Effect of nutrient intake in lactation on sow performance: Determining the threonine requirement of the high-producing lactating sow 1,2

Transcription

1 Effect of nutrient intake in lactation on sow performance: Determining the threonine requirement of the high-producing lactating sow 1,2 D. R. Cooper*, J. F. Patience* 3, R. T. Zijlstra*, and M. Rademacher *Prairie Swine Centre Inc., Saskatoon, SK, Canada, S7H 5N9; Department of Animal and Poultry Science, University of Saskatchewan, Saskatoon, SK, Canada S7N 5B5; and Degussa Hüls AG, Hanau, Germany D ABSTRACT: Reproductive performance is steadily increasing within the pork industry; logically, amino acid requirements need to be redefined for sows producing larger litters. The objective of this study was to determine the threonine requirement of the high-producing lactating sow and to determine the effect of lysine on this requirement. A total of 419 PIC C-15 sows were assigned randomly to treatment within parity groups (1, 2, and 3+) and gestation treatment at d 110 of gestation. Lactation diets were formulated to contain 0.80% total lysine (tlys) with 0.30, 0.35, 0.40, 0.45, 0.50, 0.55, 0.60, or 0.65% total threonine (tthr) or 1.06% tlys with 0.40, 0.45, 0.50, 0.55, 0.60, 0.65, or 0.70% tthr. Litters were standardized to a minimum of 11 piglets within 48 h after farrowing, and sows had free access to feed throughout lactation (lactation length = 20.1 ± 0.1 d). Sow ADFI exceeded expectation, averaging 6.90, 7.40, and 7.20 kg/d for Parities 1, 2, and 3+, respectively. Daily tlys intake was 58 g/d (47 g of apparent ileal digestible lysine [dlys] per day) and 74 g/d (59 g dlys/d) for the low- and high-lysine group, respectively. Lysine intake did not affect sow or litter performance (P > 0.10). Sows gained an average of 4.8 kg in lactation. Using regression analysis, BW gain was maximized at 0.54% tthr for all parity groups (quadratic; P < 0.05). Litter weaning weight (67.1, 67.9, and 66.2 kg for Parities 1, 2, and 3+, respectively) and litter weight gain (2.49, 2.53, and 2.44 kg/d for Parities 1, 2, and 3+, respectively) were maximized at 0.53% tthr using regression analysis, for all parity groups (quadratic; P < 0.05). Based on regression analysis, plasma urea nitrogen on d 10 and 18 was minimized at 0.54% tthr (P < 0.05). Lysine levels in excess of 58 g of tlys/d did not benefit sow or litter performance. The requirement for threonine to minimize sow tissue mobilization was 37, 40, and 38 g tthr/d (28, 30, and 30 g of apparent ileal digestible threonine [dthr] per day) for Parities 1, 2, and 3+ sows, respectively. The threonine required to maximize litter performance was 36, 39, and 38 g of tthr/d (28, 30, and 29 g of dthr/ d) for Parities 1, 2, and 3+ sows, respectively. Alternatively, the requirement can be expressed as 14.3 g tthr (11.8 g dthr) per kilogram of litter gain. Key Words: Lactation, Lysine, Sows, Threonine 2001 American Society of Animal Science. All rights reserved. J. Anim. Sci : Introduction The productivity of the breeding herd contributes heavily to the overall success of a pork production system. Milk is generally the sole source of nutrients avail- 1 The authors gratefully acknowledge Degussa-Hüls AG for funding of the project, donation of synthetic amino acids, and amino acid assays. The authors also gratefully acknowledge the strategic program funding provided by the pork producers of Saskatchewan, Alberta, and Manitoba. We further acknowledge with gratitude the assistance with statistical analysis provided by Brian Thompson. 2 Presented in part at the ASAS Midwestern Section Mtg., Des Moines, IA, March 13 15, 2000, J. Anim. Sci. 78(Suppl. 1):30 (Abstr.). 3 Correspondence: P.O. Box 21057, th St E (phone: ; fax: ; patience@sask.usask.ca). Received October 19, Accepted May 4, able to the suckling pig; thus, sow lactation diets need to be formulated to maximize milk production and thereby litter weight at weaning, as well as to allow the sow to maintain body condition throughout lactation, supporting rapid and successful rebreeding. Failure to meet the AA requirements of the sow in lactation can lead to decreased milk production and litter weaning weights (King et al., 1993) and to increased reproductive failure (King and Williams, 1984). Threonine is often the second most limiting AA in Western Canadian growing pig diets (Fuller et al., 1989). There is limited research on the threonine requirement of the lactating sow. The threonine requirement of multiparous lactating sows has been estimated to be 31 g of total threonine/d to minimize sow body weight losses (Westermeier et al., 1998) and 28 g of total threonine/d to maximize milk production (Paulicks et 2378

2 Threonine requirement of the lactating sow 2379 al., 1998). In these studies, sows nursed a maximum of 10 piglets, and the ADG of piglets was 188 g/d, lower than that often achieved in commercial practice. According to the model proposed by Pettigrew (1993), the sow requires 41 mg of threonine/kg BW 0.75 for maintenance. The ratio of threonine to lysine in milk protein (0.58) was then used to determine the threonine required for milk production. The Pettigrew (1993) model was based on high producing sows and thus more accurately reflects the industry s expectation of the modern sow. The objective of the present study was to determine the threonine requirement of the high-producing lactating sow. The requirement was determined across parities based on milk production and sow body composition. Two levels of lysine were used to ensure that the response to threonine was not compromised by inadequate lysine intake. Experimental Design Materials and Methods In total, 419 sows (Camborough 15, Pig Improvement Canada, Acme, AB), including 99 first parity, 106 second parity, and 214 third-to-ninth parity sows were bred to terminal line boars (Canabrid, Pig Improvement Canada). A total of 450 sows were used to achieve the total number of sows needed for this experiment. Parity 1 sows were required to weigh at least 120 kg at breeding to be included in the experiment. Sows were assigned randomly within parity at mating to one of two gestation diets that contained either 0.44% or 0.55% total lysine and 3,100 kcal of DE/kg. A description of the diets, sow management procedures, and sow and litter performance in gestation and at parturition are reported in the companion article (Cooper et al., 2001). The experiment was arranged in a 2 2 (7 or 8) factorial with two levels of lysine in gestation, two levels of lysine in lactation and either eight (low lysine) or seven (high lysine) levels of threonine in lactation (0.30 to 0.70% total threonine). Two additional treatments, consisting of conventional wheat-barley-soybean meal formulations, were added as positive controls. The experiment was conducted across three replicates of 139 sows each, with 2 extra sows. Lactation diets (Tables 1 and 2) were assigned randomly within parity and gestation diet at d 110 of gestation. The lactation diets were formulated to be in excess (at least 110% of the ideal AA ratio relative to lysine) of all total indispensable AA except threonine and valine (ARC, 1981; NRC, 1988). L-Valine was added to the diets to obtain a valine-to-lysine ratio of 1.15% to ensure that valine was not limiting sow performance (Richert et al., 1997). Diets were formulated to contain 0.80% or 1.06% total lysine: one equal to and one above the NRC (1988) recommendations to determine the effect of lysine intake on the threonine requirement. Within each series, the lowest level of threonine was set at 37.5% of lysine. L-threonine was then added to the basal diets in increments of 0.05% to maximum total threonine levels of 0.65 and 0.70% at 0.80 and 1.06% total lysine, respectively. Minerals and vitamins were added to be in excess of NRC (1988) requirement estimates. One positive control diet at each lysine level, containing barley, wheat, and soybean meal as well as synthetic L-valine, was fed to determine whether performance on the semisynthetic diets was equivalent to that of sows fed diets with minimal synthetic AA. Lactation diets were analyzed for total lysine and threonine levels (Table 2). For diets high in threonine (0.50 to 0.70%) the analyzed threonine concentration was close to the formulated threonine concentrations. For diets low in threonine, unexplained discrepancies between assayed and formulated levels were observed and therefore considered very carefully. The agreement between formulated and analyzed threonine concentrations in most of the experimental diets indicated that the threonine basal diet had to contain the correct concentration of threonine; if they had contained excessive threonine, then all of the diets would have been above expected in threonine content and this was not the case. In addition, based on feed mixing quality assurance standard operating procedures, daily L-threonine inventory sheets always agreed with formulation usage, confirming that excess L-threonine was not inadvertently used in some diets. Therefore, formulated lysine and threonine concentrations were used for the purposes of statistical analysis and reporting of the data. An additional experiment was conducted using barrows (Camborough Canabrid cross, Pig Improvement Canada, Acme, AB; 25 ± 2 kg) to determine the digestibility coefficients of energy and amino acids in the two positive control lactation diets and in the two basal lactation diets, one containing 0.80% total lysine with 0.65% total threonine and the other containing 1.06% total lysine with 0.70% total threonine (Table 2). It was assumed that added L-threonine was 100% digestible. All digestibility coefficients described herein are apparent ileal digestible values and will be expressed as digestible. Description of the study, including the materials and methods and resultant apparent ileal digestibility coefficients for energy, lysine, and threonine in the gestation diets are reported in the companion article (Cooper et al., 2001). Animal Management Sows were moved to the farrowing room on d 110 of gestation and fed 1.25 kg of their respective lactation diet twice daily until farrowing (farrowing day = d 0). Sows had ad libitum access to feed and water from d 1 to weaning. Feed spillage and refusals were collected, dried for 24 h at 105 C and weighed to accurately determine net feed intake as opposed to feed disappearance. Lactation feed intake was determined for d 1 to 7, 8 to 11, 12 to 15, 16 to 19, and 19 to weaning. Sows were

3 2380 Cooper et al. Table 1. Ingredient composition of experimental and positive control lactation diets, as fed basis a 0.80% lysine, 1.06% lysine, 0.80% lysine, 1.06% lysine Ingredient, % 0.30% threonine b 0.40% threonine c positive control postive control Wheat Hulless barley Soybean meal (44% CP) Cornstarch Beet pulp Oat hulls Canola oil Vitamin premix d Mineral premix e Dicalcium phosphate Limestone Salt Choline chloride Sodium bicarbonate Potassium carbonate L-Lysine HCl L-Valine L-Leucine L-Isoleucine L-Phenylalanine L-Histidine DL-Methionine L-Tryptophan a Based on total calculated lysine and threonine values; analyzed values presented in Table 2. b Threonine was added at 0.05% increments to the following six diets to a total of 0.65% in the last diet at the expense of cornstarch. c Threonine was added at 0.05% increments in the following seven diets to a total of 0.70% in the last diet at the expense of cornstarch. d Provided the following per kilogram of diet: vitamin A, 82,500 IU; vitamin D 3, 825 IU; vitamin E, 40 IU; menadione, 4 mg; thiamine, 1 mg; riboflavin, 5 mg; D-pantothenic acid, 15 mg; niacin, 35 mg; vitamin B 12, mg; D-biotin, 0.2 mg; folic acid, 2 mg. e Provided the following per kilogram of premix: zinc, 100 mg; iron, 80 mg; manganese, 25 mg; copper, 50 mg; iodine, 0.5 mg; selenium, 0.1 mg. weaned on the Thursday closest to 21 d after farrowing but never sooner than d 19 of lactation (mean lactation length = 20.1 ± 0.1 d). Piglets were treated according to routine management practices that included teeth clipping, tail docking, ear notching, and subcutaneous iron dextran injections (200 mg/pig) within 1 d of parturition. Piglets were castrated within 7 d of birth. Litters were standardized to at least 11 piglets within 48 h after birth by cross-fostering. When not enough piglets were available from sows on the present study, piglets from other sows were fostered onto litters. Creep feed was not offered to the piglets during the lactation period, and piglets had free access to water through nipple drinkers. The University Committee of Animal Care and Supply at the University of Saskatchewan reviewed and approved the animal care protocol for the experiment to ensure adherence to the Canadian Council of Animal Care (UCACS protocol #970004: Canadian Council on Animal Care, 1993). Measurements Sow BW and backfat thickness were measured on d 110 of gestation, 24 h after farrowing (d 1), and on d 10 and 18 of lactation and weaning. Backfat thickness was measured over the last rib 2 cm off the midline on the left and right sides using A-mode ultrasound (Renco Lean Meater type 7; Renco Corporation, Minneapolis, MN). The average backfat thickness of the left and right sides was calculated. Weaning-to-mating interval, subsequent litter size (total born and born alive) and whether a sow was culled after the experiment were obtained from the farm s PigCHAMP (University of Minnesota, Minneapolis, MN) records. Piglets, including still-borns, were weighed individually within 12 h of birth. The total numbers of piglets born and born alive were recorded. Litter weights were recorded on d 2 after farrowing once cross-fostering was completed, on d 7, 11, 15, and 19, and at weaning. Blood was collected from sows before morning feeding (0800) via jugular venipuncture to determine plasma urea nitrogen concentrations on d 10 and 18 of lactation. Most sows did not have feed in the feeder at this time. Blood samples were centrifuged, and the plasma was separated and stored at 20 C until analyzed for urea nitrogen. Plasma urea nitrogen was determined using the Abbott Spectrum urea nitrogen test (Abbott Spectrum Series II, Abbott Laboratories, Dallas, TX).

4 Threonine requirement of the lactating sow 2381 Statistical Analysis Data were analyzed as a randomized complete block design by using the general linear models procedure of SAS (SAS Inst. Inc., Cary, NC). Sows were assigned to lactation diet within parity and gestation lysine level in each of three replicates. There was no effect of replicate (P > 0.10); therefore data were pooled. Each replicate consisted of one first-parity, one second-parity, and two third- and higher-parity sows for each treatment combination. Lysine effects in gestation and interactions between gestation lysine level and lactation lysine and threonine levels were not observed (P > 0.10; Cooper et al., 2001); therefore, data in lactation were pooled across gestation diet. Gestation diet did not affect litter birth weight (P > 0.10; Cooper et al., 2001); thus, BW of the litter after fostering was used as a covariate for litter performance. Furthermore, lactation length was used as a covariate in the analysis. The initial analysis included data for threonine levels between 0.40 and 0.65%, that is, the levels of threonine that were common to both lysine levels. As well, each lysine level was analyzed sepa- rately using all of the threonine levels within that level of lysine. Single-degree-of-freedom contrasts were performed using the solution statement in the GLM procedure of SAS to determine linear and quadratic treatment effects. The inflection point was calculated using the solution option of SAS to determine the threonine requirement for sow BW changes, litter growth, and plasma urea nitrogen. Pearson s correlation coefficients were calculated using SAS to test the degree of association between lactation feed intake, lactation weight changes, and litter growth. The relationships between variables of interest were studied using linear regression techniques. Chi-square analysis was used to determine the frequency of culling by treatment. All values are reported as least square means. Results Overall, sows gained an average of 4.8 ± 0.7 kg in lactation and had an overall ADFI of 7.10 ± 0.10 kg/d. Average daily litter growth was 2.49 ± 0.1 kg/d, indicat- Table 2. Nutrient composition of experimental and positive control lactation diets, as fed basis 0.80% lysine, 1.06% lysine, 0.80% lysine 1.06% lysine Nutrient 0.30% threonine a 0.40% threonine a positive control a positive control a Calculated analyses DE Mcal/kg CP, % Lysine Total, % Apparent digestible, % Threonine Total, % Apparent digestible, % Methionine Total, % Apparent digestible, % Tryptophan Total, % Apparent digestible, % Ca, % Total P, % Nutrients, analyzed DE (Mcal/kg) CP, % Lysine Total, % Apparent digestible, % b Threonine Total, % b Apparent digestible, % b Methionine Total, % Apparent digestible, % deb, meq/kg c a Calculated total and apparent digestible values derived from least-cost feed formulation program. b Analyzed apparent ileal digestibility values derived from digestibility trial using grower pigs. c Dietary electrolyte balance; (Na + +K + Cl = meq/kg).

5 2382 Cooper et al. Table 3. Performance of sows fed experimental and positive control diets containing similar lysine and threonine levels Positive control diet Experimental diet Item 0.80% lysine 1.06% lysine SEM 0.80% lysine a 1.06% lysine b SEM ADFI, kg/d BW change (farrow to wean), kg Average daily piglet gain, g/d Average litter gain, kg/d a Lactation experimental diet containing 0.80% total lysine and 0.65% total threonine. b Lactation experimental diet containing 1.06% total lysine and 0.70% total threonine. ing that the goal of using high producing sows in lactation was met. Because of the high concentration of synthetic amino acids in the experimental lactation diets, a concern about the effect of synthetic amino acids on sow productivity was addressed by adding one positive control diet at each level of lysine. The two positive control diets were then compared with the highest threonine diets in the 0.80 and 1.06% total lysine experimental diets. In doing this, lysine and threonine levels were similar between the positive control and synthetic experimental lactation diets. There were no differences in ADFI, BW changes, and litter growth between sows fed synthetic diets and sows fed positive control diets (P > 0.10) in the present study, indicating that high levels of synthetic amino acids in the lactation diets did not impair performance (Table 3). Sow Characteristics Sow lactation BW changes were not affected by lactation lysine level (P > 0.10), and there was no lactation lysine threonine interaction for sow BW changes (P > 0.10). Therefore the data for the two levels of lysine were pooled for threonine levels 0.40 to 0.65% (Table 4). Overall, lactation BW gain increased quadratically (P < 0.05) with increasing dietary threonine. There was a parity lactation threonine level interaction (P < 0.05); threonine level did not affect sow BW gains in Parity 1 or 2 sows (P > 0.10) and sow BW gains increased quadratically (P < 0.001) with increasing levels of threonine for Parity 3+ sows. Based on regression analysis, sow BW gain in parity 3+ sows was maximixed at 0.54% total threonine, equivalent to 38 g of total threonine/d (27 g of digestible threonine/d). Neither lactation lysine nor threonine level affected backfat thickness or backfat change (Table 5; P > 0.10). Lactation lysine and threonine level did not affect overall ADFI (Table 6; P > 0.10). However, in the first 7 d of lactation, sows on the low level of lysine consumed 300 g more feed per day than sows on the high lysine diets (P < 0.05). This effect was not observed after d 7 of lactation. Sow ADFI was lowest in the first week and reached a maximum between d 12 and 15 of lactation. Parity 1 sows had ADFI similar to those of Parity 2 and 3+ sows for the first 7 d of lactation, but their maximum ADFI was less than that of the higher-parity sows (P < 0.05). Litter Characteristics Overall, dietary lysine did not affect total litter weight at weaning (Table 7; P > 0.10). Threonine level did affect total litter weight at weaning (P < 0.05), and there was a lysine threonine interaction (P < 0.05); sows on the low level of lysine had an overall quadratic (P < 0.05) response to threonine. Based on regression analysis, litter weaning weights were maximized at 0.53% total threonine, equivalent to 37 g of total threonine/d (27 g of digestible threonine/d). There was a par- Table 4. The effect of dietary threonine on sow BW change in lactation a Dietary total threonine, % P-value for threonine Item Parity No Mean SEM Linear Quad. Sow BW, kg (d 1) Mean Sow BW change, kg b <0.001 <0.001 Mean a Results were pooled across the two lysine levels. b Sow BW change from d 1 to weaning in lactation.

6 Threonine requirement of the lactating sow 2383 Table 5. Initial backfat thickness and change during lactation as influenced by dietary threonine treatment Dietary total threonine, % Item Parity No Mean SEM Initial backfat, mm a Mean Lactation backfat change, mm Mean a Parity effect (P < 0.01). ity threonine level interaction for sows fed the lowlysine diets (P < 0.05); only Parities 2 and 3+ sows had a response to threonine level (P < 0.05). Dietary threonine did not affect litter weaning weights when sows were fed high dietary lysine diets (P > 0.10). Lactation lysine level did not affect total litter weight gain over lactation (Table 7; P > 0.10). Litter gain had an overall linear and quadratic (P < 0.01) response to threonine. There was a lysine threonine interaction (P < 0.05). Litters on sows fed the low-lysine diets had a quadratic response to threonine (P < 0.05). Litter weight gain was maximized at 0.55% total threonine, equivalent to 39 g of total threonine/d (28 g of digestible threonine/d). Litters on sows fed the high level of lysine had no response to threonine (P > 0.10). Unlike total weaning weight, Parity 2 sows were the only parity group fed low-lysine diets to exhibit a response to threonine (quadratic, P < 0.05), and there was a tendency for threonine to affect litter weight gain in Parity 1 sows (P = 0.12). Overall, litters on Parity 3+ sows gained less weight over lactation than litters on Parity 1 or 2 sows (P < 0.05). Average piglet weights were highest for Parity 2 sows (Table 8; P < 0.05). Piglet ADG was lowest in the first 7 d of lactation and continued to increase until weaning. Sows suckled an average of 11 piglets during lactation, and the mortality rate of piglets in lactation was 3.8% (Table 9). Mortality rate was measured after all cross-fostering was completed and was low because lowviability piglets were replaced with high-viability piglets. Plasma Urea Nitrogen Sows fed high-lysine diets had higher plasma urea nitrogen (PUN) values at d 10 and 18 of lactation (Table 10; P < 0.05). Overall, a quadratic response to dietary threonine was observed at both d 10 and d 18 of lactation (P < 0.01); based on regression analysis, PUN was minimized at 0.54% total threonine, equivalent to 38 g of total threonine/d (27 g of digestible threonine/d). There was a lysine threonine interaction (P < 0.05); a quadratic response to threonine was detected only in sows on the low level of lysine (P < ). Overall, threonine did not affect PUN for d 10 and 18 for sows on the high level of lysine (P > 0.10). There was a parity threonine interaction (P < 0.05) at both d 10 and 18. By d 18 of lactation, Parity 1 sows on the low level of lysine did not respond to threonine (P > 0.10). Table 6. Average daily feed intake (kilograms per day) by parity, lactation lysine level, and time period (as-fed basis) a Day of lactation Total lysine level,% Parity No. 0 7 b Mean c SEM x y y 0.1 Mean x y xy 0.1 Mean a No effect of threonine treatment (P > 0.10); therefore results were pooled across threonine level. b Lysine treatment effect (P < 0.05). c Parity effect (P < 0.05). x,y Means lacking a common superscript within a column differ (P < 0.05).

7 2384 Cooper et al. Table 7. The effect of parity, dietary lysine, and dietary threonine on litter growth performance a P-value for Total Total threonine level, % threonine lysine Item level, % Parity Mean SEM Linear Quad. Litter weight at weaning, kg b,c Mean Mean Mean Litter gain from 0 to weaning, kg b,d Mean b Mean Mean a Days of lactation (20.1 ± 0.1 d) and initial litter weight after fostering used as covariates. b Lysine treatment threonine treatment interaction (P < 0.05). 3 Parity threonine treatment interaction (P < 0.05). d Parity effect (P < 0.05). Subsequent Sow Performance The average weaning-to-estrus interval (WEI) for the experiment was 5.7 ± 0.3 d (Table 11). Parity 3+ sows had the shortest WEI (P < 0.05). There was no effect of lactation lysine or threonine level on the WEI (P > 0.10). Lysine level in lactation did not affect the subsequent piglets born or born alive (P > 0.10). Threonine level in the diet affected subsequent piglets born and born alive (Table 11, P < 0.05). Parity affected the subsequent number of piglets born (P < 0.05). Parity 1 sows farrowed an average of 2.7 piglets fewer than Parities 2 and 3+ sows in their subsequent litter. The frequency of culling after the sow came off experiment was not affected by dietary lysine or threonine (Table 11; P > 0.10). Discussion In the present study, sows fed the 0.80 and 1.06% lysine diets consumed, on average, 58 and 74 g of total lysine/d (47 and 59 g of digestible lysine/d), respectively, during the lactation period. These intakes were greater than expected because sow ADFI was substantially higher than anticipated. The exceptional feed intakes reported herein demonstrate the capacity of the modern sow for nutrient intake, provided factors that suppress feed intake are minimized. They also demonstrate that the feeding of semisynthetic diets, containing modest levels of crude protein, can support a high level of animal performance equivalent to that obtained with conventional formulations. Lysine level in the lactation diet did not affect (main effects) sow ADFI, lactation BW and backfat changes, Table 8. Average piglet weights (kilograms) by time period over lactation (d 1 to 19) Day of lactation Parity 2 a SEM x 2.65 x 3.58 x 4.56 x 5.59 x y 2.88 y 3.97 y 5.10 y 6.16 y x 2.71 x 3.66 x 4.69 x 5.76 x 0.01 Mean a Average weight (kg) of piglets after cross-fostering was completed. x,y Means in the same column followed by different superscripts differ (P < 0.05).

8 Threonine requirement of the lactating sow 2385 Table 9. The number of piglets suckling in lactation and the percentage of mortality after completion of cross-fostering (d 2 to weaning) Parity Item Mean SEM No. piglets suckling No. weaned % Mortality a a Mortality measured after all cross-fostering of piglets completed. Cross-fostering was completed within 48 h after farrowing, standardizing litters to a minimum of 11 piglets. Low-viability piglets were removed from the sow within the first 48 h after farrowing and replaced with piglets of higher viability. litter ADG, WEI, or subsequent litter size. Recent studies support most of these findings. Sow ADFI, backfat changes, and litter size at weaning were not affected by increases in lysine from 36 to 57 g of total lysine/d (Johnston et al., 1993) or 36 to 56 g of total lysine/d (Richert et al., 1997). Johnston et al. (1993), Monegue et al. (1993), and Knabe et al. (1996) reported that increased dietary protein, through the addition of lysine, either as synthetic AA or through soybean meal, did not affect days from weaning to estrus. Other research has reported that increasing lysine in lactation results in reduced BW loss during lactation (Johnston et al., 1993; Monegue et al., 1993; Richert et al., 1994). Daily lysine intakes in these studies were lower than in the present study. Knabe et al. (1996) reported that increasing dietary lysine from 34 to 51 g of total lysine/d did not affect sow BW changes in lactation. Contrary to other studies (Knabe et al., 1996; Dourmad et al., 1998; Touchette et al., 1998), sows in the present study gained BW during lactation, suggesting that high ADFI in lactation prevented sow BW loss in lactation. Sow BW gain was maximized at 38 g of total threonine/d (27 g of digestible threonine/d). The National Research Council (1998) recommends 39 g of total threonine/d (29 g of digestible threonine) for a 175-kg sow at farrowing, producing a litter of 10 piglets gaining 250 g/d and losing no weight in lactation. The high ADFI, BW gain, and backfat thickness gain in lactation indicate that a net mobilization of body tissues to support milk production was unlikely. However, given the low accuracy of P2 backfat measurements in predicting body composition, although unlikely, it is possible that body protein or lipid was utilized by the sows in support of lactation. Litter growth was 2.49 kg/d and was higher than that reported by Dourmad et al. (1998; 2.14 kg/d) and Richert et al. (1997; 2.12 to 2.39 kg/d). Richert et al (1997) fed 27 and 44 g of total threonine/d with 36 and 56 g of total lysine/d, respectively. The total threonine levels were comparable to the threonine levels fed in the present study although ADFI was higher in the present study (7.1 vs 4.6 kg feed/d). Sows fed the low level of lysine in lactation were the only sows to show a response to increasing levels of threonine. The threonine:lysine ratio of both the high- and low-lysine diets was similar for the low-threonine diets. However, sows fed the high level of lysine were consuming more threonine per day on the lowest threonine treatment accounting for the lack of response to threonine. The level of threonine consumed by the sows in a day on the high-lysine diets obviously met the threonine requirement for those sows. Plasma urea nitrogen concentrations can be used as an indicator of protein mobilization in lactating sows (Coma et al., 1996). Lactating sows that consume dietary AA below requirement catabolize body protein in an attempt to supply nutrients needed for milk production (Coma et al., 1996) and metabolize other AA in excess of requirement. As the first-limiting AA reaches requirement, muscle tissue catabolism decreases and there are fewer dietary AA available for oxidation (Etienne et al., 1989). Furthermore, PUN is also an indication of excess total protein, explaining why PUN is higher with the 1.06% lysine diet (15% CP) vs the 0.80% lysine diet (12% CP). Plasma urea nitrogen levels at d 10 and 18 of lactation were minimized at 38 g of total threonine/d (27 g of digestible threonine/d). The requirement to maximize sow BW gain was 38 g of total threonine/d (27 g of digestible threonine/d), similar to the value obtained from the PUN measurements. The threonine requirement of multiparous lactating sows has been estimated to be 31 g of total threonine/d to minimize sow body weight losses (Westermeier, et al., 1998) and 28 g total threonine/d to maximize milk production (Paulicks et al., 1998). This estimate of the threonine requirement is much lower than that found in the present study. However, sows in these studies had an average daily feed intake of 4.3 kg/d, were nursing a maximum of 10 pigs, and were losing an average of 25 kg BW in lactation. The productivity of sows in the present study was much higher than reported in these studies, indicating that the threonine requirement may actually be higher for high-producing lactating sows and sows that do not lose BW in lactation. The threonine requirement of the lactating sow can be expressed in many different ways. The present study indicates that to minimize sow tissue mobilization, the threonine requirement is 37, 40, and 39 g of total threonine/d (29, 31, and 30 g of digestible theonine/d) for Parities 1, 2, and 3+, respectively. The threonine requirement to maximize litter performance is 37, 39, and 38 g of total threonine/d (28, 30, and 29 g of digestible threonine/d) for Parities 1, 2, and 3+ sows, respectively.

9 2386 Cooper et al. Table 10. Effects of parity, lysine level, and threonine level on plasma urea nitrogen (PUN) of sows P-value for Total lysine Total threonine level, % threonine PUN, mm level, % Parity Mean SEM Linear Quad. Day 10 a,b,c Mean Mean Mean Day 18 a,b,c Mean Mean Mean a Lysine level effect (P < 0.05). b Lysine treatment threonine treatment interaction (P < 0.05). c Parity threonine treatment interaction (P < 0.05). Alternatively, the requirement can be expressed as 0.54% total threonine in the diet if sows are consuming 7.1 kg feed/d. Using regression analysis and knowing the daily litter gain at each level of threonine, the threonine requirement in the present study was calculated as 14.3 g of total threonine per kilogram of litter gain. A maintenance requirement of 41 mg of total threonine/kg BW 0.75 (Pettigrew, 1993) is then added to the litter gain requirement to obtain the total requirement for threonine in the diet. Recommendations for lysine indicate that more lysine is required to minimize sow BW loss than is needed to maximize litter growth, contrary to the current findings for the threonine requirement (King et al., 1993; Everts and Dekker, 1994). However, previous studies found that the dietary lysine requirement for maximum litter growth rate was similar to the requirement for minimum sow BW loss (Lewis and Speer, 1973; Chen et al., 1978). Because BW and backfat thickness data in the present study indicate that sows were probably not mobilizing Table 11. Effect of threonine level and parity on the weaning to estrus interval (WEI), subsequent piglets born, subsequent piglets born alive, and culling percentage of sows Dietary total threonine, % Item Parity Mean a SEM WEI, d c c d 0.4 Mean Subsequent total born c d d 0.4 Mean 12.8 x 12.0 x 12.2 x 12.7 x 10.8 y 11.6 xy Subsequent born alive c d d 0.3 Mean 12.1 x 10.8 y 10.9 y 11.6 x 10.0 y 10.8 y Culled, % Overall a Parity effect for subsequent piglets born and born alive and WEI (P < 0.05). c,d Means lacking common superscripts within column differ (P < 0.05). x,y Means lacking common superscripts within row differ (P < 0.05).

10 Threonine requirement of the lactating sow 2387 body tissues, one can reasonably conclude that the threonine supplied in the diet was the only source available to the sow. Indeed, some threonine was likely required to support tissue accretion. However, in the absence of body composition data, it is not possible to quantify. Using performance data of the sows and litters obtained from the present study, and estimating the threonine maintenance requirement based on the model proposed by Pettigrew (1993), the requirement would be 40 g of total threonine/d (28 g of digestible threonine/d). The requirement proposed by Pettigrew (1993) using a modeling approach is comparable to the threonine requirement determined by production parameters herein. Because there was no response to lysine in the present study, it can be assumed that the lysine requirement was no greater than 58 g of total lysine/d (47 g of digestible lysine/d). This is slightly lower than the 64 g of total lysine/d (52 g of digestible lysine/d) one would calculate from the current NRC (1998) model. On this basis, the threonine:lysine ratio would be about 0.69, using total amino acid values, and 0.60 using apparent digestible amino acid values. This is slightly higher than the ratio of 0.61 (total basis) and 0.58 (digestible basis) calculated from NRC (1998). Given the generally higher requirement for threonine for maintenance purposes, one would expect the threonine:lysine ratio to decline as productivity rises; this does not appear to be the case in the above calculations, although it must be recognized that the maintenance requirement for threonine is based on limited experimental data (Pettigrew, 1993). Implications The present study confirms the threonine requirement recommended by the NRC (1998). The study also supports the estimation of threonine requirements based on litter performance (i.e., 14.3 g of threonine per kilogram of litter gain) with a maintenance requirement of 41 mg of total threonine/kg BW Sows of modern genotypes are capable of much higher levels of productivity than previously reported, indicating the need to further evaluate the requirements for other AA. Given the ease with which litter growth rates can be measured, the factorial approach to defining nutrient requirements on a specific farm is well supported for its ease, precision, and economic benefit. Finally, the observation that the performance of the sows fed the experimental diets with high levels of synthetic amino acids was equivalent to that of the more conventional positive control diets confirms that significant quantities of synthetic amino acids can be added to the lactation diet without detrimental effects on sow or litter performance. Literature Cited ARC The Nutrient Requirement of Pigs. Commonwealth Agricultural Bureaux, Farnham Royal, Slough, England. Canadian Council on Animal Care Guide to the Care and Use of Experimental Animals. 2nd ed. p 211. Canadian Council on Animal Care, Ottawa, ON. Chen, S. Y., J. P. F. D Mello, F. W. H. Elsley, and A. G. Taylor Effect of dietary lysine levels on 21-day lactation performance of first-litter sows. Anim. Prod. 27: Coma, J., D. R. Zimmerman, and D. Carrion Lysine requirement of the lactating sow determined by using plasma urea nitrogen as a rapid response criterion. J. Anim. Sci. 74: Cooper, D. R., J. F. Patience, R. T. Zijlstra, and M. Rademacher Effect of energy and lysine intake in gestation on sow performance. J. Anim. Sci. 79: Dourmad, J. Y., J. Noblet, and M. Etienne Effect of protein and lysine supply on performance, nitrogen balance, and body composition changes of sows during lactation. J. Anim. Sci. 76: Etienne, M., J. Noblet, and B. Desmoulin Study of the lysine requirement of sows during lactation. Rech. Porcine Fr. 21:101. Everts, H., and R. A. Dekker Effect of nitrogen supply on the excretion of nitrogen and on energy metabolism of pregnant sows. Anim. Prod. 59: Fuller, M. F., R. McWilliam, T. C. Wang, and L. R. Giles The optimum dietary amino acid pattern for growing pigs. 2. Requirements for maintenance and for tissue protein accretion. Br. J. Nutr. 62: Johnston, L. J., J. E. Pettigrew, and J. W. Rust Response of maternal-line sows to dietary protein concentration during lactation. J. Anim. Sci. 71: King, R. H., M. S. Toner, H. Dove, C. S. Atwood, and W. G. Brown The response of first-litter sows to dietary protein level during lactation. J. Anim. Sci. 71: King, R. H., and I. H. Williams The effect of nutrition on the reproductive performance of first-litter sows. 1. Feeding level during lactation, and between weaning and mating. Anim. Prod. 38: Knabe, D. A., J. H. Brendemuhl, L. I. Chiba, and C. R. Dove Supplemental lysine for sows nursing large litters. J. Anim. Sci. 74: Lewis, A. J., and V. C. Speer Lysine requirement of the lactating sow. J. Anim. Sci. 37: Monegue, H. J., G. L. Cromwell, R. D. Coffey, S. D. Carter, and M. Cervantes Elevated dietary lysine levels for sows nursing large litters. J. Anim. Sci. 71(Suppl. 1):67(Abstr.). NRC Nutrient Requirements of Swine. 9th ed. National Academy Press, Washington, DC. NRC Nutrient Requirements of Swine. 10th ed. National Academy Press, Washington, DC. Paulicks, B. R. V., C. Westermeier, and M. Kirchgessner Milchmenge und Milchinhaltsstoffe bei Sauen in Abhangigkeit von der Threoninversorgung. 2. Mitteilung zum Threoninbedarf laktierender Sauen. J. Anim. Physiol. Anim. Nutr. 79: Pettigrew, J. E Amino acid nutrition of gestating and lactating sow. Biokyowa Technical Review 5. Nutri-Quest Inc., St. Louis, MO. Richert, B. T., M. D. Tokach, R. D. Goodband, J. L. Nelssen, J. E. Pettigrew, R. D. Walker, and L. J. Johnston Valine requirement of the high producing lactating sow. J. Anim. Sci. 74: Richert, B. T., M. D. Tokach, R. D. Goodband, J. L. Nelssen, R. G. Campbell, and S. Kershaw The effect of dietary lysine and valine fed during lactation on sow and litter performance. J. Anim. Sci. 75: Touchette, K. J., G. L. Allee, M. D. Newcomb, and R. D. Boyd The lysine requirement of lactating primiparous sows. J. Anim. Sci. 76: Westermeier, C., B. R. V. Paulicks, and M. Kirchgessner Futteraufnahme und Lebendmasseentwicklung von Sauen und Ferkeln wahrend der Laktation in Abhangigkeit von der Threoninversorgung der Sauen. 1. Mitteilung zum Threoninbedarf laktierender Sauen. J. Anim. Physiol. Anim. Nutr. 79:33 45.