Short Communication: Prepartum Photoperiod Effect on Milk Yield and Composition in Dairy Cows

Abstract

In a previous paper we analyzed the effects of day length, the daily change in day length, and heat load prevailing on test days, and on milk yield and composition of dairy cows in hot weather. For this analysis we used milk tests of three herds in Israel between 1994 and 1996. We used the same database to analyze the effects of the day length and the daily change in day length 3 wk prepartum. The prepartum day length effect was negatively related to milk yield and to milk fat and protein percentages. The daily change in day length was negative for milk yield and lactose percentage and positive for protein content and did not affect fat content. The difference of 4 h between the shortest and the longest day, plus the seasonal change in day length, accounted for the addition of 1.9 kg of milk/d for a cow calving after the shortest day compared with a cow calving after the longest day. The difference in milk composition between these two cows was estimated to be 0.27 and 0.08% of fat and protein, respectively.

Full text

Short Communication: Prepartum Photoperiod Effect
on Milk Yield and Composition in Dairy Cows
Y. Aharoni,* A. Brosh*, and E. Ezra†
*Department of Beef Cattle, Agricultural Research Organization
Newe Ya’ar Research Center, P.O. Box 1021,
Ramat Yishay 30095, Israel
†Israel Cattle Breeders Association,
P.O. Box 3015, Qesarya Industrial Park 38900 Israel
ABSTRACT
In a previous paper we analyzed the effects of day
length, the daily change in day length, and heat load
prevailing on test days, and on milk yield and composi-
tion of dairy cows in hot weather. For this analysis we
used milk tests of three herds in Israel between 1994
and 1996. We used the same database to analyze the
effects of the day length and the daily change in day
length 3 wk prepartum. The prepartum day length ef-
fect was negatively related to milk yield and to milk
fat and protein percentages. The daily change in day
length was negative for milk yield and lactose percent-
age and positive for protein content and did not affect
fat content. The difference of 4 h between the shortest
and the longest day, plus the seasonal change in day
length, accounted for the addition of 1.9 kg of milk/d
for a cow calving after the shortest day compared with
a cow calving after the longest day. The difference in
milk composition between these two cows was esti-
mated to be 0.27 and 0.08% of fat and protein, respec-
tively.
(Key words: milk yield, milk composition, photo-
period, preparturition)
Abbreviation key: DC = daily change in day length
(min/d) at the test day, DCP = daily change in day
length (min/d) 21 d before parturition, DL = day length
(h) at the test day, DLP = day length (h) 21 d before
parturition, HL = heat load (index) at the test day.
INTRODUCTION
The seasonal effects of photoperiod and heat load
on milk yield and composition of lactating cows in hot
weather were reported previously (Aharoni et al,, 1999).
In that paper, we referred to the heat load (HL), day
length (DL) and the daily change in day length (DC)
Received May 8, 2000.
Accepted August 29, 2000.
Corresponding author: Y. Aharoni; e-mail: yoavah@netvision.
net.il.
2000 J Dairy Sci 83:2779–2781 2779
during lactation. Recent evidence (Miller et al., 2000;
Petitclerc et al., 1998) suggests that the DL prepartum
affects the milk yield of cows in the subsequent lacta-
tion. Furthermore, this effect was negative, i.e., short
days in the prepartum period were associated with in-
creased milk yield thereafter, in contrast to the positive
effects of DL and DC on lactating cows. The use of
appropriate regression models to analyze large data-
bases of milk test records in commercial herds enables
detection of seasonal effects on milk composition, in
addition to their effects on milk yield (Aharoni et al.,
1999). Therefore, in the current report, we used the
database from our previous study to test the effects of
prepartum DL and DC on milk yield and composition
in the subsequent lactation.
The database (Aharoni et al., 1999) comprised 28,029
milk test records of 2029 cows in three herds, collected
in a 3-yr period, January 1994 to December 1996. Only
records of less than 271 DIM were included to avoid
the possible effect of stage of pregnancy on yield. Day
length (h), DC (min/d), and HL (arbitrary index units)
for each test day were calculated as described pre-
viously (Aharoni et al., 1999). The database included
identification of each cow by cow ID, herd, year, lacta-
tion number, test date, DIM, milk yield (kg/d), and per-
centages of fat, protein and lactose in the milk, and the
seasonal variables at the test date: DL, DC, and HL.
To these, we added records of the prepartum day length
(DLP, h) and daily change in the day length (DCP,
min/d), which were calculated for a date that precedes
parturition by 21 d.
We compared two regression models, the first (model
1) was used in the previous paper for the common analy-
sis of the three herds, and the second (model 2) included
the DLP and DCP variables.
The equation of the regression model was:
Y
ijklm
= A
i
+ YR
j
+ L
k
+ b
1k
∗DIM + b
2k
∗DIM
2
+ b
3k
∗DIM
3
+ b
4k
∗DIM
4
+ c
1
∗DL + c
2
∗DC
+ c
3
∗(HL∗H)
m
{+ c
4
∗DLP + c
5
∗DCP}{+d∗MY}+e
where:

AHARONI ET AL.2780
Table 1. Seasonal effects on milk yield. Common photoperiod, and separate in-herd effects of heat load on
milk and covariate-adjusted percentages of milk fat, protein and lactose. Model 2 (M2), which included
effects of the day length and daily change in day length 21 d before parturition, is compared with model 1
(M1), which did not include these effects.
Milk yield, kg/d % Fat in milk % Protein in milk % Lactose in milk
Effects M1 M2 M1 M2 M1 M2 M1 M2
DL
a
Value 0.394 0.372 −0.0189 −0.0218 −0.0349 −0.0357 0.0121 0.0122
t Value 13.3 12.5 −5.34 −6.19 −30.0 −30.7 10.62 10.71
P *** *** *** *** *** *** *** ***
DC
b
Value 0.567 0.559 −0.0209 −0.0217 0.0204 0.0203 0.0019 0.0019
t Value 13.2 13.0 −4.08 −4.25 12.1 12.0 1.17 1.16
P *** *** *** *** *** *** NS NS
DLP
c
Value −0.442 −0.0634 −0.0156 0.0013
t Value −9.63 −11.63 −8.70 0.74
P *** *** *** NS
DCP
d
Value −0.089 0.0037 0.0088 −0.0082
t Value −2.06 0.72 5.19 −4.92
P * NS *** ***
HL1
e
Value −0.054 −0.054 −0.0046 −0.0047 −0.0011 −0.0011 −0.0006 −0.0006
t Value −11.7 −11.7 −8.4 −8.6 −6.0 −6.3 −3.3 −3.2
P *** *** *** *** *** *** *** ***
HL2
e
Value −0.036 −0.036 −0.0098 −0.0099 −0.00010 −0.00016 0.0006 0.0006
t Value −8.7 −8.7 −19.9 −20.1 −0.6 −1.0 3.8 4.1
P *** *** *** *** NS NS *** ***
HL3
e
Value −0.023 −0.023 −0.0075 −0.0076 0.0002 0.00015 −0.0003 −0.0002
t Value −4.95 −4.95 −13.8 −14.0 1.15 0.86 −1.17 −1.34
P *** *** *** *** NS NS NS NS
a
Day length, h.
b
Daily change in day length, min/d.
c
Day length 21 d before the last parturition.
d
Daily change in day length 21 d prior to the last parturition.
e
Heat load index in each of herds 1, 2, and 3.
Y = daily milk yield or percentages of fat, pro-
tein, and lactose of cow i that calved at
year j on lactation k, at milk test l, in herd
m.
A = absorbing effect of the individual cow i
YR = effect of year j
L = effect of lactation grade (k = 1, 2, 3 or 4
for L 1 through 4, respectively, and k = 5
for L = 5+)
DIM = days in milk
DL = day length (h)
DC = daily changes of day length (min/d)
DLP = day length (h) 21 d prepartum (only in
Model 2)
DCP = daily change in day length (min/d) 21 d
prepartum (only in Model 2)
HL = heat load index
MY = milk yield, when Y is fat, protein or lactose
percentages
b
1,2,3,4(k)
= regression coefficients for lactation k
Journal of Dairy Science Vol. 83, No. 12, 2000
c
1,2,3,4,5
= regression coefficients for seasonal effects
d = regression coefficient for milk yield
e = random residual effect.
H
m
was the herd effect, and c
3
was the regression
coefficient for HL in herd m. The seasonal effects on
milk yield and composition, which were estimated by
model 2, are compared to those estimated by model
1 in Table 1. In conclusion, all the effects that were
significant by model 1 were significant by model 2 as
well, with values and probability measures very similar
to those of model 1. The DLP was estimated by model
2 to have a very significant (P < 0.001) negative effect
on milk yield, with good agreement with the results
reported by Petitclerc et al. (1998). Miller et al. (2000)
reported an addition of 3.2 kg of milk/d in the first 16
wk after parturition, for short-day cows, compared with
long-day cows, as a response to a difference of 8 h in
day length between the groups. Our estimation of a
difference of 1.9 kg of milk/d between winter and sum-

SHORT COMMUNICATION: PREPARTUM SHORT DAY 2781
Table 2. Estimated peak dates for the photoperiod combined effect
on milk yield (kg/d) and fat, protein, and lactose percentage in milk.
Peak dates for the effect on milking cows is compared with the peak
dates for the effect on cows before parturition.
State of cow Milk yield % Fat % Protein % Lactose
Milking May 5 Oct 10 Jan 20 Jun 20
Before parturition Dec 5 Dec 25 Jan 20 Sep 5
mer calving cow refers to 270 d rather than 112 d, and
to combined effects of DLP and DCP for an annual
amplitude of only 4 h in day length between winter
and summer. Therefore, we suggest that our estimation
agrees with this report. Fat and protein, but not lactose
percentage in milk were also negatively related to DLP
(P <0.001). Milk yield (P <0.05), and lactose percentage
(P < 0.001) were negatively affected, whereas protein
percentage was positively affected (P < 0.001), and fat
percentage was not affected by DCP. The effects of the
prepartum DL and DC on milk yield were in contrast
to their effect at the test day, which were both positive.
On the other hand, the effects of DL and DC on fat
and protein percentage were negative whether before
parturition or at the test day. The combined effect of
DL and DC, either for the milking period or for the
prepartum period, on milk yield and composition may
be described as an annual sinusoid response for each
of the dependent variables. On this sinusoid, the peak
date (the date of the maximum positive response) can
be detected.
Table 2 presents the estimated peak dates of milk
yield and composition for either milking or prepartum
cows. Peak dates for milk yield differed by approxi-
mately 6 mo between lactating and prepartum cows,
whereas those for fat and protein percentages were sim-
ilar for the two states. The difference for the dates of
the peak effect on lactose differed by approximately 3
mo between the two states. The contrast in response of
Journal of Dairy Science Vol. 83, No. 12, 2000
milk yield to photoperiod between milking and prepar-
tum cows on one hand, and the similarity in responses
of milk composition between these states, may suggest
different pathways in the response of yield and of com-
position. Long days are associated with elevated prolac-
tin (Miller et al., 2000; Petitclerc et al., 1998). Still the
response of milk yield to these elevated hormone levels
was positive when it occurred in lactation, and negative
when it occurred before parturition. Therefore, it is sug-
gested that the difference in the response of milk and
milk components to the photoperiod should be consid-
ered in factors that are affected by the hormone level.
The difference in day length between the shortest
and the longest day in Israel (latitude of 32°) is approxi-
mately 4 h. Such a difference could account for an addi-
tional 1.9 kg of milk/d, 0.27% fat, and 0.08% protein
for a cow calving in January, compared with a cow
calving in July, in the first 270 d of the lactation. The
summer calving cow is estimated to have 0.03% more
lactose than a winter calving cow.
In contrast to manipulation of the day length during
lactation, which will induce adverse effects on milk
yield and milk composition, decreased day length dur-
ing the prepartum period will result in increased milk
yield and improved milk content in the subsequent lac-
tation. Because this induction period is relatively short,
it could be beneficial to keep prepartum cows in a short-
day regimen.
REFERENCES
Aharoni, Y., A. Brosh, and E. Ezra. 1999. Effects of heat load and
photoperiod on milk yield and composition in three dairy herds
in Israel. Anim. Sci. 69:37–47.
Miller, A.R.E., R. A. Erdman, L. E. Douglass, and G. E. Dahl. 2000.
Effects of photoperiod manipulation during the dry period on
dairy cows. J. Dairy Sci. 83:962–967.
Petitclerc, D., C. M. Vinet, G. Roy, and P. Lacasse. 1998. Prepartum
photoperiod and melatonin feeding on milkproductionandprolac-
tin concentrations of dairy heifers and cows. J. Dairy Sci.
(Suppl. 1)81:251.(Abstr.).