ICA, indicating the persistency of the cows to
consume the concentrates available in the
FW.
As for individual cows (Table 2), in trial
1,
consumption, total 28 2.4
28 1.6
>50%
23c
2.0
9d
.6
40%
5
.4
17 1.0
80.9% did not miss any
FW;
16.7% missed no
more than two out of 28
FW
in 7 d. In all,
97.6%
of
the
cows
either
did
not
miss
any
FW
or missed no more then two out of 28
FW.
In
trial
2,
59.5% of cows did not miss any
FW
and 21.4% missed no more than two
FW
in 7
d. In all,
81%
of the
COWS
attended all or
missed no more than two of the 42 available
FW
during the Same
period.
The maximal
FW
was
10.7%
(3
out
of 28
FW)
for
one cow in trial
1
and 21.4% (9 out of 42
FW)
for
two
cows
in
trial
2.
Two
consecutive
FW
were missed only in a few cases in trial 2.
TABLE
1.
Usage of feeding opportunities proposed
by
two fixed-time feeding routines.1.2
Trial
1
(no.)
(96)
~,.ial
2
(no')
(%)
Complete
Incomplete
UvbSignificant differences between trials
in
the distribu-
tions of events between attendance and nonattendance of
feeding windows
(P
<
,001;
chi-square test).
C4ignificant differences between trials in the distribu-
tions of incomplete consumption
cases
between consump-
tion
>50%
and
<50%
of individual allotments for feeding
windows
(P
<
.01;
chi-square test).
lThe
6-h
(trial
1)
and
4-h
(trial
2)
feeding windows.
Data for seven 24-h observations in each trial for the
same
42
cows.
2Data relate to consumption of the individual concen-
trate allotment of this feeding window and not
orts
from
the previous feeding window.
3Trial
1:
42 cows
x
7
d
x
4
feeding windows
=
1176;
trial 2:
42
cows
x
7
d
x
6
feeding windows
=
1764.
study; however, the milk production during the
experiment was monitored to follow possibly
damaging effects of introducing new FR de-
sign. The recorded increase in mean milk
production (25.7 and 26.5 kgld per cow in
trials
1
and 2, respectively), although statisti-
cally nonsignificant, was nevertheless convinc-
ing for the fanner, so the 4-h
FW
became the
FR on this dairy after completion of the ex-
periment.
Differences between means were evaluated
by a paired
t
test; differences between distribu-
tions were evaluated by a chi-square test
(5).
RESULTS
Regularlty
of
Usage of ICA FW Allowancea
Regularity
of
Feeding Events
Within an FW: Herd
Feeding Station
Work
Load.
The ICA feed-
ing station work load pattern was similar in
both
trials (Figure 2) and consisted of
1)
a
period
of high continuous feeding station ca-
pacity utilization (first
3
h in
trial
1
and fiist 2
h
in trial 2) with very low idle time and total
duration of unrewarded visits; 2) a period of
low feeding station work load (last
3
h in trial
1
and last 2 h
in
trial 2); and 3) the last hour of
the
FW
in both trials was
also
characterized by
significant increase in feeding station
oc-
cupancy by unrewarded visits.
ICA
Consumption.
Both trials followed the
same pattern of consumption of concentrates
(Table 3): the quantity of concentrates con-
sumed was highest in
1st
h of
FW,
declined
consistently thereafter, and decreased sharply
after consumption of most (approximately
80%) of the total herd
FW
ICA.
Table
1
summarizes the total usage of avail-
able feeding opportunities by the herd. The
available FW ICA were completely consumed
:!::;
&FE:;:iF::ws
in 96.5 and 94.9% of cases in trials
1
and 2,
respectively, and partly consumed in 2.4 and
1.6% of cases in trials
1
and 2, respectively.
As is shown in Figure
3,
in both trials,
individual feeding intervals were distributed
around
FW
duration (SD 1.42 and
1.11
h for
Journal of
Dairy
Science
Vol.
78,
No.
2,
1995
300
LIVSHIN ET
AL.
TABLE 2. Distribution of cows by completeness of usage of feeding opportunities proposed by
the
two fixed-time
feeding routines.'
Trial
1
Trial 2
Incomplete Non- Incomplete Non-
attendance Cases consumutionz attendance consumotion
(no.) (no.)
(96)
(no.)
(W
(no.)
(%)
(no.)
(S)
None
2Za
52.4 34. 80.9
25a
59.5
25b 59.5
1-2
18
42.8 7 16.7
14
33.3 9 21.4
34
2
4.8
1
2.4
2
4.8 2 4.8
5-6
0
0
0
0
1
2.4 2 4.8
7-8
0
0
0 0
00
2 4.8
9-10
00
00
0
0
2 4.8
Total 42 100.0 42 100.0 42 100.0 42
100.0
akSignificant differences between trials in cow distribution by number of cases of incomplete consumption of
'The
6-h (trial
1)
and 4-h
(trial
2) feeding windows.
Data
for seven 24-h observations
in
each trial for the same
42
ZFeeding window attendance with incomplete consumption of concentrates allotment.
concentrate and
missed
feeding windows marked with different superscripts (chi-square test;
P
e
,001).
cows.
trials 1 and 2, respectively). Figure 4 shows the
distribution of deviations of the relative (from
the
FW
start)
feeding time in current
FW
from
that time in previous
FW.
Differences between
the two trials were significant
(P
<
.001). but
the overall distribution pattern was the same:
in almost 40% of the cases (36.1 and 38.2% in
trials 1 and
2,
respectively), the deviation from
the feeding time in previous
FW
was <30 min
and, in approximately
80%
of the cases (76.6
and 82.6% in trials 1 and 2, respectively), 4.5
h. This deviation exceeded
2
h only
for
13.7%
of cases in trial 1 and 7.2% in trial 2.
Visiting Behavior Adaptation
to
ICA Feeding Routine
Number
of
Visits.
The mean diurnal number
of visits to the feeding station per cow in-
creased 33% after the number of diurnal
FW
increased from 4 in trial
1
to
6
in trial
2
(Table
4). The difference between trials in number of
unrewarded visits per
FW
was not significant.
Hence, the general level of visiting activity, as
indicated by unrewarded visits, was almost
proportional to the number of ICA allotments
(number of
FW).
The mean number of re-
warded visits per cow per
FW
decreased
15%,
indicating a substantial (almost double) de-
crease in interrupted feedings.
Timing
of
Visits.
To characterize the in-
fluence of the
FR
design on the time patterns
Journal
of
Dairy
Science
Vol.
78,
No.
2,
1995
of cow visits, we compared
the
feeding station
attendance by cows during the 4-h
FW
in trial
2 with the attendance of the same 4-h intervals
in trial 1, when these intervals did not coincide
with the
FW
(Table
5).
In trial 1, only
3
cows
(7%) did not miss any of such 4-h intervals for
the seven 24-h periods, compared with 25
cows
(63%)
in trial 2. Importantly, during the
same period in trial 1, 81% of the cows did not
miss
any of the
6-h
intervals coinciding with
the
FW
of this trial (Table
2).
DISCUSSION
In both trials, ICA feeding was regular in 1)
degree of usage of available feeding opportuni-
ties,
as
shown by the data on
FW
attendance
and FW ICA consumption (Tables 1 and 2); 2)
patterns of herd feeding within
FW,
as shown
by the data on feeding station work load
(Figure 2 and Table 3); and 3) orderliness of
the individual cow's feeding within a
FW,
as
evidenced by the distribution
of
individual
feeding intervals and relative feeding times
(Figures
3
and
4).
The mean intensity
of
visiting behavior was
strongly correlated with the number of feeding
opportunities in each trial (Table 4); the per-
centage of unrewarded visits was approxi-
mately the same for both trials (52.1 and
49.4% in trials 1 and 2, respectively). The
observed feeding behavior regularity was
FEEDING
REGULARITY
AND FEEDING ROUTINE
301
1
23456
r
30
20
10
1
2
3
4
HOURS
OF
FEEDING WINDOW
Figure 2. Concentrate feeding station work
load
within
a feeding window (in percentage of
time)
under the
two
fixed-time feeding routines: 6-h (A) and 4-h
(B)
feeding
windows. Herd data (49 cows) for
three
24-h observations
for comparable feeding windows (1800 to 2400
h
in
A
and
2000 to 2400 h in B) with mean consumption of concen-
trates of 131.3 and 90.6 kg. respectively.
The
striped bar
indicates rewarded visits; open bar, unrewarded visits; and
solid bar, feeding station idle (intervisit) time.
Error
bars
represent standard errors.
based on cow adaptation, not only to number
of feeding opportunities, but also to the time
frames of those opportunities. This adaptation
could be concluded from
1)
visiting behavior
(number and timing of visits) orientation to
feeding station attendance in accordance to the
FW
system (Table
5);
2)
herd feeding patterns
within a
FW
(absence of feeding station idle
time after start of the
FW,
continuous herd
feeding, and a sharp increase in visiting ac-
tivity and feeding station occupancy before the
start of a new
FW
in
both
trials);
and
3)
consistency
in
the individual cows feeding or-
der within
FW
(as
indicated by distributions of
individual feeding intervals in both
trials).
Generally, the
ICA
feeding behavior ob-
served may be described
as
a two-stage proc-
TABLE
3.
Concentrate consumption within feeding win-
dows under the two fixed-time feeding routines.'
Concentrates consumedZ
Hour Trial
1
Trial 2
1
2
3
4
5
6
Total
-
(W
-
X
SE
30.9 .8 46.1
1.5
28.7
1.5
31.7 .9
17.2 1.2
22.8
.8
12.0 1.2
5.0
.8
X
SE
3.9
.6
......
1.7 .6
......
100.0 100.0
'The 6-h (trial 1) and 4-h (trial 2) feeding windows.
Herd data for 3 d in each trial (last three 24-h observations
in trial 1 and first
three
24-h observation in trial 2) for
three feeding windows (opened at 2400, 1200, and 1800 h)
for trial
1
and four feeding windows (opened at 2400,
1200, 1600, and 2000 h) for trial 2. Total mean daily ICA
consumption was 547.4 and 555.7 kg in trials 1 and 2,
respectively.
;?Both distributions differ significantly from uniform
(P
<
,001; chi-square test).
ess: recognition by the herd of the
start
of
a
new feeding window through observing
ex-
plorative attempts of individual cows and con-
tinuous consumption of
ICA
in which feeding
order of individual cows played an important
part. Therefore, regularity of feeding behavior
-.1
<
5
1-15
2-25
3-35
4-45
555 665
7.73
685
595
!&IO5
INDIVIDUAL FEEDING INTERVALS
(h)
Figure 3. Distribution of individual feeding intervals
under the two fixed-time feeding routines: 6-h (striped bar)
and 4-h (solid bar) feeding windows. Herd data (49 cows)
for 3-d periods, not including feeding intervals after nonat-
tendance at feeding windows.
Journal of Dairy Science
Vol.
78, No. 2,
1995
302
LIVSHIN ET
AL.
I
DEVIATIONS
(h)
Figure 4. Deviations of
the
cow’s relative (from the
feeding window’s
start)
feeding time from
this
time
in
the
previous feeding window under the two
fixed-time
feeding
routines: 6-h
(striped
bar)
and 4-h (did
bar)
feeding
windows. Herd
data
(49 cows)
are
for 3-d
periods,
not
including feeding intervals after nonattendance
at
feeding
windows.
was the outcome of cow response to regularity
of feeding opportunities proposed by the FR
and cow feeding order within the
FW,
which
was introduced by the cows themselves under
conditions of limited access to a feeding sta-
tion.
The results indicate that the proper design
of ICA FR may ensure effective regulation of
consumption of concentrates via self-feeders
both during one feeding and during ICA daily
distribution. The more even consumption of
diurnal ICA may contribute to better rumina-
tion and fermentation, thus, improving ruminal
environment. The ICA feeding station location
in the feeding bunk area also provided any
cow that attended the feeding station with the
opportunity to
feed
undisturbed from the feed-
ing bunk by turning her body (Figure
1).
More
frequent visits to the ICA feeding station (re-
warded, unrewarded, interrupted,
or
not
materialized because of limited access) could
result in more feed intake in the bunk in trial 2
(not measured in this experiment). The
recorded increase in milk production during
trial
2,
although not conclusive statistically,
can
be
explained by the possibility of higher
DMI during trial 2 as well as by an improved
ruminal environment.
Maintenance of rather stable individual
feeding intervals through proper choice of
FW
duration is of particular significance for auto-
matic milking systems based on rewarding
with concentrate
(2).
Ensuring feeding regular-
ity and a predictable feeding environment
seems to be, in itself, important for cow well-
being; therefore, the attainable feeding regular-
ity may serve as a criterion
for
comparison and
choice of
FR
type and design.
Any FR design apparently has to be under-
stood by the cows to achieve the desired re-
sponse in
FW
ICA usage, feeding station work
load, and individual feeding intervals. Lack
of
adaptation to fixed-time FR may be related to
impediments in
FR
design, facilities used, or
both. Wierenga and Hopster (12,
13),
when
TABLE 4. Mean number of visits to the concentrate feeding station under the two fixed-time feeding routines.*.*
Trial 2
as
Visits
a
percentage
per cow Trial
1
Trial 2 of
trial
1
(no.)
X
SE
X
SE
For 24
h
-
-
Rewarded 5.74
.06
7.33*** .13 127.8
Unrewarded 5.78 .23 8.00**
.I8
138.5
Total 11.51 .22 15.33***
.23
133.1
Per feeding window
Rewarded
Unrewarded
Total
1.43 .02 1.22..
.02
85.2
1.44
.06
1.33 .03 92.3
2.88
.05
2.55*
.04
88.8
‘The
6-h
(trial
1) and 4-h
(trial
2) feeding windows. Data for Seven 24-h observations in each trial
for
the same 42
2Significant differences between
trials
(***P
<
.001;
**P
e
.01;
and
*P
<
.05;
t-test).
cows.
Journal of
Dairy
Science
Vol.
78,
No.
2, 1995
FEEDING
REGULARITY
AND FEEDING ROUTINE
303
TABLE
5.
Feeding
station
attendance in the 4-h intervals
that coincide with trial 2 feeding windows.12
Missed Attendance
in
4-h intervals3
intervals
Trial
1
Trial
2
4-h
(no.)
(no.)
(96)
(no.)
(%)
None
3
7.1 25
59.5
1-2 6 14.2 9 21.4
34
7
16.7 2 4.8
5-6
9 21.4 2 4.8
7-8 8 19.0 2 4.8
9-10 7 16.7 2 4.8
11-12 2 4.8
0
0
Total 42
100.0
42 100.0
IIntervals: 2400
to
0400,
O400
to
0800,
0800 to 1200,
1200
to
1600, 1600
to
2000, and 2000 to 2400
h.
2Data
for seven 24-h observations in
each
trial for the
same 42 cows.
3Significant difference between trials
in
cow distribu-
tions by number of missed 4-h intervals
(P
<
.001; chi-
square
test).
comparing feeding behavior with various types
of ICA FR, employed a cubicle modified into
a feeding station in one end of a free-stall
cubicle line. The
FR
used by them was of
unequal
FW
during which unequal amounts of
concentrates were allocated
(3
kg per cow for a
4-h
FW
followed by allotment of
1.5
kg for an
8-h
FW).
Hence, their conclusion that the cows
did not understand the proposed FR might be
related to the conditions of their experiment. In
our case, the corral-type shed provided feeding
station location that enabled good visibility
from any point in the dairy. Good visibility
and a simple
FR
design (equal durations of
FW
that are properly adjusted to the daily routine
and equal sizes of ICA in each
Fw)
are
impor-
tant factors for feeding behavior adaptation to
FR design; however, this subject requires
fur-
ther comparative investigations.
The number of self-feeders in this experi-
ment was sufficient for cyclic herd feeding in
both trials, as was shown by the data on feed-
ing station work load (Figure
2).
The signifi-
cant excess supply
of
self-feeder service over
demand seems to be imperative for cows to
adapt to fixed-time
FR
and to display regular
feeding behavior. In this experiment, feeding
station employment increased about
5
h/d in
trial
2
compared with that of trial
1,
because
the mean number of unrewarded feeding at-
tempts was proportional to the diurnal number
of FW (Table 4). Hence, sufficient excess ca-
pacity of ICA feeding station may be the limit-
ing factor for increasing the number of diurnal
Fw.
The observed feeding order of cows within
a
FW
reflects the social relationships within
the
dairy
herd, which
are
often discussed in the
literature in terms of feeding rank and social
dominance (4,
8,
12).
Our paper does not deal
with this issue; however, the significant num-
ber of interrupted visits and differences be-
tween trials in social pressure for access to
feeding station
(N.
Livshin and
E.
Maltz, 1994,
unpublished data) indicate the need for further
research regarding the influence of FR on
agonistic behavior in herds and feeding
regularity.
From the findings about easily attained
regularity
of
ICA feeding with fixed-time FR,
we cannot yet conclude that
this
FR
should be
preferred over variable-time
FR.
The regularity
and related aspects of cow feeding behavior
with variable-time FR deserve separate study.
CONCLUSIONS
High regularity of cow feeding behavior is
attainable with proper design of the ICA FR
and feeding environment,
This
regularity was
observed in relation to
1)
feeding station atten-
dance by cows in accordance with a pro-
grammed system of
FW,
2)
time patterns of
herd feeding within a
FW,
and
3)
time patterns
of individual cow feeding within a
FW.
Regularity of feeding behavior was based on
adaptation of cows to the proposed
FR
rather
than being a result of periodic visits to the
feeding station, regardless of the feeding pro-
gram.
Cows were capable of adapting quickly
to the new FR (to attend the feeding station in
accordance with a new
FW
system).
This
result
indicates controllability of
this
behavior by
ICA
FR,
which can be exploited as a tool to
achieve specific goals related to cow well-
being or to farm management considerations.
These goals may include not only ICA distri-
bution throughout the day,
or
limitation of
one-time consumption
of
concentrates, but also
a predictable feeding environment, regular
feeding behavior, and regular feeding station
attendance. Regulation of feeding station atten-
Journal
of Dairy Science
Vol.
78, No. 2, 1995
304
LIVSHIN ET
AL.
dance is
of
particular significance for auto-
matic milking systems based on concentrate
rewards.
ACKNOWLEDGMENTS
The authors thank
S.
L.
Spahr (University
of
Illinois) for critically reading the manuscript
and A. Antler for assistance in data collection.
REFERENCES
1 Cassel, E.
K.,
W.
G.
Merrill, T. V. Bui. R. Milligan.
and R. Guest. 1982. Evaluation of systems for feeding
supplemental concentrates to group fed cows. Anim.
Sci. Mimeo Ser. No. 59 rev., Cornell Univ., Ithaca,
NY.
2 Devir,
S.,
J. A. Renkema, R.B.M. Huime, and A. H.
Ipema. 1993. A new
dairy
control and management
system in the automatic milking farm: basic concepts
and components. J. Dairy Sci. 76:3607.
3 Humik, J.
F.
1992. Ethology and technology: the role
of ethology in automation of animal production
processes. Page
401
in
Proc. Int. Symp.
Prospects
Automatic Milking, Eur. Assoc. Anirn. Prod. Publ.
No.
65,
Pudoc Sci. Publ., Wageningen, The Nether-
lands.
4Metz-Stefanovska,
J.,
and
S.
L.
Spahr. 1989. Be-
havioural observations on a group
of
cows using a
dual computer-controlled concentrates dispenser. Res.
Rep. No 89-1, Inst. Agric. Eng., Wageningen, The
Netherlands.
5 Number Cruncher Statistical System, Version
5.1
Edi-
tion. 1988.
J.
L.
Hintze, Kaysville, UT.
6Olson, T. 1950. Elements of Dairying. rev. ed. Mac-
Millan, New York, NY.
7 Pirkelmann, H. 1992. Feeding strategies and automatic
milking.
Page 289
in
Proc. Int. Symp. Prospects Auto-
matic Milking, Eur.
Assoc.
Anim. Prod. Publ. No.
65,
Pudoc Sci. Publ., Wageningen. The Netherlands.
8
Rutter,
S.
M., D. A. Jackson, C.
L.
Johnson, and
J.
M.
Forbes. 1987. Automatically recorded competitive
feeding behaviour
as
a measure of social dominance
in dairy cows. Appl. Anim. Behav. Sci. 17:41.
9 Tennessen, T. 1989. Coping with confinement-
features of the environment that influence animals’
ability to adapt. Appl. Anim. Behav. Sci. 22:139.
10Wiepkema.
P.
R. 1988. Control and cognition: key
concepts in applied ethology. Page 128
in
Proc. Int.
Congr. Appl. Whol. Farm. Anim. Scara, KTLB,
Darmstadt, Germany.
11 Wierenga, H. K., and H. Hopster. 1990. The sig-
nificance
of
cubicles
for
the behaviour
of
dairy cows.
Appl. Anim. Behav. Sci. 26309.
12 Wierenga, H. K., and H. Hopster. 1991. Behaviour of
dairy cows when fed concentrates with an automatic
concentrates feeding system. Appl. Anim. Behav. Sci.
30:223.
13Wierenga, H.
K.,
and
H.
Hopster. 1991. Timing of
visits
to
the concentrates feeding station by dairy
cows. Appl. Anim. Behav. Sci. 30247.
Journal
of
Dairy
Science Vol. 78, No.
2,
1995