| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ANIMAL GENETICS |
Division of Animal Sciences, University of Missouri, Columbia 65211
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
Key Words: Genetic Parameters • Pigs • Placental Efficiency • Selection
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
). Studies of the reproductive physiology of the highly prolific Meishan breed suggest that increasing the ratio of piglet weight to its placental weight (placental efficiency) could be an effective way of increasing uterine capacity, and thereby litter size, in commercial breeds. Divergent selection for placental efficiency resulted in an extraordinary response in litter size after one generation (Wilson et al., 1999), but the small size of the lines used in that study leaves open the possibility that the effect was the result of genetic drift and is not reproducible in larger populations.
Selection for reproductive components has successfully increased the number of fully formed piglets, but the correlated decrease in piglet birth weight may be responsible for the smaller response in number born alive, the increased number of stillborn piglets, and the higher preweaning mortality (Johnson et al., 1999). Placental weight and efficiency were subsequently evaluated in this population (Mesa et al., 2003). Placental weight also decreased as a correlated response to increased litter size and, unexpectedly, placental efficiency was higher in the control line. Thus, litter size increased in the selected line by mechanisms other than improved placental efficiency.
An index was chosen as the selection criterion because it has been shown that selection for a trait defined as a ratio of two positively correlated component traits puts unequal pressure on the components, whereas selection on a linear index places predetermined selection pressure on the component traits (Gunsett, 1984). The objectives of the present study were to estimate the response to selection for a linear index designed to alter placental efficiency, and to evaluate the effect of placental efficiency on litter size in swine.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
.
|
Data collected on piglets from 193 females were as follows: birth weight (BRWT) = 2,111, placental weight (PW) = 2,006, placental efficiency (PE) = 1,677, and selection index (SI) = 1,677. Placental efficiency was defined as the ratio BRWT:PW. Litter traits recorded (n = 193) were total number born (TB), number born alive (NBA), number of mummified fetuses (NMF), number weaned (NW), number of stillborns (NSB), litter birth weight (LBRWT), litter placental weight (LPW), and preweaning survival (SURV) calculated as (NW/NBA) x 100. Survival records from fostered piglets were assigned to the birth mother.
Animal Management
All piglets were processed and ear notched within 24 h of farrowing. Fostering was minimal. Piglets were weaned when the oldest litter was 21 d old (average 18 d) and maintained in environmentally controlled nursery facilities. At 5 mo of age, pigs were transferred to a modified open-front building (Generation 0), pasture lots (Generation 1), or dirt lots (Generations 2 and 3) until a new breeding cycle began. Animals were fed a corn-soybean meal-based diet formulated to meet or exceed nutritional requirements at every stage (NRC, 1998). Management of animals at all stages was in accordance with approved procedures at the University of Missouri South Farm Swine Pasture Unit (IACUC Protocol No. 3198).
Beginning at 160 d of age, gilts were exposed to direct contact with a mature, intact boar for 10 min daily for detection of estrus. After at least 75% of gilts had expressed their first estrus, a breeding period of 2 wk was initiated with the intention of synchronizing estrous cycles. Synchronization was accomplished with two i.m. injections of 10 mg of dinoprost (Lutalyse, Pharmacia and Upjohn Co., Kalamazoo, MI; 5 mg/mL) given 12 h apart to pregnant females 14 d after finalization of the breeding period.
After synchronization of estrus, once-daily detection of estrus was resumed and females were inseminated with fresh diluted semen (3 x 109 cells/dose) from males in the same line and replicate, avoiding half- and full-sibling matings. Females were bred 12 and 24 h after the first signs of behavioral estrus. At 107 d of pregnancy, females were transferred to the Animal Sciences Research Center farrowing facility (University of Missouri, Columbia) where parturition was supervised 24 h/d. In the morning of the fifth day of the 7-d farrowing period, parturition was induced in remaining females with the same protocol used for synchronization of estrus.
Selection Procedures
A selection index was constructed that included TB in the litter where the piglet was born, and individual BRWT and PW adjusted for differences in TB and gestation length, with the additional adjustment of parity number only in the base population. The index was designed to modify PE and was constructed using the information obtained from the first 20 litters in the base population. With the objective of giving each component equal weight, each trait was divided by its standard deviation to remove the effect of the units used and the inherent differences in variability. Total born was further adjusted to take into account that its heritability was estimated to be half that of the other two traits and that individual selection for TB using the dam’s record halves the selection differential.
The selection index was calculated as follows:
 |
The resulting index was
 |
Within each replicate, the 24 highest- and lowest-indexing female progeny and the seven highest and lowest indexing male progeny were chosen to create divergent lines with either high (H) or low (L) PE, respectively. This selection scheme was continued within replicate in H and L for high and low index values, respectively. The only selection criterion in Generations 0 and 1 was the high or low SI. To avoid selection of runts in subsequent generations, piglets weighing less than one-third of their litter average and less than 1,000 g were culled. Four individuals from H and one from L were culled in Generation 2, whereas seven from H and nine from L were culled in Generation 3.
Statistical Analyses
(Co)variance components, heritabilities, and genetic correlations were estimated with an animal model by using the MTDFREML programs described by Boldman et al. (1995). Estimated breeding values for individual and litter traits were calculated similarly. The pedigree file represented seven generations comprising 2,236 individuals. Litter traits were considered a trait of the dam. All individual and litter traits were analyzed with a single-trait model that included the fixed effects of contemporary group (all animals born within a week), with the addition of sex for individual traits or parity for litter traits. Litter was fitted as an uncorrelated random effect for all traits, and TB was used as a linear and quadratic covariate for BRWT, PW, and PE. Two-trait analyses were performed by using the same model used for single-trait analyses, with the exception that no covariates were used.
Estimated breeding values for all traits were outputted from MTDFREML for further analyses to test line effects and estimate genetic trends. Statistical analyses were performed using the GLM procedure of SAS (SAS Inst., Inc., Cary, NC). The model used to test line effects for both the phenotypes and the breeding values included the effects of generation, line nested within generation, and replicate nested within line. Replicate within line was used as the error term. Genetic trends were estimated by regression of the line-generation least squares mean EBV on generation number.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
. Line differences were not significant at any generation for BRWT and SI. Similarly, line differences were not significant between lines at any generation for TB, NBA, NSB, NMF, NW, SURV, and LBRWT (means were 11.0 ± 0.2, 10.6 ± 0.2, 0.6 ± 0.1, 0.8 ± 0.1, 9.7 ± 0.2, 90.5 ± 1.0%, and 15.26 ± 0.27 kg, respectively).
|
Variance component estimates from single-trait analyses are presented in Tables 2 and 3. Direct heritability estimates from single-trait models were 0.03, 0.25, 0.18, 0.11, and 0.08 for BRWT, PW, PE, SI, and TB, respectively. Maternal heritability estimates were 0.13, 0.09, 0.03, 0.00 and 0.07 for BRWT, PW, PE, SI, and TB, respectively. The observed direct-maternal correlations of 1 and –1 are likely inaccurate because correlations of high magnitude are often scaled to those values when using a relatively small sample size; however, the signs of those correlations are assumed to be correct.
|
|
. Two-trait analyses involving TB represent the correlation between a dam’s individual traits for BRWT, PW, PE, and SI and the litter she subsequently produced. The direct additive correlations of TB with BRWT, PE, and SI were –0.33, –0.63, and –0.18, respectively. Phenotypic correlations were 0.61 between BRWT and PW, 0.09 between BRWT and PE, and 0.29 between BRWT and SI (P < 0.001). Phenotypic correlations were –0.67 between PW and PE, and –0.56 between PW and SI (P< 0.001). The phenotypic correlation between PE and SI was 0.86 (P < 0.001). None of the phenotypic correlations between a sow’s individual traits and her subsequent litter were significantly different from zero.
|
. Genetic divergence trends were 20.7 ± 2.7 g, 0.24 ± 0.03, and 0.11 ± 0.02 per generation for PW, PE, and SI, respectively (P < 0.01), but not significant for BRWT. Least squares mean EBV by line for litter traits are plotted against generation in Figure 3. Among litter traits, genetic divergence trends were 24.0 ± 5.8 g, 16.7 ± 2.1 g, 0.07 ± 0.02, and 0.06 ± 0.02 per generation for LBRWT, LPW, TB, and NBA, respectively (P < 0.01), but were not significant for NSB, NMF, NW, and SURV. Regression coefficients of least squares mean EBV on generation for each line are presented in Table 5.
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
), but was somewhat lower than those ranging from 0.13 to 0.18 in other studies (Southwood and Kennedy, 1991; See et al., 1993; Johnson et al., 1999; Ruíz-Flores and Johnson, 2001). Similarly, the heritability estimate for NBA (0.08) was similar to the 0.10 reported for U.S. breeds (Chen et al., 2003), but was lower than the 0.13 to 0.17 reported elsewhere (Southwood and Kennedy, 1991; See et al., 1993; Irgang et al., 1994; Johnson et al., 1999). Maternal heritability for NBA (0.08) was higher than the 0.03 to 0.05 reported for other populations, where the maternal effect was at most one-third of the direct effect (Southwood and Kennedy, 1991; See et al., 1993; Irgang et al., 1994). In contrast, direct and maternal additive genetic variances were of the same magnitude.
Direct and maternal heritability estimates for NW (0.03 and 0.02, respectively) were considerably smaller than those obtained in Yorkshire (0.09 to 0.12) and Landrace (0.04 to 0.08) crossbred gilts (Southwood and Kennedy, 1990). The heritability estimate for NSB (0.08) was lower than the 0.17 obtained from a population under long-term selection for litter size, whereas the heritability estimate for NMF (0.10) was similar to the 0.12 obtained in that population (Johnson et al., 1999).
Estimates of variance components for individual birth weight were adjusted for TB. Direct (0.03) and maternal (0.13) heritabilities for BRWT were lower than comparable estimates reported for German pigs (0.08 and 0.22, respectively; Roehe, 1999). On the other hand, our estimate for litter effect (0.18) was higher than the estimate (0.09) reported by Roehe (1999). Our heritability estimate for litter birth weight (0.08), unadjusted for litter size, agreed with that reported by Roehe (1999), but it was lower than the 0.32 from a selection experiment for litter size (Johnson et al., 1999). Heritability estimates for BRWT (0.03), PW (0.25), and PE (0.18) were similar in magnitude to those obtained from pigs selected for ovulation rate or uterine capacity (0.05, 0.18, and 0.29; Vallet et al., 2001). Similarly, when the heritability for BRWT was estimated by using a direct-maternal animal model, estimates of direct (0.03) and maternal (0.13) heritabilities were very close to those obtained with European pigs (Grandinson et al., 2002; Knol et al., 2002).
Heritability of preweaning survival, a trait of the dam, was low (0.05), agreeing with results obtained in other populations (Lamberson and Johnson, 1984; Knol et al., 2002). This suggests a slow expected response to direct selection for preweaning survival.
Selection for a linear index of total born, birth weight, and placental weight was effective in increasing PW and PE in H and decreasing them in L. Contrary to expectation, the correlated genetic trend for total born was negative in H and positive in L. Despite the difference in PE being larger in the present study than in that of Wilson et al. (1999), we did not detect a significant phenotypic difference in litter size, and the genetic trend was in the opposite direction. The genetic divergence trend observed was 0.07 and 0.06 piglets per generation for TB and NBA, respectively, in favor of the line with lower placental efficiency, which agrees with the negative sign on the genetic correlation between litter size and placental efficiency. The observed genetic trend agrees with the lower placental efficiency observed in a line with higher litter size relative to its control (Mesa et al., 2003). Interestingly, in an evaluation of a line selected for increased uterine capacity, placental efficiency was not different from the control line, but it was lower than in a line selected for increased ovulation rate (Vallet et al., 2001). If placental efficiency is a key component of the increased fetal survival and uterine capacity observed in the prolific Meishan pig, these results suggest that the positive association between placental efficiency and litter size might not hold in Western breeds.
The selection index was designed to allow more space in the uterus by decreasing placental size, while not allowing the unfavorable changes in BRWT that have been reported to accompany effective selection for increased litter size. The goal of minimizing the decrease that naturally occurs when litter size increases is justified by the need to avoid the undesired effects that reduced BRWT would have on postnatal survival (Johnson et al., 1999; Quiniou et al., 2002).
The lack of an unselected control line in this experiment makes the evaluation of asymmetry of response difficult. Nevertheless, a between-lines comparison of genetic trends can still have value. The absolute genetic trend for placental weight was 2.6 times greater in the L than in the H line. Similarly, the genetic trend for placental efficiency was 1.6 times greater in the L line. Considering that birth weight was not different between lines, the asymmetry of the response suggests a lower limit in placental weight, after which fetal survival may be compromised, possibly explaining the unfavorable trend observed in litter size.
Although litter size has the biggest effect on the economic efficiency of the swine industry, problems associated with other reproductive traits, such as weak estrus symptoms and high piglet mortality, should be kept in mind when setting goals for long-term selection programs (Rydhmer, 2000). It should be remembered that after NBA, preweaning viability has the greatest effect on economic efficiency (Legault, 1983; Tess et al., 1983). The direct and maternal genetic variability observed for birth weight suggests that selection schemes for increased litter size can be designed to control reduction in birth weight and consequently improve piglet survival to weaning.
Selection for increased TB has been effective, but the process requires complicated surgical procedures to measure components of TB such as ovulation rate, embryo survival, and uterine capacity (Christenson et al., 1987; Johnson et al., 1999). These approaches are possible in the academic field, but in practice are not desirable or cost effective (Webb, 1998). Other successful selection systems, such as the hyperprolific scheme (Bidanel et al., 1994), require large centralized sets of performance records and are difficult to implement. In addition, the success of increasing TB is associated with an increase in NSB to the point that only half the response obtained in an experiment was represented by live pigs (Johnson et al., 1999). The increased time from beginning to end of parturition in large litters may be the cause of the increased NSB (Johnson et al., 1999). The selection procedure used in the present experiment could alleviate the negative environmental effect on piglet survival, although it requires the constant presence of a technician during parturition. This person can supervise farrowing and increase the probability of survival of small and weak piglets. In this experiment, 3.5% of the pigs were stillborn. This proportion was significantly lower (P < 0.001) than the 7% average observed in first parity gilts from the Maternal Line National Genetic Evaluation Program (Moeller et al., 2004), and the 8.2% (mummies plus stillborn) of the National Animal Health Monitoring System survey of commercial farms (USDA, 2002).
The complexity of the interactions among the component traits of litter size makes the manipulation of this trait difficult. According to the current understanding of the physiology of litter size, this trait will increase only when the most limiting of its components is enhanced (Bennett and Leymaster, 1989, 1990a,Bennett and Leymaster, b). A comparison of the results of the present experiment and the literature available on the Meishan breed and other selection experiments suggests that in any specific population, prolificacy can be increased, but through different physiological mechanisms. Additional comparisons of litter-size components (including ovulation rate, embryo survival, uterine capacity, and placental efficiency) in those populations would allow for a better understanding of their relationships.
Fetal development and survival are ultimately dependent on the transfer of nutrients and removal of waste materials through the placenta. A possible explanation for the negative genetic correlation between placental efficiency and litter size found in this study is that, without compensatory increases in physiologically relevant traits for placental function, a decrease in placental weight compromises fetal survival. This type of compensatory increase has been shown in the Meishan breed, in which placental vascularity increases in the third trimester of pregnancy to allow survival of the fetus (Biensen et al., 1999). If these compensatory mechanisms do not exist in Western breeds, attempts to increase litter size must focus on traits relevant to uterine capacity. Relevant traits that merit attention include fetal erythropoiesis, endometrial gland function, placental hormone secretion, and placental vascularity (Vallet and Christenson, 1993; Pearson et al., 1998; Vallet, 2000; Kim et al., 2001).
Based on the negative direct genetic correlation between litter size and placental efficiency found in this experiment, it seems unlikely that litter size can be increased as a correlated response to higher placental efficiency. However, determining whether a population expressing high placental efficiency can be successfully used in crosses with lines in which uterine capacity is limiting would be of interest. Alternatively, selection for an index of litter size, birth weight, and placental weight can be a valuable tool to use in populations where preweaning viability potentially can be increased by selecting for heavier piglets.
|
Implications |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
|
Footnotes |
|---|
2 Supported by a Colciencias-Fulbright-LASPAU program scholarship.
3 Correspondence: 159 Animal Sciences Center, 920 East Campus Dr. (phone: 573-882-8234; fax: 573-884-7827; e-mail: LambersonW{at}missouri.edu).
Received for publication October 6, 2004. Accepted for publication February 17, 2005.
|
Literature Cited |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Rosendo, L. Canario, T. Druet, J. Gogue, and J. P. Bidanel Correlated responses of pre- and postweaning growth and backfat thickness to six generations of selection for ovulation rate or prenatal survival in French Large White pigs J Anim Sci, December 1, 2007; 85(12): 3209 - 3217. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. Vallet and B. A. Freking Differences in placental structure during gestation associated with large and small pig fetuses J Anim Sci, December 1, 2007; 85(12): 3267 - 3275. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. A. Freking, K. A. Leymaster, J. L. Vallet, and R. K. Christenson Number of fetuses and conceptus growth throughout gestation in lines of pigs selected for ovulation rate or uterine capacity J Anim Sci, September 1, 2007; 85(9): 2093 - 2103. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Rosendo, T. Druet, J. Gogue, L. Canario, and J. P. Bidanel Correlated responses for litter traits to six generations of selection for ovulation rate or prenatal survival in French Large White pigs J Anim Sci, July 1, 2007; 85(7): 1615 - 1624. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Rosendo, T. Druet, J. Gogue, and J. P. Bidanel Direct responses to six generations of selection for ovulation rate or prenatal survival in Large White pigs J Anim Sci, February 1, 2007; 85(2): 356 - 364. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Mesa, T. J. Safranski, K. M. Cammack, R. L. Weaber, and W. R. Lamberson Genetic and phenotypic relationships of farrowing and weaning survival to birth and placental weights in pigs J Anim Sci, January 1, 2006; 84(1): 32 - 40. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
for P < 0.01, < 0.05, and < 0.10, respectively.
