Interrelationships Among Ambient Temperature, Day Length, and Milk Yield in Dairy Cows Under a Mediterranean Climate

Abstract

We examined the effect of calving month (CM) on the production of milk and milk protein by Israeli Holstein dairy cows located in the main climatic zone of Israel during their third and fourth lactations, and found it to be significant. Cows that calved in December produced the highest milk and milk protein yields, and those that calved in June produced the lowest, 92.8% of the maximum. The combined effect of the environmental average temperature and day length accounted for 0.96 of the variability in average milk production during lactation and 0.93 of that in average protein production during lactation. Average milk production was reduced by 0.38 kg/degree C and average protein production was reduced by 0.01 kg/degree C. Elongation of daylight increased average milk production by 1.2 kg/h and average protein production by 0.02 kg/h of daylight. Analysis of the temperature pattern effect on milk and protein yield during lactation indicated that cows at the second month (the pike of their milk yield) are more vulnerable to the negative temperature effect than cows on the ninth month of lactation.

Full text

J. Dairy Sci. 84:2314–2320
 American Dairy Science Association, 2001.
Interrelationships Among Ambient Temperature, Day Length,
and Milk Yield in Dairy Cows Under a Mediterranean Climate
H. Barash,* N. Silanikove,* A. Shamay,* and E. Ezra†
*Institute of Animal Sciences, Agricultural Research Organization
The Volcani Center, Bet Dagan 50250 Israel
+Israel Cattle Breeders Association, PO Box 3015, Qesarya Industrial Park 38900, Israel
ABSTRACT
We examined the effect of calving month (CM) on the
production of milk and milk protein by Israeli Holstein
dairy cows located in the main climatic zone of Israel
during their third and fourth lactations, and found it
to be significant. Cows that calved in December pro-
duced the highest milk and milk protein yields, and
those that calved in June produced the lowest, 92.8%
of the maximum. The combined effect of the environ-
mental average temperature and day length accounted
for 0.96 of the variability in average milk production
during lactation and 0.93 of that in average protein
production during lactation. Average milk production
was reduced by 0.38 kg/°C and average protein produc-
tion was reduced by 0.01 kg/°C. Elongation of daylight
increased average milk production by 1.2 kg/h and aver-
age protein production by 0.02 kg/h of daylight. Analy-
sis of the temperature pattern effect on milk and protein
yield during lactation indicated that cows at the second
month (the pike of their milk yield) are more vulnerable
to the negative temperature effect than cows on the
ninth month of lactation.
(Key words: cow, milk-yield, temperature, light)
Abbreviation key: CM = calving month, MY = milk
yield.
INTRODUCTION
Milk yield (MY) of lactating cows in Israel in the hot
summer and autumn months is lower than during the
winter and early spring (Bar-Anan et al., 1981; Barash
et al., 1996; DeBore et al. 1989; Kahn, 1991; Wolfenson
et al., 1988). High ambient temperature is the main
cause of this phenomenon (Berman et al., 1985;
Wolfenson et al., 1988). Heat stress is well known to
depress milk yield (MY) and appetite in dairy cows
(Armstrong, 1994; Sharma et al., 1988). However, at
Received March 1, 2001.
Accepted June 1, 2001.
Corresponding author: H. Barash; e-mail: Baras@agri.huji.ac.il.
2314
least two major questions regarding seasonal effects
on MY remain open: 1) Do cows at different stages of
lactation respond similarly to heat stress, or are cows
in early lactation more vulnerable? 2) MY is positively
affected by photoperiod (Bilodeu et al., 1989; Dahl et
al., 1997, 2000; Peters et al., 1981; Philips et al., 1989;
Stanisiewski et al., 1985). However, it is not clear
whether the long summer days in Israel (the difference
between the shortest and longest day in Israel is only
about 4 h) have such a positive effect on MY. Analysis
of the data from three herds in the Yizre’el valley in
Israel by Aharoni et al. (1999) suggests that the in-
crease in day length positively affects MY and milk
protein percentage. However, that study, which covered
only three herds in a very small area of Israel, is the
only one to have addressed the question.
The aim of the present study was to answer these
questions by analyzing production records of Israeli
Holstein cows from large herds located throughout
Israel.
MATERIALS AND METHODS
General
The production parameters analyzed in this study
were based on monthly records of milk production and
percentages of milk protein, fat and lactose, as pub-
lished in the Israeli Holstein Herd Book. The records
comprise data from 107 herds of cows in their third and
fourth lactations that calved from 1992 to 1996. Data
were discarded in the following cases: 1) The herd was
located in mountainous areas in which heat stress is
below average, or in the eastern valleys in which cows
are exposed to extreme heat stress; 2) data came from
herds with less than 200 cows, and 3) the calving inter-
val was less than 278 d, because such data might in-
clude cows that were culled due to illness.
The average monthly temperatures were represented
by the averages of the monthly maxima and minima
from 1992 to 1996, as recorded by the Meteorological
Service of Israel, which also provided the data on day
lengths throughout the year (Tables 1 and 2). High
correlations between temperature maximum and mini-

EFFECT OF LIGHT AND TEMPERATURE ON MILK YIELD 2315
Table 1. The monthly temperature (°C) minimum, maximum, and averages for 1992 to 1996.
Minimum Maximum
Month Average SE Average SE Average
Jan 7.1 0.9 17.8 0.4 12.5
Feb 6.9 0.9 18 0.8 12.5
Mar 8.8 0.8 20.2 0.2 14.5
Apr 11.2 0.4 25.1 1.2 18.2
May 15.3 0.5 27.6 0.8 21.5
Jun 19.0 0.1 30.5 0.7 24.8
Jul 21.6 0.2 31.6 0.3 26.6
Aug 21.9 0.1 32.3 0.5 27.1
Sep 20.2 0.5 31.9 0.7 26.1
Oct 16.9 0.6 30.1 1.4 23.5
Nov 11.7 0.7 23.5 0.7 17.6
Dec 8.0 0.8 19.9 1.2 14
mum and heat load in the area monitored in the present
study have been observed by previous studies (Aharoni
et al., 1999, Gat et al., 1998).
Effect of Calving Month
The analytical model used to study the effect of calv-
ing month (CM) was:
Y
ijklmn
= R
i
+ H
j
+ CM
k
+ YR
l
+ PRG
m
+ DIL
n
(1)
+ DIL*CM*PRG+ e
ijklmn
,
where:
Y
ijklm
= daily milk production or protein, fat, and
lactose percentages in milk;
R
i
= effect of region i (Coastal, west Yizre’el
and Negev);
H
j
= effect of herd j;
CM
k
= effect of calving month k;
YR
l
= effect of year I;
PRG
m
= effect of pregnancy m;
DIM
n
= effect of day in milking n; and
e
ijklmn
= random residual effect.
The analysis was carried out according to the SAS
PROC GLM (SAS, 1988). Type III analysis of signifi-
cance was used to determine the significance of produc-
tion parameters. Observation = The average milk pro-
duction (kg/d) of cows in their third and fourth lacta-
tions from the same herd, which calved in the same
month of the same year and whose monthly/daily milk
yield records are for the same time fraction from their
calving day. The number of observations in the dataset
= 103,084.
GLM LSMeans were used to calculate the magni-
tudes of the effects of CM and DIM on MY and the
percentages of milk protein, fat, and lactose.
Journal of Dairy Science Vol. 84, No. 10, 2001
Isolating the Seasonal Effect
To analyze the seasonal effects per se, we organized
the GLM LSMeans given by equation (1) to create vir-
tual lactation curves for each calendar month, covering
all months during lactation. For example, for the virtual
lactation of June, all the recorded data were sorted so
that cows calving in June were grouped as mo 1 of June
lactation, those calving in May and April were grouped
as mo 2 and 3 of June lactation, and so on. This proce-
dure was applied for each calendar month.
The effects of day length and average monthly tem-
perature on the average milk and milk protein produc-
tion in the calendar month were calculated by using
the Jump Multiple Regression Model (JMP, 1995). The
model used for the data analysis was:
Y
I
= AMT
I
+ DL
I
+ e
I
(2)
where:
Y
I
= milk or milk protein production,
AMT
I
= the effect of the average monthly tempera-
ture,
Table 2. Average
1
day length throughout the year.
Month Hours of light
Jan 10:41
Feb 11:33
Mar 12:25
Apr 13:27
May 14:17
Jun 14:43
Jul 14:32
Aug 13:50
Sep 12:54
Oct 11:55
Nov 11:03
Dec 10:35
1
The average day length = the hours of sunlight + 10 min, on d 15
of each calendar month.

BARASH ET AL.2316
Table 3. Effect of calving month on average production of milk pro-
tein, fat, and lactose throughout lactation.
Month Milk (kg) Protein (kg) Fat (kg) Lactose (kg)
Jan 37.2 1.09 1.16 1.68
Feb 37.2 1.09 1.16 1.67
Mar 36.7 1.08 1.16 1.65
Apr 35.9 1.06 1.11 1.61
May 36.1 1.06 1.12 1.61
Jun 35.0 1.03 1.09 1.57
Jul 35.6 1.05 1.10 1.59
Aug 35.7 1.05 1.10 1.60
Sep 36.4 1.08 1.12 1.64
Oct 37.1 1.10 1.16 1.68
Nov 37.4 1.10 1.17 1.69
Dec 37.7 1.11 1.18 1.71
DL
I
= the effect of average monthly day length,
and
e
I
= the random residual effect.
The partial effects of temperature and day length
on milk and milk protein production were analyzed
according to Genizi (1993).
RESULTS
The model in equation (1) accounts for 63.7% of the
variability in MY, 43.7% in protein, 20.0% in fat and
12.9% in lactose (P < 0.0001). Cows calving in December
produced the highest MY and protein, fat, and lactose
contents, whereas those that calved in June produced
the lowest MY and its components (Table 3).
The lactation curves of cows that calved in April,
August, and December are depicted in Figure 1a. In
general, daily MY was higher in cows milking between
February and May comparison with daily MY of cows
at the same lactation part milking between June and
December. Protein yield was also affected by CM (Fig-
ure 1b); however, in this case, the highest daily yield
was obtained if that part occurred between December
and April, and the lowest yield if it occurred between
May and November.
The lactation curves of MY and milk protein content
that were analyzed for the calendar month effects of
April, August, and December (by the virtual curve
method), were more homogeneous than those based on
CM (Figure 2a and b vs. Figure 1a and b). The average
slopes of the linear regression of the last 9 mo of lacta-
tion were the same for the virtual as for the CM curves.
However, the r
2
was larger and the SE in the virtual
curves was half that in the CM curves (Table 4). Ac-
cording to the production curves of milk and protein
during the virtual lactation, the highest MY occurred
in April, and the highest protein yield in February (Fig-
ure 3a and b). The lowest MY (0.90 of the maximum)
Journal of Dairy Science Vol. 84, No. 10, 2001
occurred in September, and the lowest protein yield
(0.88 of the maximum) in August. The combined effect
of average temperature and day length (the model in
equation 2) accounted for 0.96 of the variability in milk
production and 0.93 of that in protein production (P
< 0.0001). The partial effect of average temperature
accounted for 53% of the variability in milk production
and 79% of the variability in protein production,
whereas the partial effect of day length accounted for
38% of the variability in milk production and 14% of
the variability in protein production. The calculated
average effects of hour of day length and temperature
degree on average milk and protein production are pre-
sented in Table 5.
Daily milk and protein yields of cows in the mo 2 and
9 of their virtual lactation at the different calendar
months are presented in Figure 4a and b. The highest
daily MY of cows in their second month of lactation
was in February (47.0 kg/d), and the lowest was in
September (40.8 kg/d, 86.8% of the maximum). The
largest decline in MY occurred from June to September
(summer), and recovery occurred from October to No-
vember (autumn). The highest daily MY of cows in 9
mo of lactation was in mo 5 (May) (30.7 kg/d), and the
lowest was in September to October (27.6 kg/d, 90.0%
Figure 1. Lactation curves of (a) milk yield (kg/d) and (b) milk
protein yield (kg/d) of cows calved in April (䊏 –––䊏), August
(䊉 䊉), and December (▲— ⴢ —▲).

EFFECT OF LIGHT AND TEMPERATURE ON MILK YIELD 2317
Figure 2. Lactation curves of milk yield (MY) (kg/d) and milk
protein yield (kg/d) analyzed according the calendar effect (i.e., the
virtual curve method) of cows calved in April (䊏 –––䊏), August
(䊉 䊉), and December (▲— ⴢ —▲). Virtual lactation curve =
curve that includes all the MY data of the relevant month. For exam-
ple: the virtual lactation curve for June includes all the MY data
recorded for that month such that cows calving in June, May, April,
etc. were grouped as lactation mo 1, 2, 3, etc., respectively. This
procedure was applied for each calendar month.
of the maximum). The protein yield of cows in their
second month of virtual lactation was maximal in mo
1 (January)-February (1.28 kg/d) and minimal in Sep-
tember (1.10 kg/d, 86% of the maximum). Thus, the
protein yield of these cows was highest during the win-
ter and declined sharply during the summer, with a
sharp recovery in November. The maximum protein
yield of cows in their ninth month of virtual lactation
maximum yield was in November to February (0.95 kg/
Table 4. The differences in the slopes (kg/m ± SE) of the decaying parts
1
of the normal and virtual lactation
curves and their r
2
.
Milk yield Milk protein
Normal curve Virtual curve P< Normal curve Virtual curve P<
Slope −2.25 ± 0.44 −2.25 ± 0.22 −0.044 ± 0.014 −0.044 ± 0.008
r
2
0.98 ± 0.009 0.992 ± 0.005 0.002 0.912 ± 0.089 0.964 ± 0.028 0.066
1
The decaying portion of the lactation curve = the part of the lactation curve starting at the second month
of lactation (the lactation peak) and terminating at the ninth month of lactation.
Journal of Dairy Science Vol. 84, No. 10, 2001
Figure 3. Patterns of (a) average daily milk (kg/d) and (b) protein
yields (kg/d) during the virtual lactation.
d), and the minimum protein yield in August to Septem-
ber (0.88 kg/d, 93% of the maximum). The pattern of
protein yield during the ninth month of virtual lactation
differed from that of MY. While the MY slowly increased
during spring and early summer, the protein yield
slowly decreased during late spring and early summer.
To analyze the interaction between environmental
effects and stage of lactation, we calculated the ratio
between MY in the ninth month of the virtual lactation
and MY in the second month (the lactation peak) of the
same virtual lactation for each calendar month, then
expressed it as percentage of the yield in the second
month. The same approach was taken with protein yield
(Figure 5a and b). The following behavior was observed,
1) the largest difference between MY in the ninth and

BARASH ET AL.2318
Table 5. The effects of the average monthly temperature and photoperiod on average daily milk (MY) and
protein yields.
Milk yield Day length effect
(kg/°C ± SE) P< (kg/h ± SE) P<
MY −0.38 ± 0.02 0.0001 1.157 ± 0.870 0.0001
Protein yield −0.011 ± 0.001 0.0001 0.017 ± 0.03 0.0017
second months was during the winter and early spring
(December to April), and it was minimal in the summer,
with the smallest difference in July. 2) The pattern of
protein yield differences the ninth and second months
of the virtual lactation differed from that of MY, mainly
in the following points: protein yield was less affected
by stage of lactation than MY. The average difference
in the protein yield was 23.3 ± 1.1%, where as the differ-
ence in MY was 34.1 ± 0.74%, (P < 0.0001); the difference
between protein yields in the ninth and second months
of virtual lactation was almost constant during the late
summer (July to September), whereas that of MY
was variable.
Figure 4. Yield patterns of (a) milk (kg/d) and (b) protein (kg/
d) in the second (䊉 –––䊉) and ninth (䊏 䊏) months of the
virtual lactation.
Journal of Dairy Science Vol. 84, No. 10, 2001
DISCUSSION
The Seasonal and CM Effects
The rations in Israeli herds of 200 or more cows are
provided, in most cases, by large feed centers. These
rations are quite homogeneous all year round in terms
of components and composition. In the present study,
each monthly figure for milk and milk protein produc-
tion was based on yield data from at least 200 cows.
Over the last 20 yr, about 90% of the Holstein cows
from the Israeli dairy herd have been descendents of
about 10 bulls (J. I. Weller, personal communication),
suggesting minor genetic sire effects. The effect of CM
on MY and protein production could be separated from
that of the cow’s birth month. The effect of birth month
is similar to that of CM only at the first parity, and is
greatly reduced from the second parity onwards (Bar-
ash et al., 1996). Thus, the effects of CM on milk and
protein yields observed in this study can be attributed
to variations in ambient conditions of heat load and
photoperiod rather than to seasonal variations in feed
composition or to genetic variations. In support of this,
97% of the variations in milk production and 93% of
the variations in protein production attributed to the
effect of CM were explained by the variations in temper-
ature and photoperiod throughout the year. The model
in equation 2 indicated that Israeli high-yielding cows
respond negatively to temperature in the early winter
and spring in the Mediterranean zone. Thus, high-
yielding cows became vulnerable to temperatures con-
sidered to be well within the thermoneutral zone, con-
sistent with Silanikove et al., (1997) who showed that
cows sweat at 13 to 14°C. Enhanced MY increases the
overall thermal load on the cows, because of increased
metabolic heat production (West, 1994).
The Antagonisict Effects of Heat and Photoperiod
The maximal average temperature for the year oc-
curred in August and was 14.6°C above the minimum
average temperature, in January and February. The
temperature effect on MY (−0.38 kg of milk/°C) should
produce a difference of −5.55 kg of milk/d between the
milk production during the virtual lactation occurring

EFFECT OF LIGHT AND TEMPERATURE ON MILK YIELD 2319
Figure 5. Patterns of the virtual lactation’s 9-mo daily yields as
percentage of second-month daily yields of (a) milk and (b) milk
protein.
in January and that occurring in August. The difference
in day length between January and August is 3.09 h.
In terms of the light effect (1.157 kg of milk/h), should
result in a difference of 3.65 kg of milk/d between the
virtual lactations occurring in these two months. How-
ever, the actual combined effect of these two opposing
factors was −1.9 kg of milk/d, which accounts for the
difference between milk production during the virtual
lactation in January (36.8 kg of milk/d) and that in
August (34.9 kg of milk/d). Thus, the seasonal effect
appears to be the actual effect out of these two oppos-
ing factors.
Increasing the exposure of cows to light, from less
than 12 h of light/d (short-day photoperiod), to 16–18
h of light/d (long-day photoperiod) enhances milk pro-
duction on average by 2.5 kg/cow per day (Dahl et al.,
2000). Thus, the effect of4hofadditional daylight is
larger than that of 8 h of artificial light supplemen-
tation.
Conventional and Virtual Lactation Curves
The conventional lactation curves differ among them-
selves, due mainly to the effects of temperature and
daylight variations during of the lactation period. By
Journal of Dairy Science Vol. 84, No. 10, 2001
definition, the virtual lactation curves are obtained un-
der uniform temperature and daylight and are there-
fore expected to better represent the response of MY to
the seasonal effect. Indeed, our results suggest that the
latter approach does better reveal the cows’ response
to seasonal effects. Heat and light affected mainly milk
and milk protein yields, and they had less effect on
the pattern of lactation decline. However, the virtual
lactation curves were not completely parallel, possibly
reflecting interactions between the seasonal effects and
the level of milk and milk protein yields. This observa-
tion is consistent with West’s (1994) and the observa-
tion of Silanikove et al. (1997) that high-yielding cows
have increased metabolic heat production and are more
vulnerable to heat stress.
Interactions Between Seasonal Effect
and Milk and Protein Yields
In the case of a constant effect of temperature on MY,
a constant ratio between the yields of the ninth and
second months of virtual lactation throughout the year
would be expected. However, in the case of MY, the
smallest percent was observed during the winter, with
a maximum difference of 20 kg/d between MY in the
ninth and second months of virtual lactation. The pro-
duction of 20 kg of milk would be expected to increase
heat production by ∼9 Mcal/d (calculated according to
NRC, 1989). It seems that the low temperature in the
winter enables a cow in the second month of lactation
to express her full MY potential. This potential cannot
be expressed in the summer because of the high ambi-
ent temperatures. In the ninth month of lactation, cows
produce less heat because of the natural decline in MY.
This situation presumably enables cows to express their
MY potential even in summer.
The production of metabolic heat caused by the pro-
duction of milk could also interfere with the photoperiod
effect. As it has already been argued, MY in the ninth
month of the virtual lactation is less vulnerable to the
seasonal temperature changes than that during the sec-
ond month. This enables the cow in the ninth month
of lactation to better respond with increased MY to the
increased day length, resulting in the smallest ratio
between the MY of the ninth and second the months of
virtual lactation in July (the longest day).
The differences between the patterns of protein yield
percentage and the MY in the ninth month versus the
second month of the lactation could be explained by the
following phenomena: 1) The decline in protein yield
during lactation is slower than that in MY (Auldist et
al., 1998). This slow decline in protein yield results in
ninth-month protein yield being a higher percentage of
that in the second month of lactation. 2) As already

BARASH ET AL.2320
shown, protein production is more sensitive than milk
production to temperature and less to light. These dif-
ferences result in the highest percentage of protein yield
during June, July, August, and September, the hottest
months of the year in Israel, and the almost unchanging
low percentage of protein yield during the period of
December through April.
CONCLUSION
The effect of CM on milk and protein yields observed
in this study could be explained by the direct effects of
temperature and day length. Milk and protein yields
were affected by both temperature and day length: neg-
atively by temperature, −0.38 kg of milk/°C and −0.01
kg of protein/°C and positively by day length, 1.16 kg
of milk/h and 0.02 kg of protein/h. Protein yield was
more sensitive to temperature and less sensitive to day
length than MY. These differences in sensitivity explain
the CM effect on milk protein percentage. The negative
effect of heat load on milk or protein yield during lacta-
tion was not proportional and depended on the quantity
of milk yield.
REFERENCES
Aharoni, Y., A. Brosh, and E. Ezra. 1999. Effect of heat load and
photoperiod on milk yield and composition in three dairy herds
in Israel. Anim. Sci. 69:37–47.
Armstrong, D. V. 1994. Heat stress interaction with shade and cool-
ing. J. Dairy Sci. 77:2044–2050.
Auldist, M. J., B. J. Walsh, and N. A. Thomson. 1998. Seasonal and
lactational influences on bovine milk composition in New Zealand.
J. Dairy Res. 65:401–411.
Bar-Anan, R., and A. Genizi. 1981. The effect of lactation, pregnancy
and calendar month on milk records. Anim. Prod. 33:281–289.
Barash, H., N. Silanikove, and J. I. Weller. 1996. Effect of season of
birth on milk, fat, and protein production of Israeli Holsteins. J.
Dairy Sci. 79:1016.
Berman, A., Y. Folman, M. Kaim, M. Mamen, Z. Hertz, D. Wolfenson,
A. Arieli, and Y. Graber. 1985. Upper critical temperatures and
forced ventilation effects for high-yielding dairy cows in a subtrop-
ical climate. J. Dairy Sci. 68:1488–1495.
Journal of Dairy Science Vol. 84, No. 10, 2001
Bilodeu, P. P., D. Petitclerc, N. St Pierre, G. Pelletier, and G. J. St
Laurent. 1989. Effects of photoperiod and pair-feeding on lacta-
tion of cows fed corn or barley grain in total mixed rations. J.
Dairy Sci. 72:2999–3005.
Dahl, G. E., T. H. Elsasser, A. V. Capuco, R. A. Erdman, and R. R.
Peters. 1997. Effect of a long day photoperiod on milk yield and
circulating concentrations of insulin-like growth factor-1. J. Dairy
Sci. 80:2784–2789.
Dahl, G. E., B. A. Buchanan, and H. A. Tucker. 2000. Photoperiodic
effects on dairy cattle: a review. J. Dairy Sci. 83:885–893.
De Bore, J. A., J. I. Weller, T. A. Gipson, and M. Grossman. 1989.
Multiple analysis of milk and fat yield curves of Israeli Holsteins.
J. Dairy Sci. 72:2143–2152.
Gat, I., Z. Zemel, Z. Hurovitz, I. Flamenbum, and E. Ezra. 1998. The
dairy cattle heat load in the different zones of Israel. Meshek
Habakar va Halav 280:34–46. (in Hebrew).
Genizi, A. 1993. Decomposition of R
2
in multiple regression, with
correlated regressors Statistica Sinica 3:407–420.
JMP User’s Guide: Statistics, Version 3.1. 1995. SAS Inst, Inc.,
Cary, NC.
Kahn, H. E. 1991. The effect of summer decline in conception rate
on the monthly mil production pattern in Israel. Anim. Prod.
53:127–131.
Lacy-Hulbert, S. T., M. W. Woolford, G. D. Nicholas, C. G. Prosser,
and K. Stelwagen, 1999. Effect of milking frequency and pasture
intake on milk yield and composition of late lactation cows. J.
Dairy Sci. 82:1232–1239.
National Research Council. 1989. Nutrient Requirements of Dairy
Cattle. 6th rev. ed. Natl.Acad. Sci., Washington, DC.
Peters, R. R., L. T. Chapin, R. S. Emery, and H. A. Tucker. 1981. Milk
yield, feed intake, prolactin, growth hormone and glucocorticoid
response of cows to supplemented light. J. Dairy Sci. 64:1671–
1678.
Philips, C. J. C., and S. A. Schofield. 1989. The effect of supplementary
light on the production and behavior of dairy cows. Anim. Prod.
48:293–303.
SAS User’s Guide: Statistics, Release 6.03 Edition. 1988. SAS Inst.,
Inc., Cary, NC.
Sharma, A. K., L. A. Rodriguez, C. J. Wilcox, R. J. Collier, K. C.
Bachman and F. G. Martin. 1988. Interactions of climatic factors
affecting milk yield and composition. J. Dairy Sci. 71:819–825.
Silanikove, N., E. Maltz, A. Halevi, and D. Shinder. 1997. Metabolism
of water, sodium, potassium and chlorine by high yielding dairy
cows at the onset of lactation. J. Dairy Sci. 80:949–956.
Stanisiewski, E. P., R. W. Mellenberger, C. R. Anderson, and H. A.
Tucker. 1985. Effect of photoperiod on milk yield and milk fat in
commercial dairy herds. J. Dairy Sci. 68:1134–1140.
West, J. W. 1994. Interactions of energy and bovine somatotropin
with heat stress. J. Dairy Sci. 77:2091–2102.
Wolfenson, D., I. Flamenbaum, and A. Berman. 1988. Dry period
heat stress relief effects on prepartum progesterone, calf birth
weight, and milk production. J. Dairy Sci. 71:809.