Search
Author
Title
Vol.
Issue
Year
1st Page

Journal of Animal Science - Animal Growth, Physiology, and Reproduction

Effects of a controlled heat stress during late gestation, lactation, and after weaning on thermoregulation, metabolism, and reproduction of primiparous sows1

 

This article in JAS

  1. Vol. 91 No. 6, p. 2700-2714
     
    Received: Nov 01, 2012
    Accepted: Mar 06, 2013
    Published: November 25, 2014


    2 Corresponding author(s): lucym@missouri.edu
 View
 Download
 Share

doi:10.2527/jas.2012-6055
  1. A. M. Williams,
  2. T. J. Safranski,
  3. D. E. Spiers,
  4. P. A. Eichen,
  5. E. A. Coate and
  6. M. C. Lucy 2
  1. Department of Animal Sciences, University of Missouri, Columbia 65211

Abstract

Heat stress (HS) causes seasonal infertility in sows and decreases reproductive efficiency. The objective was to examine thermoregulation, metabolic responses, and reproduction in sows exposed to HS or thermoneutral (TN) conditions during different phases of a production cycle (gestation, lactation, and breeding). Fifty-eight first-parity Landrace (n = 26) or Landrace × Large White F1 (n = 32) sows were rotated through environmental chambers for 57 d beginning in late gestation. The ambient temperature sequences included either TN (18°C to 20°C) or HS (24°C to 30°C) for each production phase with the following treatment groups: TN-TN-TN (n = 15), TN-HS-TN (n = 14), HS-TN-HS (n = 14), and HS-HS-HS (n = 15) for gestation-farrowing-breeding (20, 24, and 13 d, respectively). Regardless of the temperature treatment, rectal temperatures were greater (P < 0.001) during lactation (39.36°C ± 0.01°C) than during the gestation (38.27°C ± 0.01°C) or the breeding period (38.77°C ± 0.01°C). The increase in rectal temperature (P < 0.001) and respiration rate (P < 0.001) in response to the HS was greatest during lactation. There was an effect of day (P < 0.001) on serum IGF-1 and insulin concentrations because both insulin and IGF-1 increased after farrowing. Compared with HS sows, the TN sows had greater feed intake (P < 0.001) and greater serum concentrations of insulin (early lactation; P < 0.05) and IGF-1 (late lactation; P < 0.05) when they were lactating. The effects of HS on sow BW, back fat, and loin eye area were generally not significant. Average BW of individual piglets at weaning was approximately 0.5 kg lighter for the sows in the HS farrowing room (P < 0.05). Weaning-to-estrus interval, percentage sows inseminated after weaning, subsequent farrowing rate, and subsequent total born were not affected by treatment. In summary, regardless of ambient temperature, sows undergo pronounced and sustained changes in rectal temperature when they transition through gestation, lactation, weaning, and rebreeding. The effects of HS on rectal temperature, respiration rate, feed intake, and metabolic hormones were greatest during lactation. The controlled HS that we imposed affected piglet weaning weight, but rebreeding and subsequent farrowing performance were not affected.



INTRODUCTION

Summer heat stress (HS) causes seasonal infertility in sows (anestrus, long weaning-to-estrus intervals, low farrowing rates, and depressed litter size; Bertoldo et al., 2012). Primiparous sows are especially affected (Love, 1978). The origins of seasonal infertility are believed, in part, to reside within the lactation phase of the production cycle. During lactation, heat-stressed sows reduce feed intake as a mechanism to reduce metabolic heat production (Teague et al., 1968; Prunier et al., 1997; Renaudeau et al., 2001, 2012). The decrease in feed intake leads to a prolonged period of negative energy balance and greater body condition loss (Black et al., 1993). Heat stress will also reduce milk production in lactating sows perhaps through an indirect effect associated with the reduction in feed intake or a direct effect of high temperature on mammary gland metabolism (Silanikove et al., 2009). The reduction in milk production then decreases piglet BW gain (Quiniou and Noblet, 1999; Renaudeau and Noblet, 2001).

Heat stress during lactation has carry-over effects on reproduction after weaning (Wettemann and Bazer, 1985; Koketsu et al., 1996), perhaps through a mechanism that involves reduced ovarian follicle size (Lucy et al., 2001). Changes in circulating hormones and metabolites (decreased blood insulin, IGF-1, and glucose concentrations as well as increased blood NEFA concentrations) caused by reduced feed intake and energy balance may be responsible for the reduction in follicle size (Hunter et al., 2004). When the follicle is smaller, a delay in return to estrus and an increase in the incidence of anestrus in sows may occur (Lucy et al., 2001).

Sows experience changes in environmental conditions as they move through different swine barns during a typical production cycle (breeding/gestation to farrowing and then back to breeding/gestation). Producers may opt to construct barns with greater capacity for cooling during the summer (evaporatively cooled barns, for example) or may retrofit existing barns with better cooling systems. If they choose to do so, then they have options in terms of which barns that they cool (breeding/gestation, farrowing, or both). The objective was to examine thermoregulation and reproduction in sows exposed to HS or thermoneutral (TN) conditions during gestation, lactation, and breeding. Sows were either maintained in a continuously HS environment (HS-HS-HS) or a continuously TN environment (TN-TN-TN) or switched between environments (HS-TN-HS or TN-HS-TN) during gestation, lactation, and breeding, respectively. The effects of the environmental treatments on the thermoregulation, metabolism, and endocrinology of sows were assessed. Growth of the piglets nursed within the respective environments was measured and reproductive outcomes from the breeding period after weaning were determined.


MATERIALS AND METHODS

Animal procedures were reviewed and approved by the University of Missouri Animal Care and Use Committee.

Animals and Facilities

Fifty-eight primiparous, pregnant Landrace (n = 26) or Landrace × Large White (n = 32) sows (n = 58; Newsham Choice Genetics, West Des Moines, IA) were brought into the University of Missouri’s Brody Environmental Center between 89 and 93 d of gestation (mean = 91 ± 0.1 d) in 5 groups from December 2007 through April 2008. The sows were reared, housed, artificially inseminated (Landrace × Large White or Large White pooled boar semen). and confirmed pregnant at the University of Missouri’s Swine Research Complex (Columbia, MO) before being brought into the Brody Environmental Center. The center contains 4 environmental chambers (each 9.3 × 5.2 × 3 m). The chambers were ventilated with 100% outside air that was exhausted to the outside (air was not recycled). There were a minimum of 10 air changes per hour (range of 10 to 14 air changes per hour) in the rooms. Two chambers were used for breeding/gestation and 2 were used for farrowing. The breeding/gestation chambers had 12 breeding/gestation stalls (2.4 × 0.6 m) with nipple valve drinkers. In breeding/gestation, the pigs laid directly on a concrete floor and there was a 0.7 m wide metal grate in the rear of the stall. The farrowing chambers had 6 farrowing crates (2.1 × 1.5 m; Rohn Agriproducts, Peoria, IL) with watering cups and feeders mounted to the front of the crate. The farrowing crates were erected approximately 0.4 m above the floor and the sows laid on plastisol coated wire mesh (Tenderfoot flooring, Tandem Products, Minneapolis, MN). The light cycle in the rooms was 15 h light and 9 h dark.

Experimental Design

Sows were in the chambers for 57 d beginning in late gestation, continuing through farrowing/lactation, and finishing during the breeding phase (Fig. 1). Sows were housed in a breeding/gestation room for 20 d and then moved into a farrowing room (approximately 111 d of gestation; sows were moved on d 20). Piglets were weaned on d 44 of the trial (average lactation length = 20.5 ± 1.4 d). Sows were moved back into the breeding/gestation room on d 44, where they were housed for the remaining days of the study. Sows were assigned to 1 of 4 treatment sequences that were denoted as TN-TN-TN (n = 15), TN-HS-TN (n = 14), HS-TN-HS (n = 14), and HS-HS-HS (n = 15), where the series of abbreviations represent the environmental temperature that the sow experienced in gestation-farrowing-breeding (20, 24, and 13 d, respectively). The daily ambient temperature cycle for TN was 18°C to 20°C and for HS was 24°C to 30°C (Fig. 2). Humidity was not added to incoming air and remained below 55% in TN chambers and below 45% in HS chambers. Maintenance of relative humidity at these levels would ensure an adequate water vapor pressure gradient for evaporative heat loss and establish the thermal gradient (i.e., animal core temperature to environmental temperature) as the primary factor affecting heat transfer.

Figure 1.
Figure 1.

Experimental design used to treat sows with different temperatures when in breeding/gestation and farrowing rooms. Treatments are denoted as TN-TN-TN, TN-HS-TN, HS-TN-HS, and HS-HS-HS, where the series of abbreviations represent the thermoneutral temperature (TN; 18°C to 20°C) or heat stress (HS; 24°C to 30°C) that the sow experienced in gestation-farrowing-breeding. Sows were in the breeding/gestation room from d 0 to 20 and the farrowing room from d 20 to 44 and were returned to the breeding/gestation room after d 44. Sows moved out of the environmental chambers on d 57.

 
Figure 2.
Figure 2.

Ambient temperature averaged across all study days inside each environmental chamber used in the study [2 heat stress (HS) chambers, 1 for farrowing and 1 for breeding/gestation, and 2 thermoneutral (TN) chambers, 1 for farrowing and 1 for breeding/gestation]. Data were collected by using 2 Hobo Data Loggers (Onset Computer Inc., Pocasset, MA) positioned at sow head level within each room. The chambers were programmed to achieve daily ambient temperature cycles of 18°C to 20°C for TN and 24°C to 30°C for HS.

 

Data from d 1 to 53 were used for the analyses. Data collected on d 0 (i.e., arrival day) were not used because sows were agitated due to the move into the facility. Data collected on or after d 54 were not used because the sows were vaccinated on d 54 and the vaccination caused an increase in rectal temperature.

Reproductive Management

Sows were allowed to farrow without intervention until d 22 of the trial. On d 23 (approximately d 114 of gestation), sows that had not farrowed were induced by injecting 10 mg of PGF [2 mL Lutalyse, intramuscularly (i.m.); Pfizer Animal Health, New York, NY), followed by injecting 40 units of oxytocin (2 mL i.m.; Vedco Inc., St. Joseph, MO) 24 h later if the sow did not show signs of impending farrowing (n = 37). If sows showed signs of farrowing on the day after injection of PGF2α, then no oxytocin was administered (n = 9 sows). Piglets were cross-fostered within treatment in an attempt to standardize litter size to approximately 10 pigs per litter.

Estrous detection was done twice daily (0900 and 1600 h) in the presence of a boar beginning 3 d after weaning. Sows in estrus were artificially inseminated once per day in the morning. Sows were transported to a commercial farm after the completion of the trial. Sows farrowed and subsequent litter data were collected that included number born alive, stillborn, preweaning mortality, and number of piglets weaned.

Thermal Measurements

Thermal response measurements were taken 4 times each day at 0800, 1200, 1600, and 2000 h. Rectal temperature was measured with a calibrated thermistor thermometer (model 8110–20; Cole-Parmer Instrument Company, Vernon Hills, IL). Respiration rate was calculated by counting breaths per minute (bpm) for 1 min duration. Movement of the side of the animal was used for this determination.

Sow BW and Back Fat Measurements

Back fat depth (BF), loin eye area (LEA), and BW were measured on d 0 (start of trial), d 20 (moved to farrowing room), d 44 (weaning; moved to breeding/gestation room), and d 57 (moved out of rooms; end of trial). Sows were weighed using a crate scale with an electronic load cell (Paul Livestock, Duncan, OK). Back fat depth and LEA were measured using an Aloka 500V real-time ultrasound machine (Corometrics Medical Systems, Wallingford, CT) fitted with a 3.5-MHz linear array transducer. Back fat depth and LEA measurements were made by using internal calipers within the ultrasound and at the dorsal midline at the 10th rib, as previously described (Thiel-Cooper et al., 2001). The hair was shaved from the area, and the skin was marked with a permanent marker at the vertical boundaries of the shaved area so that the same location could be used for subsequent measurements.

Piglet Management and BW Measurements

Piglets were weighed and underwent routine processing procedures (ear notching, tail docking, castration, and supplemental iron injection) 2 d after farrowing. Piglets were weighed individually again at d 10 and at weaning.

Ovarian Ultrasound Examination of Sows after Weaning

Transrectal ultrasonography was performed to monitor follicular development using procedures described by Bracken et al. (2003). An Aloka 900 ultrasound machine (Corometrics Medical Systems, Wallingford, CT) fitted with a 7.5-MHz linear array transducer was used. Once the ovary was identified, a series of images were captured and used to measure the diam. of 5 follicles on either ovary at their maximum size. The mean follicular diam., median follicular diam., and largest follicle size were determined. The examination began on the day of weaning and continued until ovulation or until d 54 (10 d after weaning). Day of ovulation was defined as the day when preovulatory follicles disappeared from the ovary.

Feeding and Feed Intake Measurement

Sows in gestation were housed individually and were also fed individually by using rubber feed pans placed on the floor. The sows were fed a corn–soybean meal–based diet once daily at 0615 h (Table 1). Feed offered and refused was recorded by using an AccuWeigh scale (model BD11-200PK, Metro Equipment Company, Sunnyvale, CA). Sows were offered 1.8 kg per day during gestation. After 45 min, the feed pans were removed, and the remaining feed (including any spilled and unconsumed feed) was weighed. Individual feed intake was calculated by subtracting the weight of feed refused from the weight of the feed offered.


View Full Table | Close Full ViewTable 1.

Composition of lactation and breeding/gestation diets (% expressed on an as-fed basis)

 
Item Lactation diet Breeding/gestation diet
Ingredient
    Corn 64.6 69.4
    Soybean meal (48%) 28.3 15.0
    Soy hulls 10.0
    Choice white grease 2.5 1.0
    Dicalcium phosphate 2.4 2.3
    Limestone 0.8 1.0
    Salt 0.5 0.5
    Lysine 0.1
    Vitamin premixes2 0.9 0.9
Calculated nutrient composition1
    ME, kcal/kg 3,373 3,150
    NE, kcal/kg 2,460 2,310
    CP, % 18.9 14.1
    Digestible lysine,% 0.98 0.59
    Calcium, % 0.93 0.99
    Phosphorus, % 0.82 0.74
2Supplied per kilogram of complete diet: vitamin A, 11,000 IU; vitamin D3, 1,100 IU; vitamin E, 44 IU; vitamin B12, 0.03 mg; vitamin K as menadione sodium bisulfite complex, 4.0 mg; riboflavin, 8.25 mg; D-pantothenic acid as D-calcium pantothenate, 28.1 mg; niacin, 33.0 mg; choline as choline chloride, 550 mg; folic acid, 1.65 mg; biotin, 0.22 mg; iron as iron sulfate, 165 mg; zinc as zinc sulfate, 165 mg; manganese as manganese sulfate, 33 mg; copper as copper sulfate, 16.5 mg; iodine as ethylenediamine dihydroiodide, 0.3 mg; selenium as sodium selenite, 0.3 mg; zinc as zinc methionine, 100 mg

Sows that had farrowed were fed a lactation diet that was offered 2 times a day at 0615 and 1400 h (Table 1). The amount of feed offered after farrowing was increased by 0.9 kg per day increments depending on feed consumption. Sows that consumed the previous meal in its entirety were offered additional feed. Sows that failed to consume the previous meal in its entirety were offered less feed. Sows that had been moved into the farrowing room but had not farrowed were fed 1.8 kg of feed (equally divided into 2 feedings) before farrowing.

The weight of feed offered was recorded at both the morning and afternoon feeding times. Feeders were cleaned once daily in the afternoon. Any feed remaining in the feeder at the time of cleaning was weighed before sows were fed in the afternoon. The weight of feed consumed for a given day was the sum of the feed offered at the previous afternoon feeding and the feed offered in the morning minus any remaining feed that was found when the feeders were cleaned.

Sows were weaned and moved from the farrowing rooms into the breeding rooms where they were fed and housed individually. Weaned sows were fed the breeding/gestation diet (Table 1) once daily. The amount of feed offered was based on body condition (thinner sows receiving additional feed and heavier sows receiving less feed). Sows with a BCS of less than or equal to 2 (based on subjective determination; Patience et al., 1995) were offered 2.9 to 3.2 kg of feed per day. Sows with an average BCS (equal to 3) were offered 2.7 kg per day. Sows with a BCS of greater than or equal to 4 were fed 1.8 to 2.3 kg per day. Sows were fed by using rubber feed pans on the floor. Feed intake was calculated by subtracting the weight of refused feed 45 min after feeding from the weight of offered feed.

Calculation of ME Intake and Prediction of Energy Output in Milk of the Sow

Metabolizable energy of intake (MEI) was calculated by using the energy in the feed multiplied by the feed consumed by the sow. The energy output of the milk of the sow was calculated by using the growth performance of the piglet (Noblet et al., 1990). The equation used was (Noblet et al., 1990) ME for milk production (Mcal/d) = (((2.54 × piglet ADG)+(78.7 × piglet BW)+153) × number of piglets in litter)/1000.

Blood Sampling and Processing

Blood samples were collected from sows using jugular venipuncture beginning at 0900 h. A 9-mL Luer Microvette Z 92 × 16.5 mm tube (Sarstedt Inc., Newton, NC) and a 4-inch hypodermic needle (Air-Tite Products Company Inc., Virginia Beach, VA) were used. Samples were stored temporarily on ice. Samples were obtained on d 15 (gestation room; approximately 1 wk before farrowing), on d 29 and 43 (farrowing room; first and third weeks of lactation), and in the breeding room (d 47; 3 d after weaning). Blood was centrifuged (15 min at 1,500 × g at 4° C), and serum was collected and frozen at -20°C.

Hormone and Metabolite Assays

Glucose concentrations were determined for d 15, 29, 43 and 47 using a liquid glucose reagent set according to the manufacturer’s instructions (Pointe Scientific, Inc., Canton, MI). The intra- and inter-assay CV were 3.4 and 3.0%, respectively. Nonesterified fatty acid concentrations were measured for d 15, 29, 43 and 47 using the Wako NEFA-HR(2) Microtiter Procedure (Wako Diagnostics, Richmond, VA). The intra- and inter-assay CV were 10.7 and 9.2%, respectively. Serum estradiol concentrations were measured on d 43 and 47 by radioimmunoassay as previously described (Kirby et al., 1997) and validated for porcine serum (Liu et al., 2000). The intra- and inter-assay CV were 8.2 and 15.6%, respectively. Insulin-like growth factor 1 concentrations were measured for d 15, 29, 43 and 47 using radioimmunoassay (Liu et al., 2000). All IGF1 samples were assayed in a single assay with an intra-assay CV of 9.8%. Serum insulin concentrations were measured for d 15, 29, 43 and 47 by using the Mercodia Porcine Insulin ELISA (Mercodia AB, Uppsala, Sweden). The intra- and inter-assay CV were 5.5 and 7.9%, respectively.

Statistical Analyses

Data with 1 observation per sow (e.g., gestation length, born alive, weaning to estrus interval) were analyzed by using the general linear models procedure (PROC GLM; SAS Inst. Inc., Cary, NC). The model included the main effects of treatment, group, genotype, and interactions of the main effects. The effects of treatment were partitioned into 3 contrasts denoted as contrast 1 (TN-TN-TN and TN-HS-TN vs. HS-TN-HS and HS-HS-HS; sows in a TN breeding/gestation room vs. sows in a HS breeding/gestation room), contrast 2 (TN-TN-TN and HS-TN-HS vs. TN-HS-TN and HS-HS-HS; sows in a TN farrowing room vs. sows in a HS farrowing room), and contrast 3 (TN-TN-TN and HS-HS-HS vs. TN-HS-TN and HS-TN-HS; treatment interaction). Genotype and group were generally not significant, so the results of a simplified model that only included treatment are reported.

Rectal temperature and respiration rate data were analyzed with a repeated measures design by using a mixed models procedure (PROC MIXED, SAS Inst.). Sow nested within treatment, group, and genotype was included in the model as a random effect. Day was included in the model as a repeated variable. Covariance structures [CS (compound symmetry), CSH (heterogeneous compound symmetry), AR(1) (first order autoregressive), etc.] were tested and the most appropriate covariance structure [based on lowest Akaike’s information criterion (AIC), Akaike’s corrected information criterion (AICc), and Bayesian information criterion (BIC)] was used for each analysis (Littell et al., 1998). The original model included the effects of treatment, group, genotype, day, hour, and all interactions. Genotype and group were generally not significant, so the results of a simplified model that did not include genotype and group are reported.

Data that were collected once daily on repeated days (for example, feed intake) were analyzed with a repeated measure design by using PROC MIXED. The approach used was identical to the mixed model described in the previous paragraph except that “hour” was not included in the analysis (single observation per day). In the analyses of rectal temperature, respiration rate, and feed intake, the effect of day was expressed as “day” representing the day of trial or as “day” representing the day relative to farrowing (d -2 to 20 were analyzed). Day relative to farrowing was defined as the calendar date minus farrowing date (i.e., farrowing = d 0, 1 d after farrowing = d 1, etc.). An effect of “period” was included in some analyses, where period represented d 1 to 20 (period 1; pregnant sows in breeding/gestation), d 21 to 44 (period 2; lactating sows in the farrowing room), or d 45 to 53 (period 3; weaned sows). In some cases, PROC GLM was used for the analyses of repeated measures by using a “split plot in time” analysis (Littell et al., 1998) so that means could be separated by using the Duncan’s multiple range test option in PROC GLM. Means are reported as least square means ± SEM unless stated otherwise. Proportions were tested by using χ2. Statistical significance was declared at P < 0.05. Statistical tendency was defined as 0.05 < P < 0.10.


RESULTS

Farrowing and Litter Data

Overall, 9 out of 58 sows (15.5%) were induced to farrow with PGF alone, 37 out of 58 sows (63.8%) were induced to farrow with PGF + oxytocin, and 12 out of 58 sows (20.7%) were not treated (Table 2). The HS-HS-HS sows were at a greater frequency in the PGF-alone group compared with the other treatments (P < 0.016). A greater number of sows in TN farrowing were induced with PGF + oxytocin (contrast 2, P < 0.014). Contrast 1 tended to be significant for gestation length (P < 0.056) because sows in the HS gestation rooms had a shorter gestation length than sows in the TN gestation rooms (Table 2). Ten sows (17%) required minor assistance at farrowing, and 5 sows (9%) required major assistance at farrowing. There was no effect of treatment for minor or major assistance at farrowing.


View Full Table | Close Full ViewTable 2.

Farrowing and litter data for sows that were treated with 1 of 4 temperature treatment sequences during a production cycle

 
Treatment1
Item TN-TN-TN TN-HS-TN HS-TN-HS HS-HS-HS P2
n 15 14 14 15
Trial day at farrowing 23.5 ± 0.4 24.0 ± 0.4 23.7 ± 0.4 23.0 ± 0.4 NS
Gestation length,3 d 114.7 ± 0.3 114.9 ± 0.3 114.2 ± 0.3 114.1 ± 0.3 NS
Induction protocol
    PGF only, n 2 1 0 6 <0.02
    PGF + oxytocin,4 n 11 10 12 4 <0.01
Litters at birth
    Total born, n 11.5 ± 0.7 11.7 ± 0.7 11.6 ± 0.7 11.8 ± 0.7 NS
    Born alive, n 10.9 ± 0.6 10.8 ± 0.7 11.3 ± 0.7 11.0 ± 0.6 NS
    Stillborn, n 0.40 ± 0.20 0.47 ± 0.21 0.13 ± 0.21 0.60 ± 0.20 NS
    Mummies, n 0.27 ± 0.16 0.50 ± 0.16 0.20 ± 0.16 0.20 ± 0.16 NS
Piglets weaned 10.3 ± 0.4 10.6 ± 0.4 10.3 ± 0.4 10.2 ± 0.6 NS
Piglet BW
    BW at d 2, kg 1.47 ± 0.13 1.40 ± 0.11 1.40 ± 0.11 1.57 ± 0.11 NS
    BW at 10 d, kg 3.45 ± 0.14 3.13 ± 0.14 3.22 ± 0.12 3.13 ± 0.12 NS
    BW at weaning,5 kg 6.37 ± 0.25 5.73 ± 0.24 6.05 ± 0.21 5.80 ± 0.21 NS
Litter BW gain, kg/d 2.36 ± 0.16 2.24 ± 0.15 2.27 ± 0.13 2.16 ± 0.14 NS
ME intake,6 Mcal/d 10.9 ± 0.5 9.6 ± 0.5 12.3 ± 0.5 9.5 ± 0.4 <0.001
ME milk,7 Mcal/d 9.3 ± 0.7 9.2 ± 0.7 9.8 ± 0.7 9.1 ± 0.7 NS
1The treatments are denoted as TN-TN-TN, TN-HS-TN, HS-TN-HS, and HS-HS-HS, where the series of abbreviations represent the thermoneutral environmental temperature (TN; 18°C to 20°C) or heat stress (HS; 24°C to 30°C) that the sow experienced in gestation-farrowing-breeding.
2P-value for the main effect of treatment. NS = not significant (P > 0.10).
3Contrast 1 (TN-TN-TN and TN-HS-TN vs. HS-TN-HS and HS-HS-HS; sows in a TN breeding/gestation room vs. sows in a HS breeding/gestation room), P < 0.056.
4Contrast 2 (TN-TN-TN and HS-TN-HS vs. TN-HS-TN and HS-HS-HS; sows in a TN farrowing room vs. sows in a HS farrowing room), P < 0.014.
5Contrast 2, P < 0.05.
6Least squares mean ± SEM for ME intake for d 0 to 20 of lactation. Contrast 2, P < 0.001.
7Least squares mean ± SEM for ME of secreted milk for d 0 to 20 of lactation.

Total born, born alive, stillborn, mummies, and piglet BW at processing were not affected by treatment (P > 0.10; Table 2). Litter size after cross-fostering was similar for treatments (10.6 ± 0.2), and the number of piglets at weaning was not affected by treatment. Piglet BW at d 10 was not affected by treatment. Contrast 2 was significant (P < 0.05) for weaning weight because piglets from sows in a TN farrowing room (TN-TN-TN and HS-TN-HS) were heavier than piglets from sows in a HS farrowing room (TN-HS-TN and HS-HS-HS; Table 2). Daily litter BW gain (kg/d) and the daily energy output of the milk of the sow (Mcal of ME/d) were similar for treatments (Table 2).

Thermal Responses

Rectal Temperature.

There was a treatment by hour interaction for rectal temperature of pregnant sows housed in breeding/gestation (Fig. 3A; P < 0.01) and for lactating sows housed in farrowing (Fig. 3B; P < 0.001). The treatment by hour interaction was not significant for weaned sows (Fig. 3C; P > 0.10). During pregnancy and lactation, rectal temperatures increased with increasing ambient temperature within the rooms during the day, and for lactating sows, the daily change in rectal temperature in response to the HS was the greatest (Fig. 3).

Figure 3.
Figure 3.

Least squares means with SEM (value at lower right of each panel) for rectal temperature during the day (0800, 1200, 1600, and 2000 h) for sows that were in (A) gestation (late pregnant sow), (B) farrowing (lactating sow), and (C) breeding (weaned sow). Treatments are TN-TN-TN, TN-HS-TN, HS-TN-HS, and HS-HS-HS, where the series of abbreviations represent the thermoneutral temperature (TN; 18°C to 20°C) or heat stress (HS; 24°C to 30°C) in gestation-farrowing-breeding. There was a treatment by time interaction (P < 0.01) for gestation and farrowing but not breeding.

 

Rectal temperature data were analyzed for the entire trial (d 1 to 53) and are shown for 0800 h (lowest point of the temperature cycle; Fig. 4A) and for 2000 h (highest point of the temperature cycle; Fig. 4B). Rectal temperatures were lowest (P < 0.001) when sows were pregnant and in the breeding/gestation room (d 1 to 20 of the trial), underwent an approximately 1°C increase after the sows farrowed (d 20 to 44 of the trial), and then decreased to a temperature that was intermediate between gestation and farrowing after they were weaned [d 45 to 53 of the trial; Figs. 4A (0800 h) and 4B (2000 h)]. The change in rectal temperature caused by HS was most pronounced in lactating sows at 2000 h (Fig. 4B; d 20 to 44).

Figure 4.
Figure 4.

Least squares means with SEM (value at lower right of each panel) for rectal temperature (A and B) during each day of the trial and (C and D) on each day relative to the day of farrowing. Data are for (A and C) 0800 h and (B and D) 2000 h. Treatments are TN-TN-TN, TN-HS-TN, HS-TN-HS, and HS-HS-HS, where the series of abbreviations represent the thermoneutral temperature (TN; 18°C to 20°C) or heat stress (HS; 24°C to 30°C) in gestation-farrowing-breeding. For (A) and (B), sows were in breeding/gestation from d 0 to 20, were farrowing from d 20 to 44, and were weaned and returned to breeding/gestation on d 44. There was an effect of day (P < 0.001) in (A), (B), (C), and (D) and a treatment by day interaction (P < 0.001) in (B). There was an effect of treatment (P < 0.001) in (D).

 

When data were analyzed relative to the day of farrowing, the increase in rectal temperature after farrowing occurred within 1 d after farrowing and was observed at each time of day [for example, 0800 h (Fig. 4C) and 2000 h (Fig. 4D)]. Sow rectal temperature increased (P < 0.001) to a maximum by 11 d after farrowing and then progressively decreased until weaning. Rectal temperatures were similar for sows in different rooms and exposed to different treatments at 0800 h (Fig. 4C). At 2000 h, sows in the HS farrowing room had greater rectal temperature than sows in the TN farrowing room (Fig. 4D; P < 0.001). The TN-HS-TN sows appeared to have a greater rectal temperature at 2000 h when compared with HS-HS-HS (Fig. 4D), but this difference was not significant (P > 0.10).

Respiration Rate.

There was a treatment by hour interaction for respiration rate for pregnant sows housed in breeding/gestation (Fig. 5A; P < 0.001), lactating sows housed in farrowing (Fig. 5B; P < 0.001), and weaned sows housed in breeding/gestation (Fig. 5C; P < 0.001). At 0800 h in breeding/gestation (end of low temperature; Fig. 5A), average respiration rates for pregnant sows in the HS and TN rooms differed by 13.9 bpm. The difference in respiration rate between HS and TN in breeding/gestation was greater at 1200 h (beginning of maximum temperature), 1600 h (midpoint of maximum temperature), and 2000 h (end of maximum temperature period; Fig. 5A). For lactating sows (Fig. 5B), the increase in respiration rate in response to the afternoon high ambient temperature was greater than that observed for either pregnant (Fig. 5A) or weaned (Fig. 5C) sows. At 0800 h in the farrowing room (Fig. 5B), the magnitude in the difference between respiration rates for HS compared with TN sows was the least. At 1200, 1600, and 2000 h, lactating sows in the HS farrowing room had respiration rates that were nearly twice that observed for sows in the TN farrowing. Relative to pregnant sows and lactating sows, the effect of HS on respiration rate of weaned sows was less. At 0800 h in the breeding/gestation room (Fig. 5C), respiration rate in weaned sows was slightly greater in HS compared with TN sows (P < 0.02). Likewise at 1600 and 2000 h, weaned sows in the HS breeding/gestation room had a slightly greater respiration rate than weaned sows in the TN breeding/gestation room (P < 0.05).

Figure 5.
Figure 5.

Least squares means with SEM (value at lower right of each panel) for respiration rate during the day (0800, 1200, 1600, and 2000 h) for sows that were in (A) gestation (late pregnant sow), (B) farrowing (lactating sow), and (C) breeding (weaned sow). Treatments are TN-TN-TN, TN-HS-TN, HS-TN-HS, and HS-HS-HS, where the series of abbreviations represent the thermoneutral temperature (TN; 18°C to 20°C) or heat stress (HS; 24°C to 30°C) in gestation-farrowing-breeding. There was a treatment by time interaction (P < 0.001) for gestation, farrowing, and breeding.

 

Respiration rate data were analyzed for the entire trial (d 1 to 53) and are shown for 0800 h (lowest point of the temperature cycle; Fig. 6A) and for 2000 h (highest point of the temperature cycle; Fig. 6B). Relative to pregnant sows housed in the breeding/gestation room, respiration rates underwent an increase after sows were moved into the farrowing room and farrowed (P < 0.001). After sows were weaned and moved back to the breeding/gestation room (d 45 to 53), respiration rates decreased to a frequency that was lower (P < 0.001) than during gestation or lactation. When expressed relative to the day of farrowing [Figs. 6C (0800 h) and 6D (2000 h)], respiration rates initially increased (P < 0.001) before farrowing (d -1) and then decreased until d 4. There was a subsequent increase (P < 0.001) in respiration rates from d 4 to 20 after farrowing. Respiration rates were slightly increased in HS sows at 0800 h (Fig. 6C). The difference in respiration rates for HS and TN sows was greater at 2000 h (Fig. 6D).

Figure 6.
Figure 6.

Least squares means with SEM (value at lower right of each panel) for respiration rate (A and B) during each day of the 55 d of the trial and (C and D) on each day relative to the day of farrowing. Data are for (A and C) 0800 h and (B and D) 2000 h. Treatments are TN-TN-TN, TN-HS-TN, HS-TN-HS, and HS-HS-HS, where the series of abbreviations represent the thermoneutral temperature (TN; 18°C to 20°C) or heat stress (HS; 24°C to 30°C) in gestation-farrowing-breeding. For (A) and (B), sows were in the breeding/gestation from d 0 to 20, were farrowing from d 20 to 44, and were weaned and returned to the breeding/gestation on d 44. There was a treatment by day interaction (P < 0.001) for (A) and (B). There were effects of both treatment (P < 0.001) and day (P < 0.001) for (C) and (D).

 

Feed Intake

Sows were limit fed during gestation and averaged 2.3 ± 0.1 kg/d feed intake (Fig. 7A). There was an effect of day (P < 0.001; sows progressively increased feed intake during gestation) but no effect of treatment on feed intake during gestation (Fig. 7A).

Figure 7.
Figure 7.

Least squares means with SEM (value at lower right of each panel) for feed intake (A) during each day of the 55 d of the trial and (B) on each day relative to the day of farrowing. Treatments are TN-TN-TN, TN-HS-TN, HS-TN-HS, and HS-HS-HS, where the series of abbreviations represent the thermoneutral temperatures (TN; 18°C to 20°C) or heat stress (HS; 24°C to 30°C) in gestation-farrowing-breeding. Sows were in the breeding/gestation from d 0 to 20, were farrowing from d 20 to 44, and were weaned and returned to breeding/gestation on d 44. There was a treatment by day interaction (P < 0.001) for (A) and (B).

 

There was an effect of treatment (P < 0.001) for feed intake in farrowing (Fig. 7A). The HS lactating sows had lower feed intake than TN lactating sows. There was also an effect of day (P < 0.001). Feed intake increased from a minimum of 1.2 ± 0.1 kg/d to a maximum of 4.9 ± 0.1 kg/d when sows were in the farrowing room [least squares mean (lsmeans) ± SEM for all treatments combined]. Feed intake was also analyzed according to days after farrowing (Fig. 7B). Sows ate the least amount of feed on the day of farrowing (0.8 ± 0.1 kg/d; lsmeans for all treatments combined). Afterward, feed intake increased (P < 0.001) to 5.1 ± 0.1 kg/d on d 20 of lactation. Sows that came from the HS gestation room and went into the TN farrowing room (HS-TN-HS sows) had the greatest daily lactation feed intake (4.1 ± 0.2 kg/d; Fig. 7B). Sows that stayed in TN from gestation to farrowing (TN-TN-TN) had greater feed intake (3.7 ± 0.2 kg/d) compared with sows that were housed in a HS farrowing room during lactation [HS-HS-HS sows (3.1 ± 0.2 kg/d) or TN-HS-TN sows (3.3 ± 0.2 kg/d)].

Feed intake after weaning was less than feed intake during gestation or after farrowing. There was an effect of day (P < 0.001; sows increased feed intake from a nadir of 1.4 ± 0.1 kg/d at weaning to 2.1 ± 0.1 kg/d on 8 d after weaning) but no effect of treatment on feed intake after weaning (Fig. 7A).

Body Weight, BF, and LEA

Sows entering the trial (d 0) were similar for BW (185 ± 1 kg), BF (2.4 ± 0.1 cm), and LEA (52.0 ± 0.5 cm2). The gain in BW during gestation (d 1 to 20) was greater for sows in the TN room compared with sows in the HS room (14.6 ± 1.0, 13.8 ± 1.1, 11.4 ± 1.1, and 12.0 ± 1.1 kg for TN-TN-TN, TN-HS-TN, HS-TN-HS, and HS-HS-HS, respectively; contrast 1, P < 0.022). Back fat depth (2.5 ± 0.1 cm) and LEA (51.2 ± 0.7 cm2) were similar for HS and TN sows leaving gestation (d 20). Body weight, BF, and LEA were not affected by treatment during lactation [i.e., sows leaving the farrowing room and returning to breeding/gestation room on d 44 were similar for BW (168 ± 2 kg), BF (2.0 ± 0.1 cm), and LEA (47.1 ± 0.6 cm2)]. When sows left breeding (end of trial), they were similar for BW (160 ± 2 kg) and LEA (47.7 ± 0.6 cm2), but there was a tendency for an effect of treatment on BF (P < 0.095). Contrast 1 was significant (P < 0.021) for BF because BF was slightly greater for HS sows (1.7 ± 0.1, 1.7 ± 0.1, 2.1 ± 0.1, and 1.9 ± 0.1 cm for TN-TN-TN, TN-HS-TN, HS-TN-HS, and HS-HS-HS, respectively). There was also an effect of treatment for BF change (P < 0.001). The change in BF from weaning to the end the trial was lower for sows that were housed in HS during breeding (-0.24 ± 0.05, -0.26 ± 0.05, -0.04 ± 0.04, and -0.07 ± 0.04 cm for TN-TN-TN, TN-HS-TN, HS-TN-HS, and HS-HS-HS, respectively; contrast 1, P < 0.001).

Return to Estrus after Weaning, Follicular Development after Weaning, and Subsequent Reproduction

The proportion of sows that returned to estrus and were inseminated (48/56; 85.7%) after weaning was not affected by treatment (Table 3). There were 2 TN-TN-TN sows and 1 TN-HS-TN sow that did not ovulate after weaning. Two TN-TN-TN sows, 2 HS-TN-HS sows, and 1 HS-HS-HS sow ovulated as detected by ultrasound but did not express estrus (silent ovulation). The number of sows with silent ovulations (5/56; 8.9%), weaning-to-estrus intervals (4.7 ± 0.1 d), and length of estrus (2.2 ± 0.1 d) were not affected by treatment (Table 3)


View Full Table | Close Full ViewTable 3.

Breeding and subsequent farrowing data for sows that were treated with 1 of 4 temperature treatment sequences during a production cycle

 
Treatment1
Item TN-TN-TN TN-HS-TN HS-TN-HS HS-HS-HS P2
n3 14 13 14 15
Estrus and inseminated, n (%) 10 (71) 12 (92) 12 (86) 14 (93) NS
Anovulatory, n (%) 2 (14) 1 (8) 0 (0) 0 (0) NS
Silent ovulation, n (%) 2 (14) 0 (0) 2 (14) 1 (7) NS
Weaning to estrus, d 4.7 ± 0.3 4.8 ± 0.3 4.8 ± 0.3 4.7 ± 0.2 NS
Length of estrus, d 2.3 ± 0.2 2.2 ± 0.2 2.1 ± 0.2 2.3 ± 0.2 NS
Second farrowing
    n4 9 12 11 14
    Number farrowing (%) 8 (89) 10 (83) 9 (82) 11 (79) NS
    Total born, n 10.4 ± 0.7 11.5 ± 0.6 10.7 ± 0.7 10.6 ± 0.6 NS
    Born alive, n 9.8 ± 0.6 10.5 ± 0.5 10.3 ± 0.5 10.2 ± 0.5 NS
    Stillborn, n 0.63 ± 0.30 1.00 ± 0.27 0.33 ± 0.28 0.45 ± 0.26 NS
1The treatments are denoted as TN-TN-TN, TN-HS-TN, HS-TN-HS, and HS-HS-HS, where the series of abbreviations represent the thermoneutral environmental temperature (TN; 18°C) to 20°C) or heat stress (HS; 24°C) to 30°C) that the sow experienced in gestation-farrowing-breeding.
2P-value for the main effect of treatment. NS = not significant (P > 0.10).
3Excludes 1 TN-TN-TN sow that died in farrowing and 1 TN-HS-TN sow that was lame (euthanized at breeding).
4Excludes 1 HS-TN-HS sow that died on the farm of the producer and 1 TN-TN-TN sows that had a rectal tear (euthanized at breeding).

On the day of weaning (d 0), mean follicular diam. and median follicular diam. were not affected by treatment. Largest follicle size on d 0 did depend on treatment (P < 0.047), and contrast 2 tended to be significant for the largest follicle (P < 0.065; sows in the TN farrowing room had a slightly larger largest follicle diam. on d 0). On d 2 and 4 after weaning, mean diameter (Fig. 8A), largest follicle, and median diam. were not affected by treatment. Serum estradiol concentrations increased (P < 0.001) from 1 d before weaning to 3 d after weaning but were not affected by treatment (P > 0.10; Fig. 8B).

Figure 8.
Figure 8.

Least squares means and SE (bar) for (A) average diameter of ovarian follicles and (B) serum estradiol on day relative to weaning. Treatments are denoted as TN-TN-TN, TN-HS-TN, HS-TN-HS, and HS-HS-HS, where the series of abbreviations represent the thermoneutral temperature (TN; 18°C to 20°C) or heat stress (HS; 24°C to 30°C) that the sow experienced in gestation-farrowing-breeding. There was an effect of day (P < 0.001) for (A) and (B).

 

The number of sows that farrowed their second litter (82.6%) was not affected by treatment (Table 3). Total born (10.8 ± 0.3), born alive (10.2 ± 0.3), stillborn (0.6 ± 0.1), and weaned (9.6 ± 0.2) were not affected by treatment (Table 3).

Serum Glucose, NEFA, IGF-1, and Insulin Concentrations

Serum glucose concentrations were not affected by day (Fig. 9A). There was an effect of day on serum NEFA concentrations (Fig. 9B). Serum NEFA were least on d 15 (gestating sow) and were greatest on d 29 (sows approximately 1 wk after farrowing; P < 0.001). There was an effect of day (P < 0.001) for serum IGF-1 concentrations (Fig. 5C). Serum IGF-1 concentrations were greatest on d 43 (day before weaning) and 47 (3 d after weaning). There was an also an effect of day (P < 0.001) on serum insulin concentrations (Fig. 5D). Relative to other days, serum insulin concentrations were lowest on d 15.

Figure 9.
Figure 9.

Least squares means for (A) serum glucose, (B) NEFA, (C) IGF-1, and (D) insulin concentrations for sows at d 15 (during gestation), d 29 (after farrowing), d 43 (day before weaning), and d 47 (3 d after weaning). Treatments are TN-TN-TN, TN-HS-TN, HS-TN-HS, and HS-HS-HS, where the series of abbreviations represent the thermoneutral temperature (TN; 18°C to 20°C) or heat stress (HS; 24°C to 30°C) that the sow experienced in gestation-farrowing-breeding. x–zMeans with different superscripts differ across day of trial (P < 0.05). a,bWithin means, different superscripts differ (P < 0.05).

 

There was no effect of treatment on serum glucose or serum NEFA. There was an effect of treatment on IGF-1 concentrations on d 43 (day before weaning; P < 0.05). The HS-TN-HS sows had the greatest serum IGF-1 concentrations on d 43. The lowest serum IGF-1 concentrations on d 43 were found in TN-HS-TN and HS-HS-HS sows (Fig. 9C). There was an effect of treatment on insulin concentrations on d 29 (1 wk after farrowing; P < 0.05). The HS-TN-HS sows had the greatest serum insulin concentrations on d 29, and the lowest serum insulin concentrations on d 29 were found in TN-HS-TN sows.


DISCUSSION

The sows that were studied in this trial had a large increase in rectal temperature (1°C to 1.5°C) after farrowing. This increase in rectal temperature occurred regardless of treatment and far exceeded any treatment effects on rectal temperature. The uniform and large increase in rectal temperature was apparently caused by the increase in energy intake and metabolic rate for the lactating sow. Earlier studies also reported a large increase is sow rectal temperature within 5 d after farrowing (Kelley and Curtis, 1978). The data collected during this study extend the findings of Kelley and Curtis (1978) by reporting rectal temperature 4 times daily for 3 wk before farrowing, for an entire lactation, and also for approximately 10 d after weaning. We have performed similar studies in typical production facilities and found an increase in rectal temperature after farrowing (Martin et al., 2012). In these subsequent production studies we also found that the increase in rectal temperature is for the entire lactation and that weaning leads to lower rectal temperature.

Although the largest changes in rectal temperature appear to be a consequence of lactation in the pig, there were also effects of the HS on rectal temperature. These HS effects were observed during gestation and lactation but were considerably less after weaning (breeding period). Data from previous studies are in agreement with our results that show that HS sows have greater rectal temperatures during gestation (Omtvedt et al., 1971; Liao and Veum, 1994) and lactation (Prunier et al., 1997; Messias de Braganca et al., 1998). During gestation, there was a small increase in rectal temperature (approximately 0.1°C) in sows in the HS rooms. The effect of heat stress in lactating sow was greater (approximately 0.3°C increase) and was primarily manifested in later lactation perhaps because sows were consuming more feed, were producing more milk, and were more metabolically active at that time. Our data agree with the generally accepted premise that lactating sows have a lower upper critical temperature than gestating or weaned sows. Black et al. (1993) estimated that the TN zone for the lactating sow was 12°C to 22°C. The HS temperature cycle that we employed was well above this TN zone, and lactating sows could not control body temperature at the high point of the temperature cycle. After weaning, rectal temperature decreased to a temperature that was intermediate between what we observed during gestation and during lactation. This intermediate temperature may indicate that weaned sows experience a carry-over effect of the lactation. Perhaps the mammary gland continues to be metabolically active shortly after weaning, and this metabolic activity generates heat. In a recent study of sows in a controlled environment, however, we observed that rectal temperature of sows slowly decreased during gestation and reached basal values during the final month of pregnancy (Lucy et al., 2012a). The difference in rectal temperature between gestation and breeding, therefore, may be because sows have inherently low rectal temperature in late gestation (a process that occurs over time in the pregnant sow for an unknown reason).

We did observe a greater respiration rate in sows that were housed under HS conditions. This effect existed across the 3 phases of production, and our observations agree with other studies (Machado-Neto et al., 1987; Liao and Veum, 1994; Spencer et al., 2003). For example, both Johnston et al. (1999) and Spencer et al. (2003) found a 2-fold increase in respiration rate in lactating sows exposed to HS. The increase in respiration that they observed was similar to what we observed in this study for lactating sows. Gestating sows that were exposed to HS conditions had greater respiration rates than during TN exposure, and there was an increase in respiration rate for both groups within 5 d after farrowing. The difference between HS and TN in respiration rate was greatest after farrowing (lactating sow) and was least after weaning (nonlactating sow). The further increase in respiration rates for HS compared with TN during lactation is in agreement with our general understanding of the lower upper critical temperature for lactating sows (Black et al., 1993). Lactating sows have a greater metabolic rate (with associated heat production) and attempt to cool themselves through an increase in respiration rate. An intriguing aspect of the data, however, is that the change in respiration rate for sows under TN condition was relatively small (approximately 10 bpm despite the fact that they had a nearly 1°C increase in rectal temperature associated with lactation. Apparently, the large change in rectal temperature in the lactating sow does not invoke a major compensatory thermal response in terms of respiration rate. Respiration rate in TN sows only increased in the second half of lactation, and this increase perhaps explains the slow decline in rectal temperature that was observed at that time. When the sows were weaned, respiration rates decreased, and HS sows retained a greater respiration rate when compared with TN sows. The HS and TN sows in breeding had lower respiration rates when compared with HS and TN sows in gestation despite an apparently greater rectal temperature (relative to gestation) for sows in the breeding phase. It is possible that there was a shift in the thermoregulatory set point as the sow progresses from gestation to lactation and after weaning. Although this has significant implications for heat stress management of this species, the underlying mechanisms that control this change in set point are completely unknown.

Sows increased feed intake after farrowing, and sows in TN conditions (TN-TN-TN and HS-TN-HS) consumed more feed than HS sows (HS-HS-HS and TN-HS-TN). The reduction in feed intake for HS vs. TN in our study was approximately 20% (last week of lactation). This decrease was considerably less than that reported by others, including Johnston et al. (1999), Renaudeau et al. (2001), and Spencer et al. (2003), whose studies showed an approximately 40% reduction in feed intake during HS. The difference may be explained by the HS conditions that were imposed (e.g., actual temperatures, temperature cycle, humidity) or perhaps because we used primiparous sows, whereas mixed parity structures were used in the other studies. Regardless of treatment, sows seemed to have reduced feed intake between d 4 and 7 after farrowing. Previous studies have determined that feed intake for HS sows was less than for TN sows but made no mention of the decrease in feed intake during early lactation (Teague et al., 1968; Prunier et al., 1997; Renaudeau et al., 2001; Spencer et al., 2003). Perhaps the sow has a delayed response to the lactation-induced increase in body temperature and reduces her feed intake in an attempt to decrease internal heat production.

The HS sows weaned piglets that were approximately 0.5 kg lighter. This represented an approximately 7% decrease in weaning weight. The decrease in BW gain is less than that reported by Spencer et al. (2003) for primiparous sows (11% decrease). Primiparous sows are apparently less susceptible to heat stress–induced reductions in piglet weight gain because Johnston et al. (1999), Renaudeau and Noblet (2001), and Spencer et al. (2003) all reported 17% or greater reductions in piglet BW gain when multiparous sows were used. We estimated the milk production of the sows in this study by using the equation derived by Noblet et al. (1990). We found that there was no effect of HS on sow milk production. One collective interpretation of the previously published work and our work is that the primiparous sow has a lower reduction in milk production during HS. The HS-induced reduction in feed intake without a decrease in milk production may exacerbate lactation BW loss and explain reproductive problems in first-parity sows during the summer. The underlying physiology and endocrinology that lead to sustained milk production during HS conditions in a primiparous sow but not a multiparous sow is an important topic for future study.

The concentration of blood NEFA increased after parturition and then declined as lactation progressed. The increase in NEFA for all sows after farrowing suggests that these sows were mobilizing energy for milk production. As lactation progressed, NEFA concentrations decreased, perhaps because sows were mobilizing less body fat when they were in more positive energy balance. Weaning led to a further decrease in NEFA that coincided with lower milk production and endocrine changes in the sow. Blood glucose was not different among treatments or across days. Apparently, the sow is fully capable of adapting to the glucose requirements for lactation and homeostatically regulating her blood glucose.

Serum IGF-1 concentrations increased in the present study as sows moved from gestation through farrowing and into breeding. There were treatment differences for IGF-1 on the day before weaning (d 43). Sows that went from HS gestation to TN farrowing had the greatest concentration of IGF-1 on d 43. This treatment group had the greatest feed intake, indicating that the increase in IGF-1 could perhaps be explained by the increase in feed intake. The sows in HS had the lowest concentrations of IGF-1, and these sows had lower feed intake than TN sows. Other studies have also reported that undernourished sows have decreased concentrations of IGF-1 (van den Brand et al., 2001; Hunter et al., 2004). This may be the first report showing that an environmental treatment can affect IGF-1 concentrations in late lactation. Serum insulin concentrations increased after farrowing and were also affected by treatment (TN sows having greater insulin than HS sows on d 29). The greater insulin concentrations may be explained by greater feed intake in lactating sows. The effects of treatment on insulin and IGF-1 were similar (TN sows having greater hormonal concentrations than HS sows) but occurred on different days (d 29 for insulin and d 43 for IGF-1).

The HS-TN-HS sows had the greatest concentrations of IGF-1 and insulin and also had larger follicles at weaning. Blood concentrations of insulin and IGF-1 are believed to be stimulatory to follicular growth through a synergistic effect on gonadotropins (Quesnel, 2009). Blood gonadotropin concentrations are reduced before weaning, and their activity may be increased at the cellular level through the actions of insulin and IGF-1. Alternatively, the increase in feed intake may have increased gonadotropin secretion (theoretically, LH in this case; Kemp and Soede, 2012), and this increase in LH may have led to an increase in follicular growth. Regardless, the follicular populations quickly normalized after weaning when insulin IGF-1 concentrations were also similar.

There was a clear metabolic and thermoregulatory response to the HS that we imposed. We were surprised, therefore, to find that HS had minimal effects on most aspects of reproduction that we studied. We did find that sows moving from a HS gestation to HS farrowing (HS-HS-HS sows) were less likely to be treated with both PGF and oxytocin to induce parturition. This perhaps indicates that these sows were starting to farrow earlier than the other groups. There were no treatment effects on total born, born alive, stillborn, mummies, or piglet BW at processing. Sows were HS for the final 20 d of gestation, and this relatively short treatment may be 1 reason for the lack of differences in litter data. In a different study, we did observe effects on piglet birth weight and subsequent piglet weights when HS was applied across an entire gestation (Lucy et al., 2012b).

Gross measures of reproductive performance after weaning were not affected by treatment. There were no differences in weaning-to-estrus intervals, the number of anestrus sows, or the number of sows with silent ovulations. Previous studies have reported prolonged weaning-to-estrus intervals (Cox et al., 1983; Clark et al., 1986) and an increase in the incidence of anestrus (Teague et al., 1968; Johnston et al., 1999) for HS sows, but we did not observe these responses in this study. We did measure follicular growth by using ultrasound to determine if the treatments applied had an impact on the development of ovarian follicles. There was an effect of treatment on follicular growth on the day of weaning (sows exposed to TN during lactation having larger mean largest follicle diam.). But by d 2 after weaning, the advantage in terms of follicular development in the TN sows had disappeared. A previous study indicated a delay in follicular growth for HS sows (Lucy et al., 2001). The sows in the aforementioned study were subjected to a more severe HS. Sows were inseminated at the end of the trial, and there were no effects on the subsequent litter. This again was unexpected because 1 consequence of seasonal infertility is low farrowing rates and low litter size (Bertoldo et al., 2012).

An important question that arises from this work is why the HS sows did not have a loss in reproductive performance. One obvious consideration was whether or not an adequate stress was applied to the sows. The maximum temperature applied was 30°C for approximately 8 h. We do know that the HS that we applied increased rectal temperature and respiration rates, but perhaps the stress that we applied was not great enough to recapitulate observed changes in reproduction that occur under field conditions. It is also possible that the regular temperature cycle that the sows experienced within the environmental chambers enabled an acclimation to high ambient temperatures that could not be achieved under natural weather conditions, which can create erratic ambient temperature cycles inside swine barns. There are farms that do not experience seasonal infertility. For example, in the study by Auvigne et al. (2010), approximately one-third of the herds studied did not experience lower fertility in summer. It is possible that the intense management that we imposed to keep pigs clean and healthy in the environmental chambers overcame the effects of the HS.

In summary, heat stress caused an increase in rectal temperatures and respiration rates. Sows seemed to undergo changes in their thermoregulatory set points as they progressed from gestation to farrowing to breeding. In some cases, these changes in set point led to the complete uncoupling of thermoregulatory responses (respiration rate) from the internal body (rectal) temperature of the animal. Heat stress during lactation caused a decrease in feed intake. Although the sows were HS and had reduced feed intake, the effects of the treatments were minimal in terms of the typical measures of reproduction in sows (weaning-to-estrus intervals, the incidence of anestrus, and subsequent farrowing performance). Effects on piglet weaning weight were small, perhaps because HS did not cause reduced milk production in the sows. In terms of cooling sows within a production system, lactating sows had the greatest rectal temperature and appeared to be more sensitive to HS. This finding agrees with our knowledge of a lower upper critical temperature in lactating sows. Producers choosing to alleviate heat stress should focus their cooling efforts during lactation. Future investigations on heat stress are needed to further characterize reproductive responses for sows in a commercial production system and determine what aspects of commercial production when combined with HS lead to summertime infertility.

 

References

Footnotes


Comments
Be the first to comment.



Please log in to post a comment.
*Society members, certified professionals, and authors are permitted to comment.