Interactions of Heat Stress and Bovine SomatotropinAffecting Physiology and Immunologyof Lactating Cows1

FRANCOIS ELVINGER,2 ROGER P. NATZKE, and PETER J. HANSEN3Dairy Science Department

University of FloridaGainesville 32611

ABSTRACT

During summer, 34 cows receiveddaily injections of placebo or 25 mg ofbST and were placed in a thennoregu­lated or a heat stress environment. Heatstress increased rectal temperatures,respiration rates, and plasma cortisolconcentrations and decreased milk yield.Four of 9 bST-treated cows and none of8 control cows became atactic on the 1std of heat stress. When exposed to heatstress, cows treated with bST experi­enced higher rectal temperaturesthroughout the trials than cows treatedwith placebo. Nonetheless, bST in­creased milk yields in both environ­ments. The major effect of heat stress onimmune function was decreased migra­tion of leukocytes to the mammary glandafter chemotactic challenge. This effectof heat stress was not altered by bST. Insummary, hyperthermia induced by heatstress and associated changes weregreater for cows treated with bST. De­tected effects of heat stress on the im­mune system were few and were notalleviated by bST. Use of bST duringsummer in subtropical climate zones re­quires careful management to avoid

Received June 28, 1991.Accepted September 26, 1991.IFlorida Agricultura1 Experiment Station Journal Se­

ries Number R-Q1695. Supported in part by a grant fromthe Milk Checkoff funds provided by the daily farmers ofFlorida.

2Present address: Veterinary Diagnostic and Investiga­tional Laboratory, College of Veterinary Medicine, Uni­versity of Georgia, Tifton 31793.

3Reprint requests: IFAS 0701, University of Florida,Gainesville 32611-0701.

overexposure of bST-treated cows toheat stress.(Key words: bovine somatotropin, heatstress, immune system, milk yield)

Abbreviation key: DPBS = Dulbecco's PBS,DPBSS = DPBS + 10% gamma-globulin-freehorse serum, HS = heat stress, TR = ther­moregulated.

INTRODUCTION

With the demonstration that bST improvesmilk yield in hot environments (9, 11, 13, 14,19, 21, 24), it is important to understand therole of bST in physiological adjustments toheat stress (HS). There are discrepancies in theliterature as to whether bST increases meta­bolic rate (13, 14, 18, 20); such an effectwould make thermoregulation in hot environ­ments more difficult. Similarly, reports areconflicting as to whether bST increases bodytemperature during the summer or during ex­perimental HS (11, 13, 14, 19, 21, 22, 24).There may also be interactive effects betweenHS and bST on immune function, because HScan affect components of the immune system(6, 8, 12), and bST has been reported to en­hance aspects of the immune system in cattle(3, 4, 7, 10). Additionally, the inhibition ofproliferation caused by culturing lymphocytesat elevated temperature was less for cells ob­tained from bST-treated heifers than for cellsfrom heifers treated with placebo (7). Takentogether, these results indicate the possibilitythat bST could alleviate certain reductions inimmune function caused by HS. The objectiveof the present studies was to characterize phys­iological and immunological effects of HS andbST in lactating cows with particular emphasison whether bST alters effects of HS.

1992 J Dairy Sci 75:449-462 449

450 ELVINGER ET AL.

TABLE 1. Panel of mouse monoclonal antibodies to bovine leukocyte differentiation antigens.

Cell line Isotype Antigen specificity

CH128A IgG] BoTI (CD2), anti-sheep red blood cell receptor (Le., T-cell marker)CACT83B IgM BoT4 (CD4), helper T cell, MHcI Class II restricted T cellsCACTS8C IgG3 BoTS (CD8), suppressor/cytotoxic T cell, MHC Class I restricted T cellsBAQ155A IgG] Receptor on B cells

lMajor hislocompatibiIity complex.

MATERIALS AND METHODS

Reagents

Recombinantly derived methionyl bST(sometribove) was obtained from MonsantoCo. (St. Louis, MO). McConkey agar waspurchased from BBL Microbiology Systems(Becton Dickinson & Co., Cockeysville, MD).Other materials for bacterial cultures werefrom Difco (Detroit, Ml). Mouse monoclonalantibodies (Table 1) to bovine leukocyte dif­ferentiation antigens were purchased fromVMRD Inc. (Pullman, WA). The fluoresceinconjugated anti-mouse IgG was obtained fromUSB Corporation (Cleveland, OR). MouseIgGJ, IgG3, and IgM for isotype controls werefrom Sigma Chemical Co. (St. Louis, MO).The cortisol radioimmunoassay kit was ob­tained from Ventrex Corp.· (Portland, ME).

Staphylococcus aureus vaccine, suspendedin dextran sulfate for adjuvant, was donated byD. L. Watson, CSIRO (Commonwealth Scien­tific and Industrial Research Organization), Ar­midale, New South Wales, Australia. Dulbec­co's PBS (DPBS), Histopaque 1077 (Sigma),oyster glycogen, and p-nitrophenyl-J}-N­acetylglucosaminide were from Sigma Cllemi­cal Co. (St. Louis, MO). Gamma-globulin-freehorse serum was purchased from GmCO(Grand Island, NY) and was added at a con­centration of 10% (vol/vol) to DPBS(DPBSS).

Animals (Experiment 1)

Protocols used were approved by the Uni­versity of Florida Animal Care Committee.Thirty-four lactating Holstein cows were usedin the experiment. Parity ranged from 1 to 8and DIM from 30 to 209. Four of the cows hadbeen purchased in New York 7 wk previously,but these had been placed in an environment

lonrnal of Dairy Science Vol. 75, No.2, 1992

without shade before the experiment began andwere acclimated to Florida conditions. An ad­ditional 3 cows had been purchased in NewYork 7 mo previously and were also adaptedto Florida conditions. During the entire trial,cows were fed a total mixed ration based oncom silage and with a calculated value of 1.54Meal of NErJkg of DM Feed was given twicedaily in quantities sufficient to achieve at least5 to 10% orts. Water was available for adlibitum consumption.

From d 1 to 9 of bST treatment, all cowswere maintained in a thermoregulated (TR)environment, i.e., were maintained in a shedopen on the sides for ventilation and equippedwith an evaporative cooling system (sprinklersand fans) that ran continuously from 0800 to1800 h each day. The shed was open to anadjacent drylot to which the cows had freeaccess. Sixteen randomly chosen cows, as­signed to the placebo group, received dailysubcutaneous injections of 2.5 ml of .075 Msodium bicarbonate; the other 18 cowsreceived daily injections of 25 mg of bST in2.5 ml of sodium bicarbonate. Treatment withbST started on d 1 (August 8) and lasted for 29d. On d 10 of bST treatment (August 17), halfthe cows in each group were randomly as­signed to one of two environmental treatmentgroups: cows either stayed in TR environmentor were placed in an HS environment. Expo­sure to the HS environment was for 15 d. Theenvironmental treatments for cows in the HSenvironment varied. On d 1 of HS, cows wereplaced in a lot with no access to shade or othercooling facilities from 0700 h until the evening(ca. 1700 to 1800 h). At this time, severalcows displayed severe symptoms of HS (seeResults), and all cows were moved back intothe TR environment. From d 2 to 6 of environ­mental treatment, cows were maintained in thelot without access to shade or other coolingfrom 0900 until 1400 to 1500 h. During the

HEAT STRESS AND SOMATOTROPIN 451

remainder of the day, cows were allowed ac­cess to an area covered with shade cloth. Fromd 7 to 15 of environmental treatment, HS cowswere allowed access to shade cloth (but nofans or sprinklers) continuously. Environmen­tal treatment lasted for 15 d, at which time HScows were placed back into the TR environ­ment.

On d -12 and 11 relative to bST treatment(d 2 of environmental treatment), all cowsreceived an intramuscular injection of 1 ml ofStaphylococcus aureus vaccine in dextran sul­fate solution. On d 19 (August 26; d 10 ofenvironmental treatment), 10 m1 of a .1% (wt/vol) oyster glycogen solution in DPBS wereinfused in one front quarter of all cows withlow see (i.e., cows for which Californiamastitis test reactions taken on July 24 andAugust 6, 13, 20, and 26 were repeatedlynegative; 28 cows received infusions). Thepurpose of oyster glycogen infusion was tomonitor treatment effects on the chemotacticmigration of leukocytes into the mammarygland.

AnImals (ExperIment 2)

Seventeen cows, which in the previous ex­periment were in the TR environment (i.e., notsubjected to HS), and which had been treatedwith placebo or bST for 29 d, were maintainedon sodium bicarbonate or bST treatment for 12d more. Cows were maintained in the TRenvironment until the start of Experiment 2 ond 35 of bST treatment (September 11) when all17 cows were exposed to the HS environment.One cow was removed after 1 d of HS becauseof lameness, but the other 16 cows were main­tained in the HS environment for a total of 7 d.Each day of HS, cows were maintained in a lotwithout access to shade or other cooling facili­ties from 0900 until 1200 to 1300 h. Duringthe remainder of the day, cows were also givenaccess to an area covered with shade cloth.

Weather Evaluation

For Experiment 1, dry bulb, wet bulb, andblack globe temperatures and air movementwere monitored at 10-min intervals for theduration of the experiments at an outdoorweather station located 10m from the HSenvironment lot as well as in an adjacent bam

of structure similar to the TR environmentbam. Dewpoint temperature and black globehumidity index were calculated (2). Daily av­erages (from 0900 to 1800 h) of black globetemperatures and of black globe humidity indi­ces were calculated. For Experiment 2, blackglobe temperature was recorded between 1230and 1300 hid in the outdoor lot used for theHS environment.

Milk Weights, Rectal Temperatures,and Respiration Rates

For Experiment 1, milk weights were mea­sured daily at the a.m. and p.m. milkings,starting on d -8 relative to start of bST treat­ment. Rectal temperatures were measured be­tween 1400 and 1500 h on d 9, 10, 11, 13, 14,15, 18, 19, 23, 24, and 29 of bST treatment (d-1, 1, 2, 4, 5, 6, 9, 10, 14, and 15 of environ­mental treatment and d 5 after environmentaltreatment). Respiration rates were measured atthe same time on d 14, 19, 23, and 24 (d 5, 10,14, and 15 of environmental treatment).

For Experiment 2, milk weights were mea­sured daily at a.m. and p.m. milkings. Rectaltemperatures and respiration rates were mea­sured at 0930, 1100, and 1230 h on d 1 ofenvironmental treatment and daily between1200 and 1300 h thereafter.

Blood and Milk Samples

Samples were collected only for Experiment1. Blood samples for harvesting of plasmawere obtained by venipuncture, using evacu­ated heparinized blood collection tubes. Bloodsamples for peripheral blood cell counts andisolation of lymphocytes were collected on d 9,10, 11, 15,24, and 29 after start of bST (d -1,1,2,6, and 15 of environmental treatment andd 5 after environmental treatment). Blood sam­ples for cortisol determination in plasma werecollected on d 9, 10, 11, 13, 15, 18,23,24, and29 after start of bST (d -1, 1, 2, 4, 6, 9, 14,and 15 of environmental treatment and d 5after environmental treatment). Milk. samplesfor bacteriological evaluation were collectedon d -14, -2, 6, 13, and 27 after start of bST.Milk: samples for SCC and evaluation of NA­Gase activity were obtained on d 19, 19.5, 20,21, 22, and 27 relative to start of bST (d 0, .5,1,2, 3, and 8 after oyster glycogen infusion; d10, 10.5, 11, 12, 13, and 18 of HS).

Journal of Daily Science Vol. 75, No.2, 1992

452 ELVlNGER ET AL.

Analysis of Milk samples

Foremilk samples for bacteriological evalu­ation were obtained aseptically. Tenmicroliters of milk per quarter were spread onone-fourth blood agar plate and examined forbacterial growth after incubation for 24 and 48h at 38°C. Bacterial growth was identified byevaluation of colony morphology, hemolysispattern, and Gram staining. The coagulase testwas performed on colonies of Staphylococcusspp. Colonies of Streptococcus spp. weretested for CAMP reaction and esculin hydroly­sis. Intramammary infection was diagnosedwhen samples from two or more consecutivesampling dates were bacteriologically positive,except for the last sampling date (d 27); apositive sample on that date alone was consid­ered diagnostic of intramammary infection.

For the microscopic evaluation of somaticcell numbers, 10 ~ of milk were expanded ona surface area of 1 cm2 of a microscopic slideand stained with methylene blue. Twenty mi­croscopic fields were counted using a 40 xlens of a light microscope, and SCC werecalculated per milliliter of milk.

To determine NAGase in milk, triplicatedeterminations of enzymatic activity weremade on each milk sample. For the assay, 80JlI of 5 roM p-nitrophenyl-p-N-acetyl-glu­cosaminide in .05 M citrate buffer (pH 4.4)were added to 20 ~ of skim milk diluted 1:1in citrate buffer and incubated for 30 min at38°C. The reaction was stopped by the additionof 150 ~ of .5 M carbonate buffer (pH 10.3),and absorbance was read in a MicroplateEL309 autoreader photometer (Biotek Instru­ments, Wmooski, VT) at 405 om.

Cortisol Assay

The radioimmunoassay was performed withan animal cortisol radioimmunoassay kit fromVentrex Laboratories (portland, ME). Crossre­activity data provided by the manufacturer in­cluded values for ll-deoxycortisol (18.3%),cortisone (5.4%), and 17-hydroxyprogesterone(1.1%). Twenty-five microliters of plasma and1.0 mI of 125I-labeled cortisol conjugate weremixed in polypropylene tubes coated with anti­body against cortisol Tubes were incubatedfor 45 min in a 37"C water bath. The reactionmixture was then decanted, and tubes were airdried before measuring radioactivity in a

Iournal of Daily Science Vol. 75, No.2, 1992

gamma-counter. Standards ranged from .625 to500 ng per assay tube and were prepared inhuman cortisol-free serum. Using 25 ~ ofsample, the detection limit was .625 ng/mI. Allsamples of individual cows were assayed induplicate in the same assay. Pools of bovineplasma were assayed at 5 to 30 ~ per tube.Parallel displacement of labeled cortisol andquantitative recovery of added cortisol in low,medium, and high (10, 30, and 78 ng/mI) poolsof bovine plasma were demonstrated. Intraas­say and interassay coefficients of variationwere less than 8%.

Peripheral Blood leUkocyte Count

Blood (20 ~) was diluted with 380 ~ of3% (voVvol) formic acid to lyse red bloodcells. Leukocytes in the resulting suspensionwere counted in a hemocytometer.

Isolation of Mononuclearcells from Peripheral Blood

Blood (10 mI) was diluted 1:3 in DPBSwithout Ca++ or Mg++. The diluted blood waslayered on 10 mI of Histopaque 1077 andcentrifuged for 30 min at 400 x g. Cells in theinterface between Histopaque 1077 and plasmawere collected, washed twice in DPBS withoutCa++ or Mg++, and suspended in DPBSS at 5 x1Q6 cells/mI.

Monoclonal AntibodyStaining for Flow Cytometry

All reagents were maintained at 4°C; centri­fugation steps and incubations were also per­formed at 4°C. Mononuclear cell suspensions(150 ~) were placed into wells of 96-wellround-bottom microtiter plates. Plates werecentrifuged at 350 x g for 3 min, and thesupernatant fraction was discarded. Plates weregently vortexed to loosen cell pellet, and 50 JlIof primary antibody (10 Jlg/ml in DPBSS)were added to appropriate wells (Table 1).Plates were then gently vortexed to mix cellsand antibody and incubated for 15 min. Next,150~ of DPBSS were added to each well, andthen plates were gently vortexed and centri­fuged at 350 x g for 3 min. Supernatant wasdiscarded, and the washing step with DPBSSwas repeated. Subsequently, 50 JlI of secondantibody [anti-mouse IgG from sheep, F(ab'h

HEAT snmss AND SOMATOTROPIN 453

fragment, coupled to fluorescein OOthio­cyanate; dilution was 1:100 in DPBSS] wereadded to each well, and plates were gentlyvortexed to mix cells and antibody.Mononuclear cells were then incubated for 15min in the dark before being washed twicewith DPBSS. The supernatant was discarded,and 40 J1l of 1% (wt/vol) parafonnaldehyde insaline were added to each well. Plates wereplaced into the dark at 4'C until analysis byflow cytometry. hmnediately before analysisby flow cytometry, cells were washed twicewith DPBSS to remove excess paraformalde­hyde solution. The time span between stainingand flow cytometry did not exceed 10 h. Con­trols employed in each staining procedure weresecond antibody alone (for background fluores­cence) and appropriate ootype controls inplace of primary antibodies (for nonspecificbinding).

Flow Cytometry

A Facstar flow cytometer (Becton­Dickinson Immunocytometry Systems, Moun­tain View, CA) was used for examination ofcells. The data on 5000 cells selected asmononuclear cells based on light scatter prop­erties (forward and side scatter) were acquiredin list mode. Cells were evaluated for fluores­cence intensity. Cells with fluorescence inten­sity above intensity for background and iso­type control were considered as positive, andtheir percentage was recorded.

Statistical Analysis

Data analysis was performed using the gen­eral linear models procedure of SAS (17).Analyses were first perfonned using data fromthe entire experimental period, and then addi­tional analyses were performed for the pre-HS,HS, and post-HS periods separately. The math­ematical model in Experiment 1 generally usedwas

= J.1+lli+ bj+Ck+ dl+ I ijkl+ C(abcd)ijldm + en + eIijkln+ £ijldmno

where J.1 is the population mean; aj is environ­mental treatment effect (fixed); bj is the bSTtreatment effect (fixed); Ck is the parity effect

(parity 1 and parity 2 or greater, fixed); dl isthe DIM effect (DIM <140 d and DIM >140 d;fixed); Iijkl represents effects of two- and three­way interactions between environmental treat­ment, bST treabnent, parity, and DIM;C(abcd)ijklm is the effect of cow nested withinenvironmental treatment, bST treatment, andparity and DIM (random; main plot errorterm); en is effect of day of observation (fixedeffect), eIijkln represents interactions between eand other treatments; and Eijldmno is the ran­dom element associated with the observation 0

in the subclass ijklmn. The mathematicalmodel for Experiment 2 did not include theenvironmental treatment effect. For mostanalyses, DIM and interactions with DIM werenot significant sources of variation, and datawere reanalyzed without this effect. Addition­ally, in some models, some of the two- andthree-way interactions for parity or DIM werenot included in the model because of emptycells. Milk yield was analyzed with or withoutlinear and quadratic effects of preinjectionmilk yield and. separately, linear and quadraticeffects of rectal temperature on day ofmeasurement as covariates. Similarly, effectson rectal temperature were analyzed with andwithout linear and quadratic effects of milkyield on day of measurement of rectal tempera­ture as a covariate.

RESULTS

Environmental Conditions

Daily averages (from 0900 to 1800 h) ofblack globe temperatures and of black globehumidity indices for Experiment 1 are shownin Figure 1. Note that data for the HS groupwere recorded at a location exposed to fullsunlight. Thus, recorded data do not corre­spond to the actual environments to which HScows were exposed when they were allowedac<:ess to shade (partial access on d 2 to 6 ofHS and full access on d 7 to 15).

Physical Responses

On d 10 (d 1 of environmental treatment), 4cows from the HS bST group became atactic(i.e., lost muscular coordination or ability tomaintain a standing position) in the late after­noon. One of the 3 atactic cows, cow 1063,

Journal of Dairy Science Vol. 75, No.2, 1992

ELVINGER ET AL.

HS bSTEnd End

i i

28 4September

217 14August

Environmento-othermoregulated.-.heat streIis

-o~50.0.,.----------------------! bST HS:1 Start Start~ ! !eD 45.0Q.

E.sD 40.0.0o

""6t~ 35.0o:0De30.0~.o2' 25.0+----+---.......---4---.......--~----J·0C

454

bSTEnd

!

HSStart

!

Environmento-othermoregulated.-eheat stress

100.0.,.-.-------------------.bST

Starti)(

~ 95.0.5

.:tilU

~ 75.0

b&~ 90.00·-b ED :1~.&: 85.0~D=.0o 0c ""6t 80.0

70.0+----i---~~--~--_I_--~----17

August14 21 28 4

September

Figure 1. Average black globe temperature and black globe humidity index in lhermoregu1ated environments and heatstress (lIS) during Experiment 1. Results are daily averages of readings taken every 10 min from 0900 to 1800 h. Notethat data for the lIS group were recorded at a location exposed to full sunlight Thus, recorded data do not correspond toacn.aJ environments to which HS cows were exposed when !bey were allowed access to shade (partial access on d 2 to 6of HS and full access on d 7 to IS).

Journal of Dairy Science Vol. 75, No.2, 1992

HEAT STRESS AND SOMATOTROPIN 455

was removed from the trial. She was 8 yr old,217 d in lactation, and had a milk yield thataveraged 19.1 kg/d before environmental treat­ment. The rectal temperature at 1500 h of d 1of environmental treatment was 42.4·C. Afterfalling down, she was observed to have amusculoskeletal dysfunction in her hind legs,and she died 2 d later. Two quarters of themammary gland were infected with Strepto­coccus uberis, 1 quarter with Streptococcusdysgalactiae, and 1 quarter with Escherichiacoli. A second cow that collapsed, cow 8740,had a rectal temperature at 1500 h of 43.4·C.Cow 8740 was 56 d in her first lactation andproduced an average of 28.9 kg of millc/d priorto initiation of HS. Before becoming atactic,cow 8740 displayed aggressive behavior. Twoquarters were infected with Strep. dysgalac­tiae. The third cow that became atactic on d 1of HS, cow 8709, experienced rectal tempera­tures of 42.0·C and had pre-HS milk yieldsaveraging 23.0 kg/d. A fourth cow, cow 8712,became atactic for about 30 min during bloodcollection. She experienced a rectal tempera­ture of 42·C and had average pre-HS milkyields of 24.3 kg/d. Overall, pre-HS milkyields averaged 23.8 ± 2.0 kg/d for cows ex­periencing ataxia and 26.9 ± 2.4 kg/d for otherHS bST cows. None of the cows that wereatactic were of New York origin.

As a result of this unexpected occurrence,all cows were maintained in TR conditionsduring the evening and night of d 1 of environ­mental treatment. Additionally, the protocolfor causing HS was changed to reduce theprobability of causing severe HS. From d 2 to6 of environmental treatment, cows were main­tained in the lot without access to shade orother cooling from 0900 until 1400 to 1500 h.At other times, cows were allowed access to anarea covered with shade cloth. At approxi­mately 1730 h of d 17 (d 6 of HS), cow 8740(bST and HS) collapsed again and died within15 min. On this day, cows were given accessto shade at 1415 h, when the rectal temperatureof cow 8740 was 42.00C. From d 7 to 15 ofenvironmental treatment, environment wasagain altered for the HS group so that cowswere allowed continuous access to shade cloth(but no fans or sprinklers). For the remainderof Experiment 1 and for Experiment 2 (inwhich cows received access to shade each dayfrom 1200 until 0900 h the next day), no

clinical signs of ataxia or heat prostration wereobserved. Data of cows 1063 and 8740 up tothe day of removal from treatment were in­cluded in data analysis.

Rectal Temperatures, RespirationRates, and Packed Cell Volume

Rectal temperatures and respiration rateswere measured throughout the environmentaltreatment periods of Experiment 1. For Experi­ment 1, rectal temperatures and respirationrates were higher in HS cows than in un­stressed cows (P < .01; Table 2). The HS cowstreated with bST had higher rectal tempera­tures than HS cows treated with placebo,whereas rectal temperatures in cows main­tained in the TR environment were not af­fected by bST treatment (environment x bST:P = .01; Figure 2). When used as a covariate,there were no significant linear or quadraticeffects of milk yield on rectal temperature.There were no effects of bST or environment xbST on respiration rate (Table 2). Packed cellvolume was affected by an environment x bSTinteraction (Table 2). Heat stress caused anincrease in packed cell volume in cows in­jected with placebo but not in cows treatedwith bST.

In Experiment 2, all cows were exposed toHS. On d 1 of HS, the change in rectal temper­ature over the course of the sampling periodwas greater for bST-treated cows than for cowstreated with placebo (treatment x time; P =.04; Figure 3). Over the entire trial, milk yieldexerted linear (P = .09) and quadratic (P < .01)effects on rectal temperature. Cows treatedwith bST tended to have higher rectal tempera­tures than cows treated with placebo whetherrectal temperature was (P = .09) or was not (P= .07) used as a covariate (Table 3). There wasno effect of bST on respiration rates.

Cortisol ConcentrationsIn Blood Plasma

During the environmental treatment period,seven plasma samples were collected and ana­lyzed for cortisol. Concentrations of cortisolwere higher (P = .02) in HS cows than inunstressed cows (Table 2). There was an envir­onment x bST x day interaction affecting con­centrations ?f cortisol (P = .04); HS onlyaffected cortlsol on d 1 and 6 of environmental

Journal of Daily Science Vol. 75, No.2, 1992

456 ELVINGER ET AL.

TABLE 2. Least squares means of milk yields, rectal tempemtures, respiration rates, packed cell volumes, and plasmacortisol concentrations of cows treated with or without bST and maintained in thennoregulated (TR) or heat stress (liS)environments.

Treatment group

Variableand period

TRPlacebo

TR HSbST Placebo

HSbST SEM

Milk yield, kg/d

Pre-bST, pre-environmental treatment 22.4bST, Pro-environmental treatment 22.8bST, Environmental treatment' 21.7bST, Environmental treatment (adjosted)b,c 22.5

Rectal temperature,d,e 'C 38.6Respiration rate, breaths per minuted,f 68Packed cell volume,d.g % 33.0Cortisol,h,i ng/ml 9.4

23.425.224.724.738.75933.810.0

25.125.020.118.8405

11635.412.7

22.825.621.021.341.1

11332.616.1

1.31.31.1

.15

.6

.1

ltJ!ffects of environment (P = .01) and bST (P = .10).

boata adjusted for pre-bST pre-environmental period milk yield.

~ffects of environment (P < .01), bST (P < .01). and enviroIllIJent x bST (P = .09).

dLeast squares means for bST and environmental treatment period.

emtects of environment (P < .01), bST (P = .04), and environment x bST (P = .01).

fEffeet of environment (P < .01).

8Effeet of environment x bST (P < .01).

hDala subjected to log transformation before statistical analysis. Reported values are the antilogs of the least squaresmeans.

iEffect of environment (P = .02).

treatment, and the increase associated with HSon those days was greater for bST-treated cows(Figure 4).

Milk Yields

In Experiment 1, milk yields were not dif­ferent for treatment groups before bST treat­ment started (fable 2; Figure 5). After initia­tion of bST treatment, milk yields increased in

cows treated with bST, and, to a lesser extent,in cows treated with placebo. During the envi­ronmental treatment period, milk yields werelower in HS cows (P < .01), but HS cowstreated with bST had higher milk yields thanHS cows treated with placebo (after adjust­ment for pre-bST treatment milk yields: bST:P < .01; environment: P < .01; environment xbST: P = .09; Table 2 and Figure 5, panel B).In Experiment 2, milk yields were greater (P =

TABLE 3. Least squares means of milk yields, rectal tempemtures, and respiration rates of cows treated with placebo orbST and maintained in heat stress (liS) environment (Experiment 2).

Variable Treatment

and period Placebo bST SEM P Value

Milk yield, kg/d

Pro-HS 21.1 22.4 1.1 .44HS 19.8 19.6 1.0 .32HS (adjusted)8 19.3 20.1 .02

Rectal temperal1!re,b,c 'C 40.8 41.1 .1 .09Respiration rate.b breaths per minute 127 122 3 .45

8Adjusted for the period before Experiment 1.

b:Least squares means for HS period.

COata are adjusted by using 1inear and quadratic effects of milk yield as a covariate. Meet of bST treatment; P =,(17for unadjusted data, and P = .09 when data are adjusted for milk yield.

Journal of Dairy Science Vol. 75, No.2, 1992

HEAT snmss AND SOMATOTROPIN

43.0bST HS HS bST 41.5Start Start End End

I I I I a-One bST T0' 42.0

~e-ebST

~e..- u 41.0~ 41.0 ~

.a ~

~ .a 40.5It 40.0 l?E X.~ E

Cl 40.0 /C 39.0 thermonogulatood t-

Q O-Ono bST 2Cl .-ebST u'" 38.0 heat otrea Cl 39.5

6-Ano bST '".a.-AbST i37.0 39.07 14 21 28 4 9:30 11:00 1~:30

August September TIme of day

457

Figure 2. Rectal temperatures of cows treated withplacebo or bST (2S mg/d) and cqlOSCd to a thermoregu­lated environment or heat stress (HS) (Experiment 1).During the bST enviromncnta1 treatment period, rectaltemperatures were higher for cows in the HS environment(P < .01), were higher for bST-treated cows (P = .04), andwere affected by environment x bST intmlCtion (P =.01).Results are least SQ\WeS means. 1be pooled standard errorof the means = .17.

.02) for cows treated with bST when data wereadjusted for milk yields prior to Experiment 1(fable 3).

Bacteriological Analysisof Milk Samples

Prior to the start of bST treatment, of 133quarters examined, 6 quarters (from 3 cows)were infected with Staphylococcus spp. and 7quarters (from 3 cows) with Streptococcus spp.During the bST treatment period (including theenvironmental treatment period), 1 of 32 quar­ters (one cow) in the TR placebo group(Staphylococcus spp.), 3 of 36 quarters (2cows) in the TR bST group (1 Staphylococcusspp., 2 Streptococcus spp., 1 Gram-negative), 1of 31 quarters (one cow) in the HS placebogroup (Staphylococcus spp.), and 2 of 34 quar­ters in the HS bST group (cows 1063 and8740) became infected. Throughout the trial, 2of 8 TR placebo, 4 of 9 TR bST, 2 of 8 HSplacebo, and 4 of 9 HS bST cows experiencedat least 1 infected quarter (differences in inci­dence of mastitis between treatments primarilyreflects mastitis that preexisted before applica­tion of treatment).

SOmatic cells and NAGase In Milk

Data for sec and NAGase were log­transformed for statistical analysis. Peak sec

Figure 3. Rectal temperatures of cows treated withplacebo or bST on the 1st d of exposure to heat stressenvironment (Experiment 2). Rectal temperatures tendedto be higher in cows treated with bST (P = .09). Addition­ally, the change in rectal temperature over time wasgreater for bST-treated cows than for placebo-treated cows(bST x time: P = .04). Results shown are least squaresmeans ± standard error of the means.

70bST HS HS bST

Start Start End End60 I I I I

E~ 50..s.(; 40D

1:a 30u

" thermoregulotedE 20.. a-Ona bST

" .-ebSTii:

10

__A-Ana bST·-.bST

07 14 21 28 4

August September

Figure 4. Concentrations of cortisol in plasma of cowstreated with placebo or bST (2S mg/d) and exposed to a1hcnnoregulated environment or heat stress (HS). Antilogsto least squares means of log-transformed data are repre­sented. Standard error of the mean of log cortisol =.223.During the bST environmental treatment period, cortisolcoocenJrations were higher for cows in the HS environ­ment (P = .02). There were no differeDCCS because of bSTtreatmeot or the interaction between environment and bST.1bere were interactions between day of treatment x envir­onment (P < .01), day of treatment x bST (P < .01), andday of treatment x environmcntal treatment x bST treat­ment (P = .(4).

lomnal of Dairy Science Vol. 75, No.2, 1m

458 ELYINGER ET AL.

A30

bST HSstart start

I I

-~ 2501~-"0li0>,

~ 20°ethermorwgulat8d

O-Ono bST·-.bST

157 14 21

August

HS bSTEnd End

I I

28 4September

B.bSTEnd

I

HSEnd

I

HSStart

I

30.,..--------------------bST

startI

25

20

heat strae.-.no bST"'-"'bST

15+----+-----li----+-----f------li------J7

August14 21 28 4

SeptemberFigure 5. Mean milk yield of cows treated with placebo or bST (15 mg/d) and exposed to a thc:rmoregulated

environment or heat stress (liS). Panel A represents cows maintained in the thc:rmoregulated environment duringexperimental treatment period, and panel B represents cows maintained in the HS environment during the environmentaltreatment period. For the pro-bST treatment period and for the bST period before environmental treatment, least squaresmeans for milk yield were not different (P > .10). During the bST environmental treatment period, milk yields werehigher for cows in the thermoregulated envirooment (P =.01) and tended to be hiF for bST-treated cows (P - .10).When milk yields for the bST enviromnenta1 treatment period were adjusted for pro-bST milk yields (data not shown; seeTable 2), milk yields were higba for cows in the thmooregulated environment (P < .01) and for bST-treated cows (P <.01). There was also a tendency (P = .09) for an environment x bST interaction for adjusted data.

Journal of Dairy Science Vol. 75, No.2, 1992

HEAT STRESS AND SOMATOTROPIN 459

Figure 6. Somatic cell counts in milk from mammaryquarters infused with 10 m1 of a .1% (wt/vol) oysterglycogen solution at time 0 from cows treated withplacebo or bST and exposed to a thermoregulated environ­ment or heat stress (lIS). Somatic cell counts from cowsmaintained in a thermoregulated environment increased toa higher level 12 h postinjection than sec from HS cows(time x environment; P =.02). There were no significanteffects of bST treatment or environment x bST. Resultsshown are antilogs to least square means of log­transformed data ± antilog of standard error of the meansof log-transformed data.

after oyster glycogen infusion were higher forcows in thennoregulated environment than forcows in HS environment (40.8 x 106 vs. 17.1x 106, SEM =3.9 x 1()6 somatic cells/ml; P =.01) and were 38.4 x 106, 43.5 x 106, 24.1 x106, and 12.2 x 106 somatic cells/ml for 1Rplacebo, TR bST, HS placebo, and HS bSTcows, respectively. Differences in see be­tween 1R and HS cows were present only at12 h after oyster glycogen infusion (environ­ment x time postinfusion: P =.02; Figure 6).The NAGase content was higher in multi­parous cows (17.3 ± 1.5 nmol/ml per min) thanin primiparous cows (13.1 ± 1.0 nmol/ml permin; P < .05). There were no main effects ofenvironment or bST on NAGase in milk(results not shown). Pearson's coefficient ofcorrelation for see and NAGase was .58 (n =107; P < .01).

TR-no bSfITR-bSf~

HS-no bSfFJHS-bSfI

HS PostHS

PreHS

o

20

20,---------------~1. Periphe~1 Blood

([;' 1. U1COCyte81

614...- 12X"'-'" 10

E II

~.'ii 4U 2

40

10CD2

10

30

Figure 7. Number of leukocytes, CD2+ lymphocytes (Tcells), and B lymphocytes in peripheral blood from cowsIreated with placebo or bST and exposed to a thennoregu­lated (TR) environmeot or heat slress (lIS). Bars representleast squares means of unttansformed data (± standarderror) for cows sampled once in the period prior toenvironmental treatment; on d I, 2, 6, and 15 of eoviron­mental treatment; and on d 5 after the end of environmen­tal treatment. There were no differences between treatmentgroups in total number of leukocytes in the period prior toenvironmeotal treatmeot or in the environmental treatmentperiod. Mer the environmental period, leukocyte countsin bST-Irealed cows were lower than in placebo-treatedcows (P = .06). For CD2+ lymphocytes and B lympho­cytes, there were no significant differences between treat­ment groups in any period.

0-0 thermoregulated environment.-.heat stress environment

o 12 24 48 72 192

lime Postinjection (h)

50

u 20u

'"

~40

E;;;-I 30a

Leukocyte Counts andLymphocyte Subpopulatlons

These data were log-transfonned beforeanalysis. Least squares means of nontrans­fonned data are shown in Figures 7 and 8.Leukocyte counts in peripheral blood weregenerally unaffected by treatment, except that

bST-treated cows had lower numbers of leuko­cytes than cows treated with placebo at 29 d ofbST treatment (P =.06; Figure 7). This differ­ence was not caused by hemodilution becausepacked cell volume on d 29 averaged 32.4%for bST-treated cows and 33.9% for cows

Journal of Dairy Science Vol. 75, No.2, 1992

DISCUSSION

As has been shown previously (2,5, 8, 11,13, 14, 16.23,24). HS caused an increased in

treated with placebo. There were no differ­ences detected because of environmental orbST treatment or their interaction for propor­tions of CD2+ cells. CD4+ cells. CD8+ cells.CD4+:CD8+ cell ratio. or for the proportion ofB cells (Figures 7 and 8).

body temperature, respiration rate, and plasmaconcentrations of cortisol while causing adecrease in milk yield. A major finding ofExperiments 1 and 2 was that the hyperthermiacaused by HS was of greater magnitude forcows treated with bST than for cows treatedwith placebo. Respiration rates in HS cowswere not affected by bST, but this probablyreflects the fact that, as rectal temperature in­creases, respiration rates increase to a maxi­mum and then decrease (1). In Experiment 1,treatment with bST was also associated withoccurrence of ataxia and death upon exposureto severe HS. Heat stress did not cause ataxiaand mortality in Experiment 2, but the protocolof Experiment 2 was designed to prevent suchresponses.

The present results that bST increased rectaltemperatures in experimentally HS cows isconsistent with reports that bST-treated cowshad higher body temperatures than controlcows in lactation trials conducted in hot envi­ronments (21, 22. 24). In contrast, no effect ofbST on HS-induced hyperthermia was seen inseveral experiments conducted in environmen­tal chambers (11, 13, 14) and in one Floridastudy conducted in free stalls (19). One possi­ble explanation for these different results isthat bST-treated cows are more likely to be­come more susceptible to HS when the stressis severe. For example, average rectal tempera­tures of HS cows in the present experimentswere often over 4O.soC, whereas rectal temper­atures in experiments in which bST did notincrease rectal temperatures were 40 to 4O.soCor less (11, 13, 14, 19). Also, HS in the currentstudy was caused in large part because ofincident solar radiation, whereas this source ofenvironmental heat was not a major factor instudies in which bST did not increase rectaltemperature (11, 13, 14, 19).

The ataxia and deaths associated with thecombination of bST and HS have not beenreported previously. The small numbers ofanimals involved make it possible that theincreased occurrence of these severe symptomsfor cows treated with bST was due to chance.Nonetheless, the higher incidence of ataxia anddeath in the bST-treated group is consistentwith the finding that bST-treated cows hadhigher rectal temperatures than cows treatedwith placebo. It is not known whether vaccina­tion with Staph. aureus vaccine contributed to

ELVlNGER ET AL.

TR-no bSTITR-bSTI!\I

HS-no bSTfJHS-bSTI

Pre HS PostHS HS

1.0

0.0

460

50CD4

40

;m

~- 20

10

0COB

20

"'"'Cle- 10

4.0CD4/CDB

Figure 8. Percentage of CD4+ and CD8+ lymphocytesand the CD4+:CD8+ ratio in peripheral blood from cowstreated with placebo or bST and exposed to a thermoregu­lated (TR) enviromnent or heat stress (lIS). Bars representleast squares means of untransformed data (± standarderror) for cows sampled once in the period prior toenviromnental treatment; on d 1.2,6. and 15 of environ­mental treatmcot; and at d 5 after the end of enviromncotaltreatment. There were no significant differences betweentreatment groups in any period.

Journal of Dairy Science Vol. 75. No.2. 1992

HEAT STRESS AND SOMATOTROPIN 461

the response to HS. Effects of vaccine seemremote, however, given the fact that all cowswere vaccinated and that the first vaccinationoccurred 22 d before HS.

The mechanism by which bST increased thesusceptibility of cows to HS is unclear. It isunlikely that the increased susceptibility is dueto the effect of bST on milk. yield. because theabsolute difference in milk yield between con­trols and bST-treated cows at the initiation ofHS was very small. It is unclear whether bSTincreases metabolic rate (13, 14, 18, 20). If so,bST may increase the degree of hyperthermiain HS cows because metabolic heat productionis increased. It is also possible that bSTreduced dissipation of heat from the cowthrough some unknown action. In anotherstudy (13), however, increased heat productionin bST-treated cows was accompanied by anassociated increase in heat loss.

In the present study, packed cell volumewas greater for HS cows than for cows in theTR environment only if they were not alsoreceiving bST. Changes in hematocrit presum­ably reflect the balance between HS-associatedchanges in water intake and HS-associatedchanges in water utilization (increased waterfor evaporative heat loss versus decreased wa­ter if milk yield is depressed). Perhaps bSTaltered some of these relationships between HSand water balance. Treatment with bST hasbeen reported to increase water intake (11, 13)and decrease hematocrit (11, 24) under bothcool and hot conditions.

Concentrations of cortisol in plasma wereelevated in HS cows, particularly if cows werealso receiving bST. The HS-associated in­crease in concentrations of cortisol was notsustained, however, and greatest differenceswere seen on d 1 and 6 of HS. It had beenreported previously that concentrations ofplasma cortisol return to basal levels afterchronic HS (5), although other studies reportedcontinuously elevated cortisol levels in cows inHS for a duration similar to that for the currentstudy (16, 23). West et al. (22) found no effectof bST on cortisol concentrations in cowsmaintained in a hot, humid environment.

A major hypothesis of the present studywas that bST would reduce effects of HS onimmune function. In a previous study (7), theinhibition of mitogen-induced proliferation oflymphocytes caused by culture at elevated

temperature (42°C) was less for cells frombST-treated heifers than for cells from controlheifers. The major effect of HS on immunefunction in the present study was a decrease inmigration of leukocytes to the mammary inresponse to a chemotactic challenge of oysterglycogen. In a previous study, random migra­tion and chemotaxis by polymorphonuclearleukocytes in vitro were inhibited by culturetemperatures of 42°C, and chemotaxis of poly­morphonuclear leukocytes from HS cows wasalso depressed (8). These effects on function ofpolymorphonuclear leukocytes could explainthe decreased response to oyster glycogen inthe present study as well as the increasedincidence of mastitis in the summer reported ina Florida study (15). In spite of the literatureindicating effects of bST on immune function(3,4, 7, 10), there was only one effect of bSTon immune function and no bST X environ­ment interactions affecting immune function.This effect of bST was a reduction in numbersof peripheral blood leukocytes after treatmentwith bST for 29 d. This latter finding, whichcorresponds to previous results in which bSTreduced peripheral blood leukocyte counts inheifers treated for 100 d with bST (8), was notassociated with hemodilution in the presentstudy. There was no detectable effect, howev­er, of bST on percentages of peripheral bloodB cells, CD2+ cells (T cells), CD4+ cells,CD8+ cells, or on chemotactic responses ofpolymorphonuclear leukocytes in the mam­mary.

In spite of the fact that bST increased rectaltemperatures of HS cows, bST increased milkyield in HS cows of Experiment I, a resultsimilar to other studies conducted under HSconditions (9, 11, 14, 19, 21, 24). Thus, bSTdoes offer the potential for increasing milkyield in hot environments. The fmding thatcows treated with bST were more adverselyaffected by HS than cows treated with placeboemphasizes the importance of ensuring that useof bST during summer in subtropical climatezones be implemented in conjunction withgood management to avoid overexposure toHS.

ACKNOWLEDGMENTS

The authors wish to thank Sandy H. Meanand Ray A. Bucklin for collection of climatic

Journal of Daily Science Vol. 75, No.2, 1992

462 ELYINGER ET AL.

data; the Flow Cytometry Laboratory of theInterdisciplinary Center for Biotechnology Re­search, University of Florida. especially DavidJ. Hurley, Melissa Chen, and Neil Benson forassistance with data collection and analysis atthe flow cytometer; R. Luzbel de la Sota,Manuel F. Lander Chacin, Marie V. Leslie,and Boon G. Low for assistance with collec­tion and processing of samples; Joyce Hayen,Paulette Tomlinson, and H. H. Head for con­ducting cortisol assays; Monsanto Co. for pro­viding bST; and D. L. Watson, CSIRO, Armi­dale, New South Wales, Australia for donatingmaterials for Staphylococcus vaccination.

REFERENCES

I Bianca, W., and I. D. Findlay. 1962. The effect ofthermally-induced hypemea on the acid-base status ofthe blood of cows. Res. Vet Sci. 3:38.

2 Buffington, D. E., A. ColIazo-Arocho, G. H. Canton,D. Pitt, W. W. Thatcher, and R I. Collier. 1981.Black globe-humidity index (BGHI) as comfort equa­tion for dairy cows. Trans. Am. Soc. Agric. Eng. 24:711.

3 Burton, I. L., B. W. McBride, B. W. Kennedy, J. H.Burton, T. H. Elsasser, and B. Woodward. 1991.Influence of exogenous bovine somatotropin on theresponsiveness of peripheral blood lymphocytes tomitogen. J. Dairy Sci. 74:916.

4 Burton, I. L., B. W. McBride, B. W. Kennedy, I. H.Burton, T. H. Elsasser, and B. Woodward. 1991.Serum immunoglobulin promes of dairy cows chroni­cally treated with recombinant bovine somatotropin. J.Dairy Sci. 74:1589.

5 Christison, G. I., and H. D. Johnson. 1972. Cortisolturnover in heat-stressed cows. J. Anim. Sci. 35:1005.

6 Downing, J. F., and M. W. Taylor. 1987. The effect ofin vivo hyperthermia on selected lymphokines in man.Lymphokine Res. 6: 103.

7 Elvinger, F., P. J. Hansen, H. H. Head, and R P.Natzke. 1991. Actions of bovine somatotropin onpolymorphonuclear leukocytes and lymphocytes incattle. I. Dairy Sci. 74:2145.

8 Elvinger, F., P. J. Hansen, and R. P. Natzke. 1991.Modulation of function of bovine polymorphonuclearleukocytes and lymphocytes by elevated temperaturein vitro and in vivo. Am. J. Vet Res. 52:1692.

9 Elvinger, F., H. H. Head, C. J. Wilcox, R P. Natzke,and R G. Eggert. 1988. Effects of administration ofbovine somatotropin on milk yield and composition. I.Dairy Sci. 71:1515.

10 Heyneman, R., C. Burvenich, M. Van Hoegaerden,and G. Peeters. 1989. Influence of recombinantmethionyl bovine somatotropin (rBSn on blood neu­trophil respiratory burst activity in healthy cows. J.Dairy Sci. 72(Suppl. 1):349.(Abstr.)

II Johnson, H. D., R. Li, W. Manalu, K. I. Spencer­lones, B. A. Becker, R I. Collier, and C. A. Baile.

lournal of Dairy Science Vol. 75, No.2, 1992

1991. Effects of somatotropin on milk yield and phys­iological responses during summer farm and hotlabo­ratory conditions. I. Dairy Sci. 74:1250.

12 Kelley, K. W., C. A. Osborne, I. F. Evermann, S. M.Parish, and C. T. Gaskins. 1982. Effects of chronicheat and cold stressors on plasma immunoglobulin andmitogen-induced blastogenesis in calves. J. Dairy Sci.65:1514.

13 Manalo, W., H. D. Iohnson, R-Z. Li, B. A. Becker,and R. J. Collier. 1991. Assessment of thermal stressof somatotropin-injected lactating Holstein cowsmaintained under controlled-laboratory thermoneutral,hot and cold environments. J. Nutr. 121:2006.

14 Mohammed, M. E., and H. D. Iohrtson. 1985. Effectof growth hormone on milk yields and related physio­logical functions of Holstein cows exposed to heatstress. I. Dairy Sci. 68: 1123.

15 Morse, D., M. A. DeLorenzo, R P. Natzke, and D. RBray. 1988. Characterization of clinical mastitisrecords from one herd in a subtropical environment. J.Dairy Sci. 71:1396.

16 Roman-Ponce, H., W. W. Thatcher, and C. J. Wilcox.1981. Hormonal interrelationships and physiologicalresponses of lactating dairy cows to a shade manage­ment system in a subtropical environment. Therioge­nolO~ 16:139.

17 SAS User's Guide: Statistics, Version 5 Edition.1985. SAS Inst., Inc., Cary, NC.

18 Sechen, S. I., D. E. Bauman, H. F. Tyrrell, and P. I.Reynolds. 1989. Effect of somatotropin on kinetics ofnonesterifJed fatty acids and partition of energy, car­bon, and nitrogen in lactating dairy cows. J. Dairy Sci.72:59.

19 Staples, C. R, H. H. Head, and D. E. Darden. 1988.Short-term administration of bovine somatotropin tolactating dairy cows in a subtropical environment. J.Dairy Sci. 71 :3274.

20 Tyrrell, H. F., A.C.G. Brown, P. I. Reynolds, G. L.Haaland, D. E. Bauman, C. J. Peel, and W. D. Stein­hour. 1988. Effect of bovine somatotropin on metabo­lism of lactating dairy cows: energy and nitrogenutilization as determined by respiration calorimetry. I.Nutr. 118:1024.

21 West, J. W., B. G. Mullinix, I. C. Johnson, Jr., K. A.Ash, and V. N. Taylor. 1990. Effects of bovine somat­otropin on dry matter intake, milk yield, and bodytemperature in Holstein and Iersey cows during healstress. J. Dairy Sci. 73:2896.

22 West, I. W., B. G. Mullinix, and T. G. Sandifer. 1991.Effects of bovine somatotropin on physiologicresponses of lactating Holstein and Jersey cows duringhot, humid weather. J. Dairy Sci. 74:840.

23 Wise, M. E., D. V. Armstrong, I. T. Huber, R Hunter,and F. Wiersma. 1988. Hormonal alterations in thelactating dairy cow in response to thermal stress. I.Dairy Sci. 71:2480.

24 Zoa-Mboe, A., H. H. Head, K. C. Bachman, F. Bac­cari, Ir., and C. I. Wilcox. 1989. Effects of bovinesomatotropin on milk yield and composition, dry mat­ter intake, and some physiological functions of Hol­stein cows during heat stress. I. Dairy Sci. 72:9CY7.