Effects of exposure to high ambient temperature and dietary protein level on performance of multiparous lactating sows1

D. Renaudeau*, N. Quiniou, and J. Noblet*,2

*Station de Recherches PorcinesInstitut National de la Recherche Agronomique, 35590 Saint Gilles, France and Institut Technique du Porc, 35650 Le Rheu, France

ABSTRACT: Fifty-nine multiparous Large White Landrace sows were used to determine the effects of high ambient temperature and level of dietary heat increment on lactation and reproductive performance. During a 28-d lactation and the 14-d postweaning period, ambient temperature was maintained constant at 20 or 29C. Experimental diets fed during lactation were a control diet (NP; 17.6% crude protein) and two low-protein diets obtained by reduction of CP level (LP; 14.2% CP) and both reduction of CP and addition of 4% fat (LPF; 15.2% CP); the NE:ME ratio was 74.3, 75.6, and 75.8% for NP, LP, and LPF diets, respectively. All diets provided 0.82 g of digestible lysine/MJ of NE, and ratios between essential amino acids and lysine were above recommendations. During the ad libitum period (i.e., between d 7 and 27 of lactation), ADFI and NE intake decreased, respectively, from 7.63 to 4.22 kg and 82.0 to 43.3 MJ of NE when the temperature increased from 20 to 29C. Exposure to 29C reduced litter BW gain (2,152 vs 2,914 g/d) and increased lactation BW

loss (34 vs 16 kg) but increased postweaning BW gain (2 vs 4 kg). Expressed per kilogram of feed intake, water intake increased from 4.0 to 7.5 L between 20 and 29C. Compared with the NP diet, low-CP diets (LP or LPF) maintained performance and reduced total N excretion (22.5%) in lactating sows at thermoneutrality and attenuated the negative effects of high temperature on NE intake and BW loss (40.9 vs 47.9 MJ of NE/d and 41 vs 30 kg for NP and LP + LPF diets, respectively). Temperature and diet composition did not affect the reproductive performances, and the mean weaning-to-estrus interval was 4.9 d. The respiratory rate and skin, udder, and rectal temperatures increased markedly at 29C (105 vs 33 breaths per min and 37.8 vs 36.1C, 38.9 vs 38.2C, and 39.5 vs 38.8C, respectively) without any effect of diet. In conclusion, low heat increment diets (i.e., with reduced CP/NE ratio) did not affect lactation performance of sows at thermoneutrality and attenuated the effects of high ambient temperature on energy intake and BW loss. At any ambient temperature, N excretion can be markedly reduced.

Key Words: Dietary Protein, Feed Intake, Heat Stress, Lactation, Sows 2001 American Society of Animal Science. All rights reserved. J. Anim. Sci. 2001. 79:12401249

Introduction
During the last decade, selection for improved prolicacy of sows has resulted in increased litter size at weaning and a concomitant increase of the sows nutrient requirements. Simultaneously, selection for leaner pigs at slaughter weight was associated with a decrease of body reserves and voluntary feed intake. Consequently,

The authors gratefully acknowledge Degussa-Huls (Hanau, Ger many) and Ajinomoto Eurolysine (Paris, France) for their nancial support; Ajinomoto Eurolysine (Paris, France) for measurement of amino acid contents in feeds; Institut Technique du Porc (Paris, France) for the grant of D. R.; J. Van Milgen for critical evaluation of the manuscript; and P. Bodinier, S. Daniel, S. Dubois, J. Gauthier, and H. Renoult for their technical assistance. 2 Correspondence: E-mail: noblet@st-gilles.rennes.inra.fr. Received June 19, 2000. Accepted January 5, 2001.

the nutritional decit of lactating sows has been accentuated with subsequent excessive weight loss during lactation and extended weaning-to-estrus interval or successful mating (Whittemore, 1996). Apart from genetic selection, high ambient temperature is one of the many factors that may inuence the voluntary feed intake of lactating sows (OGrady et al., 1985). In summer, lactating sows are frequently exposed to high ambient temperatures and their feed intake is reduced (Black et al., 1993; Prunier et al., 1997; Quiniou and Noblet, 1999). This decreases heat production due to the thermic effect of feed (TEF). The reduction of energy intake is more important than the change in energy requirement due to decreased milk yield, which results in an accentuated nutritional decit. Performance of sows exposed to high ambient temperatures can be maintained by increasing heat loss to the environment (McGlone et al., 1988). An alternative approach is to reduce heat production of the sow. The TEF

1240

High temperature and diet composition

1241

of digestible fat, starch, protein, and dietary ber averages 10, 18, 42, and 42% of the ME content, respectively (Noblet et al., 1994). Therefore, diets with reduced crude protein (CP) level and(or) fat addition result in lower heat production (Noblet et al., 1987; Le Bellego et al., 1999) and seem to be better tolerated under hot climatic conditions (Stahly et al., 1979). The objective of our study was to evaluate the effects of diets with reduced TEF on performance and feeding behavior of lactating sows and performance of piglets when exposed to high or conventional ambient temperatures. The present paper will focus on performance of sows.

Table 1. Composition of experimental diets

Item Ingredient, g/kg Wheat Corn Soybean meal Wheat bran Sugar beet molasses Soy oil L-Lysine HCl DL-Methionine L-Threonine L-Tryptophan L-Isoleucine L-Valine Dicalcium phosphate Calcium carbonate Salt Vitamins and trace minerals mixture Chemical composition, % (as fed) DM Ash Crude protein Crude fat Starch Crude ber NDF ADF ADL Lysine Methionine + cystine Threonine Tryptophan Isoleucine Leucine Valine Histidine Phenylalanine Digestibility coefcients, %b DM N Energy Energy value, MJ/kgbc Digestible energy Metabolizable energy Net energy NE/ME, % Digestible lysine, g/MJ of NEd NP 350 353 207 30 20 1.7 0.4 0.9 LP 403 406 91 30 20 5.5 1.5 2.5 0.6 1.3 2.2 16 11 5 5 87.4 5.0 14.2 2.6 49.5 2.0 9.4 2.3 0.4 0.95 0.60 0.67 0.20 0.61 1.07 0.80 0.33 0.60 89.5 88.9 90.9 14.4 14.0 10.6 75.6 0.82 LPF 364 366 129 30 20 40 5.3 1.7 2.7 0.6 1.4 2.3 16 11 5 5 88.1 5.5 15.2 6.1 44.7 1.9 8.9 2.4 0.4 1.04 0.64 0.74 0.21 0.66 1.15 0.84 0.35 0.65 89.6 90.6 90.9 15.2 14.8 11.2 75.8 0.82

Materials and Methods

bred Large White Landrace sows divided into 11 groups of four to six animals were used in the experiment. Two farrowing rooms were used for the experiment, one set at 20C (thermoneutral) and one set at 29C (hot). Each group of sows was maintained either at 20C or at 29C during the 28-d lactation period and the subsequent 14d postweaning period. The ambient temperature was maintained constant over the 24-h period. Within each group, sows were allocated to three dietary treatments (one or two sows per treatment) in order to obtain, for the total experiment, similar parity number and body condition and BW at farrowing for the six treatments. The three experimental diets differed in their NE:ME ratios; differences were achieved by a reduction of the CP level (LP for low protein level) or both reduction of CP and addition of fat (LPF for low-protein and fatadded diet); these diets were compared to a normal CP level diet (NP). The three diets supplied the same levels of lysine (0.82 g digestible lysine/MJ of NE), and levels of digestible essential amino acids relative to digestible lysine were similar and higher than values recommended by Dourmad et al. (1991). Mineral and vitamin contents met or exceeded the recommendations of INRA (1989). Composition, chemical characteristics, and nutritional values of the diets are given in Table 1. Care and use of animals met the requirement of the certicate of Authorization to Experiment on Living Animals, no. 04739 (certied by the French Ministry of Agriculture to J. Noblet). Animal Management. In order to standardize body condition at farrowing, pregnancy feed allowance was calculated for each sow according to the model proposed by Dourmad et al. (1997) and described previously by Quiniou and Noblet (1999). The gestation diet contained 13.0 MJ of DE/kg, 13% CP, and 0.55% lysine. Fourteen days before farrowing, the animals were moved to the farrowing rooms and kept at 24C with a photoperiod of 14 h of articial light (0830 to 2230) and a minimum ventilation rate of 25 m3/h per sow. Within each temperature, sows were allotted to the three dietary treatments according to BW and backfat thickness on d 100 of gestation. Farrowing room conditions and the feeding plan at the beginning of lactation have been

Experimental Design. Fifty-nine multiparous, cross-

16 11 5 5 87.3 5.1 17.6 2.5 43.2 2.2 9.3 2.6 0.4 0.96 0.60 0.69 0.20 0.70 1.40 0.82 0.45 0.82 89.9 90.0 91.7 14.7 14.2 10.5 74.3 0.82

a NP = normal protein, LP = low protein, and LPF = low protein and added fat. b Values measured on nonpregnant adult sows (four measurements per diet). c Adjusted for measured DM on the pooled samples. ME was estimated from DE content and the ME/DE ratio was assumed to be equal to those measured in growing pigs fed the same diets (96.2, 97.2, and 97.2% for NP, LP, and LPF diets, respectively; Lebellego et al., unpublished data). Net energy was estimated from measured digestible energy content (MJ/kg DM) and chemical component (g/ kg DM) according to Noblet et al. (1994). d Estimated from analyzed lysine content and estimated digestibility coefcient of lysine calculated by additivity from characteristics of raw materials according to Eurolysine and ITCF (1995) tables.

described by Quiniou and Noblet (1999). Litter size was standardized to 12 pigs by cross-fostering within 48 h after birth. When the number of available piglets was too low, piglets from nonexperimental sows were fostered.

1242

Renaudeau et al.

Experimental temperature was established on d 1 (i.e., 24 h after the mean farrowing day) over a 4-h period. Water was supplied for ad libitum consumption to the sows from a low-pressure nipple drinker connected to a 55-L graduated water tank. Creep feed (17.3 MJ of GE/ kg, 21.4% CP) was offered to the pigs after d 21 following farrowing. The day before weaning (d 27), refusals were collected at 1600 in order to weigh the sows with an empty digestive tract on the following morning. After weaning, sows stayed in the farrowing room at the same temperature for 14 d to record postweaning reproductive performance. During this period, they received 3.0 kg/d of the gestation diet and were checked visually twice a day for estrus using an aerosol reproducing the odor of the boar (Boarmate, ANTEC, France). Measurements. Live weight and backfat thickness were measured at farrowing, at weaning (d 28), and d 14 after weaning. Backfat thickness was determined ultrasonically 65 mm from the midline at the last rib (P2 site). Piglets were individually weighed at birth, every 7 d during lactation, at weaning, and at d 14 after weaning. Every morning, refusals were removed between 0830 and 0900 and new feed was immediately offered as pellets. Average daily feed intake and water consumption were determined as the difference between feed allowance and refusals collected on the next morning. The weaning-to-estrus interval was determined visually (i.e., immobilization while pressing on the back) and conrmed by blood progesterone tests on blood samples taken on d 8 and d 13 after weaning. Nutrient digestibility and energy value measurements of experimental diets were carried out in a Latin square design during 10-d periods of excreta collection using three nongravid sows fed at maintenance as described by Noblet and Shi (1993). Body temperatures were measured every Monday morning from the Monday before farrowing to the Monday of the 2nd wk after weaning. The ambient temperature was established on Friday (i.e., the day after the mean farrowing day), so temperatures were measured at 4, 3, 10, 17, 24, 31 and 38 d of exposure. The respiratory rate was measured on the same days, but only during lactation and only for some sows (n = 18). A digital thermometer and a type K probe were used to measure the temperatures on the back (i.e., at the P2 position), on the mammary gland, and on an intermediate point between the back and the udder (i.e., ank). A digital thermometer was used to measure rectal temperatures. To determine respiratory rate, ank movements were counted during one minute between 0900 and 1100, but only for resting sows (Quiniou and Noblet, 1999). Chemical Analyses. Feed and feces were analyzed for moisture, ash, and fat according to AOAC (1990) methods, for N according to the Dumas method, and for crude ber and cell wall components according to Van Soest and Wine (1967). Gross energy was measured using an adiabatic bomb calorimeter. Total amino acid contents of the diets were analyzed at the Ajinomoto Eurolysine laboratory (Amiens, France) using ion-exchange chroma-

tography, except for tryptophan, which was analyzed using high performance liquid chromatography. Nitrogen in urine was measured on fresh material and urinary energy content was obtained after freeze-drying 50 mL in small polyethylene bags. Concentration of blood progesterone was measured by radioimmunoassays; sows were considered to be in estrus when blood concentration of progesterone exceeded 3 ng/mL (Martinat-Botte et al., 1996). Calculations and Statistical Analyses. Digestibility coefcients (%) and DE contents (MJ/kg) of experimental diets were calculated according to routine procedures (Noblet and Shi, 1993). Nongravid sows used for digestibility measurements had close to zero N retention; ME values obtained under these conditions underestimate the ME value of similar diets fed to lactating sows, which retained about half of the dietary N. Therefore, the ME content (MJ/kg) was estimated from DE content measured in the sows and the ME:DE ratio measured on the same diets in growing pigs (96.2, 97.2, and 97.2% for NP, LP, and LPF diets; Lebellego et al., unpublished data). The NE content (MJ/kg) was estimated from DE content and chemical composition of diet (g/kg MS) according to the equation proposed by Noblet et al. (1994). Chemical composition of lactation empty BW loss (g/d) was estimated from variations in BW and P2 backfat thickness using the equations proposed by Dourmad et al. (1997).3 Lactation N loss from body reserves was assessed from estimated mobilization of body protein, assuming a N content of 160 g/kg in body protein. Nitrogen from creep feed (Ncf) intake was calculated from creep feed intake per litter over the last week of lactation. Nitrogen output in milk (Nm) was estimated for all lactation periods from total N retained by the litter (Renaudeau and Noblet, unpublished data); Ncf and Nm were assumed to be used by suckling piglets with efciencies of 90 and 87%, respectively (Noblet and Etienne, 1986). Fecal N was estimated using the digestibility coefcient of N for each experimental diet. Urinary N was calculated as the difference between N supplied by feed and body reserves and N output in milk and feces. The effects of ambient temperature, diet composition, and their interaction on results were tested through analysis of variance (General Linear Models procedure, SAS Inst. Inc., Cary, NC). The effect of group of sows was tested within the effect of ambient temperature. Contrasts were used to compare the NP diet with the low-CP diets (LP and LPF). The effect of stage of lactation on feed intake was analyzed as repeated measures

Lipids (kg) = 26.4 (4.5) + 0.221 (0.030) empty BW + 1.331 (0.140) P2 (RSD = 6.1); Protein (kg) = 2.28 (2.22) + 0.178 (0.017) empty BW 0.333 (0.067) P2 (RSD = 1.9); Water (kg) = 23.6 (4.4) + 0.551 (0.031) empty BW 0.919 (0.135) P2 (RSD = 5.2); Ash (kg) = 0.58 (0.61) + 0.037 (0.005) empty BW 0.081 (0.018) P2 (RSD = 0.5); Energy (MJ) = 1,075 (159) + 13.67 (1.12) empty BW + 45.98 (4.93) P2 (RSD = 208). Empty BW (kg) = a BW1.013 (kg), with a = 0.912 at farrowing and a = 0.905 at weaning. P2 = P2 backfat thickness (mm).
3

High temperature and diet composition

1243

Table 2. Effects of ambient temperature and diet composition on litter growth and daily consumption of sows during lactation (adjusted means)
20C Item No. of sows Parity Duration of lactation, d Litter size Litter weight gain, g/d Intake from farrowing to weaning Feed, g Water, L Intake from d 7 to 27 Feed, g DE, MJ ME, MJ NE, MJ Water, L Water/feed, L/kg
a b

29C LPFa 9 3.0 29.0 10.8 2,938 6,727 30.9 7,630 116.7 113.4 85.9 34.2 4.5 NP 11 2.9 28.1 10.8 2,150 3,560 28.7 3,874 56.3 54.2 40.2 31.5 8.2 LP 11 3.0 27.9 10.8 2,242 4,054 25.1 4,513 65.0 63.1 47.7 27.1 6.1 LPF 9 3.2 27.8 10.1 2,064 3,862 31.3 4,264 65.2 63.3 48.0 34.3 8.2 RSDb 0.8 0.7 1.0 494 790 7.6 910 13.4 12.9 9.5 8.9 1.8

NPa 10 2.6 28.3 10.8 2,933 6,707 25.5 7,736 112.5 108.2 80.4 29.0 3.7

LPa 9 2.8 28.4 10.6 2,872 6,507 25.4 7,541 108.5 105.4 79.7 29.2 3.8

Statistical analysisc G* T**, G** G* T***, G* T*** G*, D T*** T*** T*** T***, D T***

NP = normal protein diet, LP = low-protein diet, and LPF = low-protein and added fat diet. Residual standard deviation. c From analysis of variance including the effects of ambient temperature (T), diet composition (D), interaction between temperature and diet composition (T D), the linear contrast between NP and LP + LPF diets (C), and the effect of group of sows within temperature (G). Statistical signicance: ***P < 0.001, **P < 0.01, *P < 0.05, P < 0.10.

with temperature as main effect (Repeated procedure, SAS Inst. Inc.). Data for body temperatures and respiratory rate were analyzed with a multifactorial analysis of variance (split plot) taking into account the effects of temperature, diet, duration of exposure to ambient temperature, and interactions. Data from d 4 of exposure were not taken into account for testing the interaction between ambient temperature and duration of exposure because experimental temperature was set on d 1 after farrowing. The effects of location of measurement (only for body temperatures) were also taken into account in the preceding model without the effect of duration of exposure. Number of sows coming into estrus before and 5 d after weaning were compared using a chi-square test (SAS Inst. Inc.).

Results
The sow parity number was 2.9 on average, and the mean litter size during lactation was 10.6 (Table 2). The lactation length was lower for one group of sows kept at 29C with a subsequent shorter lactation at 29C than at 20C (27.9 vs 28.6 d); no difference was observed between dietary treatments. Compared with an exposure to 20C, exposure to 29C depressed ADFI during the whole lactation period and over the ad libitum intake period (i.e., between d 7 and 27) (3,825 vs 6,647 and 4,217 vs 7,635 g/d, respectively; Table 2). The effect was similar whether based on daily ME or on NE intakes. At 29C, even though dietary treatments had no effect (P > 0.10) on ADFI and daily NE intakes, those values were similar (P = 0.36) for the LP and LPF diets and numerically higher than for NP diet (47.9 vs 40.2 MJ of NE/d). However, at 20C, daily

ME or NE intakes were similar (P > 0.10) for NP and LP diets and lower than for the LPF diet (80.0 vs 85.9 MJ of NE/d). As a consequence, NE intake was less depressed by exposure to high ambient temperature when sows were fed the LP diet (minus 40%) than when they were fed the NP diet (minus 50%); the LPF diet was intermediate (minus 44%). Ambient temperature had no inuence (P > 0.10) on daily water intake either over the total lactation or during the ad libitum intake period (Table 2). This resulted in a water:feed intake ratio that was higher at 29C (7.5 vs 4.0 L/kg at 20C, P < 0.001). At the beginning of lactation period, the sows were restrictively fed, so that ADFI increased similarly at both ambient temperatures until d 5 (Figure 1). At 29C, between d 5 and 6, ADFI decreased by 677 g/d (P < 0.001). After d 6, ADFI increased progressively with the advancement of lactation (+752 and +587 g/d between

Figure 1. Daily feed intake during the lactation period (20C = ; 29C = ).

1244

Renaudeau et al.

d 6 and 26 at 20 and 29C, respectively; P < 0.05) (Figure 1). Lactation BW loss was lower at 20C than at 29C (16 vs 34 kg, P < 0.001, Table 3), but it was not affected by dietary treatment. However, despite a tendency of interaction between temperature and diet (P = 0.10), sows kept at 29C and fed the low-CP diets lost less BW than those fed the NP diet (30 vs 41 kg). Consequently, when temperature increased from 20 to 29C, the BW loss variation was more pronounced with the NP diet (+ 160%) than with the LP or LPF diets (+80%). At farrowing, backfat thickness tended (P < 0.10) to be lower at 29C than at 20C (19.8 vs 21.4 mm) and variation over lactation was not affected by temperature and dietary treatment (3.7 mm on average). Over the postweaning period (i.e., between weaning and d 14 after weaning), sows BW increased by 2 kg at 29C and decreased by 4 kg at 20C; backfat thickness change was not affected by ambient temperature (+0.5 mm on average). The latter criteria were not affected by dietary treatment during the previous lactation. The empty BW loss was higher (P < 0.001) at 29C than at 20C (1,260 and 633 g/d, respectively), in connection with higher (P < 0.001) water, lipid, protein, ash, and energy mobilization (Table 4). The water, protein, and ash losses at 29C tended to be higher (P < 0.10) for the NP diet than for the average of the LP and LPF diets (680, 214, and 43 g/d vs 503, 157, and 31 g/d, respectively). Effects of ambient temperature and diet composition on the sows N balance are presented in Table 5. The increase of temperature from 20 to 29C resulted in a decrease (P < 0.001) of N intake (44%) and total excreted

N (35%) and an increase (P < 0.001) of N mobilized from body reserves (+130%). Reduction of dietary CP level markedly decreased (P < 0.001) the amount of N excreted by the sow for both temperatures (minus 7% for each 1% reduction of dietary CP). When the sows and their litters were considered together and contribution of N from body reserves was taken into account, N excreted, expressed as a percentage of total N supplied (i.e., N from feed + N from creep feed + N from body reserves), was lower (P < 0.05) for the LP diet (62%) than for the NP diet (65%); the corresponding value for the LPF diet was intermediate (63%). Reproductive performances were measured for a total of 58 sows because one sow on the NP treatment was slaughtered after weaning. The weaning-to-estrus interval was unaffected by temperature or diet composition (Figure 2). Most sows (49 out of 58) came into estrus within 5 d following weaning; one sow did not show estrus signs before d 14 after weaning. The mean weaning-to-estrus interval was 4.9 d for the 57 sows displaying estrus before d 14 after weaning. Ovulation was conrmed in the 57 sows by a blood sample concentration of progesterone higher than 3 ng/mL. For technical reasons, BT were not recorded for one group of sows at d 4 and 24 of exposure and for two groups of sows at d 31 and 38 of exposure; 18 sows were measured for respiratory rate at all stages of exposure. Over the lactation period, respiratory rate and back, ank, udder, and rectal temperatures increased (P < 0.01) at 29C without effect (P > 0.10) of dietary treatment (Table 6). Because no effect of location was found for values obtained on the back and the ank, skin temperature calculated as the mean of back and ank tem-

Table 3. Effects of ambient temperature and diet composition on body weight and backfat thickness changes of sows over the 28-d lactation and postweaning period (adjusted means)
20C Item No. of sows Body weight, kg At farrowing At weaning At d 14 after weaning Lactation BW change Postweaning BW changed Backfat thickness (P2), mm At farrowing At weaning At d 14 after weaning Lactation P2 change Postweaning P2 changed
a b c

29C LPF 9 264 247 243 18 4 21.0 18.5 19.0 2.5 0.5
a

NP 10 260 244 238 16 6

LP 9 258 242 239 15 3

NP 11 268 227 233 41 1 19.1 14.6 14.7 4.5 0.1

LP 11 264 235 236 29 1 20.5 16.4 16.6 4.0 0.3

LPF 9 257 227 231 31 5 19.7 16.1 16.9 3.6 0.8

RSDb 15 18 16 12 5 3.5 3.4 3.4 2.2 1.1

Statistical analysisc G*** T**, G*** G*** T*** T***, G* T, G* T**, G* T**, G

21.3 17.6 17.9 3.7 0.3

22.0 18.5 19.5 3.5 1.0

NP = normal protein diet, LP = low-protein diet, and LPF = low-protein and added fat diet. Residual standard deviation. From analysis of variance including the effects of ambient temperature (T), diet composition (D), interaction between temperature and diet composition (T D), the linear contrast between NP and LP + LPF diets (C), and the effect of group of sows within temperature (G). Statistical signicance: ***P < 0.001, **P < 0.01, *P < 0.05, P < 0.10. d Two sows were culled between weaning and d 14 after weaning (in treatment NP at 29C).

High temperature and diet composition

1245

Table 4. Effects of ambient temperature and diet composition on composition of sows BW loss over a 28-d lactation (adjusted means)a
20C Item No. of sows Empty body weight loss, g/d Backfat thickness loss, (P2) mm/d Mobilization of chemical components Water, g/d Lipid, g/d Protein, g/d Ash, g/d Energy, MJ/d
a b c

29C LPFb 9 676 0.09 292 266 91 18 13.3 NP 11 1,503 0.16 680 547 214 43 27.9 LP 11 1,107 0.14 477 438 149 29 21.8 LPF 9 1,179 0.13 529 435 166 33 22.1 RSDc 419 0.08 196 178 62 13 8.5

NPb 10 627 0.13 223 316 67 12 14.7

LPb 9 596 0.12 216 295 65 12 13.8

Statistical analysisd T***

T***, C T** T***, C T***, C T***

Empty BW and chemical composition were calculated at farrowing and weaning using equations of Dourmad et al. (1997; see text). NP = normal protein diet, LP = low-protein diet, and LPF = low-protein and added fat diet. Residual standard deviation. d From analysis of variance including the effects of ambient temperature (T), diet composition (D), interaction between temperature and diet composition (T D), the linear contrast between NP and LP + LPF diets (C), and the effect of group of sows within temperature (G). Statistical signicance: ***P < 0.001, **P < 0.01, *P < 0.05, P < 0.10.

peratures was used as a single criterion to compare with udder and rectal temperatures. At 20C, skin temperature was lower (P < 0.001) than udder temperature (36.2 vs 38.2C), and both temperatures were lower than rectal temperature (38.8C). At 29C, all these values were higher (P < 0.001) than at 20C, but the differences between locations were reduced: the gradient between skin and rectal temperatures was lower (1.6 vs 2.6C at 20C, P < 0.001), whereas the gradient between udder and rectal temperatures remained constant (0.6C). During the lactation period, no effect (P = 0.15) of duration of

exposure was found on skin temperature. The respiratory rate increased from 33 to 105 breaths/min when the ambient temperature changed from 20 to 29C, but it was not affected by dietary treatment. Both skin and rectal temperatures decreased after weaning (i.e., between d 24 and 38 of exposure) (Figure 3). The skin temperature reduction was greater at 20C than at 29C (2.9 vs 1C between d 24 and 38, P < 0.01), whereas rectal temperature change was the same for both temperatures (0.9C on average). Consequently, after weaning the gradient between skin and rectal tem-

Table 5. Effects of ambient temperature and diet composition on nitrogen balance over a 28-d lactation (adjusted means)
20C Item No. of sows N supplied, g From feed (Ni) From body reserves (Nr)d From creep feed (Ncf) N output, g In milk (Nm)e Excreted (N urine + N feces) Sowsf Pigletsg Total Total, % (Ni + Ncf + Nr)h
a b

29C LPF
a

NP

LP

NP 11 3,269k 968k 115 j 1,725 j 2,512 l 223 i 2,736 l 63 ij

LP 11 2,950k 665 l 92 j 1,671 j 1,945m 214 i 2,159m 58 j

LPF 9 2,985k 738 l 78 j 1,530 j 2,193m 196 i 2,389m 63 ij

RSDb 662 285 47 424 527 52 513 7

Statistical analysisc T***, D**, C*, G T***, C* T**, G** T***, G T***, D***, C T***, G T***, D***, C D*

10 6,137 i 305 i 61 i 2,369 i 4,073 i 297 i 4,370 i 67 i

9 4,811 j 299 i 65 i 2,300 i 2,809 j 289 i 3,098 j 60 j

9 5,377 j 422 i 41 i 2,431 i 3,369k 303 i 3,672k 63 ij

NP = normal protein diet, LP = low-protein diet, and LPF = low-protein and added fat diet. Residual standard deviation. c From analysis of variance including the effects of ambient temperature (T), diet composition (D), interaction between temperature and diet composition (T D), the linear contrast between NP and LP + LPF diets (C), and the effect of group of sows within temperature (G). Statistical signicance: ***P < 0.001, **P < 0.01, *P < 0.05, P < 0.10. d Nr = body protein mobilized during lactation/6.25; body protein calculated from equation of Dourmad et al. (1997); see Table 4. e Nm = (total N retained by piglets Ncf 0.90) efciency of N milk utilization (i.e., 1.14; Noblet and Etienne, 1986). Litters N retention was calculated according to the comparative slaughter technique (Renaudeau and Noblet, unpublished data). f N feces = (1 digestibility coefcient of nitrogen) N in feed, N urine = Ni + Nr Nm N feces. g N excreted = Nm (1 efciency of N milk utilization). h N total excreted expressed as a percentage of total N supply (Ni + Ncf + Nr). i,j,k,l,m Effect of temperature and diet composition. Within a row, means not followed by the same superscript differ (P < 0.05).

1246

Renaudeau et al.

Figure 2. Effect of ambient temperature and diet on number of sows displaying estrus after weaning (white portion = estrus 5 d after weaning and black portion = estrus > 5 d after weaning). Mean weaning to estrus interval (5.2 d at 20C and 4.9 d at 29C) was calculated without one sow not displaying estrus before d 14 after weaning (in LP treatment at 29C). NP, normal protein; LP, low protein; LPF, low protein and added fat. peratures was 2.1C at 20C but remained the same as during lactation at 29C (1.7C).

Figure 3. Effect of ambient temperature (20C = ; 29C = ) and duration of exposure on rectal () and skin ( ) temperatures).

Discussion Heat Dissipation at High Ambient Temperatures in Lactating Sows

Independently of the ambient temperature during lactation, skin and rectal temperatures increased after far-

rowing, remained constant thereafter during lactation, and decreased after weaning. Similar variations in rectal temperature over the successive periods of the reproductive cycle were described by Prunier et al. (1997). This observation can be compared with the direct relationship between heat production and energy supply in lactating sows (Noblet and Etienne, 1987). We can then suppose that elevated body temperatures during lactation compared with those in the prefarrowing and postweaning periods is related to the increase of metabolic heat production due to increased feed intake and milk synthesis.

Table 6. Effects of ambient temperature and diet composition on body temperature and respiratory rate in sows (adjusted means)
Duration of exposure, d Item Observations, n 4 54 3 59 36.8e 36.9e 38.4e 39.2e C 2.4e 2.3d 0.7e 10 59 37.0e 37.2e 38.6e 39.2e 2.3e 2.1d 0.6e 18 70 17 59 36.8e 37.0e 38.5e 39.1e 2.2e 2.1d 0.6e 18 66 24 54 36.9e 37.2e 38.6e 39.1e 2.2e 1.9d 0.5e 18 74 31 49 35.6d 35.7d 37.1d 38.3d 2.7d 2.6e 1.2f 38 50 35.1f 35.1f 35.5f 38.2d 3.0f 3.1f 2.7f Temperature, Ca 20 111 36.1i 36.2i 38.2h 38.8g 2.7 2.6 0.6 36 33 29 120 37.6i 37.9i 38.9h 39.5g 1.8 1.6 0.6 36 105 RSDb 0.8 0.7 0.6 0.5 0.7 0.7 0.4 T**, T**, T**, T**, E**, E**, E**, E**, Statistical analysisc T T T T E**, E**, E**, E**, A** A** A** A**

Body temperature, C Back 35.5d Flank 35.8d Udder 36.7d Rectum 38.1d Gradient of body temperature, Rectum back 2.6de Rectum ank 2.3d Rectum udder 1.4d

T**, E**, T E**, A** T**, E**, T E**, A** T**, E**, T E**, A**

Respiratory rate, breaths per min Observations, n 18 Rate 66

a b

17

T**, A*

Means calculated only with lactation data. Residual standard deviation. c From multifactorial analysis of variance (split plot) including the effects of ambient temperature (T), diet composition (D), duration of exposure to the temperature (E), interactions, and effect of animal (A). Statistical signicance: **P < 0.01, *P < 0.05. d,e,f Within each line, adjusted mean values with different superscripts are different (P < 0.05). Data from d 4 of exposure were not taken into account for testing the interaction between temperature and duration of exposure because experimental temperature was set on d 1 after farrowing. g,h,i Effect of site of measurement (back, ank, udder, or rectum) was taken account in the model (see footnote b). Within each column, adjusted mean values with different superscripts are signicant (P < 0.05, RSD = 1.1).

High temperature and diet composition

1247

One of the effects of high ambient temperatures is an increase of skin, udder, and rectal temperatures; our results (+0.18, +0.08, and +0.08C per extra degree of temperature, respectively) are in agreement with those of Quiniou and Noblet (1999) (+0.15, +0.05, and +0.11C between 22 and 29C). The elevation of peripheral temperature can be explained by an increase of blood ow in the skins vessels for dissipation of body heat. Moreover, elevated ambient temperature induced a rise in respiratory rate (+8.0 breaths/min/C) similar to values reported in other studies (+7.5, +7.1, and +8.9 breaths/min/C, according to Schoenherr et al., 1989, Lorschy et al., 1991, and Quiniou and Noblet, 1999, respectively). Similarly, in cattle given ad libitum access to feed, Hahn (1999) reported an increase of respiratory rate (+4.3 breaths min1 C1) between 18 and 32C. Cattle have more sweat glands than pigs, and the contribution of the lungs to evaporative heat losses is less important. In contrast, taking into account the low cutaneous evaporative capacities in pigs, an important rise in respiratory rate is required in pigs for an effective increase of evaporative heat losses. All these processes correspond to a short-term adaptation to hot climatic conditions via increased heat dissipation. The reduction in the gradient between rectal and skin temperatures (i.e., 1C) observed in the present experiment when ambient temperature increased from 20 to 29C is in agreement with results of Quiniou and Noblet (1999). No signicant effect of duration of exposure at 29C was observed on skin and rectal temperatures and respiratory rate in our study. This means that, in agreement with Quiniou and Noblet (1999), lactating sows adapt quite quickly to hot temperatures (i.e., within 3 d, Figure 3), with no more adaptation on a long-term basis. At a higher temperature (i.e., 32C) and with lower productivity sows, Schoenherr et al. (1989) reported a reduced rectal temperature and respiratory rate in lactating sows as the lactation progressed. This suggests a tendency for long-term adaptation in heat-stressed sows with an increase of efciency of evaporative heat loss. The discrepancy between both sets of data could be explained by differences in sows productivity and its change over lactation.

more pronounced at high ambient temperature (Stansbury et al., 1987; Quiniou and Noblet, 1999). Similar conclusions have been obtained in growing pigs (Giles and Black, 1991; Quiniou et al., 2000a). Our value (331 g d1 C1 and 4 MJ of ME d1 C1 between 20 and 29C) is higher than the average value suggested by Black et al. (1993) (170 g d1 C1 and 2.4 MJ of ME d1 C1) but is quite similar to those obtained in the recent study of Johnston et al (1999) (289 g d1 C1 and 3.9 MJ of ME d 1 C1 between 19 and 27C) or calculated from the equation proposed by Quiniou and Noblet (1999). It can also be suggested that the higher negative effect of hot temperatures on ADFI of lactating sows observed in recent studies can be related to the higher production level at thermoneutrality and the associated higher rate of heat production in highly productive sows with subsequent higher susceptibility to heat stress. At 29C, sows were restrictively fed at the beginning of the lactation period and seemed not to be affected by high ambient temperature; consequently, ADFI linearly increased until d 5. However, ADFI was signicantly reduced from d 5 to 6 (Figure 1), suggesting that sows compensated for previous feed restriction on d 5 and they had a delayed response to ambient temperature. Similar results have not been reported in the literature.

Body Composition in Heat-Stressed Lactating Sows

The BW loss measured at thermoneutrality in our study was considerably higher than the value reported by Johnston et al. (1999) in multiparous sows (560 vs 60 g/d). In that study, ADFI was rather similar to ours (i.e., 6,375 vs 6,647 g/d) but litter weight gain and litter size were lower (2,110 vs 2,914 g/d and 9.9 vs 10.7 piglets, respectively). This suggests that the ad libitum supply of high-quality diets under favorable climatic conditions was unable to meet requirements of highly productive sows. The BW loss contained 46% lipids, 38% water, and 12% protein; lipids then represented the major source of energy from body reserves (86%). For multiparous sows studied over a 21-d lactation, Quiniou and Noblet (1999) reported similar values of body nutrient change (15% and 33%, respectively, for protein and lipid content in BW loss) in sows fed diets comparable to ours (i.e., 0.70 g/MJ of ME). Moreover, Sauber et al. (1998) showed that more protein was mobilized when the lysine:ME ratio was below 0.85 g/MJ of ME. However, reduction of energy supply at constant daily protein supply was associated only with accentuated body fat mobilization (Noblet and Etienne, 1987). Consequently, BW loss composition depends to a large extent on the ratio between protein (i.e., amino acids) and energy in the feed. Between 20 and 29C, BW loss increased greatly (34 vs 16 kg), whereas backfat thickness loss increased slightly. The lack of signicant effect of temperature on backfat thickness loss may be due to a great variability between sows (CV = 60%) and partly by a lower backfat thickness before farrowing in sows exposed to 29C.

Thermogenesis at High Ambient Temperatures in Lactating Sows

The effects of high ambient temperature on voluntary feed and energy intakes have been well described in the literature. As reviewed by Quiniou and Noblet (1999), in multiparous and mixed-parity sows, each single degree of rise in farrowing house temperature causes a decrease in ADFI ranging between 100 and more than 200 g/d (or 1.3 to more than 2.6 MJ of ME/d); this variability can be explained by differences in lactation length, parity number, milk production, and temperature range considered. Indeed, in contrast to what has been considered in the review of Black et al. (1993), the effect of temperature on ADFI in lactating sows becomes

1248

Renaudeau et al.

Although postweaning ADFI was similar for all sows (i.e., 3 kg/d), those kept at 20C lost BW, whereas the BW of those at 29C increased during the 14-d postweaning period. Such a difference was also reported by Quiniou et al. (2000b) for sows exposed either at 20 or 26C during lactation and 21C after weaning. Hughes (1993) reported that feed restriction (3 kg/d) during lactation induced a BW gain (+260 g/d), whereas sows fed to appetite (6 kg/d) lost BW (1,111 g/d) over the 7-d period following weaning. This suggests that the postweaning BW change is dependent on the difference in feeding levels between the lactation and the postweaning period. Under conventional conditions (i.e., high level during lactation and lower after weaning), BW loss is in fact associated with an important reduction in gut ll and weight of viscera (e.g., gut and liver), whereas lean and adipose tissues start to be reconstituted. Conversely, in sows fed restrictively during lactation, BW change during the postweaning period corresponds to a reduction of mammary gland, positive gain in body reserve tissues, and minor changes in weight of viscera (Ziopoulos et al., 1983; King and Dove, 1987).

with a subsequent marked reduction of N excretion. Our results and those of Everts and Dekker (1994) and Dourmad et al. (1998) indicate that N excretion can be reduced by 7% for each 1% reduction of dietary CP. In agreement with Tilton et al. (1995), there was a tendency for higher NE intake for the fat-supplemented diets (i.e., 8%) without reduction of ADFI and BW loss.

Low Heat Increment Diets at High Ambient Temperatures

Results of the present study suggest that the negative effects of high ambient temperature on energy intake and mobilization of body reserves are attenuated when low-CP diets are fed. There are few studies on the effect of low-CP diets on performance of lactating sows exposed to high ambient temperatures. In response to a reduced protein supply (i.e., 17 vs 14%), Quiniou and Noblet (1999) reported no improvement in performance of lactating sows kept at 29C. In that study, the lack of interaction between temperature and diet was attributed to the low number of observations and(or) the lower supplies of sulfur amino acids and tryptophan expressed as a percentage of lysine in the 14% CP diet. However, Lynch (1989) observed an increase of feed consumption (+700 g/d) in multiparous, lactating sows fed a low-CP diet (14 vs 20%) at 28C. Moreover, increasing the energy concentration of the sows diet by addition of fat resulted in an increase of daily ME intake without any effect on BW loss in hot conditions (Schoenherr et al., 1989; Quiniou et al., 2000b). However, according to our results, the combination of reduced CP level and increased fat content (i.e., with 4% of vegetable oil) attenuated the negative effects of high temperature on overall performance of lactating sows, including BW loss. These results would conrm the benet of reducing metabolic heat production under high ambient temperature with a low heat increment diet (low CP with or without fat).

Reproduction at High Ambient Temperature

This experiment showed no inuence of ambient temperature on weaning-to-estrus interval, and only one sow at 29C did not display estrus before d 14 post after weaning. For multiparous or mixed-parity sows, Lynch (1977), McGlone et al. (1988), and Schoenherr et al. (1989) also found no effect of high ambient temperature on weaning-to-estrus interval. However, Prunier et al. (1997) and Johnston et al. (1999) reported a lower proportion of multiparous sows displaying estrus when kept under high temperature during lactation, even though lactation BW loss was lower than in the present study. In agreement with Mullan and Close (1991) and Prunier et al. (1996), this suggests that, unlike in primiparous sows (Mullan and Williams, 1989; King and Williams, 1984), the relationship between body reserves losses during lactation and reproductive performance remains unclear in multiparous sows.

Implications
The present experiment demonstrates that the reduction of CP level in a wheat-corn-soybean meal diet combined with supplementation of crystalline lysine and other essential amino acids (methionine, threonine, isoleucine, tryptophan, and valine) allows the maintainance of performance and reduction in N excretion in lactating sows kept under thermoneutral conditions. The detrimental effect of exposure to high ambient temperature on feed intake and lactation performance can be attenuated by using diets with low thermic effect of feed.

Utilization of Low-Protein Diets in Lactating Sows

In thermoneutral conditions, feed intake and lactation performance were quite similar for the sows on the NP and LP diets. This is comparable to the results reported by Dourmad et al. (1998) and Libal et al. (1997). However, Johnston et al. (1999) reported a slight increase of BW loss with reduction of dietary CP from 16.7 to 13.3%, whereas ADFI was not affected. This latter result can be partly explained by the probable shortages in threonine, tryptophan, and valine relative to lysine supply in the low-CP diet (58, 15, and 82%, respectively). Our results with lactating sows combined with those of Canh et al. (1998) with growing pigs conrm that a reduction of CP intake when essential amino acid supplies are kept constant allows maintaining the level of performance

Literature Cited
AOAC. 1990. Ofcial Methods of Analysis. 15th ed. Association of Ofcial Analytical Chemists, Washington, DC. Black, J. L., B. P. Mullan, M. L. Lorschy, and L. R. Giles. 1993. Lactation in the sow during heat stress. Livest. Prod. Sci. 35:153170. Canh, T. T., A. J. A. Aarnink, J. B. Schutte, J. D. Sutton, D. J. Langhout, and M. W. A. Verstegen. 1998. Dietary protein affects nitrogen

High temperature and diet composition

excretion and ammonia emission from slurry of growing-nishing pigs. Livest. Prod. Sci. 56:181191. ` Dourmad, J. Y., M. Etienne, and J. Noblet. 1991. Contribution a letude des besoins en acides amines de la truie en lactation. J. Rech. Porc. Fr. 23:6168. Dourmad, J. Y., M. Etienne, J. Noblet, and D. Causeur. 1997. Prediction ` de la composition chimique des truies reproductrices a partir ` du poids vif et de lepaisseur de lard dorsal: Application a la determination des besoins energetiques. J. Rech. Porc. Fr. 29:255262. Dourmad, J. Y., J. Noblet, and M. Etienne. 1998. Effect of protein and lysine supply on performance, nitrogen balance, and body composition changes during lactation. J. Anim. Sci. 76:542550. Eurolysine and ITCF, Paris, France. 1995. Ileal digestibility of aminoacids in feedstuffs for pigs. p 53. Everts, H., and R. A. Dekker. 1994. Effect of nitrogen supply on nitrogen and energy metabolism in lactating sows. Anim. Prod. 59:445454. Giles, L. R., and J. L. Black. 1991. Voluntary food intake in growing pigs at ambient temperatures above the zone of thermal comfort. In: APSA Committee (ed.) Manipulating Pig Production. pp 162 166. Australiasian Pig Science Assoc., Animal Research Institute, Werribee, Australia. Hahn, G. L. 1999. Dynamic responses of cattle to thermal heat loads. J. Anim. Sci. 77:1020. Hughes, P. E. 1993. The effects of food level during gestation and early gestation on the reproductive performance of mature sows. Anim. Prod. 57:437445. INRA. 1989. Lalimentation des animaux monogastriques. Institut National de la Recherche Agronomique, Paris, France. Johnston, L. J., M. Ellis, G. W. Libal, V. B. Mayrose, W. C. Weldon, and NRC-89 Committee on swine management. 1999. Effect of room temperature and dietary amino acid concentration on performance of lactating sows. J. Anim. Sci. 77:16381644. King, R. H., and H. Dove. 1987. Composition of the weight change in sows after weaning. In: APSA Committee (ed.) Manipulating Pig Production. p. 134. Australiasian Pig Science Assoc., Animal Research Institute, Werribee, Australia. King, R. H., and I. H. Williams. 1984. The effect of nutrition on the reproductive performance of rst-litter sows. 1Feeding level during lactation, and between weaning and mating. Anim. Prod. 38:241247. Le Bellego, L., J. van Milgen, M. Rademacher, S. Van Cauwenberghe, and J. Noblet. 1999. Effect of low protein diets on energy utilization in growing pigs. J. Anim. Sci. 77 (Suppl. 1):196197(Abstr.). Libal, G. W., D. J. Uttecht, and C. R. Hamilton. 1997. Tryptophan needs of lactating sows fed diets supplemented with crystalline lysine. J. Anim. Sci. 75:417422. Lorschy, M. L., L. R. Giles, C. R. Smith, J. M. Gooden, and J. L. Black. 1991. Food intake, heat production and milk yield of lactating sows exposed to high temperature. In: APSA Committee (ed.) Manipulating Pig Production. p 81. Australiasian Pig Science Assoc., Animal Research Institute, Werribee, Australia. Lynch, P. B. 1977. Effect of environmental temperature on lactating sows and their litter. Ir. J. Agric. Res. 16:123130. Lynch, P. B. 1989. Voluntary food intake of sows and gilts. In: A. T. Smith and T. L. J. Lawrence (ed.) Pig Housing and the Environment. pp 7177. British Society of Animal Production, Edinburgh, U.K. Martinat-Botte, F., H. Quesnel, A. Prunier, J. Tournut, and M. Terqui. 1996. Reproduction de la truie: Bases physiologiques et matrise. ` 1ere partie. Rev. Med. Vet. 147:3346. McGlone, J. J., W. F. Stansbury, and L. F. Tribble. 1988. Management of lactating sows during heat stress: Effects of water drip, snout coolers, oor type and a high energy-density diet. J. Anim. Sci. 66:885891.

1249

Mullan, B. P., and W. H. Close. 1991. Metabolic endocrine changes during the reproductive cycle of the sow. In: APSA Committee (ed.) Manipulating Pig Production. p 32. Australiasian Pig Science Assoc., Animal Research Institute, Werribee, Australia. Mullan, B. P., and I. H. Williams. 1989. The effect of body reserves at farrowing on the reproductive performance of rst-litter sows. Anim. Prod. 48:449457. Noblet, J., and M. Etienne. 1986. Effect of energy level in lactating sows on yield and composition of milk and nutrient balance of piglets. J. Anim. Sci. 63:18881896. Noblet, J., and M. Etienne. 1987. Metabolic utilization of energy and maintenance requirements in lactating sows. J. Anim. Sci. 64:774781. Noblet, J., H. Fortune, X. S. Shi, and S. Dubois. 1994. Prediction of net energy value of feeds for growing pigs. J. Anim. Sci. 72:344354. Noblet, J., Y. Henry, and S. Dubois. 1987. Effect of protein and lysine levels in the diet on body gain composition and energy utilization in growing pigs. J. Anim. Sci. 65:717726. Noblet, J., and X. S. Shi. 1993. Comparative digestibility of energy and nutrients in growing pigs fed ad libitum and adult sows fed at maintenance. Livest. Prod. Sci. 34:137152. OGrady, J. F., P. B. Lynch, and P. A. Kearney. 1985. Voluntary feed intake by lactating sows. Livest. Prod. Sci. 12:355365. Prunier, A., M. Messias de Braganca, and J. Le Dividich. 1997. Inu ence of high ambient temperature on performance of reproductive sows. Livest. Prod. Sci. 52:123133. Prunier, A., H. Quesnel, M. Messias de Braganca, and A. Y. Kermabon. 1996. Environmental and seasonal inuences on the return-tooestrus after weaning in primiparous sows: A review. Livest. Prod. Sci. 45:103110. Quiniou, N., S. Dubois, and J. Noblet. 2000a. Voluntary feed intake and feeding behaviour of group-housed growing pigs are affected by ambient temperature and body weight. Livest. Prod. Sci. 63:245253. Quiniou, N., D. Gaudre, S. Rapp, and D. Guillou. 2000b. Inuence de la temperature ambiante et de la concentration en nutriments de laliment sur les performances de lactation de la truie primipare. J. Rech. Porc. Fr. 32:275282. Quiniou, N., and J. Noblet. 1999. Inuence of high ambient temperatures on performance of multiparous lactating sows. J. Anim. Sci. 77:21242134. Sauber, T. E., T. S. Stahly, N. H. Williams, and R. C. Ewan. 1998. Effect of lean growth genotype and dietary amino acid regimen on the lactational performance of sows. J. Anim. Sci. 76:10981111. Schoenherr, W. D., T. S. Stahly, and G. L. Cromwell. 1989. The effects of dietary fat or ber addition on yield and composition of milk from sows housed in a warm or hot environment. J. Anim. Sci. 67:482495. Stahly, T. S., G. L. Cromwell, and M. P. Aviotti. 1979. The effect of environmental temperature and dietary lysine source and level on the performance and carcass characteristics of growing swine. J. Anim. Sci 49:12421251. Stansbury, W. F., J. J. McGlone, and L. F. Tribble. 1987. Effects of season, oor type, air temperature and snout coolers on sow and litter performance. J. Anim. Sci. 65:15071513. Tilton, S. L., P. M. Ermer, A. J. Lewis, P. S. Miller, and C. K. Wolverton. 1995. The effect of dietary tallow on milk production and milk composition in rst and second parity sows. J. Anim. Sci. 73(Suppl. 1):86(Abstr.). Van Soest, P. J., and R. H. Wine. 1967. Use of detergents in the analysis of brous feeds. IV. Determination of plant cell-wall constituents. J. Assoc. Off. Anal. Chem. 50:5055. Whittemore, C. T. 1996. Nutrition reproduction interactions in primiparous sows. Livest. Prod. Sci. 46:6583. Ziopoulos, P. E., J. H. Topps, and P. R. English. 1983. Losses in weight and body water in sows after weaning. Br. J. Nutr. 50:163172.