J. Anim. Sci. 2003. 81:1772-1780
© 2003 American Society of Animal Science
An evaluation of the NRC (1998) growth model in estimating lysine requirements of barrows with a lean growth rate of 348 g/d1,2
R. Wei3 and
D. R. Zimmerman
Department of Animal Science, Iowa State University, Ames 50011
3 Correspondence:
Harvard School of Public Health, Department of Nutrition, 1633 Tremont Street, Boston, MA 02115.
 |
Abstract
|
|---|
Two experiments were conducted to evaluate the NRC (1998) growth model in predicting lysine requirements of high lean gain barrows by using plasma urea nitrogen as a rapid response criterion. In Exp. 1, 20 PIC barrows with an initial BW of 19.1 kg were used to estimate total lysine requirements at approximately 30 and 50 kg of BW in two separate randomized complete block designs. Another set of 20 PIC barrows with an initial BW of 59.0 kg was used to estimate total lysine requirements at about 70, 90, and 110 kg of BW in three separate, completely randomized designs. Pigs were individually penned and had free access to feed and water. Results indicated that total lysine requirements at 33, 52, 72, 93, and 113 kg of BW were 0.96 ± 0.01, 0.85 ± 0.02, 0.76 ± 0.05, 0.66 ± 0.03, and 0.49 ± 0.21% of the diet (18.6 ± 0.2, 20.1 ± 0.5, 22.6 ± 1.5, 18.7 ± 0.8, and 16.8 ± 7.2 g/d), respectively. The precision of the estimation decreased when pigs reached 70 kg of BW. To increase the precision, Exp. 2 was conducted in which 20 PIC barrows with an initial BW of 45.2 kg were repeatedly used in Latin square designs to estimate total as well as true ileal digestible lysine requirements at BW ranges of from 60 to 80, 80 to 100, and 100 to 120 kg, respectively. During the three BW range periods, the individually penned pigs were limited in feed intake to 2.6, 2.8, and 3.0 kg/d, respectively, and fed once daily. The estimated requirements in the three BW ranges were 21.8 ± 0.5, 18.8 ± 0.5, and 20.2 ± 0.7 g/d in total lysine and 19.9 ± 0.6, 17.0 ± 0.5, and 18.1 ± 0.6 g/d in true ileal digestible lysine, respectively. Total lysine requirements at approximately 30, 50, 70, 90, and 110 kg of BW were about 102, 98, 106, 92, and 99% of the NRC (1998) recommendations, respectively. The close agreement validated the NRC growth model in predicting lysine requirements of high lean gain barrows over the growing-finishing period.
Key Words: Growth Models • Lysine • Nutrient Requirements • Pigs
|
Introduction
|
|---|
The NRC (1998) has incorporated a growth model to calculate the lysine requirement of pigs over the growing-finishing period from a lean growth curve. The shape of the curve is fixed; it simply moves up or down in accordance with the average lean growth rate input by the user. In the face of research evidence that a lean growth curve may differ in response to genetics, sex, and environments (Whittemore et al., 1988; Schinckel and de Lange, 1996; Van Lunen and Cole, 1998), the NRC (1998) growth model needs to be evaluated.
When pigs are fed diets with lysine concentrations ranging from deficient to excessive, plasma urea nitrogen (PUN) concentrations decrease until the lysine requirement is met. The concept has been applied to estimate the lysine requirement of pigs (Lewis and Speer, 1973; Braude et al., 1974; Brown and Cline, 1974). Plasma urea nitrogen reequilibrates within 3 d in response to a change in dietary CP or lysine concentration (Kaji and Furuya, 1987; Coma et al., 1995a). Coma et al. (1995a) demonstrated that the PUN response allowed an estimation of lysine requirement within a 7-d period, and the estimate was consistent with the requirement determined from nitrogen balance. In addition, PUN is easy to measure and the experimental procedure is simple. The rapid response, accurate estimation, and simple analytical procedure make PUN an ideal response criterion to determine lysine requirements of pigs over the growing- finishing period with the objective of evaluating the NRC (1998) growth model. Therefore, in this study, we chose PUN as the response criterion to estimate total lysine requirements of high lean gain barrows at approximately 30, 50, 70, 90, and 110 kg of BW, and then used the estimates to evaluate the validity of the NRC (1998) model in predicting lysine requirements of the pigs.
|
Materials and Methods
|
|---|
The study consisted of two experiments. In Exp.1, total lysine requirements of high lean gain barrows at BW of approximately 30, 50, 70, 90, and 110 kg were estimated. In Exp. 2, total as well as true ileal digestible lysine requirements at BW ranges of 60 to 80, 80 to 100, and 100 to 120 kg were determined.
Dietary Treatments
At each estimation, dietary treatments (Diet 1 to 5) were five equally spaced total lysine concentrations; the third one was approximately the same as the total lysine requirement (%) predicted by the NRC (1998) growth model for pigs with an average lean growth rate of 350 g/d. In each treatment set, a basal diet containing the lowest lysine concentration with other essential AA at or above their ideal ratios (NRC, 1998) to the highest lysine concentration was formulated to ensure that lysine was the first-limiting amino acid. To avoid using crystalline valine, some wheat was used in diets of Exp. 1 because of its higher valine:lysine ratio compared with corn (NRC, 1998). Because the ideal valine:lysine ratio increases with BW (NRC, 1998), the proportion of wheat in the basal diet changed dramatically across BW classes. In Exp. 2, diets were supplemented with crystalline valine. Each basal diet was sampled and ground through a 1-mm screen for total AA analysis with a high-performance cation exchange resin column after acid hydrolysis (Beckman Systems, Fullerton, CA). Corn and soybean meal used in Exp. 2 were also sampled and analyzed for total AA concentrations. The compositions and analyzed total lysine concentrations of the basal diets used in Exp. 1 and 2 are shown in Tables 1
and 2, respectively. The true ileal digestible lysine concentrations of the basal diets in Exp. 2 were calculated by using the NRC (1998) coefficients and were listed in Table 2 as well.
Lysine additions were achieved by supplementation with L-lysine•HCl and it replaced L-glutamic acid on an equal nitrogen basis. Chloride and sodium concentrations were held constant in diets by adjustments of sodium chloride and sodium carbonate. Therefore, five diets in each treatment set were isonitrogenous and identical in electrolyte balance (Na + K - Cl), thus preventing the confounding effects of dietary protein concentration (Eggum, 1970, 1972) and acid-base balance on PUN concentration (Cai and Zimmerman, 1995).
In each treatment set, the analyzed lysine concentration of the basal diet with the graded additions of crystalline lysine was used in quantifying the dietary total lysine concentrations. In calculating the true ileal digestible lysine concentrations of the treatment sets in Exp. 2, the efficiency of the utilization of crystalline lysine was assumed to be 100%.
Animals and Experimental Designs
In this set of experiments, a total of 60 pigs was used. The experimental protocols were approved by Iowa State University Animal Care and Use Committee.
Experiment 1.
Two trials were conducted to estimate total lysine requirements of PIC barrows at approximately 30, 50, 70, 90, and 110 kg of BW. Pigs were individually kept in pens that were 0.60 x 2.2 m and which had steel-slatted flooring. Each pen contained a stainless steel self-feeder and a nipple drinker. The room was mechanically ventilated and cleaned every 4 d. Room temperature was maintained between 17 and 24°C. Animals had free access to feed and water. In Trial 1, 20 PIC barrows (327 x C22) with an initial BW of 19.1 kg were grouped into four blocks based on litter and initial BW. During a 2-wk adjustment period, pigs were fed a diet containing 1.23% lysine. The first estimation started when pigs reached approximately 30 kg of BW. Five diets (0.840, 0.915, 0.990, 1.065, and 1.140% total lysine) were randomized to pens within each block (i.e., each diet was assigned to a total of four pigs). The estimation was a 7-d procedure and was repeated at other prescribed BW. On d 1 and 2, pigs were bled between 0730 and 0800 via an orbital sinus, collecting about 10 mL of blood from each pig. Immediately after bleeding on d 2, pigs were weighed and the dietary treatments started. After a 3-d period for PUN to reequilibrate, pigs were bled on d 6 and 7 and weighed after bleeding on d 7. ADFI over the dietary treatment period was recorded. Pigs were fed a diet with 1.03% lysine until they reached approximately 50 kg of BW, and then the second estimation started. Another set of diets (0.615, 0.690, 0.765, 0.840, and 0.915% total lysine) was randomized to pens within blocks, which remained the same as in the first estimation.
In Trial 2, a new set of 20 PIC barrows (327 x C22) with an initial BW of 59.0 kg was used for the estimation of total lysine requirements at about 70, 90, and 110 kg of BW. Dietary total lysine concentrations ranged from 0.508 to 0.828, 0.425 to 0.745, and 0.380 to 0.700%, respectively, with a spacing of 0.08% in each treatment set. Because of uniform BW and lack of litter control among the pigs, a completely randomized design was used at each of the three estimations. Pigs were fed a diet of 0.84% lysine during the adjustment period (1 wk) and a diet of 0.64% lysine between the estimation periods.
Experiment 2.
To increase the precision of lysine requirement estimation for finishing pigs, a repeated Latin square design was used. In addition, pigs were limited in feed intake to approximately 90% of the ad libitum amounts predicted by NRC (1998) in order to reduce PUN variation caused by feed intake differences among animals (Coma et al., 1995c). Twenty PIC barrows (327 x C22) with an initial BW of 45.2 kg were used repeatedly to estimate total as well as true ileal digestible lysine requirements at BW ranges of 60 to 80, 80 to 100, and 100 to 120 kg, respectively. Pigs were individually penned with the same environmental control as in Exp. 1. During a 1-wk adjustment period, pigs had free access to a diet containing 0.84% lysine. The first estimation started when pigs reached about 60 kg of BW. It consisted of 20 d (divided into five 4-d periods) and was conducted as a repeated 5 x 5 Latin square design with pigs as rows and periods as columns. There were four squares; each was formed when five pigs reached approximately 60 kg of BW. Dietary lysine concentrations ranged from 0.585 to 0.905% in total lysine and from 0.500 to 0.820% in true ileal digestible lysine, both with a spacing of 0.08%. Feed intake was limited to 2.6 kg/d and remained constant during the 20-d estimation period. Every day at 0800, pigs were fed the assigned rations and feed waste was collected. On the last day of each 4-d period, right before feeding, pigs were bled via an orbital sinus. Body weights at the beginning and the end of the 20-d estimation period were recorded. Right after the first estimation, the second estimation started. Dietary total lysine and true ileal digestible lysine concentrations ranged from 0.421 to 0.741 and 0.352 to 0.672% (spaced at a 0.08% interval), respectively. During this 20-d estimation period, feed intake was limited to 2.8 kg/d and remained constant. The third estimation was conducted with a feed intake limitation of 3.0 kg/d and a dietary lysine range of 0.399 to 0.719% in total lysine or 0.332 to 0.652% in true ileal digestible lysine. At the end of the third estimation, pigs were measured by ultrasound at the 10th rib for fat depth and longissimus muscle area (Loin-O-Matic, Critical Vision, Atlanta, GA). Then the average fat-free lean growth rate over the growing-finishing period (20 to 120 kg of BW) was calculated by using the procedures of NRC (1998). Dressing percentage was assumed to be 74% (NPPC, 1991).
PUN Analysis
Plasma was harvested from blood samples by centrifugation and stored at -20°C until analyzed. Each sample was analyzed for PUN concentration by colorimetrically measuring the product formed in the direct reaction of urea and diacetyl monoxime, as described by Marsh et al. (1965).
Statistical Analysis
Experiment 1.
PUN concentrations on d 1 and 2 were averaged for each pig, and the average was used as pretreatment PUN. Similarly, the average on d 6 and 7 was considered as treatment PUN. Data were analyzed by ANOVA using GLM procedures of SAS (SAS Inst., Inc., Cary, NC) with pretreatment PUN as a covariate to correct PUN variation not related to lysine adequacy (Coma et al., 1995c). Type-III sum of squares was used to test lysine effect on PUN.
Dietary total lysine requirement (%) was estimated by fitting the corrected PUN means vs. dietary total lysine concentrations to a two-slope, broken-line model by using the NLIN procedures of SAS (Robbins, 1986). Daily total lysine requirement (g/d) was calculated as requirement (%) x ADFI.
Experiment 2.
Data were analyzed using the MIXED procedures of SAS to test square, lysine, period, and lysine x period interaction effects. To estimate total lysine requirement, PUN response vs. total lysine intake was fitted to a one-slope, broken-line model and a two-slope, broken-line model with NLMIXED procedures of SAS. Pig was the random subject and a compound symmetry covariance structure (Neter et al., 1996) was used to describe the correlation among PUN measurements within each pig. Likelihood ratio test (LRT; Agresti, 1996) was used to select between the one-slope, broken-line model and the two-slope, broken-line model for lysine requirement estimation. True ileal digestible lysine requirements were estimated by fitting PUN response vs. true ileal digestible lysine intake to the selected model.
The wald test (Agresti, 1996) was used to test if the total lysine requirements estimated from Exp. 1 and Exp. 2 were different.
|
Results and Discussion
|
|---|
Experiment 1
The average BW at the five estimations were 33, 52, 72, 93, and 113 kg, and the ADFI were 1.94, 2.37, 2.97, 2.83, and 3.43 kg, respectively. PUN responses to dietary total lysine and corresponding estimated total lysine requirements are summarized in Table 3. Total lysine requirements at 33, 52, 72, 93, and 113 kg of BW were 0.96 ± 0.01, 0.85 ± 0.02, 0.76 ± 0.05, 0.66 ± 0.03, and 0.49 ± 0.21% of diet; 18.6 ± 0.2, 20.1 ± 0.5, 22.6 ± 1.5, 18.7 ± 0.8, and 16.8 ± 7.2 g/d, respectively. Dietary total lysine requirement decreased with BW, whereas daily total lysine requirement increased with BW until approximately 70 kg and then decreased.
View this table:
[in this window]
[in a new window]
|
Table 3. The responses of plasma urea nitrogen (PUN) to dietary total lysine concentrations and estimated total lysine requirements in Exp. 1
|
|
An examination of the SE of the requirement estimates across BW revealed that the precision of estimation decreased when pigs reached 70 kg of BW, especially at 110 kg. Even though two groups of pigs were used, the comparison across BW was defensible by the following evidence of uniform variation among the pigs between the two groups. First, BW records at the start of the five estimations (31.0 ± 2.2, 54.5 ± 5.2, 69.0 ± 3.6, 90.5 ± 3.2, and 110.0 ± 3.6 kg) did not show a difference in BW variation between the two groups of pigs; Second, in ANOVA analyses for the estimations at 30 and 50 kg of BW, when only block and lysine terms were in the model, the block effects were significant (P = 0.05 for both estimations); however, when the pretreatment PUN term was added to the model as well, the block effects disappeared (P = 0.52 and 0.67 for estimations at 30 and 50 kg of BW, respectively). The results implied that the statistical inference would not change with groups of pigs after individual PUN variation was corrected by pretreatment PUN. Third, the experimental errors in ANOVA (1.78, 1.82, 1.59, 1.61, and 0.93 for the estimation at 30, 50, 70, 90, and 110 kg of BW, respectively) did not indicate a difference in PUN variation between the two groups of pigs. The increase of the SE of the requirement estimates revealed that the fit of the broken-line model decreased when pigs reached about 70 kg of BW. This might be expected, because in finishing pigs, the restriction of voluntary energy intake on protein deposition was relaxed (Campbell et al., 1984, 1985; Dunkin et al., 1986); therefore, the variation in protein deposition and lysine requirements among the pigs increased. As reflected in PUN, the variation in PUN response to dietary lysine concentrations increased. This kind of variation cannot be corrected by pretreatment PUN, which explained the difference in model fitting between ANOVA and broken-line regression.
It was not only this study that showed that the precision of AA requirement estimation decreased in finishing pigs when PUN was used as the response criterion. Coma et al. (1995b) found that the broken-line regression did not yield a lysine requirement in one dietary treatment set in finishing pigs. It was also reported by Parr et al. (2001) that PUN response did not allow an estimate of isoleucine requirement in finishing pigs. Even though the SE at 70 and 90 kg of BW in Exp. 1 were reasonable, we felt that an investigation to increase the precision of lysine requirement estimation for finishing pigs was necessary.
To increase the precision of estimates, one solution would be to increase the number of pigs, which would, however, increase the experimental cost and the requirement for facilities. Instead, we addressed this challenge through an experimental design approach. Because PUN reequilibrates within 3 d after a change in lysine concentration (Kaji and Furuya, 1987; Coma et al., 1995a, 1996), it is possible to feed each animal over a 20-d period the five treatment diets in a sequence. In this way, the variation in PUN response to lysine additions was controlled within each pig. In addition, because each animal receives all treatments, the number of experimental units is increased without an increase in the number of animals. The control of PUN variation and increase in the number of experimental units should increase the precision of lysine requirement estimation. Based on these assumptions, Exp. 2 was designed as three repeated Latin square designs to estimate lysine requirements of finishing pigs. To reduce PUN variation caused by feed intake difference among animals (Coma et al., 1995c), pigs were limited in feed intake to approximately 90% of the ad libitum amounts predicted by NRC (1998). The decision was based on the results that the lean growth rate of finishing pigs reached a plateau at about 80% of voluntary energy intake (Campbell et al., 1985; Dunkin and Black, 1987; Möhn and de Lange, 1998), and that restricting feed intake to 80% of the ad libitum level did not affect daily lysine requirement of finishing pigs (Coma et al., 1995b). We assumed that the feed limitation in Exp. 2 did not reduce the lean growth rate and therefore the daily lysine requirements of the pigs.
Experiment 2
Estimation at 60 to 80 kg of BW.
During the adjustment period, pigs had free access to feed, but they were restricted in feed intake when the dietary treatments started. The change in feeding regimen caused a reduction in feed intake in some pigs, especially during the first week. A number of pigs did not consume all of their daily allotment of 2.6 kg of diets. The reduced feed intake may have resulted in a reduction in lean growth rate. Because the study was to estimate the lysine requirement for maximal lean growth, data from pigs with feed intake less than 2.0 kg/d were removed, resulting in 65 observations for data analysis. The average BW at the beginning and the end of the 20-d estimation period was 59.1 ± 0.9 and 78.0 ± 1.6 kg, respectively. The average BW was 69 kg and ADFI was 2.54 kg. Variance analysis indicated an effect of lysine on PUN (P = 0.0001) with no square, period, and lysine x period interaction effects (P = 0.54, 0.46, and 0.61, respectively). LRT gave a P-value of 0.25, indicating that a one-slope, broken-line model was adequate for lysine requirement estimation. The PUN response vs. total lysine intake and the fitted broken-lines are illustrated in Figure 1. Total lysine requirement was estimated to be 21.8 ± 0.5 g/d. Fitting the PUN response vs. true ileal digestible lysine intake to a one-slope, broken-line model, true ileal digestible lysine requirement was estimated to be 19.9 ± 0.6 g/d.
View larger version (13K):
[in this window]
[in a new window]
|
Figure 1. The response of plasma urea nitrogen to dietary total lysine intake and fitted one-slope, broken-line models in PIC barrows weighing from 59 to 78 kg in Exp. 2. Total lysine requirement was estimated to be 21.8 ± 0.5 g/d.
|
|
Estimation at 80 to 100 kg of BW.
During this 20-d estimation period, most pigs consumed their dietary allotment; however, one pig died and two had feed intakes of less than 1 kg/d because of sickness. Therefore, a total of 85 observations were used for data analysis. The BW ranged from 78.2 ± 1.7 to 95.4 ± 2.3 kg, with an average BW of 87 kg. The ADFI was 2.70 kg. A lysine effect (P = 0.0001) but no effects of square, period, and lysine x period (P = 0.83, 0.33, and 0.79, respectively) were detected. LRT (P = 0.27) led to the choice of a one-slope, broken-line model to estimate lysine requirement, which was 18.8 ± 0.5 g/d in total lysine (Figure 2), or 17.0 ± 0.5 g/d in true ileal digestible lysine.
View larger version (13K):
[in this window]
[in a new window]
|
Figure 2. The response of plasma urea nitrogen to dietary total lysine intake and fitted one-slope, broken-line models in PIC barrows weighing from 78 to 95 kg in Exp. 2. Total lysine requirement was estimated to be 18.8 ± 0.5 g/d.
|
|
Estimation at 100 to 120 kg of BW.
At this estimation, one pig died and another was unthrifty; therefore, a total of 90 valid PUN measurements were obtained. The initial and final BW were 96.0 ± 1.7 and 115.3 ± 2.2 kg, respectively. The average BW was 106 kg and ADFI was 2.98 kg. Variance analysis indicated that there were both lysine and period effects (P = 0.0001 and 0.002, respectively), but there were no effects of square and lysine x period interaction (P = 0.75 and 0.49, respectively). The one-slope, broken-line model was selected from LRT (P = 0.50). PUN response vs. total lysine intake is shown in Figure 3. The requirements in total lysine and true ileal digestible lysine were estimated to be 20.2 ± 0.7 and 18.1 ± 0.6 g/d, respectively.
View larger version (13K):
[in this window]
[in a new window]
|
Figure 3. The response of plasma urea nitrogen to dietary total lysine intake and fitted one-slope, broken-line models in PIC barrows weighing from 96 to 115 kg in Exp. 2. Total lysine requirement was estimated to be 20.2 ± 0.7 g/d.
|
|
From the ultrasound measurements, the average fat-free lean growth rate of pigs over the growing-finishing period was calculated to be 348 g/d.
Validity of the Lysine Requirement Estimates from Experiment 2
Is Lysine Requirement Different from Period to Period?
In each Latin square, each pig was fed five diets in a sequence, with each diet fed for a 4-d period. Besides responding to lysine intake, PUN concentration might change from period to period (Coma et al., 1995a,c; Chen et al., 1999). However, the period effect, if it existed, would be averaged out by the balanced structure of the Latin square design. The critical issue was the lysine x period interaction. If PUN response to dietary lysine depended on period, lysine requirement would differ from period to period, and missing data might bias the requirement estimated. Variance analyses, as discussed above, did not show any evidence for lysine x period effects in the three estimations (P = 0.61, 0.79, and 0.49, respectively). The results were in agreement with those of Fuller and Garthwaite (1993), who found that rate of nitrogen retention (expressed as g/kg of BW0.75) did not change significantly over a 60-d period in growing pigs. However, just as a nonsignificant difference from an F-test does not guarantee lack of difference between any two groups (Neter et al., 1996), a non-significant lysine x period effect only indicated that the lysine requirements were not different among the five 4-d periods. Therefore, we cannot conclude that lysine requirements at the first and the last periods were the same. Each lysine requirement from Exp. 2 was the average requirement over the respective 20-d period.
Effect of Feed Limitation.
It was assumed in Exp. 2 that the feed limitation did not reduce the daily lysine requirements of pigs. However, this assumption might not be valid in view of the research showing that in pigs with a high lean growth potential, energy intake might still limit protein deposition in the finishing period. Campbell and Taverner (1988) observed a linear increase in protein deposition with increased energy intake up to ad libitum level in boars with a high genetic potential for lean growth and weighing from 45 to 90 kg. The work of Rao and McCracken (1991, 1992) showed that in Landrace boars weighing from 33 to 88 kg, the protein deposition increased linearly with energy intake up to levels above ad libitum. Their data also indicated that within the normal range of practical energy intake, each megajoule reduction in energy intake led to an approximately 6-g reduction in protein deposition. Gómez et al. (2002) reported that in high lean gain barrows weighing from about 30 to 82 kg, protein accretion rate increased with each increase in feeding level from 80 to 100% of ad libitum. In Exp. 2, the ADFI during the three estimations were 2.54, 2.70, and 2.98 kg, approximately 0.89, 0.85, and 0.88% of the ad libitum amounts predicted by the NRC (1998) growth model, respectively. The estimated average lean growth rate (348 g/d) indicated that the pigs had a high lean growth potential. Hahn and Baker (1994) reported that digestible lysine requirements of PIC barrows (26 x C15) were 20.3 and 19.6 g/d at BW ranges of from 50 to 95 and from 90 to 110 kg, respectively. Our estimates in true ileal digestible lysine (19.9 g/d for pigs weighing from 60 to 80 kg and 17.6 g/d for pigs weighing from 80 to 120 kg) are in agreement with theirs for early finishing pigs, but are lower than their value for late-finishing pigs. In the face of limited information about the lean growth rate and lysine requirements of PIC barrows, it is unclear whether the obtained lysine requirements from Exp. 2 represented the requirements for maximal lean growth of the pigs.
Effect of Feeding Regimen.
In a once-per-day feeding regimen, lysine requirement may be overestimated because of different absorption rates between crystalline lysine and intact amino acids from feed ingredients (Batterham, 1974; Batterham and O’Neill, 1978; Batterham and Murison, 1981). However, in Exp. 2, pigs at the BW range of 60 to 100 kg had access to feed for more than 8 h after feeding. In addition, animals may conserve lysine during a period of inadequate lysine ingestion and thus prevent a reduction in the efficiency of lysine utilization (Yamashita and Ashida, 1969; Chu and Hegsted, 1976). Therefore, the feeding regimen should not have resulted in an overestimation of lysine requirements. However, at the BW range of 100 to 120 kg, pigs consumed feed more rapidly than in the earlier BW range; most pigs finished their allotted feed in less than 4 h. As a result, the lysine requirement at this stage may have been overestimated.
Lysine Requirements from Experiments 1 and 2
In estimating lysine requirements for finishing pigs, Exp. 1 and 2 each used 20 pigs of the same genetic background and used similar environmental conditions. Pigs in Exp. 1 had free access to feed, whereas pigs in Exp. 2 were restricted in feed intake and were fed once per day. Exp. 1 used a completely randomized design, but Exp. 2 employed a Latin square design in an effort to increase the precision of lysine requirement estimation. A comparison of lysine requirement estimates and the precision (SE of the estimates) between the experiments is of interest. In addition to the arguments we used to defend the comparison of the SE across BW in Exp. 1, the nonsignificant square effects in Exp. 2 (P = 0.54, 0.83, and 0.75 for the three estimations) could be used as further evidence to support the comparability between the two experiments. For the purpose of comparison, the average BW during each estimation period was used. Total lysine requirements of pigs weighing approximately 70, 90, and 110 kg were 22.6 ± 1.5, 18.7 ± 0.8, and 16.8 ± 7.2 g/d in Exp. 1, and 21.8 ± 0.5, 18.8 ± 0.5, and 20.2 ± 0.7 g/d in Exp. 2, respectively. No differences in total lysine requirements at 70, 90, and 110 kg of BW were detected between the two experiments (P = 0.58, 0.92, and 0.22, respectively). The large numeric difference in the requirement estimates at 110 kg of BW between the experiments (16.8 vs. 20.2) may be because 1) the precision of the estimate in Exp. 1 was poor and/or 2) the requirement may be overestimated in Exp. 2 because of the once-per-day feeding regimen. Comparison of SE of the estimates between Exp. 1 and 2 (1.5 vs. 0.5, 0.8 vs. 0.5, and 7.2 vs. 0.7 at 70, 90, and 110 kg of BW, respectively) indicated that the precision of the lysine requirement estimation increased in Exp. 2. The increase may be because of the Latin square design and/or feed intake limitation. However, if pigs had had free access to feed in Exp. 2, the PUN variation caused by the voluntary feed intake difference among the pigs would have added, in a large part, to the individual PUN variation. Because individual variation was controlled within each pig by the Latin square design, the increased PUN variation would not have had much influence on the precision of lysine requirement estimation. A repeated Latin square design would have performed well at ad libitum feed intake situation.
Validation of the NRC (1998) Growth Model
Because the estimates of the total lysine requirements in Exp. 1 and 2 were relatively similar, we pooled the estimates to evaluate the NRC (1998) growth model. The requirement vs. BW, along with the total lysine requirement curve from the NRC (1998) growth model with a lean growth rate of 348 g/d, is shown in Figure 4. Even though there was a large uncertainty at the estimate at approximately 110 kg of BW, the empirically determined requirements fit the NRC curve well. Total lysine requirements of PIC barrows at approximately 30, 50, 70, 90, and 110 kg of BW were approximately 102, 98, 106, 92, and 99% of the NRC recommendations, respectively. The good fit of our estimates to the NRC curve validated the NRC (1998) growth model in predicting lysine requirements of high lean gain barrows over the growing-finishing period.
View larger version (11K):
[in this window]
[in a new window]
|
Figure 4. Total lysine requirements of PIC barrows vs. the total lysine requirement curve predicted by the NRC (1998) growth model with a lean growth rate of 348 g/d. The squares and triangles represent the estimates from Exp. 1 and 2, respectively. The requirement at approximately 110 kg of BW from Exp. 2 may be overestimated.
|
|
|
Implications
|
|---|
The close agreement of the empirically determined total lysine requirements with the NRC (1998) recommendations validated the NRC (1998) growth model in predicting lysine requirements of high lean gain barrows over the growing-finishing period. Utilization of a repeated Latin square design can increase the precision of lysine requirement estimation when plasma urea nitrogen is used as a rapid response criterion.
|
Footnotes
|
|---|
1 Journal paper No. J-19530 of the Iowa Agric. and Home Econ. Exp. Stn., Ames, project No. 3812, and supported by Hatch Act and State of Iowa funds. 
2 Research supported by the Natl. Pork Prod. Council.
Received for publication June 12, 2002.
Accepted for publication March 10, 2003.
|
Literature Cited
|
|---|
Agresti, A. 1996. An Introduction to Categorical Data Analysis. John Wiley & Sons, Inc., New York.
Batterham, E. S. 1974. The effect of frequency of feeding on the utilization of free lysine by growing pigs. Br. J. Nutr. 31:237–242.[Medline]
Batterham, E. S., and R. D. Murison. 1981. Utilization of free lysine by growing pigs. Br. J. Nutr. 46:87–92.[Medline]
Batterham, E. S., and G. H. O’Neill. 1978. The effect of frequency of feeding on the response by growing pigs to supplements of free lysine. Br. J. Nutr. 39:265–270.[Medline]
Braude, R., R. J. Fulford, K. G. Mitchell, A. W. Myers, and J. W. G. Porter. 1974. Performance and blood plasma amino acid and urea concentrations in growing pigs given diets of cereals and groundnut meal and supplemented with graded amounts of L-lysine. Livest. Prod. Sci. 1:383–400.
Brown, J. A., and T. R. Cline. 1974. Urea excretion in the pig: an indicator of protein quality and amino acid requirements. J. Nutr. 104:542–545.
Cai, Y. J., and D. R. Zimmerman. 1995. Relationship of plasma urea nitrogen and urea-cycle amino acid concentrations in swine to dietary electrolyte balance and water intake. Nutr. Res. 15:1517–1524.
Campbell, R. G., and M. R. Taverner. 1988. Genotype and sex effects on the relationship between energy intake and protein deposition in growing pigs. J. Anim. Sci. 66:676–686.
Campbell, R. G., M. R. Taverner, and D. M. Curic. 1984. Effect of feeding level and dietary protein content on the growth, body composition and rate of protein deposition in pigs growing from 45 to 90 kg.Anim. Prod. 38:233–240.
Campbell, R. G., M. R. Taverner, and D. M. Curic. 1985. Effects of sex and energy intake between 48 and 90 kg live weight on protein deposition in growing pigs. Anim. Prod. 40:497–503.
Chen, H. Y., A. J. Lewis, P. S. Miller, and J. T. Yen. 1999. The effect of excess protein on growth performance and protein metabolism of finishing barrows and gilts. J. Anim. Sci. 77:3238–3247.[Abstract/Free Full Text]
Chu, S. H. W., and D. M. Hegsted. 1976. Adaptive response of lysine and theronine degrading enzymes in adult rats. J. Nutr. 106:1089–1096.
Coma, J., D. Carrion, and D. R. Zimmerman. 1995a. Use of plasma urea nitrogen as a rapid response criterion to determine the lysine requirement of pigs. J. Anim. Sci. 73:472–481.[Abstract]
Coma, J., D. R. Zimmerman, and D. Carrion. 1995b. Interactive effects of feed intake and stage of growth on the lysine requirement of pigs. J. Anim. Sci. 73:3369–3375.[Abstract]
Coma, J., D. R. Zimmerman, and D. Carrion. 1995c. Relationship of rate of lean tissue growth and other factors to concentration of urea in plasma of pigs. J. Anim. Sci. 73:3649–3656.[Abstract]
Coma, J., D. R. Zimmerman, and D. Carrion. 1996. Lysine requirement of the lactating sow determined by using plasma urea nitrogen as a rapid response criterion. J. Anim. Sci. 74:1056–1062.[Abstract]
Dunkin, A. C., and J. L. Black. 1987. The relationship between energy intake and nitrogen balance in the growing pig. Page 110 in Energy Metabolism of Farm Animals. Eur. Assoc. Anim. Prod. Publ. 32. Rowman & Littlefield, Totowa, NJ.
Dunkin, A. C., J. L. Black, and K. J. James. 1986. Nitrogen balance in relation to energy intake in entire male pigs weighing 75 kg. Br. J. Nutr. 55:201–207.[Medline]
Eggum, B. O. 1970. Blood urea measurement as a technique for assessing protein quality. Br. J. Nutr. 24:983–988.[Medline]
Eggum, B. O. 1972. The levels of blood amino acids and blood urea as indicators of protein quality. Pages 317–327 in Proteins in Human Nutrition. J. W. G. Porter and B. A. Rolls, ed. Academic Press, New York.
Fuller, M. F., and P. Garthwaite. 1993. The form of response of body protein accretion to dietary amino acid supply. J. Nutr. 123:957–963.
Gómez, R. S., A. J. Lewis, P. S. Miller, H.-Y. Chen, and R. M. Diedrichsen. 2002. Body composition and tissue accretion rates of barrows fed corn-soybean meal diets or low-protein, amino acid-supplemented diets at different feeding levels. J. Anim. Sci. 80:654–662.[Abstract/Free Full Text]
Hahn, J. D., and D. H. Baker. 1994. Ideal digestible lysine levels for early and late-finishing swine. J. Anim. Sci. 72(Suppl. 2):68. (Abstr.)[Abstract]
Kaji, Y., and S. Furuya. 1987. Plasma urea nitrogen as a response criterion for determining the amino acid requirements of growing pigs. Jpn. J. Zootech. Sci. 58:737–742.
Lewis, A. J., and V. C. Speer. 1973. Lysine requirement of the lactating sow. J. Anim. Sci. 37:104–110.
Marsh, W. H., B. Fingurhut, and H. Miller. 1965. Automated and manual direct methods for the determination of blood urea. Clin. Chem. 11:624–627.[Abstract]
Möhn, S., and C. F. M. de Lange. 1998. The effect of body weight on the upper limit to protein deposition in a defined population of growing gilts. J. Anim. Sci. 76:124–133.[Abstract/Free Full Text]
Neter, J., M. H. Kutner, C. J. Nachtsheim, and W. Wasserman. 1996. Applied Linear Statistical Models. WCB/McGraw-Hill, New York.
NPPC. 1991. Procedures to Evaluate Market Hogs. 3rd ed. Natl. Pork Prod. Council, Des Moines, IA.
NRC. 1998. Nutrient Requirements for Swine. 10th ed. Natl. Acad. Press, Washington, DC.
Parr, T. M., I. Mavromichalis, B. J. Kerr, and D. H. Baker. 2001. Isoleucine requirement of growing (25–40 kg) and finishing (90–105 kg) pigs. J. Anim. Sci. 79(Suppl. 2):64. (Abstr.)
Rao, D. S., and K. J. McCracken. 1991. Effect of energy intake on protein and energy metabolism of boars of high genetic potential for lean growth. Anim. Prod. 52:499–507.
Rao, D. S., and K. J. McCracken. 1992. Energy:protein interactions in growing boars of high genetic potential for lean growth. 1. Effects on growth, carcass characteristics and organ weights. Anim. Prod. 54:75–82.
Robbins, K. R. 1986. A method, SAS program, and example for fitting the broken-line model to growth data. Exp. Stn. Res. Rep. 86-09. Univ. of Tennessee, Knoxville.
Schinckel, A. P., and C. F. M. de Lange. 1996. Characterization of growth parameters needed as inputs for pig growth models. J. Anim. Sci. 74:2021–2036.[Abstract]
Van Lunen, T. A., and D. J. A. Cole. 1998. Growth and body composition of highly selected boars and gilts. Anim. Sci. 67:107–116.
Whittemore, C. T., J. B. Tullis, and G. C. Emmans. 1988. Protein growth in pigs. Anim. Prod. 46:437–445.
Yamashia, K., and K. Ashida. 1969. Lysine metabolism in rats fed lysine-free diets. J. Nutr. 99:267–273.
Top
Abstract
Introduction
