|
|
||||||||


* Department of Animal Sciences, University of Illinois, Urbana 61801;
and
Department of Animal Sciences, Purdue University, West Lafayette, IN 47907; and
and
PIC USA, Franklin, KY 42134
1 Correspondence:
216 Animal Sciences Laboratory, 1207 W. Gregory Dr. (phone: 217-333-6455; fax: 217-333-7861; E-mail:
m-ellis7{at}uiuc.edu).
| Abstract |
|---|
|
|
|---|
Key Words: Floor Space Growth Pigs Sires
| Introduction |
|---|
|
|
|---|
| Materials and Methods |
|---|
|
|
|---|
Experimental Design and Treatments
The treatments were arranged in a 2 x 2 x 2 factorial consisting of two sire lines, two floor spaces, and two genders. Sire lines were a Pietrain-based line (Line A) and a synthetic line comprised of Large White, Landrace, Duroc, and Pietrain (Line B) from the Pig Improvement Co. U.S.A. (Franklin, KY) chosen for this study to represent the range in growth rate among PIC sire lines (Miller et al., 2000). Pigs in the unrestricted floor space were in pens of four with floor space of 0.93 m2/pig for the entire period. Recommended floor space allowances for pigs of 45 to 68 kg and greater than 68 kg of BW are 0.56 and 0.74 m2/pig, respectively (Fritschen and Muehling, 1986). Pigs in the restricted floor space were in pens of 12 with floor space of 0.37 m2/pig from 40 to 80 kg of BW and 0.56 m2/pig from 80 to 120 kg of BW. The first replicate was conducted from June to September, the second from December to March, and the third from September to December. However, only pigs from the first and second replicates were utilized in the growth curve analysis.
Animals and Management
The pigs were the progeny of mating eight Line-A and nine Line-B sires with PIC Camborough 22 dams. A total of 92 pens (736 pigs) were placed on test from 40 to 120 kg of BW. These consisted of 32 pens in the first and second replicates and 28 pens in the third replicate, with 45 and 47 pens of Line A and B pigs, respectively, and 46 pens on both the restricted and unrestricted floor space. All lines used in this study were tested as free of the detrimental alleles of both the Halothane and RN genes. The study was conducted in a mechanically ventilated building at the University of Illinois Swine Research Center that had part-solid, part-slotted concrete floors. Pigs were housed in like-genotype, like-gender groups and were allocated to treatment on the basis of sire and BW. Progeny from a minimum of three sires were represented in each pen of pigs. At a mean pen BW of 80 kg, the restricted floor space pens were enlarged by widening the existing pen, keeping a constant ratio of solid to slotted floor.
Pigs were given ad libitum access to feed from a two-hole feeder and were fed on a three-phase dietary program formulated to meet or exceed NRC (1998) recommendations; diets were based on corn and soybean meal (Table 1
). The first-phase diet was fed between 40 and 70 kg of BW and was formulated to supply 18.2% CP, 1.05% lysine, and 3,384 kcal of ME/kg. The second-phase diet was fed between 70 and 100 kg of BW and was formulated to supply 16.5% CP, 0.90% lysine, and 3,390 kcal of ME/kg. The third-phase diet was fed between 100 and 120 kg of BW and was formulated to supply 14.3% CP, 0.82% lysine, and 3,386 kcal of ME/kg.
|
![]() |
Statistical Analysis.
Pen was used as the experimental unit for the growth data, which were analyzed using the GLM procedure of SAS (SAS Inst., Inc., Cary, NC). The model included the effects of sire line, floor space, gender, replicate, and all two- and three-way interactions. The residual mean square error was used as the error term to test the main effects and interactions. Means were evaluated using the PDIFF and STDERR options of SAS GLM.
The BW and compositional growth curve data were analyzed using the NLMIXED procedure of SAS. Body weight mass data were fitted to age using Bridges equation (Bridges et al., 1986; Craig and Schinckel, 2001):
![]() |
where C equals an estimate of mature BW, WT equals BW minus birth weight (1.4 kg), t is days of age, M is the exponential growth decay constant, and A is the kinetic order. The exponential growth decay parameter was close to zero so the function was reparameterized:
![]() |
where M' = ln M. The mixed model had the form:
![]() |
where ci is the random effect for pig, and ci is assumed to be normal with mean 0 and variance
2c, and eit is assumed normal with mean 0 and variance
2e (Craig and Schinckel, 2001). The predicted BW of each pig was given using the pig specific ci and m'i values. The i values are for the individual pigs, and the individual pigs are nested within replicate. The mean predicted weights of each pig were predicted for each 18-d age period from 90 to 180 d of age. The predicted ADG for each pig for each age period was predicted from the pig-specific equations. These predicted values were analyzed by the PROC MIXED procedure of SAS with the observation of each 18-d age period as the repeated measure. The model included the random effect of replicate and fixed effects of line, gender, floor space, and their interactions. The model was fitted with a compound symmetry covariance structure.
Equations including BW, 10th rib backfat depth, and longissimus depth determined by ultrasound were used to predict empty body protein (EBP, kg) and empty body lipid (EBL, kg). Different equations were used to predict body composition at different weight ranges: 40 to 55, 55 to 80, 80 to 100, and 100 to 120 kg. These prediction equations were developed from two studies that used pigs of five genetic lines that were scanned serially and harvested (Thompson et al., 1996; Wagner et al., 1999). Predicted empty body mass data were fitted to allometric (EBP = aXb), augmented allometric [EBP = aXb(700 - X)c], and generalized nonlinear [(EBP = M[1 - exp(b0 + b1x + b2x2)] functions, where x is live weight and M is an estimate of mature body protein (Wagner et al., 1999). The generalized nonlinear function was solved by linearizing the function ln [1 - (EBP/M)] = b0 + b1x + b2x2 and identifying the value of predicted empty body protein mass (MTP: 20, 25, 30, or 35 kg) that resulted in the highest R2 values. These parameter values were used as initial values for an interactive solution by PROC NLIN in SAS. Predicted empty body lipid data were fitted to allometric, augmented allometric, and exponential (EBL = eb0 + b1x + b2x2 +b3x3) functions (x) of BW (Wagner et al., 1999). The b3 coefficient of the exponential function was deleted if P > 0.10. The R2 values were calculated as the squared correlation coefficient between the predicted (
i) and the observed values (Yi) for each component. The residual standard deviation (RSD) for all functions was calculated, and the equations with the lowest RSD were used to calculate their respective curves. The RSD was calculated by the equation:
![]() |
where ei is the residual value for the ith observation, n = number of observations, and p = the degrees of freedom of the equation. For almost all cases, generalized nonlinear (empty body protein) and exponential (empty body lipid) functions minimized the RSD values; as a result, these equations were used for all curves. Daily gain and protein and lipid accretion relative to BW gain were determined by the derivative of each function. Average daily gain was determined by ADG =
LW/
T. Daily protein accretion and lipid accretion rates were determined by
C/
T = [(
C/
LW) x (
LW/
T)] (Schinckel and de Lange, 1996), where C is the body component mass, T is the time, and LW is the live weight.
| Results and Discussion |
|---|
|
|
|---|
|
|
|
|
|
There were significant (P < 0.001) line x age period, line x floor space x age period, and line x floor space x gender x age period interactions for predicted BW and ADG (Table 4). These interactions between age period and the main effects indicate that the shape of the curve changed over time and was different with respect to sire line. The relationship of predicted ADG curves to BW (Figure 1b
) were similar for the two lines and showed a similar pattern to other studies (Schinckel and de Lange, 1996; Smith et al., 2002), increasing to approximately 80 kg of BW, and then remaining relatively constant thereafter. Predicted BW and ADG were higher for Line-B vs. Line-A pigs throughout the growth period (Figures 1a and b
). Line-B pigs had greater predicted protein accretion (Figure 1c
) throughout the trial than did Line-A pigs. Protein accretion rate had a curvilinear response, with rates increasing to a maximum at approximately 75 kg and subsequently decreasing, similar to other studies (Schinckel et al., 1996; 2002; Schinckel and de Lange, 1996). In contrast, lipid accretion rates were relatively similar for the two sire lines (Figure 1d
). Thus, the compositional growth curves highlight that the differences in growth performance between these two lines resulted largely from the greater protein accretion rate of Line B than Line A, with little difference between the lines for lipid deposition rate.
Effect of Floor Space
Floor space did not affect the coefficient of variation of BW at the end of the study (Table 2
). There was a trend (P = 0.08) for the removal rate owing to morbidity and mortality to be higher for pigs reared in the restricted than the unrestricted floor space. A number of studies have shown that morbidity levels increase with a decrease in floor space and/or an increase in group size (McGlone and Newby, 1994; Brumm and Miller, 1996; Wolter et al., 2002), whereas others have shown limited effects of crowding on these measures (Brumm et al., 2001; Wolter et al., 2002). Pigs in the restricted floor space had lower (P < 0.05) ADG and ADFI during Phase 1 (122 and 212 g/d, respectively) and Phase 2 (91 and 347 g/d, respectively) and the overall test period (106 and 280 g/d, respectively) than pigs reared in the unrestricted floor space (Table 3
). The gain:feed ratio was also lower (P < 0.05) for the pigs in the restricted floor space during Phase 1 and the overall test period, but did not differ between the floor spaces during Phase 2 (Table 3
). This decrease in growth performance was likely the result of the combination of decreased floor space and group size. A number of studies have shown that a decrease in floor space allowance results in a decrease in ADG and ADFI, with little effect on feed efficiency (NCR-89, 1993; McGlone and Newby, 1994; Brumm and Miller, 1996). Gonyou and Stricklin (1998) reported a decrease in performance with increasing group size from 3 to 15 pigs/pen and Petherick et al. (1989) also found lower ADG, ADFI, and gain:feed for pigs stocked at 36 pigs/pen than those stocked at 6 or 8 pigs/pen. In contrast, Randolph et al. (1981) and McGlone and Newby (1994) reported no effect on growth in group sizes from 5 to 40 pigs.
There were significant (P < 0.001) floor space x age period, line x floor space x age period, and line x floor space x gender x age period interactions for predicted BW and ADG (Table 4). The interactions between floor space, age period, and gender indicate that the effects of crowding were different over time. More specifically, the effects of the restricted floor space increased as age and weight of the pig increased. Predicted BW (Figure 2a
), ADG (Figure 2b
), and protein accretion rate curves (Figure 2c
) were greater over the entire test period for pigs reared in unrestricted vs. restricted floor space. In addition, predicted lipid accretion rate (Figure 2d
) was greater for pigs reared in the unrestricted floor space. Lipid accretion rates increased linearly with BW with the difference between the two floor spaces increasing at heavier weights. Holck et al. (1997) also found that lipid accretion rate increased as pigs became heavier, and that pigs reared in an unrestricted floor space deposited lipid at a higher rate than pigs reared in a more restricted floor space. The higher lipid accretion of pigs reared in the unrestricted floor space is due to their higher feed intake, which, when in excess of the energy requirement for protein deposition and maximal lean gain, results in increased ratio of lipid accretion:protein accretion (Schinckel, 1999).
Effect of Gender
Gilts had less (P < 0.05) variation in BW at the end of the study than did barrows (Table 2
). There was a trend for more barrows than gilts to be removed from test (P < 0.05), a result that has not been reported elsewhere. Barrows grew faster (P < 0.05) during Phase 2 and the entire test period, consumed more feed (P < 0.05), and had poorer feed efficiency (P < 0.05) than did gilts (Table 3
). Gilts had less (P < 0.05) backfat, greater longissimus depth, and a higher predicted lipid-free soft tissue percentage than did barrows (Table 3
). These results are in agreement with others that have compared barrows and gilts over the weight ranges used in this study (Cisneros et al., 1996a; Ellis et al., 1996).
There were significant (P < 0.001) gender x age period, floor space x gender x age period, and line x floor space x gender x age period interactions for predicted BW and ADG (Table 4). The floor space x gender x age period interaction indicates that the difference in predicted ADG for barrows was relatively constant between the two floor spaces for each age period, but the difference in ADG between the two floor spaces for each age period increased for gilts. Predicted ADG was greater for barrows than gilts, with the difference tending to decrease at heavier weights (Figure 3b
). Predicted protein accretion rate curves for barrows and gilts were curvilinear (Figure 3c
). Protein accretion rate for barrows peaked at approximately 75 kg of BW, whereas the peak for gilts occurred at approximately 85 kg of BW and the rate tended to decline more slowly for gilts than barrows. Predicted lipid accretion was similar for both genders until approximately 65 kg of BW, when barrows began depositing lipid at a higher rate than gilts (Figure 3d
). Schinckel et al. (2002) also found that barrows deposited lipid at a higher rate than gilts as BW increased.
Gender and sire line x floor space interactions were not significant (P > 0.05) for any trait. However, a number of other studies have investigated the effects of various genetic lines reared at several test stations and reported the occurrence of genotype x environment interactions in swine populations (Merks, 1989; Bidanel and Ducos, 1996). Research involving more diverse genetic lines and a wider range of environments than those used in the present study is required to establish the extent of genotype x environment interactions on commercial operations.
| Implications |
|---|
|
|
|---|
|
Received for publication June 4, 2002. Accepted for publication January 20, 2003.
| Literature Cited |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. G. Young, M. D. Tokach, F. X. Aherne, S. S. Dritz, R. D. Goodband, J. L. Nelssen, and T. M. Loughin Effect of space allowance during rearing and selection criteria on performance of gilts over three parities in a commercial swine production system J Anim Sci, November 1, 2008; 86(11): 3181 - 3193. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. P. Schinckel, D. C. Mahan, T. G. Wiseman, and M. E. Einstein Growth of protein, moisture, lipid, and ash of two genetic lines of barrows and gilts from twenty to one hundred twenty-five kilograms of body weight J Anim Sci, February 1, 2008; 86(2): 460 - 471. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Lebret, M. C. Meunier-Salaun, A. Foury, P. Mormede, E. Dransfield, and J. Y. Dourmad Influence of rearing conditions on performance, behavioral, and physiological responses of pigs to preslaughter handling, carcass traits, and meat quality J Anim Sci, September 1, 2006; 84(9): 2436 - 2447. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. C. Brumm The effect of space allocation on barrow and gilt performance J Anim Sci, August 1, 2004; 82(8): 2460 - 2466. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |