5269

Effects of high fiber intake during late pregnancy on sow physiology, colostrum production, and piglet performance1

F. Loisel,*†‡ C. Farmer,§ P. Ramaekers,‡ and H. Quesnel*†2

*INRA, UMR1348 PEGASE, F-35590 Saint-Gilles, France; †Agrocampus Ouest, UMR1348 PEGASE, F-35000 Rennes, France; ‡Nutreco R & D, 5832 AE Boxmeer, The Netherlands;

and §Agriculture and Agri-Food Canada, Dairy and Swine R & D Centre, Sherbrooke, Quebec J1M 0C8, Canada

ABSTRACT: Dietary fiber given during pregnancy may influence sow endocrinology and increase piglet BW gain during early lactation. The aim of the current study was to determine whether dietary fiber given to sows during late pregnancy induces endocrine changes that could modulate sow colostrum production and, thus, piglet performance. From d 106 of pregnancy until parturition, 29 Landrace × Large White nulliparous sows were fed gestation diets containing 23.4 [high fiber (HF); n = 15] or 13.3% total dietary fiber [low fiber (LF); n = 14]. In the HF diet, wheat and barley were partly replaced by soybean hulls, wheat bran, sunflower meal (undecorticated), and sugar beet pulp. After parturition, sows were fed a standard lacta-tion diet. Colostrum production was estimated during 24 h, starting at the onset of parturition (T0) and ending at 24 h after parturition (T24) based on piglet weight gains. Jugular blood samples were collected from sows on d 101 of pregnancy, daily from d 111 of gestation to d 3 of lacta-tion, and then on d 7 and 21 of lactation (d 0 being the day of parturition). Postprandial kinetics of plasma glucose and insulin concentrations were determined on d 112 of pregnancy. The feeding treatment did not influence sow

colostrum yield (3.9 ± 0.2 kg) or piglet weight gain during the first day postpartum to d 21 of lactation. Colostrum intake of low birth weight piglets (< 900 g) was greater in litters from HF sows than from LF sows (216 ± 24 vs. 137 ± 22 g; P = 0.02). Preweaning mortality was lower in HF than LF litters (6.2 vs. 14.7%; P = 0.01). Circulating concentrations of progesterone, prolactin, estradiol-17β, and cortisol were not influenced by the treatment. Sows fed the HF diet had greater postprandial insulin concen-trations than LF sows (P = 0.02) whereas the postprandial glucose peak was similar. At T24, colostrum produced by HF sows contained 29% more lipid than colostrum pro-duced by LF sows (P = 0.04). Immunoglobulin A concen-trations in colostrum were lower at T0 and T24 (P = 0.02) in HF than LF sows (at T0: 8.6 ± 1.1 vs. 11.9 ± 1.1 mg/mL; at T24: 2.5 ± 0.7 vs. 4.8 ± 0.7 mg/mL). In conclusion, dietary fiber in late pregnancy affected sow colostrum composition but not colostrum yield, increased colos-trum intake of low birth weight piglets, and decreased preweaning mortality, but these effects were not related to changes in peripartum concentrations of the main hor-mones involved in lactogenesis.

Key words: colostrum composition, colostrum yield, dietary fiber, endocrinology, piglet preweaning mortality, sow metabolism

© 2013 American Society of Animal Science. All rights reserved. J. Anim. Sci. 2013.91:5269–5279 doi:10.2527/jas2013-6526

INTRODUCTION

According to animal welfare legislation in the European Union (Council of European Union, 2001), sows have to be provided with sufficient quantities of bulky or high-fiber feed during pregnancy to reduce feeding frustration. Bulky diets given during preg-nancy were also shown to increase sow voluntary feed intake during lactation and to have beneficial effects on piglet performance (Matte et al., 1994; Guillemet et al., 2007). When given for 3 mo, a bulky diet increased

1The authors gratefully acknowledge F. Legouevec, M. Lefèbvre, D. Boutin, Y. Surel, A. Chauvin, J. Georges, V. Piedvache, A. Condé, Y. Jaguelin, A. Starck, R. Comte, F. Thomas, G. Guillemois, J. F. Rouaud (INRA, Saint-Gilles, France), A. Bernier, and L. Thibault (Agriculture and Agri-Food Canada) for their technical efficient assistance. They also thank J. Y. Dourmad and J. Noblet (INRA, Saint-Gilles, France) for their helpful advices. Financial support was provided by Nutreco R & D (Boxmeer, The Netherlands).

2Corresponding author: Helene.Quesnel@rennes.inra.frReceived March 28, 2013.Accepted July 31, 2013.

Published November 24, 2014

Loisel et al.5270

litter weight gain during the first week of lactation (Guil-lemet et al., 2007). Moreover, a bulky diet given during the last 20 d of pregnancy and the first week of lactation increased piglet weight gain during the first 5 d of lacta-tion (Oliviero et al., 2009). The mechanisms underlying the effects of dietary fiber on piglet performance have not yet been elucidated.

Colostrum intake by piglets is a major factor influ-encing piglet weight gain and survival during early lac-tation (Dyck and Swierstra, 1987; Devillers et al., 2011). Piglet colostrum intake depends, in part, on their ability to extract colostrum from teats but also on the ability of sows to produce a high yield of colostrum (Devillers et al., 2007). In swine, lactogenesis is under endocrine control and both progesterone and prolactin are key hor-mones for this process (Taverne et al., 1982; Farmer et al., 1998). Farmer et al. (1995) and Quesnel et al. (2009) showed that dietary fiber given to sows during preg-nancy tended to increase prepartum prolactin concen-trations. Moreover, in lactating cows, dietary fiber de-creased insulin concentrations, thereby increasing pro-gesterone hepatic clearance and presumably decreasing plasma progesterone (Lemley et al., 2010). We hypothe-sized that a high fiber intake during late pregnancy could induce peripartum endocrine changes that increase sow colostrum yield and, in turn, influence piglet growth and survival during lactation. The present study aimed at in-vestigating this hypothesis.

MATERIALS AND METHODS

The study was conducted at INRA, Saint-Gilles, France, and animals were cared for according to the French regulations for the humane care and use of ani-mals in research. The experimental protocol was ap-proved by the local Ethics Committee in Animal Experi-ment of Rennes, France.

Animals and Experimental Design

Thirty-one Landrace × Large White nulliparous sows were used in 5 replicates of 6 or 7 females. At approxi-mately 285 d of age, sows were inseminated with semen from Piétrain boars. After insemination, sows were housed in groups of approximately 6 in a pen with concrete floors (5 by 3.5 m) without straw. The room was equipped with individual feeding stalls. Sows were moved to the farrow-ing room on d 105 of pregnancy and were kept in indi-vidual farrowing crates (2 by 2.5 m) thereafter.

From d 1 to 91 of pregnancy, sows were fed 2.6 kg/d of a conventional gestation diet (as-fed basis: 9.54 MJ of NE/kg, 13.3% CP, 0.6% Lys, and 5% crude fiber). Sows were then allocated to 1 of 2 experimental diets accord-ing to backfat thickness and BW on d 91. The experi-

mental diets contained 13.3 or 23.4% total dietary fiber (TDF) [low and high fiber (LF and HF, respectively)]. The LF diet was based on wheat, barley, and soybean meal (Table 1). In the HF diet, which contained 32% fi-brous ingredients, wheat and barley were partly replaced by a mixture of soybean hulls, wheat bran, sunflower meal (undecorticated), and sugar beet pulp. The LF and HF diets contained 11.4 and 20.6% insoluble fiber and 1.9 and 2.8% soluble fiber, respectively [analyzed ac-cording to the methods reported by Prosky et al. (1985);

Table 1. Composition of the experimental diets containing 13.3 [low fiber (LF)] or 23.4% [high fiber (HF)] total dietary fiber, which were given to sows during late pregnancyItem LF HFIngredient, % (as-fed basis)

Barley 41.69 31.31Wheat 41.69 31.31Soybean meal 11.50 1.45Soybean hulls – 8.00Wheat bran – 8.00Sunflower meal (undecorticated) – 8.00Sugar beet pulp – 8.00Rapeseed oil 2.00 1.45l-Lys HCl 0.05 0.16l-Thr – 0.05Calcium carbonate 0.72 0.23Dicalcium phosphate 1.40 1.20Salt 0.45 0.34Vitamin and mineral premix1 0.50 0.50

Chemical analysis2

Total dietary fiber, % 13.3 23.4Insoluble fiber, % 11.4 20.6Soluble fiber,3% 1.9 2.8Crude fiber, % 3.3 7.9NDF, % 13.1 20.3ADF, % 4.2 9.4ADL, % 0.6 1.3CP, % 13.8 13.0Lipids, % 3.4 3.2Starch, % 46.1 36.3Ash, % 4.6 4.6DM, % 87.5 87.5GE, MJ/kg 16.13 16.10

Nutritional value4

NE, MJ/kg 9.94 8.81Lys, % 0.63 0.601Supplied the following amounts per kilogram of diet: vitamin A, 10,000

IU; vitamin D3, 1,500 IU; vitamin E, 45 mg; vitamin K3, 2 mg; thiamine, 2 mg; riboflavin, 4 mg; nicotinic acid, 15 mg; D-pantothenic acid, 10 mg; niacin, 0.02 mg; pyridoxine, 3 mg; D-biotin, 0.2 mg; folic acid, 3 mg; vitamin B12, 0.02 mg; choline, 500 mg; Fe, 80 mg as ferrous sulfate and ferrous carbonate; Cu, 10 mg as copper sulfate; Mn, 40 mg as maganous oxide; Zn, 100 mg as zinc oxide; Co, 0.1 mg as cobalt carbonate; I, 0.6 mg as potassium iodide; and Se, 0.25 mg sodium selenite.

2Analyzed values except for soluble fiber content.3Calculated as the difference between total dietary fiber and insoluble fiber.4Calculated from INRA-AFZ (2004) tables.

Dietary fiber in sows during late pregnancy 5271

Table 1]. The transition from the conventional diet to the experimental diets was progressive. From d 92 to 98 of pregnancy (step 1), sows were fed a mixture that con-tained 33% of 1 of the experimental diets and 66% of the conventional gestation diet. From d 99 to 105 (step 2), the mixture contained 66% of 1 of the experimental diets and 33% of the conventional gestation diet. From d 106 of pregnancy until the farrowing (step 3), sows were fed 100% of 1 of the experimental diets. Sows were individ-ually fed. The daily feed allowance during the transition and the treatment periods was greater in HF than in LF sows to provide the same amount of NE (25.0, 25.3, and 25.8 MJ/d for steps 1, 2, and 3, respectively). The day after farrowing, sows were fed 2.5 kg of a conventional lactation diet, which provided 9.52 MJ of NE/kg, 17.5% CP, 0.9% Lys, and 4.3% crude fiber (as-fed basis). Feed allowance during lactation was increased by 1 kg/d until ad libitum feeding, which was reached approximately on d 4 or 5 of lactation. During pregnancy and lacta-tion, feed refusals were weighed daily and actual feed intakes calculated. Water was freely available to sows throughout the experimental period. Sow water intake was recorded daily from d 106 of pregnancy to d 21 of lactation using individual water meters. From insemina-tion until d 105 of pregnancy, sows received a daily meal at 0900 h. From d 105 of pregnancy to 21 of lactation, sows were fed twice a day at 0900 and 1430 h. During ad libitum feeding, feed troughs were filled twice a day so that feed was always available.

Fasted sows were weighed on d 91 of pregnancy and d 2 and 21 of lactation. On those same days, backfat thick-ness was measured ultrasonically (Vetko plus; Noveko, Boucherville, QC, Canada) at the level of the 10th rib on each side and 65 mm from the midline. Farrowings were attended. Piglets were weighed at birth, 24 h after birth of the first piglet (T24), and at 7 and 21 d of age. Piglets had free access to water throughout lactation but had no access to creep feed. They were weaned on d 21. Ambient temperature was maintained between 22 and 25°C. One heat lamp was available per farrowing crate.

Farrowing and Piglet Supervision during the First Day Postpartum

Parturition was induced on d 113 of pregnancy (d 0 being the day of first insemination) by an intramuscu-lar injection of 2 mL of alfaprostol (Alfabédyl; Céva Santé Animale, Libourne, France). Farrowing duration was estimated as the time between births of the first and the last piglets. Time elapsed between birth and the first suckling was recorded for each piglet. When this latency exceeded 45 min, the piglet was placed on the dam to suckle. Piglets weighing less than 600 g at birth were euthanized immediately after birth. During the first 24 h

postpartum, the original litter was kept with the sow. Be-yond 24 h, litters were standardized to 12 ± 1 piglets by cross-fostering within treatment.

Surgery and Sampling

Surgery. On d 105 of pregnancy, a catheter (2.16 mm o.d. and 1.02 mm i.d.; Silastic Dow Corning, Midland, MI) was inserted through a collateral vein in the right external jugular vein. The surgery was performed under general anesthesia induced by an intramuscular injection of 5 mg/kg BW of tiltamin and zolazepam mixture (Zolé-til 100; Virbac Santé Animale, Carros, France) and main-tained by 2 to 5% of isoflurane (Aerane; CSP, Cournon d’Auvergne, France) in oxygen (3 L/min). The catheter was tunneled under the skin, externalized on the dorsal surface of the neck, and put in a small tissue bag sutured in the skin. Sows were fasted from the evening before surgery and were fed again when they were moved to the farrowing crates. Catheters were flushed 3 times weekly with 10 mL of a saline solution (154 mM NaCl) contain-ing 50 IU/mL of heparin.

Blood Sampling. Blood samples (15 mL) were col-lected from the fasted sows before feeding (0830 h) on d 101, daily from d 111 of pregnancy to 3 of lactation, and on d 7 and 21 of lactation. Additional blood samples (15 mL) were collected daily at 1630 and 0030 h from d 112 of pregnancy until the day of farrowing. The day of sampling was calculated related to the day of farrowing a posteriori. Feed troughs were emptied at 2100 h the day before blood sampling. Blood samples were collect-ed into heparinized (40 IU of heparin/mL) tubes (10 mL) or in tubes containing no anticoagulant (5 mL). Samples collected for plasma assays (heparinized tubes) were kept on ice and centrifuged within 20 min for 10 min at 2,600 × g at 4°C. Samples for serum assays (tubes con-taining no anticoagulant) were left at room temperature for 4 h, stored overnight at 4°C, and then centrifuged for 10 min at 2,600 × g at 4°C. Serum and plasma samples were frozen at –20°C until they were assayed.

Postprandial Kinetics of Glucose and Insulin. Postprandial kinetics of plasma concentrations of glu-cose and insulin were evaluated on d 112 of pregnancy. Sows were deprived of feed from 2100 h the previous evening and were fed at 0900 h the following day. The LF sows received 0.62 kg of the LF diet and HF sows received 0.70 kg of the HF diet. The meal provided 6.1 MJ of NE to all sows. The duration of meal ingestion was recorded. Blood samples were collected at –15, –2, 15, 30, 45, 60, 75, 90, 120, 150, 180, and 240 min where time 0 corresponded to the meal delivery.

Colostrum and Milk Sampling. Colostrum was collected immediately after birth of the first piglet (T0) and T24. Milk was collected on d 7 and 21 of lactation.

Loisel et al.5272

At T24, colostrum was collected after an intramuscular injection of 20 IU of oxytocin (Ocytovem; Céva Santé Animale). On d 7 and 21, milk was collected after an intravenous injection of 10 IU of oxytocin. Milk and colostrum samples (45 mL) were manually collected from all functional teats. They were immediately filtered through gauze and stored at –20°C.

Biological Analyses

Hormonal Assays. Prolactin, progesterone, estradiol-17β, and cortisol were measured on the daily blood samples and insulin on the serial postprandial samples in duplicate within single assays. Prolactin was assayed in serum and progesterone, estradiol-17β, cortisol, and insulin were assayed in plasma. Plasma concentrations of progesterone, estradiol-17β, and insulin were measured by RIA using double antibody commercial kits validated in pigs [references IM1188 and A21854 for progesterone and estradiol-17β, respectively (Beckman Coulter, Rois-sy, France), and Insulin-CT for insulin (Cisbio Bioassays, Codolet, France)]. The intra- and interassay CV were, re-spectively, 6.3 and 8.9% for progesterone, 4.4 and 8.8% for estradiol-17β, and 6.4 and 14.4% for insulin. Average sensitivity was 0.1 ng/mL for progesterone, 6 pg/mL for estradiol-17β, and 3 µIU/mL for insulin. Serum prolactin concentrations were assayed using a homologous double antibody RIA according to Robert et al. (1989). The intra- and interassay CV were 4.8 and 3.5%, respectively, and sensitivity was 1.5 ng/mL. Plasma cortisol was measured using a fluorometric competitive enzyme immunoassay technique (AIA-1800 Automated Immunoassay Ana-lyzer and ST AIA-PACK CORT Methodology; TOSOH Bioscience, Tokyo, Japan).

Metabolites. Glucose, NEFA, lactate, urea, and creatinine concentrations in plasma were determined in blood samples collected before the morning meal on d 101 of pregnancy, from d 110 of pregnancy to 3 of lactation, and on d 7 and 21 of lactation in duplicate within single assays. Plasma glucose concentrations were also analyzed on the serial postprandial samples. Automated enzymatic methods using an automatic ana-lyzer (Konelab20i; Thermo, Cergy, France) were used to determine plasma concentrations of glucose (reference 61269; bio-Mérieux, Marcy l’Etoile, France), NEFA (NEFA C, reference 754664; Wako, Dardilly, France), lactate (lactate PAP, reference 61192; bio-Mérieux), urea (Urea Unimate 5, reference 07-3685-6; Roche, Neuilly-sur-Seine, France), and creatinine (Créatinine cinétique, reference 61162; bio-Mérieux).

Colostrum and Milk Composition. Dry matter, ash, GE, CP, lipids, and lactose were assayed in colostrum at T0 and T24 and in milk on d 7 and 21 of lactation. Dry matter and ash were measured according to the Asso-

ciation of Official Analytical Chemists (AOAC, 1990). Gross energy was measured using an adiabatic bomb calorimeter (C5000; IKA, Staufen, Germany). Nitro-gen was determined according to the method of Dumas (AOAC, 1990) based on sample pyrolysis and direct de-termination of N2 using an automatic device (Rapid N cube; Elementar, Hanau, Germany). Crude protein was estimated to be N × 6.38 (Gordon and Whittier, 1965). Total lipids were measured (LECO TFE-2000; Leco, St. Joseph, MI). Lactose was assayed using an enzymatic method (reference 01766303, Lactose/d-galactose test combination; R-Biopharm, Darmstad, Germany).

Immunoglobulin G and IgA were analyzed in colos-trum at T0 and T24 and IgA was measured in milk on d 7 and 21 of lactation. Both IgG and IgA were analyzed by ELISA using commercial quantification kits for por-cine IgG and IgA (references E100-104 and E100-102, respectively; Bethyl Laboratories, Montgomery, TX). Both immunoglobulins were assayed in triplicate. The intra- and interassay CV were, respectively, 2.9 and 7.0% for IgG and 4.0 and 5.8% for IgA.

Estimation of Colostrum Production

Individual colostrum intake by piglets was estimated from the piglet BW gain between birth and T24 accord-ing to an equation developed in the present experimental herd (Devillers et al., 2004): CI = –217.4 + (0.217 × ti) + (1,816,019 × BW24/ti) + BWB × (54.8 – 1,816,019/ti) × [(0.9985 – 3.7 × 10–4 × tFS) + (6.1 × 10–7 × tFS2)], in which CI is individual colostrum intake (g), BW24 is BW at T24 (kg), BWB is birth weight (kg), ti is the time elapsed between the first and the second weighing (min), and tFS is the interval between birth and the first suck-ling (min). Colostrum production during the 24 h after the onset of farrowing was calculated as the sum of in-takes by each piglet of the litter.

Statistical Analyses

Two sows, which had only 3 and 5 piglets, were excluded from the experiment. Data, except for mor-tality rate and percentage of piglets that did not suckle within 45 min of birth, were analyzed by ANOVA us-ing the MIXED procedure (SAS Inst. Inc., Cary, NC). The sow or the litter represented the experimental unit in the model, which included diet (LF or HF) as the main effect. Performance of piglets during the first day after birth was analyzed for all the piglets nursed dur-ing that day and then specifically for piglets born with a low BW (600 to 900 g, low birth weight piglets). Lit-ter size after cross-fostering was used as a covariate for piglet and litter performance from d 1 to 21 of lactation. Time-related variations in concentrations of hormones

Dietary fiber in sows during late pregnancy 5273

and metabolites in blood were analyzed using repeated measure analyses. The model included the effects of diet, sampling time, and their interaction. When sampling time was significant, differences between least squares means were assessed with Scheffe’s test. Colostrum or milk composition as well as IgG and IgA concentrations was analyzed within sampling day. Rates of stillbirth, mortality between birth and T24, mortality between d 1 and 21 of lactation (preweaning mortality), and the per-centage of piglets that did not suckle within 45 min of birth were analyzed using a logistic regression (GEN-MOD procedure), using a binomial error distribution with the diet as a fixed effect. The link function was a logit-transformation. Results of the logistic regression were then converted back to (original) natural units and expressed as means with confidence intervals. Values are expressed as least squares means in tables and as least squares means ± SEM in text and figures.

RESULTS

Sow PerformanceAt the onset of the experiment (d 91 of pregnancy),

sow BW and backfat thickness were similar for both treatments and averaged 206.2 ± 3.4 kg and 20.8 ± 1.1 mm, respectively (Table 2). From d 91 of pregnancy un-til parturition, sows had no feed refusals. The daily feed allowances and, thus, feed intakes of LF and HF sows

were, respectively, 2.60 and 2.72 kg from d 91 to 98 of pregnancy, 2.60 and 2.82 kg from d 99 to 105, and 2.60 and 2.93 kg from d 106 until the day of farrowing. Aver-age water intakes from d 106 of pregnancy until the day of farrowing were 16.4 ± 3.6 and 19.0 ± 3.6 L/d for LF and HF sows, respectively. During lactation, average feed and water intakes did not differ between treatments. Sow BW and backfat thickness on d 2 and 21 of lactation were not influenced by the treatment. The LF and HF sows lost 17.7 ± 2.7 and 15.7 ± 2.2 kg BW and 3.9 ± 0.6 and 3.6 ± 0.8 mm backfat thickness, respectively, during lactation.

Pregnancy length was similar for the 2 treatments (113.5 ± 0.2 d; Table 3). Farrowing duration ranged from 67 to 478 min and did not differ between treatments. Av-erage interval between births was 14.5 ± 2.2 and 12.0 ± 2.1 min for LF and HF sows, respectively. Birth interval was shorter (P < 0.05) between the first and the third born piglet for HF than for LF sows and tended to be shorter (P = 0.10) between the third and the fifth piglet (Fig. 1).

Litter Characteristics during the First Twenty-Four Hours and Colostrum Production

Litter characteristics at birth and at T24 did not dif-fer between LF and HF sows (Table 3). The average

Table 2. Body weight, backfat thickness, and water and feed intake for sows fed diets containing 13.3 [low fiber (LF)] or 23.4% [high fiber (HF)] total dietary fiber dur-ing late pregnancy

Item

Diet Root MSE1

P-valueLF HF

No. of sows 14 15BW, kg

d 91 of pregnancy 206.5 205.8 8.2 0.84d 2 of lactation 203.2 201.9 12.3 0.77d 21 of lactation 185.1 186.5 12.4 0.77

Backfat thickness, mmd 91 of pregnancy 20.9 20.6 1.7 0.70d 2 of lactation 22.8 22.7 3.3 0.94d 21 of lactation 19.0 19.0 3.2 0.98

Average water intake, L/dd 106 to farrowing 16.4 19.0 12.6 0.62wk 1 of lactation 17.7 20.5 6.4 0.30wk 2 of lactation 23.8 27.0 6.5 0.27wk 3 of lactation 27.3 29.3 8.1 0.55

Average feed intake, kg/dwk 1 of lactation 4.2 4.2 0.6 0.89wk 2 of lactation 5.5 5.7 1.1 0.76wk 3 of lactation 6.0 6.1 1.0 0.841MSE = mean square error.

Table 3. Piglet and litter characteristics, and colostrum production during the first 24 h postpartum for sows fed diets containing 13.3 [low fiber (LF)] or 23.4% [high fiber (HF)] total dietary fiber during late pregnancy

Item

Diet Root MSE1 P-valueLF HF

No. of sows 14 15Pregnancy length, d 113.4 113.6 0.6 0.62Farrowing duration, min 202.6 177.6 105 0.55Litter size at birth

Total born 14.8 14.9 2.5 0.92Born alive 14.2 14.5 2.3 0.82

Stillborn piglets,2 % 4.3 3.1 – 0.52Litter birth weight,3 kg 18.1 18.6 3.17 0.73Mean piglet BW,3 kg

At birth 1.28 1.27 1.55 0.92At T244 1.39 1.37 1.57 0.81

Mean piglet BW gain between T05 and T24, g 100 84 31 0.49Colostrum intake by individual piglets, g 308 282 46 0.48Litter weight gain between T0 and T24, kg 1.2 1.1 0.4 0.63Colostrum production between T0 and T24, kg 3.9 3.8 0.6 0.75Birth to suckling interval, min 29 25 5.0 0.29Mortality between birth and T24,2 % 8.5 6.9 – 0.54

1MSE = mean square error.2Data were analyzed using a logistic regression and root MSE was not

applicable.3Piglets born alive.4T24 = 24 h after the onset of parturition.5T0 = onset of parturition.

Loisel et al.5274

numbers of piglets per litter were 14.9 ± 0.8 (total born) and 14.4 ± 0.8 (born alive). The rate of stillborn piglets averaged 4.3 [2.3, 8.1] and 3.1% [1.5, 6.4] in litters from LF and HF sows, respectively. Mean BW of piglets born alive were 1.28 ± 0.67 at birth and 1.38 ± 0.68 kg 24 h after birth. The CV for piglet birth weight was 21.0 ± 2.1 and 21.4 ± 2.1% for LF and HF sows, respectively. The treatment did not influence birth to suckle interval, piglet BW gain between T0 and T24, average colostrum intake by piglets, or estimated colostrum yield. The proportion of piglets that did not suckle within 45 min of birth was lower (P = 0.01) in HF than in LF litters (12.0 [8.3; 17.0] vs. 23.7% [18.2; 30.3]). The mortality rate of piglets be-tween birth and T24 was 8.5 [5.3, 13.3] and 6.9% [4.2, 11.1] in litters from LF and HF sows, respectively.

Low Birth Weight Piglets. The number of low birth weight piglets born alive per litter was similar in both treatments (2.14 ± 0.56 and 2.07 ± 0.57 in LF and HF lit-ters, respectively) and represented 14.5 ± 3.7 and 14.0 ± 3.7% of born alive piglets per litter for LF and HF sows, respectively. The low birth weight piglets from HF sows consumed more colostrum than low birth weight piglets from LF sows (216 ± 22 vs. 137 ± 24 g; P = 0.02; Table 4). Their BW gain for 24 h averaged 60.9 ± 17.6 g whereas low birth weight piglets from LF sows lost weight (–1.3 ± 17.6 g; P = 0.02). Expressed as grams per 100 grams birth weight, colostrum intake was 76% greater in the HF treat-ment (29.4 ± 2.3 vs. 16.7 ± 3.4 g/100 g birth weight; P = 0.01). The birth to suckle interval of low birth weight pig-lets was not influenced by treatment.

Litter Characteristics during Lactation

After litter standardization on d 1 postpartum, litter size averaged 12.3 ± 0.3 piglets and was similar for both

treatments (Table 5). On d 21 of lactation, litter size did not differ between treatments (10.7 ± 0.4 and 11.2 ± 0.4, respectively, for LF and HF sows). Preweaning mortali-ty between d 1 and 21 was lower in litters from HF sows than in litters from LF sows (6.2 [3.4, 10.8] vs. 14.7% [10.2, 20.7]; P = 0.01). Mean piglet weights on d 7 and 21 of lactation did not differ between treatments and av-eraged 2.59 ± 0.09 kg on d 7 of lactation and 6.24 ± 0.13 kg on d 21 of lactation. Mean litter weight gain averaged 13.2 ± 0.7 kg between d 1 and 7 and 53.1 ± 2.0 kg between d 1 and 21 of lactation. These did not differ between the 2 treatments.

Colostrum and Milk Composition

Dry matter, ash, lactose, and GE in colostrum and in milk were not influenced by treatment (Table 6). Com-pared with LF sows, colostrum from HF sows contained 29% more lipid (P = 0.04) and tended to contain less protein and IgG (P = 0.10). Colostrum from HF sows also had less IgA at T0 (8.6 ± 1.1 vs. 11.9 ± 1.1 mg/mL; P = 0.02) and at T24 (2.5 ± 0.7 vs. 4.8 ± 0.7 mg/mL; P = 0.02) and milk tended to have less IgA on d 7 (P = 0.10) and 21 of lactation (P = 0.09).

Sow Endocrine and Metabolic Status

Hormones. Variations in circulating concentrations of progesterone and prolactin around farrowing are shown in Fig. 2. Plasma concentrations of estradiol-17β and cortisol are not shown. Hormone profiles fluctuated over time (P < 0.001) but were not influenced by treat-ment. For all hormones sampling time × diet interaction was not significant.

Postprandial Kinetics of Glucose and Insulin. All sows ingested the meal within 5 min after meal delivery. Profiles of plasma glucose concentrations did not differ between HF and LF sows (Fig. 3a). Plasma glucose con-

Figure 1. Kinetics of birth for the first 15 piglets born from sows fed diets containing 13.3 [low fiber (LF)] or 23.4% [high fiber (HF)] total dietary fiber during late pregnancy. Treatment: †P < 0.10 and *P < 0.05.

Table 4. Performance of low birth weight (< 900 g) pig-lets during the first 24 h postpartum for sows fed diets containing 13.3 [low fiber (LF)] or 23.4% [high fiber (HF)] total dietary fiber during late pregnancy

Item

Diet Root MSE1

P-valueLF HF

No. 30 31Mean piglet BW,2 g

At birth 790 736 104 0.18At T243 788 796 137 0.84

Colostrum intake by individual piglets, g 137 216 112 0.02Birth to suckling interval, min 40 33 18 0.16

1MSE = mean square error.2Piglets born alive.3T24 = 24 h after the onset of parturition.

Dietary fiber in sows during late pregnancy 5275

centrations reached a peak 45 min after the meal for LF sows (5.90 ± 0.13 mM) and 60 min after the meal for HF sows (5.99 ± 0.14 mM). Profiles of insulin concentrations differed between treatments (P < 0.01; Fig. 3b). Plasma insulin concentrations reached a peak at 60 min in both treatments and then decreased. Plasma insulin concentra-tions were lower in LF than in HF sows 60 min (60.2 ± 5.2 vs. 79.7 ± 5.5 µIU/mL; P < 0.05) and 75 min after the meal delivery (38.1 ± 5.2 vs. 60.4 ± 5.5 µIU/mL; P < 0.05) and tended to be lower at 90 min (P < 0.10).

Metabolites. From d 111 of pregnancy to 21 of lacta-tion, preprandial concentrations of glucose, NEFA, lac-tate, and creatinine fluctuated over time (P < 0.001) with-out treatment effect or sampling time × diet interaction (data not shown). Plasma concentrations of urea were in-fluenced by treatment. They were greater in LF than in HF sows at d –1 (2.60 ± 0.22 vs. 1.73 ± 0.22 mM; P = 0.01; Fig. 4) and tended to be greater at d –13, –2, 0, and 1 of lactation (P < 0.10). Plasma concentrations of urea fluctu-ated over time (P < 0.001); they progressively increased during lactation (between d 2 and 21; P < 0.05).

DISCUSSION

In sows, the initiation of structural and metabolic differentiation of the mammary epithelium as well as the initiation of milk-specific components synthesis is observed around d 105 (Bazer et al., 2001). The treat-ment in the present experiment was, therefore, applied at d 106 of pregnancy. Nevertheless, an adaptation period,

during which sows were fed increasing amounts of ex-perimental diets, started on d 92 of pregnancy for sows to gradually adapt to the high fiber intake. The level of crude fiber incorporation was based on common prac-tice and literature. The HF diet contained 7.9% crude fi-ber, which is greater than the dietary fiber inclusion rate commonly used in gestation diets (around 5% in France). To prevent feed refusals, this incorporation rate was less than the rate used in most experimental studies. More-over, 7% crude fiber in late gestation diets was shown to be an adequate level to increase piglet BW gain during early lactation (Oliviero et al., 2009).

Colostrum Yield and Sow Endocrine and Metabolic Status

A high dietary fiber supply during late pregnancy did not influence estimated colostrum yield, which was with-in the range previously estimated for primiparous sows in the present herd (Foisnet et al., 2011). Colostrum pro-duction is under hormonal control and, notably, might be regulated by peripartum variations in progesterone and prolactin concentrations in swine (Quesnel et al., 2012). In the present study, prolactin concentrations were not affected by the treatment, contradicting with previous findings (Farmer et al., 1998; Quesnel et al., 2009). The lack of effect of dietary fiber on prolactin concentrations could be related to the shorter treatment duration in the present experiment or to the level or source of fiber used. The HF diet contained a mixture of soybean hulls, wheat bran, sunflower meal (undecorticated), and sugar beet pulp and thus provided a large proportion of insoluble fiber (20.6%) and moderate proportion of soluble (2.8%) fiber. Farmer et al. (1995) provided wheat bran and corn cobs, which are sources of insoluble fiber, and Quesnel et al. (2009) provided nearly 20% of sugar beet pulp (solu-ble fiber). Different types of fiber have different physico-chemical properties, which determine their physiological effects (Bach Knudsen, 2001).

The current HF diet did not influence plasma progester-one concentrations either, which is in agreement with find-ings by Farmer et al. (1995). However, in cows, high fiber diets were shown to reduce progesterone half-life (Lemley et al., 2010). This effect was likely due to the decrease in circulating concentrations of insulin induced by the high fiber diet. Indeed, in ruminants, elevated concentrations of insulin decrease hepatic cytochrome P4503A and P4502C activities, which are responsible for hepatic clearance of progesterone (Lemley et al., 2009) and thereby increasing plasma progesterone concentrations (Selvaraju et al., 2002). In the present study, high dietary fiber supply to sows dur-ing late pregnancy increased postprandial concentrations of insulin without significantly influencing glucose concentra-tions. These results are not consistent with previous experi-ments, which reported that sows fed a high-fiber diet had

Table 5. Piglets and litter performance during lactation for sows fed diets containing 13.3 [low fiber (LF)] or 23.4% [high fiber (HF)] total dietary fiber during late pregnancy

Item

Diet Root MSE1

P-valueLF HF

No of sows 14 15Litter size

At d 12 12.6 12.0 1.0 0.16At d 213 10.7 11.2 0.7 0.38

Mean piglet BW, kgAt d 73 2.61 2.56 0.29 0.69At d 213 6.32 6.15 0.48 0.41

Litter BW, kgAt d 73 29.3 29.8 3.9 0.76At d 213 68.6 69.7 6.0 0.76

Litter BW gain, kgBetween d 1 and 73 13.2 13.3 2.0 0.86Between d 1 and 213 52.4 53.8 4.6 0.64

Preweaning mortality,4 % 14.7 6.2 – 0.011MSE = mean square error.2After cross-fostering.3Litter size on d 1 was introduced as a covariate.4Piglet mortality rate from d 1 to 21 of lactation was analyzed using a

logistic regression and root MSE was not applicable.

Loisel et al.5276

smaller and delayed postprandial peaks of insulin and glu-cose (Rushen et al., 1999; Ramonet et al., 2000; Quesnel et al., 2009). On the other hand, Oliviero et al. (2009) did not show any influence of dietary fiber on insulin and glucose concentrations 1 h after the meal when sows were fed 7.0 or 3.8% crude fiber for the last 3 wk of pregnancy. Again, discrepancies may be the result of the dietary fiber sources used in the different studies.

The HF diet during late pregnancy had no influence on most markers of sow metabolic status, except for urea, which decreased preprandially. Urea constitutes the larg-est and the most readily available source of nitrogen for bacteria (Levrat et al., 1993). In pigs, both soluble and in-soluble dietary fibers are fermented and soluble fiber has a greater capacity to increase microbial mass and activ-ity at the level of the gut in comparison with insoluble fi-ber (Bach Knudsen, 2001). In rats fed fermentable fiber, such an increase in microbial mass was related to a greater transfer of urea from the blood to the gut and to reduced plasma urea concentrations (Younes et al., 1995). Whether this could explain the decrease in urea in HF sows remains unknown. Alternatively, lower plasma urea concentra-tions may be indicative of a lower protein oxidation in the liver that may be related to a lower intestinal absorption of protein when sows were fed the HF diet. These results

Table 6. Colostrum and milk composition of sows fed diets containing 13.3 [low fiber (LF)] or 23.4% [high fiber (HF)] total dietary fiber during late pregnancy

Item

Diet Root MSE1

P-valueLF HF

No. of sows 14 15DM,2 %

T03 25.7 26.7 1.0 0.43T244 21.6 22.6 1.0 0.48d 7 20.8 20.0 0.8 0.31d 21 24.2 22.1 2.4 0.24

Ash,2 %T0 0.69 0.73 0.02 0.20T24 0.74 0.73 0.02 0.43d 7 0.81 0.79 0.02 0.40d 21 0.92 0.91 0.02 0.80

Proteins,2 %T0 16.4 16.4 0.9 0.99T24 8.2 6.7 0.6 0.10d 7 5.2 5.1 1.4 0.67d 21 6.1 5.7 0.5 0.38

Lipids,2 %T0 5.4 6.0 0.5 0.23T24 8.3 10.7 0.8 0.04d 7 9.5 8.7 0.8 0.30d 21 12.5 10.5 2.1 0.24

Lactose,2 %T0 2.5 2.5 0.2 0.97T24 3.4 3.5 0.2 0.86d 7 4.5 4.6 0.4 0.70d 21 3.8 4.2 0.4 0.21

GE, kJ/gT0 6.7 7.0 0.3 0.33T24 6.1 6.7 0.4 0.24d 7 6.0 5.7 0.3 0.33d 21 7.3 6.4 0.9 0.23

IgG, mg/mLT0 51.9 52.4 6.2 0.95T24 10.4 6.5 2.3 0.10

IgA, mg/mLT0 11.9 8.6 1.1 0.02T24 4.8 2.5 0.7 0.02d 7 2.2 1.8 0.2 0.10d 21 4.1 3.1 0.5 0.091MSE = mean square error.2Grams per 100 grams of whole colostrum or milk.3T0 = onset of parturition.4T24 = 24 h after the onset of parturition.

Figure 2. Concentrations of plasma progesterone (a) and serum prolac-tin (b) for sows fed diets containing 13.3 [low fiber (LF)] or 23.4% [high fiber (HF)] total dietary fiber during late pregnancy.

Dietary fiber in sows during late pregnancy 5277

indicate that varying the amount of dietary fiber in the diet during late pregnancy mildly affects sow endocrine and metabolic status during the peripartum period, especially by influencing plasma insulin concentrations.

Consistent with the lack of impact of HF diet on co-lostrum yield, mean colostrum intake by piglets was not influenced by the treatment. Special focus was given to low birth weight piglets because they usually consume less colostrum than heavier littermates and are more at risk of dying within a few days (Milligan et al., 2002). The thresh-old of 900 g BW to describe a low birth weight piglet was chosen because it represented around 70% of the mean piglet birth weight in the present study and, according to Le Dividich (1999), these piglets have a substantial risk of mortality. Relative to their birth weight, these piglets con-sumed 76% more colostrum in HF than in LF treatment.

Colostrum intake by piglets depends on the capacity of the sow to produce colostrum but also on various factors in-cluding piglet BW and vitality at birth, litter characteristics, and maternal behavior (Quesnel et al., 2012). In the present study, litter size, weight, and heterogeneity were similar in both treatments. The time elapsed between birth and the first suckling, which is an indicator of piglet vitality at birth (Herpin et al., 1996), was numerically shorter for low birth weight piglets from HF than from LF sows. This indicates that a greater ability to reach the mammary gland could be responsible for this increased colostrum intake. Piglet vitality at birth is known to be influenced by farrowing duration (Herpin et al., 1996). The duration of the whole farrowing process was not affected by the HF diet but the beginning of parturition was quicker in HF than in LF sows. Sow behavior during the peripartum period was not studied in the present experiment. However, the literature shows that sows fed bulky diets during 2 consecutive pregnancies spend more time lying on their side and less time standing during the peripartum period (Farmer et al., 1995). Thus, the time during which the mammary glands are accessible to the piglets could be greater for piglets from HF sows. Nevertheless, such an effect of high-fiber diets on maternal behavior was not observed by Guillemet et al. (2007). In the present study, therefore, the greater BW gain of small piglets in HF litters could be related to a treatment effect on the ability of piglets to suckle rather than to a greater inher-ent ability of sows to synthesize colostrum.

Litter Characteristics during Lactation

Feeding sows a HF diet during late pregnancy did not increase piglet BW gain during the first week of lacta-tion. In the literature, the impact of high dietary fiber dur-ing part of or late pregnancy on piglet weight gain varied

Figure 3. Postprandial kinetics of plasma concentrations of glucose (a) and insulin (b) on d 112 of pregnancy in sows fed diets containing 13.3 [low fiber (LF)] or 23.4% [high fiber (HF)] total dietary fiber during late pregnancy. Treatment: †P < 0.10, *P < 0.05, and **P < 0.01.

Figure 4. Plasma concentrations of urea during late pregnancy and lactation in sows fed diets containing 13.3 [low fiber (LF)] or 23.4% [high fiber (HF)] total dietary fiber during late pregnancy. Treatment: †P < 0.10 and **P < 0.01.

Loisel et al.5278

among studies, observing either positive (Guillemet et al., 2007; Oliviero et al., 2009) or no effects (Matte et al., 1994; Holt et al., 2006). Discrepancies are likely related to differences in the experimental design (i.e., treatment du-ration, source of dietary fiber, etc.) Although the HF diet did not influence piglet BW gain during the first day after birth and throughout lactation in the current trial, prewean-ing mortality was lower in litters from HF sows. Mortality rate in LF sows was similar to the average mortality ob-served in French commercial herds in 2011 (13.8 ± 3.9%; IFIP, 2011). The most important causes of piglet mortal-ity during lactation are low colostrum intake and crush-ing by the sow (Dyck and Swierstra, 1987). The increase in colostrum intake of low birth weight piglets from HF sows could partly explain the decrease in piglet mortality. Indeed, mortality of low birth weight piglets represented 18 and 31% of mortality during lactation in litters from HF and LF sows, respectively. The risk of crushing increases when the sow performs many postural changes (Weary et al., 1998) and sows fed high fiber diets were shown to exhibit less frequent postural changes during pregnancy (de Leeuw et al., 2005). To our knowledge, however, the influence of dietary fiber on the frequency of sow postural changes during lactation has not been studied.

The effect of dietary fiber on piglet survival could be also related to colostrum and milk composition. Indeed colostrum from HF sows at T24 contained more lipid than colostrum from LF sows. Lipids in colostrum play an im-portant role in piglet thermoregulation, which influences piglet survival (Le Dividich et al., 1997). The origin of the greater lipid content in colostrum from HF sows is not clear. Lipids in colostrum arise from dietary lipids but also from sow adipose tissue and de novo fatty acid syn-thesis in the mammary gland (Patton and Keenan, 1975). The LF and HF sows ingested daily 89 and 94 g of dietary lipids, respectively, during the experimental period (start-ing at d 106 of pregnancy). The 5% greater daily intake of dietary lipids by HF sows does not seem to be large enough to induce an increase of 29% in lipid content of colostrum. Dietary fiber increases the plasma concentra-tion of short chain fatty acids (Serena et al., 2009), which can be used by the mammary gland as precursors for de novo synthesis of lipids (Theil et al., 2012). Whether the increase in colostral lipid content could be partly due to a greater catabolism of lipids from adipose tissue or to a greater synthesis in the mammary gland, or both, cannot be inferred from the present findings.

The HF diet also affected immunoglobulin content in colostrum. More specifically, the HF diet tended to de-crease colostral IgG at T24, which, together with the re-duced IgA, may partly explain the trend for lower protein concentrations in colostrum from HF sows at T24. Di-etary fiber decreased IgA concentrations in colostrum by one-third at T0 and by half at T24 and tended to decrease

IgA concentrations in milk as well. The explanation for this effect is unclear. In sows, there is strong evidence of an immunological link between the gut and the mammary gland. Indeed, specific IgA antibodies are present in milk after stimulation by the corresponding antigens at the level of the gut (Evans et al., 1980). Dietary fiber influ-ences gut microflora (Varel and Yen, 1997) and leads to a greater protection of the intestine from pathogenic micro-organisms because of the increased secretion of intestinal mucins (Montagne et al., 2003). Furthermore, IgA are transferred from the mammary gland to the colostrum by the specific polymeric immunoglobulin receptor and the expression of this receptor in the mammary gland is influ-enced by numerous factors, such as hormones, cytokines, and metabolites (Lejan, 1993). Thus, the decrease in IgA concentrations could be related to a change in gut micro-flora of sows or to a change in polymeric immunoglobulin receptor expression in the mammary gland. The decrease in IgA concentration in colostrum or milk had no conse-quences on piglet growth and survival. Furthermore, the incidence of diarrhea during the experiment was very low, with only 2 litters (1 from each treatment) suffering from diarrhea. Therefore, the decrease in IgA concentration had no consequence on the incidence of piglet diarrhea before weaning in the present experimental conditions, yet possible consequences of such a treatment on piglet health requires further investigation at a larger scale.

In conclusion, progressively increasing the amount of dietary fiber up to 23.4% TDF in sow gestation diet from d 92 of pregnancy onward, as compared with pro-viding 13.3% TDF, affected sow colostrum composi-tion but not colostrum yield, increased colostrum intake of low birth weight piglets, and decreased preweaning mortality. These effects were not related to peripartum profiles of the main hormones involved in lactogenesis, yet the potential role of maternal behavior and piglet vi-tality at birth warrants further investigation.

LITERATURE CITEDAssociation of Official Analytical Chemists (AOAC). 1990. Official

methods of analysis. 15th ed. AOAC, Washington, DC.Bach Knudsen, K. E. 2001. The nutritional significance of “dietary

fibre” analysis. Anim. Feed Sci. Technol. 90:3–20. Bazer, F. W., J. J. Ford, and R. S. Kensinger. 2001. Reproductive physi-

ology. In: W. G. Pond and H. J. Mersmann, editors, Biology of the domestic pig. Comstock Publishing Assoc., Ithaca, NY. p. 150–224.

Council of European Union. 2001. Directive 2001/88/EC. Minimum standards for the protection of pigs. Official J. European Union, Brussels, Belgium.

de Leeuw, J. A., J. J. Zonderland, H. Altena, H. A. M. Spoolder, A. W. Jongbloed, and M. W. A. Verstegen. 2005. Effects of levels and sources of dietary fermentable non-starch polysaccharides on blood glucose stability and behaviour of group-housed pregnant gilts. Appl. Anim. Behav. Sci. 94:15–29.

Devillers, N., C. Farmer, J. Le Dividich, and A. Prunier. 2007. Vari-ability of colostrum yield and colostrum intake in pigs. Animal 1:1033–1041.

Dietary fiber in sows during late pregnancy 5279

Devillers, N., J. Le Dividich, and A. Prunier. 2011. Influence of colos-trum intake on piglet survival and immunity. Animal 5:1605–1612.

Devillers, N., J. Van Milgen, A. Prunier, and J. Le Dividich. 2004. Estima-tion of colostrum intake in the neonatal pig. Anim. Sci. 78:305–313.

Dyck, G. W., and E. E. Swierstra. 1987. Causes of piglet death from birth to weaning. Can. J. Anim. Sci. 67:543–547.

Evans, P. A., T. J. Newby, C. R. Stokes, D. Patel, and F. J. Bourne. 1980. Antibody-response of the lactating sow to oral immuniza-tion with Escherichia-coli. Scand. J. Immunol. 11:419–429.

Farmer, C., S. Robert, J. J. Matte, C. L. Girard, and G. P. Martineau. 1995. Endocrine and peripartum behavioral responses of sows fed high-fiber diets during gestation. Can. J. Anim. Sci. 75:531–536.

Farmer, C., S. Robert, and J. Rushen. 1998. Bromocriptine given orally to periparturient or lactating sows inhibits milk produc-tion. J. Anim. Sci. 76:750–757.

Foisnet, A., C. Farmer, C. David, and H. Quesnel. 2011. Farrowing induction induces transient alterations in prolactin concentra-tions and colostrum composition in primiparous sows. J. Anim. Sci. 89:3048–3059.

Gordon, G. O., and E. O. Whittier. 1965. Proteins of milk. In: B. H. Webb and A. H. Johnson, editors, Fundamentals of dairy chem-istry. Avi Publ. Co., Westport, CT. p. 54–90.

Guillemet, R., A. Hamard, H. Quesnel, M. C. Père, M. Etienne, J. Y Dourmad, and M. C. Meunier-Salaün. 2007. Dietary fibre for gestating sows: Effects on parturition progress, behaviour, litter and sow performance. Animal 1:872–880.

Herpin, P., J. Le Dividich, J. C. Hulin, M. Fillaut, F. DeMarco, and R. Bertin. 1996. Effect of the level of asphyxia during delivery on viability at birth and early postnatal vitality of newborn pigs. J. Anim. Sci. 74:2067–2075.

Holt, J. P., L. J. Johnston, S. K. Baidoo, and G. C. Shurson. 2006. Effects of a high-fiber diet and frequent feeding on behavior, reproductive performance, and nutrient digestibility in gestating sows. J. Anim. Sci. 84:946–955.

IFIP(Institut de la FIlière Porcine). 2011. Average national GTTT re-sults from 1970 to 2011. www.ifip.asso.fr/PagesStatics/resultat/pdf/retro/00gttt.pdf. (Accessed 1 March 2013).

INRA-Association Française de Zootechnie (AFZ). 2004. Tables of composition and nutritional value of feed materials. Pigs, poultry, cattle, sheep, goats, rabbits, horses, fish. 2nd ed. INRA-AFZ, Paris, France.

Le Dividich, J. 1999. A review – Neonatal and weaner pig: Manage-ment to reduce variation. In: P. D. Cranwell, editor, Manipulat-ing pig production VII. Australasian Pig Science Association, Werribee, VIC, Australia. p. 135–155.

Le Dividich, J., P. Herpin, E. Paul, and F. Strullu. 1997. Effect of fat content of colostrum on voluntary colostrum intake and fat utilization in newborn pigs. J. Anim. Sci. 75:707–713.

Lejan, C. 1993. Secretory component and IgA expression by epithe-lial-cells in sow mammary gland and mammary secretions. Res. Vet. Sci. 55:265–270.

Lemley, C. O., J. M. Koch, K. P. Blemings, and M. E. Wilson. 2009. Alterations in progesterone catabolic enzymes, CYP2C and CYP3A, in hepatocytes challenged with insulin and glucagon. J. Anim. Vet. Adv. 8:39–46.

Lemley, C. O., K. A. Vonnahme, L. R. Tager, K. M. Krause, and M. E. Wilson. 2010. Diet-induced alterations in hepatic progesterone (P4) catabolic enzyme activity and P4 clearance rate in lactating dairy cows. J. Endocrinol. 205:233–241.

Levrat, M. A., C. Remesy, and C. Demigne. 1993. Influence of inulin on urea and ammonia nitrogen fluxes in the rat cecum: Con-sequences on nitrogen-excretion. J. Nutr. Biochem. 4:351–356.

Matte, J. J., S. Robert, C. L. Girard, C. Farmer, and G. P. Martineau. 1994. Effect of bulky diets based on wheat bran or oat hulls on reproductive performance of sows during their first 2 parities. J. Anim. Sci. 72:1754–1760.

Milligan, B. N., D. Fraser, and D. L. Kramer. 2002. Within-litter birth weight variation in the domestic pig and its relation to pre-weaning survival, weight gain, and variation in weaning weights. Livest. Prod. Sci. 76:181–191.

Montagne, L., J. R. Pluske, and D. J. Hampson. 2003. A review of interactions between dietary fibre and the intestinal mucosa, and their consequences on digestive health in young non-ruminant animals. Anim. Feed Sci. Technol. 108:95–117.

Oliviero, C., T. Kokkonen, M. Heinonen, S. Sankari, and O. Peltonie-mi. 2009. Feeding sows with high fibre diet around farrowing and early lactation: Impact on intestinal activity, energy balance relat-ed parameters and litter performance. Res. Vet. Sci. 86:314–319.

Patton, S., and T. W. Keenan. 1975. The milk-fat globule membrane. Biochim. Biophys. Acta 415:273–309.

Prosky, L., N. Asp, I. Furda, J. DeVries, T. Schweizer, and B. Harland. 1985. Determination of total dietary fiber in foods and food prod-ucts: Collaborative study. J. Assoc. Off. Anal. Chem. 68:677–679.

Quesnel, H., C. Farmer, and N. Devillers. 2012. Colostrum intake: In-fluence on piglet performance and factors of variation. Livest. Sci. 146:105–114.

Quesnel, H., M. C. Meunier-Salaün, A. Hamard, R. Guillemet, M. Etienne, C. Farmer, J. Y. Dourmad, and M. C. Père. 2009. Di-etary fiber for pregnant sows: Influence on sow physiology and performance during lactation. J. Anim. Sci. 87:532–543.

Ramonet, Y., S. Robert, A. Aumaître, J. Y. Dourmad, and M. C. Meunier-Salaün. 2000. Influence of the nature of dietary fibre on digestive utilization, some metabolite and hormone profiles and the behaviour of pregnant sows. Anim. Sci. 70:275–286.

Robert, S., A. M. B. de Passille, N. St-Pierre, P. Dubreuil, G. Pel-letier, D. Petitclerc, and P. Brazeau. 1989. Effect of the stress of injections on the serum concentration of cortisol, prolactin, and growth-hormone in gilts and lactating sows. Can. J. Anim. Sci. 69:663–672.

Rushen, J., S. Robert, and C. Farmer. 1999. Effects of an oat-based high-fibre diet on insulin, glucose, cortisol and free fatty acid concentrations in gilts. Anim. Sci. 69:395–401.

Selvaraju, S., S. K. Agarwal, S. D. Karche, S. K. Srivastava, A. C. Ma-jumdar, and U. Shanker. 2002. Fertility responses and hormonal pro-files in repeat breeding cows treated with insulin. Anim. Reprod. Sci. 73:141–149.

Serena, A., H. Jorgensen, and K. E. Bach Knudsen. 2009. Absorp-tion of carbohydrate-derived nutrients in sows as influenced by types and contents of dietary fiber. J. Anim. Sci. 87:136–147.

Taverne, M., M. Bevers, and J. M. C. Bradshaw. 1982. Plasma con-centrations of prolactin, progesterone, relaxin and oestradiol-17β in sows treated with progesterone, bromocriptine or indo-methacin during late pregnancy. J. Reprod. Fertil. 65:85–96.

Theil, P. K., M. O. Nielsen, M. T. Sørensen, and C. Lauridsen. 2012. Lactation, milk and suckling. In: K. E. Bach Knudsen, N. J. Kjeldsen, H. D. Poulsen, and B. B. Jensen, editors, Nutritional physiology of pigs. Danish Pig Research Centre, Copenhagen, Denmark. p. 1–47.

Varel, V. H., and J. T. Yen. 1997. Microbial perspective on fiber utili-zation by swine. J. Anim. Sci. 75:2715–2722.

Weary, D. M., P. A. Phillips, E. A. Pajor, D. Fraser, and B. K. Thomp-son. 1998. Crushing of piglets by sows: Effects of litter fea-tures, pen features and sow behaviour. Appl. Anim. Behav. Sci. 61:103–111.

Younes, H., K. Garleb, S. Behr, C. Remesy, and C. Demigne. 1995. Fermentable fibers or oligosaccharides reduce urinary nitrogen-excretion by increasing urea disposal in the rat cecum. J. Nutr. 125:1010–1016.